Tweak Fuzz

[jayp5150]

I couldn't find much here searching...

I'm working on a Tweak Fuzz for a friend (modding a factory one).

Found the schem here: http://analogguru.an.ohost.de/193/schematics/SeymourDuncan_TweakFuzz.gif

One thing to point out first is that Q1 and Q2 are swapped on the board. SD marked Q1 as Q2, and vice versa; so the schematic actually has the tranny types flipped.

Q1 (as we know it in the schem) is actually the 2XD, which I can't really find any info on... marked as a "4401 ?" on the schem, and Q2 is the 1AM, which should be a 3904-alike.

The issue I'm having is getting the bias down to 4.5V on Q2. I've actully bumped R4 to 2k at this point, and have a 10k trim on R5. This, maxed out, still has me at 5.25v. that seems odd to me. Should I have that much resistance to bias a FF?

So, I'm wondering if I should just junk what's in there and start swapping trannies? Once that SMD tranny is off the board, it's pretty much gonna be dead lol. OR, just up the resistance til it biases and use my ears?

The other question is that if Q1 IS a 4401, that doesn't make sense in the traditional FF circuit, as they hover around an hfe of 250, and the 3904 would be around 150...

Just looking for insights from some folk who are smarter than I lol

Thanks!

[Gus]

Use this circuit with the stock transistors, keep the switchable input cap section.  FWIW the schematic has been here for years and it was up before the axis face(the lowpass, 1K and .01 cap) and Jack emitter resistor at Q1 that does not do as much with a 33K as a 10K collector resistor for the first transistor.
http://www.diystompboxes.com/pedals/gusFuzzFace.gif

Why are you modding it anyway?  What was wrong with it?

[petemoore]

  My buddy bought one. Dog...I looked and checked, decided it'd actually be hard to top the FF in this one, I guess I've been through the FF mods enough.
  Maybe put a booster in front...but I think getting the right' pickups and amp for it'd be where I'd start.
  There are plenty of FF derivatives which will go where the basic FF with all the whistles like this one won't go, but they're OT really and best drawn from extensive researching [probably something like bias starve or whatever else is seen around with a Face in it [I guess the feedback resistor has to stay in the 2 transistor circuit to call it a FF, I suppose a cap or pot added here might still let it qualify?] to find the ballpark sound type desired, populated on a breadboard or something like the tweekfuzz [the capacitor switch voicings are pretty slick and should cover most of the gamut of TB/FF/other FF sounds.
  To improve'' on the cap value selections for the voicing switch or put some new twist on the FF beyond the capability of the TF...I would suggest actually warrants the 'other routes' as compared to a re-mod of the already quite modded tweekfuzz.
  Guitar or amp change of course will reveal Tweek sounds beyond the scope of the guitar/amp now used, beyond that a new circuit board seems like a small addendum to either find the new Fuzz or find the Tweek that the TF is somehow missing.

[jayp5150]

Well, it was actually inoperable when he sent it too me (gain pot wires had been sheared due to a loose pot and what seems to be some serious twisting lol)

The only complaint he had was that it lost sustain at the fatter settings, which I think was due to the high bias voltage stock.

I actually threw a 4.7k in r4 a few minutes ago, and with the 10k trim, got it down to 4.5v. It's super loud now, I might swap in a 15k trim and move r4 back to 1k

Honestly, it doesn't really need anthing else.

One of my faves is to put the 100k feedback resistor on a switch to go between that and a 1n914. That tames it down and seems to add more touch sensitivity.

Now that it's biased properly, it's actually very nice. I was just thrown by the high resistance needed to get it there.

[PRR]

> marked as a "4401 ?"

2N4401 is a very standard large cheap NPN Si transistor. Really more than this circuit needs. Neither transistor is fussy. "Si NPN with hFE over 100." The circuit is locked-in by NFB, as long as hFE is not "small".

> start swapping trannies?

As long as they are not -dead-, or amazingly weak, tranny swap make no difference. hFE from 150 to 800 will all work very nearly the same. This rig does what the resistors tell it to do.

