Hagerman Technology WebLog 2006.Comments?
December 29, 2006
Holy moly. Brick wall. I just crashed into reality. Started laying out the
ARCHIVER circuit board and discovered there ain't no room for parts. The
rotary switches take up all of the useful real estate. Hmmm. I don't really
want to make this taller, I like the appearance as it is. Have to get clever.
Unfortunately, some things gotta go. I have to lose the RCA output jacks, go
with XLR only. That opens up some critical space. Not sure where I'm putting
the big polypropylene caps, maybe on backside. In a way, this is good because
it forces me to simplify. Anyway, you can see for yourself how crowded this is
going to be.
Oh my, it still ain't gonna go. Filling in the layout, trying to wire it up,
no way. Ok, start over. How about making this a 1.5U panel? Is there such a
thing? Well, maybe I'll be first, as I don't want it to grow to 2U. So thin
it out at 2.5" height. Try again. Ok, this time plenty room, what a
difference an extra 1/2 inch makes per channel. Traces and placement coming
out really clean. I'll post nes pics next week. Other good outcome is that I
could put the XLR connectors over/under, which I think makes more sense. Panel
price goes up a little. Evolving into apiece of art. Now everything fits on
top except for the connectors and big polyprops. I might be changing all of
the 0.1uF to 0.22uF. Seems the extra 0.1" length lets it hang over a 1/4W
You know, I wonder how anything can get designed by committee. Seems
impossible. I go in so many circles, feedback between disciplines. I can't
imagine what I would get if I just handed a schematic over the wall. So much
of good design is interactive. Layout changes circuit. Circuit changes panel.
Panel changes knobs, chassis, etc., which then impacts specs. It all has to
happen and evolve in parallel, each compromising to help to the other. The
result is a synergy and holistic integrity you can't get any other way. Not
December 28, 2006
Doh! It hit me on the way home from buying milk. I should use the balanced
JFET buffer from the ARCHIVER in the RGP instead of those differential opamps.
Duh. And so I spent the time calculating all of the gain resistor values
needed for ARCHIVER. Going for a nominal gain of about 58dB at 0.5mV input. I
use the sensitivity control as gain for the headamp to cover the full range of
carts from 0.1mV up to 5mV. Then, the level control range is from -6dB to +4dB
in 2dB steps. That should get me about 72dB maximum gain. I didn't want to
push it too hard in order to keep really good headroom. Made a few other value
changes to bring the -3dB point to 20Hz. Anyway, schematic is done. Just need
to get a layout going. TONE audio is planning on doing some articles on 78s
and their playback. I was hoping to send them a BUGLE PRO, but a working proto
of this machine would be way better.
FRY GOLD up and running. Only one small mistake. I forgot to include an ac
coupling cap between the generator and the driver chips. The dc component
screwed up output bias. Anyway, only took a few minutes to dial in gain such
that signals looked just right. Not much headroom with the LM386 at 9V, it
starts to clip a little. But hey, that's probably good for burn-in. Should
have panel in by Monday and will have a complete unit ready for photos. Then
the tedious task of writing up a new manual and web page.
Oh, here's a link to the construction details of DN's CHIME. Has free panel
files you can use.
Dan's CHIME build page
December 27, 2006
FRY GOLD boards came in, time to start building. Panel shipped today.
Meanwhile I whipped out schematics for ARCHIVER. I decided to do a very simple
balanced power supply like the TRUMPET. Use a noisy dc-dc inverter to create a
-15V supply from the +15V wall wart. I do a LOT of RC filtering here to quiet
things way down. It worked on the TRUMPET. Like 100dB noise and ripple
rejection by the time it gets to the inputs, each stage successively filtered.
The balanced constant current stages do a lot for supply rejection. Why would
I want to add feedback regulators to such a clean no-feedback design? Turned
out to be a lot more expensive than I had anticipated. Retail price is now
looking like $900. The panels themselves run $110. The many rotary switches
add up too. Then a ton of polypropylene caps. Heck, the cheapest section is
the USB output.
December 26, 2006
The more I watch things play out, the more I think I was wrong. Early this
year I predicted the demise of the CD transport. That we were witnessing the
exodus to computer-based or hard drive only playback. Yes, it's a big movement
in that direction, but it appears there will now be two camps. The CD player
is not going away after all. Seems there will always be users who prefer the
simpler method. Shove a disk in and sit back. No menus, no software hassles,
no keyboard or mouse. Just play.
Was thinking the weak part of the ARCHIVER was the output stage. Here I'm
using the same low noise matched JFETs for output duty that I have at the
input. The standing current is quite low at 2mA. Is this really going to
provide a robust drive? So I did some more sims. Wow. Guess I shouldn't
worry. With 50k or higher loads the distortion is -100dB. Ok, that's not real
life, just a mathematical sim. Still, it even drove a 1k load at 1Vac. Gain
dropped from 1.0 down to 0.74, giving a dc impedance of 350 ohms. So I loaded
it with 1uF and ran a step. This time it slowed down to 550us time constant,
or a 550 ohm ac impedance. Not sure why the discrepancy. But I'd guess this
gives a decent 500 ohm (or so) output drive. That will push any cable. Heck,
and even with a nasty 10k load, you only need 1mA output drive to hit 20Vpp.
Looks like I was worried over nothing. Hmmm, maybe I have neglected JFETs for
too long. I've used them many times over the years in non-audio circuits, and
in the TRUMPET MC as my only audio amplifier. And it sounded beautiful. Son
of a gun, I may never use an opamp again.
December 25, 2006
So I get some weblog feedback (I can't thank you enough!) from folks who say
that having the ADC in variable EQ phono is a good thing. Separates me from
the competition with a useful feature. Ok, I can live with that. If you're
going to take the time to write, I'm easily persuaded. I guess I go back to
the ARCHIVER name. Looks like the ADC will fit. They said look at the KAB
unit. So I did. Hmmm, seems to have all of the features I was thinking of.
Even the dual inputs. I suppose the way they did with their canned curves
makes more sense, but I'm going to keep the independent controls that I already
have. The only drawback is that you need a chart handy to remind you of
I still keep thinking this could lead to a compact little half-kit. The basic
circuit is really quite simple, and without all of the controls and EQ stuff,
it could be quite small and low cost. Could be a fabulous balanced phono kit.
But I think I should pump out the headamp kit first. Aye, so much on my plate
Ok, so here's sort of what I think it will look like. Same sandwich
construction as the COMPRESSOR. Similar appearance to FRY GOLD. Everything is
layed out in terms of circuit board simplicity and optimal wiring. Hence, the
extra level control is in between the the two eq knobs. Not logical, but
practical. The level is needed to trim gain for ADC input. I'm leaving out
the DAC. USB will be run as ADC only. That removes another level of switching
and amplification. For the XLR outputs, I added another parallel buffer. One
buffer for XLR, one for RCA and ADC. That way cabling for one system does not
affect the other. Two inputs, one mono, one stereo. Power jack is for 15Vdc.
One blue LED indicates power. Both cartridge loading and gain are adjustable.
I don't do the vertical/lateral switching other units use, as I think this
should be properly handles by the cartridge, not some fake stereo cross wiring.
December 23, 2006
Realized there was one very important switch I forgot! The mono-stereo one.
So let's add that, but also make it switch inputs. Maybe 3 inputs. One stereo
and one or two mono inputs. Have to look at wiring up the switch. Makes sense
as this will be a terrific modern phonostage, so it can go with 2 arm tables.
Looking at power supply, I think I can make a simple inverter chip work ok with
a wall-wart. Lots of RC filter stages, no feedback regulators. Keep it like
the TRUMPET and take advantage of the constant current operation. The pentode-
like curves of the JFETs make PSRR quite good as it is. So forget the ADC?
Let's make this a pure phono. I'm even thinking of an extra machined back
panel. Yeah, that pushes retail cost up towards $600, but I'm thinking the
extra shielding in just the right places makes a difference in hum and noise
pickup. Good, things falling into place. Part of that is because of
understanding exactly where HAGERMAN fits into the marketplace. My main focus
has generally been phonostages, but certainly far from being the market leader.
I've been able to offer value and innovation. That has brought me from zero to
being A market leader. Certainly not in terms of revenue, but in sonics,
value, and innovation. And so that is where I aim the BUGLE GOLD, nah, let's
call it the FLUGELHORN. It is a big step up from a BUGLE. The machine aims to
take on a leadership role in the archiving and 78 markets. It's not just about
variable EQ, but also high-end flexibility and sonics. It is not a second or
backup machine. It is for front-row operation.
I should reflect some more on stuff I came up with years ago. Maybe their time
has come. Good examples above are the name FLUGELHORN, the JFET version of
TRUMPET, and the PICCOLO JFET headamp. These have all been on the back burner
for years. I wonder what else I can ressurect.
Sort of like this marketplace thinking. Really helps to focus in on what I
should be building. Aim at a segment or niche. Build what is needed, not what
I feel like. It might seem like I am a bit spread out. Leadership in vacuum
tube testers, cable burn-in, phono, and now USB (CHIME, THE RIPPER). Is that a
crazy mix or what? Well, no. It may have happened sort of haphazardly, but I
think I can bring it all together to make sense. And that's with the RGP! It
combines the best of phono and tube circuits, has built-in burn-in capability,
and an optional USB interface (that's the ADC/DAC I designed the Hagerworth
filter for). And to actually produce it, I need the VACUTRACE to match tubes.
All pieces now fit. Many technologies had to be developed before I could get
started. The missing pieces are remote control and volume circuits. Well, I
came up with the volume stuff earlier this year, and remote control is quite
easy (plus I did one a decade ago). Everything is in place, technically.
Meanwhile, working on financing and distribution plans.
December 21, 2006
I just don't know what to do with this new machine. Drew up the schematic to
turn it into the ARCHIVER, but I'm not convinced that's what I want to do. Or
more importantly, what you guys want to buy. Basically, the concept is to
replace the aging BUGLE PRO. Ok, use the FRY GOLD style chassis. Use a wall
wart (and make it quiet). And this JFET implementation of the TRUMPET circuit
just keeps looking better and better to me. It should sound spectacular. A
huge upgrade from the BUGLE, so maybe it should not be called such. Where I am
stuck is because it was super easy to add the USB ADC chip. While one output
goes the RCA, the opposite phase goes to ADC. Perfect. Only now I've turned
this into a RIPPER! Do I really want to do that? I'm thinking no. Convince
me otherwise. The ARCHIVER name is good for such a machine, record any LP or
78 straight to USB. I dunno, this seems to actually take away from the great
phono section that it will be. So here's the other option, put the dang XLR
connectors back in. Use the balanced outputs the way they should be. I could
even put them on the front panel with RCAs out the back. Sort of like pro
gear. In fact, this will make a superb reference phono stack for a studio. I
can't tell you how excited I am at how circuit the circuit itself came out. My
only problem is deciding how to package and sell it.
December 20, 2006
And so I'm looking at a possible layout for BUGLE GOLD, and it looks to be a
wiring problem. The BUGLE uses dual opamp gain stages, one side for each
channel. And down the middle of the circuit board these go, in the same place
that the rotary switches go. This works ok for the EQ, but not for gain. I
don't want the long traces on the inverting input of the opamp. The other
option of running separate channel strips is even worse, as the board height is
too small to fit a pair of channels. At least not the way I like to place
parts. And so maybe I will look again at a discrete JFET solution. I came up
with a topology the other day on the forum talking about headamps. With
discrete parts, I think I can fit two channels. Easier than if using opamps.
This circuit is exactly a TRUMPET, but with JFETs and not valves. I should
think it sounds real good. It also offers balanced outputs. But I don't
really want to install XLRs. So maybe I just add a phase switch instead.
That's a good feature. Anyway, do I still call this thing a BUGLE? Or is it
time to whip out the FLUGELHORN name? Or maybe I just call it THE ARCHIVER?
This project may also lead to the PICCOLO headamp half-kit. Seems like a lot
of folks are waiting for that. In fact, I don't see anyone selling such kits.
It's something I started years ago and was thinking of using the 6.3Vdc heater
supply to power it (from CORNET). Problem was creating a -6V supply. Wasn't
sure I wanted to use an inverter chip. Yet here I am, now actually thinking of
using a wall-wart. I guess it is a problem I need to tackle. On the other
hand, it looks like the headamp runs at low current, and would be good for
battery operation. The only issue is that people always tell me they forget to
turn it off. Oh, here's the concept I am looking at for a headamp:
Hmmm, you know, having the PICCOLO board fit inside a CORNET ain't such a bad
idea. Instantly change your existing machine to MC status at low cost. Design
the board such that it can be used inside an existing machine (no connectors),
or in a metal box with connectors.
Oh, I just programmed in my new "HagerWorth" steep low pass filter into the
design theory page. This filter was designed for use in anti-aliasing
applications such as a non-oversampled DAC. And yes, I am going to try it out.
Filter was modified for really good transient performance, yet still give a
reasonable out of band rejection. None of that elliptic shit.
Ok, so I just ran some sims on the PICCOLO. Wow, seems I had guessed quite
well at the values. Everything coming out right where I wanted. Drain current
at 2mA, +/-6V supplies, 20dB gain. In order to measure distortion, I had to
run the signal level way up, 100mV at 1kHz. This allowed me to look at the
harmonics. Came up with interesting results as a function of supply voltage.
Supply (+/-V) Gain (dB) 2nd (dB) 3rd (dB)
------------- --------- -------- --------
6 19.4 -36 -47
9 21.7 -40 -42
12 22.9 -42 -38
As supply voltage increased, so did gain. By just a little. And 2nd order
distortion went down. However, 3rd order increased! Hmmm, seems to me that
the 6V operation will actually sound the best. Shoot, this circuit is ready
for prototyping. BTW, model I used for the JFET was:
.model _2sk389 njf(vto=-0.7 beta=0.02 lambda=3)
Now did a full sim on the ARCHIVER (BUGLE GOLD). Found several trouble spots.
Not sure I got the EQ right, something seems off. But this is getting close.
Had to distribute gain because there wasn't any headroom. After some tweaking,
I got headroom to about 20dB, which isn't super good. But it clips softly, and
there is a bit of 2nd order distortion. Ran supplies up to +/-12V to make sure
it could spit out a solid signal. Starts to clip about 5V. Was able to keep
input stage resistances low for good SNR. Basically, it should give 66dB gain
at very low noise for MC carts. Now I am worried a bit about what the variable
EQ might do to overload. Oh wait, that's not a problem if input is proper
inverse. Duh. Just gotta keep reminding myself it's about sonics, not numbers.
Tweaked some more here late at night. Sometimes you just can't put things
down. Manuevered values around until I got the same exact EQ values as the
BUGLE PRO. This makes it super easy to calculate all of the other values for
the various turnovers. Because it's already done! So I get RIAA working here
and it all falls into place. Calculations on paper meet reality. Or at least
the simulator. Nominal gain is 60dB. Headroom 20dB. I can run headamp stage
from 0dB up to 30dB. Power consumption is quite low at 8mA per 12V.
Differential topology gives me constant current operation (like an active shunt
regulator, only faster). Clipping is very soft. I get 10Vpp out of this. EQ
came out well with an IEC rolloff at 20Hz to remove rumble.
December 16, 2006
Somebody mentioned using a rechargeable battery in the BUGLE GOLD. Not a bad
thought. Maybe so. One issue I face is how to create two supplies. Since
this is an open style chassis, I don't want to do a regular 120V supply. So
maybe I can use a wall-wart like everyone else. Perhaps I use two warts, same
as the ones I use for FRYBABY. Or perhaps just one wart and do a simple dc-dc
inverter to make a negative supply. These are all doable, just need to think
about performance and what it takes to make each version quiet enough. And
then maybe a new low dropout version of a super regulator. Came up with a new
set of controls. I'd like to make it easily do MC carts too, so a pre-gain
switch will select headamp ratio. The cool thought is to label the control
per cartridge (0.5mV, 1mV, 2mV, 5mV) versus gain (0dB, +10dB, +20dB), which is
too arbitrary. Then I add a post-gain control. Why would I do that? Well, I
was thinking, since most archivist want to record, why not just throw in the
USB ADC from the RIPPER? The post-gain control tweaks in 2dB steps to set
proper recording level. There will also be a resistive loading control. A
nice feature-laden version of the BUGLE PRO. Variable EQ and all. Oddly, this
is almost turning into a RIPPER itself. Maybe I need to differentiate it more.
I could use the OPA1634 (or whatever it was) differential opamp and do the
whole circuit fully balanced. More like a TRUMPET. Could then have XLR output
in addition to RCA. I dunno. Maybe I don't even need to make this product.
Not sure how to handle the headamp portion. Could easily do with another
opamp, but that just might be too noisy. Maybe I will adapt the JFET circuit
I used in the defunct TRUMPET MC? I bought a hundred of those damn matched
FETs and never used them. And folks keep asking if I will ever design such a
thing. If it does work out well, I can make a kit out of the circuit. It sort
of works that way now. My main focus is no longer on kits. I did that for a
time (constructor series), where I designed the kit first, and then also sold
it in finished form. A marketing disaster. My new approach (same as I started
with years ago) is to design a real product first. Then if there is something
that comes of the development that could be turned into a simple half-kit, I
will do so. For example, a kit headamp could come out of a BUGLE GOLD. From
the RGP I might end up with a volume control kit. Or maybe a remote control.
Or a new DAC.
December 14, 2006
Getting toward the end of the year, time to reflect a little on progress. It
has been quite a busy year, with suprising accomplishment. Sales are up right
now and I'm ready to storm into the new year. Here's a quick summary in no
1) Conceived 2 mass-spec inventions (submitted patent applications for both).
2) Developed and built 4th generation electronics for mass-spec (2 boards).
3) Designed and built lower noise electrometer box.
4) Designed and built COMPRESSOR (not yet fully debugged).
5) Designed panels for GOLD TRUMPET.
6) Re-organized website.
7) Came up with candy bag concept (now only used for FRYBABY).
8) Re-issued BLUE CORNET with ac select board.
9) Moved and re-built entire shop.
10) Started HAGLABS and HAGSCI.
11) Designed RGP concept, architecture, and circuits (not all).
12) Designed FRYBABY. Hooked up with AUDIO ADVISOR and MUSIC DIRECT.
13) Designed, built, and debugged CLARION (article accepted).
14) Designed and launched THE RIPPER.
15) Designed UFO PRO.
16) Did RMAF.
17) Designed CHIME GOLD.
18) Designing FRYKLEANER GOLD.
Meanwhile, I have FRY GOLD done, should be building the proto in 2 weeks. I'm
really happy with the way it came out. And then I had a brain fart tonight,
what if I do the same for BUGLE PRO? Using the same concept of a 1U rack
chassis, I could easily whip up a BUGLE GOLD, one with variable EQ. It wasn't
planned, but I seem to be adding a number of new GOLD products. It all started
with the COMPRESSOR. Yeah, I think it is time to retire the BUGPRO and design
a new GOLD version. Price of $500 retail is reasonable. I'll still sell the
old PRO machines in half-kit form.
Hey, not bad. I am shipping 19 products! About to be 20 and 21. Compare that
to other small shops who have been in this businees much longer. I would also
estimate that HAGTECH is the market leader for burn-in (FRYKLEANER series) and
for tube testers (VACUTRACE). A leader (one of many) in phonostages and USB
DACs. This is not a bad position to be in, and recognizing my strengths allows
me to focus efforts on those markets. The FRY GOLD and FRYBABY will solidify
that niche. These machines work and business can only grow. I really doubt
anyone else wants to try and take on the tube testing business, so that is
probably locked up. The Metasonix and Maxi-Matcher are toys in comparison. It
could be the nice shiny gold panel for the BUGLE GOLD is just what people are
looking for. One thing I have learned the hard way, appearances are incredibly
important. For BUGLE GOLD I can add variable loading and gain too. That would
put it head-to-head against a Sutherland with way more features for half the
price. Hmmm, I need to think about this some more. Circuitry is already
designed, so it is just a repackaging job, which should not take up much time
Anyway, once I lock up these last two machines, I can get started full steam on
the RGP. Been talking to a major distributor about it, so action is presently
taking place on the business side. I expect most of 2007 to be chewed up with
RGP development. It's a very big and complex machine. I'm breaking a lot of
new ground with it.
December 13, 2006
Finished up the pcb layout for the FRY GOLD. Came out quite nice. I realized
with the LM386 they had enough power that I didn't need to worry about left and
right channels. One set can power all outputs at once. That's ten per side.
That was a drawback in the original and PRO designs. They can only do one pair
at a time. The GOLD will do two pair speaker cables (LRLR), two pair RCA
interconnects, a BNC digital cable, any number of XLR cables (string 'em), and
2.5mm and 3.5mm headphone cables. All at once. And the blade design lets you
rack these things. Forget the Vidar! Will order PCBs and a panel tomorrow.
Should have a working prototype by end of month. And this is not an easy month
to do such.
FRY GOLD PCB
December 12, 2006
Ok, I did a schematic for the FRY GOLD. Borrowed pieces from both BABY and
regular KLEANER designs. Makes sense to use all the same common parts, so I
don't have to stock all new stuff. The waveform generator from the BABY is a
tweaked lower cost version that works really well. But for extra output drive
I'll use the LM386 from the KLEANER. Four of them. Easy and cheap. Also
found locations to put the chips and circuitry amongst the connector array that
seems to make sense. Waveform generatrion towards the middle, power buffers
at the outsides. Coming together nicely. Tomorrow I'll take a stab at a first
draft circtui board. If it looks promising, I'll order the front panel. Can
also order some more panels for CHIME GOLDs.
Really busy lately filling out orders. Building stuff. Almost caught up. I
seem to make a DigiKey or Mouser order at least twice per week. CLARION never
sold. Maybe I should raise the price. Nobody realizes what a steal it was.
UFO review from RSF slated for a March release. Heard some Zu Definitions
today. Wow, pretty nice.
My lovely wife did the marathon on Sunday. Injuries hampered her progress, but
she pushed on and finished without walking. What a studette.
December 9, 2006
Refining the FRY GOLD design. Trying to get various pieces to fit. The
shallowest XLR connectors I could find fit just right with a 3/4" standoff
between the panel and the PCB. Then I spent hours digging through various
other connectors to find that will fit this paradigm. The big solid binding
posts were too short, and they also don't have a provision for prongs. So
change that to a more common 3-way type, which also has a 3/4" spacing. A
little tight, but this opens up more possibilities. I re-grouped the posts and
RCAs with the new spacing and voila, more room to put back in the 2.5mm and
3.5mm stereo jacks (for iPods and headphones). I shifted these to the middle
of the RCAs and now have room for BNCs. It's pretty full up again, and with
better spacing such than when bladed in a kleaner farm it doesn't interfere
with the one above or below it. Design is also so thin (less than an inch, not
including connectors sticking out the front) that it can be wall mounted.
Thought about adding handles, but nah. The design is too clean right now.
Don't want to mess it up visually with standoffs or any other mounting stuff.
Cost is also coming out lower than expected. I might even be able to retail
this for $500.
December 8, 2006
Yeah, I got going on this FRYKLEANER GOLD. Made me a front panel, looks a bit
FRY GOLD panel
Tried to add some BNCs and 2.5mm and 3.5mm stereo jacks such that iPod cables
and headphone cables or studio patch cords could be burned. But things got too
tight. No room for fingers. So I simplified it. Instead, I'll include
adapters for these various things that connect to the RCAs. That maintains the
flexibility without compromising cleanliness of the design. Panels are cheaper
than I expected, so I could probably sell this for as $600 retail. Maybe even
$500. Wondering where I'm going to put all the circuitry. I see room for the
output power amps, but not for the waveform generation. No big deal, this is
not high SNR audio so I can spread it out across the 16 inches. The idea is to
make this very rugged and easy to build. No wires this time. I really hate
building the FRY PROs. All wires. Here I will use my sandwich concept
developed for the COMPRESSOR. Open chassis. The other possibility is to use
an existing 1U rack chassis and modify it. Won't look quite as nice, but it
has the benefit of being enclosed. You've heard of server farms? This is a
burn-in blade. Perfect for doing professional quantities.
Got new Acoustic Sounds catalog today. They do a nice job. Very good at
promoting new items. Sutherland phono on the cover. This is brilliant
marketing! I saw the real thing at RMAF. Big wide chassis with nice paint
job. Inside, the circuit board electronics section covers a tiny area in the
back corner. Looks a bit like a BUGLE. Other side of chassis has a bunch of
batteries. Maybe $100 in parts total, sells for $1000. I guarantee you this
machine won't sound any better than a stock BUGLE. Very similar in terms of
parts quality and circuitry (I don't know what topology is used). Only big
difference is the chassis. Lots to be learned from this. I bet they sell
boatloads of 'em.
December 7, 2006
A day that will live in infamy. Did I spell that right? It's also a day we
lost power for a couple of hours and I realized how helpless I am at getting
stuff done without it. The soldering iron went cold right in the middle of
wiring up a FRY PRO. Trying to ship out a bunch of stuff, can't print manuals,
can't read email. Helpless. So I paid some bills. Then started designing a
new FRYKLEANER. Actually running out of FRYPRO panels. That means it is time
for a new generation. So I think it'll be a rack mount version, very
industrial without some of the existing limitations. More robust binding
posts, every output active so you can burn in more than one set at a time. No
need for level control. Just max it out. So basically, it's the equivalent of
two FRY PROs all at once. Make a nice front panel to look like the CHIME GOLD
and voila! We have a FRYKLEANER GOLD for $750. Higher performance, more
rugged, better looking. It seems the burn-in business just keeps getting
better. I am selling more and more of these things.
Then I see the Nordost burn-in machine. They pay for full-page ads in both
stereophile and TAS. Sheesh. That's like double my salary. How is a guy like
me supposed to compete? He he he. Read the ad. One of those shops listed in
it just bought a FRYPRO this week. Yes, I think a FRY GOLD will do fine. Just
borrow circuits from the KLEANER and BABY designs, make a panel, a new board.
I should have this on the market in a month. Can't let Nordost steal my
thunder. Imagine a rack with ten of these machines. Each one is only 1.75"
And now Thom is interested in making UFO PROs. What should I do here? I
really don't feel like shelling out ten grand to get a decent price on parts
in quantity. Thinking I should focus on preamplification and just license the
design out. Besides, he'll be able to make them look even nicer. Fits very
well with his tables. And they aren't exactly flying out the door for me. I
was hoping for more press from RMAF. Everyone took a photo of it, but it
didn't show up on the websites. This could be a winner, but it takes the right
marketing and distribution, and I'm not sure I'm up to the task for either.
December 4, 2006
Ok, got the new CHIME GOLD put into the website. Will write a new manual later
December 1, 2006
FRYBABY is one of the PFO writer's choice award winners for 2006. I guess that
says it all.
PFO writer's choice awards
November 30, 2006
Came up with a new product idea last night - a vibration meter - or something
that will help you determine the effectiveness of your audio stands, turntable
plinths, or floor. Something simple, like perhaps a laser diode that just sits
on the surface and points at the wall. Or maybe using an optical interference
pattern to get really sensitive. Could be an off-the-shelf seismometer already
exists. Anyway, a simple product that helps you to tune a shelf or platter and
identify bugaboos. Then again, just placing the stylus on the LP without the
motor running seems to do a pretty good job. I like the concept of a laser, as
the light is coherent, so you can send it off two different paths and then
combine them to watch for interference. One path is a reference, the other
measures the surface in question. Unreal sensitivity.
Hmmm, a major wire manufacturer just bought themselves a FRYKLEANER. And I
don't think this is for audio interconnects. He he he.
Just finished up the first CHIME GOLD (sold). Breaking it in, but had to have
a listen. Dang that's nice! I couldn't stop listening. And this is in the
middle of my work day, with machines to build and ship. Here I sit song after
song wasting time. It was wonderful. I get so little time nowadays to do
nothing but listen. This machine sounds fabulous. Had it running from the
iMac. Very nice looking, too.
November 29, 2006
I'm rebuilding my stereo system. It's been in pieces since before RMAF (I used
all of the cables & cords). Now we have this giant (42") plasma screen, I
thought I would experiment with home theater for awhile. See what gives. So I
took the audio outputs from the TV and run it into a CLARINET, CYMBALs, and
Vandersteens. Wow, no buzz or hum. And it sounds a whole lot better than the
built-in speakers. Will have to play a DVD later and see what the impact is.
My first impression was that intelligibility of speech goes way up. Words come
through more apparent against background music and noise. Also, there is a
very different EQ to the voices. Less boominess and more articulation. Less
like a radio announcer.
This also means I moved the iMac to the shop, which now gives me a convenient
internet connection to look stuff up, and frees my laptop from USB testing
duty. Speaking of which, I have an interesting HAGUSB test to make. Since the
relationship between the 1kHz data packets and the 12MHz clock do not relate
well to the 44.1kHz word rate, I'm thinking the output frequency changes every
once in awhile. Sort of like hunting in a PLL or dither in an ADC. I'll hook
it up to a HAGDAC and accurately measure the output frequency over time, see if
it bumps up and down in very small increments every once in awhile. To me,
that is always how I guessed the circuit worked. We'll see if I was right.
What brought this to my attention was that some customers are having trouble
connecting a HAGUSB to an Altmann Attraction DAC. No other DACs on the market
have trouble (so far reported), only that one. One guy is getting clicking
noises (a relay?) every once in awhile as the Altmann thinks it loses lock.
The other didn't lock at all. So I'll dig into it.
Just ran the test. Got the new CHIME GOLD up and running (minus a few PCM1704
chips) and ran it off the iMac. The output from the VCO (recovered clock) was
a solid 11.28948, which then slowly lowered to 11.28942 after the iMac was on
for an hour. Ok, nominal redbook is 11.2896, or an error of 16ppm. That is
perfectly fine. I never saw the frequency jump, only a hunting of the last
digit between adjacent numbers. So I guess I was wrong, the PCM2704 USB chip
does not shift frequency noticeably in order to track the input. If it does,
the jumps are too small for me to detect (1ppm), and well within the redbook
standard. So I'm not sure what the deal is with the Altmann. Ok, next test,
I plug in my laptop and gazingo! Frequency drops to 11.28922 MHz. That means
I was right on all of my forum postings claiming that the output frequency is
dependent on the xtal in the PC, not the 12MHz one on the PCM2704.
CHIME GOLD under test
November 28, 2006
Just had a wierd idea. They should have put a magnetic stripe (one ring) near
the outer edge of CDs and DVDs. This could be used for temporary storage of
information, like exactly where the heck the DVD was last paused. This would
have been a super cheap non-volatile read/write buffer. Voluntary. And no
standard, so each machine could have done whatever it wanted to with the
Thinking this way because I just splurged and bought a plasma TV as the family
Christmas present. Spent the entire year without TV in the living room. I
also realized the way they make money is by selling the wall mounts. Holy cow,
$250 for ten bucks worth of metal. I once bought an entire 7-in-one workout
machine (including the weights) for less. Yikes. Where can I buy a reasonable
TV wall mount?
I'm noticing the latest trend in hi-fi boxes is curves. Large radius curves
and swoops in chassis design. Curves and bulges. Somehow I want to spend the
money on circuits and parts, not sculpture. But hey, I understand, anyone who
will for over this kind of cash wants the machine to look damn nice.
Music Direct finally has the FRYBABY in their catalog.
November 27, 2006
I don't know why, been thinking about DAC filters again. And so why not do a
straight 16-bit output at 44.1k? Yes, it has problems, but only in terms of
filtering. The 8x sync upsampling of the HAGDAC works great. But you never
know exactly what's going on in that digital filter. So what sort of filter
might work at 1x sampling? Clearly a brick wall is needed to avoid aliasing
problems. This is on both ADC and DAC. And there is such a revival of 1x that
there must be something to it. The question is, how to do it right. None of
this non-filtered stuff for me. No way can I pass the problem on down the
line. And so I am looking at various passive filters. The Bessel is my first
obvious choice, which I use a lot. It has perfect time domain performance.
Great transient response with no phase anomalies. But holy cow! I see now why
this has never been used. The attenuation near the breakpoint (-3dB) is awful,
and even adding more poles makes no difference. It is impossible to get one of
these to work for this application.
So now what? Ok, look at Butterworth. Forget the elliptics or other steep
slope oddities (which caused the original treble problems!). I played around
in SPICE for awhile, but even the Butterworth had transient problems. Too much
ringing for my tastes. In frequency domain they look pretty good. Hey, let's
make my own. So I goofed with the component values this way and that way until
I got a better transient response. Ok, this is not too bad. And the values
all normalized to basic fractions. I ended up with a 7th order Hagerworth.
Attenuation still isn't that great around turnover, but it might do. I'd like
to see how this sounds. Perhaps make the turn at 18kHz. Then it gives -7dB
at 22kHz, and -34dB at 36kHz (2x). The normalized -3dB point is at 0.144Hz.
This topology is a compromise between the filterless NONOS and the upscale
HAGDAC types. Bascially, I try to take the best features of both. The idea
is to make the machine as musical as possible but at the same time remove the
unwanted artifacts. The biggest drawback I can see is a rolled off treble.
But you maintain the detail and clarity since much of the smearing caused by
aliased components folded into the passband are removed. Oh, this is for the
digital section of the RGP. Thinking of alternate architectures to do 16-bit
in and out via USB.
I also send out this drawing for quotes on UFO PRO parts. See if I can get
the cost of this thing down.
Looking at transformers again for RGP. Been talking to Slagle about possible
custom winds. Best alternative seems to be the Jensens. My balanced
requirements manuevers me into the JT-347-AXT step-ups. Dang, these things
are $250 a pop! For outputs the fit is a JT-10K6101M. Kinda nice that they
have almost perfect devices for me. Wish they came in a PC mount version,
although maybe an extra anti-vibration level of isolation could be good. Wow,
this is $775 worth of signal iron for one RGP. Also changing to TFT caps, the
tin foil teflon film Exoticas. Unreal how the price of this machine is going
up. I may end up at $20k retail if I'm not careful.
