Si FF rant

[Gus]

Look for the OLD Vox 2 transitor circuit  WEB has lots of good circuits, search engines are your friend(sometimes)
Vox did it years ago.  Its easy

10k collector on the first transistor.

1.5K and 8.2K 2nd collector

47K feedback

1K gain  pot 22uf 33uf  or what ever you want cap on the wiper
bottom of pot connected to a 820 ohm to 1.2K to (taste) fixed resistor other side of resistor to ground
OR A 2K pot and no extra resistor.

100k volume

input and output caps to taste

Should bias up just drop in Si small signal transistors.

Si transistors you can DESIGN for.  Trims in Si collector legs is bad design IMO.

[R.G.]

You're right Gus - but being right is not necessarily what the effects world wants.

The effects world is peopled primarily with music people who can learn some electronics. People with hard sciences backgrounds thin on the ground.  This means that the fine points of design get lost in the emphasis on side effects. Things like predictability being more useful than occasional flashes of goodness. Or having to tweak something in being the sign of a poor design. Or having to design something out of solid unobtainium because you can't do it any other way.

The music world deals primarily in emotion. The best musicians are those that can emote through their sound, and that attracts an intuitive, emotive kind of person. It is very hard if you've nurtured that side of  yourself to flip over to the solid, hard line of numbers, equations and conservative design principles.

I'm glad that we have this thing off the ground to the point that it's starting to attract a few more solid sciences types than the handful that started it. And we are teaching quite a number of crossovers with open minds. We'll never reach them all, but we have given a lot of neophytes a good basis for how to look at things even if they still have to go dig out a lot of the sophistication on their own. Quite a number will make it.

I view this forum and the very few others like it as building a bridge. It's a way for the technically unschooled to cross over into what is a very foreign country. And a lot more people are making the trip. Enough that I have to keep reminding myself that I'm gathering beginners up to start across, not getting the ones that make it across to know all of the language and customs.

[tcobretti]

I respectfully play devil's advocate by pointing out the Roger Mayer uses a trimpot to bias Q1 in his Si FFs.

Granted, that's different from a half dozen trimpots biasing as many jfets in one of the amp emulators.

[MartyMart]

If you're building 50,000 units, then some 50c trimpots make a huge dent in the "cash"
If you're building two for yourself then that 1 dollar is a worthy investment for some
"fine tweaking" - which I personally do not see as "bad design" !
Recalling 175,000 Volvo's because the engine management system is shit and makes
you cut out on the freeway .... is bad design 

MM

[Gus]

I think my post did not come across in the way I meant.

I have no problem with people building and learning. I think it is good

I am suprised that even after all the years stuff has been on the web people keep building a Si FF with the Ge circuit.  Ic is maybe not the best....

If one looks at the ff type circuits in schematics with my name you will see a bias pot.  Now I dislike small bias pots but in my own builds the bias pot is a full size and external mounted and can set the ff circuit from gating to compress so it is a part of the design I could use a 1k fixed but I would not have the other sounds.

Most people here seem to want an ungated Si FF.  The stuff they find on the web and build will often gate or so it seems from posts here.  Not good if the person wanted a paint by numbers build (this is not a bad thing it can be a very good thing, one needs to start some place) they might have built it totaly correct.

A side story I make no great claims for the NPN boost.  The NPN boost started as a self set challange.  I wanted to make a booster that was not like other boosters and low parts count AND would work with almost any Si transistor.  The variable input Z and bootstrapping  are a little different  I have not seen a booster done like that before.
  After building it friends liked it and a few pros have one or two.  So after that I gave it to Aron as a beginner build.

I think a Si two transistor fuzz that works with a wide spread of transistor specs might be a good thing for people to learn on.

[orgaqualia]

You're right Gus - but being right is not necessarily what the effects world wants.

The effects world is peopled primarily with music people who can learn some electronics. People with hard sciences backgrounds thin on the ground.  This means that the fine points of design get lost in the emphasis on side effects. Things like predictability being more useful than occasional flashes of goodness. Or having to tweak something in being the sign of a poor design. Or having to design something out of solid unobtainium because you can't do it any other way.

