Headtrip2 dry signal “heartbeat” problem

[bhaneman]

Just finished a HT2 build, lovely board.  Wet channel seems fine, controls work as expected... but the dry signal has an intense "heartbeat", it almost sounds like an LFO - about 130 bpm, always the same rate regardless of effect or pot settings.  I didn't do any particularly strange substitutions and have double checked all component values and reflowed a few joints, but the problem persists.  Any ideas where to continue debugging?  I'm going to check voltages but seeing as the wet channel is working, and doesn't seem to be modulating, I doubt that will produce an obvious culprit.  I'm running at 3.3 volts (via regulator) and input voltages are pretty low, FWIW. 
  Ideas?

[jimilee]

Pictures and voltages are the next steps. I'd reflow all the joints. Also, the power source, battery, multi power adapter, or just a single voltage?

[bhaneman]

For some reason VB (as labelled on the schematic : the reference voltage from a voltage divider in a single-supply opamp configuration) is oscillating when IC1 is present.  I've tried two tle2074s and a tl084 to be sure.  This is with tails disengaged, so it's got to be somewhere in the opamp gain stages, it happens even when the FV-1 isn't getting any input.  With the opamp installed, signal at pin 5 (input to first opamp stage) has the slow oscillation superimposed on the input signal and everything in the signal chain thereafter has the low frequency oscillation.  No wonder, since VB is oscillating... strange.  It doesn't happen if the input is an open circuit, for instance a guitar cable that isn't plugged into a guitar.  But from a live pedal for instance, the oscillation is back.  Surely the reference voltage isn't meant to oscillate, is it?  FWIW all other voltages are as expected and even the weird ones oscillate around their expected values.  Am I missing some RC circuit here?

Power source is a well vetted linear supply, no daisy chaining.  It's clean. I did try another power supply just for grins, identical behaviour.  I've already reflowed everything as I initially stated.

Edited to note that even the input signal is oscillating by the time it makes it to the first opamp.  I'm going to have another look at the switch wiring to see if I've inadvertently created an LFO...

[madbean]

Maybe the VB decoupling cap is discharging? That could cause the voltage to vary.

[bhaneman]

Maybe the VB decoupling cap is discharging? That could cause the voltage to vary.

Thanks for the reply!  And the hint, at least it's a direction to search.

That sounds about right... I take it that's C18, between VB and ground.  How would I diagnose that, and what would be causing it to discharge?  Something's causing a low impedance path from VB to ground?  Clearly it's not happening when the opamp is removed... funny thing, I tried touching another cap in parallel to see if the frequency would drop, and it didn't seem to change things... hmm

[bhaneman]

I tried putting a 47uF capacitor in parallel with C18 (just scratching the leads against the solder lugs on the back of the board)...  voila!  So I hadn't made good contact when I tried this before.  Thanks madbean for totally sussing this.  The old capacitor was a nichicon which I thought was supposed to be good... anyhow I replaced it with the 47u just to be on the safe side.  Pedal sounds great now! 

But I am tempted to build another one with an LFO on the Vref  it was annoying but potentially cool if you could turn it off or vary depth and rate 

[madbean]

It was a lucky guess. Could have been like 5 other things, haha. Anyway, glad that turned out to be an easy fix for you.