November 24, 2006
Just sold another pair of VacuTraces to China Lake NAWS. That can only mean
they continue to refurbish equipment and keep it in the field. Tube equipment.
Which also means there are more military stockpiles of NOS just waiting for a
socket in which to reside. It also means they like the VacuTrace. I wonder if
Stan Ricker still works there.
November 22, 2006
Just started reading the Nov issue of stereophile. First, I wish I had some
La Scalas to try out in my home. But I was more interested in the reviews of
electronics. The Artemis Labs phono has displaced the Trumpet in a couple of
homes that I know of. Supposedly a sonic improvement, so I was curious to see
the measurments. They look quite good all around except for the distortion
curves. Wow, my power amp does better than this! And the numbers I get for
the Trumpet are almost ten times better. Not just a little, ten times! So it
again makes me wonder about sonics. Perhaps I just need to work a little more
on the unmeasurables.
And then I come across the Zanden DAC. Oh my gosh, these plots can't be right.
So I read the text, and Mikey loved it. Read the measurements, John hated it.
Here we go again! Can there be no correlation between decent measurements and
decent sound? All I can do here is give you snippets of what was said, and
this is for a $28k machine!
MF: "... hearing a compelling presentation that excited my auditory and visual
senses while producing the feel of musical continuity and delicacy that,
heretofore, only good analog has provided ... ability to produce transparency
and air wihout edge or etch created the most realistic and convincing images
I've heard ... transparency, three-dimensionality, and - especially - its
microdynamic resolve made it a credible rival to the vinyl ... sounded more
open, airy, and dynamic than I'm used to, with better-defined images, greater
transient definition, and tighter and better-defined bass ... and the cymbals
did not, for once, sound like air brakes ... I've never heard these discs sound
so open, airy, spacious, and believable ... delivered soundstaging, imaging,
dynamic nuance, tonal and harmonic purity, rythmic flow - and especially
transparency - with an effortlessness I hadn't thought possible from digital,
not to mention 16-bit/44kHz digital, while being totally - I mean, totally -
free of digital artifacts ..."
JA: "... frequency response was dissapointing with a noticeable rolloff of low
frequencies that reached -3dB at 65Hz ... this is simply poor engineering ...
the noise floor was higher than I usually find with 16-bit DACs ... relatively
high AC-supply components ... I haven't seen a DAC behave this poorly for many
years ... produces a dirty, indistinct waveform ... high level of ultrasonic
energy present in its output ... spectrum was littered with distortion
harmonics and the THD+noise was ridiculously high at 25.4% ... incapable of
reproducing high-level low frequencies without introducing high levels of
distortion ... this is pathological behaviour ... dominated by a slew of
audioband intermodulation and aliasing products, and the image of the 20kHz
tone at 24.1kHz was almost as high in level as the fundamental ... jitter level
was high, at 1018 picoseconds peak-peak, which is more than four times the
jitter of the best performing components ... this is poor performance ... it is
the worst-measuring digital product I have encountered ..."
Shit, I am speechless. Almost makes me want to quit the industry and install
lawn sprinklers instead. Seriously, I'd give a left testicle to have
stereophile publish measurements of a CHIME side-by-side with these.
November 21, 2006
Aw crap, my iMac is dying. I think it's the super-drive. Been loading stuff
onto an iPod and it keeps crashing on me. I'm not suprised as I leave this
computer off most of the time and it is right in the salt-spray airflow from a
window. Use the same spot to environmentally test my amplifiers (they still
The low voltage toroid transformer for the CHIME turned out to be really small,
so I was able to mount them right on the circuit board. I could hard-wire it
right in place. This kept things simple. Use the same jumpers for ac voltage
selection as before. Assembly is coming out cleaner than I expected. This
should make for quite a nice product.
Music Direct finally got the FRYBABY added to their Christmas catalog. And
they placed an order. Good, things are rolling on that one. Plus the super
endorsement from PFO (product of year).
And lastly, I jammed all day writing the CLARION article. Finished. I agonize
over writing articles and manuals. Just not my thing. But hey, it comes with
the territory, so I work hard to do a good job and make them readable. Now to
submit to aXp (they accepted).
November 17, 2006
Ok, time to implement my year-end plan. The idea is to clean up loose ends.
Been on a roll lately, will keep selling off any machines in inventory. A-gon
sales. Anything that isn't tied down is getting sold. Extra cash to pay for
Christmas, pay off bills, get finances solid. Maybe I have some unused test
equipment I can dump on eBay. Making room for RGP. Second, is to get those
magazine articles written (CLARION and RIPPER). Third, is to finish up the
design of the CHIME PRO (gold) and get it online (completely assembled it costs
less than the Audio Note kit). Fourth, get the rest of the pre-production UFO
PROs built. Fifth, debug the COMPRESSOR and finish the design, as I need to
prove the output stage for use in RGP. Then sell it.
Made good progress today on the articles. Built up the basic formatting and
outline, added pictures, etc. Realized that the RIPPER article is basically a
re-write of the assembly manual. Cool, now I don't have to think so much.
Running low on HAGUSB boards. Wow, I think I just went through 200 or so of
these in less than two years. Still the best deal for the money.
Just finished the RIPPER article. Will submit to Nuts & Volts magazine. Here
it is raw:
November 14, 2006
More progress on CHIME gold. Or maybe I call it the PRO. CNCed the rear panel
and have most of it built up. Looking good so far. Just ordered a ton of
Just got word of a very positive review coming for FRYBABY. The reviewer
thought it so good he picked it as a product of the year. Details later, after
they take it live.
November 13, 2006
Nice long weeked. Was getting a little burnt out. Made me this animated gif
this morning of the UFO in action...
UFO animated gif
Had my wife's birthday party at a nice restaraunt, and they took our photo. It
came out so nice that I'll use it for Christmas cards. I did touch up the
background, airbrushed in more green to replace some other stuff that was
there (like a mirror).
More on the RGP. If this machine ends up costing so much, I may as well just
pour more money it the the parts. What the heck, it's a statement product. So
been researching best capacitors and other stuff. I'm also now thinking to use
transformers on the output stage. It was going to be like a TRUMPET, but now I
realize I don't have to limit myself. Curiously, this is exactly the output
stage I put in the COMPRESSOR project. So now I've decided to finish it and
debug the proto. Hey, it was almost done anyway. Besides, I actually have a
buyer for it. If the topology works well, then I'll put it into the RGP. This
is yet another level of refinement. I really need to talk to Slagle to see if
he wants to wind me some outputs to go with the inputs. That would really be
Looking again at a CHIME in the Par-Metal box. Have three of these nice
chassis sitting around collecting dust. Will see if I can make some custom
DACs out of them. Things I put on Audiogon these days seems to sell pretty
fast. Must be the season.
We had our final audio club meeting of the year on Saturday. A shootout of
ST-70s and clones. Of course everyone upgrades these things. I think we had
6 or 7 of them. One was a clear winner, but turned out to be mostly stock
topology, just really good part upgrades. Kind of a fun test. Gets you to
really focus on listening to particular things. Try and tune your ear to be
better at detecting issues.
Oh my, I just found some design sketches in my notebook regarding the RGP
dating back to 05/05. Dang, that means I already have a year and a half put in
to this project. I was going back in my notes to see what I had thought of for
the series "one" chassis I had proposed. I'm gonna go ahead and build 3 of
these in a CHIME version. Already sold one. I think they might boost sales,
as interest inthe CHIME keeps building. Just ordered a new front panel, looks
something like this:
CHIME gold panel
November 7, 2006
Got my new Audio Advisor catalog today. Yup, there it is! FryBaby on page 27.
Looks good, they made a very nice photo. Hope they sell a ton of these.
AA FRYBABY page
Put in the extra 100uF caps in the CLARION cathode bias circuit. This helped
to extend the bass response and solidify the sound. Also reduced bandwidth of
driver stage such that it doesn't push too much subsonic info into the output.
The lower grid bias resistor keeps grid current in check. So I ran the
battery of tests. Bandwidth -3dB is from 12Hz to 65kHz, at -1dB it is flat
from 20Hz to 40kHz. Output is 3.0W at onset of clipping. Distortion was 1.1%
at 1W, almost all 2nd order. Noise was -82dBA ref 1W, or about 0.22mV. Gain
is low and requires 4Vrms to reach full output. Hence, probably need a line
stage in system. I've been using directly with a RIPPER, which has enough
gain. Changing to the 5687 input tube should help a little.
When matched with the WHITE CORNET, a 5mV cart produces 0.8V out. On peaks we
get 5x greater, or exactly 4V. So it's almost a perfect match. It won't be
very loud, but you can't clip the amplifier, either. Crank volume up to 100%
and it's just right. No sonic losses in volume control, no clipping. I'll put
the pair up on Audiogon. Cheap. Just need to cover costs. These were made
for RMAF demo and for a magazine article.
November 6, 2006
Put up another CHIME on audiogon. Cheap. Another magazine review unit. It
just doesn't seem right to sell it as new, even though it has had some
beneficial break-in. Also put up the CYMBALs from RMAF. Figured I need to
make some room in the shop for new projects. I got boxes everywhere, can
hardly move. Absolute steal on the CYMBALs, a thousand bucks off. Pretty much
the same price it costs me to make them. Next week I'll also be putting up the
CLARION and matching WHITE CORNET.
This week working on the two articles, the RIPPER one for Nuts&Volts, the
CLARION for aXp. Good cheap advertising.
Meanwhile, I had my latest embarassing moment. Was modifying a UFO to see if
it would operate at 1/2 speed, make the dots more visible. Holy crap, now that
I have a working frequency meter, I discover the original circuit runs at
150Hz. Dang. I glitched somewhere in the original schematic. Son of a gun.
Anyway, the fix is easy (for those of you with UFOs): cut pin 14 off of U1,
then re-wire that hole to pin 15. Voila. I'll fix the UFO manual tomorrow.
The good news is that now all those tiny strobe dots are actually pretty easy
November 5, 2006
Just finished the last triathlon of the season here. Pretty good results for
the lackluster training and taking a week off for RMAF (critical week). Got me
another 2nd plage (age group), and 17th overall. It was a very weak field. I
think I had too much gatorade before the race. Didn't feel that great and the
bike leg was on the slow side. Funky conditions with a Kona wind. I seemed to
lack the gas to really redline it.
But hey! I had a great swim and caught the leader (age group) on the sand.
Long run to transition and a fast T1 got me out front. Held the lead all the
way through the bike and T2. Both transitions went well. Then about 1/2 mile
into the run Dustow caught me. The guy can fly. He caught a whole slew of
racers. That's twice this year he's done that.
Time to take a break and then rev up for next year. My wife is doing the
marathon in December.
Haleiwa triathlon 2006 results
October 31, 2006
Found the old style UFO battery holders at Allied. Bought me enough to finish
this batch of 10 units. Gotta get two shipped off to Foster for a review. New
UFO PRO manual online.
October 30, 2006
I just realized Google totally screwed up the internet. Volunteered to carve
the class pumpkin at my daughter's kindergarten class, so thought I would do a
quick search for stencils. Something good for kids. As an artist, I am only
good for triangle eyes and nose. Well holy cow, it was not easy to find such.
Every one of the top ranked pages was a link to another page of links - they
all cross-linked each other! The web is now full of these auto-generated link
farms that do nothing but point you in circles! The cleverness of google ads
has devolved into an endless loop of non-content. Hello? The beauty of the
web was content and knowledge. And now it is hidden behind a swamp of greed.
Seems everyone was trying to take advantage of google's ranking algorithm,
which ironically, has put itself out of business. It's very existence has
created its own demise. I need to find another search engine that does not
use cross-linking as relevance. Anyone have a suggestion?
Progress? Five years ago a search would have brought up decent content in less
than 5 minutes. Last night, 1/2 hour to accomplish the same task. It was
sick. So now I have to re-think my use of google ads. Am I adding to the
Back by popular demand is the wooden TRUMPET. Yeah, I like it too. Kept
getting requests for the old style. So ok, I now give you the option of either
metal or wood fascia. Did a big website update. Added the UFO pro, lowered
some prices here and there. Updated the cartridge loading page by adding step
up info. Calculate the resistor you need to get desired loading.
October 29, 2006
Ok, I think I have a solution for the RGP. Again. The idea is to make the
digital record loop an option. easily upgraded. The problem I had was that the
front panel layout was a little different between the analog and digital
versions. Not good. But Al from RMAF requested an analog tape loop, of the
traditional variety. That way he could send out a line level for recording on
another rig or, more importantly, for his headphone amp. Shoot, I had never
considered that. Thank god for astute customers that like to give feedback.
So I'm now (for the moment) adding an analog tape loop with outs and ins. This
keeps front panel and operation the same for both models. The compromise is
that the tape out will have an opamp buffer. Only the non-loop audio path will
remain all-tube. That's ok, I was planning on using opamps in the digital part
anyway. Probably use the fully differential OPA1632 amps. That way I can
maintain a fully balanced signal path for all options.
The trick here is to have sub-panels on the rear. Hard to explain. But these
sub-panels can be configured for various options. For example, XLR or RCA
jacks for the line inputs. Or analog tape loop versus USB loop. Or possibly
even a future firewire digital loop (probably the best way to get 24/96). It
gives the customer more choices and is easily upgraded. The connectors are
then wired to the main board (vibration relief). The digital option sub-panel
will have the digital board actually mounted to it. Since the sub-panel is
small in size, I can probably only get a 2" x 3" board on the back of it.
That's not a problem, as I can stack boards, which only serves to make it even
more modular. For example, I can have the converters and reclocking on one
card, and the USB I/O on the other. A simple board change can get you either
USB or S/PDIF. I probably won't do an S/PDIF version, but that illustrates the
basic concept. It also gives me an easy upgrade hook. Buy the latest digital
loop 5 years from now.
The other option was to make a digital box as a separate product. More profit,
but it's just not as cool. I just can't make myself greedy enough to do such.
I'd rather offer the more innovative and less costly solution.
Second result of this new topology is that the level attenuators for the tape
are are standard. So you can equalize all input levels. The VU LED display
is there even on the all-analog model. That's cool. You can set input levels
even though the listening signal is not physically passing through the extra
attenuators. Not unless you monitor the loop, which sends signal through a
couple of opamp stages. I like this. Each input then has it's own sticky
settings for gain, loading, level, and balance. Once set, flip between inputs
and everything is tweaked in just right. So simple. Nothing to memorize, no
menus, no instructions or codes, only one button. Yes, a good day for
Also looking at different concepts for improving on beauty. Clearly, the eye-
candy factor is significant in this price range. So back to custom carved
remote control, knob, and now faceplate. I still like to keep it simple, but
the workmanship has to be top-rate.
October 27, 2006
Or maybe I can just do a separate box, the RGD. A reference grade dac? And
not put it inside the RGP. In fact, I could use my HAGDAC in a nice matching
box, sell it as a separate product. That would save some time. Not really
much I would change with the HAGDAC at this point anyway. Only upgrades I can
think of would be to add 96k reclocking or a custom digital filter. In a
separate box, I could do the gain stage in tubes. If buried in the RGP, I'd
have to use an opamp. But hey, it's only digital. Or maybe I should just do
the plug-in option card thing. The stock version would be a tape monitor.
Then I could have a USB digital loop option. It would also open the door to
other future expansion such as firewire I/O or ethernet. Probably the only
way to get reasonable 24/96 other than S/PDIF. I dunno, can't seem to decide
which way I need to go here. Suppose some more customer research is necessary.
Finally got around to checking out the ironman championship results. A few of
my friends had qualified and did it. Most amazing was Rachel, who I trained
with in 2004. Shux, I used to almost keep up with her. We even swam the same
speed. Well, she's really ramped up, nearly a pro now. In '04 we trained for
the half-ironman. I ran it, she passed up because she couldn't afford it.
This year she did it and came in 2nd overall! And now, she wins her age group
at the Kona ironman with a huge 10:05. Unreal. She's only done this for a
little over two seasons, as she has 3 kids, the youngest barely 3. Maybe I
should just retire.
Updated the CLARION design to use the ECC99 tube (or 5687) as a driver. So you
can use either it or the 6H30. Found solution to the loose bias, just add
electrolytic. My calculation for original value was in error because I
neglected the mu factor looking into the cathode. Anyway, this should make
things solid. I will upgrade the chassis I have and then sell it. Hopefully
it can sell in conjunction with the matching CORNET. Would make an awesome
combination for someone who has a turntable (volume control in CLARION).
Bought a ton of parts. Catching up on orders from RMAF. Really depleted stock
on some items. A new VACUTRACE could not ship because it was missing the small
knobs. FRYPRO can't leave until I get a plastic box. And I am way behind on
HAGDACs. And I was thinking, I have 4 Par-Metal chassis sitting around doing
nothing. Maybe I can do 3 custom CHIMEs. One of the biggest complaints with
the existing CHIME is the aesthetics. No suprise. But I can build 3 GOLD
CHIMEs and sell them at same cost. For those of you who like the gold anodized
faceplate. I might also build a GOLD CYMBAL (stereo) too, with the 4th (big)
chassis. Otherwise it'll just sit there and take up space. Oh my. That's a
lot of stuff for audiogon. Maybe I can try out that new audioXsell.com? I
still have an octal CHIME I custom modified for a customer who changed their
mind. That's gotta go too.
October 26, 2006
Been writing up the CLARION project for aXp. But I recently subscribed to Nuts
& Volts magazine, mostly geared to PIC and robots and other hobby stuff. And
I'm thinking this would be good for a RIPPER article. Seems to be right up
their alley. I bet the subscriber base is huge, far more than aXp. So I think
I'll do that. So two project aricles in the works. Two magazine. I'm curious
to see what sort of impact this might have on sales.
Mulling over a lot of stuff from RMAF. Like maybe I should make the RGP an
analog only machine. Just leave out the digital loop. Ok, so I think it's a
huge and fabulous and innovative idea, but perhaps the market is not ready for
it. Or maybe they are not interested in buying it, even if it is cool. The
top dog type customer will already have a top flight digital system. What they
really want to buy from me is a phonostage. Combining it with the linestage
and remote control only makes it more appealing. However, is it possible that
adding a digital tape loop going too far? My thoughts are now to just make it
an option or leave it off altogether. The good news is that not having to
design a new digital board will save me a bunch of time. Having it as a future
upgrade is reasonable.
I think I'll update my cartridge loading page. Was suprised at how many people
at the show came up to me and said "I love your website, so much good
information" or something along those lines. Hmmm. So I think it is time to
get back to some updates. A long time ago I was talking about popup pages with
tips and tricks. Never got around to it. Maybe it is time again. Or at least
new pages. I need one on grounding and hum tips for sure. Add some resistor
calculator stuff to the loading page.
Looking into 1/2 speed operation for the UFO. So many drill holes in the
strobe pattern make the piece very costly. And the dots are so close together
that the patterns is hard to see in operation. Problem is I don't think I end
up with integers. Think there is 135 dots for 33.3 right now. No way to
remove every other dot. Also, the battery holders that fit are now out of
stock, the replacement is too tall. Hmmm, maybe it is time for a new layout
and change to 123 size batteries.
Good new about the logo pens. Since they had wrong logo, I got a full refund.
October 25, 2006
Wow, great reader editorial on RoHS in the 10/9 issue of EE Times. I was not
able to find a link, but it hits the nail on the head. This lead-free
initiative is not only wasting billions in unnecessary costs, but will end up
killing people - all for minimal impact on the environment. In fact, in the
end it probably does more overall damage than otherwise.
And I get around to reading the TAS issue they gave away at RMAF. The class-D
roundtable article was good. It really exposed a number of things. Not that I
am a fan of Krell, but Dan D-Agostino was right on the money. In fact, I'm
suprised the magazine actually printed his comments. Sort of put Rowland into
a corner, a bit of a panderer driven by marketing and not sonics. Anyway, I
finally got to A/B a class-D amplifier. Holy cow it sounded like crap. I mean
it was really bad. In head-to-head comparison with decent amps, it got
absolutely trashed. No, I don't know what brand, but something supposed to be
good. Don't drink the koolade folks!
October 24, 2006
Ha! My logo pens came today. And they didn't even have the right logo! Said
something about a credit union in Oklahoma.
October 23, 2006
Where to start, this was a long trip. No time off, non-stop action. Lotta
work, even though much of it seemed like fun. No delays because of earthquake
and power outages in Hawaii, so I got in to Denver early. A beautiful blanket
of overnight snow made for some good photo possibilities as I headed for
Boulder. Here's a shot in the canyon.
I then hooked up with Mackris who was storing all of my boxes. He was running
way behind having nothing packed up yet. So we played around instead, working
on last-minute tuning of Lynn's amps. The Azzolinas were set up so we played
LPs into the night. Talk about getting blown away, I had my first exposure to
the Dynavector XV-1. Holy crap that's a fine cartridge. In fact, best I ever
heard. Even on old 60's LPs and non-audiophile stuff the sound was amazing.
It pulled sonic info out of the grooves like you can't imagine. Not only did
it have superb tracking, musicality and speed, but the separation between
instruments was phenomenal. I mean everything got pulled apart. Voices were
fleshed out and each individual contributor or instrument on the recording
remained individual. Just unbelievable what this cart could do with regular
And so I also had my first exposure to the Azzolinas. These are odd beasts,
from a technical standpoint. Lowther full range drivers, but horn loaded on
the front, open baffle on the back. This, mated to a La Scala like folded bass
enclosure, probably somewhat horn loaded, but also with reflex ports. A very
unusual combination of techniques. Only it wasn't just a lark, these things
really work! The quickness and direct connectedness of a Lowther, improved
with the dynamics and slam of a horn. No they didn't sound like horns, much
of the expected colorations were missing, no shouting or honkiness. They are,
nonetheless, quite directional, so the sweet spot is rather small. Bass was
equally efficient and tuneful. They seemed to be optimized for articulation
instead of sonic boom power, probably a compromise with extension.
Anyway, Thursday setup was a long day. One of the CYMBALs had smoked in Texas
prior to the show, so I knew I had to work on it. I got in the room relatively
early and had most everything unpacked and partially setup before the Azzolinas
and the Galibier table arrived. It seemed everything powered up ok. I guess
the CYMBAL had a blown output tube, and sure enough, that's what it was. Nice
to have a VacuTrace on hand (which I sold)! Yup, the shorted tube cooked the
tuning resistor in series with the choke on the B+. It was still working, but
I changed it anyway. Put in new tube, rebiased, and that was that. The P3 and
RIPPER were up and running so I had signal to apply to the speakers. Ok,
everything looked good, I had most of room hooked up and theoretically ready to
go. Frank brought in an arm and EMT cartridge, so all of the pieces came into
place. There was some crickling noise coming out of the TRUMPTET, though.
Dang. Srew it, already 1am, I'll look at it in the morning. Figured it was a
Ok, Friday morning I started working on the TRUMPET. Not a tube. Dang. Took
awhile and got it narrowed down to the left channel first stage. Then down to
a few resistors. I searched for a bad solder joint or loose tube socket.
Nothing. Tried to touch up everything in the area. Crap. Still, crackling
noise, very random. Almost 3 hours had passed and the doors were to open in 20
minutes. Ok, so I have the white CORNET warming up in the background just in
case. What a shame not to have a TRUMPET! SO I ripped out the 4 possible
resistors in the tail and measured them. All perfect. Oh well. Soldered them
back in as a last ditch effort, hoping I had eliminated an unseen solder ball
or contamination of some sort. Yes! All of a sudden the noise was gone.
Ten minutes to spare. I whipped the table I was using as a work bench to my
other room (sleeping only), put out the brochures and opened the door. Five
minutes to the start of the show and I was already packed with visitors. The
TRUMPET was operational, but stocked with brand new Sovtek tubes, not the NOS
I had brought (replaced, thinking they were the source of noise). Not burned
in, but music was playing, and the show went on.
Very busy day Friday. Saturday was non-stop action too. Quite a few of the
magazine folks came by, Jason from stereophile sat in the chair for probably 20
minutes. I was quite impressed. Overall turnout and comments were very
positive. I sold a bunch of half-kits and RIPPERs and HAGUSBs. Sold the
VACUTRACE, and the CHIME dac right out of the rack. Both UFOs went, RSF wants
two of them for HiFi+, Dorgay borrowed a RIPPER. Shoot, I can't even remember
the rest of what went out. Reaction to the UFO was wonderful. Everyone loved
it. Several folks are now on the waiting list. Most thought the RIPPER was a
Ok, most importantly, the sound. Yes, the sound was excellent. The Azzolinas
made my room. Seriously, these speakers are terrific, and they mate to the
CYMBALs like bread and butter. They coasted at maybe 1/2 watt or less most of
the time, and having the extra headroom and low output impedance keep
everything in control. Dynamics were really punchy. I've never heard a
cone/box speaker do this before. Very open, extremely clear. We also got
lucky with room placement. Having spent all the time on the TRUMPET, we never
got around to trying placement, room tunes, or anything else. Scott delivered
the speakers, figured they would sound pretty good right about here (finger
pointing to a spot on the floor), dropped them in place. They never got moved.
It was a measured guess, and there was no need to moving things around. The
sound was already there. Focus, a soundstage way behind the speakers. Ok, it
wasn't very wide, because these things are big.
So maybe we didn't have best sound of show, but must have been on the list.
And this is up against rooms with $100k+ worth of equipment. And I'm using
$2.5k worth of CYMBALs, a TRUMPET, and a CLARINET. He he he. Seriously, it
ended up a great sounding room, getting even better as the show went on and
things settled in. Then late Sunday Frank wanted to try out some new Slagle
step-ups. Stupid me figured we better not so it on saturday. Son of a gun.
The second we put these new trannys in the sound got even better. An easier
flow and more robust and heartiness in the vocal. I am now planning on using
Slagle iron in the RGP.
I have yet to read any of the online blogs or postings. No idea what has been
said by anyone else. Dozens of photographers came around taking pictures of
everything in each room. I'd say turnout had met expectations.
Then there were the after-hour antics. Tweaking, listening, dinner, drinks,
etc. The hallway conversations, so many various folks in the industry. I
have to admit, I learned a ton of new things regarding audio. Lynn was very
educational, the guys from Montana, meeting industry folks in random
conversations. Everywhere you walked (especially after hours), there was
somebody to run into. I come home re-armed with new thoughts on design,
marketing, strategy, friends, and stuff to do for the RGP. Those few I did
trust to give details on the RGP concept and features thought it was great.
One guy (who was extremely helpful in watching the room for me during breaks)
even offered to buy one. I have a lot of other stories I could add (coyote
bait), but there's just some things that are best left personal.
I guess it is just hard to put into words the positiveness of the entire show,
my room, my electronics. Nobody left mumbling "that sucked". So here's a
public thanks and big round of applause for everyone who helped make this show
great. Kudos to Thom and Scott.
More RMAF pics
October 16, 2006
Ok, final post before RMAF. Got a pair of UFO PROs completed just in time over
the weekend. Came out just fine. Lotta work, though, peeling the anodize off
of them. The bottom pieces look very nice without any finish at all. Did this
pair with the same black filler I use on the wood TRUMPET. Also tried one with
black paint to fill in the holes, which looks just as good.
I now have about 30 pounds of half-kits packed up for the show. Furiously
doing all of the last minute tasks. Plunged into darkness yesterday - the
earthquakes left us without power. That's ok, soldering pretty much done. I
only worry about the flight out. I hope not to get bumped, as a full days
worth of passenger loads was cancelled. Not sure they can fit them all out
today. Many will be on my plane. Hope I am too. Can use the setup time.
Just found out one of the CYMBALs to be used in show blew out. I think a tube.
Might require a little repair work on-the-fly.
October 13, 2006
The new TONE issue is out (www.tone.ws), but they used the wrong artwork for my
ad. Damn. And for the money I fork over for this. So I'm seeing if they can
update the file live. Now that's one cool thing about not having it on paper.
Ok, almost everything ready to go for RMAF. Just need to pack up tools and
spare parts. Everything except the UFOs and pens. Writing up a press release
October 12, 2006
Yikes, still waiting on the UFO parts. They called Monday and said they were
complete, ready to ship. Well, UPS should have had them here by now. I am in
panic mode. My fear is that they show up next week after I have left for RMAF
show. The whole point of the exercise was to have a neat item for show. Three
months in fab, should have been one. I may never use this emachineshop again.
Good concept, poor execution. Dang, two grand down the drain for nothing if
they don't get here this week.
Meanwhile, been cookin in the shop. Building lots of gear and packing up half-
kits for sale. Did the nice candy bag packaging for all of them, even RIPPERs.
Have a whole bunch of HAGUSBs for sale. Trying to pack lightly. So much to
bring. Had my pamphlets done today. All big gear shipped out as of last week,
already arrived in Denver. Everything on schedule except for the UFOs. Oh
wait, I ordered logo pens in bright orange. They're not here either. Crap.
October 5, 2006
Dang, the bottom part to my UFO PRO is still in fab. Expected delivery date is
the 12th. Or days before I leave for RMAF. No spare time. Hopefully I can
get a pair of these finished! Sheesh, I forked over two grand to have these
parts made for the sole purpose of having yet another cool item at the show.
Will not be happy if I spent all that cash for nothing.
Churchill once quipped: "When you are going through hell, keep going."
Oh yeah, right when I ship out a stack of LPs to RMAF my copy of Green Day
October 4, 2006
I can't figure out what's going on with my iMac. Tried to load Audacity awhile
ago, yeah got it working. But I don't think I loaded it right, as it made some
new folder and didn't go where applications belong (I don't know much about the
OS-X). So I think I removed it, but maybe not. Anyway, now iTunes won't play
back at 44.1k. It seems like everything gets resampled to 48k. And now my IBM
machine is doing the same via ASIO. It's no longer working either. It spits
out something, but the VCXO won't lock on a CHIME. I switch to DirectSound and
it works fine. Maybe I found an incompatibility with the new foobar and
ASIO4ALL. I dunno, but need to clear this up before RMAF. I wonder if
Audacity has something to do with it. I experiment so much with various
software and settings on these computers it's hard to get back to a normal
setup. The Mac used to be so easy, now something is hosed. Just can't seem to
Shipped out the last boxes for RMAF today. LPs and the turntable, CLARION. I
just need to load up one crate with a soldering iron, power cords, cables,
brochures, tape, etc. Basically, we're ready to go!
Then I see this comment on phonogram:
"You can find fault with anything, but in a world rich with irony,
iTunes and Airport Express have rendered the CD player obsolescent
and the readily-copied and streamed new digital technology is pushing
kids toward vinyl as the preferred high resolution format. Sure
seems like that to me anyway."
I love it! It means I am headed in the right direction. He makes a good point
that computer transports (USB audio) and MP3s have made vinyl the top choice
again for sonics. I'm also thinking the days of a CD transport are dead. Sell
yours now while you still can.
And so the RGP will focus on the concept of "simplify your life". That is, not
only is it a world-class preamp, but that the feature set and ease-of-use are
paramount and required for the sale. It's not enough to just sound good. And
so I'll need to push the novel remote control, the fact that you need fewer
boxes and (much) fewer cables for a system. Make the machine really fun to
use, a delight. That, combined with customization will give me the
differentiation I need. And so I realize, the chinese built machines are stuck
in a long supply chain. Impossible to customize, other than last minuts things
a dealer can do. Aha, so there's an advantage. Build-to-order, just like Dell
computers. The question is, can I make the supply chain fast enough? Right
now it would be 3 weeks. Place an order at your dealer for a custom machine
built with your chosen options (maybe even initials) and have it shipped
directly from the factory 3 weeks later. The others guys won't be able to
match this. Is the wait too long? Or is that typical for a hand-built
reference grade machine?
October 1, 2006
So then I come across this Scull Communications ad. Some website banner. Ok,
check it out. I find my new mantra: "Sell the brand, not the product". Which
is nice, but it took me ten years to figure out. However, this new Scull
business is worrisome. What exactly is he saying here? He can get you timely
reviews in the big magazines and press releases, etc. if you pay him money? Is
that extortion or just good middleman salesmanship? Only an industry insider
has the contacts to pull this off? If true, then the whole high-end industry
is a racket and they're gaming us. Pay to play. I never believed that to be
the case, but conspiracy theorists might take this website as an example of
Reading this "Extreme Competition" book by Fingar. Modern day business tome
talking about a new world order, Internet, cheap revolution, and everything
else. Trying to get a handle on where audio might be heading in 5 or 10 years.
Definately more net delivery of music, more super-cheap boxes. Mainly, every
thing built in China. Often designed elsewhere, but always assembled in China.
Am I headed in the wrong direction here? Is there any market for craftsmanship
and personalization? I wonder. The book does suggest that high-speed delivery
(customers want that machine now) of a customized machine is the trend. So
maybe the RGP comes in different flavors? Maybe a choice of color? I was
planning on two versions, one with digital, the other without. Perhaps the
ability to order even more choices is better? I'll have to think about this.
Maybe the entire supply chain can be tweaked to allow for customization. For
example, place an order via Music Direct. Order then comes to me where I then
get panels from one company, tubes from another, choice of capacitor yet again,
and all pieces delivered to factory for assembly and test. Perhaps then item
is shipped directly from factory to customer. All of this is invisible to the
customer. They only see Music Direct, or whatever dealer. They order an RGP,
their choice of circuits, connectors, tubes, color. It then gets delivered
right to their doorstep. Hmmm, I'm pondering this customization. I can design
the machine with a choice of coupling capacitors, choice of anodize color, XLR
or RCA, choice of NS or NOS tubes, etc. Of course, taking this too far could
completely backfire. Anyway, long way to go before I finish the book.
September 28, 2006
I read about this new juicy music preamp (stereophile). Looks pretty nice.
I'm curious to see how they do. Difficult, I find to sell e-direct. It's a
nice machine, but he brings nothing new to the table. This is all old school
retro stuff. It's even ugly. I think the aesthetics alone will kill their
sales. Fortunately, he used a CORNET phono architecture, so at least it will
sound good. Hey, for the price, it's a great preamp and I don't think anyone
will be dissatisfied. But to me, it's just a ho-hum "me-too" product.
