Weird tone control phenomenon

[Mark Hammer]

So, I was tweaking an FY-2, and I thought I would extend the range of the scoop adjustment, like I did with the Aefea.  The FY-2 normally has a 10k anbd 15k resistor in series between the Fuzz control and Volume pot, with a .1uf cap to ground from their junction.  I replaced those two resistors with a 1k5, 20k pot, and 4k3 resistor, with the pot wiper connecting to ground through the .1uf cap.

Okay, so far so good.  The premise was that the pot would allow for adjustment of the lowpass rolloff point from about 74hz to 1060hz.  This sort of arrangement worked quite well in the Aefea.

So I wire it up, and the bloody scoop-shift pot behaves weird.  Move the wiper in the direction of the 4k3 resistor and yes it rolls off treble, then mids, and finally upper bass.  But rotate it in the other direction, towards the 1k5 resistor and at first it increases upper bass and mids, as it should, but then as it gets towards the 1k5 resistor, treble starts to be cut again.  This is true in any of the tone-switch positions.

I thought maybe it was the absence of a cap, just after the Fuzz pot wiper (the Aefea feeds a similar tone control from the output of an op-amp, with an electrolytic cap on the output of the op-amp), but that did not change anything.

Essentially, once the resistance between the Fuzz pot wiper and the .1uf cap gets small enough (and I've set the lower limit as 1k5), that small resistance and cap to ground behaves as a treble bleed, rather than a lowpass filter with a higher rolloff point.

And its driving me nuts.  How do I avoid this?  I thought maybe if I made that 1k5 resistor bigger (and 1k5 is already bigger than the 820R I started out with), and chopped the value of the cap to ground, the larger resistor value might yield greater isolation, and allow the variable lowpass to behave as a variable lowpass across its entire range.  But will that work?

[PRR]

For minimum interaction, you usually make each sequential stage much-much higher impedance than the one before it.

(The opposite is possible but unconventional and typically mind-bending.)

So estimate your impedances.



The tone-section impedance is 1/10th or 1/50th of the stuff driving it.

There's too much stuff here for me to count on fingers. And my neck is stiff so I won't hand-crank my antique simulator today. But my guess is that it worked with fixed values by guess/gosh trial, but a Variable solution may need significant re-scaling and/or buffering.

[Mark Hammer]

Thanks.  I appreciate your efforts, and explanation.  What bugs me is that it worked here. https://www.youtube.com/watch?feature=player_embedded&v=CGWPSDK4Huk  The main difference was that I didn't aim for as much variation.

Would changing the value of the Fuzz or Volume pots (e.g., Fuzz lower, Volume higher) make a difference?

[earthtonesaudio]

Mark, I think buffering on both sides of the tone section would make it a lot easier to reason about. You can't change the Fuzz pot too much so try to increase the impedances to the right of it. I'd start by using a smaller capacitor than the 0.1 (on the wiper), as you suggested.

An alternative arrangement of the cap and pot might work better for your goals. The tone control in the middle of this schematic is a mid-scoop (wiper left), or a treble-cut (wiper right) or a mid boost (wiper middle). It's not a true mid boost, but in the middle position it cuts less of the mids and treble, which gives the impression of more mids.



The values above are suitable for the ~1k source and ~100k load impedances. For dropping in to the Companion Fuzz, I would start by scaling all the resistors by 10 and all the capacitors by 1/2.

[Mark Hammer]

Thanks for the feedback. 
That's an interesting circuit.    I gather R14 pans between the distorted and undistorted signal?

[duck_arse]

mark, did you ever try the "swinging hammer" mod?

http://www.diystompboxes.com/smfforum/index.php?topic=101447.msg920069#msg920069

[Mark Hammer]

Need to be more specific.  Which mod is that?

[duck_arse]

the pot wiper goes to the resistor string, and you put a diff value cap at either end, to ground.

[Mark Hammer]

Okay, I get it (if there is any accompanying graphic, I can't see it at work).  The series resistance ahead of each cap to ground is varied, with the mid-position effectively minimizing the role of both caps.  Interesting.

I actually did do that to a Muff Fuzz (essentially a silicon Fuzz Face with a fixed gain), with the pot wiper going to the emitter of Q2.  A medium-value electrolytic to ground was on one outside lug, and a smaller nonpolarized to ground on the other outside lug,  so that gain could either be applied across the entire bandwidth, or to the upper mids and highs only.  But that was an application of such a control to gain adjustment, rather than only tonal adjustment.  Worth exploring further, though.

[earthtonesaudio]

Thanks for the feedback. 
That's an interesting circuit.    I gather R14 pans between the distorted and undistorted signal?

Thanks! I like it a lot. It's a fuzz, so it's more like "distorted and more distorted". The C1/R5 to ground is rather low impedance, and shunts/divides negative feedback for both op-amps (less negative feedback yields more gain). IC1A gain goes from about 110 to 1100, while IC1B goes from about 3 to about 9. Make R5 a little bigger and it'll oscillate.