Thumb Sucker Compressor distorts instead

[puretube]

see eh la bas ma`s latest thread ...
https://www.diystompboxes.com/smfforum/index.php?topic=129237.new#new

[matopotato]

see eh la bas ma`s latest thread ...
https://www.diystompboxes.com/smfforum/index.php?topic=129237.new#new

Thanks, very interesting. Will try

[matopotato]

see eh la bas ma`s latest thread ...
https://www.diystompboxes.com/smfforum/index.php?topic=129237.new#new

Ok, disconnected pin 8 and 9 to ground so that they are unconnected to anything.
No change in the distortion. Tried on off to be sure and it had no effect on the output as far as I can tell.
Even joined them to eachother but not to ground.
Still closer to a Fuzz than a Compressor sadly. Cleans up with guitar volume even...

[eh la bas ma]

Then little or no point getting another kit.

I agree...

Edit: did you have the settimgs as in the soundclip picture throughout?
Mine might be more obvious on the bass strings, or full chord.

Maybe that's because your pickups are closer to the bass strings or something...

The soundclip is a short loop, with 3 repeats :

first, clean guitar.

Then, settings as shown on the picture.

Finally, Theshold fully CCW, Ratio fully CW, Release fully CCW . These are the settings with the worse distorted signal possible with my build.

[matopotato]

Then little or no point getting another kit.

I agree...

Edit: did you have the settimgs as in the soundclip picture throughout?
Mine might be more obvious on the bass strings, or full chord.

Maybe that's because your pickups are closer to the bass strings or something...

The soundclip is a short loop, with 3 repeats :

first, clean guitar.

Then, settings as shown on the picture.

Finally, Theshold fully CCW, Ratio fully CW, Release fully CCW . These are the settings with the worse distorted signal possible with my build.

Thanks.
I was told that the fully dimed extremes might cause distortion, so I have had mine on
Attack and Release, close to CCW but not fully in the bottom.
Ratio from noon to 3 o'clock or almost full CW, Distortion all the way now, before less at noon and most at full, but now seems bad regardless
Threshold first between noon and 9 o'clock and at 9 bad enough in combination with Ratio. Nowdays the Threshold seems to have less impact.
Volume seems to be doing its thing and would not say it has an impact.

[matopotato]

@duck_arse:

Some attempt at updated reading with breadboarded using only TL072 x 2 and LM13700 from original build.
Plus one B1M pot, but for different circuit location (i.e. no longer prime suspect)

IC1
4,65   8,83
4,42   4,40
4,02   4,27
0,00   4,20


IC2 got some variations, this was the latest read
4,30  8,84
4,46  2,69
4,0    2,02
0       --- falling

IC3 (LM13700)
1,09   0,00
0,3     0,16
4,43   1,3
4,43   0,20
4,43   0,20
0        8,84
0        0
0        0
This was before degrounding them

PNP
C: 1,13
B: 4,8
E: 4,45

[DrAlx]

If you want to try it on your PCB version, lift one end of the 10k and connect it in series with a pot. Tweak and see. If it works, take the pot out of circuit then measure it and add a resistor of the pot value or next value up.

I've a hunch that (if this is the problem) 22k instead of the 10k may be the charm, but swapping resistors to find perfection soon gets tedious so a 50k or 100k pot should give enough range to try.

Thanks, it was an interesting thing to try.
I put the 10k back on the board an added a 50k trimmer together with it. So I could dial 9.7k to 59k (ish). Sadly there was no change in the distorted sound. It gave an overall volume boost though.
Even though I have tried similar idea in other builds, it just did not strike me until you suggested it. I hope to learn by now using trimmers and pots in areas where a changed R could make a difference.

You tried changing the voltage divider ratio at the OTA inputs from 220R : 10k  to  220R : 59k  and managed to get a volume boost.

Just imagine what taking that ratio to its limit  (for example by shorting out the 220R) would do.
The OTA would then have no input signal (because the two inputs would be shorted together) and so the OTA would give no output current.
In other words it would look like an infinite feedback resistance in the opamps loop, so the gain would be massive and give distortion.

The same thing goes for my (not well thought out) suggestion of trying to increase the resistance on the PNP emitter.  The bigger you make that emitter resistance, the lower the OTA output current will be, so the OTA will act like a bigger resistor in the opamp's feedback loop, and so give distortion.