> getting the bias down to 4.5V on Q2.

P2 should be just about 2K.

What is it now? Has it been replaced?

NFB from Q2 E to Q1 B sets up about 1V across Q2 emitter resistor, here P2. If this is 2K, we flow 1V/2K or 0.5mA through Q2.

Q2 collector voltage is this current, times R4+R5, down from 9V.

If R4+R5 is about 9K, Q2 collector sits at 4.5V.

> I've actully bumped R4 to 2k

Changing R4 is not much change.... R4 is a very small part of R4+R5. If you can't change Q2 emitter resistor (because it is a pot you don't want to change), then fiddle R5 from 8K to 5K or 10K. If P2 is way-off from 2K, do what you need to do.

> It's super loud now

With Q2 collector near 4.5V, it can put out almost 3V of signal at R5. This is way more than we want. The designer added R4 so he could tap-down and get a fraction of this. The 8200:470 ratio will give about 0.15V signal on R4, which is a good ballpark. If you have increased R4 toward 4.7K to get a 4.5V bias, the ratio 8.2K:4K will give maybe 2V of output, which IS "super loud".

Check what P2 really is. You can make it work with 1K or 5K, but you will have to make R5 about 4 times higher to get near 4.5V at the collector. Then make R4 about 7X-20X smaller than R5, to get a "reasonable" maximum output level (depending what you call "reasonable", and what your amp likes).

[jayp5150]

The gain pot is a 2k. I had overlooked that.

So, are you saying that a higher value on the gain pot will increase the resistance needed to bring Q2 collector to 4.5v, or am misunderstanding?

This just proves that I am still learning, after 5 years, and still on simple circuits lol.

I appreciate the responses from all you guys.

My wife was also refinishing this, so I'll post pics and clips when we're done with it.

Thanks, again.

[petemoore]

  A large for FF chunk of perfboard, a couple 8pin IC sockets total about 1 dollar.
  The rest totals none or not much more, depending on how inclined the builder is to save on the other 6 buxx.
  Socketted transistor and cap positions [the bypass cap can be small, made bigger with parallel capacitor].
  Large value, long leg [above board] resistors can then easily be made smaller with parallel piggyback resistors.
  The 100k could be another added socket [large FF board for testing/tweeking every possible permutation of myriad possibilities can be easily reduced to something that leaves room for the extra switches and pots you may feel want for after using the personal Fuzzface tweeking board and allowing it to teach pretty much everything about a FF and beyond [see GEO and AMZ and].
  Q1 emitter does well with small [100ohm] resistor for Si smoothie fuzz, good rolloff etc.].
  The eloquence with which the sound defines itself goes well beyond what any the information any lengthy descriptive terms can translate. The FF is a perfect specimen {IME} for experimentation, coupled with the available texts, an extremely rich wealth of information can be streamed through to ears through the speaker.

[jayp5150]

  A large for FF chunk of perfboard, a couple 8pin IC sockets total about 1 dollar.
  The rest totals none or not much more, depending on how inclined the builder is to save on the other 6 buxx.
  Socketted transistor and cap positions [the bypass cap can be small, made bigger with parallel capacitor].
  Large value, long leg [above board] resistors can then easily be made smaller with parallel piggyback resistors.
  The 100k could be another added socket [large FF board for testing/tweeking every possible permutation of myriad possibilities can be easily reduced to something that leaves room for the extra switches and pots you may feel want for after using the personal Fuzzface tweeking board and allowing it to teach pretty much everything about a FF and beyond [see GEO and AMZ and].
  Q1 emitter does well with small [100ohm] resistor for Si smoothie fuzz, good rolloff etc.].
  The eloquence with which the sound defines itself goes well beyond what any the information any lengthy descriptive terms can translate. The FF is a perfect specimen {IME} for experimentation, coupled with the available texts, an extremely rich wealth of information can be streamed through to ears through the speaker.

I still need to pick up a breadboard  but I agree, I especially pearn better by just doing and testing it.

I have a spare FF board that I've been meaning to just populate with sockets, just haven't found the time.

