Balance/panning volume changes.

[stonerbox]

Having a pretty expected issue on my hands here but I can't figure out any smart way to mend or solve it.. when both fuzzes are mixed the volume is significancy higher than when one of the signals is muted. What could be done? I have tried increasing R2, R5 and R6 in an effort to prevent the soloed signal from being pulled down but it seemingly did nothing.



full image
https://i.imgur.com/VtzOJKs.png

[antonis]

https://www.diystompboxes.com/smfforum/index.php?topic=116219.0
http://www.geofex.com/Article_Folders/panner.pdf

[ElectricDruid]

+1 agree. Antonis is right.

The Geofex article tells you that the pot should be x0.707 the resistor value, so 10K for 15K resistor. You have 100K.

Alternatively, you could swap all the 15K resistors for 150K. Same thing (ish), done the other way around. 15K/10K is better from a noise point of view.

[Rob Strand]

With a panner the idea is the perceived level is the same for all positions of the pot.

A problem that arises is non-correlated signals (eg noise) add differently to correlated signals (eg a similar signal like L and R channels or a modified version of the original).

You will see variants with different resistor values for both cases and even in between cases trying to cover both cases.  There were some detailed posts about it say 5 years ago.   I put up some spice simulations.   The post linked an article by Rane or Peavey pretty much saying the same thing.

Having a pretty expected issue on my hands here but I can't figure out any smart way to mend or solve it.. when both fuzzes are mixed the volume is significancy higher than when one of the signals is muted. What could be done? I have tried increasing R2, R5 and R6 in an effort to prevent the soloed signal from being pulled down but it seemingly did nothing.

It makes more sense to simply attenuate the louder one using a simple voltage divider so the two outputs have the same level!

If the cause of the problem is the 15k's loading thing down then follow what the other guys said.

As an example.

Suppose you want to knock the level on one arm by 6dB (ie. a factor of 0.5).   If you replaced R1 (15k) with 30k then add a 30k resistor from the pot side of R1 to ground it will create a 6dB divider (30k & 30k) and also maintain a 15k impedance (30k//30k).

If you wanted 5dB (0.562) then it's a bit messier.   Replace R1 with 27k and then add 33k to from the pot side of R1 to ground it will create a -5.2dB divider (27k & 33k) and maintain a 14.9k (27k//33k) impedance.

[ElectricDruid]

I found one recent thread where we discussed this before:

https://www.diystompboxes.com/smfforum/index.php?topic=128576

The rough gist of the relationship between the pot value and the resistor value is that as the pot/resistor ratio gets bigger, you get more hump in the middle. You can reduce the hump by using a smaller pot and/or larger resistors, but then you need a lot of make-up gain in the op-amp.

In the OP's case, they've inadvertently used a pot that's much too large and given themselves a big hump in the gain in the centre. It looks like this according to my old script for this kind of thing! -



( Red = Left, Green = Right, Blue = voltage sum, Magenta = power sum )

Edit: The 15K/10K pot "traditional" equal-power values give you a +3dB bump in the middle for correlated signals. Two different fuzzes might get something close to that, but they're not identical, presumably. But +3dB isn't huge.

[Rob Strand]

I can only find this one:
https://www.diystompboxes.com/smfforum/index.php?topic=118967.0

Constant power when the resistors are the 4x resistors are 1.5 times the pot value and unity gain when feedback resistor is 7.5 times pot value. (not confirmed)

[antonis]

It makes more sense to simple attenuate the louder one using a simple voltage divider so the two outputs have the same level!

+1
(neat and clear..)

[stonerbox]

+1 agree. Antonis is right.

The Geofex article tells you that the pot should be x0.707 the resistor value, so 10K for 15K resistor. You have 100K.

Alternatively, you could swap all the 15K resistors for 150K. Same thing (ish), done the other way around. 15K/10K is better from a noise point of view.

Boy, do I feel like lazy! The info was right under my dumb nose. Thank you all for the great input, you guys are valuable, patient and kind beyond words.