The music world deals primarily in emotion. The best musicians are those that can emote through their sound, and that attracts an intuitive, emotive kind of person. It is very hard if you've nurtured that side of  yourself to flip over to the solid, hard line of numbers, equations and conservative design principles.

I'm glad that we have this thing off the ground to the point that it's starting to attract a few more solid sciences types than the handful that started it. And we are teaching quite a number of crossovers with open minds. We'll never reach them all, but we have given a lot of neophytes a good basis for how to look at things even if they still have to go dig out a lot of the sophistication on their own. Quite a number will make it.

I view this forum and the very few others like it as building a bridge. It's a way for the technically unschooled to cross over into what is a very foreign country. And a lot more people are making the trip. Enough that I have to keep reminding myself that I'm gathering beginners up to start across, not getting the ones that make it across to know all of the language and customs.

This is whats happening to me. I have a very hard time learning from books. Ohms law has no meaning for me untill I see (hear) it in action. I have a library of EE texts and I have learned more by futzing around here and making some noizemakers for me and my friends than I ever did in any class. I first started doing this because I was interested in building my own studio gear. I'm a drummer and I don't even play guitar well enough to really evaluate my builds, so I go to a friends house and we do it together.... long winded huh? sorry.
My point is that this forum is an amazing learning resource. I'm taking guitar lessons, I wanna go to back to college for EE, and cheezily enough, my life has changed because of this forum. If I couldn't be stupid about things, I would never learn.
Thanks again and again to everone here for sharing their knowledge and experience with me and everyone else with an interest.

[burnt fingers]

I personally like the idea of a trim pot.  I build a FF variant and tuned the trim pot by ear.  I heard waht I liked and then got my MM.  To my surprise I was at 6v.  At 4.5 it just wasn't the sound that I wanted. 

Scott

[Paul Perry (Frostwave)]

I guess what people are learning is that the distortion or fuzz or whhatever you want to call  it, can come from more than one point in a "fuzz face" derived circuit. It can be a transistor clipping, it can be the signal cutting off the biasing, it can be the non-linear transfer function of the transistor itself.
So no wonder 'automatic bias' or 1/2 V bias, limits the scope.
Too many knobs? Can't happen (except when you are gigging, but that's why they invented locking pots. Not that you can afford them..)

[lenwood]

Should bias up just drop in Si small signal transistors.

I am the person like RG suggested, a 2 year degree in Electronic Technology and more experience as a emotive musician.

So my question is,  even with this VOX design every Si trans would bias up but would they not sound different because of HFE, etc?

How could you account for these differences in transistors? i.e. someone putting in a Q1 hfe= 400   Q2 hfe=100

[WGTP]

I'm sure it is very frustrating for technical types, but a lot of us that come here are winy dump ass guitar players that can't change our own strings or tye our own shoes, or spell.  But, if you get addicted to enough drugs, get busted and go to rehab enough, get enough bad publicity, bang enough models and some tech's to handle your equipment, you too can be a STAR.  Is that what R.G. was talking about?   

[R.G.]

So my question is,  even with this VOX design every Si trans would bias up but would they not sound different because of HFE, etc?
How could you account for these differences in transistors? i.e. someone putting in a Q1 hfe= 400   Q2 hfe=100

That's where the design process for consistency comes in.

There are two ways to get things to be exactly as you want them. One is to tweak and adjust things to be just right. The other is to have more than enough of whatever it is, and set things up so that you only get as much of it as you ask for at the moment. That first way is done in electronics by trimmers and adjustments, the second is by the use of feedback or digital logic.

Let's dispose of digital logic first. At an analog level, digital logic elements have massively more gain than they need to make it from fully-on to fully-off on every input transition. When you put in a logic level at the input of a logic gate, the gate is massively overdriven from the analog sense. Logic gates are always banged against the power supply limits, always "clipping". That's how the waveforms stay square.

