Fuzz Face - Does transistor selection really matter?

gtfields13

gtfields13

Member
I think the answer is, not much...

I've been breadboarding the standard fuzz face circuit (see ElectroSmash Fuzz Face) with trim pots in place of R3 and R4, in advance of committing to solder on the PedalPCB Twin Face pcb. What I've been finding is that, when I iteratively adjust R3 and R4 to set the Q2 Collecter bias to -4.5 V and Q1 Collector/Q1 Base to -0.7 V, I can't hear any differences amongst my GE transistors, Russian GT309, MP21, MP26, or a salvaged Hitachi 2SB75. This includes changing either Q1 or Q2 from 40 hFE to 150 hFE. I do find that trimming R4 is what really makes a difference for setting the Q2 collector bias point, and R3 is more of a fine adjustment that helps me dial in Q1 collector. Interestingly (or not - probably reflects a fundamental attribute of the circuit and a gap in my ChEng eduction), when these two points are at the nominal targets, the other test points (Q1 Base, Q2 Emitter) are exactly -0.2 V and -0.5 V, as in the Electro Smash analysis.

I've been recording at a couple of set guitar volume and fuzz settings, and truly don't hear anything in A/B comparison.
The only thing that made a difference was swapping in a silicon PNP (2N4403).

This seems consistent with a comment I read that biasing really matters more than the transistor gain. Which makes me think I can a lot more Fuzz Face out of the lower (40-60) Ge in my stash without obvious changes in performance. It also makes me understand that, in addition to the usual bias pot at R3 (as in the Twin Face design) I am going to want to adjust R4 in each builds to suit the transistor.
 
Only people who tweak transistors around the circuit, rather than the other way around, think it matters.

And they aren’t wrong. In that way, it does matter. These are also the people who think leakage is a design requirement as opposed to a limitation.
 
Now let's not go to TGP and burst any bubbles. You wouldn't want to ruin the day for those who spent dozens of pages dissecting the tonal differences between the blue and the teal Chase fuzz, would you?
 
andare said:
Now let's not go to TGP and burst any bubbles. You wouldn't want to ruin the day for those who spent dozens of pages dissecting the tonal differences between the blue and the teal Chase fuzz, would you?
Click to expand...
But but but buttt my G(e)FF sounds bit different day to day… Yeah it could be temperature drifting, but it mostly got to do with WHERE I stand in rehersal room, how guitar feedbacks on that particular spot. And this whole sensitivity changes with how my amp gain stages are set, but the most dramatic tonal shaper of the fuzzzzz is how deep the earplugs are pressed. :D
 
ftumch said:
Always wondered if these were less temperature sensitive in fuzzy settings. Just learnt the other day that the hole is there to screw them down by
Click to expand...
Thermal mass. It all depends.

So...yeah. You can bond these things to a heat sink, but the device itself generally isn't putting out much heat. The second law of thermodynamics is the guy you'll be fighting here: a heat sink will only increase the rate at which your device adapts to ambient temperature.

Insulation is the other option, and that could definitely help slow down swings in temperature, but that's a double edged sword. Insulation only slows the passage of heat: it doesn't stop heat transfer, and by virtue of slowing the passage of heat it's also as good at keeping heat *inside* as it is at keeping heat *out*.

The other options are a dedicated heater circuit (meant to keep the transistors at a temperature range above normal ambient conditions, so that the circuit's bias doesn't drift and you have consistent performance). That's a current hog, and control of the internal temperature takes a bit of extra engineering.

Also: refrigeration. Peltier setups could be used to keep your transistors at a specific temperature range. Those are current hogs too, and even more complex than the heater setup. DX is just...no. It would be cute, but no. Evaporative would be a ridiculously complex proposition, and a dry cooler would have the same issues as a heat sink: dependency on ambient conditions.

Or, you could build The Mechanic. Chuck's little corner of this site has lots of great little goodies in it. Use a special opamp to maintain collector current across a temperature range. Pretty nifty.
 
Stickman393 said:
Thermal mass. It all depends.

So...yeah. You can bond these things to a heat sink, but the device itself generally isn't putting out much heat. The second law of thermodynamics is the guy you'll be fighting here: a heat sink will only increase the rate at which your device adapts to ambient temperature.

Insulation is the other option, and that could definitely help slow down swings in temperature, but that's a double edged sword. Insulation only slows the passage of heat: it doesn't stop heat transfer, and by virtue of slowing the passage of heat it's also as good at keeping heat *inside* as it is at keeping heat *out*.

The other options are a dedicated heater circuit (meant to keep the transistors at a temperature range above normal ambient conditions, so that the circuit's bias doesn't drift and you have consistent performance). That's a current hog, and control of the internal temperature takes a bit of extra engineering.

Also: refrigeration. Peltier setups could be used to keep your transistors at a specific temperature range. Those are current hogs too, and even more complex than the heater setup. DX is just...no. It would be cute, but no. Evaporative would be a ridiculously complex proposition, and a dry cooler would have the same issues as a heat sink: dependency on ambient conditions.

Or, you could build The Mechanic. Chuck's little corner of this site has lots of great little goodies in it. Use a special opamp to maintain collector current across a temperature range. Pretty nifty.
Click to expand...
I kind of want to build a fuzz with a fridge in it now. But I've got a feeling it'll buzz
 
Peltiers are solid state: build a big enough heat sink and it can passively radiate. No need to deal with the potential of back EMF with a fan motor.

Buttttt....well. big heat sinks are fragile. You could always use a small 12vdc fan with its own isolated power supply. Hell, you can typically run those at 9vdc too: they just run a bit slower.
 
IMO, BJT selection does make a difference in FF circuits. Primarily between silicon or germanium. Silicon BJTs will typically be rather consistent in a FF circuit. Whereas germanium BJTs will always exhibit a temperature sensitivity that silicon does not. From day to day, or even hour to hour, a Ge FF can sound different due to temperature variances. Silicon does not do this.

Among Si FF circuits, you may detect some differences between BJT part numbers. At my age, I no longer can.
Ge BJT selection can be fussy due to their thermal sensitivity and/or leakage. Even when carefully measuring Ge BJTs for matched gain & leakage, they can sound different in the exact same circuit due to Ge substrate inconsistencies. Silicon does not suffer those inconsistencies.

And in either Si or Ge selection, biasing is key to get the best operation from each. Some builders insist on depending on measuring a proper 1/2Vcc for biasing whilst others (like myself) depend solely on what is heard.

At least, in all my experience with FF circuits and BJT selection; these have been my findings. Your milage may vary.
 
Yup tubes are the same - you'll get differences, so you get used to "rough ball park" figures, build it and then fine tune. The fine tuning you know will change between tubes so in the end that has to be redone on the tube changes if you get into that level of 'detail'. Then you remind people it's sat on a non-regulated supply (+10%/-6% wall) and the random output and power transformers can be +/-10% tolerance too between each..

Generally BJTs also allow more current the warmer they get, irrespective of silicon or germanium, vs mosfets that drop in current the hotter they get.

Just stick it in your pedal, tweak the tone, and love it. Not much point going to the nth degree (says the person has spent an age researching and designing some very quiet power supplies).
 
You must log in or register to reply here.
Back
Top