Then I see the measurements of that tripath amp. Holy cow that's nasty! I
thought this was supposed to be a nice sounding amp, so many people raving
about it. Sheesh, the plots are awful. High distortion with large odd-order
products and glitches, terrible transient response, and a harsh treble. Is
this right? If so, this baby is not going to sound smooth and natural. It
will be sharp and excessively detailed. Maybe you can't hear it right away,
but I bet you can't spend hours in a row listening. This baby will wear you
down with fatigue.
Ok, just finished the CLARION resistor modification. Brought everything much
closer to spec, right where I wanted it. The output tubes are now running at
225V and 56mA. That's a little on the easy side, as I didn't want to push for
maximum power output, but rather a sonic sweet spot. All supply voltages came
right to nominal, with the 2A3 heaters a tad high at 2.6Vac instead of 2.5Vac.
This was the result of the light bulbs not being able to handle the cold inrush
current. As is, the machine is stable, powers up very softly, and is quiet as
hell. In fact, I was worried about DHT designs as almost everyone complains of
hum. The old stories from "fi" in New York were that the sound was great, as
long as you could listen through the hum. Well heck, I don't have any. Even
with my 94dB HORNS, you can't hear anything until you get within 6" of the
cones. And I didn't use any hum pots or other cancellation tricks. This is
raw amplifier and it is dead quiet. The only change I had made was adding tie
wraps to hold ac wires up against the chassis and to put the bottom cover on.
Unit is now burning in for RMAF. Check out the guts:
September 27, 2006
Here it is again, this theory of genetic evolutionary optimization algorithm.
This time in MIT Technology Review magazine. I think this is genius and will
end up big, like the cheap blade servers running Google. So what does this
have to do with audio? Simple, I look back at my business development, one
product generation to the next, and find the same rules apply. Each new
generation is meant to improve on the last. But how to go about this? That's
where the algorithm comes into play. Perhaps this has been unconciously going
on. With each new product I use what worked before and don't use what didn't.
So how can this help me now? Maybe I can experiment with artificial
generations. That is, virtual products. The trouble is measuring results.
Not sure that can be virtually done in a simulator. Or can it? That's how
these computer programs end up with an optimized antenna design or a better
postal route. Specifically, this will apply to the RGP (PROJECT X). And it
seems that I've been following the strategy well. Take pieces that work, use
them again. Combine the best from a HAGCLOCK and TRUMPET. Then a BUGLE and a
UFO. Create some offspring. Measure success. Breed the best. Eventually I
will produce an RGP that is far more optimized than your typical high-end
design. So how to measure success of each generation? This sentence from the
article says it well:
"Bonabeau believes that programs based on the Hunch Engine can help people
with a particular class of problems: the kind where they have a hard time
expressing exactly what they want, but they know a solution when they see it."
Yeah, that's right where I am. Therefore, since this method has proven to be
successful, I'll keep on it, creating yet more and more versions of virtual
RGPs. I've already done maybe ten to twenty revisions, all on paper. Each
time I experiment with a new layout or approach. Like the mute button keeps
coming and going. Presently I removed the S/PDIF and use only USB for digital.
Stuff like that. The generations are indeed converging on a solution that I
Got the new stereophile today, weeks after you did. I find the letter from
Chris Rothko to be fascinating. A clever and insightful theory. I'll have to
think about this one for awhile.
And in EE Times, we get back to the original subject of design. Some guy
pontificating about innovation and how you need a psychiatrist, salesman, and
others on your design team. Blah blah blah, but what I like was his comment
that design isn't just about the hardware, but the experience. To paraphrase:
"Don't just think about designing a better preamp, think about designing a
better audio experience."
September 26, 2006
Ok, so I finally got around to doing an A/B between the CLARION and my CYMBALs.
Same setup, source, cables. Just swap wires quickly. First thing I noticed
was how close they were. The CYMBALs have more gain, so less volume control is
needed, but the sonic differences weren't exactly astonishing. Maybe it just
took awhile for things to warmup. I know the CYMBALs take almost a friggin
hour to get completely opened up. Played various vinyl, sparse vocals and
instrument, to trumpet, to full symphony. The latter showed up the biggest
One odd thing, the CLARION seemed to have more hiss. Not amplifier noise, but
the groove noise from playing vinyl. Maybe it was the different volume
settings I had to run at, but the groove hiss just seemed more noticeable on
the CLARION. The other wierd thing was that the CLARION also had more bass. I
did not expect this from an SET. I think maybe the bass on the CYMBAL is just
tighter and more in control, and so less apparent. It almost seemed as if the
CYMBALs were vocal enhancers. That they painted the vocal track more clearly
and louder amongst the mix. I had always thought this was an attribute of the
HORNs, but the vocal boost was less on the CLARION.
Loudness-wise, the CYMBALs had a distinct advantage. They felt less compressed
and a little more effortless. The CLARION had a warmer and I would say fatter
sound. Yes, fat. That's the word. But still a more distant presentation.
The CYMBALs were more up front, with a tad more punch and clarity. Must be
that dynamic thing. Still, I liked the balanced CLARION sound with it's
definately not rolled-off bass. It played a little better with the HORNs in
terms of frequency neutrality.
In the end, though, I have to tip my personal hat to the CYMBALs. Yes, that
extra punch and the ability to separate and lay bare individual instruments
adds just enough to take the cake. Funny how the tighter bass actually tends
to have a diminishing effect, making the bottom end less noticeable. I can
definately live with either of these amplifiers. It remains to be seen which
plays better with the Azzolinas.
Was this a fair comparison? The CYMBALs are monoblock 8W push/pull versus the
CLARION 3W single-ended triode. Both were open and clear with no audible
amplifier noise floor. Neither uses any feedback. The CYMBALs retail for
$2500, the CLARIONS (if sold) would come in around $2000.
September 22, 2006
Agonized a bit over the RGP architecture last night. It wasn't perfect yet. I
had all the functions I wanted, but there was one tube stage too many. The
extra tape loop buffer bothered me. The topology allowed me to have separate
level control on each input, separate from the volume. This would make each
input roughly the same volume as the next, so when you went from one source to
the next, the volume stayed relatively constant. I was also wondering if the
mono switch belonged prior to the tape loop or after. Anyway, while eating a
hot fudge sundae the concept came to me. I was able to strip out the extra
tube stage and simplify things. I can still sort of have the same functions
logically, so the result is transparent to the user. This is good, because I
was max'ing out on heater and B+ current. Removing these tubes allows me to
switch the last phono stage to a 6CG7 instead of your 6067 (12AU7). The
simplified architecture also plays well with the new style digital section.
Now if only I could find a PCM2903 type chip that used I2S instead of S/PDIF.
Burned in the CLARION overnight (sure helps to have a FRYKLEANER on at all
times), took the edge off. Must have been capacitors. I dunno what, just that
it ends up making a huge difference. Plays loud now with no strain or
harshness. In fact, I can't believe how loud it plays on Ed's HORNs. Super
revealing. Can't wait to get the 680 ohm resistors in there.
September 21, 2006
I am now listening to my CLARION. It's amazing how direct the connection is.
Not sure how else to explain it. The tube and speaker have become one. Ok, so
there is still a bit of edge to it, nothing has been broken in. And the
biasing didn't come out right, but it still pumps out 3 watts. I made an error
in the design - forgot to consider the primary resistance of the output tranny.
This adds about 20V drop, so I don't get the tube running right in the sweet
spot. Both current and voltage on tube came out too low. Easy to fix.
But first, I tried out the light bulb trick. Put in the 2A3s and fired it up
on heaters only. Poof, went the lamp. Ok, try one tube. That's ok. Try both
tubes again and new lamp. This time ok, but the tubes when cold have such a
low resistance that the current load must be at least double or triple the hot
condition. Hence, the lamp is extremely bright for a couple of seconds. I'm
guessing this to be right on the edge. It works great, and I get both soft
start and the exact voltage drops I wanted, but the chance of blowing a bulb
is too great. So I changed it. I split the heater trannys to two lamps. Ok,
so I only get 1/2 the drop desired and the output voltage is a tad high. I
also took off the lamp from the B+ primary. It ran ok, but taking it out helps
to make up for the voltage loss from the output tranny. Now the lamps work on
the output tubes only (4 of them). And bias voltages are much closer to the
design nominals. Anyway, I need to change the 750 ohm cathode resistors to 680
and get a fuse holder. Will have to make some modifications to the chassis.
For now, though, I can play music.
The CLARION came out a lot quieter than I expected. Certain there would be hum
or buzz, I was wrong. It is actually very quiet. Only when I put my ear to
the 94dB HORNs can I hear it. And that's without any tweaking! One channel is
a little worse than the other. I have to do some more scoping around to see
if I can route wires better, whatever. Even so, at full volume you can't hear
a thing from 6 feet.
Square waves came out very fast with decent response at 10kHz. Getting super
wide bandwidth, probably in the neighborhood of 50kHz -3dB. There is a bit of
high frequency ringing from the output tranny. Will see if a snubber or
something can tame that. Probably nothing to worry about, because it is way
out of band. But you know me, I'm pretty anal about transient response.
September 19, 2006
Discovered I forgot to make a spot for a CLARION on/off switch. Oops. Will
add it right above power cord.
Updated the RGP again, this time removing all of the S/PDIF stuff. I like the
way this really simplifies the architecture. Should make the digital design a
lot easier, too. Can just focus on straightforward CD and DAT specs. At least
until TI and Microsoft come up with a PnP spec for 24/96 USB audio. Having the
board modular makes for easy upgrade. Maybe even a 3rd party card. Added a
"bypass" control that sets the volume to +0dB. So if you add a home theatre
input, it is easy to set for either mode. One touch control. Renamed the
"burn" button to "clean". This is the built-in FRYKLEANER that conditions or
cleans the audio path and cables. Will also do degauss on MC carts. Was able
to remove the digital control (since only USB input) so the analog can be
renamed to "input", which I like better. Dang, I can't wait to get moving on
this proto. But the show must go on.
September 18, 2006
Ok, so finally the new ultrafitimes issue is out. Been waiting a long time for
this issue - four months late - as it has a review of the CHIME in it. It was
nice to see that their version of a stereo 2A3 has almost the same layout as my
CLARION. Same exact back panel. I guess there's not much variation once you
logically place all of the components. That's the good news.
The bad news is that they thrashed my CHIME. The article started out ok,
basically a description ripped right from my website. But then it goes into a
philosophical tyrade of the USB implementation (completely wrong in the opinion
of the reviewer). Well, given such a blatant (and exposed) pre-disposition and
bias I'm not suprised he didn't like it. There was a self-motivation in
proving himself correct. No wonder he thought the soundstage was too tall! I
should not be suprised, I took a chance on this one. You see, UFT is located
in the outskirts of Cincinatti, minutes away from Wavelength Audio. I remember
Larry telling me he was a friend of Gordon's before I even sent the unit. It
wasn't such a good idea sending a competing unit into the lion's den. On the
other hand, they did like the HAGUSB.
Fine, so I finally get a bad review. It had to happen sooner or later. It
would have been nice though, if they had at least given me a chance to respond
to their technical concerns prior to publication. I have compelling answers
for all of them. Oh well. Maybe it's nothing to fret about. Their subscriber
numbers aren't much bigger than those reading this blog. At $75 per year, it's
a very expensive magazine, especially when you consider this latest issue
comprises a mere 11 pages. Compare that to TONE AUDIO, which has met every
deadline, is over 100 pages, and FREE!
Rocker did make one good point - that S/PDIF should no longer be used. This is
a very good and forward thinking proposition. And I think I will reconsider my
digital implementation for the RGP (no, it won't be nonos). Just throw out the
S/PDIF stuff. I sort of like that.
Meanwhile, been wiring up the CLARION. Not so painful, this P2P. At least for
a simple and wide open chassis like this one. Most of it done, should be able
to finish the driver stage tomorrow. Have to let a little silicone glue dry
tonight. Wondering what sort of noise pickup I'll get, as the grid lead to
the output tubes is long and ropes over a lot of other stuff.
September 15, 2006
Built another HAGDAC for the show. Nothing like a piece of hardware in your
hand to see the craftsmanship and superior technology. I'm in good shape here,
and can relax a bit and go back to RGP design.
Meanwhile, finally finished the first draft of my new ad. The third paragraph
took me forever. I'm very interested in your feedback. Is it too much? Wrong
direction? Will I be confused with Wilson Audio?
Full page ad
September 14, 2006
So the replacement USB drive comes in. I plug her in and it's a go. That's
it. No hassle. And it's quite fast, too. It backed up all my HAGTECH files
in no time. No more writing to CDs. Ok, that issue solved.
Still cranking along on my TONE ad. Almost done. Rewording it over and over
until it sounds good and makes sense. Right now a bit too egotistical. Will
post it here first, get your feedback before it publishes.
Shipped out 5 big boxes of gear towards RMAF. Thom is letting me store it in
his garage. Also bought my flights and a rental car. Gave myself plenty of
extra time for problem repair (like a cancelled flight). Added it up, and so
far I've kicked in exactly $4000 on this one show. Then add in the cost of
building a CLARION, getting a new P3, two grand on UFO parts, etc. and it's
an expensive thing to do. Sure wish I lived in Flagstaff or Butte so I could
just load up and drive there. I'll be passing out free HAGERMAN pens again.
You won't want to miss that.
And so I think again about my comment below, where an amplifier is like a
woman. That could truly be insightful! Maybe this comparison can shed some
more light. The topic was scultpured chassis design. It's like a beautiful
woman, or a regular one with tons of makeup. It catches a man's eye and
creates immediate desire (let's face it, 99% of audiophiles are men). The
good looking one gets the attention. But what about the nice girl who would
treat you well? She gets passed up. That's the way it is with my CORNET.
Most guys pass on it for something more showy (which they end up selling). So
what is it guys want in a woman? Answer that and you understand what an
amplifier should be.
Got my new UHF magazine from Canada. Hmmm. Looks like ASL has a new phono.
And (oh shock) it uses a pair of 12ax7s for gain and a 12au7 as an output
buffer. Does anyone out there have one of these? Please look at the schematic
in the manual and tell me if it's yet another copy of the CORNET. If so,
should I be flattered? Maybe so! Perhaps it just means all these companies
are years behind.
September 13, 2006
Picked up the CLARION chassis. Nice, this powder coating. It's thick and
glossy. Sort of like a hot-rod car. Started installing parts today, discover
I am short some of the hardware needed, like 2.5 inch #8 bolts. I like to use
only stainless or brass hardware for everything. Nothing magnetic if I can get
away with it. Capacitors glued in and drying. Looks like I left enough
spacing between parts so I can do the wiring without grinding knuckles. All
trannys and sockets mounted using rubber washers.
CLARION chassis guts
Went with blue for the tranny bells. It looks good with the white chassis.
Unlike anything else you'll find out there. So I made me a CORNET in the same
colors. This will make a nice matched set. Maybe I'll sell them after RMAF.
I'm building the CLARION with a blue velvet volume pot, so the match with the
CORNET makes for a superb little tube vinyl system. Just two pieces. Add
speakers and a turntable.
CLARION white/blueCORNET white/blue
Oh yeah, baby. Reading through the stereophile I get to Art Dudley's lament:
"We may not know what we want from life, but we damn well know what we want from
an expensive preamplifier! I look for a balance-control scheme that's easy,
repeatable, and has no deleterious effect on the sound; a good mute switch; a
mono switch; a polarity-inversion switch; and two sets of output jacks, so I
can go back to using a subwoofer if I ever lose my mind again. I'd also prefer
having a built-in phono preamplifier, and, where such a thing exists, I want
two pairs of inputs and switchable gain. I also prefer letting my moving-coil
cartridges see a transformer first, as opposed to an active device of whatever
sort. Everything else is negotiable. I'm not sure such a thing exists..."
He he he. It's coming Art, it's coming. But don't stop there. Why don't you
also ask for variable loading - that you can change while sitting on your ass?
How about a self burn-in mode and built-in cartridge degausser? How about a
one-button remote control that is so easy to use it makes an iPod seem
complicated? Don't limit yourself Art Dudley. Go ahead, ask for more.
So then I whip through a few more pages and come across this "Gryphon" thing.
I'm not sure what it is, but clearly they spent all of their time designing the
outside. Bad move. You see, a piece of audio equipment is like a woman. You
can heap on the makeup, but in the end it's what's inside that counts.
September 12, 2006
Ok, got the show pamphlet almost done. Reworked the design of last one, added
all the new products (6 major machines in 2 years). No more naked chick. Just
a CHIME. Hope to get all of these existing designs settled into solid sales
mode so I can concentrate on the RGP HAGLABS design. My plan is to devote all
of 2007 into R&D. Of course, I always seems to slip in an extra machine here
and there sort of by accident. The HAGCLOCK and HAGUSB were spinoffs of the
CHIME project. You never know when a cool idea will pop into your head. It's
possible I can take my new remote controlled attenuator circuit and make it a
separate DIY board. Anyway, doing well finishing up all current projects and
loose ends. Not sure what to do with the CLARION other than to publish it as
an article in aXp.
Got me a separate warehouse for RMAF. Will start shipping boxes this week. I
just can't wait for the last minute. Lots of stuff to move. Getting me a car,
as it will help with runs to Home Depot and elsewhere during setup. The
independence should speed installation and demolition.
Still working the next TONE ad. It is starting to sound pretty good. For me
to write prose, or good sounding text with proper pace, timing, grammar and
descriptive precision is quite difficult. I'm basically a visual mode thinker
and words are not my thing. Right now I am averaging about one sentence per
half hour. Ad copy is really difficult! You should try it yourself sometime.
September 11, 2006
CLARION chassis are ready, need to go pick them up. Wow, I lave less than a
month left to get everything packed up and shipped out to RMAF. Time is
running out. Have yet to start on the brochure and press releases. Thought I
would do the next TONE ad first. That's actually looking good, sort of going
not where I had planned. Not sure if this is a good move - but changing the
copy to first person narrative. It'll be much more personal and ungrey. Not
the typical platitudes. Again, it's all about branding the name. Focus on the
designer behind the company. That's exactly what Wilson did with their ad on
the back inside cover of stereophile. They praise the man, the driving force
behind the company. I sort of praise myself. Hmmm, maybe that won't come
across quite right.
Found a photo of the long swim last week. I sprinted the last 30 yards on sand
to jump ahead in the standings.
Swim finish line
Still a long wait for the bottom piece for the UFO. But I still managed to do
a fake photo using the parts in hand. At least the overall shape is correct.
Initial pricing will be $400. It's very clever, but I'm not sure it's worth
much more than that. Sadly, the first batch will cost me about $300 each in
parts. And then half a day's labor. So I won't be getting rich. Not until I
can buy parts in quantity of 100. Maybe next one should be gold.
UFO Pro proto
September 8, 2006
Agonizing. The USB drive is supposed to be plug and play. It comes with its
own wall-wart so I didn't expect it to suck any power from USB (maybe it
doesn't). Anyway, plugged it into a hub and no go. But it wiped out the hub.
Ok, plug directly into PC, works. I get a drive letter, copy a file to it. Ok
everything looks good, but I get an XP message that says use a faster port for
high speed operation. Ok, so I plug it into the back. High speed. Everything
works. But my hub is still dead. Ok, power down and reboot. Hub comes back,
all my old stuff working again, but USB drive not there. It says it is there,
I get the bee-beep chime from XP. Just no drive letter and no way to access
it. Sheesh, play around plugging cables in an out, every possible combination.
Plug into another PC, but no drive letter assigned, so it doesn't show up as
a usable hard drive. Crap. Two hours wasted. I'll exchange and try this
again in two weeks. The good news is that everytime I plug in a RIPPER (part
of my testing) it ran fine.
Ok, so I'm changing my prices again. Mostly by $1. I decided to no longer use
the $xx9 style pricing for assembled products. Will stay for half-kits. BLUE
CORNET will be $800. I am also doing the FREE campaign. Yup, a free FRYBABY
if you purchase any assembled tube component. But a BLUE CORNET, get a $250
FRYBABY for free. This offer is good through the end of the year. Basically,
the holiday sale starts now.
September 7, 2006
So it happens again. I get an email from someone who wants to build a CHIME,
and can I please send them all my schematics and parts lists so they can do it
alone. Uh, no. What, I'm supposed to give away all my hard work for free? No
thanks, do your own design. I might have to rethink the way I post full
schematics in my manuals, and post them online. I had always considered it to
be a more honest and up-front approach so that a buyer would know exactly what
they were getting into. I sort of liked being one of the few audio companies
that wasn't afraid to post schematics. Maybe I was wrong. Maybe Bottlehead,
Decware, Welborne, Transcendent and all the other guys have it right by keeping
the design a secret.
On the other hand, perhaps this is a very positive thing. Perhaps immitation
is not just flattery, but a stamp of approval. Maybe I should want everyone to
copy my work? Hmmm.
Got a new USB drive in. For backups and general storage. Music and all. I
expect it to be a little slow, we'll see. Mostly I'm worried my old data hard
drive is ready to go. Same for the picture one. Do you worry about disasters?
I've had many a hard drive and PC go sour. I keep multiple backups of all my
data files. Some on USB keychain, CDs in a safe, my ISP. What is your plan
at work for disasters? For me, a big danger is fire. If the office or shop
burns down, I lose a ton of stuff. All my computer stuff. Most of my hard
drives, even most of the backups. Maybe all my inventory of parts and circuit
boards go down in flames. They say you need contingency plan for all of this.
Hurricane, ok. A nuclear explosion at Pearl Harbor could be really bad. We'd
be so cutoff from our lifeline of supplies. We import nearly everything, not
that much food produced here anymore (pineapple, sugar, taro). Only thing this
state exports is phonostages. At least the wind is blowing the right way.
Interesting study I read. But during wartime or famine or recessions, people
continue to fork over 5% of their income on entertainment. I guess it is
something the human condition needs for survival or to thrive under adverse
conditions. In fact, I believe the study showed the percentage spent on
entertainment went up during bad times. I wonder how that applies to high-end
audio. It does to music. But what about the equipment? If nuclear bombs
start going off, does my business shut down or do sales go up? I'm thinking
that since the equipment is not direct entertainment (like going to a movie),
that sales would slump, as simplified logic would predict. Hmmm.
September 6, 2006
I finished the swim in a reasonable time. Should have been able to do it 5
minutes faster. Ended up in the back of the pack. Gimme some time, I'll think
up a few excuses. Finished in 77 minutes (73 was median). Not too shabby for
a first timer.
Roughwater swim times
Good news is that I'm busy building and shipping products. Thanks for all of
the new orders! I'm finding the RIPPER is a joy to build. Fast and easy. No
extra chassis work or wiring like a FRYBABY or HAGUSB. Parts all done and fit
together perfectly. I built one from start to finish in less time than it took
me to swim that 2.4 miles. The extra cost in custom parts is paying off.
One more UFO part arrived. Just the bottom strobe piece to go. I think it
might be a wait - gotta get on their case. Everything still on schedule for
RMAF. Hope to get my CLARION chassis back soon. That will be a very cool
project to build. Gotta take photos as I go so it can be turned into an aXp
article. I'm thinking that's a good way to advertise. Submit an article or
review for publication. I should think about doing this on some of the big web
sites too. I notice some of them have invited or contributed papers. Hmmm.
Speaking of advertising, working on new artwork for TONE. I like having the
full page to work with. Lots of space to throw away and not get hurt. Made a
new photo, introducing the new HAGERMAN logo. Concept of upcoming ads is
simple branding - just get people to recognize the name. Then hopefully they
come to website for detail info. Been thinking of maybe launching a new deal
on the BLUE CORNET. Or any of the CONSTRUCTOR series. Get a *free* FRYBABY
with any purchase of one of these other units. I've noticed over the past few
years that discounts or freebies work great. For example, 10% off an $800
machine will vastly outsell the same machine priced at $700. The other
effective approach is to give something away for free. So I am considering
using the FRYBABY as bait. I won't make the same margins, but overall sales
should go up enough to cover it.
And then I hear of a new tube phono for sale (elseware) at a mere $895. Looks
like a great machine, and I expect it to sound wonderful. It should, as it
started life as a CORNET - I had offered to sell blank boards (per their
request) with the agreement that it would be sold as a "hagerman inside" - with
some tweaks. It didn't quite turn out that way. They decided to bypass my
board entirely (even making their own copy for testing). Nevertheless, it
still bears a striking resemblance to the original, with an added OA3. And you
wonder why they won't publish a schematic. Fine, this happens. It's business.
It's dog-eat-dog. And I'm a puppy. Sometimes you gotta learn things the hard
September 3, 2006
Article in Business 2.0 on the 5 ways to do a successful launch. They like the
iPod, the Dyson, 787, but not the Segway. Point is to be so clever that you
end up making the product simpler to use and exciting, yet useful. The Segway
is a neat idea, but geeky and totally impractical. Especially for the cost.
It didn't serve a need. They say to start building early, make a lot of protos
and improve on each one. Don't get stuck writing specifications and doing
schematics. Get out an do it. Hmmm, that sound like how the RIPPER came
about. I had the idea for a BUGLE2 and made one. But it didn't really offer
anything new or cool. Add in the USB and recording capability and Bam! New
product. The FRYBABY is also an example of many prototypes (I went through 5
versions, I think). UFO will also be an innovative winner. Once I get these
things scrubbed clean.
Products that don't strike a nerve? The OBOE was a dud. Most of the others
just plain ho-hum. VACUTRACE is unique. The others just nicely designed
machines. A hit takes much more than that.
Oh yeah, they also covered a lot on blogs. I'm doing a lot of this wrong.
Each entry is supposed to have its own permanent URL. Not this one big long
file thing. I chose the simplest method. Suppose I need to update and make it
faster. Lemme think about it.
Big swim tomorrow (what do you do on labor day?). 2.4 miles open ocean.
Weather looks perfect.
September 1, 2006
Unbelievable. I never would have guessed this one. Asked around to all the
machinist and mechanical types I know how to remove the hard anodize from the
UFO proto pieces. Only one guy had the answer, Stu, my Hawaii audio dealer!
Must have been from his days working in the Pearl Harbor shipyards - the answer
is to use Easy Off oven cleaner (sodium hydroxide). No kidding. Spray it, let
it soften the coating. Then wash it off. Ok, so it takes a couple of tries,
but the anodize really does come off. Left behind is reasonably bare aluminum.
I'm going to make a lathe holder that these pieces can then bolt to, and then
run my drill press to spin the pieces while I give them a nice metal brushing.
This will do a final cleanup and make an aesthetic circular brush pattern.
The way it should have been in the first place.
I also wrote a quick CNC program to carve out sections making room for the
batteries. This ran great, first try.
UFO CNC machining
Well, after much experimentation, it looks like chemical cleaning is the only
way to go. I have one piece looking pretty good. Buffing with various
materials on the drill press. Tried to scrape off the anodize - no way. Only
the oven cleaner does it. And then soap and water. Once I get the junk off,
the buffing takes out most of the pitting. So with a lot of work, I think
I'll be able to salvage the project.
August 31, 2006
Serendipity, my daughter brings home a spanking new iBook G4 yesterday from
school. He he. I hook it up to a RIPPER. Much easier to set up than Windows.
Just change the audio I/O for the program you are using (ie System Preferences)
and away you go. I tried out GarageBand. A bit tricky, having to open a new
external instrument track, but it recorded an LP just fine. No need for any
other software on a Mac. Today I'll load up the newer version of Foobar onto
my IBM laptop, the one to be used at RMAF. Better start adding some songs to
Focusing on RMAF preparations. Nothing I can do with the CLARION, as the
chassis pieces are out getting powder coated (gloss white). Also doing a
CORNET. Still need to make me a new VACUTRACE. Everytime I get one built, it
gets sold. Will ship off the CYMBALs to Azzolina, so they can pre-test the
setup. Most everything else is built and running. Time to get the show ad
written. Also plane tix. Thinking I will go in a day early, just in case. I
can then visit Thom and bug him during last minute prep.
What other issues? Turns out I designed the RIPPER so well that it only takes
me 70 minutes to build one. Comparatively, the FRYBABY is just under an hour,
which is not good. Because I expect to sell plenty of them once it shows up in
the mail order catalogs. The board goes together fast, but not the chassis
work and wiring up the panel mounted jacks. Need to tweak the assembly process
so it is faster and better.
Ok, just timed myself at 45 minutes to assemble a FRYBABY. Add in testing,
burning a CD, filling up the packaging, etc., I suppose just under an hour
each. I can live with that. Now I have to figure out how best to outsource
some of this assembly. Maybe hiring some student techs part-time. Depends on
Tried to sand the finish off of one of my proto UFOs. Dang! No wonder they
call it hard coat. Sheesh, I can't even sand it off. By hand, anyway. And
now they tell me it's super expensive for them to refinish. Cripes. I have
to find someone else to do this. Maybe a local sandblaster? Maybe the same
place powdercoating my CLARION? I really down't want to sell brown UFOs.
Finished my ad for RMAF. I'm quote proud of this one. Actually looks decent.
Most of my ads fall flat. I have a concept, but just can't make it go. My big
push now is to focus on just the HAGERMAN brand, without the technology, or
audio labs, or scientific. I just want folks to start recognizing the name.
Nothing more. Simple as I can make it. That's what the new logo style is for.
As time goes on my marketing guy will start to associate qualities such as
design competence, reliability, originality with it.
RMAF directory ad
August 29, 2006
Took me most of the day writing the RIPPER setup web page. The newer Foobar
looks nice. I'll have to update the installation on my laptop. Then start
moving files to it as my RMAF playlist. Don't need much. Probably play CDs on
request. Most of the time might just be vinyl. Then again, I'll be doing live
RIPPER demonstrations so the audience can A/B the analog and digital feeds in
RIPPER setup page
Made a new UFO that works. Discover that the "hardcoat" anaodize I chose was
indeed not colorless. Shoot. Having them quote a sandblasting and refinish.
I can't let this $2000 investment in test parts go down. The LED is perfect.
Once I get it together, it's gonna be beautiful.
Meanwhile having fun with AdWords. Adding one for the VACUTRACE. I should
also do a CORNET. Looks like I can even pick certain websites to place an ad.
For example, the www.foobar2000.org start page. I can put a HAGUSB as right
there at the top, where tons of folks will see it. Question is, do we have the
cart before the horse?
August 28, 2006
Just got some PCM2705 in. They work just fine as the PCM2704 replacement. It
might enumerate differently from the 04, but you'll never notice unless you've
already used the 04 on your machine. Otherwise, install and operation appears
to be identical. So no more hold on HAGUSB. Backorders ship tomorrow.
Then some of the UFO pieces start to come in. Right after I find out they have
trouble finishing the bottom piece (strobe section). Anyway, the top pieces
are dark! Sheet. They were supposed to be the silver color of aluminum. Au
natural. Then it would look like a saucer floating in space. I'll have to see
what happened. My selection was "non-dyed chromic anodize", I just assumed
color would not change. They ended up a very dark brown. Dang, I figured this
to be a hit at RMAF. Now, not so cool, unless I can get them to fix. Beyond
that, the finishing isn't quite the quality I had hoped. Soldered a board up,
and now I have two problems to contend with. There is a short somewhere
between VCC and the battery mid-connection. Can't find it. Gotta be a solder
ball somewhere. Sheesh, slowin me down. Second problem is that my calculation
for fit on the battery holders was too tight. My CAD model must have been off
a tad. Anyway, the top cover can't quite fit all the way down, about a 0.015"
gap. I guess it needs a little machining on my CNC. Plenty thickness to work
UFO proto pieces
August 25, 2006
Pushing sales on the RIPPER. Yesterday added the Google ads. Today someone
suggested I send one to dagogo.com for review. Sheet yeah. I even sent them a
press release. Seems they've been growing and doing pretty well since I last
visited them. A site with good info (just how many can there be?). Also got a
positive note back from Harley (longshot), but he checking with a reviewer with
The Perfect Vision for interest. Cool. No response from Atkinson yet.
Email, TPV said yes! Super, this product launch is starting to come together.
August 24, 2006
Making new ads for next issue of TONE and for the RMAF show brochure. Using my
new HAGERMAN logo so I can focus on branding the name. I think it will stick in
the readers head better than my regular HAGTECH logo. The big market push now
is just for name recognition. Then to associate the name with technical
competence, ingenuity, craftsmanship, service, and reliabilty. I'll keep
posting photos of products in the ads, but they are only to catch. I think the
RMAF ad is coming along nicely. Will post it here first.
Then I realize the PCM2705 is almost the same chip as the 04! Except that it
has an SPI port or something for interface to a CPU. But since I don't use
those features, it should be totally compatible in my boards. So I took the
chance and bought ten of them. Will try it out and see if they work. That
would solve my inventory issue.
Meanwhile, a rewrite to the FRYBABY manual is in order. Something to make
Audio Advisor and the other guys happy.
Ok, just got around to doing something I should have tried a long time ago.
Signed me up for Google AdSense. I am now paying for click-through ads on
other folk's web pages. This is a test to see if I can bring in any traffic to
the RIPPER page. Wow, the signup process is superb (it wasn't a year ago when
I first looked). Very well done. So anyway, we'll see what Google can do for
sales. If decent, then I'll add the BUGLE and CHIME and maybe some other
stuff. It allows you to run multiple campaigns. Maybe that's how they get so
dang rich? Really, I think this will be more effective than a specific banner
ad for $250/month at a single site. The more keywords I throw at it, the more
selective it becomes and the less I pay.
August 23, 2006
School starts again tomorrow. Finally. Got RIPPER product announcements to be
posted at both 6moons and ETM. These are great free resources. Now I need to
get some of the big boys to volunteer for a review. Any suggestions? Maybe
Mikey, I sent an email to John. Anyone at TAS into computer audio?