It's next on the list 

[jayp5150]

> marked as a "4401 ?"

2N4401 is a very standard large cheap NPN Si transistor. Really more than this circuit needs. Neither transistor is fussy. "Si NPN with hFE over 100." The circuit is locked-in by NFB, as long as hFE is not "small".

I forgot to respond specifically to this. My main point with that was that it doesn't seem anyone knows what the "2XD" tranny actually IS. On the schem, it was marked with a "?."

I get your point, though. 

[liquids]

That schem is a bit iffy for a simply circuit, though for understandable reasons in some places.

I'm curious about what the right schematic is, in the end.  I searched SOT-23 and "1AM" and found this datasheet which says "MMBT3904LT1=1AM"  at the bottom much like the schematic. This page has a link http://doc.chipfind.ru/avictek/1am.htm.  But no such luck with searches for the 2XD.

Anyhow, for a simple circuit the schematic sure has a number of question marks.  Said question marks in the end not add up to much, all said.  Like many circuits and for those of us at home, simply to swap 2n2222/2n3904/2n4401/2n5088/mpsa18 etc for that unknown transistor, and pick one we like (or get a feel for what if any difference it makes).  Not so if you are trying to keep that 'as is' and don't know how to replace them.   If you in particular had a breadboard it would take no time to build to a/b for closest sound (to your ears) for various parts...if not a more technical approach...

I.e. - do you have an accurate enough capacitance meter to measure C1?  Big differences between 47pf and 470pf oftentimes in this type of circuit.   Do you have the patience to desolder and measure the HFE of the two (surface mount) transistors?  Have you tried flipping the two  (2XD, 1AM) transistors to reflect the drawn schematic, and compare any difference....I'm curious of the schematic was an error, or if the one you have in hand was soldered with swapped locations, or the other, etc.  But unless we have others who have opened a similar one or two, we may never know...though it may make little difference compared to other values....

Most of all, is it any difficulty to measure resistance across the outer lugs of the pot on Q2s emitter? The schem says 2.5k maybe 1k...very odd, and a big difference in this case.   Total pot resistance seems like an easy one to figure out in most cases, if you have a DMM.  And I see you are saying this pot is 2k - did you measure it?

Curious as to the need to get close to half voltage on the emitter...is it a volume thing or an attempt to make it sound better?  If volume, you could simply add up R4 and R5 (~8.7k per the schematic) and make R4 1.8k and R5 6.8k, for example or 3.9k + 4.7k, etc.  Whatever total resistance you want, split it differently so R4 is bigger but R5 is smaller but R4+R5 is the total resistance you want.  That will give more available volume.  Hope that makes sense.

[jayp5150]

I was just trying to get it to bias properly and was thrown by the large resistance needed.

It does not appear that this has ever been touched, so no issues there.

It actually sounds great with the right bias, I just need to play with the resistances to get the right volume I want. Just need to go pick up a 15k trim. Haven't had a chance.

Oh, the gain pot is a 2k.

I'll definitely post samples when done, i've just been super busy at work and home with other stuff.

Thanks for the reply!

[jayp5150]

I finally got around to getting this back together (my wife did the paint)



I got the bias set properly, and  the only actual mod was to replace R6 (the 100k) with a dpdt that throws between a stock 100k fbr and a loop of 1N914's. The diodes reduce the feedback to the 1st Q, but also increase the touch sensitivity of the circuit (although, at a lower gain). It also seems to roll of the low mids, which makes the fatter settings more useful. To be honest... I think one diode would have done the same thing, but I had done this on a previously-built sili fuzz, and it was good, so I left it alone lol.

Everything else was left stock, I was just getting this thing up and running, and biased properly, really.

I'm going to try to do a vid tonight or tomorrow.

[jayp5150]

Here's the vid I promised. apologies for the slight clipping (lol, in the tracking, not the fuzz... obviously...). I didn't hear it while recording, and afterward, the kid was already in bed, so I just went with it. You get the idea.

I think it turned out well.

http://www.youtube.com/watch?v=uwKnqn64qz8

Comments welcomed.