si FF transistor "magic gain #s"

[b_rogers]

i know that with the Ge FF the best sound is supposedly at 70-90 for q1 and 90-130 for q2. what about silicon? has anyone analyzed one of the "good"  sili FF's? i know everyone says use whatever sounds best but i was just wondering.

[brett]

IMO to get the input impedance right, you'll need hFE around 70, and no more than 100 for Q1.  For Q2 the "right" values can be higher, but it'll start to pick up interference and sound bad as the gain goes too high.

good luck

[petemoore]

Yupp,  check out the ROG site FF's..IIRC they list lots of voltages for different FF's.
 Lower Hfe for Q1 seems to have become the consensus.
 There are so many FF types to choose from, LTA, I built a FF perf/bread/jig...gotta lotta sockets, and a trim er two on it, great for tweekin' tar from countless FF types.  
 33k, 100k, and the small one up top 1k. can be that or less with a paralleled resistor on their back...I leave room atop the board so I can tack parallel R to reduce the total R on these, but it's not real useful.
 Everything else I socket, and trimpot.
 Can change it to PNP from NPN...change the battery and 22uf cap polatiry. oh I use no polar caps for input and output DC blocking/ voicing.
 I use a large [2k2 ?] gain pot then tag R values across its outer lugs till I like the way it works...some transistors may need a bit more than the stock 1k, and a twist of the ~2k2's knob can bring the gain down.

[b_rogers]

i have read the article at geofex if thats the one you mean, it really concentrated on the germ faces with not to much talk about silicon. i just noticed there isnt much about hfe of si in ff?

[stm]

As far as I remember, the first thread about Miss Piggy (7 pages, around January 2004 perhaps) stated using two silicon transistors in parallel without any resistor on the piggyback trannie, which gave hfe's in the order of 3 to 10, and was reported to sound very good.  Then someone suggested using 1k and 2k (for the first and second stage, respectively) which gave hfe's in the order of tens.  Later a version appeared using 10k resistors or so, which I guess limits gain to around 50.

All versions are supposed to soung good in their own right, and I am sure there is no consensus as of know regarding "ideal" values for silicon, as opposed to germanium devices.

So far, IMHO the best option is to replace the hfe limiting resistors with 20k pots and trim by ear.  I made a circuit myself, but there were too many degrees of freedom and wasn't able to get a sound to suit my tastes.  The ROG FF versions there seem appealing to me, especially that with higher treble content, the one stated to "sit well in a band mix", or something.

[petemoore]

Pickups, amps, caps, resistances, transistors, pots, Si Ge, etc...alot of parametere and opinions on the matter...great for guidelines...
 Texted descriptions and soundclips aside, I think a little tweekin' is in order for great FF performances, or alot.
 A flexible FF board, and some different type Q's is loads of fun to tweek on...Don't forget to try differemt input cap values, re-bias after Q swaps, and turn the gain knob, and vol knobs of the Circuit output and the amp....oh yes turn the guitar volume knob up and down while tryin x setup.

[brett]

Given that some people even liked the JH-1 and JH-2 with extra high hFE (including MPSA18s with hFE around 800), there's obviously some experimenting that can be done.  

My advice for Q1 is that if you don't get hFE down around 70, and load your pickup, you'll probably want to get hFE over 150.  There's a real flat spot around 100 to 120, which I believe sounds dull because the input impedance is around the 40 or 50k ohms mark, which is neither a large enough nor small enough load.  This might explain why the JH fuzzfaces at least sounded clear.

Piggybacking a couple of 2N3904s with a 10k emitter-bridging resistor makes a cheap low hFE, low-noise tranny.  They're easy done and give you 90% of the magic of vintage Ge or Si. :wink:

[b_rogers]

oooh Run Off Groove...r.o.g. i get it. i have a multiface that i tweak with a lot.
piggybacking is kinda cool, i was trying some combos out when that thread was new and found quite a few i liked, but i found a great combo of a bc338 in q1 which is a audio output driver, and a 2n7000 fet for q2. bc547c also sounded good in q2 and it was pretty hot, more of a heavy metal type tone. the 338 - 2n7000 combo is more of a "tubey" tone with less fuzz and EXELLENT cleanup.
then i added a modded bmp tone control and i also changed the fuzz pot cap to a 33uf. run a strat- multiface thru a clean fender style amp cranked to just breaking up and the tone is creamy jimi goodness and LOUD.
i need to make a pig face or ms piggy after i finish the RM axis. of course i wouldnt have pulled it off without a lot of help from the guys here at the forum.
someone should "borrow" eric johnsons FF and blueprint that baby! (just kidding)

[brett]

i found a great combo of a bc338 in q1 which is a audio output driver,

Yeah, back in the days before RG suggested piggybacking, I was trying mid-power trannies.  Apparently, Gus Smalley (?) was an advocate of these, too.  Lowish hFE, highish junction capacitance.  I used BD139 (NPN) and BD140 (PNP), but they often have hFE a bit too high (120-150).  Unfortunately, full-on power devices like the MJE3055 and others are a bit expensive (@$2 ea) for cheapskates like me. :oops:   The TIP series is another area that could be fruitful...

