Fuzz 3 band eq and recovery stage help.....?

[brokenstarguitar]

Hey guys! Hope youre all doing well and building awesome shit as usual! Im trying to find a design/layout for a decent sounding fuzz with a 3 band eq (lows, mids, highs). Ive been looking for quite some time without success.

Now Ive also read I could add a 3 band eq at the end of any fuzz circuit but would need to add a recovery stage after it but tbh, Im not even sure I know exactly what a recovery stage is. I mean I know what it does and why its needed but I never had to design one before. How do I calculate what Ill need? Or is it as simple as taking one like from the big muff as pictured?

[GibsonGM]

Yes.

(taking it from the Big Muff like you show! ) 

You could insert a "TMB" tone control and then follow it with that recovery stage...do you have "Duncan's Tone Stack Calculator"? (free download)

[brokenstarguitar]

Yes.

(taking it from the Big Muff like you show! ) 

You could insert a "TMB" tone control and then follow it with that recovery stage...do you have "Duncan's Tone Stack Calculator"? (free download)

I wasnt aware of the "TMB" so Im going to do some reading through all that. Thank you. I havent decided on what kind of fuzz yet, besides an OP amp version. Any advice there?

[antonis]

Im not even sure I know exactly what a recovery stage is. I mean I know what it does and why its needed but I never had to design one before. How do I calculate what Ill need? Or is it as simple as taking one like from the big muff as pictured?

My apologies for Sir Mike's lazyness..
(he covered the whole issue with a single "Yes"..)

Recovery stage "recovers" what has been "lost" through previous stage(s)..
(in case they're passive tonestacks, like yours..)
It isn't difficult to design such a stage (either discrete transistor amp or op-amp one) as long as you know/calculate/estimate the signal "loss" (attenuation, either in dB or in voltage ratio..)

If you aren't familiar with dB term [20*log(V_out/V_in)], you can calculate or measure your specific EQ voltage loss (V_out/V_in) and simply compensate it with an amp stage..
e.g. your scheme recovery stage exhibits about 13dB voltage gain.. 20*log(V_out/V_in) where V_out/V_in = R6/R4 (roughly, w/o consideration of bias configuration loading, intrinsic Emitter resistor & Volume pot R6 loading..)

Of course, passive EQ's don't exhibit "constant" load, hence fix signal loss, due to their reactive elements (caps and/or inductors) so you have to decide the frequency band loss you want to recover, which in most cases should be mid-band, to prevent large output volume fluctuations (more than +/- 6dB, say - up 'til +/- 3dB are practically inaudible..) for extreme low & high end settings..

P.S.
A fulsome recovery may result in extra - unwanded - distortion..

[antonis]

I havent decided on what kind of fuzz yet, besides an OP amp version. Any advice there?

As long as pinkjimiphoton doesn't read that, we may consider ourselves in peace of mind.. 

[brokenstarguitar]

Antonis, thanks for the explaination. Im gonna breadboard this and check the eq voltage loss as you mentioned. I will report back soon. Thanks guys, its very appreciated!

[GibsonGM]

I will come back for more than "Yes" after the OP messes around for a while, bangs his head on the wall, and now wants to know "Why".    The answer IS "yes", as the recovery stage has a volume control.    Problem solved.

I find that many people don't want to know about insertion loss and fine-tuning things like this - they want a fast fix, a quick answer.   If he looks and finds that overall, he is losing 20 dB, then we can help him tailor a BJT stage to do 25dB gain and add a volume pot.   

Can't give a book in one forum post   

[antonis]

Im gonna breadboard this and check the eq voltage loss as you mentioned.

Alternatively, you can post your EQ schematic and I ensure you some willing guys could simulate its impedance attenuation.. 

[ElectricDruid]

...or we could save OP the pain and point them at an active three-band EQ design that won't need the extra messing about?
Especially seeing as they already mentioned op-amps. A three band baxandall would do this.

I never understood why people put a passive EQ in, and then use an active stage to make up the losses. Why not use an active EQ and *avoid* the losses?

