Strategy sought for 2-in-1 pedal

[Digital Larry]

There's a lot of stuff I've never done.  Like built a fuzz pedal.     I know!     I want to correct this oversight. 

I think I would like to combine two designs I find interesting.

a) Silicon fuzz face
b) Silicon harmonic jerkulator

I looked at a fuzz face schematic and the output voltage is developed between a 330 ohm and 8.2k ohm resistor in Q2's collector circuit.  I'm a bit rusty at this stuff but even if Q2 goes completely between saturation and cutoff, that is going to be a fairly small AC voltage swing.  Is this correct?

I don't have a clue what the voltage swing of a harmonic jerkulator is either.  Some designs include clipping caps which would obviously affect that.  I was thinking of adding clipping LEDs.  I know that whatever clipping thing one does affects the level and loudness.

So I want these two things in a single design with a blend knob between them and maybe a big muff pi tone control at the end!  Hey why not?  I'd like to bring it in at 6 knobs or fewer.
E.g.
2 each for drive/adjust on the FF and HJ
Tone
Level. (probably not a 0-100, I think something like -12 dB/+6 dB would be better).
Blend!  I forgot about blend!  That makes it 7.  Humm.

The main thing I am not too sure about is the blend between the FF and HJ.  If I knew they were going to swing the same under typical conditions (e.g. right at the collector resistor point) and the impedances were comparable I might just put a linear high value pot between them.  If that was not the case I could probably run the two sides into inverting op-amp stages, where I could adjust the level balance in a fixed way prior to driving a blend knob with low impedance sources (the op amps).

Thoughts?  Suggestions? 

Thanks!

[garcho]

Bravo! Give your eyes a break from code and squint at a PCB for a while instead

how about a summing amplifier and a "level" pot for each? or you could do it one further and have a clean mix too. HJ level, FF level, dry level, FF drive, HJ drive, tone. 6 knobs

Good luck and enjoy your first fuzz pedal build! 

[GGBB]

You might need to invert phase for one of the pedals. I *think* the HP is inverting but the FF is not.

[iainpunk]

You might need to invert phase for one of the pedals. I *think* the HP is inverting but the FF is not.

they are both non-inverting, 2 gain stages each.

you can also mash together the designs, the HP is basically a NPN and PNP gain stages stacked on top of each other. you could stack two low gain Fuzz Faces, one NPN and one PNP on top of each other. if you use like 30 to 50-ish Hfe trahsistors the gain won't get too high.

cheers, Iain

[GGBB]

[the HP is basically a NPN and PNP gain stages stacked on top of each other

I must have been looking at a different schematic. It looked to me like 2 devices in parallel, not series like the FF.

[Digital Larry]

Let me be a bit more explicit.

A passive blend, linear pot between two low impedance sources with the same peak output (or RMS, or whatever level criteria you want to use) will crossfade between them.  I'm not sure how it works in the analog world, but in DSP if you do a linear fade between 2 signals there is a volume dip in the middle as it does not obey the equal power law.  If there's a separate level control it doesn't matter as much as long as it can compensate.

The fuzz face's load circuit looks like it seeks to create a guitar level output signal even though the output transistor is going between saturation and cutoff - or at least very close.  Since I'm going to have another mixer stage and EQ after this, I don't necessarily need to attenuate the signal so heavily there.  I could go for full swing off the collector if I wanted to, or split the load resistor at any point to get in between values.

I haven't peeked at the HJ schematic yet, but again I am interested in the raw voltage swing of the waveform generating element and its output impedance.  Now I know you are going to tell me to just go off and breadboard it, and I guess I will.

Comments?  Am I crazy?

[garcho]

Now I know you are going to tell me to just go off and breadboard it, and I guess I will.

How did you know?! 

A friend of mine fixes/upkeeps super high end studio gear, as in Neve consoles, SSL, etc. He's always saying, in a cartoonishly thick Quebecois accent, "first, we must try". There's only so much theory reality can accommodate before it plays a joke on you. Transistor overdrive/distortion seems like one of those cases. Maybe I just like that quote because it makes me feel better about myself not understanding half of what I build...

