breadboarding (fuzz face, tone bender)

[owlexifry]

seemed like i was dragging @Stickman393 ’s thread out of scope (germanium transistor testing and leakage), so i’ve removed my comments and started a new thread here…
>>>>>>>>>>>>>>>>>><<<<<<<<<<<<<<<<<

very recently i've just begun trying out breadboarding for the first time.
(wanted an excuse to test out some MP38 transistors i grabbed last year)

- first did an LPB-1. yep cool. it boosts. quite nicely too.
- next logical step was to try a fuzz face circuit.

- impatiently did a quick and dirty Hfe check with the DMM and picked x2 that looked stable and 'high enough' ?? (some bullshit creative logic going on here)
- biased Q1 collector to 0.7v (39K), Q2 to 4.5v (6.5K on trimmer).

result:
- fuck yeah. great sounds.
- seemed to be fuzzing nicely across the sweep of the gain
- sounded really good around 3'oclock to max on the fuzz/gain control.

So after @PedalBuilder pre-emptively crushed my dreams about the prospects of finding some real good NPN Ge, I thought yeah maybe I should actually measure these MP38s properly.

to measure gain and leakage i've been using this little veroboard circuit layout based on the rgkeen method
(http://www.geofex.com/article_folders/ffselect.htm)


so last night I did some R.G.Keening with my sample (n = 20)
- Hfe mostly around ~30.
- leakage all under 100 except for x2 (438uA, 655uA)
(i was gonna be helpful & type this up, but maybe it's a good idea to expose my potential errors)


well guess what i had running in Q1 and Q2:
Q1 - 1.082v / 1.217v : . 13.5 Hfe, 438uA leakage (#18)
Q2 - 0.108v / 0.403v : . 29.5 Hfe, 44uA leakage (#19)

so of the 2 / 20 that had significantly higher leakages, of course i picked one of them for Q1.
lol.
how is it possible that this sounded any good?

then it got more confounding.

so i thought ohh geez what if i used something in Q1 that isn't completely fucked (well 13.5hfe/438uA leak seemed pretty fucked to me...)

picked a couple different ones to try:
Q1 - #4 - 32.8 hfe/ 48uA
Q2 - #9 - 38.6 hfe/ 58uA
(re-checked bias at both collectors and got pretty much the same numbers as previously, Q1 0.73v / Q2 5.0v)

result:
- this actually sounded worse.
- more gated.
- had to pick harder to get 'the response'
- at low gain settings it almost sounded ring modulator-ish
- objectively a pretty good fuzz tone and sounded much the same i guess, but everything just felt like it was noticeably worse.

>>>>>>>>>>>>>>>>>><<<<<<<<<<<<<<<<<

so i've just read this statement:
"for a transistor to function well in an original fuzz face circuit is not the correct gain, but rather the correct amount of leakage to bias the base."

honestly i hadn't paid a lot of attention to the voltage bias at Q1 base.
seems to me, since these MP38s are so low gain (~30), perhaps the leakage may not be enough in some to bias Q1 base properly?
(and maybe that's why that super leaky transistor in Q1 i first tried sounded alright...)


i did notice Q1 base was about 0.1xx V at one point (after inserting the 'properly measured' MP38s)

so, it looks like i need something in Q1 that will leak enough to get the bias at Q1 base up to at least 0.2V.

>>>>>>>>>>>>>>>>>><<<<<<<<<<<<<<<<<

so to update from that:
- checked Q1, and yep, the base is idling at about 0.100V.
- this seemed too low. (target spec 0.2V)
- in an attempt to get Q1 bias voltage higher, last night after work i went ahead and tried every MP38 in Q1 to see what I could get.

- well, this sucked - half of them resulted in 0.08-0.09V, the other half about 0.1V, and a couple outliers.
- picked one with the highest base voltage at Q1 (0.118). Incidentally this was one of the lower gain (20 Hfe) MP38s.
- re-biased the collectors.

