Gain control pros and cons

[TimWaldvogel]

What are the pros and cons of putting the gain pot before the op amps input VS putting it in the feedback loop.

[petemoore]

  It's easier to do on an offboard wire ?
  Depends on what it depends on as usual...
  The circuit, what you want out of it, how it responds to the gain placement.
  Rod Elliot and other opamp pages describe the feedback loop gain method, and other ways to set opamp gain including frequency selective gain setting.

[TimWaldvogel]

I guess what I mean is, on a basic overdrive or distortion design would it change the way the pot controls the gain?

[GibsonGM]

You are likely to get a loss of input impedance if you place the gain pot  before the input. In this case, it would only act as an 'input volume control'.   This would result in a decrease in high frequencies (like turning down the volume knob on the guitar), and also will decrease signal-to-noise ration, leading to hiss in the output. 

With a pot in the FB loop, you can filter frequencies of interest, too, which you lose by controlling signal level at just the input or output.  That is what the caps are sometimes in the loop for.  Variably boosting some frequencies over others (usually treble).

I'm not clear on if you would have a dedicated 2-resistor feedback loop, setting the opamp at 1 gain level, and would be controlling its drive by input signal.  If so, the above is true....if you still have a gain pot, and add a volume control at the input, you'd end up with a variable input impedance plus normal gain control....which would be interactive and do strange (maybe neat??) things, but I still think would lead to unnecessary hiss/noise...

[TimWaldvogel]

Well what if the input inpediance was set by a buffer or previous gain stage?

[fuzzo]

With a pot in the FB loop, you can filter frequencies of interest, too, which you lose by controlling signal level at just the input or output.  That is what the caps are sometimes in the loop for.  Variably boosting some frequencies over others (usually treble).

You mean we can cancel the treble lost we have with a "regular" volume pot (volatge divider at the end of the circuit) in placing it on the feedback loop ? that could be good to avoid interaction betweens tone and volume control.

[cpm]

Well what if the input inpediance was set by a buffer or previous gain stage?

thats ok

you may use a pot as attenuator before an opamp in some application
For example in a non-inverting stage where gain is set by 1+R1/R2 (cant get <1), or an inverting stage where gain is set by -R1/R2 (you can attenuate but not to 0)

I also think may be useful when there are filters as part of the feedback loop, this way you can vary gain without altering RC frequency settings.

You mean we can cancel the treble lost we have with a "regular" volume pot (volatge divider at the end of the circuit) in placing it on the feedback loop ? that could be good to avoid interaction betweens tone and volume control.

also works putting a buffer right before that volume pot.
Pot of FB loop behaves different than a regular divider, as i noted ^^

[fuzzo]

Ok, I didn't think about buffer actually.

Catalinbread did that on  their new pedals (jfet drive) , a buffer after each gain stages.
 

[TELEFUNKON]

Pro in feedbackloop: 2 connections only. Can be replaced by LDR!

[GibsonGM]

Fuzzo: well, to some extent, you could regain treble lost by an input pot, but that's sort of doing work twice (parts count).   A pot in the feedback loop will just BE a gain control.  I meant that if you had, let's say, a bass-heavy input, and wanted to trim some out before distortion, you can do it in the feedback loop - it's very handy!

Adding a buffer before a gain stage just in order to use a pot on the input is a waste of extra parts and work, IMHO.  There are times it's DEFINITELY a good idea, don't get me wrong (if you need to drive something like a tone stack, or change impedances, etc), but for the simple stuff, not a lot of point.   You will get more noise with a 'volume control' in front than a 'gain control' in the FB loop, which will reject common mode noise if done correctly.
 
The output impedance of an opamp is such that it is low, anyway, with or without a buffer, and its input Z is high (unless you dial it down by putting a voltage divider, AKA volume pot, before it)....in the end, try it out and see if it's too noisy.  If not, it's a good idea!  )