For things like analog sound, you are left with either tweaking or using feedback. If you choose tweaking, then the likelihood is that everything you want to get right has to be tweaked for each new device. Feedback moves the variation in results away from the vagaries of the actual devices and into a predictable component, like a resistor.

Opamps are the perfect example, because they were designed specifically to take advantage of the principles of feedback. Without feedback, an opamp has a gain of zillions. With feedback, all of that gain of zillions is used up trying to make the output be perfectly controlled by the feedback components. Opamps have an even wider variation in forward gain than transistors. Some opamps may have an open loop gain of "only" 100,000. Some have gains of several million, at DC at least. But put them in a stock gain-of-ten feedback circuit and the differences condense down to the feedback-stabilized gain being 9.9999982347.... or 9.99999999999...

The same is true with transistors. It is possible to design transistor circuits without using feedback. It is not possible to do that consistently without pre-selecting or matching the transistors. The matching we do for JFETs in a P90 phaser is an example. That is a situation in which the characteristic we want (Rds) is not easy to stabilize with feedback. So we have to match JFETs. That's a mess, not to mention wasteful of about 2 out of three devices that don't work well. It's possible to "design" gain circuits from bipolar transistors the same way.

Emitter resistors are feedback elements, and so is the 100K from the emitter of the second transistor to the base of the first transistor in a fuzz face. The capacitor that bypasses the 100K feedback resistor wipes out the AC feedback and lets the FF run mostly open loop for AC but be DC stabilized, which is why the transistor characteristics "show through" on FF's.

But that's with gains of 100 (about) in the transistors. When you get to silicons with much higher gains, there is much more gain available for feedback, and you can take advantage of that to make the design more consistent. It's just different from what it is with germanium.

Where feedback amplifiers sound different is where they run out of feedback. When a feedback amp hits a power supply limit, the instantaneous gain goes from the open loop value to 0. There's probably some tiny range in there where the gain is no longer, say one million, but drops to 500,000, then 200,000, then 100,000, then 1,000, then 100. The open loop gain that hid the amplifier's real nature shows through in that range as it runs out of open loop gain. The higher the open loop gain, the tinier that transition region. So the higher the open loop gain, the more "perfectly" the amp acts until it can't do it any more, and the sharper the transition into clipping.

So to cover up variations in transistors, you design with feedback. In a silicon FF, the variations in DC bias are going to be much less because of the higher gain, lower leakage and smaller thermal drift. Beyond that, it's a matter of finding out exactly where the clipping takes place when it runs out of gain and messing with that to get best sound.

The coverup will not be perfect. But it will be more predictable than with lower gain devices.

[R.G.]

I'm sure it is very frustrating for technical types, but a lot of us that come here are winy dump ass guitar players that can't change our own strings or tye our own shoes, or spell.  But, if you get addicted to enough drugs, get busted and go to rehab enough, get enough bad publicity, bang enough models and some tech's to handle your equipment, you too can be a STAR.  Is that what R.G. was talking about?

It is, it is.

Quoting one of my heroes, the drummer for Spinal Tap, "I guess I could do without the rock-n-roll if I got the sex and drugs..."

[WGTP]

That's what I'm talking about.  I will probably never understand what R.G. just posted in this life time, maybe the next.

Did Hendrix understand all this stuff???  Where would he have been without Roger Mayer, electronics geek???

Stereotype Warning.  I notice there are lots of Bass players here, and in my experience, they seem to be slightly less flacky and more technically minded than us guitar players.  Plus they have better social skills because they don't spend as many hours in their bedrooms practicing alone.   

[R.G.]

Stereotype Warning.  I notice there are lots of Bass players here, and in my experience, they seem to be slightly less flacky and more technically minded than us guitar players.  Plus they have better social skills because they don't spend as many hours in their bedrooms practicing alone.

You must get a better grade of bass player than I got back in the day. Our bass player was pretty much carted to gigs in a cage and let out to play. The hard part was getting him back in the cage when it was over. 'course, sometimes he was, um... chemically enhanced... into a pleasant, nodding state. This made it much easier.