If the distortion we are trying to solve is due to the OTA looking like too big a resistor (due to it giving out too little current) then I would

1) Leave the 1k resistor on the PNP emitter alone.  Certainly do not lower it or you could kill the OTA if you are not careful.
2) Try to decrease the 10k resistor in the divider at the OTA input, not increase it as you tried above. Easiest way to experiment is to add resistance  in parallel to it. Try add 47k in parallel to the 10k (which effecively turns the 10k into an 8k2). If 47k does not help much try 22k instead.

[DrAlx]

If the distortion problem is due to the OTA acting like an infinite resistor (i.e. sometimes giving zero output current) and this happens while strumming hard (i.e. despite the input control current to the OTA being large), then the only way I can see for that to happen is that the OTA input signal is zero. That's because 
          OTA output current = OTA input signal * Transconductance determined by control current

The opamp responsible for the circuit gain might have a DC output voltage that is not exactly at the bias voltage (due to opamp imperfections), but the scaling resistors at the OTA input are referenced to the bias voltage.  I am wondering if such a mismatch would be a problem ?

You can imagine an extreme case when the difference between those two DC voltages is large.
So a sinusoidal signal at the opamp output would get scaled down to give an OTA input signal with a non-zero offset.
In an extreme case, you could imagine that the OTA input signal (i.e. the voltage between the two OTA inputs)
looks like a sinusoid that is always negative and just touches zero at its peaks, rather than being a sinusoid that oscillates about zero.
In that extreme case, the OTA would act like an infinite resistor (i.e. give no output current) for the very top of the sinusoid where it touches zero, but at the bottom of the sinusoidal it would be acting like a smaller resistor (and give output current).
So the top and bottom of the sinusoid would be treated differently.
For the OTA to act like a resistor, I think we'd want the top and bottom of the sinusoid to be treated in the same way (i.e.  produce equal and opposite OTA output currents).  They would be treated in the same way if there was no mismatch in the DC levels.

Maybe this simplistic thinking is not correct and mismatch in DC levels is not a problem?

[DrAlx]

So if DC offset could be a problem (and I dont know if it is) then maybe block the DC ?

What would happen if instead of tying the opamp output direct to the scaling resistors at the OTA input, a large capacitor (many uF) is used there instead to let the signal through but block any mismatch in DC.

If we think of OTA inputs like they were opamp inputs, then we only need one of the inputs to set the input bias voltage.  Not both of them.  Just a thought.


matopotato:  Try this.  The 10k resistor that is part of the 220R : 10k scaling resistor section at the OTA input.  Disconnect that 10k resistor from connecting directly to the opamp output, and instead connect it to the opamp output through a 10uF capacitor.   Before disconnecting anything though, measure the voltage at the opamp output and see if it is larger or smaller than the bias voltage.  The capacitor will probably be electrolytic and you want to orient it so that its +ve end is connected to the higher voltage.

[matopotato]

Thanks a lot @DrAlx,
I will try the 10k with 47k in //.

And also the 10k plus a 10uF cap. If I first measure voltage on the OTA connecting pin and then the other side of 10k, and set minus of 10uF to the lower potential, the is "safe"?

I DO read what you wrote, but have to confess that my electronics understandings are quite limited.
Both from functional aspects (but learning every day) as well as physics/theoretical aspects (started reading in semiconductors though, so given enough time..)
I really appreciate the effort you put in.
Again thanks.

[merlinb]

Write the voltages on the circuit diagram
You have given us conflicting voltages on the same wire. How can IC1 pins 6 & 7 give different readings when they are shorted?
IC2 pins 6 & 7 are way low, they should be at Vref or higher when idle. Either IC2 pins are connected wrong or something in the side chain is incorrect (backward diode?).

[Kevin Mitchell]

Conflicting voltages or unstable power supply?
Hmmm....

[duck_arse]

IC2 got some variations, this was the latest read
4,30  8,84
4,46  2,69
4,0    2,02
0       --- falling

pin 5 falling would indicate the rectifier had input signal applied. 2V69 at pin 7 would mean "compressing". but not while you have 4 something volts that should be coming from pin 1. I think.

[edit :] deleted - see reply #69.

[DrAlx]

If I first measure voltage on the OTA connecting pin and then the other side of 10k, and set minus of 10uF to the lower potential, the is "safe"?

