Tonebender Mk. II transistor clarification

[Fama]

Yes, I know, there's a bunch of threads and a whole lot of information online. And that's really my issue.

I've seen some advice which aims for hFE around ~100 or so, with leakage around 200uA or so, like here: https://forum.pedalpcb.com/threads/tone-vendor-mk1-transistor-options.26965/#post-347091 (second post has advice with slightly less leakage and more hFE instead).

PedalBuilder had a set here https://forum.pedalpcb.com/threads/tone-bender-mk-ii-clone.22653/ with hFE around 200-350uA instead. That's ~double what Small Bear suggested as the maximum leakage for Q2 and Q3.

mybud posted here https://forum.pedalpcb.com/threads/two-tone-fuzz.28081/ about their unit with much higher leakages up to 1.2mA, but also higher hFE's. I did also find a mention of a vintage unit with leakage up to 900uA, so it's not that unheard of.

I forget which thread it was, but over on these forums someone mentioned aiming for ~200-300uA of leakage for all three (I think?), don't mind the hFE.

Then there's the approach from https://www.thegearpage.net/board/index.php?threads/right-voltages-on-tone-bender-mk2.1351035/ that I've also seen over on the DIYpedals subreddit which focuses on getting the bias in a certain range - around ~-8V for Q1, something like -0.2V for Q2 and either the same as Q1 for Q3, or maybe -4.5V?

So my question is, does the leakage and hFE matter or is the voltage what's important in the end? At least everyone seems to agree that some amount of leakage is needed, but how much and what is too much seems like a mystery to me. And if "voltage is king", what would be the values to aim for?

 

[Scruffie]

Yes it does matter and no it doesn't matter.

Is there a perfect Tone Bender in existence? Yes, but only to that individual user... without going back in time and measuring originals, we'll never know what the ideal gain and leakage was or if there ever was one.

You have to experiment and decide for your self, I've built Mk.II's with low gain transistors and they sounded nice and I've built them with 'high' gain transistors and they sounded nice.

Here's the readings of two units I built I remember being particularly nice, brackets indicating transistor position;
OC78D(2) OC81D (1)
0.072mA 118hFe
23hFe <500uA

OC81(1) OC81D (2)
0.566mA 65hFe
220hFe >300uA

OC45(3) OC81D (3)
0.004mA 106hfE
13hFe >400uA

And some voltage readings I have of original units from somewhere;
OC75 MkII (10K on Q1 base, 47K on Q2 collector)
Battery -9.01
Q1 c -8.50 b -0.038 e 0
Q2 c -0.13 b -0.076 e 0
Q3 c -8.29 b -0.13 e -0.08

OC81D MkII #1 (100K on Q1 base, 100K on Q2 collector)
Battery -9.55
Q1 c -8.51 b -0.06 e 0
Q2 c -0.13 b -0.068 e 0
Q3 c -8.72 b -0.13 e -0.07

OC81D MkII #2 (100K on Q1 base, 100K on Q2 collector)
Battery -9.50
Q1 c -8.76 b -0.062 e 0
Q2 c -0.22 b -0.075 e 0
Q3 c -8.13 b -0.22 e -0.14

If you breadboard and listen to the input stage in isolation trying out different parts you'll find a lot of variation in hiss, frequency response, gain... just remember that whatever his is generated in the first position is going to be amplified by the rest of the circuit but then some 'unusable' parts can be made usable with a 10nF to ground at the input.

There is no rule to follow, just use your ears.

 

[andare]

My explorations with the circuit (and the 1.5) led me to believe that what's important is getting around -8V at Q3c (Q2c in a 1.5) because that's what gives the TB its signature brighter tone than the FF while preserving the fuzz quality, as FFs lose it when the voltage climbs up from -4.5V, and high leakage is what yields such high voltages. Faking the voltages by manipulating bias resistors lways gives me dull tones/low gain.

I have a prototype on a breadboard that sounds great, goes a bit crazy when maxed out, has a good range of tones throught the sweep of the knob and cleans up. No filtering, a good amount of noise but it sounds dangerous and that's what counts. Just haven't boxed it up yet.

I've tried dozens of transistors and they all give different results so I've given up on trying to sound true to the OG units. I just want a fuzz that has a similar character but is usable for my needs. Sometimes that means it only works when maxed out, sometimes I want to be able to turn the knobs.

 

[Fama]

Thanks to both of you!

Partially for helping me realize that "breadboard it" is probably the best way to approach it in the future if I build more Tone Benders, and partially to help me realize that aiming for the "perfect Tonebender in every aspect" is probably not helpful.

I was somewhat comparing it to an Otsola Mk. 1 fuzz which sounds amazing (technically it's a capture of one, and I suspect not everyone will believe me, but the capture tech is so good these days it sounds identical), but then there is probably a reason he only sells a handful of pedals every now and then and they cost 400€ or whatever.

This was all inspired by me trying to get a set of transistors in the Two Notes Fuzz which worked well for me, and listening to the last ones I had in place with fresh ears, I think I'll stick with these. In Mk. 2 mode with full sustain and full guitar volume it can get pretty choked sounding depending on the fat/stock/mids setting, but there are great combinations of settings in there, and that's probably good enough for me.

I think I might try to do a "pure" Mk. 2 (probably with a charge pump as an inverter though, not a fan of batteries) at some point, but definitely going the breadboard route with that one if I get down to it.

 

[andare]

I'm familiar with Otsola and the price is not absurd given he probably buys his enclosures from Fuzzcases in the UK (60-90 pounds a pop plus shipping and import) but what's capture tech?

I follow Seeker Effects on Youtube and he has a video where he builds an MKI and he explains the System tuning and the tricks he uses to make it work with less than ideal transistors. No link ATM but it's easy to find. There's no way these builders have enough proper transistors as they're almost all gone (or Sola Sounds builders have hoarded them all by now) so they have to resort to tricks to simulate leakage.

 

[Fama]

I'm familiar with Otsola and the price is not absurd given he probably buys his enclosures from Fuzzcases in the UK (60-90 pounds a pop plus shipping and import) but what's capture tech?

I was referring to modeling stuff like the Kemper or NAM, but in this case to the Quad Cortex capture version 2. It's spot on for the ones I've done personally (the Sasquatch II, a Muff with a germanium transistor which gave it a touch of gating, plus a toggleable green ringer was captured perfectly), and it came with Otsola Mk. I captures. So I'm quite confident it sounds exactly like that, although obviously you can't capture the full behaviour with volume pot perfectly (due to the impedance not reacting with the actual pedal) - they do clean up quite nicely, but just in a "less gain way", the tone coloration that happens with an actual guitar (or at least passive pickups) is not there.

I follow Seeker Effects on Youtube and he has a video where he builds an MKI and he explains the System tuning and the tricks he uses to make it work with less than ideal transistors. No link ATM but it's easy to find. There's no way these builders have enough proper transistors as they're almost all gone (or Sola Sounds builders have hoarded them all by now) so they have to resort to tricks to simulate leakage.

Thanks, I'll keep that in mind or try finding this again once (if) I get around to building a more "basic" TB.

 

[AmplifiedPartsTom]

With the stock circuit, Q1 is the only stage that needs leakage. It won't work without some unless the stage is modified.

It can work with a wider range of transistor specs for Q2 and Q3, but it's easiest to get it to bias just like originals if you use transistors that have some.

I've measured a decent number of OC75s and would say they generally tend to be 60-90 hfe with 200-300uA of leakage. Some will fall outside that range, but if you build the stock circuit using transistors with specs in that range there's a good chance it will bias up right and sound good right away.