Shin Ei FY-2 germanium build report #2

[LucifersTrip]

I decided to revisit this thing since it's one of the most underrated, loudest and nastiest ge fuzzes...and it's insanely rare. Most just know of the silicon version with the mid-scoop, which ironically outputs too low volume.



Last time I took the easy way out, set the "fuzz" at max and put an external bias pot on Q2. Since I'm fortunate enough to have a large selection of transistors to select from, I took the hard road....Instead of the bias pot, I  took the time to choose the "right" transistors (NPN this time) that gave the hottest, best fuzz with nothing changed (except for the pots)...so, this isn't a typical build report where I tweak values.

I worked in reverse and it worked like a charm as it has many times before. Procedure: go thru a bunch of transistors to get a really good fuzz (good starting point: use hfe's 50 - 100). Then, sub pot for Q2 collector resistor. Adjust until fuzz is the hottest. Measure pot. If 22K, stop. If not, try next transistor and repeat. Once Q2 is chosen, do the same for Q1.

It's harder and takes a bit longer and supposes you have a good selection to choose from but I've had amazing success...and this is not an exception.

Not surprisingly, Q1 turned out to be low leak, but a bit lower gain that expected...and Q2 turned out to have much higher leakage than expected.  

Peak Atlas measured:
Q1: 2N388 (hfe 55, leak .04mA)
Q2: 9007558 1326 (153, leak .5mA+)

I ran into the exact 2 problems I did with my last build of this.

1) Volume is insanely loud (which I can deal with), but the bass heavily increases along with volume....and gets boomy and a bit muddy when 50-60% and higher.  The simple solution is to decrease the size of the volume pot. You can get away with 20K if you want to go for the Vox Tonebender treble punch...but, I suggest somewhere in the 30-40K range. I settled midway at 35K with a 120K resistor across the 50K pot. Note: there's a big difference between 20K, 25K, 30K, 35K, etc, so choose wisely....

2) Fuzz control...it's not a typical fuzz control. With the pot at 1K, it's a loud, nasty fuzz with excellent sustain. As you turn towards 50K, it slowly gets buzzier, a bit thinner and more saturated...and moves towards gating. For me, at 26K it's max fuzz with just a hint of gate starting to appear (used 32K on a previous build). Turn past ~30 - 35K, it'll be quiet, gated (< 2V on collector)...no hum, buzz or anything...So 55K resistor across 50K pot. If you want super-gate, leave the 50K in there.

...since this can get pretty damn loud & nasty with the vol turned up, a small cap across BC of Q1 knocks out a little harshness without disturbing the tone. Make it larger or put it on Q2 if you want to tame it more.

Here's the final version with details:

[joegagan]

wow, great guild report. i agree, this circuit has lots of potential.

i took inspiration from your first fy-2 thread ( http://www.diystompboxes.com/smfforum/index.php?topic=92714.0 ) when i made the death by treble fuzzed wah. it ended up as npn-   Si 2n2222/Ge ac187 combo with fixed output vol and diodes (?)  - it was designed to be on all the time with the wah, not independent of it.

this is my fuzzy build note:





and how it sounds:

[LucifersTrip]

^ sounds very cool...dark & heavy.  You used diodes in there?

What type of volume/tone did you get without the wah. I guess the diodes would bring it down and knock out some of the boom....

[joegagan]

i never tried it without the wah, believe it or not. since the two were destined to always be together, i simply tuned the whole thing as a unit. i can't recall why, but i must have added the diodes for more grit going into the wah. thanks for  taking a listen!

i know my drawing is  bad, it was just my build notes. hard to see on there, but there are silicon diodes as shown. i don't have the fuzzed wah here any more, it has gone to clif taylor in tucson ( maker of the FUZZ movie)

[Mark Hammer]

Although the name says FY-2, the circuit is certainly very different than the Companion Fuzz we all know and love.

So, a naive question:  How is this circuit shown different in behaviour from a Fuzz Face?

I'm also curious about the pot between the Q1 base and Q2 emitter.  Zach Vex uses that path for the Wooly Mammoth, and it produces some interesting and musically valid tonal changes, interacting quite a bit with the traditional Fuzz Face Gain pot.  Here, the AC bypass cap on Q2 has a fixed resistance from the emitter.  Gotta wonder what happens to the circuit when the 1k5 emitter resistor is replaced with a 2k pot (10uf going to the pot wiper), or simply a 1k pot in series with the 10uf cap.

[LucifersTrip]

Although the name says FY-2, the circuit is certainly very different than the Companion Fuzz we all know and love.

yep, the si is a Fuzz Rite rip off and the ge is a Fuzz Face rip off...

So, a naive question:  How is this circuit shown different in behaviour from a Fuzz Face?

louder, nastier, no mush...pretty brutal.  which is why, as a joke, I basically made a Fuzz Face into one...remove the 470, wire the output directly off the collector, set collector to lower voltage, heavily decrease size of vol pot (note: .01 out cap works with ~5K -10K and .00X works with ~20K - 50K...~ten-fold change in both), add BC taming cap to either Q
http://www.luciferstrip.com/fuzz/fuzzface-fy2-mod-variable.jpg

I'm also curious about the pot between the Q1 base and Q2 emitter. 

see above...it's not typical (from low distortion to a fuzz). Instead it goes from fuzz to a slightly thinner saturated fuzz that gates

Gotta wonder what happens to the circuit when the 1k5 emitter resistor is replaced with a 2k pot (10uf going to the pot wiper)

Is there any reason you wouldn't wind up with a standard Fuzz Face fuzz control?

[Mark Hammer]

Gotta wonder what happens to the circuit when the 1k5 emitter resistor is replaced with a 2k pot (10uf going to the pot wiper)

Is there any reason you wouldn't wind up with a standard Fuzz Face fuzz control?

None.  The point, however, is that the 50k emitter-to-base pot becomes more interesting because of how it interacts with the more traditional emitter-to-ground pot.

[LucifersTrip]

Gotta wonder what happens to the circuit when the 1k5 emitter resistor is replaced with a 2k pot (10uf going to the pot wiper)

Is there any reason you wouldn't wind up with a standard Fuzz Face fuzz control?

None.  The point, however, is that the 50k emitter-to-base pot becomes more interesting because of how it interacts with the more traditional emitter-to-ground pot.

...and why not make it a 4-knobber. throw a 20-50K pot on Q2C and you can get interactions galore...

either way, this one's a cool FF variant that's not much like a normal FF.