Correct. I would try that capacitor change on its own, leaving the resistor at a value of 10k.
EDIT:  If the voltages written on the circuit diagram are correct, the + side of the 10uF cap should connect to the opamp output, and the - side of the 10uF cap should connect to R7.

I would try to experiment with just one circuit change at a time, especially since most people do not have a problem with their builds of this circuit.  I have successfully built the ET three times in the past (versions 1,2 and 4) and never had a problem (except for the version 4 which had a minor problem if an LM13600 was used instead of a LM13700 due to how the output buffers work).

[matopotato]

If I first measure voltage on the OTA connecting pin and then the other side of 10k, and set minus of 10uF to the lower potential, the is "safe"?

Correct. I would try that capacitor change on its own, leaving the resistor at a value of 10k.
EDIT:  If the voltages written on the circuit diagram are correct, the + side of the 10uF cap should connect to the opamp output, and the - side of the 10uF cap should connect to R7.

I would try to experiment with just one circuit change at a time, especially since most people do not have a problem with their builds of this circuit.  I have successfully built the ET three times in the past (versions 1,2 and 4) and never had a problem (except for the version 4 which had a minor problem if an LM13600 was used instead of a LM13700 due to how the output buffers work).

Thanks.
Yes, I have the same reflection.
I tried the 47k parallel, no change
And 10uF in series. It was 4.6 something at the junction with 220R, and 5.7 something on the other end, so I put minus closest to the OTA. No change.
Took a leap of fait and swapped, no change.
Sorry.

[matopotato]

A reflection: I have been trying so many things by now, and lately it got all bassy and muddled on me. Some less distortion. Seems the breadboarding is coming apart from all the tearing and swapping. Might have to redo that.
After some re-grounded pins I had forgotten left out, it got back to its "normal" scrape-distortiony self.
In the trial I sometimes forgot to disconnect etc, and I notice that I can remove or cut off several things and there is no change to the distortion.
So rebuild might be in order.

[matopotato]

Write the voltages on the circuit diagram
You have given us conflicting voltages on the same wire. How can IC1 pins 6 & 7 give different readings when they are shorted?
IC2 pins 6 & 7 are way low, they should be at Vref or higher when idle. Either IC2 pins are connected wrong or something in the side chain is incorrect (backward diode?).

There is no reason to shout. I can still read....
I can understand you are frustrated that some of us posting here do not follow praxis. And as per the picture above I was told in another thread to post it the way I have in my two reading sessions in this thread. Which leads me to think that there are more than one ways to present readings.
So sorry about that, I was following what I thought was the norm.

Yes, how can they give different values on pins 6 and 7. I took a DMM and place the red one on pin 6, read it, then pin 7 read it while have black to a common ground. I wrote it down.
I realize I shouldn't have. I should have gone on until they were the same. Until reality decided to confine to theory. But I thought "That's odd, but I will not judge at this point. If I am asked to re-measure I will".
But Now I think I should re-breadboard it first.
And then I can re-measure. And present in two ways for the greater benefit, one on the schema and one in table.

[matopotato]

Conflicting voltages or unstable power supply?
Hmmm....

I don't know. I use a Truetone 1 Spot CS12 or CS6. And I feel it has been stable. Then I have used the wall wart I got from some pedal purchase in some cases, one from EHS and another small I think from Source Audio.
Is it strange? Yes.
I will rebuild and then re-measure again. And hope to be brave enough to present whatever readings I get.
Got me thinking of
"...then whatever remains, however improbable, must be the truth"
But if all this gets to "said and done" most likely there is some error on my side.

[matopotato]

IC2 got some variations, this was the latest read
4,30  8,84
4,46  2,69
4,0    2,02
0       --- falling

pin 5 falling would indicate the rectifier had input signal applied. 2V69 at pin 7 would mean "compressing". but not while you have 4 something volts that should be coming from pin 1. I think.

Thanks. I take this as another vote for "Re-build. Re-measure" (Title of a Roxy Music album that never got released...)
I hope you liked the pictures though, despite amateur quality shots.

[matopotato]

Write the voltages on the circuit diagram
You have given us conflicting voltages on the same wire. How can IC1 pins 6 & 7 give different readings when they are shorted?
IC2 pins 6 & 7 are way low, they should be at Vref or higher when idle. Either IC2 pins are connected wrong or something in the side chain is incorrect (backward diode?).

(I still appreciate that you respond. Thanks)