[WGTP]

Yes, I know some of those too, but I think they are really rhythm guitar players that never learned the chords.   

[modsquad]

ROFLMO 

I have to go with RG on that one.  My bass players were always like bad chicken you got rid of them as soon as you could.  Don't take me wrong guitar players and lead singers have their own issues (Axl Rose).

I understand what R.G. is saying.  I fell into this addictive hobby by modding a DS-1.   When I opened it up, I was like is that all.  Then I found this site and realized I could build anything I damn wanted for pennies on the dollar.   Now I have a software engineering background so electronics was foreign to me.  But my analytic side made me go out to the half price store and buy every book I could understand.  I still don't have a clue what I'm doing but at least now I can understand the terminologies and theories enough to debug, substitute, come up with ideas, etc. 

[WGTP]

I'm really sorry, but I had to do this.

What is the difference between a blues band and a moose?

With the moose, the horns are in front and the ass hole is in the back.   

[Doug_H]

Re. hfe differences in a Si FF- I suspect you could hear slight tonal differences but the circuits should at least bias up consistently.

I think the trimpot in the drain/collector/etc circuit has been done to death and misapplied quite a bit. It's really a stupid way of biasing a circuit, if for nothing else, the way it will change gain and make the gain of a stage inconsistent from device to device. I suspect it got started with the JFET "amplike" circuits as a quick & dirty way of dealing with inconsistent JFETs. There are app notes on JFETs that describe many alternate methods for biasing. I suppose some might be more suitable for dealing with the inconsistent JFET but I haven't researched it. In the end it will further complicate a circuit that is already more complicated than is necessary IMO.

There's a lot of hacky sh*t going on in pedal circuits, even in the pro and boutique stuff. Some of the "designers" don't seem to really know much about electronics. Nothing new there. So I think a good sounding pedal circuit that is device-independent and is a product of good design principles is a great idea for beginners. People who want to do something akin to a kit-build can trust that it will work with a wide variety of devices and will maintain stability regardless of weather, sunspot activity, or what their horoscope said on a given day. And for people that want to learn how to design, it might help them learn how to do it correctly instead of just hacking their way through.

[Gus]

I only wanted consistency in that the two transistor fuzz would work without gating.

  One can pick parts to get it in an area that will sound like it is not broken.  This will hopefully help the beginner because it works, maybe not the sound they want but then they can change things one at a time and learn and find what they want.  Sometimes you need a good reproducible starting point.

If it is working then changing transistors and operating points could/should be done.  Lots to be learned from this even high Hfe transistors have there own sound in a simple circuit like the FF type.

I never posted 1/2 battey voltage is the spot.  A ff type I built has the external bias knob marked gate to compress players understand that and have fun trying the different sounds.  When you get deeper into circuit operation you will find sometimes you want the output offset for a few reasons.  I did post 820 ohm to 1.2K as starting points.

EAR/BRAIN and MATH can work together.

[JHS]

Due to the part tolerances those old Vox-distorters, FF and other fuzzes sound all a bit different, some sound more gated than others.
Even trannies of the same type and the same lot can sound totally different and any trannie-type has it's "own" sound.

There's no need for a BIAS-pot as long as both transistors have the "correct" hfe. Considering that most circuits are taken from the datasheets nobody cared about biasing those trannies in the sixties. Producing a high gain Si-trannie with hfe of more than 350 was impossible at that time and most old fuzz-circuits from the web work quite good with original old trannies and the values from the schem.
Those old Si-trannies from the sixties often sound very soft compared to new ones and a slightly misbiased trannie won't ruin the sound.

If you use new Si-trannies some part values must be slightly adapted. I don't like any biasing pots in my circuits, a pot produce more signal and highend loss than any resistor (the same with trannie and IC sockets). For biasing I put in a larger resistor, mesure the voltage at the trannie and wire another resistor parallel to get the correct Bias-voltage (it's no big problem to calculate the value of the parallel resistor).

JHS