Fuzz Dog Big Fluff (BMP) Blender oscillating

[soggybag]

I built two of these Big fluff blenders, they are Big Muffs with a clean blend. Both have the same problem they both oscillate. Other than the oscillation everything else seems to be working correctly.

The sound definitely coming the distortion side, the blend works and tunes out the oscillating.

The tone control seems to filter out the oscillation or tune it in.

How can I solve this? So far I'm thinking I should get out the scope and find the transistor stage where the oscillation starts.

[antonis]

Make ALL feedback caps 1nF..

[soggybag]

Probing with the scope shows small oscillation at Q1 and it just gets louder from there.

[soggybag]

Solved! I had to bump C2 (across B C of Q1) from 500p to 1000p.

This circuit has J201 buffer that added in front of Q1 in the BMP circuit. I wonder of this changes impedance (?) of the Big Muff input stage since it's not seeing guitar pickup (10k impedance?) And now sees a low impedance buffer?

Here is a link to the build doc with a schematic.

http://pedalparts.co.uk/docs/BFPBlender.pdf

[antonis]

This circuit has J201 buffer that added in front of Q1 in the BMP circuit. I wonder of this changes impedance (?) of the Big Muff input stage since it's not seeing guitar pickup (10k impedance?) And now sees a low impedance buffer?

Let's consider Q1 stage gain as R3/R1 (open input)...
(for the sake of simplicity - it actually is (R5/R4) / [1 + (R5/R4) x (R1/R3)]..)

Without buffer, gain is R3/(R1 +10k)..
With buffer, gain is R3/(R1 + 1/gm)..
1/gm is much lower than 10k, so buffered input gain is much higher than unbuffered one..
(about 12 vs 9.5 with resistor values as per BOM..)

P.S.
In practice, gain difference is lower than above shown but, as you've realized, is enough for Q1 stage oscillation..
By enlarging C2, you've lowered Q1 stage gain at high frequencies hence settled things back to calmness..)