[b_rogers]

what is the with  TIP? is the difference just package, or are they a different device? i have noticed them at radio shack before.. anyone used a TIP in a FF? what #s?

[phillip]

Use a PN2369A in Q1 and a 2N3904 in Q2, and put a 100pF capacitor across the collector-base of Q2 to help roll of some of the highs that silicon transistors seem to introduce into the circuit.  With that arrangement you're very close to the Germanium sound.  The Germanium transistors are still a little warmer sounding, but not by much!

I like to adjust the values of the input and output capacitors a little to help get rid of that "buzzy" bass, since the Fuzz Face is a little too muddy sounding for me.  I also like to use a 100KA volume pot, which lets out some more highs.

Phillip

[b_rogers]

i like the 100k for volume also.

i am really satisfied with my "bluesface" it sounds really unique and not really fuzz like. i even tried it with a "marshall"  tone stack and adjustable diode clipping for a while but it sounds best just a 2 knobber.

i will try the 2369A if i can find a few.  the 2n3053's sound really good too if i recall. that multiface can give you a nervous breakdown with all the tweaks.

[phillip]

The PN2369A is a pretty low gain device.  All 30 that I ordered from Mouser had gains of between 50 - 53, but they work great for Q1.  The "white face" 2N3904 that I got from Steve Daniels a while back all had gains in the 120 range, just the ticket for Q2.

The PN2369A transistors from Mouser are $0.09 I think...the Fairchild Semiconductor type.

Phillip

[brett]

anyone used a TIP in a FF? what #s?

The TIP series includes mid-power (rated at a few watts) mid hFE devices.  I've used a TIP41 in a fuzzface, and it worked well.  There's also TIP31 and 32, and others.  Another device that I didn't mention is the MJE340.  From memory it was quite good, too.

[brett]

Use a PN2369A in Q1

I just got some from mouser, and they are spot on.  Consistent hFEs of 65.  I also got some BD139s, and by chance they were all hFE 130 to 140 (I say by chance because sometimes I've bought them with hFE around 150).

I'll surely be building some fuzzy boxes this weekend.    

[phillip]

I think you've got the makings of a winner on your hands with the Silicon transistors that have gains right at what the Germaniums should be  

If it's not "warm" enough, try putting a 100pF capacitor across the base-collector of Q2...I like to use Silver Mica...Ceramic adds some annoying "fizzy" sounds.

Phillip

[b_rogers]

has anyone tried sending a little of the clean input signal to the volume control after the fuzz out?

what i mean is a separate signal like a "blend" of the two sounds???

would it be as easy as a resistor/cap to a pot with fuzz input on lug 1 , out to vol control on wiper, and clean input on lug 3?  

would a diode be needed to keep the clean signal from getting fuzzified?

is it do-able? let me know what u think..this may just be a newbie retarted question so excuse me if it is... :oops:


Brent

[brett]

has anyone tried sending a little of the clean input signal to the volume control after the fuzz out?

The Maestro FZ-1 and 1a have "blend" controls that do just this.  It uses a dual gang pot.  Very simple.  A search for "Maestro" and "schematic" should find it.

[b_rogers]

hey thanks for the info brett!

[Fret Wire]

With such a small range of hfe to work with, even silicon's can be almost as hard to match as Ge's. Sometimes you get lucky, sometimes not. Silicon's good consistancy usually means that all will be useable, or none will.

I just got a Mouser order in for some FF trannies, and I hit 2 out of 3. I've tested 30 of 100 of each so far. The PN2369A's are testing out at 69-74. The 2N4400's are running at 107-115. The disappointment came with the PN2222A's. So far, they're testing at 200-210. The last batch of PN2222's I had ran from 110-140. No big loss though. Judging from the data sheets, it looks like I can use them anywhere you'd use a 3904. Another recent strikeout was TO-18 metal can 3053's, which all tested out at 200-225.

So, testing and matching wise, Si's are no less work than Ge's. But....the difference is that the 2369A's, 2N4400's, and PN2222A's were $.05 apiece, not $2 apiece. I also found some 2.2uf metal films that are smaller than a dime. I may sell some of the extra's in the For Sale section.

BTW, you know you're a pedal freak when it bothers you that your great sounding Fuzz Face has plastic TO-92's inside, not metal can TO-1's or TO-18's. :shock: