HFE, leakage and tuning a germanium fuzz - pointers needed for a Zonk Machine

[holio.cornolio]

Hi All,

I have a few questions, all vaguely interlinked as I'm about to start building 3 or 4 fuzzers for my own experimental use, but my understanding of trannies is very limited and may be confusing my plans!
I plan building a Zonk Machine circuit, a Marshall Supa Fuzz type thing, and also messing around with some silicon trannies in those circuits. I've been reading up on germanium fuzzers, and all the circuit notes I've read imply that these circuits kind of rely on the leaky nature of germanium to get the circuit working.
I've also noted that typically the hfe of germs, seems to be way below that of silicon equivalents, so I wondered if the relatively low hfe was the key to these trannies sounding good in these circuits or is it just measured differently? I ask because I have some very low gain 2n5133s which I got from small bear for a big muff build, but were just too low gain to cut it (typical hfe way below 100) and I'd thought that I could just pop these into the Zonk circuit, switch the polarity and hey presto - a decent sounding silicon zonker? But I guess that's not the case? Would a bias adjustment compensate for the lack of leakiness of the silicon, or am I just confused?

With the Zonk, everything points to trannie selection being paramount. I have a few old PNP germ trannies to try, some OC72 (hfe between 56-98), a few AC128 (hfe around 120) and some 2sc270, OC44, OC71. My question here is, is the actual transistor selection important for each position, or is it more important just to get the hfe and leakage right? From what I've read, the zonk seems to prefer a low gain (low noise) trannie for Q1, and higher gain in Q2 and something in the middle for Q3. So I could use my OC71 or OC72 in Q1, and AC128 in Q2 and a higher gain OC72 in Q3?

Of course I'll socket and experiment, but the experience of others always helps to tune that experimentation!

[mac]

I've also noted that typically the hfe of germs, seems to be way below that of silicon equivalents, so I wondered if the relatively low hfe was the key to these trannies sounding good in these circuits or is it just measured differently?

Most of these circuits are very old and so the germs used, which were low gain in those days.

I ask because I have some very low gain 2n5133s which I got from small bear for a big muff build, but were just too low gain to cut it (typical hfe way below 100) and I'd thought that I could just pop these into the Zonk circuit, switch the polarity and hey presto - a decent sounding silicon zonker?

Yes. You could get very good sounding Tonebenders, FFs, Zonks, etc.
Many members are very happy with their low gain silicons clones.
YOu can build these pedals with high gain silicons but I prefer low gain ones, more germanium feel IMHO.

But I guess that's not the case? Would a bias adjustment compensate for the lack of leakiness of the silicon, or am I just confused?

Yes.

See Tonebender II Q1 or Zonk Q3, they use leakage to bias. Every 100ua of leake through a 10k resistor is 1v drop so you need a moderate leaky germ to bias properly. Silicons leakage is almost zero so you need to feed the base with a big resistor to place the collector to a suitable value, 7v - 8v. For example, in a Tonebender II, a 1M resistor from collector to base will put the collector at about 7.8V if hfe=100. Note that very low leakage germs, say below 50ua will need also a big resistor to bias properly!

Zonk Q1, same as above.

The FF is less tricky. Just use some low gain silicons and a 10k trimmer at Q2 collector. Most likely it will bias at 4.5v near 5k.
Same for Tonebender II FF stage.

Tonebender 3knob, I wrote about biasing the germanium thing but I cant find the link! 
YOu can read these links to know how to bias Q3 if using a silicon:
http://www.diystompboxes.com/smfforum/index.php?topic=80590.0
http://www.diystompboxes.com/smfforum/index.php?topic=84315.0

With the Zonk, everything points to trannie selection being paramount. I have a few old PNP germ trannies to try, some OC72 (hfe between 56-98), a few AC128 (hfe around 120) and some 2sc270, OC44, OC71. My question here is, is the actual transistor selection important for each position, or is it more important just to get the hfe and leakage right? From what I've read, the zonk seems to prefer a low gain (low noise) trannie for Q1, and higher gain in Q2 and something in the middle for Q3. So I could use my OC71 or OC72 in Q1, and AC128 in Q2 and a higher gain OC72 in Q3?

RG has a great article about the Zonk.
I'd say that leakage is the issue here. And the weather!!!

I was thinking about modding bias resistors of the zonk or TB I  to make them more independent of the characteristics of the transistors without altering the tone.

mac

[holio.cornolio]

Thanks Mac,

that's really useful info. It's also nice to know that I'm gradually gaining a bit of understanding about these circuits. And I do mean a bit...

[LucifersTrip]

The reason I haven't done the Zonk yet is because I wasn't sure which was the real schematic...When researching it for a build, I found at least 2 diff versions.

Which ones are you guys using?





Your ge/si question leads me to another question: "I've also noted that typically the hfe of germs, seems to be way below that of silicon equivalents, so I wondered if the relatively low hfe was the key to these trannies sounding good in these circuits or is it just measured differently?"

We know that ge's have leakage so they are measured differently and the leakage effects how the circuit is biased. But, lets say you have 2 transistors (1 ge/1 si) with identical hfe's and leakages, what is the next most important attribute that affects the circuit? The forward voltage between BC, BE ? What else...

[holio.cornolio]

I think they're both potentially accurate, as I understand that there were some variations between units. the only thing  that I'm aware of is that the 0.47uf cap on the first schem should actually be 0.047, but apart from that I think both schematics represent variations present in original units. I think....

only think is good luck finding the A02 650. I think any low noise / low gain PNP should be ok.
Based on my limited understanding of trannies, you would be unlikely to find similar leakage levels in any Silicon transistor.

[Electric Warrior]

If you don't know which schematic is correct, just trace a pedal and make your own schematic! Pics can be found here: http://decadeoffuzz.com/index.php?/project/john-hornby-skewes-pedals/

[LucifersTrip]

I think they're both potentially accurate, as I understand that there were some variations between units. the only thing  that I'm aware of is that the 0.47uf cap on the first schem should actually be 0.047, but apart from that I think both schematics represent variations present in original units. I think....

yes, probably both variations of originals...I wonder if there's even more. I'll probably wind up just trying both.

only think is good luck finding the A02 650. I think any low noise / low gain PNP should be ok.
Based on my limited understanding of trannies, you would be unlikely to find similar leakage levels in any Silicon transistor.

that's right...I've only measured a few si's that have had any considerable leakage (10-50ua), and they were probably bad ones. the lowest leak ge's I have are in that range (10-50ua).

tho my question still remains:
We know that ge's have leakage so they are measured differently and the leakage effects how the circuit is biased. But, lets say you have 2 transistors (1 ge/1 si) with identical hfe's and leakages, what is the next most important attribute that affects the circuit? The forward voltage between BC, BE ? What else...

Actually, it doesn't have to be 1 si and 1 si. Let's say it's 2 si's or 2 ge's with identical hfe & leak. What's the next most significant attribute? And does that attribute effect biasing or just the sound?