Ok, didn't mean to irk you guys with the post from yesterday. Yes, I realize
96k is a good feature, even if you don't have the software. My point was that
it was something I missed in my market research. It is not about logical
thought. Sales is about filling emotional desires. The final votes come in
the form of opening your wallet. All I am trying to say is that I think
Meridian made the same mistake I did. They fell into the same logical trap.
August 22, 2006
Aha, I did locate some PCM2704 chips, but the broker wants $14 a pop for them,
in large quantity. Sheesh. I'll keep looking. Seems to be no stock left
anywhere in the Americas or Europe. Only in Asia. At least I have RIPPER
So I'm looking at TAS #164 and notice the new Levinson stuff looks just like
the new BAT stuff, and even Rotel is now copying the style. So many on this 3D
silver & black two tone bandwagon. Then again, it looks really nice! I'm
putting all of these guys in the crosshairs. Then I see some new expensive
cables, and they have that godawful fabric platic sheeth that all the amateurs
use. Doesn't anyone realize the poor and non-uniform dielectric this stuff
adds to your conductor fields? Seriously, take the crap off. And way too much
heat shrink. That stuff is even worse. Finally I read what I am looking for
in the Meridian 808 review: "At present, with SACD as well as DVD-Audio nearly
defunct, the 808's only mission in life is to extract every last bit of
information possible from the millions of titles that are available now - and
for many years to come - on CD." Bingo! That was exactly my reasoning behind
the CHIME. Funny, though, I get a lot of complaints that it doesn't do 96k or
other non-CD rates. And if you ask those same people if they have any music
other than CD, the answer is no. This is where logic fails and marketing
becomes an art. It is not about needs, but emotional desires. I wonder if the
response to Meridian will be the same. Sounds like it should be a good player,
it has most of the features of the CHIME. Or you could just buy a CHIME and
spent $11,000 on a transport, get the same thing but with tubes.
Did all the machining to the CLARION chassis today. All by hand. Drill, file,
and hole punch. Now I just need to get it powder coated. I'm also doing a
CORNET just to see if it looks any better. Both will be at RMAF.
CLARION chassis work
August 21, 2006
Woohoo! Audio Advisor has decided to carry the FRYBABY in their catalog. This
is a big step forward in promotion of HAGTECH. Hopefully some extra sales too.
Changing price to $250. Soon, all orders will be redirected to dealer sites.
Ordering parts for various machines, discovered DigiKey is out of the PCM2704
used in the HAGUSB. New stock in 12 weeks! I have only two left in my bins.
Will have to look around, so far, it doesn't look like any of the major players
have them on hand. Might need to put HAGUSB orders on hold for two months.
August 19, 2006
Got back in this morning. Rented an RV and had driven all over Alaska.
Perfect for kids in tow. Wow, so much to see up there. This ended up being
more of a scouting trip than anything else. Now we know what is up there and
where to go / what to do. My original route was done basically by picking RV
sites in towns not too far apart that we could drive and do stuff in same day.
That worked, but there were a few spots I could have skipped and others that
deserved more time. I really liked Homer. What had been a dot on a map turned
out to have perhaps the best scenic overlook I've ever come across.
Homer, AK panorama
These photos do not do it justice. Across the bay hidden by clouds, are the
most spectacular volcanos (one of which blew big not too long ago). Then some
open ocean, then another shoreline of mountains and glaciers (and Seldovia).
Close in the artsy town and the famous "spit", which we camped on. This is
someplace I could retire. Just a bit cold.
Weather was cold and rainy the entire trip. Oh well, you make the most of the
cards you are dealt. So we had fun despite. Got to see plenty of wild animals
including about 6 bears in Denali.
Me, Portage glacier
On the flight up I grabbed some magazines I hadn't read before. Entrepreneur
and Fast Company. Hey, these turned out to be pretty interesting and useful.
The latter had an article on the evolving Chinese/USA economy. The gist of the
copy was that the "eBay" of China was just starting to take off, previously
hampered by payment issues, and going to have quite the ripple effect. My
"... China's impact on worldwide retailing is going to be astonishing. As with
e-commerce in the United States, the Internet cuts out the middleman, turning
China's 111-million-and-growing Netizens loose on the planet, each scrabbling
for something that will move online ... The idea of all of them swarming online
is almost too much to contemplate ..."
Basically, once the payment (PayPal) and shipping (UPS) issues are tweaked, it
is predicted that the middlemen will be eliminated by e-stores. That means
prices will drop under fierce competition. In high-end audio terms, that means
the Upscale Audios and anyone else importing low cost gear are about to be
bypassed. It also means those shiny $1500 50W tube amps will soon be available
for about half that price. Will the market really change that much? I don't
know, but it will be a frenzy as the attempt is made. For me, it is carnage to
stay away from. It will be impossible to compete in the $1000 and under
I also came across a good find for my remote control, an OTS plastic case that
I can modify. It's a way cool oval shape with LED hold and AAA battery slots
built in. All I have to do is add a circuit board. But of course, I will take
it further and add a metal faceplate to match the RGP. Can still use my custom
knob. It also cuts development time and costs. Yes, the pieces of the puzzle
are falling into place.
OKW remote control housing
Ok, just put the top 10 photos from the Alaskan trip. Nothing spectacular, as
most were shot in the rain. But it does give you a good view of what the place
sort of looks like.
Top 10 Alaskan photos
August 7, 2006
More chassis study. I notice there is a difference in the UK market than US.
Components sold and designed there seem a bit less extravagent. Maybe a tad
smaller and perhaps more practical. Lots of solid state stuff. Anyway, the
big trend I see in casework is the two-tone color scheme. Everyone is now
mixing silver and black. The other trend is away from flat faceplates,
everything is 3D now, with grooves, notches, flaps, ridges, and bolted-on
extras. Each manufacturer is trying to outdo the other in sculpture and
radical curved protrusions. All for naught, in that this visual drama adds
absolutely nothing to the sonics or product value. This too will be an
advantage for me. Manufacturers will be distracted by this false need for
visual impact. Meanwhile, I can focus on innovation, sonics, and features that
actually provide value. Casework is one of them, but not the superfluous
contest of artifacts going on. Again, I use the iPod as my guide. Simple is
better. Beauty can be practical too.
Cool, get this. I can also run a FRYKLEANER to the LINE inputs too. So burn
in is not limited to phono stage. I can also fire up the line inputs, which of
course does linestage (so did the phono burn), but notice that it will also
break in the cable! The RGP works in reverse here. And when the source (CD
player or tuner) is turned on, then it's low output impedance is seen by the
cable and the burn-in turns from voltage mode to current mode. Cool. The
volume through the linestage will be very low, but that's ok, we're doing a
cable burn. Hey, this will even burn in the output stage inside the line
source. All of this can be done without tearing apart your system. Just
select the input, go to BURN, and then crank up the amplitude. When done, turn
it back down. I will also do this automatically for you. For example, if you
press the knob, operation of controls goes back into select mode. Rather than
jump out of burn, the CPU will crank amplitude down slowly for you and then
switch modes. Same goes for changing inputs. Each change gets an automatic
mute, volume down and volume back up. Very quickly, but a fast fade and mute
keeps the pops and tics away.
Yippee! The RIPPER chassis came in today. The absolute last possible moment
before I go on vacation. So I'm working all night to get the manual finished
and all backorders filled. RIPPERs are shipping tomorrow. Oh, and the panels
came out perfect. Everything is a super fit. Chassis look very nice. When I
get back, the marketing blitz will begin. RIPPERs out to reviewers, notices,
ads, the works.
August 6, 2006
New stereophile, pretty funny article by Dudley. I like his "woodies" and
"fuzzies". Also notice that the new Cary line has some things in common with
RGP. They added an RS-232 port for computer control. I had this. Then I had
an ethernet port. Took them off as too much feature. Do folks really want a
serial port on their linestage? It's easy to do. I can make simple text
commands for everything. Even a windows program with buttons and sliders for
complete remote control. But will anyone use it?
Ok try this. Got rid of mute control. Heck, just turn the damn volume down.
It is super fast, no need to sit there and push a down button fifty times. I
changed the RGP panel again, moving things around logically. Added a "burn"
control, which maybe should be changed to "degauss". This initiates a built-in
FRYKLEANER that degausses the selected phono input and burns in the phono,
digital, and line sections all at once. The output is muted, but not zero. So
when burn is selected, the volume starts at zero, you dial it up to a level you
like (maximum is good in power amps are off), then dial back down when you are
done. This way you can do a 30 second degauss, or an overnight burn. I also
moved the balance control to the input side. That way it is sticky. It is set
independently for each input. Normally, this is always set to zero, but in
some cases you may want to correct for an unbalanced cart on a specific input.
I also removed the LEDs showing gain and clock rate. Like the other controls,
the center LED ring will be the indicator of setting. Gain is either high or
low. The clock rate will be the same as that if the selected digital input.
So 44.1k for USB. Anything for DIG1 and DIG2, and a default 24/96k for the
INT internal input. The lock LED can go, I'll just blink the selected input if
there is an error or problem with the signal. Moved the error red LED to under
knob. It will blink for all errors, clipping, etc. Anyway, I like the
assymmetry and cleaner look of the panel. The added FRYKLEANER feature will be
great. For the analog-only version, the LEVEL, DIGITAL, and MONITOR controls
Bought me the latest issues of Hi-FI Choice, UK magazine. Nice because the
photography is really good. Big pictures so you can see everything. Makes
market research easier. And then I spot the new Krell knob is just like the
one I thought up! Shoots. Everyone will think I copied them. Been looking a
lot at chassis design and casework. Colors, everything. I like the look of
titanium with blue LED. It's possible I could make the RGP in the color of
your choice. Then I look at top plates. Most have slots for vents. The Krell
has holes, kinda nice. Some have wire screens. I'll consider that one, as the
stainless screen is a little different, offers much better airflow, and lets
you see tubes inside. Only drawback is that it's not quite as rugged for
August 5, 2006
TRUMPET reviewed in that fabulous new online audio rag TONE. It's free and as
good as anything in print.
My P3 arrived. Got it installed and all setup with the Goldring 1042 that
Wayne gave me. Wow, this thing sounds good. Very transparent and neutral, no
issues. Just clean and natural sound. The Grado was much thicker in the
middle with a overdramatic and organic midrange, but it seemed a bit wooley to
me. Maybe it was the TD124. So yeah, I traded the set in. Mostly, I needed
the ability to run an MC cart, and the iron platter on the TD124 (early model)
would not allow it. Right off the bat the P3 was enjoyable. I sat in the shop
and played two album sides when I should have been spending just 5 minutes, was
supposed to be getting dinner (they waited). Suprised the 1042 was so natural.
This was one of the carts I ran a FRYKLEANER into. Yeah, I powered it like a
motor. Probably works just like one of those degaussing things. Hmmm, maybe I
need to add this to RGP? Anyway, the cart sounds just fine. For the MC, I'm
leaning towards a Benz ACE or Dyna 20XL. Something in that ballpark.
Hey, think of this. How about I connect a FRYKLEANER-like signal to the phono
inputs (RIAA corrected) of the RGP. Normally disconnected. But, it doubles as
a burn-in device and a cart degausser. Hmmm. I like it. The beauty is that
the cart is connected, so the signal goes both ways. It drives the cart, burns
the phono cable, and burns the phono amplifier too! Even step-up if it is in
the selected path. And that's not all. It self-burns the linestage and/or
digital section too! The linestage is strong enough that it ends up burning in
the interconnects. Since the machine is CPU controlled, it is possible to have
the machine do an entire self-burn on itself. Or on any selected channel when
the user feels it is necessary (like after months of non-use). The amplitude
can be controlled for a smooth release, important for degauss. Now, the
question is how to trigger this. Running out of room on the front panel. It
should not be as easily accessible as, for example the volume control. I have
a similar problem with the cartridge loading, it's very easy to get to and
adjust. Right next to input select. Could end up getting twiddled by mistake.
August 4, 2006
Trying to figure out how to place all of the parts into the CLARION box. Boy,
it sure helps to have the actual pieces in front of you. Chassis design is
such a 3-dimensional jigsaw puzzle. More than one answer. You have to not
only fit the parts so that it makes sense, but also visualize the wiring and
grounding as you go. Not to mention imagining where the stray magnetic and
electric fields will be. That's why I think I'll route the inputs via coaxial
cable all the way around the outer rim, put the power supply stuff in the
middle. There is some unavoidable crossovers, as the filament trannys for the
output tubes are in the signal path. The photo shows an approximate fit. I
then realized the electrolytics are too tall for the chassis! So now I guess
I will lay them down. Need a terminal strip in middle so I can have a
reference grounding point and run a bus bar between the big capacitor ground
taps. Then I found an ALPS blue velvet volume pot I had for a terminated
project. I think I'll put that in to, just for the heck of it.
The DIYer of the month award goes to Bob Gutz for a tricked out BUGLE.
Ugh, I think I just messed up my iMac. Played around downloading a few songs
from iTunes. Why do I do that? Played em back and they sound like shit. The
bass is woofy and everything is dead and dull. It sounds worse than my car
radio. Damn encoding. Anyway, tried to hook up my wife's iPod, never figured
out how to tranfer songs. Sheesh. The played with the RIPPER. It is
recognized properly and immediately by the Mac. Just have to actually select
the USB AUDIO CODEC for both input and output. Then I downloaded Audacity.
Works great on Mac too. Well, after some trouble, I was able to configure it
for the RIPPER. Made a track or two and could play it back. Unfortunately, my
new turntable is still at the shop, I'll pick it up tomorrow. So no ripping of
actual vinyl tonight. Then I screwed something up by moving the Audacity
folder. Didn't like it on the desktop. Ok, so I have no idea how to properly
install programs on the Mac. Now everything hung. I had to eventually yank
the power cord. Most working again, but now I lost my ability to playback from
iTunes to my CHIME. I think it is doing resampling, but I can't find the
dialog box. It boot ok, but once iTunes is launched, the PLL lock LED on the
CHIME goes off. Can't play CDs, nothing. So at the moment, my Mac is screwed.
At least I know it works with the RIPPER. But now my home audio system is
August 3, 2006
Ugh, just finished a pair of FRYKLEANER EXTREMES, a custom job I did as a favor
for a customer. Never again! Holy crap, it took me a full day just to do the
metalworking. Two and a half days in all, plus a lot of extra cash. What is
an EXTREME? It's a bullet-proof, I mean a metal chassis FRYPRO. I realize the
standard product is a little on the lightweight side. The heavy duty chassis
is nice. Thought I might turn it into a product, but no way. I'd have to
charge double the retail price to break even. Way too much work. But it looks
really nice and I'm done.
Want to do it yourself? Ok. Start with an #HM338 box from DigiKey. I also
picked out some more robust binding posts, but forget where they came from.
Step 1) Cut the wings off from the front & back panels. Not the side part,
just where it wraps around. 2) Apply a thin film of expoy and glue down the
plastic FRY PRO panels to the metal panels. Let sit overnight. 3) Trim off
the excess plastic on top and bottom. 4) Using your 3-axis milling machine,
cut out the XLR holes and outlet cutout. Step drill won't work here as the
metal heats up too much and melts the plastic (I learned the hard way). 5)
Drill 1/8" pilot holes in the center of each control or connector hole. 6)
Using a step drill, cut holes for all of the RCA jacks. Unfortunately, these
need to be insulated from the metal chassis panel, so you have to overdrill
just enough to get the plastic insulator ring through the plastic, but not the
metal. Add an insulating washer to the other side and secure. 7) Do the same
for the binding posts and the rest of the panel mounted parts (steps 3-7 take
about 4 hours). 8) Once all parts are mounted, epoxy the LED and ac outlet in
place for a more secure fit. 9) Finish up mounting boards and wiring as in a
regular FRY PRO.
Speaking of design, style, and fashion, I saw a Dyson vacuum cleaner for the
first time. Wow, beautiful. And the coolness and unusualness was not just
for the sake of sculpture and art, but driven by practicality. That's where
design is good. When you can marry sculpture and usefullness into one. That's
why the iPod is a winner. That's where much of high-end audio is a loser. So
many companies are focusing too much effort on visual style and appeal with no
practical benefit. All style and no substance. It becomes meaningless. You
know what I'm talking about, there are so many examples. It has to be about
the sound first. Second the simplicity and convenience of operation. Fashion
is great, but it has to be in harmony with sonics and customer integration.
August 2, 2006
Here's what I was thinking of for a knob. Regular rounded knob, but you know
the scroll type that have the dimple? This one has 6 of them, just to be cool.
Fancy machining puts the cost up at $9 in large quantity. That's ok I suppose,
as this is the center of the universe. The remote will be home base for a
customer and it may as well feel expensive. The skirt will be located behind
the front panel.
Just got notice that the RIPPER chassis ship tomorrow. I should have them in
time to complete all pre-orders on Monday. Sorry for the delay, I had promised
to ship in July.
August 1, 2006
Aw crud. I can't get the good times font into front panel express. Not sure
if I can use it for the RGP. We'll see, I asked if they can do a conversion
for me. Did find me the rotary/push control. Nice piece, very rugged and
solid. Should be able to take all sorts of abuse, which I expect from guys
sitting on the remote control. Only odd thing is that it takes a size 12mm
shaft. Now where the heck do I get a knob for that? So I made up one in the
emachineshop program. Can make it look like anything I want. My opportunity
to have my own brand and style of knob. Pricing looks to be about $7 if I buy
100 of them. The closest thing I can find is a nice aluminum knob at DigiKey
for - $7. I hope to stay away from custom parts. They really jack up the
retail price. Could just drill out the middle of the DigiKey knob.
Wondering if I should move the FRYBABY under the HAGLABS banner. That is, if
I can get this retail giant to pick it up for their catalog. I forget why I
left it where it is. Might want to place all dealer-only items in HAGLABS.
Need to keep it non-confusing.
Tried an experiment today. Modified my personal HAGDAC to bypass the reclocker
so it can also run at 48k. Maybe I didn't do it right. Trying to enter this
home theatre mind-set by connecting my CHIME to the el-cheapo DVD player we
have. It has a digital audio output. Was guessing it to be 48k. Anyway,
didn't work, not sure why. All I know is that when we play DVDs on my tiny
iMac screen I can run the sound on a CHIME.
July 31, 2006
Here's an investment for you. Buy up the PCM1704s. Looks like TI has classed
this an obsolete part - not recommended for new designs. Too bad, it's perhaps
the ultimate multibit converter ever made. Maybe the last one. If so, then
some day it will have cult status like the CS8412 and TDA1541 chips used by all
the DIYers. Note the CS8412 sells for $30. It's a gamble and longshot, but
could the same thing happen to the PCM1704? Will it someday catch $100 a pop?
Hey, pretty good guess. I added up $625 in parts for the CLARION. Using the
cheap tubes. Then I realized they don't sell the 6H30! So I'll change the
design to use either an ECC99 or a 5687 for a driver. Fine, I keep the 6H30
Finally solved my chassis problem while putting in a hill run. It's so simple
I'm embarassed. All of you would have come up with this already. Been doing
mental gymnastics trying to come up with something clever, but always trying to
do too much with too few pieces. Stupid. Anyway, it comes down to a simple
little L-bracket. Now everything falls into place. Best yet, I can apply the
architecture to the COMPRESSOR. Now I can have common parts for all of my new
chassis designs (everyone else does it). Mainly, I wanted no visible screws on
the faceplate. That simple requirement made assembly and disassembly a
nightmare. How to get a screwdriver and knuckles into certain invisible
spaces. The topology allows me to manufacture very high quality appearance in
minimal numbers (crucial during ramp up). Then, if sales volume permits, buy
in bulk for some significant cost savings. Even the back panel is just as
beautiful as the front. The remote matches too. I almost hate to admit I've
been working on this issue for several years.
Boy, realized I have been spending a fortune lately. Investing a ton of cash
on R&D, production, and now advertising. Bought me a new P3 travelling show
table, getting two new carts for it. Just ordered parts for my CLARION, got
big bucks invested in UFO and RIPPER production parts. Fully paid for RMAF
already. And now my full page ad hits the street in a few days. Like they
say, ya gotta spend money to make money.
July 30, 2006
Been thinking about logos. I like my HAL graphic. Nice and simple. But it
means nothing to someone not already familiar with it. Somehow I have to tie
it in with the words "hagerman audio labs". Or maybe I can do something else?
How about just the hagerman name in a cool font. It would then apply to both
companies (audio labs and technology). In fact, I've been pondering adding a
third (scientific), to spin off all of my non-audio related stuff. I do have
some patents pending that might be worth cash if played right. Can put these
under the scientific banner. That would make for easy sale. Really, audio is
my true love and that's where I want to stay, even if I can make more dough
doing the other stuff. Maybe an unbrella logo like the following?
I really like this "good times" font. Powerful, modern. Easy to engrave. Now
I just need a way to integrate the three division names.
July 29, 2006
Ok, I fixed the stupid mistake in my CLARION 2A3 design. Ended up doing a self
bias on the output, regular capacitive coupling from 6H30 driver stage. Full
choke rectification, finely tuned for perfect power supply step response.
Seven big chunks of iron, four tubes, five lamps. A fairly basic design, with
not much new. Except my one contribution to the art - lamps. I solved a few
problems by using incandescent lamps. Mainly, the power trannys are rated for
117V and 115V respectively. Running these at 120V nominal makes them run hot,
as the primary inductance starts to creep down and saturate. The output
voltages also run high. So my trick is to add a lamp in series with the 120V
line on tranny primary! That way, I get near perfect input voltages at load
and the lamp doubles as a fuse. Third, it adds a bit of resistance that when
used with an 0.01uF capacitor across the primary gives me a low pass filter to
help cut out line noise before it even gets to the tranny. Very helpful on
heaters. Now with 5 lamps, you get nice indications of operating status. If
the tubes are not installed, the fuse/lamps do not turn on. They also show
heater voltage ok and output tube bias current. Basically, everything has to
be up and running properly for all lamps to be on. Now I just need to sort out
what colors they should be. I'm going to try a neat powder coat combination
that Stu came up with, white gloss chassis with blue gloss tranny bells. It
looked really deco and cool.
Ok, market research again. This time stereophile. I see McIntosh has a new
preamp to compete with RGP. About $17k. Great specs, clearly it was designed
for superb measurable performance. Lots of feedback no doubt. But what caught
my eye was the comment by Fremer; "It's easy to get lost in the maze of
possibilities, and once you've made a change, it's easy to forget what you did,
and damned difficult to figure out how to get back to where you once were."
Exactly my point! Just because you can do something doesn't mean you should.
Having a CPU in a preamp (RGP) can get out of hand. Flexibility is great, but
it has to make sense. It has to be practical and easy to use. Even for a four
Does anyone read the manufacturer's showcase ads in the back? What about the
classified ads? I once ran a 1" classified for the Trumpet for a year. Not
sure the response was good, yet had no way to measure it. Could be a bargain
if folks actually read it.
July 28, 2006
Just got my issue of TAS in (it comes by boat). The letter from D.R. is
hilarious! This guy has a way with words. Then as I scour the pages to do my
market research, I notice several things. One, the cover says "giant killer".
Ok, these magico speakers look very nice, but at $22k a pair, just who is the
giant they are slaying? Exactly where is the threshold now between david and
goliath? Anything under $25k is now the little guy? Then there is the full
page ad from AVA. How can these guys afford this? This has to be $3k a pop.
And $5k in stereophile. Do they really sell enough amps to pay for this? It
boggles my mind. And if you're going to fork over this much cash for a color
ad, can't you at least use a decent photo? Not only does the machine look like
a 50's vintage dynaco, but the color is oversaturated, it's blurry, and the
wine glass casts a shadow across the faceplate. What are you guys thinking?
This reminds me of the turntable ads a year or two ago promoting a shiny ball
bearing enhanced table with stars and cosmic galaxies and shouting words all
over. I can't remember the name, but they also made astronomy equipment.
Purchased a year of expensive full-page ads and then disappeared. Let me take
this as a lesson that there is more to marketing than blowing cash on full page
Something else puzzling me. I see a number of new construction projects going
on around town. And they are ALL storage places. I swear, four or five brand
new rental storage facilities. What the heck is going on? We must have 4x the
storage space we had two years ago. Who is using it? What is so different now
that we need these? What changed?
Been working the CLARION design. Coming out really nice. Easy to do this now
after all I have learned on the other tube designs. Hmmm, I wonder if I can
turn this into a product. For now, I'll make a P2P proto and publish the info.
Will make for a super DIY project. Part cost about $600. I'm using all the
best parts available from AES. One stop shopping! Adding some nice touches,
like jewel bulbs to indicate heaters and output tube current. If your output
tube fails, the light goes off. Comes on with B+ if everything else is ok.
Did a full choke input, critically tuned supply. Split off LCRC decoupling
separately for each channel, so this is mostly a dual mono design. Oops, made
a big mistake in tube biasing. Doh! Easy to fix. Maybe that original power
tranny I bought is the right one after all. Lots of good iron, simple circuit,
polyprop caps, expensive electrolytics, all mounted to a 10x17 aluminum
chassis that I plan to have powder coated.
July 27, 2006
Wow. Talk about letters to the editor! This one to EE Times magazine really
bowled me over. Much to ponder. I really appreciate the huge vision this guy
demonstrates, a long-term view covering maybe a century. To me it is food for
thought. Am I doing the right thing? Am I positioning my company for success?
We are in for a long dark night...
And then this one to Forbes, talking about the estate tax:
"The fundamental lesson of Adam Smith, rooted in common sense, and confirmed in
the laboratories of history, is that an economic system must align with
ordinary moral principles to allow society to flourish ... The death tax comes
to the industrious, the thrifty and the altruistic. It spares the unproductive
the spendthrift and the selfish."
Holy crap. That's a lot to digest. Never thought of the connection to moral
and ethical principles. But it does indeed make sense. Why punish those that
create and reward those who destroy? Ok, so this is not exactly audio. But to
me it is. It's all about how to run a successful long-term business. Loosely.
Meanwhile, I jammed on the RIPPER kit manual and have it mostly complete. It
is now online minus the photos. Forgot to include changes for the line-level
So what sells? Philosophically speaking. In this hyper-competitive world with
China tugging the bottom of hi-end audio into commodity pricing, what will
work? Yeah, I keep asking the same question over and over. My latest answer
is "must-have pizzaz". That's what the iPod has. Style, simplicity, and fun.
But without sacrificing practicality. It has that "hey, let me try that thing"
appeal. It looks cool and it's different. That's the draw I need to go for in
my products. And I think I've stumbled into it already with some. The RIPPER
will have that must-have pizzaz. So will my assembled UFO. And my latest
round of changes to the RGP are also in that direction. I took some more LEDs
off. Simplified. One knob control. When in use, everyone will want a go at
that remote. Really, it's going to be a joy to play with.
July 26, 2006
Realized I need some fancy test equipment for debug of the RGP digital section.
Need S/PDIF sources for all clock rates up to 192k. And a DAC that can read
what the RGP ADC spits out. Or maybe I just use two RGPs?
I modified my original CLARION design to remove the coupling capacitor which
drives the 2A3. Using a long self-bias tail on the output tube. Discovered I
can use the same power tranny as the Trumpet. Have the new pencil design done.
Uses a 6922 driver to 2A3 to UBT3 tranny. Choke input on B+. Changed parts
such that everything is available from AES, probably about $600 for a stereo
3W SE machine. Part choices are limited, but I think I can make this sound
pretty good. I already have most everything in stock.
Finished my new ad yesterday. Doesn't quite look right. Had to send out for
deadline, though. I'll revise it for next issue by replacing the timeline with
text. The idea here is to demonstrate company viability, integrity, solidity,
and innovation. The ad campaign is focused on "branding", as most people have
never heard of us. Forget the individual products, just get the name out
there. The photo is totally unposed. It was the first test shot just to see
if the camera was set up properly. A dozen subsequent "posed" shots all looked
Full page ad
July 25, 2006
Interesting editorial regarding RoHS in Electronic Design. He brings up this
"take-back" program, where manufacturers have to assume responsibility for
recycling their products. Hmmm. Maybe I should do this. Whenever you decide
you don't need a Trumpet or VacuTrace or whatever anymore, send it back to me!
Don't sell it on audiogon. Don't keep it for your children. Let me recycle it
Maybe I just need a sticker that says "RoHS exempt - military grade
electronics - high reliability". Or perhaps medical grade. Isn't quality
music a theraputic event? Or does music fall under telecommunications?
Couldn't figure out how to engrave my logo in frontpanelexpress. No line or
arc drawing features. But it does have an HPGL input capability. Ok, so go
learn HPGL - the classic pen plotter language. Pen down, pen up, etc. A web
search gave me what I needed and came up with a working answer in about two
hours. I did it in simple integers so I could end up scaling it to any size.
Put it on the proto concept for my remote control, which came out the same size
as an iPod (kharma?). But 1" thick, like my cell phone. So if you hold an
iPod upside down, then the scroll wheel is in roughly the same position as my
encoder/button wheel. Less bend to the thumb, so it should be more comfortable
Ran across a whole bunch of CLARION parts in the shop. Extra chassis, tubes,
trannys. Shoot, why not just build one? Original plan was for fancy ultra-
sonic heating, but why not just hand wire up an ac heated version. It won't
be a product, but would make for a nice project paper. It can also be my
backup unit for RMAF. Would be nice to compare PP to SE in an A/B situation.
July 24, 2006
Touch race yesterday. Got hot and I dehydrated on the final mile. Was able to
dig and finish well. Took home a 6th place, not bad for the amount of training
I've been doing. Actually finished 3 minutes faster than last year. Pushed
the bike leg hard as that's where I am weakest. Got off to a good start with a
30 second lead on Griffith. But a slow T1 put him right on my tail and he
passed me at the fountain. Held off Huddleston for lot longer. Three others
went by too. Caught Reinke at Triangle Park. By the time I got to the run I
was back in 6th place. Quads were cramping and I was unable to get into high
gear. Hot sun, not enough salt, too much pain. Halfway through the run I
caught Huntington. Celebrated in the photo below. Seconds later I hear a
cheer for Reinke, who had pulled up within seconds. I didn't look and gave my
last surge. Held him off for a mile, until I finally caved at Triangle Park
(again). Not a bad run at 45:18.
Tinman run photoTinman Hawaii results
RIPPER boards came in. First one mostly built. Waiting for chassis. They
had contacted me on Friday with a graphics problem. I adjusted the artwork.
They should be able to finish now, all of the milling had been done. Still
expecting them in this week.
Meanwhile a customer wants a ruggedized FryK Pro. Wants a heavier and more
durable enclosure. So trying to work that out. Found a steel chassis that
actually fits really well. Upgraded binding posts and other parts should
make this nice. If it works out well, maybe I'll offer it as a product. A
"FryK Pro X" for extreme. He also wants an octal CHIME. So looking into that
possibility as well. No, I usually don't do custom work. Only when it might
lead to a new product or a nice application note. Otherwise I try and stay
away from it - way too much development time burend up.
Must work on my new ad.
Did some quick design exercises on a remote control for RGP. It's all custom,
one knob, machined from solid aluminum. Top piece would match the front panel
of RGP, bottom a black anodized shell. Trying to shape it nice with rounded
edges and smoothness. Looks like I can build these for maybe $100 in parts.
Not too bad in larger quantities. But hey, it's a class act. If it's going to
be a reference class product, it needs one of these.
The web stats show we're getting 30 hits/day on this blog. Not bad, really.
Website averages over 1000/day.
July 21, 2006
Maybe I need a break from this. Realized the new RGP topology fit in nicely
with my original control concept of one knob. So I have yet another possible
configuration for front panel and remote. This is sort of where I wanted to go
all along, kind of an iPod-like interface. The drawback is that you need
visual feedback from the box when using the remote. Otherwise easy to get
lost. Or maybe not. Once the panel is memorized, you can do it blind.
Anyhow, this concept has one knob that is both rotary encoded and a momentary
pushbutton. Much like a car radio knob. First mode is to select a control,
scroll left or right (CCW, CW) until you light up the desired control. Push
once to change mode. No the wheel operates on that contorl. For example, 6
clicks to the right is volume, select it, and now the knob is a volume control
with 31 steps. Push and you go back to select mode. In this fashion you can
adjust everything via the remote too. All with one button. This is the iPod
paradigm I wanted.
RGP one knob concept
I still have the nuclear power plant syle LEDs for everything. You always
know what the settings are. Hmmm, come to think of it, I don't need to limit
myself to 6 loading values. I can now make this 31. Well, I wanted to be
different. One cool thing is when the recording level control is selected, the
big round LED display turns into two channels of VU meters, with a peak reading
dot. Such will make it super easy to set proper recording levels for the ADC.
Stylish, fun, practical. That's what I have to aim for. Something so
intruiging that you just can't wait to give it a try. In that respect, this
latest alternative might be best. Gives you the best combination of control,
simplicity, and fun. And this example converts to an all-analog model easily.
The one drawback I can see is that it loses the balance on input side prior to
recording. I'll think about what I can do about variable EQ.
Big race this Sunday. And not just the TDF. Must get sleep.
July 20, 2006
Thinking about alternative front panels for RGP. This time organize not around
hardware, but by function. Have an idea to move balance to the input side, not
on output. That way, each input can have it's own level and balance setting
such that volume stays constant as you switch inputs. That is, you don't have
to monkey around with volume and balance when switching between CD and phono.
However, it appears this would require an extra buffer. I like the concept, as
my recording on the RIPPER shows a balance problem with the cartridge. Do NOT
want to use sliders in Windows to fix this. So a balance capability prior to
recording is good. Could still have balance on listen side too. This is kind
of nice as it equalizes all inputs. On the other hand it adds more circuitry
in the signal path. Adding an opamp would be unforgivable. Hmmm, and now
switching back and forth between analog and digital while recording would be at
same volume. Dang, just finished the power supply design. How can I add more
Nothing like necessity to bring out invention. Came up with an interesting
solution, a modified cathode follower. I was trying to stay away from these.
It adds no capacitors. Eliminates miller capacitance on the analog inputs too.
