Tone Bender Q3 Leakage, work around possible for a low leakage germ?

[Locrian99]

Hello all,

So I put together a soul bender (ppcb tone bender mkIII or the fulltone version of it basically) on my breadboard.  Using a couple GT310B in the low 70's for HFE and minimal leakage, and a 3ax31c hfe 134 and a leakge of 175ua.   It's absolutely fantasic!  But I only have one of those 3ax31c's, I've got a bunch of GT310b's right in that hfe range so I tried one of those in there, it's not terrible but its kind of shrill and well not inspiring me to break into good times bad times the moment I hit the first chord like the 3ax31c is.   Is there any work around for the leakage, to make a lower leakage transistor sound better here?  I realize probably not, but hoping

[bmsiddall]

You might try a high value resistor from C-B (plus a small cap in parallel if it's still too shrill) to simulate leakage. Take a look at General Guitar Gadgets' silicon 3 knob tonebender schematic- uses a 1M trimpot to dial in q3.

[mzy12]

If it works, a better way to simulate leakage would be to get a germanium diode and reverse bias, going from the base to connector. Whether the transistor is PNP or NPN will dictate the direction of the diode.

[Locrian99]

Thank guys I'll try those out.

[andy-h-h]

Are you sure it's the leakage?  I sometimes use an OC45 as Q3, which are both low gain and low leakage, and they often sound really good.

Is the circuit that you're using stock?

[mzy12]

Are you sure it's the leakage?  I sometimes use an OC45 as Q3, which are both low gain and low leakage, and they often sound really good.

What counts as low-leakage here? Just out of curiosity
Tonebender MkIII's, according to Aion FX's documentation, require about 100uA leakage to work. A lot of those Russian GT transistor are (reportedly) pretty good for leakage, as are your OC45's. I wonder what the OP's leakage measurements actually are?

[aion]

Tonebender MkIII's, according to Aion FX's documentation, require about 100uA leakage to work.

"Require" is probably not the right word... I just collected this somewhere along the way as a good ballpark recommendation and it's never failed me! I know you can go lower, but I don't know how far. I haven't personally done any tests to figure out where the cutoff is, or what happens as you approach it.

[Big Monk]

Tonebender MkIII's, according to Aion FX's documentation, require about 100uA leakage to work.

"Require" is probably not the right word... I just collected this somewhere along the way as a good ballpark recommendation and it's never failed me! I know you can go lower, but I don't know how far. I haven't personally done any tests to figure out where the cutoff is, or what happens as you approach it.

I've gone down to 10 microamps with my favorite General Electric 2N169 units in "No Bias" stages.

[Locrian99]

I'd have to go head out to the garage to where all my transistor testing papers are by the russian transistors I was trying to use were testing at VERy low leakage.   They were the GT310B's.   Work awesome for both Q1 and Q2 in a FF.   I was comparing to a 3AX31C with a leakage around 175ua (134 hFE) as q3 in the Tonebender and just was shrill, not absolutely terrible.   Ive recently acquired some different russian germs I'll try out, who knows maybe its just the type as was mentioned leakage is about the same as the 310b's very low.  Haven't had time to try them in the breadboard yet...