Dr Quack squealing sound

[differo]

Hi, I built a Dr Q (using Tayda pcb) and it works but on the treble side of the sweep i get these really unpleasant squealing sounds. I tried different transistors, OP amps but the problem remains. The sound is similar to amp feedback sound. Its more prominent when both sensitivity is increased and even range trim pot. Don't know if its a coincidence but I believe its a bit better with 1458 opamp instead of 072 or 4558. Any ideas?

[antonis]

It seems you need a couple of feedback caps..

If you post the schematic we might suggest you proper caps values..

[differo]

I believe this is the Jack Ormans version:

https://www.taydakits.com/instructions/dr-q-autowah/pages/designators-and-components--44

I double/triple checked all the solder joints, values for components etc. It sounds worst with TL072 (I  mean this undesired effect is most pronounced then).

[antonis]

OK..

Try, at first, a 10 to 22 pF across R5 ..

[differo]

No difference with cap over r5. :/

[Rob Strand]

No difference with cap over r5. :/

You might be able to add a resistor in series with C6 (1nF).   Maybe try 220 ohm to 470 ohm.  If that doesn't work try 100 to 2k2.  Without analyzing the circuit in detail it's hard to come up with a precise value.    It's never a good idea to put caps across opamp inputs.

[amptramp]

You may benefit from a cap across R13.  The use of C6 and C7 allows AC signal to be shunted across the op amp inputs resulting in reduced feedback as the frequencies get higher.  A resistor in series with C6 may solve the oscillation but may change the operation of the circuit.

[antonis]

Ron, isn't C7 just a humble +V/2 stabilizer..??

Also, U1B gain is always less than unity (for both SW settings, either Q1 1/g_m in series with R12 or R11 in series with R12) so I don't think a cap across R13 could help..
(on the other hand, it could do no harm..)

If we like to focus on feedback caps, I'd propose a Miller one between Q2 C-B..

[Mark Hammer]

1) Do we know that the JFET is installed right?  I never fail to be surprised by the pinouts of JFETs I wasn't expecting.  I don't know that it would be responsible for squealing, but let's confirm it's in there properly.

2) You can play with the attack and decay times by varying the values of R8 and C3.  However, increasing R8 to lengthen the Attack time will eat up drive current to Q2, so the standard amendment to the circuit is to increase the value of R5 to 2M7 or even 3M3 for more drive.

Half-wave rectifiers like the DQ uses, are prone to more audible envelope ripple, that can sound like "distortion" under some circumstances.  It does so less when the time constants (attack and decay) are either short enough for things to pass quickly so as to be unnoticeable, or long enough to smudge/blur the envelope.

There are limits to how much the Attack time (R8) can be varied, and the modest practical differences may not justify using a pot.  I'll suggest use of a "default" value of 390R for R8, and a SPDT on-off-on toggle to add two other resistors in parallel for smaller values and faster attack.  With added parallel resistors of 180R and 56R, you end up with parallel effective values of 123R and 49R, which will give you faster, close to stock, and slower than stock attack.  Swap the 10uf C3 for 15-22uf, but wire up another 3-way toggle to place two different resistors to ground in parallel with it.  Middle toggle provides no parallel resistance, but side positions place either 33k or 100k in parallel with C3. That will provide fast, medium, and slow decay, as they bleed off the charge stored in the cap, or leave it to discharge on its own.

[m4268588]

Why does C6 exist?

[differo]

Thanks for all suggestions, I'll tinker with it and try different values. I can hear a little bit of warble/distortion/ripple in the end but that's fine I guess for the design of this circuit. However this squailing is really weird. I built it stock and I could not a single report with similar behavior. What's the easiest way to upload a sound sample?

[Rob Strand]

Why does C6 exist?

Only in the link posted by the author in post #2.

[amptramp]

Ron, isn't C7 just a humble +V/2 stabilizer..??

Also, U1B gain is always less than unity (for both SW settings, either Q1 1/g_m in series with R12 or R11 in series with R12) so I don't think a cap across R13 could help..
(on the other hand, it could do no harm..)

If we like to focus on feedback caps, I'd propose a Miller one between Q2 C-B..

Yes, that's why I only proposed a resistor in series with C6.  If there is any feedback in the circuit ground, C7 may be involved.

[Rob Strand]

Thanks for all suggestions, I'll tinker with it and try different values. I can hear a little bit of warble/distortion/ripple in the end but that's fine I guess for the design of this circuit. However this squailing is really weird. I built it stock and I could not a single report with similar behavior. What's the easiest way to upload a sound sample?

Some people put up youtube videos.

You could also try adding a 100uF across the power rails.  Maybe some signal is feeding back through the buffer via the power rails.

I'm assuming it occurs in both modes?

[Rob Strand]

Something else, the sweep has the potential to push frequencies way into the high frequencies.
When it does that there's an enormous gain which could promote oscillation.

The basic way of limiting that would be to tweak the 22k resistor on the base of the transistor.

The thing then is it might not sound right (maybe it will, I don't know).   So a way around that would
be to say put a zener across the range pot (pins 1 and 3, or pins 1 and 2) maybe something like 4.7V to
start.   There's plenty of other ways to mod it.   You might need to tweak the 22k resistor for best sweep,
then after  that set the zener voltage to prevent weird effect with range maxed out.

Another trick would be to add a small resistor in series with the collector but I recommend against it
as it will affect the nature of the sweep.  150 ohm might be OK but if you need 470 ohm then maybe
try the zener.
What's interesting is the Nurse Quacky variant adds 220 ohm in series with the emitter.  (adding in series with the collector is better)
https://www.home-wrecker.com/nurse-quacky.html