Hmmm, this might work. Adds a lot of features and capability for the cost of
a pair of tubes. Operating point not too critical so maybe a choice of tubes
for plugging in here. Found a way to add the extra current in the power
supply. Was already pushing the limits, but a simple trick there gets the job
done. Adds a little thermal margin to the power trannys. Who said innovation
in tube circuits is dead? It's a monster with 10 tubes and 9 chokes and 4
Speaking of which, great article in FORTUNE 7/10 page 53. Very uplifting. It
means I got it right.
Ok, so here's the alternate proposal. Add in the CFs, move balance to inputs.
The LEDs for balance and level can display in two modes, one the amount of
attenuation or balance, second as a live peak reading VU meter. The latter
might be better, although some might find annoying. It'll help to set nominal
inputs levels, perfect for recording. Then each input uses same volume (now
listening) setting for a given loudness. Th RGP will equalize your system such
that proper recording level is always set. Sure, this will vary from LP to LP,
or one FM station to the next, but on the whole keeps you from having to
fiddle. Very flexible arrangement. In normal use, super simple. Just select
input, volume (listen), and phase. Hmmm, maybe I move mono control to input
RGP1 (analog only)RGP2 (with digital)
July 19, 2006
Got an idea for my book. The RGP will be a work of art with lots of
proprietary tricks thrown in. So the product manual will probably not come
with a schematic. Nor should I publish it online for free. However, maybe
there is some value to making a technical manual describing everything in
detail. Say, a 120-page book on lulu.com. Could be as good as a Morgan Jones
book, even though it focuses on one product. A lot of design theory could be
included for application to other designs. Make it general, yet focused. I
still haven't given away details on my proprietary passive regulation scheme.
Depends on how sales go. Such an endeavor will chew up a lot of time.
Some news I've been holding back on. Mostly it's just difficult to face. But
one of our Honolulu Audio Club members was killed about two weeks ago. He and
his wife were trying out a new digital camera at a hilltop city overlook when
some lunatic came up and shot them. Taxi driver shot too. Been in the news
here. Very disturbing. I didn't really know Jason, but do remember talking
with him at several of our club meetings. Jolly sort. Really, I don't know
what to say here and apologize for the rambling. Words not my forte.
RIPPER boards shipped today! Chassis to be ready tomorrow. I should be doing
photos and manual next week. Shipments by end of month. Sorry for the long
Hey, check out the nice font at the clarisonus blog. Georgia. Nice. I might
have to use that somewhere. Very readable, yet vintage.
Can't figure out a workable architecture for the RGP chassis. U-shaped top or
bottom? How to attach panels, connectors? What to do about servicability? I
did pretty good on the TRUMPET chassis, with mostly flat pieces, one U cover.
It's hard to work with though, because of the blind screws holding the panels.
A lot of folks just bolt through the front panel. I don't like that visually.
So for me I use blind screws. Trouble is getting access to such. Once a board
is installed there ain't no room for fingers. Also trying to design such that
costs are ok in small quantities. I've always overestimated the sales of my
products, certainly don't need to overinvest in parts again. Chassis design
goes hand in hand with business design. Do I want an acrylic see-thru top?
Kinda neat. At least for demo units. The guts will be beautiful. Do I offer
your choice of color? Rack mount version? In small quantities, the machine
can be customizable. That could backfire in the long run. Hmmm.
I am considering a non-digital version. Even though it's way cool, I think
some would like to save money and buy just an analog-only RGP. So is that RGPa
and RGPd? Analog will be built with fewer parts, simpler panels. Upgrade from
analog would be expensive, but possible. Retail hovering around $7500 and
$9000. Will try to push that down.
RGPa analog-only build
July 18, 2006
Found some cool oval blue LEDs, adds some extra class to the faceplate. Now if
I can only find oval pushbuttons to match. Making the digital as generic as
possible, as it is a daughter card with possible future expansion. Therefore,
I do not want to limit myself on data rates. Looking at the CS8427 instead of
the more common CS8416. It's very similar but adds S/PDIF output (from ADC I2S
input). Thinking the ADC clock will be same as selected digital input. Added
an "int" option that allow you to record to S/PDIF output at any rate using an
internal reference clock - and feeds through to the DAC with internal loopback.
Had a great brain fart last night! Been having trouble trying to work out the
RGP1 power supply, what to use for transformers. In order to separate B+ from
heaters, I needed two trannys. Then a third low power one for the CPU. Plan
to make all of them choke input types, for super low noise and inherent passive
regulation. Hammond doesn't sell a good plate only tranny. And I want a 400V
or so B+, meaning I need 500-0-500 secondary. Options quickly became limited
and very expensive. Then the fart, to use two 117V trannys! I can ties the
secondaries together to form a center tap, that way, any imbalances aren't
really a problem. And I can hook for for either 115V or 230V operation. Two
birds with one stone. Now I can use something like the 260C, much lower cost.
And I can turn off B+ by disconnecting the center tap! As a result, I can use
the 5V and 6V windings for heater use. Three birds, as it eliminates one
tranny. This is fallout from using all balanced and complementary circuits.
Way cool. The cores do not have to be coupled. Four birds, as I can now mount
them in anti-phase such to null external magnetic fields. I sure hope this
works. Will cut cost and increase performance at same time. Requires three
trannys and nine chokes in this design. Should fit easily.
And taking this a step further, I found a simple way to make the 5V CPU output
to be a relatively constant current load, even with as LEDs turn on and off.
At time many can be on, others almost all off. Got this canceled out too, load
is essentially constant.
While digging around Hammond, I discovered they now make a 373DZ tranny. That
is a dual primary version of the tranny in the TRUMPET. Wow, wouldn't it be
great if I could finally make that machine in 240V version. Alas no, the core
is a little bigger and won't fit. Dangit.
So I thought maybe there was a chance I could shoehown this supply into the
preamp box, save on chassis costs. Holy crap, no way. Turns out the iron I
selected is huge. In fact, I can barely fit it onto a 12" x 16" board. It's
going to take a bit of jigsaw just to get it into its own box. Was hoping to
eliminate the unbilical, but it looks like it will stay. Going to be one heavy
Searching around for other chassis options, came across protocase.com. Looked
at them before. Downloadable chassis design software. Hmmm, might check into
this. The Par-Metal stuff I have is good, but nowhere near the quality needed
for an RGP. Then there is the iitmetalfab.com place. Hard for those of us who
generally make small quantities.
Finally, inspired by anothe website, came up with the idea for awarding a
"DIYer of the month". Can post a special page just to highlight the extra
special efforts from some of my customers. Many of these are already on the
forum, but I might as well show them off in concentration. Hmmm, I think I'll
give out prizes. Wonder what they should be. Maybe a one paragraph interview
to go with each photo.
July 17, 2006
Hey, maybe I'm not so whacked out. Found the Boulder preamps have a lot in
common to my proposed RGP. All balance I/O, buttons and one rotary knob, plus
built-in phono, cool remote. Very nice. I chose multi-LEDs over the digital
display. All XLR! I was worried about that, as BAT has switched to both XLR
and RCA, which I think causes electrical problems.
Also reading review of AR machines in HiFi+. Less performance than RGP for
about double the price. Good. These are the guys I have to compete against.
Wrote me a new business plan over the weekend. Yeah, even micro companies need
them. Good advice from one of my informal advisors. I had a fancy business
plan from 1999, 25 pages and all. Graphs, charts, numbers. Unfortunately, the
predictions were so far off from reality that it was a joke. Really, I had
thought I could sell hundreds of VacuTraces. Writing it was a good experience
as it makes you try and think of everything. What are your goals? What is
your product? Who is your customer? How will you market it? That sort of
thing. And lately, you'll read in this blog that I've been doing a lot to
focus and shift business strategy. Hence, the introduction of HAGLABS and the
upcoming RGP. Anyway, there is only one paragraph from that original business
plan that it still valid, the mission statement. Here is is:
"To create unique and exceptional products for the vacuum tube audio market.
Uniqueness and market differentiation are achieved by exploiting perceived
weaknesses and limitations as advantages. That is, invent new architectures
and techniques rather than copy traditional methods in the pursuit of lower
cost and higher performance."
Looking at various microcontrollers. The Atmel is nice. Was thinking of a
Rabbit, as it has the ethernet and all sorts of goodies. But in then end it
seems a straight PIC is a bargain. I don't really need computer I/O, nor a
complete operating system. And I love the Harvard RISC architecture of the
PIC. Their new ones can also be programmed in C, so I'm leaning towards the
PIC18F2220 or similar. It's a complete computer with FLASH memory, nonvolatile
storage, analog, I2C, timers, everything. I won't need to add anything. It'll
be able to control LEDs, relays, analog, and talks directly to a CS8416. An
$8 28 pin package has everything I need. KISS. Now if I can only figure out
how to rip cost out of the chassis.
July 15, 2006
RGP. Yup, that could be the name for PROJEXT X. Reference Grade Preamplifier.
Or maybe I should call it RGP1? Lots of stuff to think about yet. Like having
0dB gain loop-through channel for home theatre input. I can have such assigned
to any input by holding down the volume button for 1 second. That would then
set proper gain each time that input was selected. And then hold it there. It
could be deassigned in exactly the same fashion. Boy, it sure comes in handy
to have a microcontroller on board. So much functionality can be added. Also
need to architect the digital section. Wondering how I can share clocking with
AD and DA sections. And what about future upgrades? I the tube analog path
will be good for 20 years (maybe much more), but the digital will obsolete
sooner. That's why it is on daughter card. When TI ever comes out with a 24/
96 version of USB, I'll immediately put it into the product. Older machines
can be upgraded at minimal cost. For now, I can't do everything, so must set
some design limitations. Can't be stuck in R&D mode forever. So maybe I'll
have AD default to 16/44.1k USB mode when USB is active, otherwise go to 24/96k
via S/PDIF. And this won't be your garden variety 16-bits, either. But top of
the line chips (24 bit) with low jitter clocks. The best 16/44.1k performance
you can get. This isn't a recording studio machine, so I probably don't need
the flexibility of 192k or 88.2k AD. It's a two-channel machine for home use
for those who want the best. The DA is actually easier, so I'll be able to
lock onto any rate. Definately need more thought, I just want to make sure I
don't bite off too much and delay introduction by a year because of some silly
So what's all this nonsense about RoHS? That's the "Reduction of Hazardous
Substances" that went into effect on 7/1. Basically, the idea is to reduce
toxic metals from electronics (lead, cadmium, mercury). Only it's a farce.
If you look at the exempted products, two things become clear. First, car
batteries are exempt. What? I understand this makes sense because there are
no practical alternatives. However, ONE lousy battery has more lead than a
lifetime's worth of HAGTECH products. There are other similar exempted
products. China has exempted itself entirely, of course. Secondly, certain
categories are exempt, telecommunications, military, networks, medical,
transportation. Basically anything that requires reliability (this is where
hi-end audio belongs). You see, lead-free electronics has a guaranteed failure
mechanism in the form of tin whiskers. All electronics you buy now are
guaranteed to fail. Especially anything with high density surface mount
packages. Like a cell phone. Ok, I understand that many consumer products
today go obsolete anyway. Many are so cheap that they are just throw-aways.
When my VCR dies, which they do often, I just buy a new one. People get new
cell phones every couple of years, solely for new features. These are the
products that should be targeted by RoHS. But not high-end audio.
So it becomes obvious that completely lead-free products are so bad that
critical industries (medical, military, telecommunications) must be exempted.
And no teeth in the law, in regards to the majority producers (car battery).
The actual reduction of harm to the environment is minimal, almost
hypocritical. Big business loves it, though, as it guarantees more purchases.
Think about it. Do you want to invest in an heirloom quality tube linestage
(that previously lasted forever) now guaranteed to die on you in 10 or 20
years? Sort of eliminates the possibility of anything new from ever becoming
"vintage". Is that good? No. We build machines with near mil-spec quality.
We build them to last forever. Or at least 100 years. We should be on the
same exemption list as those other critical industries. That's the category
HAGTECH is putting itself in - "critical reliability". We will use as many
RoHS qualified components as possible, but not to the point of reduced service
lifetime. Our customers will not suffer from this legislation. Legal
petitioning is in the works for our own high-end exemption. Stay tuned.
July 14, 2006
Stayed up late last night pretending to use a PROJECT X preamp in my head. See
if it made sense under normal operating conditions. I have to admit, this is
not at all how I envisioned the result. That faceplate really took a turn on
me. But it's the result of driven design, parallel processing of requirements,
desires, and limitations. It's really wierd looking. On the other hand, by
placing such high demands on performance and usability, it could be a wonderful
solution. Just not what I expected.
You see, I'm a bit anal regarding analog controls. I hate push button volume.
Neither can I stand a numerical readout. Give me something analog! I want
pure tube analog circuits and an analog-like big knob for volume. Well, I got
it, I guess. Will have to mull this concept over for awhile. Now is not the
time to rush down a path. Must make sure it is acceptable to the audience.
As far as parallel processing goes, that's my mode of system design. Take
everything into account at the same time. That is, think about marketing,
manufacturability, cost, reliability, sonics, etc. for each piece of the
design. For example, coming up with the balanced LDR attenuators, I wasn't
just focused on performance, but also on cost and ease of programming. Sure,
there's a merketing aspect as well. Basically, this is the opposite of what
many companies do - chopping up a design and assigning small chunks to various
designers. Each doing their own thing. It leads to a loss of synergy and
coherence in the overall product. That's where a good system architect comes
in, to pull it back together.
In fact, I've already thought up a new ad campaign. The title reading "And Now
For Something Completely Different", borrowing from MP. "Control everything
whilst seated, even with your eyes shut".
Anyway, I've organized the back panel by function and ease of signal routing.
Decided to just go straight XLR for everything. At this price, there's no
point compromising with RCAs. Except for the phono inputs, where it will
remain standard. Remember, I'm aiming for the top of the hill. That place
held by the Steelhead. My price estimates are giving an MSRP of maybe $8500.
So for about a grand more, you get a comparable (theoretically better) phono
section, plus a remote control and digital box. Twice the machine for almost
the same price.
PROJEXT X chassis
Oh yeah, also tried to squeeze the box size down a little to standards. Can
then be ordered with a rack size front panel. Hey, maybe I'll even offer your
choice in color. Checked gain, 75dB phono at max settings, for a total of
90dB. That gives a 3V output from an 0.1mV cart. Volume in 1.5dB steps,
balance in 0.5dB steps. Loading in 3dB steps. All can be changed in software.
Decided to drop the ethernet port. Long way to go yet. Must do chassis
design, pick a CPU, locate buttons, remote control bottom, etc.
Am I insane giving away all my secrets?
Working with a customer of some residual TRUMPET hum. In the process of toying
with some ideas, I ended up discovering two things. First, the o-ring tube
dampers I've been supplying work really well. Been using some NOS GE 12AX7s in
the first stage on my personal GOLD TRUMPET. But holy cow, they're noisy. And
super microphonic. The stock Sovteks kick their butt in this regard. Swapped
with some NOS Sylvania and it's working better. Tried various damping stuff,
including the dip paint, HALOs. Well, my fingers grabbing on tight work the
best. Next in line is the dual o-rings. Then the HALO right on their tail.
The second discovery was an experiment in the STEPUP internal wiring to try and
break a small ground loop in the L-R channel coax shields. There was always
some residual magnetic pickup from the power transformer. Well, the new wiring
worked great. Noise floor dropped significantly to a new low. Unfortunately,
it's a pretty extensive modification and requires adding a new ground lug and
ground wire. However, it is an improvement. So anyone out there with a STEPUP
can send it in for a free upgrade. All of them are still under warranty. Even
if you weren't the original owner.
July 13, 2006
Made some big strides today in defining the architecture of PROJEXT X. Working
the phono inputs, I came up with some interesting configurations. One thing I
really wanted to do is to have a truly balanced phono input using XLR. And the
ONLY way that can be done is using a center-tapped tranny. So far, I don't
think anyone in the industry is doing this. But it is the right way to go for
maximum performance. Anyway, trying to get at least two phono inputs, I ended
up getting three. And the logic fell into place. It's not just 2 MC and one
MM, but LOW, MED, and HIGH. Found that I can special order from Sowter the
tranny I need, with center taps on both input and output. Turns out you can do
a lot with this, feeding a balanced tube stage. The HIGH input is same as a MM
having a 1:1 step-up ratio. The XLR gives 1:10 fully balanced and isolated, or
1:5 dropping the output balance. LOW configures for 1:10 or 1:20, for those
super low output carts. How handy! Three phono inputs, each configurable to a
range of cartridges. Then I have my variable loading section. Although it is
one LDR, it is programmable and each input can have it's own setting. And I am
pushing the range in both directions. Not just 47k and down, but also up to
100k. No capacitive loading, as that is nothing more than a marketing gimmick.
Switching and routing topology makes sense, and I don't nee a whole lot of
tubes to do it. Relays and LDRs. Moving the digital section into the analog
box was a great move, it really simplifies a lot of things, and greatly reduces
cost. On the other hand, it increases complexity of the control interface. So
much flexibility and capability in this machine. My old ideas on controls just
wasn't working out. Not intuitive and it didn't work on remote. Tried
something new, arranging controls the same as the signal path. Start at the
left side and do phono input select. Then tranny ratio, then loading. Then
analog input select (phono or 3 line levels). Next section is digital tape
loop followed by the linestage. Having useful features makes for a lot of
buttons. Like mute, mono, phase, etc. The user can record while listening,
monitoring either the analog original or the digital playback. Very much like
the RIPPER. The digital section can record to USB at 16/44.1k or to S/PDIF at
24/96k. Playback source determines clock rate. Anyway, I ended up using a
select button for each control, sort of like the CJ preamps. Then one big
rotary knob. I hate up/down volume controls. Way to slow for me. Just can't
stand it. So once you select volume (default), the big rotary knob takes over.
The big knob also does balance, record level, and will walk through the input
sources and loading. The front panel looks complicated, but it isn't. Uses
maybe $100 in blue LEDs.
PROJEXT X front panel concept
The remote works the same way. You can scroll through the loading resistances
but hitting the loading button. Or hit it once and then use the rotary knob
to go up/down. Yes, adjust loading on the fly from your seat. That's the way
I want this full function preamp to be. My original idea was for a joystick,
like a game console. Then maybe a scroll wheel like the iPod. But I think the
knob is more me. Don't think anyone else is doing it.
PROJEXT X remote control concept
Ok, so they're both pretty ugly. I need to work on that. This is just one
concept - the logical one. Keep in mind this is doing the work of a
phonostage, a DAC, an ADC, a linestage, and a whole bunch of cabling. It is
for vinyl lovers who want the best of the best yet still enter the 21st century
of AIFF and FLAC. Theoretically you can record phono to 24/96k files. And
play it back. Thinking I can make the record level setting a dual display.
Like a VU meter, continually monitoring the recording level. Dancing LEDs.
Can do some other visual stuff on power up/down.
And now to my real question. Was thinking I'd put an ethernet port on this
machine. Not sure how it would work, probably have its own IP address. You
could connect a laptop and control the machine via a browser interface. Sort
of a different type of remote control. Ethernet is galvanically isolated so no
ground loop problems. USB is not, but I make it so using isolation trannys.
My question to you is, would such an Ethernet port on a preamp make sense?
Email me your thoughts!
Nice to see that my feature set is a lot in common with the wish list from
Thorsten's survey. One guy even asked for the ADC. So PROJECT X is on the
right track. Three phono inputs, variable loading, remote control, phase, and
mono. Yeah, it's what they want. If it just weren't so darned funny looking.
July 12, 2006
The emachineshop guys rejected my design files. Guess I have to use a polished
finish. Must resubmit.
Working on the architecture of PROJEXT X, had some breakthroughs. Noodling
with the LDRs for volume control, I forgot everything had to be done on both
phases, as the circuitry is all balanced. Shoot, that's double the attenuators
needed. Fortunately, I was not a problem. The LDRs will work fine. And then
I realized I can do the same with cartridge loading! Duh. This simple and
accurate circuit eliminates a whole bunch of relays and contacts from the phono
input circuitry. And it provides nearly continuous choices in resistive
loading. Nobody else can do that. Cool. I'll have it on a remote control
rotary knob where you can sit in the chair and dial it in while playing. No
going back and forth. That's the beauty of PROJECT X. Remote control of
everything without loss of sound quality. This is not your father's preamp.
The dual LDRs also make great output mutes. Soft, no clicks, take less power.
They seem to be helping more than I expected. Not quite right for input
selection or output phase though. For that I think the relays will perform
better. Oh yeah, the one wierd thing about LDR volume control is that it does
not present a constant resistive load to the previous stage. Input impedance
to preamp will be 50k minimum, 500k maximum, depending on volume setting. I
really don't see this as a problem. In fact, it might even help.
But not quite everything fits. Have this very nice concept for +/-6V heater
supply, but the tubes I am looking at are 6V only. I like the dual supplies,
if I can get them balanced, because I can tap off of them for the DAC circuits.
That's part of system design, ripping out the cost. Repartitioning and making
topology and part changes to improve performance and cut cost at the same time.
One example is moving the digital section into the analog box. This eliminates
some very expensive external cables and buffering requirements. The tradeoff
was to put the power supply in the other box. That, ended up having advantages
too. So it is coming together. Had some sharing of tube sections going on,
but nixed it. KISS, but don't sacrifice anything important. Around and around
in circles I go. Hey, I even had this crazy idea of using the screen grid on a
pentode as input for a buffer stage. Nah. But I would like to use a 6CG7.
July 11, 2006
Cool, looks like Thorsten did some work for me. He asked what folks wanted in
a phonostage. Interestingly, it includes a LOT of the stuff I am putting in
HAGLABS machine. What I didn't think about was switchable EQ for 78s. Hmmm.
This can be done, but now we're getting a lot of relays in there. Yes, remote
control will do it all. I'm glad to see more and more people demanding the
polarity switch. Good info. Confirms I am on the right track.
Ouch. I just spent $1,900 on machined parts for the UFO. God I hope these
work out ok. 37 day lead time. Should arrive right after summer vacation.
Very exciting times indeed!
July 10, 2006
An old memory popped up the other day. One that may have been the origin of
HAGTECH. It was Harvard square subway station in 1986, in that little
connecting tunnel that goes around a corner before you head up the stairs.
Some guy was cranking away on electric guitar, and I noticed he was plugged
into an outlet. So the location was not the greatest in any other respect.
But that got me to thinking, why not make a battery powered guitar amp? I
knew I could do that. Not sure if tubes had entered my head yet. Soon I was
making a paper mache horn over a soldered wire frame. It was supposed to be
exponential. Sounded like doo doo. But the audio and business bug caught.
Somehow, I knew I would end up doing a real HAGTECH.
Did another quick web site statistic check. Page hits still creeping upwards,
even in the slower summer months. Averaging well over 1000 hits per day. Then
I noticed I get a lot of errors for a "robots.txt" file. What the heck it
that? Turns out it's like the favicon file. But it's a simple text file to
give instructions to webcrawlers (like google) in what to do. I think you can
use it to hide directories or files. Anyway, I made one that instructs bots
not to dig into my personal pages.
And finally the production panels for my GOLD TRUMPET arrived. Took some
photos to update site. Also have panels for a step-up tranny. Although I
don't list it as a product anymore, I can build these special order. It takes
a while because I have to order the parts from Bent Audio.
Thought of another movie to make. This time of a RIPPER. Basically, a quick
walk through of recording an LP to playing the song back on an iPod. Sort of
a 30-second commercial that rips bing bang boing through all of the steps. If
only I can get some good industry sites to link to it.
Also decided to go back into the advertising business. But not like before.
This time much more focused, with a message. I couldn't pass up this deal.
Turns out you can get a year of ads in the new TONE AUDIO magazine for about
the same price as a one-time one-page ad in stereophile. Ok, which will be
more effective? I was suprised to hear how large their circulation is already.
So, we're back thinking up ads. Only this time I have a full page to work
with. That could be a lot of words! Anyway, I'll push my USPs - unique
selling points, in marketing speak. I really had to spend some time to come up
with these (getting my MBA the hard way), with some mentoring help. And it
ain't price. It ain't kits. They are simple design competence and innovation.
My engineering background and experience with a huge array of products and
technologies gives me a real edge when it comes to designing new audio
hardware. I am not limited to copying others. I am not limited to just tubes
or just digital. It is this hard-earned competence that gives me a bit of an
edge over much of the competition. Secondly is innovation. Or inventiveness.
And that's where my right brain comes in. The brain farts. Resourcefulness to
come up with novel or imaginative solutions. Again, I don't have to copy. I
can do something that breaks new ground. The UFO is a good example of this.
So somehow I have to impress this into an ad. As the feedback from my
marketing question on the forum suggests, it is the designer and the company
that matter a lot. What would Marantz be without Saul, or Hafler without
David, or Klipsch without Paul? You get the idea. Look at Arnie Nudell or Bob
Carver? Customers tend to follow the designer, and that's an important lesson.
July 6, 2006
Ah, the world of audio. Here's what it's like to be me: I get an email today
from some guy on the other side of the world who bought a cut-rate chinese USB
DAC. He wants to modify it to add a transformer S/PDIF output just like the
HAGUSB - because he came across my schematic! But because he doesn't want to
pay extra for shipment from DigiKey, he wants ME to sell him a tranny. Like I
need to support piracy. Ok, maybe if he sends me a fake Rolex.
Been awhile since I put up a quote, this one from Forbes:
"The doubters, the faint-of-heart and the safety-obsessed are always noisier
than are the innovators and visionaries."
Thinking again about marketing and advertisements. I have a good opportunity
to place some major ads in a major publication. Full page. So what should I
do? What is really effective? Do I focus on one product at a time? Or sell
the brand? As I probably mentioned somewhere below, my "unique selling points"
are design competence and innovation. They are my sole advantages in the
marketplace. I can't compete on price. So I came up with an ad that promotes
the brand HAGTECH. I may not use it, but just in case, figured I need to back
up statements that claim inventiveness. So I added a section to the patents
page that outlines my more significant inventions over the past dozen years.
Proof there is some pudding. Also have a lot of "firsts" for the audio
industry, but haven't listed yet.
July 5, 2006
Oh my, I actually did some work yesterday. Got so excited with this
emachineshop program I had to figure it out. Quirky, because it draws from the
machinist's point of view. But once you catch on it is pretty easy. I was
able to input all 4 pieces pretty quickly, minus the strobe pattern. Cost will
be about $150 for the set in quantities of ten. I don't know if that is good
or bad, but it gives me a whole new capability. Am very pleased with the
results and quality from frontpanelexpress, so maybe this will be the same. I
have to try it out. This UFO is going to be so cool looking! Here's some
screen shots of the bottom piece in progress.
UFO top viewUFO bottom view
Took me several hours to type in the coordinates for all of the strobe marks
(392 of them!). But it looks great! Having a tone of trouble getting the text
function to work. Keep getting errors for milling operations. Very odd.
Would like to at least put "UFO" and my website on the bottom. Found a button
for selecting premium appearance (no flaws). That upped the price another $25.
Total weight came out to 1.17 pounds using all non-magnetic aluminum.
Batteries and electronics will add a touch more. So maybe 20 oz. Hoping most
of the hi-end table bearings can handle this.
UFO strobe pattern
July 3, 2006
Since it is sort of a half day of work, I decided to put some time into the
UFO. I wanted to make a cool one for RMAF show. Using the basic design I came
up with 2 years ago, I tweaked it a little to make it manufacturable and, more
importantly, that it could actually be put together. It's made up of 4 pieces
that get screwed together. Had a lot of trouble finding a way to get board
mounted and all pieces together, yet still change batteries. Had one solution,
but it required complete disassembly. So I bit the bullet and exposed some
flathead screws on the top. This way, only the top pops off to change batts.
Got the top two pieces to share same internal cutouts. Didn't want to do this
to bottom, insuring maximal weight. It wasn't easy fitting the existing UFO
board in there. The batteries chew up a lot of room. Still, I wanted to make
the cool UFO shape, such that when it spins it looks like a flying saucer,
rotating with engine exhaust. Meanwhile, it doubles as a very accurate speed
strobe. Concept is complete. It uses 3/4" x 3.5" aluminum blanks for the
saucer and bottom, stem is a 2" x 1" brass cylinder. Internal cuts are in red,
circuit board and standoffs in brown.
UFO machining design
Hopefully this works out and I can find a place to do the lathe & CNC machining
cheaply. My home CNC doesn't have the precision required. Wish I could find a
vendor like FrontPanelExpress. Their software is super easy and they do a great
job on flats. Can't seem to find a similar operation that does thicker
material. Retail price might be high because of costs. But hey, this is a one
of a kind piece.
Dug up raw metals from McMaster, looks like about $75 for three discs and one
pipe. I can live with that. It's the machining that will be killer. Maybe
I'll just have to try and make one on my own machine. But wait! I just found
that site I was looking for, emachineshop.com. What the heck, will try it out.
July 1, 2006
Got my living room installation back up and running. Going full minimalist. I
took out the phono (too fragile with wild kids and dogs) and have just a CHIME
and a pair of CYMBALs. I like this combination as the volume control at 100%
is right at the onset of clipping. So I can crank it up without worry. No
linestage is the way. Just an iMac as a front-end. I use it to play CD, DVD,
and downloads. Mostly via iTunes. Hardware-wise, its very solid. A slick
combination of machines that produces no glitches. Operation is seamless.
The big problem I did have was one of dirt. This location collected huge
amounts of dust and animal hair. My amplifiers were covered in a thick coating
of gook. It even put a film on the circuit board. Much of this is due to the
salt-laden air spray from an open window. I left it that way for two years,
using it as a torture zone to run reliability tests. Well, nothing broke, it
just kept running. Anyway, I pulled it out and cleaned everything up. Checked
the bias on the CYMBALS (one of them shifted a few mV). I cut the rack down in
size, dipped all of the tubes in paint. Minimal. Nothing but a volume control
and the Vandersteens (my HORNs are in the shop system). Everything else is
done on the iMac. Someday I'll get a new TV and integrate that.
Living room sound system
So now I am doing a life test on the paint-dipped tubes.
June 29, 2006
Ok, so I tried to implement a "no tubes" check box on the order form, but the
dang code got messy. And then, for $10 off per tube, it really didn't look
like much of a bargain. So maybe it's not worth it. Sure, I'll do it by
special request. I'm also thinking that maybe keeping them in is a good way
to prove the machine is not DOA, instead of somebody's unknown tube. Ok, so
not a good idea.
Some very clever viral marketing going on over at Asylum. These posts by
"drlowmu" are brilliant. He "reprints" emails from the mysterious and unknown
"Dennis". He spews a lot of gibberish and genius, hard to tell sometimes. I
sort of gave up trying to follow it, as he's dead wrong on a number of things.
However! It is probably very entertaining to others, who now think he must be
some guru. Not knowing if "Dennis" even exists adds to the excitment. This is
a classic and superbly implemented piece of marketing. My hats off to them.
Just had a late breakthrough. Whipped up a quick prototype of my new auto-
volume control. The one using a dual vactrol in a feedback loop. Ok, I built
half of it. Anyway, fired up and ran perfect! I used a pot to generate a
reference voltage, which the loop then tried to recreate using the variable
resistor in the vactrol. Guessed at some component values, and the circuit was
stable. Response wasn't bad too. Just a little bit of lag when slewing up to
maximum volume. I can't see that as a problem. Seems to be able to get about
40dB maximum attenuation. Wow, now I need to check out the sonics.
Theoretically it should be really good, as all of the classic super-fi
recordings you are listening to have already been through one of these. If
anything, it makes a perfect attenuator for the front-end of the ADC. Sure do
like the flexibility of this. I can achieve repeatable 0.1dB steps. All the
way up and down. Hmmm, if I don't end up using this for PROJECT X, it would
make a nice half-kit.
Auto volume proto
Have a great weekend everyone!
June 28, 2006
So I wake up today and say "smoke". She says "huh?". Yeah, "wood smoke". Far
off there must have been a fire, the wisp of wood smoke was distinct and sure.
And she missed it! What? It's been known that a woman's sense of smell is
many times greater than that of a man, especially during pregnancy. And we've
had countless examples to prove it. Maybe there is more to the story. Indeed
we are wired different. But perhaps that super olfactory of a woman is tuned
to odors of a type. Like doo doo and rotting food. Cleanliness issues. The
men honed for protection, hence the detection of smoke. This begs the question
of why audiophiles are 99% male. Is it we are wired for tinkering (playing
with hardware)? Or the music itself? A primitive repetitious and mesmerizing
drum beat preparing us for war?
Hmmm, still have three CHIMEs out there for reviews. PFO, ETM, UltraHiFi. Got
to keep track of these things. Difficult sometimes with so many pieces on the
road. Hate to lose expensive hardware. Hey look, first review came out!
Had a neat idea last night. I should sell my products with a "no tube" option.
Most customers roll them anyway, so perhaps there is no need to sell tubes that
just get tossed. Or unused. Yeah, I think I'll put that into the order form,
save you a few bucks.
June 27, 2006
Wow, this is the coolest thing! I just discovered lulu.com, a self-publishing
site. It is for authors to sell real books (just like amazon.com), only one at
a time. This is a great idea. I've read about it in a business magazine, but
going to the site was very interesting. Found it by clicking links from a
forum, ended up to this book:
I almost bought it just to check out the quality of publishing. But then
realized it's a mere 36 pages. Not exactly a book. More of a brochure. And
definately not worth the cover price. Anyway, such a neat site may end up
moving to write a book myself. A few folks have suggested such. It sure might
helps sales, too. Yeah, one of these days I might just get started. This link
Meanwhile, thinking again about the COMPRESSOR. The more I read of Mix and
Recording magazines, the better I see my chances of this being a competitive
product. Need to run some numbers again. Heck, I hate to completely cancel
such a project that I've spent so much time and money on. It's very close to
operational. I'll try to have this proto running for RMAF. Certainly could
use another money maker to help pay for PROJECT X R&D, which is definately
going to be costly. Not sure how much I can squeeze out of the FRYBABY.
