Fuzz Face on the fritz

[CJ]

got a nice piece in that some lucky bloke found in his Granddad's basement, one of those stories,

 

[CJ]

somebody had the battery terminals reversed,  maybe they thought this was a Germanium 32 model.

this one had BC130c transistors,

 

[CJ]

this one had a tropical fish output capacitor, maybe that's why it sounds so nice, never seen any pics on the web with this cap,

 

[CJ]

so we reverse the batt terminal, the thing works but has some scratch in the pot, we spray the pot, now it starts picking up every radio station in the northern hemisphere,  some fuzz face boxes do this alright, but this is pretty bad,

we sub in a new pot, no difference,  wtf?  thought we had this ready to go, dang,

turns out we had moved one of the transistors a bit, there was a bad solder joint on the base of one of the transistors,

no wonder the box was in the basement,

bad joint = blue arrow

 

[CJ]

schematic with voltages, AC red DC blue,  big voltage divider on top of Q2,  Fuzz control determines NFB ratio,

 

[CJ]

waveform at output

 

[CJ]

kept forgetting the circuit so we went down to Spider Monkey tattoo shop >

 

[rock soderstrom]

LOL! ;D

 

[PRR]

> kept forgetting the circuit so we went down to Spider Monkey tattoo shop

The Monkey got the 330 value wrong.

Thanks for Granddad's voltage notes.

 

[user 37518]

schematic with voltages, AC red DC blue,  big voltage divider on top of Q2,  Fuzz control determines NFB ratio,

Jesus christ, its simply two cascaded transistors and this thing costs  more than a thousand https://www.ebay.com/itm/VINTAGE-ORIGINAL-1968-DALLAS-ARBITER-FUZZ-FACE-FUZZFACE-GUITAR-EFFECT-PEDAL/323036833977?hash=item4b367ee8b9:g:uw4AAOSwMfhabgzP unbelievable.

BTW I just noticed in the schematic that there is a 100k resistor in shunt-series feedback but no series resistance at the input, from the input perspective, the 100k resistance is in paralell with the emmiter resistance times beta (beta*re) of Q1 and the result gets divided by the loop gain, if my calculations are correct and considering maximum gain setting (with the 1.8K resistor fully bypassed by the capacitor), the open loop gain at the point of feedback is around 30 (the 8.2k and 330 ohm resistors are out of the FB loop and that, I believe, is how most of the distortion is achieved, so the 100k resistor is mostly DC feedback), the beta factor is 1/100K, so the loop gain is roughly 1, that means that the input impedance is 105(100)||100K/1=9.5K (a beta of 100 was assumed)  isnt this a low impedance for a guitar input? or is that the trick to achieve a fuzz sound?

 

[CJ]

there is a lot of hi end rolloff with this box, maybe that is why people like it, get's rid of that harsh buzz saw sound that bites at the ear drums, this rolloff could be becayse of that Lo Z you are talkin bout,

noticed the 330/470 resistor change, note that the tattoo version has the PNP NKT276 parts, thus they must have changed the divider to compensate for more gain of the Si parts,

wow, $1600,  that guy is really fishin hard,

 

[Newmarket]

Jesus christ, its simply two cascaded transistors and this thing costs  more than a thousand https://www.ebay.com/itm/VINTAGE-ORIGINAL-1968-DALLAS-ARBITER-FUZZ-FACE-FUZZFACE-GUITAR-EFFECT-PEDAL/323036833977?hash=item4b367ee8b9:g:uw4AAOSwMfhabgzP unbelievable.

BTW I just noticed in the schematic that there is a 100k resistor in shunt-series feedback but no series resistance at the input, from the input perspective, the 100k resistance is in paralell with the emmiter resistance times beta (beta*re) of Q1 and the result gets divided by the loop gain, if my calculations are correct and considering maximum gain setting (with the 1.8K resistor fully bypassed by the capacitor), the open loop gain at the point of feedback is around 30 (the 8.2k and 330 ohm resistors are out of the FB loop and that, I believe, is how most of the distortion is achieved, so the 100k resistor is mostly DC feedback), the beta factor is 1/100K, so the loop gain is roughly 1, that means that the input impedance is 105(100)||100K/1=9.5K (a beta of 100 was assumed)  isnt this a low impedance for a guitar input? or is that the trick to achieve a fuzz sound?