[pinkjimiphoton]

look for the stiffler's mom project i posted.
big muff pi with 3 band active bax eq, and a final gain/recovery stage. possibly the loudest big muff on the planet.

boards for it, if ya want one, should be available from phil and dino at deadendfx.com

[antonis]

I never understood why people put a passive EQ in, and then use an active stage to make up the losses. Why not use an active EQ and *avoid* the losses?

I never understood the difference (if any) between the two of them.. 

[ElectricDruid]

I never understood why people put a passive EQ in, and then use an active stage to make up the losses. Why not use an active EQ and *avoid* the losses?

I never understood the difference (if any) between the two of them.. 

You're a smart guy, Antonis. I bet you do

The active versions tend to have less interaction between the controls, which makes them easier to understand and use, in my view. I suspect (but don't know) that active designs are quieter because you'd typically build an active design for unity gain overall (so no net gain in noise really, except perhaps at some boosted frequencies) whereas a passive design loses a ton of volume which then has to be replaced by boosting everything significantly, including the noise floor. Not losing 10dB of treble is *not* a 10dB treble boost!

Aside from that, to me it seems like better engineering. Rather than bodge a passive control on the end of something, discover it doesn't work, so stick a buffer in front of it, and then discover its too quiet and stick a gain stage after it, you could actually design something that works straight away.

Ultimately though..both ways work, so it *is* just a matter of taste.

[Fancy Lime]

Hey there,

does "besides an OP amp version" mean you want an opamp version or you do not want an opamp version? Sorry my English.

...or we could save OP the pain and point them at an active three-band EQ design that won't need the extra messing about?
Especially seeing as they already mentioned op-amps. A three band baxandall would do this.

I never understood why people put a passive EQ in, and then use an active stage to make up the losses. Why not use an active EQ and *avoid* the losses?

Well, depending on what comes before the tone stack and, to a lesser degree, what comes after, active or passive may be more useful. Active Baxandalls usually have a gain of 1 when the controls are at noon and then you have symmetrical boost and cut of each, say, 20dB (can be more or less of course depending on the design). If such a thing runs on 9V single supply with a TL072 and you feed it with a signal that is already close to the rails (as many discreet transistor fuzzes do), then you will get horrible sounding distortion from the tone stack and no ability to boost anything. If you install a James tone control here (functionally pretty much a passive implementation of the Baxandall), with a buffer (not a recovery stage!) after it, you get a useful tone control and still dBs galore on the output. You could of course also design the active Baxandall to reduce the overall volume in such a way that you get a gain of 1 with the controls at max (as discussed here: https://www.diystompboxes.com/smfforum/index.php?topic=122679.0) but that is a good deal more difficult than just using one of the many James tone control designs that have been used for half a century.

If, on the other hand, you have something with diode clippers to ground, then you have a very small signal swing on the output of your clipping stage. So even 20dB boost in the tone stack are not going to make it clip. In this case you either want a passive tone stack with a recovery stage after it, or an active tone stack. I would prefer the latter as it is easier to simulate and therefore easier to predict during the design process.

Also important: When using an active tone stack, consider placing the volume control before the tone stack and make the tone stack opamp (or whatever active element) work double shifts as output buffer as well. Also helps with avoiding clipping.

Here's a good tool for designing a James tone stack:
http://www.guitarscience.net/tsc/james.htm
Unfortunately, I don't know one for a Baxandall. If you can program Julia, I can send you some code I am working on that is ultimately supposed to be able to do that but it is far from finished. So much to do, so little time...

BTW, I would go with a 2-band Baxandall or James. These are effectively 3-band controls with 2 knobs. You get more sonic flexibility with these than from a Fender/Marshall 3-band. A third band gives little added benefit to a Bax or James in my opinion. The law of diminished return on investment really bites hard here. I have never run into the situation where I missed a mid band on a Bax in an *amp*, so having one in a *fuzz* seems not worth the trouble to me. To make the third band really have added value, you might want to consider making it (semi)parametric. That would earn you extra nerd points on a fuzz and might indeed increase the sonic flexibility.

Cheers,
Andy