Looking into some new tubes for PROJECT X. This is not just to be different,
but to try and push the performance envelope a little farther. With previous
products I purposely limited myself to the common and popular devices like the
12AX7, which also happened to be perfect for the job. Still, I wonder if I can
do better using a 6072. Less gain means more headroom. More linearity. For
other stages maybe the 6CG7. Output buffer a 5687. All of these have variants
in current production. As I've stated before, the only way I can compete
against piracy (copycats and IP thieves) and low-cost producers is to leap-frog
via innovation. That's my new game plan. Stay ahead of the curve.
Ok, just finished me a new VACUTRACE. This one is for the RMAF show. The new
paint job is nicer than the old one. May as well show it off. Getting
everything ready way early. Building new CHIME as well. Need to finish up
June 26, 2006
Trying out a new experiment here. My idea is to dip a tube into paint, just
the top part, in order to reduce microphonics. Theoretically this should work
like a tube damper. So I picked up some high temperature paint (not spray),
some rustoleum for painting BBQ grills. I'm thinking that should be able to
handle the small preamp tubes. So I dipped a couple of 12AU7. Normally this
should only be done up to the top mica, as heat is dissipated from the plate
via radiation. There should be no problems covering up the top getter area.
I hung the tubes upside down so they dripped a very clean result. Not very
thick, so maybe a couple of dippings is better. The first test went well,
absolutely no problems handling the temperature. Microphonics were definately
changed for the better, with a noticable reduction in "ping". Ok, maybe this
is nowhere as good as a HALO, but it looks really cool and gets you quite a
bit. Next I'll try some really smokin' hot 6H30s in my Cymbal. If my tests
work out well, I'll consider this as a permanent manufacturing change for all
products. On a similar note, I should get a rubber stamp with my logo on it
so I can re-mark stock tubes. That'll confirm they've been tested and graded
(matched) on a VacuTrace.
Realizing I have a lot of topology issues to tackle in PROJECT X. One of them
is the volume control. As everything will be computer controlled (via remote
or front panel), I can only use a regular pot by adding a motor to it. You can
buy some ALPs that do this. But this isn't very consistent nor repeatable. My
original plan was to do a stepped rotary switch driven by a stepper motor.
Such would have distinct positions and use classic fixed resistors. Other
folks do this, but my twist was to use the circuit board itself as the gold
contact pads for the switch. Then build the rotary wiper onto a bearing.
Basically, the switch and board were one and the same. The motor would then
mount to board and drive via gears or chain. This would have worked really
well, however my recent disovery of dust buildup on the insides of my home
CLARINET causes me to rethink such an open-frame design. Not happy about
possible crusty buildup over time. Another choice is to use lots of relays and
fixed resistors. I think C-J does this. It works great. Relays are gas
filled and sealed. You can use any type of resistor you want. It's a much
better solution than a multiplying DAC. I'll also look into the many digital
volume controls by the likes of MAXIM. But they worry me, using CMOS switches.
The specs look great, but it just doesn't seem right. Another optiona that is
really appealing to me now is the vactrol - a light dependent resistor. These
are used expensively in the recording chain (so your signal has already been
through one!) and appear to be sonically transparent. In fact, I use vactrols
in my COMPRESSOR, so I've become familiar with their peculiarities. Their main
problem is controlling them accurately. But that is simple to solve, as they
now sell them in duals. Use one side for volume, the other in a feedback loop
to maintain control. This allows the use of a DAC to very accurately set the
"gain". Steps can be made arbitrarily small for a very analog-like feel. The
potential problem is amplitude modulation by an unstable or noisy control loop.
Anyway, all of these will be carefully studied, taking into consideration
system cost, size, and performance. Some things just cannot be compromised.
Other issues include the number of circuit boards to use. Should I make one
giant one? Or split it into a few smaller ones. What about controls? I like
simplicity, but confusion is worse. Sometimes more knobs is actually easier.
It all has to be super-intuitive and make sense. Some knobs (like volume) will
be optical encoders and merely talk to the CPU. They can have a very analog
feel to them though, if the software controls the hardware in proper
proportion. Also, LED feedback to show actual setting will help. Volume
settings should be very repeatable. Not sure what to do about other controls,
like a button. Should I do something normal, or a touch sensitive type? Yeah,
it has to be cool. One thing I have learned is that slick sells. Then there
is the remote control. It will be total custom. But do I use an iPod-like
wheel or a joystick? Or several? I don't want it to look like a Playstation,
but there is some awesome functionality in a joystick. Regular buttons just
don't cut it anymore. Maybe I'll just use a rotary knob. Really, wouldn't it
be cool to remotely control volume with a classic rotary knob instead of
hitting a damn up button twenty times? Oh my, I haven't even gotten to the
ethernet stuff. I could add a port merely for control purposes (hey, a laptop
will be part of your sound system), or perhaps I can figure out a way to pump
audio through it. Hence, no need for USB. Another issue is how to get 24/96k
out of it. S/PDIF I/O isn't too hard. But does that mean I have to recommend
a particular sound card to make it work? This is dangerous territory.
Nevertheless, I'd like to have the ADC section do 24/192k or 24/88.1k and other
formats for recording. Such a high-end machine should not be limited to 16 bit
Like I said, lots of architectural issues to work out. Many more I haven't
June 23, 2006
Going back to the drawing board with PROJECT X. Most of it remains the same,
but I am re-thinking the partitioning. Starting to lean towards a different
two-box approach, the common separate power supply. This would put both analog
and digital sections in the other box. I'd rather not have to go to three, as
the price goes up. Maybe the digital can be a plug-in card upgrade. This
topology simplifies cabling and reduces connectors. Much more room in the box
without the huge friggin supply. Things I ponder, should I use an off-the-
shelf CPU card or do my own microcontroller? If I want the ethernet connection
the answer is COTS. Maybe a Rabiit 3000. I've used their stuff before. Works
quite well. Much easier to write code in C rather than assembly. The other
big question is do I do the whole machine with XLR and keep the fully balanced
in and out? Or go RCA? Or design it such that it can be made either way. My
personal choice is to go XLR, but I have to make what you want to buy, not what
I want to sell. Anyway, I don't think this will be that hard of a project.
It's all stuff I've done before. Combine the TRUMPET's circuits and power
supplies with HAGDAC and RIPPER digital stuff. Add in a custom remote control
and it's ready to go. Tweak it all with innovative improvements. I might even
be able to offer it in different colors.
Finished building my GOLD TRUMPET. I had tweaked the values of some of the
special PRP resistors to more convenient values. The EQ is just slightly off
above 10kHz. I'll dig into that, see if there is an easy fix. Running the
tubes at slightly less current. Plate voltage dropped a few volts too. I'll
play with sonics for awhile to see if any of this helps. Or makes a difference
at all. I still need to order new panels. Screwed up these prototypes a
little. How do I get a logo carved in there? Must have perfection before I
start making production units.
GOLD TRUMPET proto
June 22, 2006
More flattery. Check out the new Cayin headphone amp (stereophile). They seem
to like the TRUMPET's architecture, supply on bottom amp on top. Tubes on top
of board and electrolytics on bottom (they don't get heated). Looks all too
familiar. A lot of big amplifiers have been done this way since the
introduction of the TRUMPET, but this is the first small box that I am aware
Ok, so now the entire chassis is included in the RIPPER half-kit. Not just
panels. You also get a couple of special screws and nuts to complete the job.
The hard part is done for you - a super tight and clean board layout and a cut
and printed chassis. The rest is as easy as building a regular BUGLE. I went
even further. On this final layout I added an optional S/PDIF output, so
DIYers have the choice of drilling and installing an extra RCA or BNC jack.
This allows you do use a RIPPER as a HAGUSB. Record on the RIPPER, but
playback to an external DAC. It won't be in the assembled product, but I did
add the holes and traces on the board for you guys. For those who already have
a phonostage, you can also DIY it for line level input. That will also save
you a few bucks. Ordering production parts today. Lead time is 3 weeks, so
RIPPER will be available early July (I was aiming for the 4th!). I've added
the RIPPER to my order page, so you can pre-order. Those who do will get a
free bonus UFO half-kit.
Wondering if I need to get yet another turntable. A traveling model for shows.
Sure, I get to borrow a super hi-end Galibier for RMAF. But it looks like I
will want a separate one to connect to a RIPPER. I think the best demo to give
is to record an LP live. Then switch back and forth to the digital playback.
Ok, ok, ok, so the analog side sounds better. What the heck did you expect?
But it's really sort of eerie, as the digital playback comes with groove noise
and pops and tics (if you ain't cleaned the LP). It sounds quite convincing,
but issing some of the warmth and organics of pure analog. Anyway, this means
I need a TT that can travel well and not get busted up like my MMF-5 did.
Perhaps I just need to keep the original packaging. Or maybe I can borrow yet
another table from someone at the show. That would make more sense. At RMAF
I'll be spinning two tables and a laptop (for CDs).
June 21, 2006
Aw damn. Just got a quote back for the RIPPER chassis end panels. The
machining and printing adds up, price came out way higher than I expected.
Maybe they misquoted? Anyway, I realize my part cost ends up totalling more
than the ART retails! And that's not even getting the parts shipped here. It
is quite the disadvantage to build machines in small quantities. Aha, seems
the quote was for complete chassis. I may have to include these in 1/2-kits.
Turns out this is the only way I can get all black chassis. Otherwise minimum
order is 1000 pieces.
Sheesh, now paging through the Acoustic Sounds catalog I stumble across the KAB
Precon - an inverse RIAA filter just like the one I make. Guess they liked
the idea. Hoping I can get these guys to carry the FRYBABY. They're big into
vinyl and would be a great outlet for new products.
Just completed my RIPPER testing. This is cool. I was able to run a RIPPER
and HAGUSB at the same time. Can select one for recording, the other for
playback. Or just let RIPPER do both. No glitching. And now I admit having
the select switch on the front is very convenient. I can switch between analog
and digital on the fly during recording. What a great way to A/B! Just want
to add one more LED to indicate when USB is operational.
Ok, this is more like it. The USB Matchbox / Plus from Henry Engineering is
more like it. Basically a RIPPER but without the phonostage (just line input)
for $695. Simpler model without headphone is $495. Now that is comparable
Since I am ordering production pieces, I will start taking orders for THE
RIPPER now. Anyone putting in a pre-order will get a free bonus UFO!
June 20, 2006
Oh, did I mention that if you buy an Audio Note Dac kit 2.1 that the USB input
option is supplied by a HAGUSB? It looks like a very nice machine, but we have
very different approaches in terms of digital playback. Someone ought to buy
both this and a CHIME and do a shootout. Anyone out there with bucks to spare?
AN DAC 2.1
And so rummaging around various online forums I came across some guy who
replaced his TRUMPET with an Artemis Labs. This was, of course, raving about
the TRUMPET for a year. He thinks the Artemis sounds better, more dynamic.
Hmmm, dynamics is one of my strong points. What's going on here? Also, at
ETF05 the Artemis fared quite well in a shootout. Ok, I know John, he's a very
competent designer. He even did the VacuTrace review for Glass Audio. So
maybe it's time to dig a little and see what's up. What can I learn here?
The specs aren't anything remarkable. The TRUMPET has wider bandwidth, perhaps
too much. TRUMPET distortion is about 5x better, and TRUMPET noise floor is
12dB better. Is maybe the TRUMPET too clean? My power supply is far more
elaborate with tons of iron and chokes. My circuits are differential. So what
could it be? I notice a battery bias on the first stage. Very clever. And
then a plate choke output circuit instead of my cathode follower. His circuit
looks very clean and well thought out. I don't think putting all of the EQ
in one stage is a good idea, at least in terms of noise. But it does help
improve the overload characteristics of the second stage. So why does this
thing sound better? Is it the battery bias, the better overload margin, or the
choke loaded output stage? Or maybe it's because he winds his own iron? And
does it really sound better? I don't know, but this is a useful exercise.
Must always think, always find ways to improve. Maybe the TRUMPET needs to be
sweetened up a bit. It definately lacks "tubiness".
Anyway, I'll use this info to design the next phono stage. I should call it
project "X" or something. It started as the HAGLABS "analog" and "digital",
then the "two series". I may end up using these names in the end, but for now
I will refer to it as PROJEXT X. It is back on the front burner! The RIPPER
is essentially done. Just ordering production pieces and then to write a
manual. This gets me all caught up on R&D except for the COMPRESSOR, which is
on indefinate hold. I will get the one prototype working for the RMAF show.
Nothing like demonstrating flexibility. Oh yeah, RIPPER page is coming along.
RIPPER web page
June 18, 2006
Went out this morning and did the Tinman (upcoming triathlon) course. Windy
and choppy. I was slow, but still ended up doing a 48 on the run. Which gave
me some time to sort out things in my head. Two and a half hours of self-
propelled motion, you can't just think about form and technique the whole time.
Anyway, I realized the obvious. And this also sort of came out in a recent
discussion with a major hi-end distributor; that you can't sell a DIY version
and a retail version at the same time. Unless, of course, it is factory
direct. The pricing and margins just don't work out. In other words, no
retailer will push a product that can also be built by DIYers. Ok, a big
collective duh. Like I said, sometimes I am slow. The obvious isn't when
you're in the middle.
And this impacts the RIPPER. I was seriously considering offering it as a
candy bag item. But then, this has also been a much anticipated half-kit from
the start. I just can't bail on the big HAGTECH DIY customer base. And so,
the assembled price ends up at a mere $399. That's a lot of value for the
money, folks. You'd spend that much retail on just a hi-end phono stage,
competitive with the BUGLE. And here I offer not only a tweaked BUGLE, but
also a nice chassis and AC power supply. Ok, now add in the high performance
regulators. Now add in the headphone output. And finally add in the price of
a USB ADC/DAC. All in one compact feature-laden package.
Which brings me to my next story. Last night I popped into B&N to pick up a
new stereophile magazine (my subscription ran out and they never informed me!).
I also grabbed a Sound on Sound and an Electronics Musician. Yeah, I'm still
mulling over what to so with this fantastic COMPRESSOR I designed. Anyway,
paging through the SoS I come across a RIPPER! Only it's really an ART. There
it is in front of me. A review of a small box that does phono to USB. It has
the same gain knob, a clipping LED (like mine used to), a select switch, and a
headphone output. These folks are clever. Ok, someone beat me to it. Then I
realize Rolex doesn't give a rip how many watches Timex sells. Duh. The ART
is a box with the same functionality, but it's likely got your basic shit one
opamp phono section. It is powered from the USB bus (or wall wart). And I'm
sure the component quality ain't exactly top rate. So basically it is like
putting a Li'l Rat up against a BUGLE. We're talking smoke city. Hey,
sometimes you get what you pay for. If you want top-notch sonics and build
quality, you gotta get the RIPPER.
June 17, 2006
RIPPER boards came in. Built one up today. Looks great. Got the mechanical
fit perfect. Just need to fire it up. Then test the daylights out of it,
check every voltage and node for proper operating cinditions and stability. So
far, total gain is 58dB with an SNR of 72dB (ref 5mV input). Basically, that
means its loud enough to be a linestage and quiet as a mouse. Once I confirm
operation of the select switch, production pieces can be ordered.
THE RIPPER circuit board
Got website updated. It may look the same, but is now organized around product
lines. Menu on left is shorter. Got rid of OBOE and RACK. Maybe a few other
things too. Funny, right after I remove it, someone tries to order an OBOE.
Ok fine. I still have parts in stock. Just that it's not a very good seller.
I am focusing on my core strengths and not promoting the unpopular items. In
the end the customer ends up making the call. Even if indirectly. I have to
make what you want to buy, not what I want to sell. By the way, now is the
time to invest in a wood Trumpet ($1999), before they're gone. It is destined
to become one of those classic machines to forever retain its value. Like a
TD124 or MC225 or ST70.
June 15, 2006
Ok, getting drastic here. Looking at my website and product line. It is a
confusing mess. What have I done? And where am I going? I keep looking at
the new HAGLABS ONE series and wonder if it's a dead end. Yeah, technically it
is an easy way to introduce a new product line. But is it worth it? Really,
it's just a repackaging and tweak on the CORNET2 and CLARINET. I'm thinking
this will not end up being a shortcut, but perhaps a time and money drain. The
actual motivation was to generate additional revenue to pay for the TWO series.
But the numbers aren't really adding up. Been going through markups and
margins and costs, business stuff. And the numbers only work if I reach a
certain break-even volume. Well, that's always the case. More to the point,
I look around at the competition. New vendors are popping up daily. And some
good stuff too. I need to compete. So I ask myself, why do I compete building
the same machines these other guys can? The ONE series is really nice, low
cost with great performance and value, but still middle of the market. It's
not unique. Why waste my time with these when I can run rings around most of
the designers out there? Seriously, what are my unique selling points? One is
innovation. Two is that I can use my vast design experience from non-audio
industries to leapfrog the other guys. Why should I design a phonostage that
competes head-to-head against a whole bunch of other machines? I should not.
I should make one that smokes their ass. The Trumpet is such a beast. But it
needs to be taken to the next level. A level of innovation and uniqueness that
your average professional audio designer cannot touch. I'm going to stop
beating dead horses. The ONE series is kaput.
Ok, so back to the website. It's a mess. I'm taking an interim step to
organize products. And regarding the above, soon we will have two new
products. The RIPPER and the GOLD TRUMPET. Yes, my very own UBER TRUMPET that
I am building for myself will become the standard product. Upgraded parts and
those seriously good looking gold panels. Design is complete, so I will just
make it the new standard product. Have 2.5 wood ones left in stock. Really, I
am so tired of woodworking. Those wooden panels took up an entire day by
themselves. That relegates the classic wooden Trumpet to vintage cult status.
I don't think they will ever loose their value. Talk about an investment!
Anyway, maybe I can get the website changes done tomorrow. Newly organized.
Some pieces will be missing. I am discontinuing the OBOE. Maybe I will give
some away. HAGDAC will only be available as part of CHIME. And I am repricing
the HAGUSB back down to $129. Decided it wasn't going to sell via dealers,
will stick to factory direct mode. HAGCLOCK will also change to $129. Anyone
who paid the higher price in the past few months will get a refund. By
focusing on selling my assembled professional audiophile machines (GOLD
TRUMPET, BLUE CORNET, RIPPER, FRYBABY, and HAGUSB) I think I can pay for new
R&D development. The DIY stuff will all remain.
HAGTECH really needed to make this focus. We're going to forge our future on
our strengths, and not bridge out as much. In fact, it may be a long time
until I design another new power amp. There is only so much time and money
available. It's about optimization. Unique selling points. Therefore, I can
finally push ahead again on my HAGLABS TWO series. The ANALOG and the DIGITAL.
These are both groundbreaking products with features and technology nobody else
has. I may as well be a phonostage guru. Focus. Play to our strengths.
Oh, and don't think I'm making this all up as I go along. I'm getting
mentoring and advice in business and marketing by one of the best. That's one
very fine aspect of having lots of customers, you'd be amazed by the people you
get to meet!
June 14, 2006
Well, the enjoy the music contest is over. All the prizes and equipment have
been awarded. Congrtulations. Except the guy (whom I shall not name) who won
the free Cornet2 has not claimed his prize. I tried sending email, but no
response. It's the only contact information I was given. Take your time.
Moved my horns into the shop. About time I got some music in there. Wasn't
any good place to put them, though. At least not sonically. Then it hit me,
I can place them right on my assembly bench. The Fostex single driver units
work just fine as near field monitors. I'm curious to see how this works out.
I can also stand back and have a semi-convincing setup. And for once, they'll
have a rear wall. Building up my RMAF rack already, did some capacitor
upgrades to my CYMBALS. Will make sure everything is burned in good. Yeah,
I'm thinking way ahead. But I think planning is crucial. Made too many
mistakes in the past by last minute decisions. Much better if I can ponder for
awhile. Same goes for HAGLABS and any new products. Like putting the
COMPRESSOR on hold. Sometimes you just have to run the numbers and put things
into perspective and priority. Anyway, I'll soon have full racks running in
both my house and shop. Can finally start playing some music. Plus, it will
be handy when I use the RIPPER to make a bunch of MP3s for the kids.
The gold step-up tranny is done. Doing a long burn on it. Also just received
the power transformers for my GOLD TRUMPET. Should have that done in a day or
two. Then, unfortunately I have to buy some new gold panels. Two minor errors
make them imperfect. This will be the RMAF trumpet.
June 7, 2006
Completed the relayout of the RIPPER. Mostly this was just removing the auto-
select circuit and relays, replacing them with a slide switch. Although not as
cool, it is actually a little more flexible. Tweaked on the chassis fit too,
so everything is perfect. I'll build up this new proto to confirm it is
production ready with no bugs or faults. Measurements looks perfect. Even the
USB signals came out really good, even though I have short lengths of unmatched
impedance lines. Fortunately they are way shorter than the wavelengths
Got the gold panels for my UBER TRUMPET in. Dang! I screwed these up too.
Put some of the blind screw holes on the wrong side (they weren't supposed to
be visible). Then I realized the XLR connectors were made for the 3/8" thick
wood panels. The 4mm aluminum will leave a gap. Shoot. I should have gone
with RCA only. Shux. I'll probably end up buying new panels. There goes
another $200. That's one of the problems with perfection. It seems you have
to go through quite a few revisions and prototypes until you get everything
right. Then again, maybe I'll make this into my standard product. I've run
out of step-up transformer stock. Maybe I should drop it from the product
line? Just sell super-Trumpets? Or perhaps I can buy the trannys from Bent
Did some capacitor upgrades on the CYMBAL amps I'm going to use for RMAF.
Checked the bias drift. These amps are about a year old and were used for the
Kennedy review. Wondering how well the tubes held up. One channel had zero
drift, the other about 5mV, which is still very small. I should check my home
units too. Always curious to see how tubes age in my equipment.
Just got in the fancy resistors for the UBER TRUMPET. And they're red, too.
These are the PRP9372 audio type. I've used specialty resistors from this
company before with superb results. Excellent quality. And I think a few
other top tier manufacturers have started using them too. Only thing better
would be tantalums.
June 5, 2006
Review of Trumpet coming up in next issue of Tone Audio magazine. I really
like the visual style of that magazine. Really good photography. Sure wish I
had some studio equipment to make some decent product shots. Basically all I
have now is a piece of white cardboard used for the background.
Started some auctions at audiogon. Selling that CLASSIC CORNET with the step-
ups real cheap. Also a CLARINET and a HAGUSB.
Finally got some photos of the race a month ago. It clearly shows I need a new
bike. This frame is too big and there is no way to get into a true aero
position. I'm like an air scoop. No wonder I keep getting my ass kicked. All
I need is to drop 1 minute per 10k.
June 4, 2006
Well, looking at the Vidar photos, it has a crystal, which means they're using
a microprocessor (and probably a DAC) to generate the burn-in waveform. They
chose not to burn power cords. Hmmm, I woulda thought that to be a lucrative
June 3, 2006
Well the copycats are here. The new Nordost Vidar sounds an awful lot like a
FryKleaner Pro, but with more amplifiers and connectors. Except you can't buy
one. Funny, the description of the signal, circuit, and methodology fit the
Ok, so I wasn't the first. As far as I know Alan Kafton was. But at least I
did my circuit differently, and based on different principles. I'm really
curious to dig into a Vidar to see exactly what they did. Imitation is
flattery. Most importantly, this is more evidence that FryBaby can be a hit.
It's more flexible and a cheaper investment. Too bad I can't afford full page
magazine ads like the one that caught my eye in Hi-Fi+.
May 31, 2006
So it hit me. The Lundahl step-ups I have in the Classic had a broken
secondary coil. Ok, so re-wire it and only use one! Discovered one of the
primaries was broken too. That's ok, there are four of them. So I re-did the
assembly using two primaries in parallel and just one secondary. Now both
channels are working and matched. Sure enough, waveforms look just fine. Then
upon listenting tests (not perfect, as I don't have MC cart on hand) I discover
it works even better this way. It seems more open with less treble loss.
Maybe it is just less loading on the cart. I dunno, but the sound was nowhere
near as dark as I was used to (loading my Grado MM with 500 ohms). Ok, I'm a
burnin this baby in. Shall do some more listening, maybe at Stu's. This saves
me at least a grand.
May 28, 2006
Ran a perfect race yesterday. This was third triathlon this season (one more
to go before vacation, then one around Halloween). Also one of my favorites,
probably because it is short and fast. Well, the swim course must have been
set wrong, because we smoked it. My age group had quite a battle going the
entire race. I was 2nd out of the water but had a fast T1 transition (where
you change from swim to bike) that put me into 1st about one second ahead of
Seaver. Nice flat bike segment, no wind. Just before the turnaround, Griffith
caught me and flew by. I was still in the top 20 overall as we got back to T2.
Griffith by now had 1:30 on me. I knew I had a better run, so just kicked it
up to see if I could reel him in. What I didn't know was that 15 seconds
behind me were Seaver and Dustow, coming out of T2 together. So I'm looking
ahead trying to spot Griffith, when Dustow comes smoking by. Ok, I don't feel
too bad, as he's a former olympian. Then comes Rachel, the top female
finisher (who had started 5 minutes behind!). They catch my target quick, so
I pick it up. Yes!, right after the turnaround I pass Griffith. I figure I'm
in 1st now and holding a really good pace. But duh, didn't Dustow just pass
me? Half a minute behind me Seaver catches Griffith and we all finish in a
tight spread. So it was my best run yet, laying down a 6:18 per mile pace. I
took 2nd place (age group) and went under an hour, achieving both of my goals.
And my wife took 3rd in her age group. How cool is that?
Mokuleia Triathlon Results
Oh yeah, audio. Good active discussions going on at diyhifi.org.
May 25, 2006
Did a burn-in on the RIPPER, took the edge off. Very listenable now. I'd say
an improvement over the standard BUGLE, even though they share much in common.
Suppose I ought to take some measurements. The digital playback loses some of
the openness, a bit more stale and not quite as involving. But decent for
regular 16-bit conversion. And most likely the combo will smoke anything else
out there except for professional studio solutions. Just to prove it is
working, I made an MP3 of Eva Cassidy for you. This is a one minute clip, long
lead-in groove. I might have had the level set a bit too low.
Eva on MP3 audio (via RIPPER)
May 23, 2006
THE RIPPER came together very fast. Debugging was easy, only mistake was
wiring one of the relays backwards. Even the PCM2906 came up right away. Plug
it into a PC and it is automatically recognized and configured. It plays back
music just like a HAGUSB. The new feature is that it also records. Had to
assign proper I/O in the preferences of "audacity", but then I could record
from vinyl. Made my first track already. Can convert to MP3 even. Very nice
and compact box. My biggest issue is that I misunderstood the suspend line on
the USB chip. I used it to control the relay selection of input. Thought that
when USB wasn't recording or playing back it would turn off. Nope. Seems the
regular PC noises (clicks, chimes), keep it alive. So shucks, that feature has
to go. Must find a way to put a switch on front panel. This, of course, will
require a new test layout. Headphones and volume control worked fine. Noise
floor was way down. Good news is that now I have a box that allows me to start
ripping some of those old records to disc.
THE RIPPER in action
May 22, 2006
Thinking about RMAF and how to set up the room. Prefer to play mostly vinyl,
but will also have CHIME connected to my laptop. Seems reviewers always walk
around with discs in a bag. I sold the Trumpet at last show, need a new one.
So I started making an UBER-TRUMPET. Made new panels to replace the wood using
frontpanelexpress. Same for a step-up chassis. The COMPRESSOR looks so nice,
I think it will be stunning. Same gold anodize with engraved black lettering.
Nice curved corners. Great software, I did all four panels in one morning.
Also ordered upgraded PRP resistors to use in the signal path. This will be my
personal unit. Maybe I can offer it for sale. The $3000 TRUMPET-SIGNATURE or
something. Best MM phono stage available.
RIPPER boards just came in. Off to the shop to start soldering!
May 17, 2006
Got another monkey off my back. This time a consulting commitment. Finally
finished writing the white papers describing my latest inventions in enough
detail to start patent applications. Actually, they are just improvements to
my novel mass spectrometer design. One is for an unusual construction method,
the other for a simpler way to drive it. Both offer huge reductions in cost
and complexity. I'd love to be able to parlay this into a windfall. Not
exactly my bag of tea, you know. Prefer to just be bought out. Anyway, I'll
post the PDFs on my website, but only after patents issue. Can't let this cat
out of the bag until then.
I seem to be caught up again. Got the above stuff out of the way, RIPPER is in
fab, lots of FRYBABYs and HAGUSBs and other candy bag products built and put
into stock. CLASSIC CORNETs going out the door. Only item that needs
attention is the COMPRESSOR, which I'm purposely letting sit for awhile. Found
an old pair of the Lundahl LL2906 moving coil step up trannys I had. Just
epoxied them into a CLASSIC. I might keep this unit for shows. Will post
photos on how to do. Nothing like a 64dB tube phono stage!
Now to start focussing on sales, signing up dealers, etc.
May 15, 2006
Got RIPPER done. Sent out for fab. Should have some blank boards here next
week for build and debug. Glad to have this off my shoulders. Sure took a lot
of work. If all goes well, this machine will be in production by the 4th of
May 14, 2006
First major race of year. This was an ITU (Olympic distance) qualifying event
where the top dude in each age group gets an invite to the world championships
in Switzerland. I thought for sure my training buddy Tore would get it, as my
calculations had him 5 minutes ahead of everyone else. Well, I was wrong. He
was 10 minutes ahead of everyone, except for some new ex-pro who moved to town
and passed him with a mile or two to go. Me? I discovered to be woefully
undertrained. It started poorly, as I forgot to get my timing chip. Standing
at the waters edge, I suddenly realized I would not be timed, nor official!
Crap. So I just took my own time. The goal was to break 2 1/2 hours for this
distance. Ok, so I did that with a 2:27:01. But so what, it turns out to have
been a fast course and I slacked off on the run because I was 5 minutes ahead
of pace coming out of T2.
That's two mistakes. One, forgetting the timing chip. Two, trying to hit pace
rather than just going all out. The combo sapped my motivation and results. I
let my rivals get away and beat me by a few minutes. Time I should have made
up on the run. Ok, that's the last time I look at my watch during a race. So
actually, I'm glad not to be posted in the results, as it really was a moderate
performance. Next race in 2 weeks, maybe I can redeem myself.
Honolulu Triathlon & shaka
May 11, 2006
So the VCR/DVD died the other day. I think it's from all the salt-spray in the
air here, really eats up electronics (also makes for a great test location for
my stuff). But my little one wanted to watch Spiderman. Wait a minute, this
iMac machine I have for iTunes & CDs will also play DVDs, right? In it went,
and Spiderman started playing. Ok, so the screen was pretty small, but the kid
didn't seem to care. Hmmm, there's a USB cable connected between the iMac and
a CHIME. What if? I have to be honest, this is my first foray into home
theatre. Been focusing everything on regular stereo hi-fi. Fired up the rack
and was really pleased with the sound. Coming through the HORNs it wasn't all
glitzy and exaggerated. It was quite natural. And very easy to hear dialog.
Ok, so this is what all the fuzz is about. Perhaps it was the lack of a sub-
woofer that made it seem so right.
Got the RIPPER layout placement done. Power routed. Had a hard time squeezing
in decoupling. Very, very tight. Talk about short signal paths! Going to be
tough routing the phono outputs to the gain pot and back. Otherwise, should
be done in another day or two.
May 10, 2006
Back on the RIPPER. A few days away from this and I'm able to give it a whole
new fresh look. Found a topology that would work very nicely. Just wasn't
obvious before. Will have to make more changes, simplifying again. Doesn't
seem to be room for extra regulator for USB section, so will just rely on the
existing onboard regulation. Tweak it good. Also, the clipping LED wasn't
necessary, as any recording software will do this for you. Superfluous. All
unnecessary stuff ripped out. Not even sure there is spare room for an
optional S/PDIF driver. Layout going to be very tight as is. The phono
section came out super compact. Power routing is nice.
You can also see front and rear panels. As simple as you can get. In, out,
USB, and AC. One knob on front controls analog phono gain. No power switch,
as at 2 watts drain, you can leave it on like a wall-wart. Selection between
analog and digital is automatic. Use as a very good standalone phono stage,
finally a BUGLE in a box. Need to work out costing.
Holy moly. Part cost really adds up! Looks like this machine will retail for
$599. That's a bit higher than I had hoped for. Hmmm. Premium parts
selection, finished case, slick operation. I suppose as a phonostage it would
stand up to a Black Cube or similar. Same price. But you don't have to deal
with an external power supply. Everything built in. Added value is the
headphone output, and of course, the USB record/playback connectivity. I guess
that's resonable. Probably $200 or so total cost for the DIY version.
May 8, 2006
Got the CHIME from tour back. Opened it up and discovered why the USB input
wasn't working. How embarrassing. The damn chip was soldered in backwards.
Seems I musta forgot to check out that little detail prior to shipment. Now it
is fixed and sounding wonderful. No scratches, you guys took very good care of
it, thanks. Looks new. Will sell at discount regardless.
Found some new knobs for the series ONE. Put them on COMPRESSOR, they look
great there too. Aluminum, black silk anodize. Nice touch. Will show off at
audio meeting this Saturday, get some feedback.
May 3, 2006
Changed the look a little of website by merely playing with font sizes in the
style sheet file. I never liked the big 12 point text, but it was recommended
by usability experts. Seems to me most websites are now more efficient in
cramming info onto the screen real estate. So I'll try this out. Takes about
2 minutes to undo.
FRYBABY price went up to $239, you say? Well, ok. I mailed some out yesterday
only to discover the beast is $6 to post. Heck, that eats way into my margin.
Next thing you know, the streamlined sales tax initiative will take place and
all Internet sales will require collection of local taxes. Shees, I do not
look forward to that. One, it will be a commerce killer. Two, with each
county or city having different rates, it adds up to about 4000 different tax
calculations. An absolute nightmare for us small business to contend with.
