Troubleshooting Si TBMKIII fuzz

[mdcmdcmdc]

Hi folks,

I posted a while ago looking for some assistance with my first attempt at a vero layout (thread is HERE). It took me a bit of time to build the circuit and an even longer time to box it up. I noticed some small issues that I thought I'd taken care of through swapping out transistors but after spending a bit more time with it, those issues seem to be persisting. I'm hoping you smart kind folks might be able to help me track down what's going on.

So, a quick recap...
Here's the original schematic:

(*note: there's a backwards capacitor)

Here's the layout I landed on with the forum's help:


Build pic #1


Build pic #2


I've checked pretty closely for solder bridges, and it seems clean enough. Boxing the pedal didn't alleviate any of the issues (i thought it might be RF/etc, which usually go away in an enclosure).

It's currently built up with 2N2222a in Q1 and Q2 and a 2N5088 in Q3.
Voltages as follow:
Q1
C - 6.22
B - 1.44
E - 0.95

Q2
C - 6.22
B - 0.95
E - 0.39

Q3
C - 2.69
B - 0.53
E - 0.00

And here's what's happening as best as I can describe it...

With the fuzz pot past about 2 or 3:00 there's a high frequency whine, right around 13K. This happens without the guitar being strummed. As I back off the fuzz from full CW to about 12/1:00, that whine increases in frequency until it disappears (I assume it moves out of audible range?).

When I strum the guitar with the pedal engaged, I can see a spike around 13K happening, and there's a weird, persistent kind of oscillation happening. The fuzz itself isn't great sounding, and I assume not what it's 'supposed' to sound like based on my experience with other tone benders. It sounds (and this is I believe a very technical term) kinda farty and doesn't have the amount of saturation I'd expect from a TB MKIII derivative, especially one from God City.

The transistor leads aren't trimmed, the wires to the pots are longer than they need to be, and I'm sure running the power and input next to each other is a bad move; that said, is there anything jumping out in the schematic/layout/my build that could be exacerbating this sort of thing? Any help would be greatly appreciated. Thanks y'all!

[11-90-an]

This seems quite suspicious...

[mdcmdcmdc]

I actually noticed that same thing when I posted the photo; I scraped the track and sadly no difference.

[andy-h-h]

Usually if there’s transistors involved and it sounds farty, there’s a reasonable chance that it’s related to biasing.   Try testing the signal coming out of Q1 / Q2 Darlington pair.  it should be really loud and not farty.   If it isn’t farty, then Q3 is probably the culprit.   

[Rob Strand]

Try putting a 100pF cap across R3.
Another scheme is to put a 1k to 10k resistor in series with C2.
You can try both together as well.

The idea is to find a scheme to remove the oscillation without affecting the tone.
 

[andy-h-h]

Try putting a 100pF cap across R3.
Another scheme is to put a 1k to 10k resistor in series with C2.
You can try both together as well.

The idea is to find a scheme to remove the oscillation without affecting the tone.

What he said....     I’d try Rob’s advice before mine.   

[mdcmdcmdc]

Thanks for the suggestions - the 100pF cap scheme reminds me that I should try putting something buffered in front of it to see what happens. That's solved oscillations in other pedals for me before (rats, specifically).

As an aside, I'm not sure I've ever seen a Tone Bender with a 10M pulldown resistor at the input before.

[mdcmdcmdc]

...and running it after a boss tuner solved the issue completely - fuzz is magically fuzzy, 13K oscillation is gone, weird high frequency content/gating disappears. I'll see if I can tuck a li'l buffer daughterboard into the case somewhere.

This might be the first Tone Bender I've tried that actually sounds BETTER after a buffer.

[iainpunk]

...and running it after a boss tuner solved the issue completely - fuzz is magically fuzzy, 13K oscillation is gone, weird high frequency content/gating disappears. I'll see if I can tuck a li'l buffer daughterboard into the case somewhere.

This might be the first Tone Bender I've tried that actually sounds BETTER after a buffer.

makes quite a bit of sense!
the biggest hurdle in Vero boards is capacitance, and a total zig zag of signal path.
having capacitance from a later stage of the fuzz near to the input causes signal to feed back, there is a lot of unintentional filtering going on in a fuzz, so phase changes a lot, this causes the majority of the signal to be Negative feedback, but higher frequencies can become so phase changed that it's positive feedback.
this does 2 things, reduce gain AND oscillate at those higher frequencies
taking away the input impedance makes the capacitive coupling less, its like a high pass filter, lowering the resistance makes the amount of signal lower, making the resistance higher, especially with the added inductance from a guitar, lets more signal get fed back into itself.

cheers, Iain

[mdcmdcmdc]

I built a cornish buffer (or as close to a cornsh buffer as I could with the resistor values I had laying around... cornish-ish? corny?) and spliced it in at the input, it all fits in a 125B and I'll spend some time messing with transistors again to see what's what. The cornish design seems a little more forgiving in front of low-input-impedance fuzzes than the older boss design.

I was hoping it was a layout issue and your explanation makes a lot of sense. I may well try to lay it out again in the more traditional left-to-right tone bender fashion and see if it works a little better a bigger enclosure.

FWIW, that range control is a nice design and is worth nabbing for other TB designs if anyone's interested—way more useful imo than a tone control.