High gain build popping

[PBE6]

My OCD build is driving me nuts. After rebuilding the circuit with some helpful suggestions from this forum, I've managed to cure the oscillation problem and the whine when in bypass mode. However, this thing pops something fierce when the stomp switch is engaged.

I've already tried two things:

1) I attached a 1M pot in parallel with the 1M5 pull down resistor to see if that was the problem. It didn't do anything until the value was so low it was killing some of the guitar signal too, so I don't think this resistor is an issue.

2) I added a 47uF capacitor to ground after the 6k8 LED resistor as per Jack Orman's suggestions:

http://www.muzique.com/lab/led.htm

This didn't do anything in my case.

I know that the other option is to switch out the 3PDT switch for a relay, but I don't have one so I'll have to order it.  In the meantime I had a thought - Jack's LED anti-popping scheme exploits the capacitor's charge up time to introduce a delay that avoids the popping sound. If the real problem with this build is the fact that the OCD is high gain, would adding an RC filter directly to the power rail introduce a similar delay that would prevent the opamp from applying all of its gain right away and therefore avoid the popping sound?

[Kipper4]

How's about millennium bypass?

[induction]

Most circuits already have RC filters (aka filter caps and small series resistors) on the power rails. This works pretty well to clean up the power ripple in many cases, but won't do a thing for switch pops. The circuit is powered at all times, even in bypass, so the concept doesn't apply.

You could try using RC filters to slow down the DC offset that's causing the pop, but that would require you to find the source of that offset. If you can do that, you can eliminate pop by removing the offset, so no need to muck up the circuit with extra RC filters.

There are a few versions of the OCD, so if you post some voltages, schematic, photos etc., we can help you track down the cause of the pop.

[PBE6]

Here the schematic minus the LED popping fix  (sorry about the handwriting):

[anchovie]

Have you measured DC voltage at the input and output to see how much of an offset there is?

[GibsonGM]

^  That, plus, have you 'lifted' the LED (removed one leg) to see if it pops without it in-circuit?  Just curious about that....

[induction]

The schematic looks ok. Can we see photos of the build?

[antonis]

(off topic.. just curiosity..)

Is there any particular reason for double stabilizing Vref..
(except from wiring the unused part of OpAmp..)

[PBE6]

Here are some pics it the build:

[PBE6]

@GibsonGM no I haven't tried that yet, I can give it a try tonight.

@antonis I used a TL072 and the way the circuit was laid out on the board I was going to have to exit through the pins on the other half anyway. That and I recalled a thread stating the best way to tie off an unused opamp is to turn it into a buffer with Vref as its input.

[antonis]

I used a TL072 and the way the circuit was laid out on the board I was going to have to exit through the pins on the other half anyway. That and I recalled a thread stating the best way to tie off an unused opamp is to turn it into a buffer with Vref as its input.

Never thought (or readed) it...

Thanks a lot PBE6..!! 

[PBE6]

@Kipper I forgot about those bypass schemes, I'll check out the article on GeoFex.

[PBE6]

@anchovie no I haven't measured that. How would you go about doing that? Just use a multimeter, connect the common probe to ground and measure DC voltage at the input and output? I should be expecting 0 V at both ends, correct?

[GibsonGM]

@anchovie no I haven't measured that. How would you go about doing that? Just use a multimeter, connect the common probe to ground and measure DC voltage at the input and output? I should be expecting 0 V at both ends, correct?

Correct.  If you get DC, you have a leaky cap. I'd do it with the effect switched in, AND out.  Just for grins...again, a leaky cap might be letting DC go where you don't want it.  Or some odd short somewhere, allowing DC into the signal path.  Like on the switch terminals, where the LED connects, or something (?)  Doesn't look it, but hey you never know.

I'd like to know what the DC reading is at the input, at the junction of the 20n/1 Meg resistor.  There shouldn't be any DC there (the 1Meg is a pulldown, it would allow small DC currents to go to ground).  However, when you switch, you ARE having DC enter your signal path somewhere. It's just a matter of finding out 'where'.  The 1u output cap is the same deal on the other end, just for reference (check there too).

The other cause is the forementioned LED...it will draw a rush of current when switched on, going *pop* (the cap fix usually works, but OCD is very high gain, sensitive to this stuf).  Normally its own resistor slows that up, but sometimes it's not enough.

[drummer4gc]

If all else fails, try a different switch. I think the quality control on these 3pdts isn't fantastic...I've had some that just seemed to pop.

[PBE6]

Someone mentioned that in another thread too., can't remember who. But they said the switches on either side might be make-then-break or break-then-make randomly. Maybe I can just spin the switch around?? Ha...

[PBE6]

Ok I checked the input and output voltages. The input looks fine, reads 0.000 V for the most part (fluctuates to -0.001 V every once in a while, but that seems to be within experimental error limits). The output however does show a voltage of between -0.006 V and -0.020 V, depending on the volume pot setting, that dissipates over 2-3 seconds after switching.

Does this sound like a leaky cap, or is the volume pot just too large? Or both?

(And yes, those voltages are negative with the common probe on ground and the positive probe on the switch contacts - seems odd to me, but there you go.)

[bluebunny]

Nothing really to do with your popping, but, um, isn't that pink foam conductive? 

[GibsonGM]

That's probably a leaky cap (but not VERY leaky I don't THINK), or a problem in the switch, some kind of "bleeding", possibly.  The values you post are VERY small.  You wouldn't 'hear' that UNLESS it came back thru again and was re-amplifed (?)  which I suppose COULD happen on a massively high gain pedal. Hmm....

Is the LED disconnected at this time?

[duck_arse]

@anchovie no I haven't measured that. How would you go about doing that? Just use a multimeter, connect the common probe to ground and measure DC voltage at the input and output? I should be expecting 0 V at both ends, correct?

Correct.  If you get DC, you have a leaky cap. I'd do it with the effect switched in, AND out.  Just for grins...again, a leaky cap might be letting DC go where you don't want it. 

while we're grinning, my turn for a request. try measuring the dc at input and output again, but this time plugged into your signal chain, as you use it when it pops. sometimes, the problem ain't in your box.