May 2, 2006
46. Decided to pull the PICCOLO from the back burner. This is my JFET headamp
project. Never quite figured out how best to do the power supplies for it. Or
the box, for that matter. Thinking this might make a nice candy bag product.
My original thought was to make it easily integrate with existing Cornets by
using the 6.3V dc heater line tap as a supply. Anyway, needs thought. I don't
really want to do the wall-wart thing. Batteries have problems too, but they
end up saving a lot of bucks.
Sent out some FRYBABY samples to select dealers. Curious what sort of response
Very interesting article in the latest Popular Science. It was about using big
computers (parallel?) to do creative thinking using a genetic algorithm. That
is, give the program some design criteria and formula (or simulator) and then
run optimizations. All based on Mendelian genetics and Darwinian evolution.
Way cool and extremely clever. Good results are paired up and passed on, poor
results are terminated. Like trying to breed a super Terrier. Well, after a
hundred generations, odd things develop. Sometimes the program delivered truly
innovative results. For example, they did a complicated eyepiece lens design
and an antenna. Both came up with extraordinary performance results. Hey,
maybe this can be applied to analog circutiry or audio?
April 29, 2006
Got the panels in for the COMPRESSOR. Wow! These are so nice I can't believe
it. Perfect milling, lettering, anodize, brush, everything. Bought these
from frontpanelexpress. Pricey, but the quality it stupendous. I coud use
this for production. Already looks better than most gear out there. I'll need
to make some adjustments to the VU meter window size (was supposed to be an
ellipse). Back panel just as nice.
A closer view gives you some detail of the quality.
April 27, 2006
Introduced the FRYBABY today. At least to the website. You may now order
them. I keep changing prices. Working more details into my parts lists and
realized that I can drop FRYBABY MSRP down to $229. The HAGUSB and HAGCLOCK
both went up, they are pricey to build. Oddly, the BUGLE went down! Sheesh,
to be consistent with my formulas (markups, part cost, time, shipping, value
added, etc.) I had to drop the BUGLE by $10. However, I am turning the
assembled version into a 40dB MM only. That way I can build up stock of just
one type. Anyway, updated website today with new pricing, new manuals, and
I've built up a stock of each of these. Sending out samples of FRYBABY to
various reviewers. Must concentrate a little time contacting dealers and
setting up sales. Business is only half done designing and manufacturing. Got
to sell the stuff too! Oh man, and I'm a little behind on HAGDACs.
Putting a CORNET Classic up on AudiogoN tonight. Here's some new photos
showing the ac voltage select board inside, and the cleaner wiring.
CORNET Classic gutsCORNET Classic
April 26, 2006
Our Kolea left yesterday. That's the Hawaiian name for the Golden Plover, a
way cool bird. This one lived across the street in the empty lot. Everyday in
winter you'd see him foraging. Hung out with a pair of Mynahs. No other
Plovers around. Well, he left yesterday. I didn't see it. Perhaps he is
gathering with the others at Waimea before the big flight. You see, this tiny
little bird, about the size of a large Robin, migrates all the way to Alaska
each summer. In the Fall, they find there way back to the same damn spot. I
can guarantee I will see this dude again in a few months. Beautiful plumage,
just before they leave. Never got a good photo. How the heck can they do
Got the first FRYBABYs done. Ready to ship off to dealers. Boy, do I ever
have a lot of website work to do. Decided to go with just the translucent gray
color. It looks nice and professional. Also, realized there is no point to
including the 75 ohm cable with a HAGUSB. Anyone who buys one likely already
has an external DAC. Duh. So what to do with all of these cables I bought?
I'll just send them off with CHIMEs. Makes more sense. Anyway, I was able to
get the new HAGUSB price down to $179. Similarly, pushing the BUGLE price back
down, $169. Shipping it without batteries. As these are 'transitional'
products, between the factory-direct HAGTECH business model and the upcoming
HAGLABS dealer-only model, I'm squeezing the pricing structure a bit.
FRYBABY grey color
Finished THE RIPPER schematics. Started layout. Discovering maybe this is too
tight a fit. I'd prefer a better chassis, but the DigiKey catalog is extremely
limited to choices. I know I can put all the parts on the board, just want to
keep transformer away from phono input, etc. Good progress. Might be able to
order prototypes next week.
April 23, 2006
First CORNET Classic up and running. Fit and finish even better than the
original. The added AC voltage select board simplifies wiring. Looks nice
Thought I had FRYBABY into full production, then realized someone forgot to
order the 47uF 25V caps! Darnit.
FRYBABY assembly jig
April 21, 2006
Got the new Cornet Classic chassis' in! Wow, I can't believe how well this new
cork/rubber damping material I found works. Better than the stuff I use in the
Trumpet. The ac voltage select board fits in perfectly, makes the internal
assembly and wiring very easy and nice. Clean guts and chassis. This phono is
Had a moment to catch up on some consulting work. Hoping to gain a windfall
(eventually) from some IP. Writing some white papers on new ideas for topology
and construction of my "ion centrifuge" mass spectrometer. Stuff I commited to
a long time ago. Anyway, it'll end up as a few more patents at a minimum.
Started putting RIPPER schematics into CAD. Using the PCM2906, a newer USB
chip from TI. Includes ADC. Taking out the special clock, as I don't see it
adding any real benefits. Removing the spare S/PDIF output too, trying to
simplify this machine and not just throw too many features at it. OTOH, added
a headphone output, useful for monitoring.
April 19, 2006
Parts trickling in. I have the first set of candy bags ready. Will shop these
around. Couldn't get the 9V batteries to fit in the BUGLE bag, so heck, just
leave em out. Adds to shipping weight, and I don't think it will be a deal
breaker. Bags are very thick and protect product nicely. Made to hang on a
Candy bag products
Also making some improvements to HAGUSB. Changing to 1:1 output transformer
(will be in half-kit too) and added a surface mount capacitor to bottom of
baord on VDD line, hopefully to reduce phase noise even more. Damn little cap
costs over $2.
Just signed the contract for RMAF06! Got the same room as last time, 1126.
Don't miss it. This is arguably the best show in north america. And I'll have
plenty of new stuff on display. Probably raffle off a RIPPER.
April 17, 2006
Reworked parts lists for the candy bag items. Trying to reduce costs as much
as possible by using multiple vendors for parts. It helps me out a bit when
not everything has to come from DigiKey. I can use MPJA and Jameco as
alternate sources. Sometimes they have unique parts that really help. In the
FRYBABY case, only Jameco stocks the BNC-to-binding post adapter. Anyway, I
jammed retail prices as low as I could. Spent a lot of time in the past few
months working the numbers and optimizing the best formula for markups. Sure
helps to have such a low overhead. The process efficiency improvements are all
part of this.
Most significantly, I've added the BUGLE (soon to be classic) to the candy bag
lineup. Price has to adjust accordingly, but you'll note my net margins are
actually going down! I expect dealer sales volumes to make up the difference.
Not sure why I didn't think of this before, a BUGLE for $189 at a hi-fi shop is
still quite a bargain. No chassis, but the performance smokes anything in the
neighborhood. And it fits the candybag motif really well.
Finally worked out prices to be FRYBABY $249, BUGLE $189, HAGUSB $189, and
HAGCLOCK $159. All prices are competitive, and I'm adding bonuses wherever I
can think of one. For example, the HAGUSB now includes a decent 75 ohm digital
cable. Even at this price, the HAGCLOCK might still be the cheapest high
performance clock on the market. Makes sense to have the dealers install them,
Anyway, got all four manuals written. New style, simple and direct. Visually
a little nicer than the old manuals. I print them out so the schematic shows
up on the back - seen through the candybag packaging. How many other retail
products give you the schematic? Bag folder pages for each product also done.
Parts arriving for the first production run of all bag units. Now, to sign me
up some dealers (shouldn't I have done that first?). Actually, the FRYBABY was
partially done for a certain hi-fi gear shop. They'll get first dibs. I
wouldn't mind a mail order catalog house like acoustic sounds picking it up.
Right now, thinking I have to totally redo my website. Oh my, what a load that
is. Home page should split all DIY into its own section. Assembled machines
need their own separate category. The present mix confusing. I also need to
make room for the new COMPRESSOR. Maybe I call it VaComPre? That's French for
vacuum tube compressor and mic preamp. That will also end up as a hi-end
dealer product (but not in a bag) only, no kit.
April 13, 2006
Thinking strategy again. Decided to do RMAF06 show. Can only afford one per
year, and I don't think CES is well suited for me. Plenty of small
manufacturers at RMAF. After all, it started as a spinoff from VSAC. So I fit
right in. Will probably just rent an entire room again. Must plan better so
as not to spend a fortune on shipping (and the subsequent damages). This gives
me all summer to hook up with someone else's speakers and turntable. Two very
heavy items I don't want to bring. Will have my rack of "C" series electronics
but with the CHIME added. Should also have a working RIPPER to add as part of
a demo. I'll spin vinyl as much as I can, and then play CDs on request via my
So, to implement that plan, I need to put the COMPRESSOR on the back burner.
Research shows that nobody is going to pay $1k in parts for a kit. However, at
a $4k price point, I can sell it as a premium assembled tube product through
dealers. My hope is to connect with a big catalog operation like Sweetwater to
stock and promote it. Browsing their catalog (I just bought my microphones
from them), my COMPRESSOR fits right up against the competition. I think it
can do well. New plan is to get the prototype up and running, but not focus
too much time on it. This will free me up to work on the RIPPER and the ONE
series. I need to have a working RIPPER and demo chassis of the ONEs at the
show. Oh yeah, will switch the COMPRESSOR to the HAGLABs brand name, too.
I also have to scale back on my consulting commitments. Just did a no-bid on
a new contract. Can focus on just writing up two new patents then. Best to
lock up the intellectual property first anyway. Ok, so that gets the
COMPRESSOR and HAGERMAN ION CENTRIFUGE out of the way. Finally, a plan that
makes sense. One that is achievable and brings in cash at the same time. The
result will be a great show with new kits, plenty of new products, and the
public introduction to HAGLABs. It all happens this October.
Just finished building the first RACK. It'll go on sale at $6999 complete. It
includes a CONDITIONER, two CYMBALS, a CHIME, a CORNET2, a CLARINET, and the
mahogany rack. This is a complete system minus the CD transport and turntable.
Very compact, but 57 pounds, 22 tubes, no feedback. Starting an auction at
audiogon for this first one, a thousand bucks off. Heck, you can spend more on
just a linestage.
Son of a gun. It's my anniversary, 12 years.
April 11, 2006
The CORNET CLASSIC is back! I bought a couple of crates of the original baby
blue chassis. Same old price of $699. Think about it. That's an unreal price
for such a decent tube phonostage. You're lucky to get something this good
used for that kind of money. Polyprop caps, gold plated connectors, chassis
damping, upgraded power transformer, and handcrafted in USA.
April 9, 2006
First race of year today. But too much sewage in the ocean to swim, so race
turned into a run-bike-run. I set myself up with good runs at a 7:01 pace.
Once again, though, my bike leg was subpar and put me out of the money with
another 4th place. Maybe I need a new bike?
Did me a test run of a full HAGDAC board in the toaster. Ran into some process
issues, like having trouble applying the paste, and placing down components. I
think my hand was a little too jittery to set the 0603 parts in place. Made a
mess of things. Paste was thick and spread all over the place. It looked
ugly. But then a magical thing happened in the new toaster oven. The paste
melted nicely and unshorted itself, components floated to center of pads. By
the time it came out, almost all of the problems had disappeared. I couldn't
believe it. Joints beautifully shiny. Was really worried I wasn't going to
get this to work, but I think with a little practice and process improvements
(like clamping the mask in place) I'll get this nailed. Cuts the time to build
a board in half. Or better!
April 7, 2006
So I took a quick look at what a COMPRESSOR will cost. Oh my, pencil and paper
came up with $930 in parts! Holy moly. Of course, that includes $360 just for
the Lundahl I/O trannys. Ok, well, I wanted to make it good. Didn't design to
a price point, went for all-out instead. This is what happens. Shoot, at best
I can deliver a direct sale price of $2999. That's a steal, actually, for what
you get. But HAGTECH is unknown in those parts of the world, so sales will be
difficult. I could offer it as a half-kit, my original plan. But I can't see
too many takers. DIYers don't want to pay that kind of money. Imagine forking
over $1k just for the parts. That would include nice finished panels from
www.frontpanelexpress.com and the circuit board from me (maybe at $150 or so).
Something tells me that won't sell. DIYers will hesitate and find something
else. Hmmm. That sort of changes the plan. So what to do? Maybe I should
follow the HAGLABs lead and do it that way. Be just a manufacturer and let the
experienced dealers do the selling. Why fight it? I guess you could look at
it this way - it would take the sale of 30 half-kits to equal the revenue from
just one assembled unit.
So a dealer sold COMPRESSOR (really need a new name) would probably retail for
maybe $4500. That's as low as I could cut it. They would be custom built in low
quantities, as the inventory costs would kill me otherwise. So turnaround time
of two weeks. That puts me in direct competition with some big names. But my
machine is quite unique in many ways. The only product similar is the SLAM!
from Manley. That goes for $6.6k. Or you could do the two-box approach and
buy both a mic pre and a compressor from the likes of DW Fearn for $8.3k. Or
$4.6k from UA, $7.0l from GT, and $5.3k from Pendulum. All of these machines
are terrific and represent the best of the market. So $4.5k or $5.0k from me
sits pretty well.
Interesting, scouring the magazine like MIX to research the competition. I
find that nobody is using handles anymore on their rack-mount products. Except
GT. I like the handles, and was planning on using them. Probably makes more
sense for my machine, as it has such a shallow form factor. The handles will
come in real handy during installation.
Tried the toaster oven soldering today. Wow, it worked great! I'm suprised
the oven came up to temperature so fast. Got there under 5 minutes. Out came
the gingerbread boy. Only problem was that I couldn't see through the glass,
and smoke came pouring out from the pizza goo at the bottom. So I ran right
out and got me a spanking new oven from Costco. Big stainless steel model, a
mere $55. Yet another capital investment and infrastructure tax deduction.
Too late to run a full board today, can't wait to try one. Just finishing up a
full RACK for sale at audiogon.
While typing in the new FRYBABY manual, I noticed I always use natural harmonic
ratios for everything. For example, the different font sizes used are: 12, 24
and 36. Everything is lined up on grid. Sometimes I use "pi" or the golden
ratio (about 0.618), as in the shape of the Trumpet boards and wood panels. It
just seems proper to use natural harmonics for everything, from thirds and
fifths, to octaves. I hate sloppy design. It has to be neat, clean, and
sensible. Oh my, I'm anal!
April 6, 2006
Had a fun time today playing with that FrontPanelExpress software. Wow, it
works really well. Very intuitive. I just wish I could figure out how to mill
out an ellipse. Can engrave one, but can't make the hole for VU meter as I had
originally intended. Went with rounded rectangle instead. Cost ain't cheap,
it came to $118 for the front panel and $51 for the rear. Both thick anodized
aluminum with engraved lettering. I actually wanted white background with
black letter, but closest I could come was gold.
COMPRESSOR front panel design
Just got in my solder paste and reflow temperature indicators. Ready to try
out the toaster oven soldering trick. This will be great once I get the
process tweaked in. Can then crank out SM boards fast.
Got FRYBABY manual started. Looking good. Didn't stray too far from my
existing manual design. Changed fonts, structure, titles, headers, and tables,
but mostly looks like what I had before. A little cleaner. Will have to cut
edges down to get a 5 x 7 booklet size. It fits into the candy bag better.
Will also use colored paper, maybe a simple pale gold.
Also changing warranty period on all new products to ten years. Some of these
USB chips will end up with obsolesence issues otherwise. Have to be reasonable
April 5, 2006
I kept promising photos of the new shop. Well here they are. First one shows
test and assembly benches, cabinet full of test equipment, my two big Tek scope
mainframes, etc. I have cabling entering the white cabinet on the left, where
I have a turntable, CD player, burn-in station. FryKleaner is mounted
permanently under shelf.
This shot shows the dirty side, with drill press, CNC milling machine, vise,
press, etc. You can also see the library, which I whittled way down to just
textbooks, rare manuals, and back issues of audio magazines.
Finished my 'candy bag' packaging for the FRYBABY. It will look something like
Also made labels for the FRYBABY audio CD. Cut two tracks using Audacity (this
freeware is awesome!) with a direct feed from a FRYBABY, then burned to CD.
Looks very professional. And it works, too. Fired up that CD on my iMac
system and it sounds just right. May as well put that track on auto-repeat and
do a little more break-in on my FE126 cones.
Ok, now to write a new-style manual.
April 4, 2006
Tried running the FryBaby straight into an MM cart today. I figured for sure I
would hear something. The coil/magnet generator should also be a motor in
reverse. That is, if I pump voltage and current into the cart, it should act
as a motor deflecting the stylus in proportion to the signal. Theoretically,
this might be a cool way to break in a cartridge without having to crank
through a lot of LPs. The FryKleaner signal should work the compliance.
Anyway, I tried two different Goldring MMs. Didn't hear a thing. On the other
hand, didn't notice any damage. I even turned up the power to line level! No
smoke. Coils tested fine afterwards. These carts had dc resistances of about
700 ohms, so max current was about 1.4mA. I suppose maybe it is safe to try
this out on a new cart, but at MM output levels just to be safe.
April 3, 2006
Decided to discontinue the TRUMPET "MC". Nobody buys them. Will keep selling
the original version of TRUMPET. It's an unreal steal at $1999. Have one MC
left in stock. Will auction it off at audiogon. Go look for it.
Worked out simpler code for doing the popups. This will make it very easy and
flexible. Just need to get going on the content. Thinking of pages for
soldering fine pitch surface mount parts, desoldering them, audio system
grounding, phono stage hum and buzz, etc.
Spending money like water, all on R&D and process improvements. That's ok,
sales continue to be strong. But hey, no way to fine tune a COMPRESSOR without
having actual microphones and studio cables. Not cheap. Probably get a Shure
SM57 and an AT2020.
Feeling confident about my short and long term plans and strategy. Seems to be
straightforward path to dealer sales and to continue with kits. The more I
study, the better positioned my products are. Do you read the audiophile mags?
I scour them cover to cover, researching the competition. Lately, prices have
been skyrocketing. Holy moly, what are these guys doing? Are there really
enough customers that want $20k+ machines? I don't think so. Carnage ahead.
But for me, the window for under $3k stuff is opening. It's like everything is
shifting in my favor.
Now the real question is what to do about RMAF06. I'd really like to go. Can
I do a room again? Or should I partner with someone? I'd hate to have to use
the same amplifiers I introduced two years ago. Would be great to have the new
"one" series prototype ready in time, but not sure that is possible. Not
unless I put the COMPRESSOR project on hold, which I don't want to do. Busy
now getting the 'candy bag' items to market. Next up is the RIPPER. Can I do
a room by myself? Or maybe a static display? I dunno.
March 31, 2006
Sheesh, it's like night in here. And it's midday! The clouds and rain are so
thick I have to turn the light on inside. Too many days in a row of rain. Got
first race of the season in one week, can't go out to train. Ok, maybe you
ride bike in the rain, but I don't. Rain rain rain. At least my car is clean.
Got production FRYBABY boards on order. Also bought a bunch of packaging for
them. Now to write a whole new manual - a little more customer friendly than
the DIY manuals. Need new visual design to be shared with HAGLAB machines.
Worked out retail pricing for the dealer line: FRYBABY $239, HAGUSB $189, and
HAGCLOCK $169. I actually net a bit less than before, but hoping the volume of
sales makes up for it. The HAGUSB price was going to go up anyway, as the time
for the introductory discount rate is pretty much over.
March 30, 2006
Saw 'Audacity' listed as a favorite in PC Mag. Looks to me like this will be my
first choice for recording software for the RIPPER. Simple and easy. And they
give it away free. Runs on both Mac and PC. Can't ask for more than that.
Thinking more about this dealer retail option for some products. Evidence is
mounting that that will be the best way to boost sales. Came up with nice and
simple packaging - a see-through bag stapled to a cardboard strip at top, such
that product can be hung on display. Just like a bag of candy. Why didn't I
think of this before? Very straightforward. No money wasted on excess. So it
occured to me maybe I have more to offer. Like maybe a HAGCLOCK in a bag. Buy
it and have the dealer install it! Great way to upgrade your existing CD
player. And I don't have to worry about inexperienced customers hacking up a
valuable machine at home. Maybe to the same with HAGUSB. Prices will have to
go up a little, but nothing too dramatic. It would put several fine products
into the reach of many, and not limited to just DIYers. Meanwhile, the DIYers
would still benefit from the super-cheap half-kit option. I'm liking this a
lot. Just not sure whether to make it HAGTECH or HAGLABS.
New production lot of HAGDAC boards came in. Need to get the thermal crayon
goop and solder paste so I can experiment with toaster oven soldering. Don't
laugh, word has it the results are more than professional. And the budget is
ok for me. I'll be able to whip these things out much faster.
No news yet on when the Cornet Classic chassis are due. Got the new AC voltage
select boards - so I can easily use the 370BX tranny upgrade. Now I just need
to figure out how to announce to the world that this classic is back at $699.
It's a price that seems very desirable for phono stages, and for a superb tube
machine it is stellar. But hey, without good advertising or word of mouth,
it'll just be chassis sitting in the shop collecting dust.
Continuous efficiency improvement - Japanese style manufacturing. And I am
borrowing it as much as I can. Not only am I switching to reflow solder
process for the surface mount stuff, but I just built a neat assembly jig to
help build multiple HAGUSBs and FRYBABYs at the same time. Make it real easy
to stuff and solder 4 or 5 boards at once. Faster, easier, fewer problems or
Thinking I can add a CD to the FRYBABY package (of same signal), such that you
can also burn in transports and players. Low cost item, but high value. This,
along with wall wart and handful of cable adapters will justify the price.
Similarly, HAGUSB package will include both USB *and* 75 ohm digital cable.
March 29, 2006
Ok, folks have gone home. Spent a lot of off-time entertaining and keeping kids
busy during their spring break. Was able to keep up with orders working a bit
of overtime. Almost finished building the COMPRESSOR. Chassis came out really
solid. Mostly working on mechanics; design and fit, right now. Will start to
debug sidechain next. Anyway, glad to see the sandwich concept is valid.
Back onto FRYBABY. Tuned the circuit a little for better performance. Now it
is spitting out perfect waveforms. No compromises. I'm suprised how well this
little circuit performs. Need to make a few small tweaks to circuit board and
will then order production quantities. The question now becomes, do I sell
them in red, green, or blue? I really don't want to stock all three colors.
DIYers have their choice. Wondering if maybe I should also make this a HAGLABS
product too, as I think dealers could sell a ton of these things. Was thinking
of offering an upgrade package of various connector adapters (BNC, binding
post, XLR, etx.). Maybe the HAGLABS version would have that included. Might
need a bit more professional packaging, too. Like a see-through plastic wrap
that hangs on a wall. Also include a wall-wart. Maybe a better manual, too.
I like this idea, as I usually don't think big enough. Can be sold as HAGTECH
half-kit, or basic assembled version, or premium extra via HAGLABS dealers.
Need to look at pricing, see what can be done.
Meanwhile, trying to keep up with all of the RoHS (reduction of hazardous
substances) lead-free ,andate for EU. I think it will be no problem meeting
this requirement. Almost all available parts from DigiKey et. al. are being
converted anyway. Getting difficult to buy non-compliant parts. Only
remaining issue is the lead-free solder. I've been using it on Trumpets, will
switchover to other products soon. At least the ones headed to EU. I feel it
is worthwhile to meet country and agency requirements for safety and other
regulatory issues. Seems to me not all small hi-end audio companies do. It
helps I have a long background at other companies designing machines to meet
standards (UL, FCC, etc.), plus several friends in the business.
And it continues to rain here, with no end in sight. My first triathlon of the
year is in two weeks. Hardly any swim training. Difficult to get out on bike
as roads are always wet. Not going to be my best race. Then again, nobody
else is getting out there either.
March 23, 2006
Stuffing board, keeping track of mistakes. So far, a bunch of traces on top
layer shorting to metal feet. Got the pinout of 78L08 backwards. But now I
have it mostly populated and +/-8V comes up ok. No shorts or goofed up chips.
The entire sidechain is alive. Minus a few parts. Need to order a bunch of
1/2W resistors of the right values. Then I can finish. All parts fitting
nicely. This pic is the front side, all capacitors, sockets, transformers and
connectors are on the back. Controls, ICs, resistors, etc. on front. Lotta
debug work and tuning ahead.
COMPRESSOR circuit board
March 22, 2006
Well, the folks are here and driving me nuts. Just kidding, of course. And
the Par-Metal enclosures came in for the HAGLABS one series - holy crap these
things are nice! Very happy with them. Going to be a beautiful set of
machines. Later this year.
Started putting the COMPRESSOR together. Drilled holes in front and rear metal
plates. That's a bitch. Especially cutting out the oval hole for the meter.
Anyway, things fitting pretty much as I had hoped. Discovered a problem again
with the metal feet on the rotary switches and pots, they short to traces
below. Why do I always do that? At least I caught it in time. Will fix next
version of layout, so production units to be perfect. So I got the sandwich
built, just to check out. Haven't really mounted any parts on board yet, just
checking fit. I love it. This is one way cool concept. You can see from the
top view just how compact and thin this chassis is. Tubes and tranny stick out
back. Didn't mount any knobs yet. Better pics later.
COMPRESSOR frontCOMPRESSOR top
March 21, 2006
Major cleanup work over weekend. Got most of my part inventory sorted and
collected into common sense bins. Lab is now chock full. Wow, what a
difference! When building stuff, everything is right there. So much easier to
get stuff done without having to look or think. I should take some pics.
Got FRYBABY up and running. Came right up. No bugs. Only change was to
lower a resistor value to make LED brighter. Looks really good all closed up.
I'm not happy about the remote panel mounting of connectors. Need to check out
the possibility of board mounting again. Makes life easier. Just not sure
there is room on output panel. Wall-wart option also works good. Also bought
a bunch of adapter connectors (BNC, XLR, phone, post, etc.). These will make a
great option package.
COMPRESSOR board came in. It's beautiful. No wonder it took me so long. Cut
and drilled the panels. Will start to populate tomorrow. Probably won't get
too far, as the parents have arrived and I need to entertain.
March 16, 2006
Ugh, I keep placing too many part orders. My inventory control is a mess. It
seems every time I buy some parts (for products or protos), I forget something
and have to put in another order. Shipping costs are killing me! I have to
get this under control. Used to buy too many parts, now I buy too few. Really
eating into profits (if there are any). I am now waiting for delivery of a
DigiKey box, two Mouser orders, one Jameco, one MPJA, one Par-Metal, two
circuit boards (not including COMPRESSOR yet), a stencil, and a box from New
Sensor. All in transit. Oh yeah, and classic cornet chassis, too. And I'm
still short some parts needed to finish a pair of VacuTraces! Sheesh. I am
planning on getting them both built before my parents show up next week to
visit. Once I get the new lab stocked and parts on hand sorted, I can
hopefully do a much more efficient job of inventory control and part ordering.
So must clean up lab and inventory. I'll be in great shape. Already
streamlined my design process. From concept, to design, schematic, layout, and
prototype is a very short cycle. That part is working very well. I probably
have the lowest R&D costs in the business. Plus, production assembly is a snap
too, thanks to attention in the design stage. The layouts are more difficult,
but it pays off in assembly. I can crank out machine after machine with nary a
Finally finished the COMPRESSOR layout. A work of art she is. Took me
forever. This is the ultimate statement in Hagerman design. Everything mounts
to the board. No wires (except for power tranny). Heck, hardly any chassis.
It's a revolutionary concept, but I doubt anyone will copy it. Proto board on
So why is my blog filled with all this wierd nonsense? I dunno. It just comes
out. I write about the business, what's going on. That's all. Sort of
documents the trials and tribulations of operating a small high-end audio shop.
Not much else. Oh, which reminds me. Had an awesome brain fart the other
night while driving. Not sure what caused it, seemingly unrelated to anything
going on around me. Just popped into my head. Anyway, finally came up with a
solution for how to construct a 3H mass spectrometer. Totally backwards from
what we were doing. Completely non-intuitive and non-obvious. The trick is to
flip the grid orientation 90 degrees from what one would expect. Solve several
problems at once - cost, vacuum, accuracy, and capacitive loading. Damn I love
when that happens.
Just jammed out a quick ad. Good artwork and copy is hard to come up with.
This is not my best work, did the entire thing in an hour. Deadline. Wanted
to promote the newer machines.
New aXp ad
Started working on the popups. Came up with some test code. Looks like a link
but it isn't. Needs work.
March 13, 2006
Cranked on taxes this weekend. What a pain trying to gather up every expense.
Good part is that most of them are deductable. Just takes awhile going through
the credit card bills to sort purchases into various categories. Not fun. But
would be nice to actually submit on time this year, without resorting to an
Meanwhile, the efficience quest continues. Discovered (from a guy up the
street, no less) that you can reflow solder surface mount boards in a toaster
oven. That seems to be the new DIY thing. They even sell controllers for
cycling through the proper temperatures. I think I'm going to try this with
the HAGDAC. Was planning on having them sent out for assembly once orders are
sufficient. But maybe this is a better way to control costs. Debug this
process and I can use it on future boards. I like it. Just ordered next batch
of HAGDAC boards today anyway. Will get a stencil, some solder paste, thermal
crayon and try it out. Down't want to waste too many boards doing this. Must
be perfect or I won't ship. I think this will save me a lot of time. It may
also improve quality. And I can make breakfast at the same time!
Customers suggesting I add more useful technical content to the website. Like
how to solder, how to chase down grounding and hum gremlins, etc. So I'm
looking into popups that can be linked to from any other page, even emails.
Useful tips. Maybe a free circuit once in awhile. Been neglecting content
lately. Then again, last half a year I've really pushed the posting into audio
forums. Figure it's like free advertising. All DIYers and most high-end audio
folks are computer savvy. The more noise I make (giving out advice, etc), the
more folks are invited to the website. It seems to be working. A lot of extra
hours most every night is paying off. Orders are up. Or maybe it is something
else. I'm just guessing. Of significant note is that my paid advertising is
way down. I may just cut it out altogether. The more overhead I can cut, the
cheaper I can keep my prices.
Lots of good industry discussions Saturday night at our audio club meeting (we
also played with the crossover for an old Altec). Discovered that chinese
reliability is still a big problem. Not the advances I expected. Some
machines have 100% failure rate! Holy cow. The only problem I ever had on any
of my products was the 270BXs on the original Cornets. Other than that, my
failure rate must be well below 1%. Boy am I happy about that. Anyway, it
also seems that customers still have some reluctance to the chinese-branded
machines. But not for the other-branded machines made in the same factories.
Seems the low cost machines from even established companies are now made in
China. Hmmm. So maybe I have overestimated the competitive threat. In fact,
maybe this is actually helping me. The establishment is moving upstream to
even higher prices, giving up the $1k segment. This, I think, means window is
opening up in the $2k to $3k range. That is right where me "one" series will
debut. By keeping my costs down, I can deliver more amplifier for the money.
It's been tried before. Quicksilver has been successful for many years using
this model - and apparently have never advertised. It is becoming clear to me
that advertising is what drives the cost up so high. Ok, enough rambling. I
gotta get back to the popups.
March 10, 2006
Almost done with the new shop. Added two more cabinets and another table
today. Getting very tight. Not much for square feet, but the organization and
layout make it a great place to work. Biggest improvement yet is a new lamp
above the soldering table.
Got FRYBABY boards in. Looking great! Built one up and fit into box. Gonna
be cool. The cases not as translucent as I had hoped, you can't see the
electronics too well - and the board is beautiful. Maybe when the LED is
flashing it'll take on a new aura. Maybe I can get it fired up this weekend
Case available in red, green, or blue.
March 8, 2006
This COMPRESSOR layout is taking an agonizingly long time. Just routing the
power lines without corrupting ground plane is a lot of work. Added jumpers to
get across the B+ and heater lines. Mostly done. Worked out the ground plane,
cutouts, and voltage spacing today. Just need to run the long connecting
traces between sections (like the vactrol LED).
Meanwhile, every day I get more of the new lab up and running. Found room for
yet another bench. You can never have enough countertop. Not sure where I'm
going to put all the shipping supplies. Two more cabinets to buy and set up.
I like the way it is very clean and organized. Will post pics one of these
days. The cleanup is great. Selling lots of unused equipment on eBay.
Updated the HagLabs logo so that it is stencil-like. That is, so it can be a
cutout for ventilation in a chassis or made into an extrusion. Might come in
handy someday. Bought four new chassis today from Par-Metal. Will see how
nice these look and if I can turn them into the "one" series. Thinking I will
do machining and silkscreening here locally. That way I can buy standard
chassis in bulk, and then customize to order. Efficiency and cost are
everything, especially if I want to hit the very demanding price points.
Just got some more photos in. Here's the crew, apre-ski at Park City.
March 6, 2006
Oh man, spent the entire weekend working on the new factory. Ok, let's call it
a shop, as it is really kinda small. Discovering I have less suqare footage
than before, so it's getting really tight, especially for storage. Focusing on
efficiency and keeping things cleaner. Two new benches installed for assembly
and test. I improved my test station with an extra scope (two big Tek 7000s),
built a burn-in chamber, and enclosed the turntable. Now have quicker access
to all the signals I need, whether USB, S/PDIF, MM from table, headphones,
frykleaner, variac, serial port, etc. Even put the VacuTrace into a cabinet.
But it sure is taking forever! Didn't realize I had so much crap laying
around. Am finding some good parts, though. Came across another soldering
iron, plenty of trannys, and a batch of BUGLE boards. Spending the extra time
to do this up right. Pouring bucks into it, new benches, tables and cabinets.
Finally got some pics back from skiing. Hope to get some movies soon.