(had to increase R1 a fair bit to get Q1c bias voltage down - with 39K, it was idling @ 1.2V - too high)
- so i ended up with about 0.12V @Q1b, 0.75V @Q1c, ~4.5V @Q2c.

had time for a quick test this morning before leaving the house
[Q1 = ~20Hfe, Q2 = ~ 38Hfe]

(sorry no time for SM57+pre)

so acknowledging that the transistors are very low gain (20 / 38), much below the recommended values (70 / 120), it's interesting that this seems to be working at all.
i'm no fuzz connoisseur, i have no idea what i'm supposed to be looking for, all my guitars have humbuckers, but so far this sounds alright to me.

assuming it will probably sound better if i use transistors with higher gain? anything else i should be looking at?

 

[Feral Feline]

If it sounds good to your ears it is good.

 

[Amplified Parts]

Don't worry about the exact Q1 voltage. Q1 base voltage will be equal to Q1 Vbe since the emitter is grounded (0V). The Vbe will change some depending on how much current is flowing into (or out of, depending on NPN/PNP) the base, with more current making for a larger voltage drop, but it's not really something you can easily change by adjusting the circuit resistances - it requires a drastic change to go from ~0.1V to 0.2V. You often hear 0.3V for the nominal voltage drop in germanium, but with the low currents in the Fuzz Face (and most fuzz circuits), the Vbe is much more likely to be around 0.1V or even lower. 0.2V seems high. I have an original NKT275-loaded Arbiter FF here next to me and Q1's base measures at 0.096V at room temperature (and Q2C biases to around -5V and it sounds great).

Regarding this statement:

"for a transistor to function well in an original fuzz face circuit is not the correct gain, but rather the correct amount of leakage to bias the base."

That's likely referring to current flowing into the base, rather than the base voltage. But the bias voltages depend on both (leakage and hfe), and Q1's specs have a much larger effect on the bias voltages than Q2.

I do wonder about your original Q1 measurements with the high leakage, it seems like those might be off. I think you'd have a difficult time getting that biased correctly, at least without drastic resistance changes, and I've got some doubts about whether it would sound good. When you swapped the transistors out for ones with more reasonable leakage, you said "(re-checked bias at both collectors and got pretty much the same numbers as previously, Q1 0.73v / Q2 5.0v)" - is that without adjusting the trimmers? If so, I definitely think the 13.5 Hfe, 438uA leakage measurement has to be off. Also, make sure you're driving it directly with passive pickups if you can. Any sort of buffer between your guitar and the FF input is likely to cover up some of the audible effects of using a very low hfe transistor in Q1.

I agree with Feral Feline that if it sounds good and works well, you don't really need to worry about it unless you're interested in the "why". Although high transistor leakage (and particularly Q1 leakage) does make the circuit more temperature-sensitive.

 

[owlexifry]

Don't worry about the exact Q1 voltage. Q1 base voltage will be equal to Q1 Vbe since the emitter is grounded (0V). The Vbe will change some depending on how much current is flowing into (or out of, depending on NPN/PNP) the base, with more current making for a larger voltage drop, but it's not really something you can easily change by adjusting the circuit resistances - it requires a drastic change to go from ~0.1V to 0.2V. You often hear 0.3V for the nominal voltage drop in germanium, but with the low currents in the Fuzz Face (and most fuzz circuits), the Vbe is much more likely to be around 0.1V or even lower. 0.2V seems high. I have an original NKT275-loaded Arbiter FF here next to me and Q1's base measures at 0.096V at room temperature (and Q2C biases to around -5V and it sounds great).

Regarding this statement:

"for a transistor to function well in an original fuzz face circuit is not the correct gain, but rather the correct amount of leakage to bias the base."

That's likely referring to current flowing into the base, rather than the base voltage. But the bias voltages depend on both (leakage and hfe), and Q1's specs have a much larger effect on the bias voltages than Q2.

this is quite helpful to know.
thanks for writing all of that.

