Fuzz Face: treble bleed cap on fuzz pot (NOT volume pot)

[j_flanders]

Most people seem to like running their FF with the fuzz and volume controls maxed out and then control everything with the guitar's volume knob.
A few like it better with the fuzz pot just a smidge below max, but with 90% or more of the action all bunched up at the end of the fuzz pot it's hard to tell what a 'smidge' is.
Very few report about loosing treble when the Fuzz control is below max, which is what this thread is about.

I tend to like it better with the Fuzz knob 'slightly' below max because I get better clean up, or over a better range on the guitar's volume knob but the moment I turn down the fuzz knob, even just the tiniest bit, I loose the hiss but also the treble.

I've read a lot of the FF threads here, but I can't seem to find any info on this particular subject.
Most are about treble bleeds on the pedal's or guitar's volume pot, reverse log fuzz pots, harsh treble taming caps across B-C etc.

So far this is the only thread I could find that mentions it:
https://www.diystompboxes.com/smfforum/index.php?topic=105552.0

So I'm left with the following 2 questions:

1)Why do I loose treble when turning down the fuzz knob below max. Which R and C make up for the LP filter when doing so?
-Is R simply the resistance from the top lug of the fuzz pot to the middle lug, C the emitter bypass cap? But that would feedback (to Q1 base) less treble and make it brighter, not darker... (more feedback=more attenuation, less feedback is more amplification)

2) According to the author in that thread a treble bleed cap seems to work but I can't explain why it would.

[smallbearelec]

90% or more of the action all bunched up at the end of the fuzz pot

If you are using a "B" (linear) taper pot for the fuzz, try a C (reverse audio) taper. You will get a much wider "spread" and hear subtleties that are otherwise lost.

Also, you might want to consider a "preset" pot in series with the input. You can see this in the schem of my Bear Face

http://diy.smallbearelec.com/Projects/BearFace/BearFace.htm

That pot allows you to set the fuzz level you want and just dime the guitar volume when you kick in the effect. Between the two controls, you miay find it easier to tweak for the effect you want.

[highwater]

You could try a cap from Q2's emitter to the fuzz pot's wiper. Same basic idea as a treble-bleed on a guitar's volume pot.

What size cap to use is above my pay-grade... but it's pretty easy to experiment with even if it's already built; you can just run alligator-clip wires from the pot lugs to the capacitor.

[j_flanders]

90% or more of the action all bunched up at the end of the fuzz pot

If you are using a "B" (linear) taper pot for the fuzz, try a C (reverse audio) taper. You will get a much wider "spread" and hear subtleties that are otherwise lost.

Also, you might want to consider a "preset" pot in series with the input. You can see this in the schem of my Bear Face

http://diy.smallbearelec.com/Projects/BearFace/BearFace.htm

That pot allows you to set the fuzz level you want and just dime the guitar volume when you kick in the effect. Between the two controls, you miay find it easier to tweak for the effect you want.

Thanks for your help and your suggestions.
I'm well aware of those 'solutions' and the variants that exist: voltage divider instead of rheostat at the input, various configurations of regular pots and series or parallel resistors to mimic the C taper pot behaviour.
I'll most probably try them out in the end.

None of these will explain nor prevent the treble loss when turning down the fuzz pot pot though, at best they will make me not turn down the fuzz pot.

So, the purely technical question remains: what causes that treble loss. Where 's the C and R and Out in that LP filter? And where or how would a treble bleed help there (and what other side effects could that cause) ?

PS:
A somewhat personal note to you: I've wanted to order from your website so many times but each time I start to calculate the shipping costs, import duties, taxes, handling fees etc. for Europe,  I'm paying like 75$ for 2 ge transistors. I know that's entirely beyond your control. It's just very unfortunate for me as you have so much good stuff. At the moment I'm waiting to have more to order so those fees can be spread across more items.

Also, a bit more on topic: when looking at your ge transistors for FF's you label the resistors but I can't seem to find the schematic (with those R numbers) on your site. Maybe it's there though? Can you post a link? Or do you consider some schematic with their R numbers to be generally agreed upon. I mean what do you consider R1, R2, R3, R4 etc. I can somewhat tell by their value but in some cases those R's could be either Q1 collector resistor or feedback resistor. (one being slightly larger than usual and the other being smaller.)

[j_flanders]

*accidental post, please remove or ignore. I accidentally hit enter* 

[j_flanders]

You could try a cap from Q2's emitter to the fuzz pot's wiper.
it's pretty easy to experiment with even if it's already built; you can just run alligator-clip wires from the pot lugs to the capacitor.

Yeah, I guess trying doesn't hurt and costs nothing.

Same basic idea as a treble-bleed on a guitar's volume pot.