Yes - it doesn't have the usual high input impedance of a guitar amp / pedal (typ 1MOhm).
Look at it like the pickup impedance is the series impedance to the base.
If you feed these from a low impedance source then the general consensus is that they sound bad  (and not in a good way .
Hence they don't match well with 'active' guitar / bass pickups/preamps.
In that instance it's possible to intentionally add series impedance to improve the result.
For same reason it's usually the first pedal in a chain straight after the pickup.

 

[Gus]

The input of a FF like circuit is an inverting amplifier
The first section gain is the open loop gain of the first transistor along with the feedback/Q1 bias resistor from the 2nd transistor emitter / the RLC that the guitar is. Also note the 1st stage collector is closer to saturation. On some other FF like fuzzes the 2nd stage is very close to the supply voltage (closer to cutoff).
For a Si FF I often like About 1.6VDC at the first collector and 7 or so VDC at the 2nd collector

think of the FF as a not very good(low 1st stage transistor gain compared to an opamp) inverting opamp
Find the textbook picture of the inverting opamp change the feedback to 100K and substitute the RLCs that make up the guitar. Coil resistance inductance capacitance tone network volume etc. as the input resistor to the inverting opamp

 

[CJ]

thanks Gus!  good stuff there!

how do we tweak the voltages, change the 100K?

going to put this circuit on board the SG and see what happens,

what are your favorite transistors for this box?

lots of Germanium cans on evilbay, don't know which ones are fake, looks like some good stuff from USSR,

 

[user 37518]

thanks Gus!  good stuff there!

how do we tweak the voltages, change the 100K?

going to put this circuit on board the SG and see what happens,

what are your favorite transistors for this box?

lots of Germanium cans on evilbay, don't know which ones are fake, looks like some good stuff from USSR,

I saw a youtube video that compares the Ge pedal VS the Si pedal, honestly the difference is almost inaudible or Im too deaf to hear it.

 

[Gus]

CJ
Here is a link to a post of mine at Diystompboxes. Look around that site for ideas
http://www.diystompboxes.com/smfforum/index.php?topic=114153.msg1059002#msg1059002

I would leave the fuzz you worked on stock it is worth more that way.

2N2222 and PN2222 work well for Si fuzzes. I have posted about and built fuzz circuits with Si power transistors using the added capacitance and lower hFE

 

[walter]

I found one of those in a second hand shop in Dallas 17 years ago. It had a piece of tape on the bottom, on the tape written " J H EXP". $350 I didn't bite.

 

[Wildfara]

I build these all the time with Ge pulls of many different varietys...2SB56's work great Toshiba...2SA234. Hitachi ...I've used a gain as low as 50 on Q1 and maybe 90 on Q2. But 70 or 80 and maybe 100 is recommended...I usually ignore that tho. Mismatched gains with a lower gain in Q1 seem to sound best. I use a trimmer resistor in place of of the 8.2K to set gain and that allows you to dial in many different types. Its pretty fun and a good way to use up salvage parts you don't have the heart to toss out. The three transistor Tonebender is pretty much a fuzzface with a Rangemaster tacked onto the end...Its pretty fun too.  Simple circuits fun tones slightly unpredictable and weird.

 

[Wildfara]

For the Ge Fuzzface 4.5 to 5 volts on collector seems to sound best..and again the best way to dial in different types is to put a trimmer on that 8K

 

[Wildfara]

I just put them in whatever enclosure i can scrounge or build...Anyone who pays more than 100 bucks for a Fuzzface I would think was foolish...its 10 bucks or less of parts and some labor