Kendrick Buffalo Pfuz mods/improvements

[ZehQuait]

Hey guys!

I've found the Kendrick Buffalo Pfuz schematic and built it on a breadboard. Right now I'm using the TBA221B instead of the NTE941M because I salvaged some of them from an old mixer. IMHO it sounds nice with a Les Paul and a Fender style tube amp, but maybe there's room for some improvements?

Has anybody here built this pedal? Any opinions and/or suggestions on some mods I could try? I think the way the opamp is biased and the low input impedance has a lot to do with its tone?

Thanks!

[Mark Hammer]

Not a whole lot to it, actually. The gain is not so high as to suggest that it depends a great deal on use of a 741 or equivalent. When the gain of these things is fairly high (e.g., >1000) very often the designer is relying on the open-loop bandwidth of the chip to force certain properties on the tone. But, as shown, this circuit doesn't really have enough gain to extract something different out of a 741 than it does out of a TL071 or whatever.

What it does not have is any sort of treble-cutting capacity, either in the feedback loop or prior to the output. In that regard, something like a 470pf cap in parallel with the gain pot might be useful for rounding off the edges of the sound.

[ZehQuait]

Mark, thanks for your help! I've been learning a lot from your posts on various forums. I've modified my SD-1 based on some of your ideas and it sounds much better now.

Back to this circuit, I was experimenting some more tonight and, although it sounds ok at first it is a little uninspiring after some time. Maybe I could say too "flat" sounding. It's a cool mild fat fuzz/overdrive that could be used sparingly to add different shades on recording. Sometimes it reminds me of one of the channels on the XXX album from ZZ Top.

While trying to make it more "alive" and usable I came up with this changes (image attached). I want something more "fuzzy" than our friend Tube Screamer, and I noticed that if the 22u cap on the (-) IN (33u on the original Buffalo Pfuz) has its value lowered the fuzz is gone and it starts to sound like an overdrive. At the same time the circuit doesn't have enough "bite" to my taste with only the 22u because it gives a full boost to the signal, hence flat. So I put another resistor and cap (100R / 1uF). Well, if I did the math right this will give me a frequency response and gain ratio similar to the ProCo RAT... I couldn't play it loud tonight but it sounded promising.

What you think about my idea? Any opinions on the frequencies / gain ratio I'm using?
What about the input impedance, should I make it bigger?
Should I expect to hear any difference between the original bias method and the one I used in this new version?

Thank you very much!

Christian

[Mark Hammer]

Well, you HAVE mimicked the top boost of the Rat, albeit with a bit less gain (not a bad thing).

Given the high potential gain (100k pot and 100R on the treble ground leg gets you a max gain of 1000x), you will want to use a socket for the op-amp so that you can sub op-amps to find something that gets you a desirable tone with the least instability.

I would recommend sticking a 1k resistor in series with the pot for a minimum gain. Given the 100R/1u network, that will result in a gain of 2x for the bass content and a gain of 11x for stuff over 1.5khz. That bright boost may be pleasing...or may not. I leave it to you to decide.

The ? on the cap is completed by working out the rolloff at min and max gain. So, for example, with a 100k pot in series with 1k, and the control maxed, a 220pf cap will get you a rolloff at 7.2khz, a 330pf gets you a rolloff at 4.8khz, and 470pf rolls off around 3.4khz.

Note that the rolloff moves upward as gain is decreased, and that the crazy amount of gain will get you wads of harmonic content as gain is increased. Note as well that, unlike the Rat, you are not using a compensating cap for the op-amp to limit high end, so the amount of harmonic content generated may be problematic. On top of that, note that the filtering effect of the feedback cap is fairly shallow.

All of this suggests that, with the component values shown, you might find a 680pf-1000pf feedback cap more useful. Using 680pf, with the gain pot at halfway (50k), you will still have a gain of 510x for the top end, and 50x for the bottom, and have a rolloff starting around 4.6khz (3.1khz with 1000pf). And with a 6db/oct rolloff, you'll be surprised how much bite you'll still have left.