the sound of leakage

[destro]

just wanted to toss an inquiry out there regarding the sound of germanium leakage. I have never been fortunate enough to hear a "proper" vintage circuit with leaky transistors so am unsure as to what audio characteristic the leakage contributes to. I remember having read an article about using resistor to simulate leakage and saw it recently in an AION project but have lost the article link to reference back to. Loving these reliable Russian germaniums, but wondering out if I am missing on an essential fuzz characteristic as the leakage on most varieties I have messed with are under 100. Thanks for any input!

[blearyeyes]

Good name for a song

[destro]

I was going to put no toilet humor in the title, but where's the fun in that? I've been waiting relieve myself of this question ever since using the term "tranny" became risqué. Luckily, holding it in has not prevented the concoction of a Super Hard On, nor deterred my from delving into multiple Big Muffs. No germs in either of them!

[blearyeyes]

I was thinking of Simon and Garfunkel. 


Sent from my iPhone using Tapatalk

[Bret608]

I couldn't resist due to the thread title! 

In all seriousness though, I think it depends which classic fuzz circuit you're talking about. When I built something in the MkI/Zonk Machine vein, the circuit just didn't work well if the transistors weren't leaky, especially Q3. It would just sound mis-biased and gated. With a MkII, it wasn't so much that they had to be leaky. I just felt like with low-leakage transistors (esp. for Q1) it sounded much more "modern", kind of like a Muff with a Screamer stacked into it. Still good, just different. For the MkIII, Q1 and Q2 actually need to be low leakage. Q3 is forgiving, but I found with low leakage there, it sounded a bit too raspy and bright. As you mentioned, a resistor or trimmer of 500k to 1m between the base and collector can force it to be leakier. This is what the Soulbender does. I did that on one of the ones I built and it sounded great. I did another with no B-C resistor on Q3 but a leaky transistor and it sounded great too. Leakage there seems to make it a bit smoother with more bass. I think you can do the same thing for Q1 on the MkII, but haven't heard for sure how well it works. There's some speculation that the 47k vs. 100k on Q1 on the different MkII versions was to deal with different levels of leakage.

I hope this helps!

[stringsthings]

Good name for a song

[destro]

Thanks, Bret. I just happened to build the soulbender a la pastyface but didn't put together that the bias was actually a leakage simulator. Very cool. Appreciate the other descriptions as well.

[Bret608]

Cool! I love the sound of that version too. Shoot for something between about 1.8 and 2.8 volts on Q3's collector and see what you like best. It only starts to get a bit soft or flabby under about 1.5v. Electric Warrior told me that the two vintage units he has were at 2.1 and 2.7 respectively. In short, they varied but all sounded good as far as I can tell.