Tone Bender MKII overload issue

[fuzzy645]

My tone bender clone has been sounding wonderful but all of a sudden i had an issue.  I was playing a low output bridge single coil on my Strat and when I played with a very aggressive attack the pedal kind of crapped out and the volume totally died for a second before coming back. Tried the same thing on a SG humbucker bridge and the same thing happened. It only does this on the bridge pickup, the neck pickup is fine.   Any thoughts on what is going on?  Could it be that a hard attack when playing is somehow overloading the trannys?   If that is the case  would i need to bias it a bit colder, or perhaps could it be a bad transistors?

I'm using this schematic: http://forum.musikding.de/cpg/albums/userpics/16602/MKIIOC75schem.png

Only deviations from this are that the 470 ohm resistor is 1K.  Getting about 8V at the collector of Q3.

Thanks in advance

[pinkjimiphoton]

is it germanium or silicon you used?
sounds to me like you're overloading the input of the first transistor and making it hit cutoff... then ya get that kinda swell effect sometimes. maybe turn up the bias voltage slightly and see if you can find a compromise where the pedal still sounds good but it stops doing that.

or learn to exploit the hell out of it, cuz it probably sounds cool as hell thru a univibe and a tape echo


i have noticed germanium fuzzes do that sometimes, especially in winter.

[fuzzy645]

is it germanium or silicon you used?
sounds to me like you're overloading the input of the first transistor and making it hit cutoff... then ya get that kinda swell effect sometimes. maybe turn up the bias voltage slightly and see if you can find a compromise where the pedal still sounds good but it stops doing that.

or learn to exploit the hell out of it, cuz it probably sounds cool as hell thru a univibe and a tape echo


i have noticed germanium fuzzes do that sometimes, especially in winter.

Thank you.  Yes, the transistors are germanium (OC76).  I will give that a shot.

[dwmorrin]

That 10k resistor at Q1's base looks a bit low, no?  Very low input impedance for a straight guitar.  I would expect 100k there, unless this is some special MKII variant.

[fuzzy645]

That 10k resistor at Q1's base looks a bit low, no?  Very low input impedance for a straight guitar.  I would expect 100k there, unless this is some special MKII variant.

Thanks that is a good point.  The other schematic I found was from fuzz central and they used a 100K as you say.  http://fuzzcentral.ssguitar.com/mkII/mkIIschematic.gif

I remember trying both 10K and 100K on the breadboard before the build, and the 10K sounded better to my ears at the time, however I must have not tried playing with an aggressive picking attack.  Now that I played the thing in a real playing situation with the harder picking attack I am first experiencing this problem.  

[dwmorrin]

would i need to bias it a bit colder

Maybe bias Q3 hotter.

I don't mess with the Ge transistors too much myself - I'm fine with a silicon MKII - but Q3's collector at 8V means Q3 is pretty near cutoff at idle.
4.5V is the "ideal" for maximum swing, but who's striving for "ideal" in a fuzz, right?  But I'd try to bring that 8V down some.  The classic mod is to switch the 8k2 resistor with a trim pot.

Now I see the 10k variant - it's the "Marshall Supa Fuzz" according to Fuzz Central's MKII page. http://fuzzcentral.ssguitar.com/mkII.php
Never messed with that myself.

[Arcane Analog]

I would refrain from relying on anything from Fuzz Central and I would disagree with dwmorrin's suggestions.

The 10K version is not specific to a Marshall Supa Fuzz. It is used in MKIIs with OC75 transistors - the 100K is used with OC81 transistors.

The 8V you have is good. Vintage units had around the same voltage and some are even reported as being in the 9V range. A MKII with 4.5V is not what you should be aiming for if the goal is to have something close to an original.

The "classic mod" of 8k2 to 10K trim is for a Fuzz face - it is not what you want to do in a MKII. To bias a MKII replace the 47K with a 100K trimmer. Seting it close to the middle will get you in the 8V range which you already have.

[Arcane Analog]

The "bottoming out" of the MKII is not uncommon. It is probably a temperature issue. Variations in temperature cause a wide variety of changes in tone.

If temperature is not the issue try dropping the 1K back down. Make sure your caps are the correct values.

[dwmorrin]

 
Sorry Arcane.  My favorite fuzz is my MKII silicon.  You're absolutely right that I spoke ignorantly.  I guess I need to hit the breadboard with some germaniums.

[pinkjimiphoton]

or..... another trick you can try is to splice in a ge diode between emitter and base, reverse biased... the idea was (the brit face from plate to plate) that the ge diode would leak the same way the ge transistors did and counter it.

i tried it and it DID work, tho not as well as i'd hoped.

i found two ways to deal with bias issues in ge so far...

first, add an external bias pot (or a global brown-out pot)
second, run it on a charge pump with humongous caps.

it takes 10 minutes to warm up that last way, but once it's on it's stable.

i'm betting it's the pick attack overloading the transistor into cutoff and then it coming back..
i would definitely avoid the 4.5 volt cliche, use your ears.

tho 4.5 volts will make most things fuzz, that's not optimal, and you lose dynamics and tone.

a higher voltage may seem like less fuzz at low volumes, but live, it helps you overdrive your amp, too.

i routinely run my ge fuzzes anywhere from 5-8 volts on the collectors, i try to find a sweet spot where it gives me the best sweep from clean to filth from my guitar.

sounds like you're getting on it... you'll get it. rock on bro!!

[Arcane Analog]

Sorry Arcane.  My favorite fuzz is my MKII silicon.  You're absolutely right that I spoke ignorantly.  I guess I need to hit the breadboard with some germaniums.

No big deal at all. Germs and silicon are very different beasts.

[Arcane Analog]

or..... another trick you can try is to splice in a ge diode between emitter and base, reverse biased... the idea was (the brit face from plate to plate) that the ge diode would leak the same way the ge transistors did and counter it.

i tried it and it DID work, tho not as well as i'd hoped.

Did you match the leakages?

[fuzzy645]

...first, add an external bias pot (or a global brown-out pot)
second, run it on a charge pump with humongous caps.

A couple of questions regarding this.  Wouldn't this circuit then need 3 bias pots, because technically the 3 trannys might all require bias adjustments.

Second, (and pardon my ignorance), but I do not understand what you mean by a charge pump with homogeneous caps. 

[pinkjimiphoton]

or..... another trick you can try is to splice in a ge diode between emitter and base, reverse biased... the idea was (the brit face from plate to plate) that the ge diode would leak the same way the ge transistors did and counter it.

i tried it and it DID work, tho not as well as i'd hoped.

Did you match the leakages?

naaah, was years ago at this point. i'd socket and stuff til i liked it.

[pinkjimiphoton]

yeah, you could do it that way. depends on how close your collector voltages are to what ya want. as you raise/lower the voltage you will get everything from fuzz to a compressed glassy kinda boost to a severe distortion right around cut off.

humongous, not homogenous... humongous is slang for big, really big.  i think i used like, 1000u on it instead of the 10u. just tinkering. i had figured it would just filter the power a lot more, but it appears now it acts like a reservoir, and the caps have to fill up or charge or something.

it's a trip. but it's kinda cool. when the blue light comes on, my fuzzface is ready to go.

as for a sag... if you add a sag control to the circuit, and bias with trimpots where the sag control would be about half way up, you can increase/decrease the overall voltage to the pedal , keeping the ration of the original circuit intact between the different transistor... i have done this on fuzzes that were obstinate, so i can increase or decrease the b+ overall, which raises or lowers the bias point universally.

a quick tweak of the pot, and you're in business if it's too hot or too cold.