a question on rectifying weirdness

[duck_arse]

shown is a fairly common type of rectifier circuit, used for envelope generation, as it is in my case. I've been watching the output of this (and a number of variation configurations) on the cro, and I'm confused by the behaviour of the output voltage.



with a wang on the guitar feeding the not shown buffer, the output voltage rises to a peak level of about 4V5, then falls at a rate to ~1V5, then sits for a bit, then rises again, maybe as much as a volt, sits a while longer, then decays as I would have initially expected it to.

the output shows little ripple, especially with D3//R5//C6 fitted, so I know I'm not watching ripple. as such. I've compared the output with various prior points, but can't see anything to correspond with the bounce. it doesn't seem to care about the sensitivity setting, single coil/humbucker, ordinary electros/tants/film caps, either.

can anyone explain?

[SISKO]

A build mistake maybe? Try watchnig the opamp output to see if the bump comes from there or even before that to see where it comes from. Maybe its a problem with the buffer. Try folowing the signal path to see where it may come. On the simulator, the rectifying part works ok

[duck_arse]

I've tried different types of opamp, watched the output at various points, rebuilt it a dozen times, diff value/types caps and resistors ...... the buffer is just an emitter follower. if I could get dc voltages into the sound card (and the oscilloscope program didn't fall over on slow scans), I'd record a decay for all to see.

[SISKO]

And where does this bump apperars? Only at output?

[Seljer]

What if you change the type of opamp?

[duck_arse]

indeed yes. nothing I can see out of the ordinary before the end.

changed the opamps, 741, tl072, lf351, lm1458. haven't tried an lm324 yet.

[Seljer]

What if you remove C4 and manually apply a pulse of DC voltage to junction between D1 and D2?

[Mark Hammer]

Maybe you need to "scope that wang", and see if there are any perturbations in the string output/vibration.  Strings often don't behave as linearly as we like to think they do.

[ashcat_lt]

IDK if this helps, but...

If you take a Rat and remove one of the diodes in hopes of getting asymmetrical diode clipping, you end up getting no diode clipping at all, because that end of the coupling cap floats up.  You're doing about the same thing in this scheme.  Could that be part of your issue?

[duck_arse]

@ashcat - I have observed that very thing with rat style diodes. I noticed it happens if D1 is removed, but with it fitted all the diodes have "something" either end, and work real good.

as usual, Mr Hammer points and says ..... and lo, it seems the answer is "wood" and wang (or high gained noise at the opamp output). I lean the guitar against the long wooden rail of my bed, and it stands on the strap button. if I put it [here], it might or might not find some sort of sweet spot, and resonate to beat notes from the strings as they decay.

or something. it does it here, and if I balance it on the button on the rail, but not there, and not if I hold it by the throat in mid air. that, and a feedback cap, and we'll call it done.

[PRR]

R2/R3= 200:1.

"Any" guitar input will OVER-load that op-amp.

Many opamps don't overload gracefully. The squared output may shift asymmetrical. That will mess-up the crude rectifier.

If I'm right, the voltage on C3 is shifting significantly when whanged.

It also won't "follow the input" when OVER-loaded.

Look into lower levels, then at "Precision Rectifiers" so you don't need a 1.2Vpp slam to get any output at all.

[duck_arse]

funny you should say precision, prr, I had one on the breadboard this evening, unaware of its correct name. it seemed to behave a little oddly, with its output limited by the Vref level, with V/2 producing less swing than V/3. is this correct operation, or did I build it wrong?

I mighta wasted my time playing with my wang yesterday.