new circuit - facetavia

[joegagan]

there is a young guy, 23 years old who lives in my town, has lurked here for years and never posted. his name is zac zagar. i will tell you more about him later, but he has been quietly studying and building circuits, we got together and messed with some circuits.

while i was at his shop last week he built my easyface, we tuned it and it sounded good like they usually do.

but while i was gone, he added a simple transformer ge diode octave circuit. sounded excellent. last night i went over and we put some more tweaks on it, here it is. i am very impressed with the gnarly tone of this, i do not say it lightly.
more on this guy later, i gotta go, but let me tell you , the younger generation is taking this stuff on in a big way!
breadboard one of these, let us know what you think. all comments welcome and appreciated.

[aron]

Alright!!!!! That transformer output has always been fun!

[puzzle87]

Between the central terminal of a potentiometer GAIN and a point of connection of the resistor 75К with base of the first transistor the condenser is necessary for exception of influence of a potentiometer on a DC-bias. Otherwise the device will be wrong to work.

[John Lyons]

Don't you usually put the gain pot first then the wiper to the 2u2 then the base of Q1 Joe?
This solves the issue mentioned above.
Then you can save the 2M2 for a rainy day.  

Looks fun! Why the dual 25k trimmers?

[TELEFUNKON]

crackle okay?

[DougH]

Looks cool, Joe! Octave pedals are always fun.

[joegagan]

good questions.

john, zac put the 2.2m at input out of habit. i think the various easyface schematics have the gain pot and input cap reversed, he must've used one that.

the dual 25k trimmers is to give the ability to increase the output of the straight fuzz, while still achieving correct bias. this is the only aspect of this circuit not tested ( will do so today). it will be a bit of a balancing act to get both of them optimal, but it will be worth it. the octave will also benefit from more gain IMO.

the gainstage for the octave was necessary to get comparable volume between the two operating modes. the octave side was too quiet. BTWm the .047 uf  to ground at octave input is designed to replicate turning down the guitar's tone knob. it also cut out some of the nasty high end razz we were hearing. someone may also want to try instead .022 and possibly a series resistor as well at this junction.

the 1k trim at q3 emitter allows user volume matching of the two modes. non DIY users will be discouraged from messing with the 25k trims, however. telefunkon, i did not hear it crackle, OK? will listen more closely for that tonight. it will be an internal trim so, not commonly tweaked. one of the main reason for this control is that different types of pickups react differently to the two modes, volume-wise.

back to the zac story. he was 13 yrs old when the easyface was introduced. now, he has paper copies of billions of circuits, he has studied them well. he is very familiar with the dozens of regulars who have contributed here over the years. his shop is reasonably organized, well equipped, and fully stocked to build just about anything you can think of. amazing. he is a fan of the RAT pedal and already has a modded version of his own design as well.

the funny thing about this easyface/octave circuit is that, to my knowledge no one has ever tried this combo before.  goes to show that fresh perspective can bring out the overlooked and obvious that us salty veterans sometimes miss. i hope zac will step out of lurker status and start posting.

[head_spaz]

Sorry for the interruption... but I have a question... and it's slightly off topic.

Would it make sense to A/C couple the output, and forward bias the output
diodes with DC to eliminate the crossover distortion?
I've always found crossover distortion to be kind of ratty sounding.

Just wonderin'
Carry on...

[aziltz]

Sorry for the interruption... but I have a question... and it's slightly off topic.

Would it make sense to A/C couple the output, and forward bias the output
diodes with DC to eliminate the crossover distortion?
I've always found crossover distortion to be kind of ratty sounding.

Just wonderin'
Carry on...

I was wondering about that too, eliminating crossover distortion..

[Zzagar]

Hello everybody, I am Zac Zagar - Proximity Electronics.
Joe thank you for your introduction and for shedding light on this newly developed circuit. Also thank you to all of you that have shed insight into ways to improve the Facetavia.  As Joe said, I have been studing and building circuits non stop since January, 2007.  Without these forums non of this would be possible.  You will see much more of my work within the forum in due time, as well as, Joe and I will be working on some future collaborations.  In the next coming months I will be releasing a pedal line under the name Proximity Electronics.  And will also spend more time posting then lurking   

[joegagan]

Sorry for the interruption... but I have a question... and it's slightly off topic.

Would it make sense to A/C couple the output, and forward bias the output
diodes with DC to eliminate the crossover distortion?
I've always found crossover distortion to be kind of ratty sounding.

