U-Stomp Dynacomp layout confirmed

[Mark Hammer]

I had the PCB transfer printed up and sitting around, and had a spare corner of copper board which fit it perfectly, so I etched and populated it yesterday, and I can happily confirm that it works.  I used a plastic CA3080A because the metal can ones were going to be an awkward fit.  No particular selecting of components.  Used an MPSA18 for Q1/Q2, a "matched" pair of 2N3904 (hfe 142 and 144) for Q3/Q4, and a 2N5089 for Q5.  Nice and smooth, and reasonably quiet when fed with a humbucked signal, despite not being in a case and being tried out in a room with fluorescent lights.  More than hot enough output to yield clipping in the amp.  With a 3-position recovery switch and a treble-boost switch it has lots of variety in tones attainable.

Note that this is a Dynacomp layout, not a Dynacomp-to-Ross "convertible" layout.  So, it's a tad smaller than the Tonepad or some other layouts, because it doesn't have room for the extra parts.  And like I say, I made it because the copper board remnant was just the right size.  Shoudl the anticipated chassis you are going to use have tight space requrements, consider this as an alternative.

Thank you Steve Rittmeyer.    http://ustomp.com/index.html

http://ustomp.com/files/692b970f-0000-00f4-2343-67f6181bbebb.gif

[markm]

Wowsers!
I thought everyone knew this was a verified layout.
I made my Dyna-Clone from it and I must say the layout is superb!
Nice job! 

[Mark Hammer]

It may have been common knowledge for some, but I just thought it was useful to let those who were previously unaware know that they could safely build this and expect a well-functioning circuit at the end of their efforts.

I have to say that what I found particularly pleasing about Steve's layout is that I could very easily install plastic caps with just a little extra lead (i.e., not flush with the board), and bend them over so that they lay flat over top of resistors.  That resulted in a very low-profile board that will easily accommodate big pots in a 1590B, if that's the route you go.  This sort of convenience is not always available in all layouts, but it was an absolute breeze with this one.

[343 Salty Beans]

I have to say that what I found particularly pleasing about Steve's layout is that I could very easily install plastic caps with just a little extra lead (i.e., not flush with the board), and bend them over so that they lay flat over top of resistors.  That resulted in a very low-profile board that will easily accommodate big pots in a 1590B, if that's the route you go.  This sort of convenience is not always available in all layouts, but it was an absolute breeze with this one.

You're lucky.

Somehow, I end up not having one electrolytic cap value except in the "ridiculously large" voltage. So I cram that on the board. Then, when I mount the project, the enormous cap is just in the right spot so that I can't mount a pot, and I end up doing some retarded-looking pot mounting so I can fit my frakking board in the frakking box.

[Mark Hammer]

If there are places near you that sell old computer cards, see if you can score one of those old 8-bit or 16-bit Soundblaster cards for a couple of bucks.  Those things are often a treasure trove of very small package caps in the 1-47uf range, virtually all at 16v rating or higher.  Unsoldering can sometimes be a pain but judicious use of a heat gun can often deliver a big payoff of lots and lots of caps for modest cost.  The caveat is that the lead spacing on those caps is occasionally a little too narrow relative to the board layout used.  In this instance, though, I had no problems.

[343 Salty Beans]

that's kinda nice I should try that. And I have an old computer in the garage I've been going to when I (very rarely) need SMD parts, but I'm willing to bet there's some decent stores of analog parts in there.

I order from futurlec, and they don't always have 16v or even 35v available. I have a few 50v electrolytics that cramp my style...and my enclosure.

The only problems with desoldering and reusing caps is the legs never turn out to be long enough a lot of the time...makes it a real pain in the butt to solder. The other problem is that sometimes they don't work

[Mark Hammer]

The only problems with desoldering and reusing caps is the legs never turn out to be long enough a lot of the time...makes it a real pain in the butt to solder. The other problem is that sometimes they don't work

Both of those are true.  But sometimes you get lucky, and if the stuff is just going to get rendered for the gold on the contacts anyways, better to recycle than make more garbage in the world, right?