Skiing in Utah
March 3, 2006
Getting a lot of calls for a BNC output connector option for HAGUSB. Wasn't
sure it was going to fit, but tried it today anyway. Ding! It can fit,
barely. Same hole as RCA, I put in a panel mount BNC (I think it was a 50 ohm
one). Short 1/2" wires down to circuit board. Ok, so I'll now offer this as
If you've been reading this blog, you'll know I've been fretting a long time
over the Chinese invasion of hi-end audio. Hoards of new tube products from
new companies, new brand names. Very nice looking chassis, decent classic (is
that a word for copied?) circuits, and at awesome prices. Reliability has
been sparse, but that will improve quickly. Somebody is going to make a lot of
money as I predict sales will remain brisk. Very hard for someone like me to
compete against. Or are they?
There is no way to compete on price. But what about value? Certainly there is
more to a customer's wants and needs than sheer price. So let's take stock of
potential advantages HagTech has. First, I think, is simply 'made in USA'.
Recent worldwide events, both political and economic, such as outsourcing of
many good middle class jobs, breeds a protectionist sentiment. Or whatever you
want to call it. I think the designed and crafted in USA reality will be an
intangible but significant factor in purchasing decisions. Also, craftsmanship
and personal attention will be appreciated. Secondly, high reliability with a
great warranty and return policy will help long-term. Machines of proven
quality with a low failure rate is almost expected. Follow this with
exceptional service (from a local factory) and customer support in English and
you have another edge. The final advantage lies in innovation. It appears
that the low-cost crowd are followers, not leaders. They do not push the
envelope or state-of-the-art, but rather mimic what has already been done.
This, I think, will be my greatest asset. It is time to use ingenuity to push
a new level of innovation. This is exactly what I am doing on the upcoming
'two' series. Improvements in technology, remote control, ethernet, USB,
topology, ease of use, 'digital tape loop', etc., for high-end audio will
differentiate my offerings and provide value ahead of the curve. I sort of
actually stumbled onto this already with the CHIME USB DAC. At present, I
believe there are only two commercial tube USB DACs available on the market.
The CHIME and the COSECANT (BRICK). Isn't that incredible? Out of seven or
eight hundred audio companies, there are only two such beasts. And mine is
the best! The hard drive music source is a big wave of the future. And I'm
getting into the recording side, too. It appears that pushing the new markets
and technologies and circuit innovations will keep me out front. I had started
behind, but now have caught up. In other words, there are plenty of ways to
compete. In a way, this invasion has only created more opportunity.
March 2, 2006
Ordered proto boards for FRYBABY. This will be a way cool kit. Result will
appear more professional than most things you can buy. Jammed on COMPRESSOR
for awhile. Tired of laying out boards already. Re-routed B+ down through the
middle between heater lines. I like symmetry. Tube section wired up. I think
the hard part will be the +/-8V supply lines. Prefer to keep them wide traces.
Looking real nice.
Took a look at the referral responses on my order form. Most folks actually
fill it out. For the past three months I get 30% asylum, 24% other, 14%
google, 8% audiocircle, 7% audiogon, etc. This is by pure count, not revenue,
which could paint a totally different picture. But it shows that the Internet
presence and forum traffic is a big part of my business. Suprisingly, I got
less than 2% from audioXpress. And I've been a big advertiser there for years.
Maybe I need to re-think that. My contract just expired and they're pushing
for a new one. Hmmm. Seems like maybe I actually lose money on the deal.
March 1, 2006
Ok, so I added the "power one" to the HagLabs lineup. A pair of CYMBALs in one
chassis. Wasn't as easy to fit as the others, thanks to all the iron. But
stacked it up in an 8" tall box. Looks like it will work fine. Probably won't
ever sell any, but it rounds out the product line. This is a great way to kick
start the business (through dealers) and buys me time to design the "two"
series, which I've spoken about previously. They are still in the works. Each
day I put some thought into them, trying to add innovation, features, or
simplify somehow. Lately, been looking at using ethernet as a control port,
which offers isolation from ground and power.
Back to FRYBABY. Just about finished all-new layout to fit into the red box.
Looks very good. Had to panel mount the RCA jacks so they looked right. Made
label. Gonna look way cool, even the kit ones. Should have layout done
tomorrow so I can buy boards. Ordered boards this morning for ac voltage
select inside the CORNET. That way I can easily use the 370BX and have clean
wire routing. Getting quote on new CORNET chassis. Once I get these projects
moving I can finish the COMPRESSOR layout, which is 80% complete already.
For the RIPPER, folks have suggested using "audacity" or "total recorder" as
software apps. I need to try these out and make sure they can talk to the USB
audio device. Then, I can copy the RIPPER circuit into the "digital one" and
have the same recording capability added to the new tube line.
Just realized I am designing eight machines at once. No wonder I feel bogged
down! Good thing that vacation revitalized me.
February 28, 2006
Ok, I'm back. Skiing was awesome, had a great time. Should have some pics
soon. These guys aren't slowing down. All mid-40's, yet they always head
straight for the trees and double black diamonds. Then again, that's where the
best snow is, in the shade of the trees.
Put some more thought into the HagLabs components. Added some pics to the
website. Need to design new boards for front and rear panels. Going to change
to the PCM2903 for the USB I/O. See if I can also add the ADC capability.
Ordering chassis for the CORNET CLASSIC. With seven projects (actually ten)
in R&D right now, it's easy to lose focus and not get anything done. Need to
push a few of the higher priority ones out first. Going to be one busy year!
February 22, 2006
Argh. Get a call at 5am, flight cancelled. Ski trip delayed. Yeah, like I
got time to waste. Shoot, best I can do now is get the redeye out (16 hours
later), shuttle straight to the slope and get in a 1/2 day. Somehow hook up
with my buddies so I can find the condo.
Meanwhile I took the time to ship out another CHIME and work on the chassis for
the ONE series. Figuring I can use common circuit boards for front and rear
panels. Overlap connectors, circuits. This will greatly help wiring, plus I
can optimize signal routing by using relays in stead of a very remote selector
switch. Easily mount the step-up trannys, loading switches, and a new USB
input circuit (I'll add new buffer stage) with auto-USB selection. Costs a bit
more, but will make the units easier to build and cleaner. Worked out some
images of what they might look like:
New ONE series for HagLabs
Not sure if I want to use handles or not. Main circuit boards are the ones
from CLARINET, CHIME, and CORNET2. Calling the enhanced machines LINE ONE,
DIGITAL ONE, and PHONO ONE. Line will have four inputs, two outputs. Digital
gets a filter switch and auto-USB input select. Phono gets internal MC stepups
and possibly variable loading. I'll make the LEDs blue. Power LED will blink
until B+ comes up. Fancier looking than the retro red/green. Not cheap, these
machines will be tweaked out with exotica parts (alps blue velvet, etc.) and
maybe NOS tubes. Will be very much high-end in appearance and performance.
Did I mention I was going to re-issue the CORNET CLASSIC? Yeah, same old
baby blue chassis, same old price, but with upgrade to 370BX tranny and an AC
voltage select board. Assembled only. Will be a steal at $699!
Checking out this new FRYBABY case. Gorgeous translucent red. Very compact.
Will be a portable frykleaner running off a 9V battery or wallwart. Fit in
your pocket. See the way cool electronic guts through the plastic. Gonna be
a winner. Available in half-kit, too!
February 21, 2006
Cranked the entire weekend working on the new factory. Cleaned it all out,
painted floor (was cement) and walls. Put in some new lighting, started
building benches. Those melamine cabinets from Home Depot work out well. Next
week I should be able to start moving equipment. But for now, I'm afraid I
have to go skiing. Back in a week!
February 18, 2006
Finally got placement of COMPRESSOR done. Updated link below. Parts all fit,
now trying to run traces. Had to change 1/4w resistors to 1/8w in order to get
the heater traces through the middle. Phantom power regulator still a little
odd, but everything else is nice and clean. Wish I were done already. Going
to try FrontPanelExpress for the two panels. Would make for a great half-kit
if folks could just buy pre-cut panels straight from them. Chassis work is
always a problem.
Meanwhile, had a good brainfart on jumpstarting HagLabs. One of the problems
I had was financing the operation during R&D. Designing all new machines from
scratch would take about a year. Question is how to make a living during that
time. Then I realized I already had the advice. Took a look at what products
were selling and why. Kits fly out the door, assembled machines do not. I'm
talking about the CLARINET style line. Great products at a steal. But they
sort of came out ugly. Perhaps a bit too retro and industrial. Then, look at
CORNET2 by customer
So the advice was to change the chassis and upgrade all parts (resistors, caps,
jacks, tubes, etc.) to make a super tweaked out version, like many of you have
done with the kit. Use what I already have to make a new product line! I can
bootstrap HagLabs with a phono, line, and dac in matching fancy chassis with
thickass faceplate. Something with far more WAF. Move the tubes inside so
they can be stacked. The hard part is already done, having the circuit boards.
I'm working out a new pricing structure whereas I can sell these items through
dealers. Like it or not, that's the only way to move product in this industry.
Thinking this super-cornet would have built-in stepups and the 370BX tranny.
Maybe even variable resistive loading. The gain of 64dB (or 74dB) would mate
with MC only. Tweaked out it would be very competitive at a retail of $2000
or so. If I keep my costs low and overhead efficient, this is possible. It
would launch HagLabs with a set of already proven circuits. And it should
carry me through to the next line. The DAC would probably be $2500. Business
strategy, marketing, production, and circuit design all coming together as one.
Making slow progress on RIPPER and FRYBABY. Fortunately, a steady stream of
orders keeps me busy. I'm just so happy I haven't had to do any consulting
work lately (although I'm signed up for such).
Next week, skiing in Utah!
February 10, 2006
Agonizing this layout stuff. Gives you a headache. But rewarding. A good
layout is key to a great design (even for you PTPers!). This COMPRESSOR is
only two layers, but has lots of supply voltages and has big parts on both
sides. I ended up putting resistors on top, capacitors on bottom. Took all
day to place parts and wire up. Needs proper arrangement now to clean things
up and accommodate traces. Last step is routing the traces. Worked out a lot
easier than I expected, plenty of room for the tube section. Can't wait to get
this machine built! Very fun project.
Also got in new cases for the RIPPER and FRYBABY. Nice. Both projects need
mechanical redesign and new layouts.
February 8, 2006
COMPRESSOR parts in. Had a real hard time getting a 1/4" phone jack to fit my
scheme. Tried four different ones. I need vertical panel mount. But PCB is
only 1/2" down. Finally found one (switchcraft) that could be made to work. I
just have to route out a small slot in board to get the contacts through.
Still moving parts around. Meter select and power switch had to move down,
making room for ac plug. Then the link switch didn't look right where it was.
Found a better place for it visually and logically, but it adds a bit of length
to the traces (not a big deal). At first it seemed like insanity, as I had it
right in the middle, but opposite side from the rectifier tube socket! Bad to
mix high voltage stuff with traces right next to low voltage sidechain. At
closer inspection, though, I had spacing, plus the low level signals were right
next to heater pins (+/-3Vac). Totally illogical, but I think it's going to
Pretty much decided I can't do the VTV show in May. Just too much going on
right now. Three machines in development. Hard to put together a show at same
time. I'd love to demo the CHIME and other stuff. Might even have a RIPPER
and COMPRESSOR ready in time. Could have done some live mic stuff. Or an A/B
of vinyl ripped to HD. And I really like the double room suites they chose.
Actually asked AS if maybe I could borrow some Klipsch LaScalas for the show.
No go, but that would have been a way cool match for the CYMBALS. Probably
would have gotten me kicked out of the hotel.
February 6, 2006
Orders for the HAGUSB keep flying in. This product has turned into a hit. I
can't seem to keep parts in stock. Just shipped two to Greenland, of all
places. Couple weeks ago one went to Malta. Guess you could say the word is
out? All of my annoying posts on asylum, audiogon, & others is paying off.
Even sold a pair to a very big name audio company in the UK. Surpassing the
BUGLE as my big seller.
So I did a summary report of everything sold in the last two years. Some of
it good, some bad. Revealing to dig into the data to sort out the trends.
Seems I am best at selling phonostage kits and esoteric items. Never expected
it, but I made the most money on VacuTraces. Then Trumpets. Big ticket items.
What I don't sell are the assembled versions of the "C" series. Nada. Boy was
I way off on that one. Worst market projection was the OBOE. Lost money on
that one. FRYKLEANER is still a hit. Especially the pro version. That's why
I think there is room for a FRYBABY. And based on BUGLE and HAGUSB sales, the
RIPPER should be a killer. Just bought chassis for it today. Working out all
February 2, 2006
So I'm back on the FRYBABY. Needed a bunch of circuit changes to clean up the
waveforms. Had a problem with loading, which an extra opamp fixed. Schematic
done, ready to update layout when I discovered Mouser no longer sells the way
cool plastic housing. It was rounded an modern, made for a great portable fry
machine. Ok, as they say, every bump in the road is an opportunity. Sure
enough, looking for replacements I found this transluscent plastic case. More
square, but with the see-thru gimmick it's a winner. Also adding a dc jack for
when you get tired of burning through batteries. Still, the concept is great.
And the circuit has been entirely redesigned for lower cost. I ripped it as
low as possible. Will make a nice half-kit. Anyway, now I have to wait until
I get one of these boxes in hand before I can finish layout.
Need to do same thing with the RIPPER. Not really happy with the cases so far.
Would be another opportunity for a translucent box. Only with this product I
have to go with DigiKey. Spent more time going in circles. Did have a switch
to select output mode, analog (phono) or digital (dac), but I figured why? Is
it really necesary? Simplify! So thought maybe an analog switch could be
toggled whenever the USB data was active. But no good analog switches fit my
requirements, and I'm a little worried about added distortion. So changed it
to relay. This will work. Then removed spare opamp and comparator sections by
consolitdation, use the comparator as an opamp! I've done this before, but can
be tricky. Easy to oscillate. So I built a super regulator using an LM393 as
the opamp. Lower performance, but with the addition of a cap on the output, I
don't have to worry about headroom. Don't have to pay for a RR output opamp.
Hey, I take this cost stuff seriously. Every bit adds up. Must rip out all
unnecessary stuff. And improve performance at the same time.
January 31, 2006
Discovered all sorts of issues with my compressor layout. Output too close to
input circuits, the concept of looping back on itself wasn't going to work etc.
Had to rearrange everything. Wasn't looking good, all controls shifted too far
left, panel didn't make sense. Went in a lot of circles until I finally came
up with a solution. I got parts on top of parts. No idea where I'm gonna put
those big capacitors or how I can route heaters and B+. Very tight. So far,
though, by really pushing the goals, I've been rewarded with excellent results.
As you can see, the new panel has a decent appearance, all controls have finger
room and make sense. Only ones that seem out of place are the hard/soft mode
and phase switch. Can't move these, as the input circuitry has to stay tight,
output has to stay away from it. At 45dB gain or so, we don't want feedback
instability. I even found a VU meter! Somehow, by luck, I had plenty of room
for it. Packed with performance, this will be the most compact stereo tube mic
pre or compressor ever. And it's got both! Long way to go yet. Schematic is
done, questionable circuits already prototyped and debugged, parts on order.
Compressor panel and PCB
Meanwhile, had a great bike ride the other day. Started out bad, pump blew up,
flatted the front even before I left the house. All I had was my tiny hand
pump, which put a good 80 psi in there. Ok, so I had a soft ride. Then,
halfway through I flatted the rear coming down Diamond head. Trying to remove
the tire I broke my tools. Sheesh. And I had only one spare tube. Then, out
of nowhere rode Tore, one of my training buddies. With his tools, I got the
tire fixed and back up to 80 psi (needs 120!). Anyway, I figured I'd try to
hang with him on the way back. Except he's an elite. Much faster than I am.
An age group champ. Ok, it started out fine, but then we hit the flat stretch
into Hawaii Kai. That's when he takes off. Keep in mind he's been training
for ironman New Zealand, I'm coming off my winter break. And I'm riding on
sponges. So we get up to 27. One guy to pass. Foooom at 30. Holy crap, how
long is he gonna do this? There's a friggin headwind of 15-20! I'm glued to
his wheel. Anyway, he pulls me for 7 miles on this upwind leg averaging 25.
Crap, normally I would do this at 20. It took everything I had to hang on.
Dang, what a great ride!
January 26, 2006
Cleaning up the lab, trying to get ready to move it to the other side. Rebuild
all the benches, add new cabinets, etc. Discovered I have lots of projects or
parts sitting around gathering dust. From my discmaster, to clocks, trannys.
See forum post with pics for a fire sale.
People keep asking me if they can put a HAGCLOCK on a HAGUSB. The short answer
is no, because you'd need extra +/-12V supplies that float with respect to the
audio ground. Not that big of a deal, but seems like a lot of extra work and
cost. So yes, it is possible. But I was wondering if there was a simpler
solution. The oscillator section of HAGCLOCK runs from 5V. How about using
just that? If the board were mounted intimately close to the HAGUSB, there
would be no need for the comparator and output buffer. Just tie the clean
sinewave to the USB chip xtal input. And so I did just that. Made a chopped
up version of HAGCLOCK and attached it to a HAGUSB. Ok, it works Can't tell
if it is an improvement or not. Listening comparisons with a unmodified
version were done through a CHIME. And that's the problem. If there was a
difference, the CHIME's reclocker erased it. Anyway, I'll make a forum post of
Mounted the clock board on standoffs, ran three component leads down to connect
power, ground, and clock out. Circuit will be prone to low frequency supply
noise, as there is no regulation on the 5V.
Oh yeah. Forgot to mention. I had been pretty happy and impressed with the
headphone output of the HAGUSB. Punchy and detailed, not bad sound. Until I
compared it to the CHIME. Holy zipcode, batman. Totally different planet. I
couldn't believe the difference. Absolutely jaw dropping. So folks, forget
the headphone output, just use the S/PDIF.
January 24, 2006
Got my AirPort card last night from eBay. Plug and play. Now my iMac sits in
the listening room and can play iTunes, CDs, etc. all with a fast net
connection. Glad I didn't have to install wiring.
Thinking more about the HagLab machines. I'd like to add a direct connection
to computer for remote control. Leaning towards ethernet now, as it uses an
isolation transformer, keeping my audio grounds a little cleaner. Way better
than a serial port. Wonder if I can combine a control port and audio I/O port
using one USB connection. No idea. Anyway, if I use ethernet, the linestage
would be a web server. Basically, you could pull up a web page with all of
the controls nicely arranged in tables. Of course, all of this is in parallel
with the IR remote. This is a bit of a paradigm shift, in that the listener
could sit there with a laptop and control both software and hardware.
Finished putting in the auto-DigiKey parts lists. Discovered a few new part
shortages (RoHs is doing it). Then I had another brain fart - probably
obvious to you the whole time - that I can deliver full kits! And do it
without carrying expensive inventory. Here's how: I use your credit card
number to order parts from DigiKey, AES & Lansing, have them ship directly to
you. Basically, I do the typing for you. I don't have to stock or ship. The
box comes directly from the distributor. I never touch it. I don't get the
insane shipping costs to/from Hawaii. Customer receives kit in up to four
boxes, all from different places. And each part is labeled. Money does not
travel through my credit card. Hmmm, wonder what I could charge for such a
service? Would have to be for domestic only, as export and VAT issues get
complicated. Is this worth $100? You can pay $400 for a set of Cornet2 parts,
or $500. Is this even legal? Would be if I routed the money through my
credit card. Anyway, putting up the auto-bom should make things easier for
Also debating whether or not I should do the VTV Expo show in May (NJ). Price
is right, the rooms look great. Can I be ready in time? Who can I borrow a
table and speakers from? I prefer to ship as little as possible.
January 23, 2006
In case anyone is interested, my model for the VTL5C3 vactrol is as follows:
a = 0.004056797;
b = 0.05077257;
c = -0.02730328;
R = 1 / (a + b*i + c * sqrt(i));
where i is current in milliamps. This gives a very good approximation for the
Just thought of a new marketing acronym: A.U.D.I.O. - ADC USB DAC INPUT OUTPUT,
That's what my new RIPPER will be. Or something like that. Ok, maybe that's
January 22, 2006
Someone pointed me to head-fi forums. So I took a look. Wow, they seems to
have a lot of activity! Busy place. And they cover topics of computers as
sources. So I trolled a bit, then registered and added two posts. Gave some
suggestions on how to solve some problem. Ok, fine. Hours later, I am locked
out. They've disabled my ability to post! Sheesh. I can take a hint.
January 20, 2006
Got a pile of HagUsbs built and shipped. Back to compressor. Worked out the
equations for gain, did some mathcad simulations to see what response looked
like with linear relationship between input signal and LED current. Came out
nice and smooth, exactly what I was looking for. Guess I don't need to use my
special gamma circuit to do any corrections. Makes for a simpler circuit.
I also have another limiting mode, but the equations are recursive and I didn't
feel like trying to pull them apart. Basically, the compression is dependent
on the output, not the input. So it becomes a function of itself. Thinking
that maybe I have this backwards in my head. Perhaps what you see above is the
"hard" limiting mode, and the recursive connection ends up with an even softer
knee. Hmmm. Wonder if I need more gain.
Did some more listening tests with downloaded music. Not cutting it! Sheesh,
thought I would try an old Beatles song (try that on iTunes). Downloaded a
mono of "all my loving" in FLAC. Converted to AIFF and imported to iTunes
where I converted to Apple Lossless. Is that too much to ask? You would think
going from one lossless format to another you wouldn't lose anything. Heck, I
could write an algortihm to do that. Anyway, I swear the left and right
channels were out of phase. On mono! Ok, how about an old Bee Gees. Only got
it with 192k compression. Ok, let's try it. Not only was it muddy and
lifeless, but it had a tic from a scratch on LP. This was somebody's transfer
from an LP! And a lousy one at that. Is this what I can expect for download?
If so, I may as well go out and buy the entire CD. At least they sound decent.
Seems to me iTunes in nothing more than a way to categories and shuffle your
existing CD collection on a hard drive. The downloads are ok for iPod or kids.
Well, at least I got a Mac out of it. Some great hardware. When I do another
show, it'll be with my IBM laptop and Foobar2000. And hopefully a Galibier.
January 18, 2006
Got some more DAC boards built. Another CHIME off to reviewer. One customer
suggested doing auctions on audiogon. Other places do this as advertising.
And they sell the things too. He mentioned there aren't any other reasonably
priced tube USB DACs out there (nothing used!) other than the BRICK. So I'll
get to that. Can put the review units up for discount. If they ever make it
back. Rewrote some of my web pages to de-emphasis the DIY aspect. I think
that scares some customers off. Meanwhile, thinking of doing a T-shirt for
sale. Something along the line of "DIYers do it cheaper". Need some clever
slogan you guys would actually wear.
Work on the COMPRESSOR continues. I was stuck on the attack/release circuit.
Tried to do exponential response, but then realized it's not important, as the
non-linearity of the vactrol makes everything happen in dB. Added vactrol to
my existing sidechain circuit so I could have a listen. Ran some music through
it. Could hear the effects of attenuation, even though circuit response was a
bit of a mess. Anyway, the compressor action was nice and soft, just the way
I had hoped. And this was without my special gamma circuit.
Compressor transfer function
So decided to go back to my original LM13700 based A/R circuit. Built it.
Wow, this thing worked perfect! Great response and the waveforms were very
clean. It ran faster than I hoped, but that's just a matter of R and C values.
Very simple circuit. Can move forward again. Now I just have to figure out if
I need my gamma circuit to drive LED, or go with a straight linear drive.
Anyway, the A/R response tracked a voice really well. Super fast attack times
(limited by vactrol) if I want. Will work even better when driven with an
absolute value circuit.
Sidechain attack/release circuit
Agonizing moment this morning. Just had some inductors shipped in to finish
some HAGDACs, only to discover them missing. Crap! Turns out the dogs got
ahold of them and chewed up the strip (surface mount tape), wife had swept what
was left into the trash. Sheesh. Was able to save half of them. Similarly,
I have a number of HAGUSBs on hold until more parts arrive tonight. Sorry for
the delay! Didn't expect to sell so many assembled ones versus kits.
January 17, 2006
Plenty of success today. And last night. Got my IBM laptop configured with
Foobar2000 and ASIO to play songs from HD. And it sounds friggin great! See
my forum post for instructions. Had to jump through a lot of hoops. I really
like the simplicity and power of Foobar. Same sort of synergy and elegance I
try to bring to my products. It is some really good code. I also was able to
buy some hi-res downloads from allofmp3.com. I recommend getting the FLAC.
Works perfect in Foobar and with the dBpowerAMP software (another nice piece of
code) you can convert it to AIFF, which imports nicely into iTunes. Once in,
you can convert to Apple Lossless. I will do some sound quality testing
tomorrow, to compare the shit AAC to the uncompressed.
Also working on the RIPPER mechanics. Want to simplify this machine. Make it
small and cheap, without compromising sound. Cost is high priority. Thinking
I can use an analog multiplexer to select output instead of a switch. Have it
automatic. When USB audio data comes down the pipe the DAC output is selected.
Otherwise it defaults to the analog from phono stage. That gets me down to
just one knob, for the gain. No power switch. Heck you leave your PC on all
the time, right? Only a watt or two drain. Anyway, the gain knob is to set
maximum full scale from phono into the ADC. I will add a clipping LED that
should make this pretty easy. Just play your loudest track and set gain as
high as you can without clipping. Simple. I think I can sell a lot of these.
It offers a perfect way to rip LPs to CD quality 16 bits. It also has a built
in playback DAC with USB connection (like a CHIME) or an optional S/PDIF
output. A merging of old and new. It will also gain me the knowledge needed
regarding USB recording, so I can do the super-ADC for haglabs. This is one
step on a long path.
Customer mentioned today about using a Zire 72 with bluetooth wireless with a
program called clicker [?] that can control iTunes from your seat. And not just
play/stop. It loads the entire song list! You can control iTunes with a
pocket PC or whatever. Neat. So I was thinking, maybe for my haglabs remote
control, it might be nice to add some hooks or extensions in the hardware to
offer similar capability. Maybe via IrDa? I don't know yet. I was thinking
serial port for PC communication, but IR to a Zire or pocket PC would be a
cool thing. More functionality than my regular IR remote. Looking into the
future, control of audio hardware and iTunes or Foobar server at same time
would be nice. A little integration lets you sit on your keester.
January 13, 2006
So I hooked up the iMac to my audio system. Had to go into system preferences
and actively select the USB audio device for output, instead of the built-in
speakers. Worked fine. CHIME locked up perfect. Put in an audio CD and
iTunes played it nicely. Sound was virtually indistinguishable from my other
transport. But I need to do better A/B comparison to confirm. Anyway, pulled
up the song I bought and played that. Oh my, it sounds like absolute crap!
What a piece of garbage, and I love this song. Flat, stale, no focus, and
compressed to the max. What gives? I had the iTunes "import" set to Apple
Lossless. Hmmm. Dug out the properties of the song file, AAC 128kbps. File
name was ???.m4p. No wonder! Must have bought the wrong thing. Now I gotta
figure out how to get the right file. Maybe I should chew up disk space with a
.wav? Don't have an airport card for yet, so I have to haul it back over to
the office for downloads.
Just thought you might want to know, whatever codec that was, it is
absolutely unlistenable. On the other hand, it sounded amazing for a 10:1
Also noticed my system makes a great reliability test. The dust on my rack is
so thick you wouldn't believe it. Must be a ton of crap inside the units
(CHIME, CYMBAL, etc.).
Meanwhile, working on making your life easier. Adding pre-downloaded parts
lists at digikey. That way, yoou don't have to type in every part. Just go
to the specified link and use that as a starting point. Add or delete. Should
save you time. My first up is for the HAGUSB.
HAGUSB parts list
January 12, 2006
Wow, what a great experience I just had. My iMac came in, so busy it sat in
the box for two days. Last night I unpacked it. Holy moly, it was like
unwrapping a brand new computer. I mean this thing was packed in original
carton, all documentation, boxes, bubble wrap. Unbelievable. And this thing
was used! I didn't want to take a chance with a 'bargain' machine (many go for
under $200 on eBay), so I forked over an extra hundred clams on a machine from
original owner, already upgraded to Tiger. This thing is mint. Glad I didn't
take a chance on an unknown unit from a reseller. This cost me more, but it
was plug and play out of the box. I fired it up, spanking new reload of OS-X,
it downloaded updates from net, and that was it. Not one glitch. And to my
suprise it plays DVDs, too! Shoot, I'd switch over to Macs in a heartbeat if I
didn't have to run all these engineering CAD packages (layout, schematic).
So I got myself an iTunes account. Bought my first song - Boulevard of Dreams,
by Green Day. Can't wait to crank that up. So far so good. Next up is trying
to figure out how to record. Maybe I need GaragBand? I dunno. Decided to
officially turn my BUGLE2 into THE RIPPER. I want to make this leap into
digital such that I can record vinyl. iTunes is sooo handy. There will be
times I just wanna random shuffle songs instead of putting a disc on the
platter. Oh no, I will never give up on vinyl. My plan is to extend it.
January 10, 2006
Had a ball today. Continue to ignore my consulting duties so I can concentrate
on the design of the compressor. Breaking new ground with this machine. I am
convinced it will be a hit. At least with the pro market. Spinning my wheels
in circles trying to get things to fit. I had to lose the high pass filter
adjustment to make room. Spent hours moving things back and forth. A half
inch this way, move that back up, etc. That's the critical thing with system
design, you get to make compromises back and forth between mechanics, layout,
schematic, and functionality. Doesn't fit? Change the spec! Or find another
part. It takes a lot of continual grinding with one option after another to
finally converge on a solution. I can see the design inching it's way to
reality. Here's a shot of where it's at now, both front and back.
Got the schematic done. Will buy parts tomorrow so I can work on actual board
layout. Again, will probably go in circles a few more times until everything
works out. Front panel is starting to look a little sparse, but it is packed
inside. Trying to really simplify operation and give it a really gentle touch.
Meanwhile, a late entry to the logo contest had a cool concept: make a vacuum
tube image out of the letters HAL. It inspired me to come up with a CAD
version. At least it's a good start. Might scrap it later, but I like it for
January 6, 2006
Spent a lot more time on compressor schematic today. The sandwich construction
is tight, so I have to be carefule which components I chose. Only 1.5" front
to back! Have to use axial leaded electrolytics. Thought I had a nice tranny
picked out, an Allied. Then I remembered I needed a 48V phantom supply! So
went back to the old 370BX. Besides, I got plenty in stock. The voltages are
just right for what I want to do. Came up with this neat trick using the 5V
winding. I stack it on top of the 6V secondary to get +/-10V rails with half
wave rectification. Fine at low current. Use the 6V winding for ac heating on
the 6X4 and ECC99, then do rectification a la cornet to make dc 6V for 12AX7
heater. Efficient and low noise at same time.
Then trying to make a pot control for the attack and release times. Just
couldn't cover the range nicely. Have to switch to a 4-pole switch and just
use a few discrete settings. So, like I said, spent a lot of time thinking
about the circuits and how they operate with real signals. Had this really
clever implementation for attack & release control. It's a charge pump type
that has linear ramp up/down to follow the rectified signal, similar to the
THAT sidechain, but with 1/5th the number of parts. I was proud of that piece,
but decided it might not be right. This is the hard part - how does it sound?
Without building it. A nice linear control, but I'm thinking maybe not very
musical or transparent. It has same linear charge up or down independent of
signal level. That doesn't seem right. So I changed to an exponential RC
style charge up/down. But not really down. More of an average envelope
detector combined with a peak detector for catching fast transients. It should
be a more natural sounding and less 'digital' in terms on compression control.
Anway, the pencil schematics are done. Mechanics layed out. Now it's time to
do schematics on CAD and start the component placement for layout. Big board,
16.75" x 5". Parts on both sides, overlapping.
Wonder how CES is going.
January 5, 2006
Put the BUGLE2 into a chassis. Fit is tight. With wood sides, it looks pretty
good. And it's really small. Blue LED on front looks cool. So I burned it in
using the lab frykleaner. Put on an LP this morning. Holy cow, this thing
sounds good. Quieter and clearer than a regular BUGLE. What the heck? Then I
realized it must be the super regulators. Everything else pretty much the
same, except for choice of resistors on input stage to make it lower noise.
But I have this new idea to make the RIPPER. Or should that be USB jack the
ripper? Add a USB ADC to the BUGLE2 and make it into a vinyl to CD ripper. Or
just to hard drive. That would be pretty cool. Have both a USB DAC and phono
stage and ADC all in one tiny little package. Ok, the DAC isn't the best, but
it's there for free. Anyway. I think there would be a nice market for a DIY
product that made it real easy to record LPs to your computer. Meanwhile, I
have 200+ BUGLE boards in stock, so there is time to think about this.
Ok, here's what I'm thinking about (see block diagram). Phono based on the
BUGLE2, add a line input for good measure. Needs a gain control to properly set
levels for ADC, monitor with a clipping LED. ADC goes to USB at 44.1k 16 bits
for rip. Can also playback digital. Output is selected between analog or
digital feeds, so you can use as a phono pre. Add headphone output? Could
simplify by removing line in and headphone. Don't really need optional
digital input either. Line input would be handy for recording radio.
RIPPER block diagram
My laptop came in to drive the CNC machine. iMac is enroute. Need to get my
iTunes up and running! Really interested in easy ways to do the recording.
Does iTunes record? Or do I need other software? Must get this figured out
so I can deliver a simple solution.
Also busy doing the mechanics of my compressor. Whoa, this baby is tight. If
I can get it to fit, what a breakthrough. Way cool sandwich. The vactrol is
available from Allied, so I then discover they still carry a line of power
trannys for tube circuits. And they're cheap! So that is a bonus. Only
expensive items in this design are the Lundahl I/O trannies. But hey, it's
gotta have great sound, or what's the point. All new tube design, not copied
from anyone. Same with the side chain. This machine will open up a new
segment of DIYers that hasn't been touched yet. Also gets me more in touch
with the recording side of the industry. I think my haglabs products will be
sort of crossover type components, having pieces from both pro and high-end