I do wonder about your original Q1 measurements with the high leakage, it seems like those might be off. I think you'd have a difficult time getting that biased correctly, at least without drastic resistance changes, and I've got some doubts about whether it would sound good. When you swapped the transistors out for ones with more reasonable leakage, you said "(re-checked bias at both collectors and got pretty much the same numbers as previously, Q1 0.73v / Q2 5.0v)" - is that without adjusting the trimmers? If so, I definitely think the 13.5 Hfe, 438uA leakage measurement has to be off.

not sure what happened with those two outliers but those were the numbers i was getting using the RG Keen method. idk. everything else seems to be testing normally so i’m not too bothered..
i suspect the ‘bad sound’ i had experienced was most likely improper bias voltage(s). i was rushing. after i had another look and re-checked/adjust bias, it was sounding pretty good.

>>>>>>>>>>>><<<<<<<<<<<<<<<<

update:
since I had started with NPN germaniums, I was intending to compare with silicon transistors, so I picked and measured some.

- BC108B
i don't have my notes on these available right now, but i remember these all displayed much higher gain than i had anticipated. something like 250-300Hfe.

- 2SC1815
to my surprise these are the lowest gain silicon transistors i could find in my stash.
- 2N3904 seems to hover around 200-250.
- all my BC108Bs were around 250-300 (i think..)
- and then these 2SC1815s are all around 160-170...
anyway, since i was already measuring a handful of these for a superfuzz build, i picked a couple (159Hfe, 170Hfe) for the fuzz face.

results: (BC108B vs 2SC1815 vs MP38) - adjusting bias where necessary
BC108B:
- everyone suggests these for fuzz face circuits
- that's why i even bought a bag from tayda at all
- the ones i had are very high gain.
- not a bad sound, but sounded the most compressed, splatty, crunchy, and aggressive. perhaps too much.

MP38:
- Q1 ~20Hfe, Q2 38Hfe - way below specified gains.
- despite that, offers a full fat fuzz crunch, not as aggressive or crunchy as the silicons, noticeably smoother, less high end.
- i suspect i'd like these better if they were higher gain.

2SC1815:
- no one ever suggests these for fuzz circuits
- i bought these for Wilhelm Scream + Extreme Scream builds that I haven't done yet.
- i liked the sound of these the most.
- for lack of a better description, i guess this was a happy medium between the sounds of the BC108B & MP38 described above.


made an order a couple weeks ago, gonna be trying out some MP38A transistors soon.
hopefully these will actually be in or above the 90-110 Hfe range


>>>>>>><<<<<<<

TONE BENDER TOME

of course the next logical step was to try out a tone bender.
mkii seemed to have the most appeal.
i followed this schematic:


decided to try it first with x3 2SC1815 transistors.
- had a few confounding moments when i couldn't get any sound.
- great thing about breadboarding is the input/output jumpers serve as an instant audio probe.
- the rest of circuit was working fine, but Q1 wasn't giving any sound.
- Q1 base had almost zero bias voltage, and collector was idling at 8.9x v.

- then i looked at this 'deimos (legacy)' schematic and had an 'aha' moment...

correct me if i'm wrong:
i) the original schematic depends on Q1 having a certain amount of leakage from collector to base.
ii) with a modern silicon device (2SC1815), there is none of this leakage.
iii) so there needs to be a biasing resistor. which is why AionFx suggests the above with RX1.

with an added 1M5 resistor 'leaking current' from Q1 collector to base -> this solved everything for the 2SC1815.
(during troubleshooting, i swapped in a MP38 and it worked fine at Q1 without the 1M5 resistor, which was initially very confusing until i had the moment of clarity).


- followed these specs as close as i could:


- tried different transistors, as i did with the fuzz face (BC108B, 2SC1815, MP38), and with combinations, (e.g. BC108B-Q1, 2SC1815-Q2,Q3).

result:

- meh.
- yep it's got a lot of gain.
- nothing that pleasing or memorable about the fuzz.
- idk, this is not the fuzz for me.
- fuzz face shits bricks on this.

i can now see why there are 420 billion fuzz face-inspired variants as compared to only 1 billion tone bender mkii variants