[343 Salty Beans]

Both of those are true.  But sometimes you get lucky, and if the stuff is just going to get rendered for the gold on the contacts anyways, better to recycle than make more garbage in the world, right?

Eh, it's not me that has to deal with it in the future 

I kid, I kid. Thanks for the tip, I'll check a few comp stores out.

[hilbi]

a "matched" pair of 2N3904 (hfe 142 and 144) for Q3/Q4,

I recently finished a ross clone, but I didn't know that Q3 and Q4 needed to be matched, can anyone explain why it is imported to match those?

[Mark Hammer]

a "matched" pair of 2N3904 (hfe 142 and 144) for Q3/Q4,

I recently finished a ross clone, but I didn't know that Q3 and Q4 needed to be matched, can anyone explain why it is imported to match those?

It's not anything you need to drive yourself crazy about.  Plenty of people have made completely satisfactory units without any attention to hfe or matching whatsoever.

Q3 and Q4 each form part of separate half-wave rectifier circuits.  Each of those diode/cap/resistor combinations in that part of the circuit do the same thing that you see on things like the Dr Q/Nurse Quacky to half of the input waveform.  The transistors help to buffer each HWR so that they can be added together without interacting.  Smoother control of the gain, however, is provided by using a full-wave rectifier, like the Dynacomp circuit does (and a dozen other similar commercial compressors do).  The smoothness of that control is a function of the extent to which the two separate HWR signals are balanced and roughly equivalent.

I guess one analogy you could use is that of two opera singers warming up singing alternating notes, going "Mi, mi, mi, mi, miiiiiiiii", except that that each one does an alternate "Mi".  If they are in perfect tune, then it sounds harmonious and stable.  But if one is a bit flat or sharp, the alternation sounds a bit grating.  The balancing of the two HWR portions attempts to avoid that bit-more-bit-less-bit-more-bit-less thing.  In the world of envelope/sidechain-controlled things (which a compressor is) that slight wavering of the control signal can sometimes be perceived as distortion.

The reason why you shouldn't drive yourself crazy about it is that:
a) it assumes identical amplitude copies of the audio signal coming off the emitter and collector of Q2 for rectifying (true in an ideal world but maybe not so much in the real world) and
b) it assumes identical forward voltage of the diodes and identical-value resistors (1M with 5% tolerance) and caps,

...such that when the two rectified half-waves are combined they are "in tune" with each other like our opera singers.

It IS possible to adjust the values of the 10k resistors on the emitter and collector of the 2nd transistor so that the signals found at the emitter and collector are absolutely identical in amplitude, and it IS possible to select the diodes, 1M resistors, and .01uf caps so that they are very tightly matched.  But that's a whole lotta work for something that really only shows a consequence during the chords you let ring for a while (envelope ripple is most noticeable during the decay phase).  If you are a fan of very quick recovery times and fast chicken-pickin, super-matching also won't show up all that much in terms of audible consequences.

All of that being said, it is probably a good idea to at least take a peek at the transistors being used just to assure that they aren't behaving in a dramatically different manner.  A decent high-gain, low noise transistor is probably helpful for the input transistor, and maybe for the output transistor (the one immediately after the 3080), but everything else can be plain old garden-variety 2N3904s as far as I'm concerned, since they are not really being used to supply an audio signal for listening purposes, but rather for providing a control signal.  I picked a couple of 2N3904s from my bin, measured them and found them to be very similar and plunked them into the circuit.  Your selection doesn't need to be as tight.  Just don't stick something with an hfe of 800 in one slot and something with an hfe of 50 in the other.

[hilbi]

ok thanks mark!

I used 2n5088's in all positions but didn't measure them, i will measure them and check just to sure.

[hilbi]

I Checked my Q3 and Q4, they have a hfe of 412 and 420, I think that is pretty close so i stuck with them.

Thanks again for the info!

[puretube]

err, what is that "TIE" component in the lower left corner?
(connects from input to gnd, it seems)


(sorry - dunno the schemo)

a pulldown-resistor?

[markm]

It's a pulldown resistor.
I used a 1M there and it was fine.

[ustompsteve]

yes, it's a pull-down resistor. For some reason I refer to those as tie-down resistors.

--steve