confusion over the tonebender mk II schematic/s

[mordechai]

I'm building up a tonebender with Mullard OC84 transistors, but I've seen conflicting reports over whether or not the resistor between ground and the base of Q1 should be 10K or 100K.   From what I gather, the choice has to do with leakage on Q1, and some comments I've read suggest that 100K should be used with lower leakage transistors to assist with leakage-based biasing, whereas 10K should be used with leakier transistors...and that using a 100K would cause far too much saturation.

Do I have that right? 

If I do (and even if I don't!), what would be the right choice for the OC84s I am using?  The HFe's are 80 for Q1, 93 for Q2, and 130 for Q3, and leakage on all three hover around 150 uA.

I may also build another one for a friend with Tungsram AC125s (Q1 73, Q2 80, Q3 110), all of which have lower-ish leakage (around 100 uA or a bit less).  So any advice on the proper Q1 base/ground resistor choice for this setup would be very appreciated as well.

[Arcane Analog]

All OC75 loaded pedals used 10K.

All OC81 loaded pedals used 100K.

OC75s commonly have 300+ uA of leakage and OC81s generally have much less leakage than OC75s.

Why not figure out what works best for your transistors on your breadboard?

[mordechai]

Thanks for that info.  My breadboard is currently stashed away as we are currently trying to sell our home, so the wife made me temporarily get rid of anything involving wires and little metal, plastic, or other types of messy components...so at the moment, I am working in the theoretical realm.

So -- greater leakage require a higher resistance to sort of "push" in the opposite direction?  Why not just reverse bias a germanium diode between the B and E of a leaky Q1?  Would that not help, or would that compensate too much and shut the leakag to the point where it wouldn't bias up? 

Is it possible to use a resistor value other than 100k/10k?  Maybe a compromise value might work for transistors that are kind of but not too leaky...

[Electric Warrior]

OC75s commonly have 300+ uA of leakage and OC81s generally have much less leakage than OC75s.

I don't think so. Someone on the other forum measured a couple of vintage OC81Ds. They were in the same ballpark as OC75s. Really leaky. I don't know what makes them behave differently and why they need a different bias setup, but they made those OC81D units sound very similar to OC75 and Impex S3-1T units.

I'm building up a tonebender with Mullard OC84 transistors, but I've seen conflicting reports over whether or not the resistor between ground and the base of Q1 should be 10K or 100K.   From what I gather, the choice has to do with leakage on Q1, and some comments I've read suggest that 100K should be used with lower leakage transistors to assist with leakage-based biasing, whereas 10K should be used with leakier transistors...and that using a 100K would cause far too much saturation.

The current D*A*M/Sola Sound units are set up with two 100ks, while earlier ones have the OC75 setup with a 10k and a 47k. Apparently they still sound the same, so it's possible to make them work with either setup.

I suggest you try both setups to see what works best for you and your transistors. Don't forget to swap them around until you find a sequence that works well.

Funny how OC84s work in that circuit with that little leakage. I tried low leakage trannies before with no success.
I have some Philips AC125s that are a great OC75 substitute. Good amount of leakage. I tried to get more, but I ended up with some that leaked less. Didn't do the trick... Neither did 2SB175s.

Is it possible to use a resistor value other than 100k/10k?  Maybe a compromise value might work for transistors that are kind of but not too leaky...

Sure, whatever works.

[mordechai]

Thanks for those additional insights.  A few years ago I did build up a tonebender with Tungstam Ac125s...standard gains for each transistor, but they had quite low leakage and they still worked really nicely.  I used the 100k on the Q1 base...the sound was really good, though somewhat compressed (not quite like a muff, but definitely leanjng a bit in that direction). I don't know how it would have been different with a 10k...maybe more "rumbly" and "swirly"? 

We really need to sell our home already so I can get my breadboard out again!

[Arcane Analog]

Every OC81 I have every measured (25+) has never been over 150uA. Not one. My spread sheet indicates the highest I have is 146uA. Most are 100 +/- 10uA. Also, they test lower in terms of HFE compared to OC75s. OC75s were by far wider ranging in HFE and uA whereas the OC81s are more consistent.

I know which thread you are talking about and I believe Scruffie was the only person commenting he had leaky OC81s. Considering his testing methods, I am not all that convinced that he took the time to let the transistors settle after handling them.