Weird tone stack issue in a known good design

[karbomusic]

This is build #15 so as far as I know it isn't the design. Basically a modified TS tone stack. I was just finishing up 3 builds and on the third when the tone knob hits 9.9 on a scale of 10 going CW (full treble), it suddenly drops 8dB and the tone goes to what appears to be the same as fully CCW (as per my scope). Other than that ".1" threshold and behavior, everything else in the entire build appears to function exactly correctly.

I've changed the pot, C7, C6 and the opamp. Resistors appear correct but have not pulled those yet. Any ideas what could cause this? I cannot find any bad solder joints or bridges on the PCB. This is driving me nutty...  

The output of the TLC2272 does hit another opamp which is a buffered/blend but I can't find a thing wrong with that either. When probing the output of the tone stack (with the buffer chip removed), it still does it. It's as if I'm hitting some threshold then dying. Since it works perfectly in 14 other builds, I can't help but think it is a bad value, part or soldering but I can't find it. Thusly, if any of you guys are familiar with the symptom, I'll take that back, armed with more than guesses.

[PRR]

Reading the DC on the output of that opamp will tell you if a cap (or its PCB) has failed short and is blowing-up the opamp bias.

The extreme position has large high-frequency gain and (depending on specific layout) may be going into oscillation.

[karbomusic]

Reading the DC on the output of that opamp will tell you if a cap (or its PCB) has failed short and is blowing-up the opamp bias.

The extreme position has large high-frequency gain and (depending on specific layout) may be going into oscillation.

Thanks, I'll check. It does provide enough gain to clip at the extreme and it dies a little before that point. The gain gets to a certain point then plummets. I couldn't find an indicator that it was oscillating such as current increase but that is as far as I went checking for oscillation. It'll be tomorrow after a good nights sleep before I can look again.

[karbomusic]

OK... Nothing like a good night's sleep... Apparently... my own known good design is only good if I actually follow the design. Who'd a thunk?

Got up this morning, thought about it... The "behavior" of the pot is just weird. All biases are correct and everything checks out but the "behavior" of the pot is just weird... Repeat after me...

The "behavior" of the pot is just weird.
The "behavior" of the pot is just weird.
The "behavior" of the pot is just weird.
The "behavior" of the pot is just weird.
The "behavior" of the pot is just weird.
The "behavior" of the pot is just weird.
The "behavior" of the pot is just weird.
The "behavior" of the pot is just weird.
The "behavior" of the pot is just weird.
The "behavior" of the pot is just weird.


... Maybe I should look at the pot wiring again. And.... Wait for it....

I'm an idiot, pins 1 and 2 were reversed and my tired eyes simply kept glossing over it even when visually comparing to the other two builds. Even my very consistent color coding system failed my eyes this time around. And just to drive the point home... A pic I had taken hours earlier showing the issue. My hard rule is purple is always on pin 2 yet I missed it:



Thanks for the sanity paul btw.

Side Note:

I always do a bench test, then a later play test. During the bench test, I measure multiple parameters using both the voltmeter and the oscilloscope and record the values to confirm functionality as well as create a spec sheet that goes along with the pedal such as current draw, max in/out voltage/vref readings and so on @ both 9 and 18VDC, frequency response of the clean and dirt channels etc. This includes multiple visual reviews on the scope because I know exactly what each param should look like and what each reading should be. Consider it a spec sheet combined with a QC test with everything up on the scope which almost always catches anything that is wrong. I file the results for referring back to if needed which becomes a very handy cross reference. This is what caught it, though I likely would have caught this during the play test, it saved me the trouble of boxing up, head scratching, disassembly and so on.

[thelonious]

Great troubleshooting work! I've been in that boat many times before, and it's no fun. Good job.

[bloxstompboxes]

With my colorblindness, that would be easy for me to do. I try to stick with highly contrasting wires that I can distinguish well to avoid such mistakes. Of course, I admit, it doesn't always help, lol.

[karbomusic]

With my colorblindness, that would be easy for me to do. I try to stick with highly contrasting wires that I can distinguish well to avoid such mistakes. Of course, I admit, it doesn't always help, lol.

What's ironic is I made at least 1 small error on two boards and BOTH were the tone stack wires. Not to mention my layout reference that is taped on the wall calls out the pot pins and the colors. I must have simply been in too big a hurry when I did those aka too confident and didn't check as closely as I should have. Hence the lesson that going slower at that time (couple of extra minutes) would have saved me the 2 hours I lost later during the QC check.