Is it the individual gain or the gain ratio of two transistors in Fuzz matters?

[Thecomedian]

let's say someone said an original fuzz face sounded great with a gain of 70-85 for Q1 and 120-140 Q2.

Would it be possible to say that two Tx with a gain of 20 for Q1 and 26.8 or 30 for Q2 would produce the same results? Does the gain itself have more importance over wave shaping than the gain ratios of the two Tx? if you could produce identical wave distortion shapes with a 20 gain Tx as with a 100 gain Tx, by changing resistors and caps in the circuit to sound identical, could it still be called a "vintage clone"?

[LucifersTrip]

sorry, can't explain the technical details, but very simply, if you bias low & high gain hfe sets at ~ 4.5 - 5V,  the higher hfe ones will usually be more compressed, mushy & saturated...and even more so with silicon, since hfe's can be much higher

[Kesh]

the one i made, when i was swapping them about, it was them being around 100, plus or minus 30, that seemed to matter most.

[Thecomedian]

This is all pretty interesting. I built a point to point fuzz with 2n2222's (npn Si) and it acted like how the instructional site stated: The last 10 percent of the 1k wiper is where you'll start to hear it. On the other hand I breadboarded one with 2 60-70 gain MP25 transistors (pnp Ge), and the volume/gain does get lower when fiddling with the 1k pot, but it plays with the distortion value almost perfectly, from heavily and possibly harshly clipped to only the very faintest clipping, if any. So the whole range of the 1k pot is working in it, which surprised me. That puts them just about in the +/- 30 range you are describing.

The funny thing was that I actually failed to hook up the Q2 base the first time, so I got some feedback/ticking when I turned the 1k pot down, but otherwise at full open on 1k, the pedal would still distort with only 1 transistor active.. hmm..

I guess that's where the biasing LT is talking about comes in. the Ge Tx need less bias, so they will saturate with a lower hFE, if I'm understanding this correctly.

[brett]

Hi
the fuzz face works quite differently with Si vs Ge transistors. Low gain gives low input impedance (less treble, less fuzz, more 'sag and bloom'). Hfe needs to be a minimum of around 40 to get enough input impedance (doesn't affect bass guitars or active pickups).

It's a bit of a mystery how you got a ff to work with the base of Q2 disconected. There's no bias.
cheers

[Thecomedian]

I reckon it was due to the Q1. LTspice wasn't really informative. destroying Q2 base to 33K/Q1 Collector connect does create some interesting near organic distortion but in the Picovolt range.

I went ahead and detached the Q1 C + 33k ohm from Q2 base again to make sure. The volts into the Q1 have to be pretty high (hard strum on two or more strings), but there is a bit of "noise" and in terms of some barely audible hum or other business, and the 1K pot on 100K to Q1 base and Q2 emitter has to be 90% to 100% "open" to the cap.

Distortion is still there. It's tons better with Q2 base in the circuit, though.

The amp I've been testing it with is 15w "practice amp" and I have the gain up to 66% or so on it, with my bass volume all the way up, and it does have active electronics in it, but passive pickups.

[Thecomedian]

Hi
the fuzz face works quite differently with Si vs Ge transistors. Low gain gives low input impedance (less treble, less fuzz, more 'sag and bloom'). Hfe needs to be a minimum of around 40 to get enough input impedance (doesn't affect bass guitars or active pickups).

It's a bit of a mystery how you got a ff to work with the base of Q2 disconected. There's no bias.
cheers

I have a question as to why it wouldnt affect bass. The higher 3 strings are the same as the lower 3 on a regular guitar.

[brett]

Hi
the frequencies that are 'lost' due to low impedance are high (starting at 2 to 10 kHz). These frequencies are low-order harmonics of the top 2 or 3 strings (e.g. A above middle A is 880 Hz, and a harmonic 2 octaves up = 3.2 kHz).
The reason for the loss of high frequencies are due to the combination of inductance and resistance in the pickups, and the capacitance of the cable. Passive pickups were designed for plugging into large loads (like a valve amp). The popularity of buffers is due to the high load they present to the pickups, which 'recovers' much of the 'lost' treble from passive pickups.
cheers

[Thecomedian]

I see. I'd heard a 1Meg volume pot in your instrument would preserve lost highs, while treble bleeds and 250k pots would obviously "lose" highs by making it warmer. An a side note some 1960's book seems to imply to me from what I've read that the reason "cutoff Frequencies" dont create a sheer cliff but an incline, is due to the quantum property of an electron to "jump" to the other side of a barrier, or some business.

Would higher value Caps input and output in pedal preserve highs?

[R.G.]

I see. I'd heard a 1Meg volume pot in your instrument would preserve lost highs, while treble bleeds and 250k pots would obviously "lose" highs by making it warmer.

Probably, depending on the guitar pickup and guitar wiring. Guitar pickups are heavily, heavily inductive, so any loading on them selectively removes highs. That probably means not much to you at your stage in learning, but it's a correct fact for you to remember some day. It's the reason most pedals have input impedances in the 1M range.

An a side note some 1960's book seems to imply to me from what I've read that the reason "cutoff Frequencies" dont create a sheer cliff but an incline, is due to the quantum property of an electron to "jump" to the other side of a barrier, or some business.

The first part is correct. The 'significant frequency' for a single R, single C circuit is always 1/ (2*pi*R*C), but that's the frequency where the signal loss is half the power in the signal. From there, there is an incline. The slope of the incline is always half-power for a 2:1/octave change in frequency. Adding more R's and Cs can make it steeper, but sheer-cliff filter cutoffs are impossible in the analog world.

The second part is incorrect. This behavior is not due to one of the quantum properties of the electron, it's just a part of its classical-physics properties.

Would higher value Caps input and output in pedal preserve highs?

No, it would preserve lows.

An ideal-world resistor has the same impedance to signal flow at all frequencies, from DC on up, and real resistors do show this behavior well up into the RF range. A capacitor has an impedance to signal flow that is infinite at DC, and decreases linearly with frequency. So a cap with an impedance of, say, 820 ohms at 1kHz will have an impedance of 410 ohms at 2kHz and 205 ohms at 4kHz, and so on as frequency rises.

The connection where a capacitor is in series with the signal means that the capacitor "impedes" signal flow most at low frequencies, least at high frequencies. So using bigger caps in series with the input and output lets more signal in and out, raising the bass. It has little effect if any on the highs.

[Thecomedian]

ah, thanks. I got those two reversed. That's why the Treble booster has .002 caps and the bass has 3.3 caps.