The thing is though, on a guitar the output is on the wiper and the pot is a voltage divider. Turning down the pot adds series resistance which forms a LP filter with the cable capacitance.

The Fuzz pot is a rheostat (variable resistor) however and there's no output on the wiper.
That pot seems to be
-a rheostat, adding emitter resistance to Q2, reducing gain (for all frequencies. 22µF is large to include lows)
-a variable resistance in the mixing/splitting setup that sends AC to either ground or back to Q1B.

Here's a visual representation of the difference:



Not sure I got everything right on the FF part but hopefully someone with more electronics knowledge will correct me.

[duck_arse]

the total resistance in rheostat config changes. in the fuzzface E resistance is the pot track, it doesn't vary resistance. it is more or less voltage dividing the cap connection. if you connect a small value cap from pot top to wiper [as well as standard electro to ground] you get a higher cut-off frequency emitter bypass, so boost the highs but not the lows.

[j_flanders]

the total resistance in rheostat config changes. in the fuzzface E resistance is the pot track, it doesn't vary resistance. it is more or less voltage dividing the cap connection. if you connect a small value cap from pot top to wiper [as well as standard electro to ground] you get a higher cut-off frequency emitter bypass, so boost the highs but not the lows.

Thanks! I think I get you what you mean, I hope...:

[PRR]

> Why do I loose treble when turning down the fuzz knob below max.

Because now you have Miller-type feedback which reduces the circuit input impedance which loads-down the top-ring of a naked guitar.

Try this. Purely thought-experiment. (I have a suspicion it will motorboat.)

[j_flanders]

> Why do I loose treble when turning down the fuzz knob below max.

Because now you have Miller-type feedback which reduces the circuit input impedance which loads-down the top-ring of a naked guitar.

Great to see you address the original issue or question since the treble bleed cap seems to have been explained.
I read that the impedance is 8k and further reduced to 5k depending on the Fuzz pot position.  (https://www.electrosmash.com/fuzz-face#link31)
Is it really possible I could hear a mere 3k Ohm reduction in input impedance?

I guess the best way to test this theory is to take the frequency response of the pickup while turning down the fuzz. I'll try to get it set up this week and report back.

Also, what exactly is the "top ring of a naked guitar"?

Try this. Purely thought-experiment. (I have a suspicion it will motorboat.)

As for your thought experiment. I'm working with this Dunlop pedal/schematic:

With that trimpot in the FB it will be easy to set up and test. I'll have to flip the cap for negative ground though.

[PRR]

> I read that the impedance is 8k and further reduced to 5k depending on the Fuzz pot position.

I don't know who to believe, my Idiot Assistant (backed by my gut feeling), or ElectroSmash.

My sim shows full-up Zin is 12k (already a heavy load on guitar treble), and at 95% up has already dropped to 2k, a VERY heavy load on g-treble, falling to 500r at minimum gain.

At pickup's top-resonance, around 3kHz, the naked pickup impedance is hundreds of kOhm, though guitar pots moderate this to 50k-100k half-up.

At bottom note Zin is always around 500r, which is a heavy load on the ~5k bass impedance of a pickup or g-pots.

There are several things going on. Getting cap off Q2 emitter makes Q1 gain rise. Toward Gv=200, which is why the "100K" can act-like 500r.

[j_flanders]

My sim shows full-up Zin is 12k (already a heavy load on guitar treble), and at 95% up has already dropped to 2k, a VERY heavy load on g-treble, falling to 500r at minimum gain.

Thanks for taking the time to plot this out.
12k to 2k to 500r is definitely something I could notice.

And "top-ring of a naked guitar" apparently means "resonance peak of an unloaded pickup". 

Turning down the guitar's volume knob greatly changes all this, I gather?
Because, full up, that 500r is parallel to my (Jazzmaster) 1M volume pot (excluding the || tone pot for now), making Zin +/- 500r
Down to 9, that 500r is || to 900k thus Zin is 100k + 900k||500r, and it jumps to +/- 100,5k.
The more I turn down the volume knob the more I increase the Zin. At 1 it's 900k + 100k||500r: 900,5k

[j_flanders]

Try this. Purely thought-experiment. (I have a suspicion it will motorboat.)

I tried it but I'll have to experiment with some more capacitor values. Too low and it does very little, too high and it sends too much lows too ground and then it increases overall distortion. No motorboating so far.

So, if I understand it correctly:
- when turning down the fuzz knob, not only DC but also AC gets fed back, which lowers AC gain/distortion
- the cap to ground sends the highs in the feedback to ground, excluding them from feedback, giving more gain for highs compared to lows, with the intent of making up for the treble loss, caused by the lower input impedance from turning down the fuzz pot.

Neat idea.

[PRR]

The AC feedback through 100k *lowers the input impedance*.

If fed with a buffer, this would not matter much. But guitar is sensitive to loading.