Just wonderin'
Carry on...

good question. i didn't give any thought to the reverse bias of the diodes. i am just used to seeing them biased this way in other transformer octave circuits. i liked the way the regular fuzz mode was unaffected even though theoctave output was still tied in ( allowing use of an spdt)

nice to you post, zac.

[DougH]

Gus's Octave Up Sick Box fwd biases the diodes with a little DC.

[Slade]

Nice, can't wait to try it, I have a transformer here, so I'll be giving it a try very soon. Thank you for the schem, Zach and Joe!

[PRR]

> a young guy, 23 years old

I was much smarter at his age than I am now. I "know more facts" but can't put them together like I did in my 20s.

> base of the first transistor the condenser is necessary for exception of influence of a potentiometer on a DC-bias. Otherwise the device will be wrong to work.

Yes, But, if the "gain" pot is set well above half-way, the DC bias is not very wrong. When turned down, of course the circuit goes "dead". I think this circuit will usually be worked turned-UP, so this "flaw" may not be important.

The more conventional way to do it would be like this:



With the new cap position, 2uFd may not be the optimum value; that's why you need a hot iron, a box of caps, and a good ear.

> forward bias the output diodes with DC to eliminate the crossover distortion?

The "doubler" effect is ALL about crossover distortion. And not a little lameness around zero, but a complete FLIP OF PHASE as the signal passes through zero.

> forward bias the output diodes...

With "ample" forward bias, the output is "nothing". Both diodes conduct, passing opposite polarity, and cancel-out.

There may be some very small current which optimizes very small signal doubling. But with Ge diodes and ~~500K load, that current may be very similar to diode thermal "leakage" current. And guitar effects are usually not about "small" signals. When you play LOUD, this does not matter.

There are more "perfect" ways to do this; also complicated and fussy. You have to trim-out more error the smaller-signal you demand; perfection is impossible. I know a guy been tweaking a signal rectifier for 6 months and still not happy (he's doing teeny-distortion research and needs incredible specs). Smacking several/many volts onto 0.1V diodes is a simple easy elegant trick which does more with less.

I simmed a "perfect" absolute-value circuit (what the CT winding and 2 diodes approximate) and flogged my Pentium for many minutes. It computed Total Harmonic Distortion at over a Million Percent. Huh? Sure: THD is ratio of unwanted-stuff to wanted-stuff, with the assumption that the input signal is "wanted". In the doubler, ideally the input signal is "GONE", canceled. So the "right" answer is infinite THD; the "millions" answer is rounding-error in the cancellation computation.

A more useful metric would be ratio of total output to input. Referenced to the 1KHz input, the 2KHz output is 0.424, 4KHz is 0.085, 6KHz is 0.036, 8KHz is 0.020. The output is very far from a sine, so we have to be careful how we specify. The peak-to-peak out is half the peak-to-peak input. The RMS output is 0.43 times the input. "Loudness" does not go exactly by either measure, and the complex output may tickle the ear more, but not a lot more. So there is "loss", and we have to apply gain (in transistor or transformer) to get a "same" output. (Also, in practice, to swamp the diode deadzone.)

[joegagan]

thanks, paul. very enlightening.

yes, we had already made the gain pot/input cap swap on our design, not posted here yet.


your post makes sense to what i am hearing. we modded the breadboard circuit last night to reverse bias the diodes, the result was interesting, a nice synthy sustaining fuzz ( esp when a quad of siidiodes to ground was added at output), but the distinct upper octave  gnarly fuzz was missing- so we went back to the original config.
i will re read your post to digest it better.

zac enjoyed everyone's input yesterday, this place is great.

[DougH]

Joe, I replied to your email and mentioned that I don't remember why some of these circuits apply a small DC bias to the diodes. Gus did it in his OUSB and I did it in an octave fuzz I designed a few years ago (partially inspired by the OUSB) but now I can't remember why.

[joegagan]

thanks doug. i liked what my ears told me in the end (as usual). it sounds like a bobtavia on steroids.

[joegagan]

update, zac changed the input, no more 2.2m resistor.

[petemoore]

  Hey Zzagar ! Joe ! Paul !
 Really enjoying this thread !
 Tried similar with Mosfet, there was a diagram I followed, basic Mosgain set up using GeO follies article IRC.
 
   

[Zzagar]

Anyone been able to bread board this yet?  I'm excited to hear your guys' results. 

Zac