Controlling the amplitude of the output signal from a pedal?

[GibsonGM]

Ok, yes, sure, in theory, there isn't a path to ground for AC there, so your assumption was correct.  But, for a good, well-made pedal, you'd really want the cap, so the bias resistor is still in play, as I 'related' above.    Your instinct on that was right, tho! 

They came up with these things to make audio equip. more stable...clean power means less hiss/hash/crap/radio coming into your device, and feeding opamps thru resistors can protect them as well as provide a 'stiff voltage' for the input to see.    When you get into using LFOs (low frequency oscillators, perhaps to drive a chorus or trem circuit etc), that will become far more important, too - in keeping LFO clicks and noise out of the audio.    Just good practice   

[YurkshireLad]

Using your circuit from https://i.imgur.com/bKngFu5.png, I'm seeing 2.3V DC on the non-inverting input.

[GibsonGM]

Does it work?   I'm going to remember this time...that could be due to meter loading    

[GibsonGM]

Sorry, had to run to make dinner.

As an aside, what do you have for a DMM?   Sometimes a meter can draw enough current from the circuit under test that it causes the voltage to sag (called "loading").  I'm curious what the impedance of your meter is, to see if that's the culprit.   

If the circuit is operating fine, then that's no concern...esp. if you read about 4.5v  back at the divider, and KNOW the 1Meg IS actually running right to that input, and the cap is there to keep DC from flowing back to your guitar pickups etc...

[YurkshireLad]

This is my DMM - https://www.canadiantire.ca/en/pdp/autoranging-digital-multimeter-0520052p.html#srp.

I'm definitely reading 4.5V back at the divider.

[YurkshireLad]

Maybe it is working; I put in a gain loop with a gain of 11 (10k + 1k resistors) and I'm getting a clipped output signal of about 7.8vpp. I'll plug the guitar into it tomorrow.

With a gain of 1, I'm getting an output signal of about 3vpp, which is the same as the input signal.

Thanks for your help.

[GibsonGM]

Hi Yurk,

Ok...at some point approaching V+, you're going to see clipping if you have an input signal...the thing can only amplify up to a certain point, then power supply limitations cause it  to run out of headroom.   To avoid that, some people run on 18V - but the vast majority don't seem to need to (why would we ever need to have a signal at over 7V?   18V is just for super-cleanness).   

Your meter is most likely loading...I couldn't find a spec for its INPUT IMPEDANCE (sometimes, Zin).   That will determine if it is loading the divider.   If you lift the left of the 1M resistor feeding the opamp, at the opamp end...see what it reads for voltage (for curiosity).   I'd say it's just loading, esp. if you get clean signal over any reasonable range   (and it seems you are given your results!). 

Any reason for 3V P-P input?  A guitar would be more like 1.4V p-p on average (this is debatable). I say that because the lower input signal would mean you have more headroom than you think. But coming from ANOTHER gain stage, not as much.

[YurkshireLad]

I got tired of strumming the guitar every time I wanted to check a signal, so I uploaded a simple SIN generator to an ESP32. It's fixed at 3V, hence the high input signal. I don't have a proper function generator.

[GibsonGM]

If you can find R.G. Keene's "quick and dirty oscillator", on the Geofex site linked at the forum main page, that is a very useful thing to make, and is easy.  I use that a lot.   

OR, maybe you could add a 100k output pot to your generator and dial in 1v. Not strictly necessary tho, only to have a 'standard' of 1V which makes things easy.

At one time I made a CD of myself strumming and noodling guitar for 10 minutes, and would use that as input   Same deal tho, it's not 'life like'.   

[YurkshireLad]

I originally had a resistor on the signal input to bring the 3V down, but I took it out so I could work with the original signal. I'll put a volume pot on the output before I plug the guitar in.

[GibsonGM]

Yes, just put signal in from one end, take the wiper out to your circuit, and the other end gets grounded. (you probably already know that!)

[YurkshireLad]

The op amp is still only biased at about 2.3V and the output signal doesn't sound right. I think I'll put this aside for now and move onto something else.

[GibsonGM]

There's only 1 thing I can think of that would cause that - bad opamp.  Not likely tho.  Other than that, the bias should be fine.   If the power is flowing thru the R divider (you confirm this), and the 1M resistor is connected to the Vr point...it is there.   

Lift one leg of the 1M bias resistor, the one AT the opamp. Now measure DC voltage right from that resistor leg.   Now measure the voltage on the OTHER leg of the 1M resistor.   The high input impedance (AC resistance) of the opamp is working with the 1M resistor AND your meter to fool you, I do believe.   If your meter isn't mroe than say, 3M input impedance, this could well be the case

EDIT:
Ok, I found specs for your meter, and it says you have 10M input impedance ("Z").  That's a lot, so should be minimal in terms of loading.  But there's little reason the bias voltage would be 'off' over such a short run.  I mean, there literally are only like 2 places a mistake could be.   While at it, measure the resistance of that bias R to be sure it really is 1Meg.

[YurkshireLad]

It's definitely 1M - I measured it and I also tried a different 1M resistor.

The leg attached to Vr measures 4.5V DC, the other measures around 2.3V DC. The op amp is definitely being biased by about 2V DC, which means any gain clips the bottom of the signal, but not the positive.

[GibsonGM]

I know what you're getting at. 

Maybe it's the TL082, vs a 'standard' TL072 (?).    You do have the other half of the opamp set up like a buffer, right? So it's not just floating.

If you've done that...try lowering the 1Meg bias resistor to 100k, see what you get at the chip; if the '82 has lower input z than we're used to, that should increase the bias at no real cost.  And lower R2, from 20k to 10k, see if a gain of 21 sounds better....do you have any other opamps?

[YurkshireLad]

Hmmm no its floating, I'll try to address that.

I thought I had some LM358s but I can't find them. I might have some LM386s but those have a different pin configuration, so I need to read the datasheet.

[GibsonGM]

(or another '82, just to be sure this one isn't kaput).    That opamp should have sufficient input impedance to NOT drag down the bias voltage, which is what seems to be going on.   It should be more than 1 meg...which wouldn't load that bias R very much.   If it's all built right, that should be the case.   

Trust me - once you get this going, you'll make 1 or 2 more, and then it won't be hard!  You forget about the 'physics' of it, and just build them (that's why I have to think about WHY yours isn't working right).

Take the unused opamp and terminate it. If we get the gain stage working, we'll go back and make that a buffer before it, or another gain after   For now, park it so it's not a factor...

https://paulinthelab.blogspot.com/2015/04/proper-op-amp-termination.html

[YurkshireLad]

I tried another TL082 and terminated the unused opamp with no change.

I changed the bias resistor to 100k and I'm reading 3V DC at the non inverting input and output pins. So an increase of around 0.7V.

[GibsonGM]

Says something, but it's still not right.     Is C1, the input cap, polarized? If so, the "+" leg should be going to the opamp.  If that was backwards or defective, it could allow DC to pass back to the input jack (draining it away).  You might measure for DC across the jack just for grins.  There should not be DC there * 

What is the inverting input pin reading with 3V on the others?  Should be quite similar (same).

You do have a cap from 'Vr' to ground in the divider?  What does your battery read?

* as an aside, that cap is way big for guitar work...something 1/10th that value would be more suited and thus non electrolytic..like, .1u or so.  If you can avoid using an electro it's good to do so, they are kind of - crappy.

[YurkshireLad]

"Is C1, the input cap, polarized? If so, the "+" leg should be going to the opamp."

Ah no, the cathode is connected to the opamp. This will be the first thing I try later.

Thanks again.