Adding balance input and outputs to a fuzz?

[Chugs]

I appreciate all the replies and suggestions.

How about this?

[bool]

IMHO you are throwing away one chip. Do you really need a Vb buffer? Could be done with another 3904 if really really necessary but I doubt so. And the IC2B  (or IC3A depending how you look at it) is totally redundant. IMHO you could easily make it with a quad TL074 for better looks. Personally I would reference the C7 directly to the GND and rework the snippet around the VOL pot. Also I would rather waste another RC and filter the Va a bit more just for the fuzz circuit. Also I would connect the diode across the C1 in order to not get it killed and not doing its job. My gut feeling is that the R12 will need to be a bit bigger value, but that's hard to tell until you "vibe" the unit and hear how it works in real life.

[Chugs]

Thanks for the suggestions. I will thinker with them when I get a chance.

I was thinking of sticking a tone control between IC2B and IC3A otherwise yes, one opamp could do the what both are now doing. Should have mentioned that in the previous post.

10K for R12 works pretty good in this application on similar buffered fuzzes I have built.

[PRR]

Looks correct.

Some fluff could be omitted.

R23 looks pointless.

I might like more gain from VOL to output, though this will depend on settings and studio practice.

The Vref buffer only feeds hi-Z opamp pins plus the several-K of IC2B NFB. However that NFB loop has a cap, so it could return to GND without upsetting DC levels. Now the Vref load is negligible. I would however swamp it with a big cap just to be SURE audio does not sneak-back through it. (Probably over-kill; if I was making a million of them I would have to compute and see if it could be a penny cheaper.)

IC3A does nothing that IC2B can't do just fine.

(Saved two 19-cent opamps, 8 holes.)

I like the 80Hz-6KHz cutoffs in IC2B. I suggested new values for more gain (same Hz) but this will have to be trimmed on studio-test.

[ashcat_lt]

me> some knock-down at the end

For example-- the FuzzFace:
http://fuzzcentral.ssguitar.com/fuzzface/fuzzfacepnpschematic.gif

Bueno.  I'm not so familiar with these schemes (even though ubiquitous and "essential") because...transistors...  Took me a minute to sort out what you were looking at, but I got.  Thanks.

[Chugs]

Ok, rough draft V2. I have made some changes and incorporated some of the suggestions. I still have the third IC at the moment as I have added a tone control and isolating the tone control from the fuzz circuit worked better. Just a basic low pass filter at the moment but I think I might add something with a more interesting eq response when I get a chance to tinker some more.

[bool]

D1 will still fry if you connect the power the wrong way. Move it after the R1. (If I was your boss, you'd loose the job).

Now onto the creative aspects of the circuit: The point is, that you already have tonal shaping incorporated into the IC2B/R7/C7/R8/C8 snippet. You also have a low-pass added right onto the fuzz snippet with C15. Now what would be a "good deign practice" would be to reference the cold side of the C15 to the ground instead to the Va. Try it it will sound a little different, "more effective" in filtering because you don't have all the routing trough power rails and the main filtering capacitors (C1, C2).

I would also reduce the values of R3, R4 to say 33K or 47K (you have all those in-circuit already) for a bit better "grip".

So the plan to further optimize the circuit would be to move the variable low-pass filtering directly after the "fuzz" circuit, before the C16. Just connect the 10K tone pot (and the C17) directly to the R17/C15 junction, before the C16 and tweak the values to get it to sound "the best".

IC3 goes out, you can use a TL074, with a cleaner layout etc, and you stay in business.


Ps.: For an "active" tonal shaping, you already have the wien-bridge "EQ" in IC2B. So you can easily make that part "variable" with not much added (unnecessary) complexity.

[Chugs]

I will move D1 and try C15 to ground and try the tone control variations.

What would be a good way to make the wein-bridge variable?

[bool]

- Change R7 to 2K4
- Change C8 to 4n7
- Add a 47nF cap to IC2B "neg" input
- Add 1K resistor in series with the 47nF cap
- Add a 10K pot (as a rheostat) in series to the 1K
- Connect the other side of said pot to the ground.

This should look a lot like the "Rat" gain network (dual legs) now btw.

So there is a basic "presence" control that should sound quite good.

[PRR]

> Move it after the R1.

Now R1 fries. (Unless a 2 Watt part.)

Yes, this is still better because with Diode facing the hostile world, the diode will try to clamp but a sufficiently beefy wart will fry it. Then everything else gets full reverse or AC voltage, and a LOT of parts die. With diode after R1, R1 fries, it all goes dead, the only permanent damage is a cheap resistor (also good evidence of customer abuse).

Note that the common 1300mA and 1600mA board-power warts can melt a 1N400x diode. If it really is limited to 1300mA then it may take all day to melt the 1A diode, but sometimes stuff gets left plugged-up wrong all day. And there are beefier warts, both pedal-market and strange lumps found in GoodWill.

> reduce the values of R3, R4

They feed only naked FET opamp pins. Maybe just one buffer, maybe (unbuffered) 5 or 6 pins. The total load is hundreds of Megaohms. IMHO 100K is plenty low. However it is quite un-critical, and I'd probably use any orange-stripe resistors I had two of (those pesky 75K that never get used).

[bool]

PRR ... you forgot to add that in case the R1 fries instead of a diode, no other parts will fry in the box - unless the diode is faulty to start with - because if a diode goes down, R1 will not stop chips to die and caps to bulge ... etc. Because in the circuit are parts that will also cause other types of damage when fried (electrolytic caps, for instance), frying a single resistor is ultimately a much more environmentaly friendly and responsible practice.

A 100uF cap (in the Vb biasing circuit) will usually "do" some current flow (leakage) so having a bit smaller biasing resistors will make the circuit more "future friendly" when/if caps start to get older and more temperamental. Unless you use tantalums ...

[bool]

( ... and we haven't even touched upon what could happen to a wallwart if it's a funky one ... )

[ashcat_lt]

If you want to be actually safe against reverse polarity induced explosions, put the diode in series with the resistor.

[bool]

In essence yes, but you still have to deal with inrush current. For low-power circuits, a R-into-reverse-biased-D may be better. A standard 221E 0.6W metal resistor into a 1N4148 will last very long for very cheap. Most electros will forgive the 0.6V reverse bias, tantalums included.