[iainpunk]

[the HP is basically a NPN and PNP gain stages stacked on top of each other

I must have been looking at a different schematic. It looked to me like 2 devices in parallel, not series like the FF.

its two series gain stages, with the headroom divided between them, the point where the emitters come together, the big ass capacitor makes that point DC at around 3 to 4 V.

then its just two almost-identical gain stages with feedback/miller bias



cheers, Iain

[iainpunk]

Let me be a bit more explicit.

ooohhh, sounds kinky

A passive blend, linear pot between two low impedance sources with the same peak output (or RMS, or whatever level criteria you want to use) will crossfade between them.  I'm not sure how it works in the analog world, but in DSP if you do a linear fade between 2 signals there is a volume dip in the middle as it does not obey the equal power law.  If there's a separate level control it doesn't matter as much as long as it can compensate.

in real life, you will have the same dip in the middle if the signals are different, like two separate instruments or a delayed signal but with two similar fuzzes, that difference will be smaller, however, its better to implement active mixing with two volume (or mix level) controls

The fuzz face's load circuit looks like it seeks to create a guitar level output signal even though the output transistor is going between saturation and cutoff - or at least very close.  Since I'm going to have another mixer stage and EQ after this, I don't necessarily need to attenuate the signal so heavily there.  I could go for full swing off the collector if I wanted to, or split the load resistor at any point to get in between values.

just take it off the collector, this has less impedance and thus less prone to noise issues, changing the two resistors here will also affect the bias

I haven't peeked at the HJ schematic yet, but again I am interested in the raw voltage swing of the waveform generating element and its output impedance.  Now I know you are going to tell me to just go off and breadboard it, and I guess I will.

its about 4v pk-pk and around 15k output impedance

Comments?  Am I crazy?

can't wait to hear sound samples,
am personally not a huge fan of mixing fuzzes, unless there is some filtering going on before the mixer stage, like taking the bass and high mids form the HJ and the low mids and treble form the FF

cheers, Iain

[Digital Larry]

How does the output impedance come at 15k?  From my lessons by the river I would expect it to be the collector resistor's value which is 90 or 100k based on the 2 schematics shown.

There's a few reasons I like theory.

a) It doesn't involve building anything (I'm not too big on building electronics any more... ironically)
b) Helps me understand what to expect so I don't waste tons of time messing with stuff on the breadboard
c) Reminds me of being in college 

can't wait to hear sound samples,
am personally not a huge fan of mixing fuzzes, unless there is some filtering going on before the mixer stage, like taking the bass and high mids form the HJ and the low mids and treble form the FF

cheers, Iain

That's a good point.  I may wish to explore each on its own first.

[garcho]

a) It doesn't involve building anything (I'm not too big on building electronics any more... ironically)
b) Helps me understand what to expect so I don't waste tons of time messing with stuff on the breadboard
c) Reminds me of being in college 

those are great reasons!

[iainpunk]

you talk about the Harmonic Jerkulator, which has 22k Collector resistors, instead the other values the Percolator has.

given the fact that
- the idle voltage of the 2nd transistor's bias is around 7v, and the huge capacitor inbetween the emitters is at 3,5v to 4v gives the transistor 3v to 3,5v of drop and the 22k has 2V drop. (lets use 3,75v [emitter] and 3,25v [Vce])
- we know the Collector current of the transistor is roughly the same as the resistor current.
- the resistor has an impedance of 22k, 2V gives it 0,00009 A
- 3,25 : 0,00009 gives a virtual resistance of 35,75kOhm,
- 35,75 parallel to 22k gives 13,63kOhm (i made a mistake in parallel resistances earlier) as a rough estimate of the output impedance
the voltages i used are form my own experiments, with the two particular resistors i used, if you use different ones, you will get different voltages, even if they have the same type number.

also; the output impedance is different over the swing of the gain stage, at the positive peaks, the output resistance is higher and the negative peaks have lower impedance.... its a bit of a rabbit hole to get in to.
it can be avoided by putting a buffer at both of the outputs of the HJ and FF

cheers, Iain

[Digital Larry]

you talk about the Harmonic Jerkulator, which has 22k Collector resistors, instead the other values the Percolator has.

given the fact that
- the idle voltage of the 2nd transistor's bias is around 7v, and the huge capacitor inbetween the emitters is at 3,5v to 4v gives the transistor 3v to 3,5v of drop and the 22k has 2V drop. (lets use 3,75v [emitter] and 3,25v [Vce])
- we know the Collector current of the transistor is roughly the same as the resistor current.
- the resistor has an impedance of 22k, 2V gives it 0,00009 A
- 3,25 : 0,00009 gives a virtual resistance of 35,75kOhm,
- 35,75 parallel to 22k gives 13,63kOhm (i made a mistake in parallel resistances earlier) as a rough estimate of the output impedance
the voltages i used are form my own experiments, with the two particular resistors i used, if you use different ones, you will get different voltages, even if they have the same type number.

also; the output impedance is different over the swing of the gain stage, at the positive peaks, the output resistance is higher and the negative peaks have lower impedance.... its a bit of a rabbit hole to get in to.
it can be avoided by putting a buffer at both of the outputs of the HJ and FF

cheers, Iain

Dude!  You win!  Thanks for the insight.

