Does the Marsh Fuzz really need all that filtering?

HamishR

Well-known member
I pulled out my Marsh Fuzz to use today and it really is a good sounding pedal, whether used as a boost or a fuzz. So I went back to this site and looked at the schematic - does the designer have shares in an electrolytic factory? I've seen so many fuzz pedals with next to no filtering and most of them sound just fine like that. A Big Muff has a single 100µF filter usually. Is there a reason the Marsh should be so splendidly endowed?
 
It looks like a pretty complicated circuit. Not all caps are there for filtering. Some are used to affect the signal gain without changing the Q-point (the bias), eg, C14 adds collector-base feedback for signal (but not DC) and C12 does the same but for highs only; C5 and C26 may be there to prevent oscillation; C15 allows changing the gain for signal keeping the bias stable. Some are coupling caps since there are 4 stages in the circuit. Then there are caps in the actual filter nets for highs and lows pots which makes sense. There is extra filtering on the power rail which I don’t understand and there are two separate V/2 generators for the two buffers at input and output which is odd (maybe there is too much noise or interference if shared?). I also don’t understand how the base of Q2 is biased, it looks like it’s at 0V DC to me (edit: it’s not, R13 does the biasing and also feedback loop, iirc).
 
Last edited:
Using an electrolytic (like the 220uF) and an MLCC (like the 100nF) in the power filtering is pretty common, as they're filtering different things.

Why there are multiple 220uFs in the power filtering section is beyond me, but one thought might be that they make sense on the original PCB layout. ie, on some circuits adding additional power-to-ground caps in close physical proximity to a transistor etc can help quiet down an angry circuit. No idea if that's the case here, but it's the kind of thing that can get lost pretty easily between trace-schematic-clone.

...just dug around and found the trace thread on FSB and you can see how the electros are spaced around the board, kinda close to the transistors (also interesting that all the caps look like mlcc aside from the three in the power section):

20130429-120659.jpeg
20130428-230715.jpeg
 
There is quite a bit of filtering (possibly overkill), but like @giovanni said, those aren't all filter caps.

That's an older version of the pedal in the pic.

This is the more recent version. Terrible photo but the only one I have that shows the entire board in one pic.

1652534834768.png
 
Last edited by a moderator:
Well however it's designed I like the sound. I don't find the bias pot to be useful - Is it supposed to allow for starved transistor sounds or is it a tone-enhancing thing? It doesn't seem to do much on my pedal. And the Fuzz pot almost needs to be a C++ taper. Wouldn't that be handy?? Even though it's a C-taper everything still happens at the end.

I don't mind really. I'm just surprised that a fuzz pedal needs so many big caps.
 
It’s a very unusual design for a fuzz. A lot of modern features if you will, requiring all the additional components. I bet it doesn’t clean up with the guitar volume like a FF would. I can’t wait to build it!
 
I think Deep Trip went way overboard on the filtering, but it does no harm. Just like you can't put too much water in a nuclear reactor, you can't put too much filtering on the power rails. Maybe Deep Trip's basement garage yurt lab had a really noisy power supply.

While we're critiquing the BOG design, what's with the two 100K resistors in series? Or the 12.4Ω resistor in series with the BIAS pot?

At my first job out of school, I learned a term from the Sales Manager: "Apparent Value." It means if you make it a product look like it's worth more, people will pay more. I suspect that's what all of the extra parts are about. Back in the old days (Fig will remember this), transistor radios proudly advertised the number of transistors inside. Obviously, a 10-transistor radio is better than a 6-transistor radio. Nevermind that four of the transistors on the board were not in the circuit.
 
I think Deep Trip went way overboard on the filtering, but it does no harm. Just like you can't put too much water in a nuclear reactor, you can't put too much filtering on the power rails. Maybe Deep Trip's basement garage yurt lab had a really noisy power supply.

While we're critiquing the BOG design, what's with the two 100K resistors in series? Or the 12.4Ω resistor in series with the BIAS pot?

At my first job out of school, I learned a term from the Sales Manager: "Apparent Value." It means if you make it a product look like it's worth more, people will pay more. I suspect that's what all of the extra parts are about. Back in the old days (Fig will remember this), transistor radios proudly advertised the number of transistors inside. Obviously, a 10-transistor radio is better than a 6-transistor radio. Nevermind that four of the transistors on the board were not in the circuit.
Well obviously you gotta have spares!

Yeah I was wondering about the two 100K and the 12R4 too. Actually did a double take at the double 100K. Maybe 200K resistors aren't available in Brazil? Maybe 12R4 resistors are extremely hip with the kids? Especially in series with a pot! They speak Portuguese in Brazil so maybe the circuit loses something in translation?

For me the strangest thing is that we have a bias trimmer inside and a bias pot outside and neither seem to do very much at all. Maybe I need to play with the trimmer some more. I'd be happy with a simplified version which sounds the same so maybe I'll just hardwire the bias next time.
 
Back
Top