Trotsky Fuzz mod issue take 2

[Arnoodles]

I tried the member area but haven't had much luck. Figured I'd try this to hopefully get a bit more traffic. Here's the original post:

Hello all,

First post here so I'll do my best to describe what I've done. Tongue

I've built the Trotsky fuzz from Beavis Audio http://www.beavisaudio.com/projects/TrotskyDrive/ but with a number of changes, partially to taste and partially due to available parts. The first important difference is the transistor, for which I used a 2sc2240. I also switched the 3k3 ohm resistor for a 10k and used a 100k pot for the gain instead of the 50k. I also used a switch for the diodes to switch between a pair of 1N60's and a pair of blue LED's. Lastly I omitted the bright switch and cap. The 3k3 to 10k resistor change was preference and the rest was lack of the correct parts, although I'm fairly happy with the diodes.

This brings me to the issue. The fuzz works well and sounds great... until I get to the b and high e strings at which point there is very little distortion. I'm thinking this is down to the smaller current from the thin strings not being enough for the diodes, but I could be wrong. Question is, are there any changes I can make to get the higher strings to distort a little more? Maybe the transistor doesn't give me enough gain or some of the other components could be switched? I don't have any other NPN transistors lying around so those would need to be ordered, but I do have quite a range if resistors and capacitors to play with as well as some 1N4148, 1N5817 diodes and a few LED'sy. I should also add that I have taken it to breadboard already (maybe not the best idea given the issue) and it's quite compact so I don't have much room for additional parts. I might be able to fit a resistor or capacitor or two depending on the location but otherwise swapping will have to do unless I have no alternative but to redo the breadboard entirely.

Anyone got suggestions? It's my first pedal build and I don't have much experience with other electronics so please keep that in mind, although I'm more than willing to learn. Thanks in advance.

[duck_arse]

hello, arnoodles, and welcome. they'll all be trafficing in soon, you watch.

taking it to the breadboard is not the major step, it is the working stage between 'I wonder what happens if ....' and 'do I have enough 270k resistors for this?' the breadboard is for plugging and unplugging and plugging again cause it was better the frist time.

the trotsky is not so much a fuzz, as a slight overdrive, a variant of the electra circuit. is it a fuzz you want, or just more even distortion?
saying that, if you reduce the 680R resistor, you'll get more gain, or if you tack a capacitor (min of 1uF) across the 680R, you'll get more upper boost. on the breadboard, these trials are a snip!

[GGBB]

until I get to the b and high e strings at which point there is very little distortion. I'm thinking this is down to the smaller current from the thin strings not being enough for the diodes

At what settings? With the gain all the way up? With the 1N60s or LEDs?

If this is happening with the gain all the way up (either diode,) then something is wrong.  There should be gobs of gain - plenty to clip even the LEDs even with a lowish input signal. The 10k and 100k pot substitutions you made already increase the min and max gain available.

What pcb/layout did you use? Did you account for the pinout differences between the original transistor (KT312/2N3904/2N2222) and the 2SC2240?

[Arnoodles]

Thanks for the responses. The main issue is that the distortion isn't equal across all 6 strings. The low E through D are great, G is okay and B and e hardly distort at all. It's not more fuzz I'm looking for all around, just on the lightest few strings. This is really only an issue with the LED's. The 1N60's clip pretty well all around. This happens with the gain at pretty much any setting, even at max.

The transistor pinout is correct. I was pretty careful putting it together and checked everything, values, placement and pinout, multiple times. I drew the layout myself on a piece of paper following the schematic on the Beavis site, and again, I was pretty careful checking and rechecking.

I did notice the increase in gain swapping to the 10k resistor, which is why I used it, but it didn't have much effect on the distortion. I tried higher values but that changed the sound, not for the better, without increasing distortion on the high strings.

I'll take a look at the 680R resistor and cap and see if that helps. Hope this post cleared some stuff up. I can try to draw up the layout on the computer if that helps later on.

[GGBB]

Lowering the 680 resistor is just another way of increasing gain like with the 10k and 100k pot.  Enough gain will create more distortion on the high strings, but it will also increase the distortion on the lower strings. An emitter to ground cap does the same thing (raises gain), but you can select a value so that only higher frequencies (e.g. B string and up) have increased gain. But those same frequencies have higher gain on all the strings, not just the B and E, so the lower strings will have boosted distortion at higher frequencies. Might be just what you want but might not. EDIT: note also that with an emitter cap your gain is always max for the affected frequencies regardless of where you set the gain pot.

The problem is really the balance between the strings. One way to compensate for that is with the coupling caps. If you use a smaller cap for C1 or C2, bass is reduced off before the clipping. If you're worried that equals bass loss or tinny sound, remember that clipping is a type of compression so that the level of clipped signals/frequencies is lowered and unclipped signal remains at the same level. You'd could also think about adding the bright cut switch to tame any excess high end.

[samhay]

>This is really only an issue with the LED's. The 1N60's clip pretty well all around. This happens with the gain at pretty much any setting, even at max.

The forward voltage (clipping here) of a blue LED is much higher than a 1N60 - perhaps 10 x higher. I would suggest you try red LEDs as an alternative as they have about half the forward voltage of a blue LED and will thus clip a bit earlier, while still behaving significantly different to a 1N60.

[duck_arse]

ahh, gain pot, yeah. sorry, my 'eyes' were a bit 'funny' last night.

you might try a string of series diodes instead of the leds, and try creeping up on the not-enough point.

[Arnoodles]

Thanks for the suggestions, everyone. Witched out the LED's for red ones and that helped quite a bit. Still doesn't clip a ton but it's now much more a slightly distorted boost (and very loud at that) which is much more consistent across the strings. Not really what I was going for, but it sounds nice and it's a good contrast to the fuzzy 1N60's. Kind of a happy accident.