Too many Muffs, not enough time

[Vivek]

I decided to look at the plethora of variants of the Big Muff circuitry, given that there are so many subtle (and not so subtle) differences between them.

My reference is the famous site http://www.bigmuffpage.com/Big_Muff_Pi_versions_schematics_part1.html

I input the following muff variants into LTSPICE:

1 = VD-1 Vintage Distortion 2011
2 = Hoowf
3 = BMP Pi Triangle Distortique
4 = Creamy Dreamer (No emitter degeneration resistor)
5 = Pharo Modified "47" (which has a "Hi" and "Lo" input)


First stage frequency response looks like this (With Pharo in "Hi" mode) :



and the First Stage AC voltage Transfer Function looks like this (With Pharo in "Hi" mode) :

X axis is input signal in mVp, Y axis is output of first stage in mVp

So far, it appears as follows:

The first stage of first 4 Muffs are quite similar in terms of frequency response and transfer function, despite part value changes.

The Pharo seems to have lower gain, more bass going through the first stage. And In "low" mode, it has even lower gain.

The other two stages with the clipping diodes are rather high gain stages, and they hard clip all input signal above about 20mVp.

There is a patent on multi-stage distortion which suggests that there should be gain and frequency response differences between stages, to make it more alive and have more movement when playing. Muff (and Vulcan) repeat the same stage multiple times.



For the input, I used a Sine voltage source with 5000 Ohms internal parasitic resistance to emulate a pickup. I have the nagging suspicion that this might not be correct way for low impedance fuzz circuits. Should I add a 3 - 5 Henry Inductance in series to the voltage source ?

One of the famous existing Muff Analysis web sites does not seem to have considered Pickup impedance. Given that they proved via math that the gain of the first stage (Common Emitter amplifier with Shunt Feedback) depends upon input impedance, it's a fox paw that they did not consider Pickup impedance, which is important for this type of circuit.

[antonis]

One of the famous existing Muff Analysis web sites does not seem to have considered Pickup impedance. Given that they proved via math that the gain of the first stage (Common Emitter amplifier with Shunt Feedback) depends upon input impedance, it's a fox paw that they did not consider Pickup impedance, which is important for this type of circuit.

It makes little diference between 39k//47k//470k and (10k+39k)//47k//470k, 10k been supposed pickup impedance..

After all, their Gain calculation is declared to be higher than simulated, partially due to 39k (and not 10k+39k) been set in denominator..

[EBK]

"fox paw" made me chuckle.   

Makes sense to me to add the inductance if you want a more complete picture.

You may want some capacitance too.

[Vivek]

I feel that the low impedance of the Muff will

A) Make less signal present itself at the base of the first transistor, since more voltage will be developed across the pickup's internal resistor

B) Reduce the highs that present itself at the base of the first transistor since more voltage will be developed across the pickup's internal inductance

I will try to plot first stage gain with ideal sine wave source / with a pickup simulation



Ha ha Fox paw, I love English word plays, Crosswords, Puns, Mondegreens etc

A Paraprosdokian is not a relative of Kim Kardashian !!!

[Mark Hammer]

My sense is that what a calculated frequency response of a Muff shows, is likely to be different than what an actual Muff shows.  I wish I could track down the original magazine article, but in an interview for a business magazine, some 15 years back, Mike Matthews noted that any 4 consecutive Muffs coming off the line in the '70s (and possibly '80s as well) would sound different from each other.  The principle reason would seem to be that EHX bought many components surplus, or other routes that provided "bargain basement" pricing.  Moreover, most of the components used had wider tolerances than they do now, and nobody at EHX was selecting components, based on measured value.

[Phend]

Does what I call "Blue Printing" an effect circuit make it sound (for a lack of words) "precise" to the way it is intended or should sound. Blue printing, meaning measuring each component for exact or close to exact value. Sans transistors and diodes since I don't have bags of them.

[Mark Hammer]

In theory, it should.  Certainly, as much as people kvetch about all things digital, one of the advantages of DSP-based pedals/effects is that they are consistent.

I've commented here on many an occasion that all of the issue-to-issue "shootouts" I see, on Youtube and elsewhere, compare ONE pedal from a given production year against another SINGLE pedal from a different production year, and never considers whether two or more pedals from each year of production actually sound the same.  These comparisons are done with the assumption that every single unit made has the absolutely identical component values and specs shown on the factory schematic, such that any ONE pedal from that era/issue MUST sound exactly the same as any other with the same cosmetics and year of production.  Reality is a little different.  Sometimes not dramatically different (i.e., each has the same general character), but different enough that people search for "a good one". 

Remember that pots don't always have the resistance value stated on the back.  This one might say 1meg but really be 887k while that other one that also say 1meg might be 923k or 831k.  If that pot is used to set gain, one can expect differences between examples of the same pedal.