[iainpunk]

you're welcome

cheers, Iain

[Mark Hammer]

I was thinking more in terms of the sort of balance control used in stereo amps.  As illustrated here, the mix/blend control doesn't really pan between the two subcircuits, but rather bleeds their outputs in reciprocal fashion.  The loss from this attenuation circuit is compensated for by gain in the mixer/output stage.  Of course, one is free to use dissimilar mixing resistors for the two signal sources, and add a bit more gain in the output-mixer stage.  In my case I used unequal series resistors for the mix/attenuator control (4K7 vs 22K), to compensate for the difference in output level for the two signal paths.

My own goal with this circuit was to have a lot of tonal variation with as few controls as possible.  I won't claim that it was a fantastic-sounding circuit, but it worked and illustrates the principle.

[Digital Larry]

Hi Mark,

That circuit is pretty much exactly what I had in mind, although looks like you combined TS and Rat style distortion/OD and allow to fade between them.  You say it didn't sound great, but did it sound any worse than either of the individual circuits sounded on their own?  Or was it just that the "blend" didn't bring much to the table?

Here's my main motivation, believe it or not.

I have a few SparkFun Proto Pedal boards which makes assembling things much easier (for me) as there's no jack or stomp switch free wiring required.  Pots, you still gotta wire.  They are pretty large and fit into a 1590BB.  I've used TinyCAD and VeeCAD to design Vero layouts previously and find it an amusing way to spend time.  I already made a board following the Proto Pedal.  The only thing about it that's a little weird is that the traces are internal so you cannot cut them.  Oh well!

So I just want to build something that I don't have that I would probably use.  Distortion/fuzz falls into that category.
I've heard samples of harmonic percolator/jerkulator and like that raw barbed wire fence Neil Young kind of tone for those special romantic times.

Fuzz Face is fuzz face, and while I know I should try to find Ge transistors and spend the remainder of my days on earth swapping them in and out, I have several other hobbies which are conceptually equivalent.  I know Ge is better.  I know!  I just want to see if I can make something out of the parts in my drawer that will make SOME noise that I can use once or twice then put away.   I am not too swayed by the "cleans up well with the volume knob" argument.  I don't perform and if I am using fuzz it's because I want it to sound like fuzz.  You know the drill.

"Blend" is probably the least important item on my feature list.  I just want to have a couple of different sounds available.

[Mark Hammer]

I wasn't aiming for blending two kinds of clipping.  Rather, I split the signal into two bands that I felt merited different approaches to clipping them appropriately.  The upper band used "soft" clipping and the lower band used "hard" clipping.  The Mix control serves to function as a sort of "Tilt" tone control, giving preference to upper frequencies or lower ones.

But enough about me.  If one has two distinct signal qualities and wishes to blend them in equal or reciprocal fashion, the Mix/Balance arrangement works.  The perk is that a master output volume control allows one to alter level without having to monkey with individual mixer pots to achieve the same balance at a higher or lower level.  I find it simpler.  YMMV.

[iainpunk]

when mixing signals, having several harmonic content's mixed might muddy up the tone and lose defenition, a way to prevent that is filter mixing, having the upper frequency's of one and lower frequency's of the other or vice verse.
the only downside is that there is always some of both inputs present in the output, you can't get 100% of either circuit, but that can be fixed with two individual volume controls before the buffers, or a single pot that controls both volumes.



cheers, Iain

[Digital Larry]

I wasn't aiming for blending two kinds of clipping.  Rather, I split the signal into two bands that I felt merited different approaches to clipping them appropriately.  The upper band used "soft" clipping and the lower band used "hard" clipping.  The Mix control serves to function as a sort of "Tilt" tone control, giving preference to upper frequencies or lower ones.

But enough about me.  If one has two distinct signal qualities and wishes to blend them in equal or reciprocal fashion, the Mix/Balance arrangement works.  The perk is that a master output volume control allows one to alter level without having to monkey with individual mixer pots to achieve the same balance at a higher or lower level.  I find it simpler.  YMMV.

Ah I missed the filtering aspect of it to be honest.  I saw that something was going on but didn't bother to figure it out.  I think I'm gonna start by deciding if I want to blend or just switch between two sounds.

Don't be silly, I want to hear about what you did.  Very enlightening.