Tube Fuzz Face

[anotherjim]

My favourite Americanism is "it don't make no nevermind". And it don't that's true.

[mozz]

When you wire the filaments in series, it's 12.6v, 150ma.(pins 4 and 5) When you wire the filaments in parallel it's 6.3v 300ma. So i guess you are wiring it in series and running at reduced voltage. Cathode poisoning, probably not in a small audio tube though.

[pinkjimiphoton]

or herbal jazz cigarettes will help. i actually talk like i write, so people must be sufferin' all over.

still in this thing. still messing with it. still running on 9v input. plenty of volume, finding where to tweak to get fuzz. the BITCH IS, every @#$%ing time you adjust anything at all, you need to @#$%ing wait for the @#$%ing bias to @#$%ing settle in again on the mother@#$%ing tube, which can be really super @#$%ing annoying, wayyyyyy more @#$%ing annoying than @#$%ing transistors are, which are really @#$%ing easy to @#$%ing work with.

extra points if ya figure out what @#$%ing word just got censored above 35 times.

and of course, at this point, the prototype has been molested so much that it is literally falling apart just about constantly, which is always my FAVORITE part of the whole damned process.

in the meantime, if anyone is dumb enough wants to follow me down this rabbit hole, take a peak at the wolf computer schematic for potential mods... many of them will work with this, as well.

i've committed to different resistances than originally used. typical scaling ya need to do up or down to find the sweet spots.

right now its volume, fuzz <gain,really> tone <valvecaster> and voice. voice is adjusting part of the feedback resistance between v1 grid and v2 cathode. lets ya go from cleanish compressed to a fairly decent distortion, to ripping velcro at min settings.

voltages are @#$%ing crucial. and they change with everything. this is even more of a pia than a ge fuzzface is. you can get lost with the biasing for literally days... i've been messing with it more than a week!! insane shit.

i am also committing to not worrying about keeping the circuit as identical as i had... seems to be subtle improvements made by changing fixed resistances or even variable ones with a combo of the two.

anyways... onwards n upwards.

i'll try and make this easier.


see jimi.
see jimi build.
see jimi build fuzzface.
see jimi use a tube instead of trans sisters.
see jimi work on it.
work, jimi, work.
jimi is very stupid to waste his time on this.
waste, waste, waste.
but jimi is crazy,
SO jimi makes fuzzy goodness possible.
be like jimi. make the world fuzzier.
gotchem go bye-bye.

[pinkjimiphoton]

When you wire the filaments in series, it's 12.6v, 150ma.(pins 4 and 5) When you wire the filaments in parallel it's 6.3v 300ma. So i guess you are wiring it in series and running at reduced voltage. Cathode poisoning, probably not in a small audio tube though.

yep. series, but running the heaters on 9volts instead of 12. pins 4 and 5, yep. cathode poisoning. i like it.

i don't see cathode being damaged run at 9v. reduces emission slightly, tube lasts longer, slightly less transconductance possible.

https://www.diyaudio.com/forums/tubes-valves/31770-heater-voltage-range-hi-low.html

https://el34world.com/Forum/index.php?topic=20238.0

"

Here's the straight-poop on this topic:

The tube was designed to be operated at its rated heater voltage (6.3vac) and will deliver "most satisfactory service" (the common term in old texts), meaning "best chance of meeting/exceeding claimed specs, for as long as possible" when you apply 6.3vac.

"If your running at a lower voltage, I would assume you would have a slight rise in current." The unstated assumption for this assertion is that heater-power stays constant. In fact, it doesn't, but rather follows Ohm's Law. So, if you drop the 6.3vac to some lower voltage, less current flows through the same-resistance of the heater, and the heater consumes less power (volts*amps) from the winding.

The cathode was designed with a certain planned power consumption by the heater, which then heats the cathode. Less power consumption then means less cathode-heating. The possible effect has two facets:
-  Normally, a new cathode has an excess of electron emitting capability in its cathode coating. So under-powering results in no real change of actual electron emission, and the tube operates as it would with less heater power, though the tube envelope may be very slightly cooler.
-  A tube which has already had a long period of use may have partially depleted some of its cathode coating, and under-powering the heater results in less emission than the tube would have under normal operating conditions.

If your tube falls in the latter category, then for the same operating point, transconductance (Gm) will be reduced by some amount. That might translate to a measurable reduction in gain. Whether the measurably-reduced gain is audibly-obvious in use depends on the amp and how severe the gain reduction (or really, tube age/wear) is.

"

[anotherjim]

With heater power, there is something to do with the Getter element inside the tube. Most of us know there is a thing called a Getter in there and can recognise it, but because it doesn't have a schematic connection, we may not know what it's for.
As I understand it...
The Getter is coated with something that really likes bonding with oxygen. Everything else in the tube must not be allowed to react with oxygen since obviously, oxidising will be bad for reliability.
The idea is that any oxygen left in the tube will get stuck in the Getter.
Nothing is perfect, so there is still oxygen hanging around, some of it hiding in the materials but releasing over time.
Unless the heater is running at its designed temperature, the stray oxygen particles may not have enough energy to reach the Getter and may react with the heater itself or other elements. I think the tube is treated to a dose of microwave or induction energy at manufacture to heat things up so that the Getter can sweep up most the looser stray oxygen before use.

So underrun heaters can reduce tube life.
That we don't notice a reliability problem due to underrun is probably because we don't run tubes 24/7 like Military/Scientific/Industrial users did/do.

[Ben N]

extra points if ya figure out what @#$%ing word just got censored above 35 times.

Well, Jimi, it helps that you @#$%ing spelled it the same @#$%ing way every @#$%ing time, all 35 @#$%ing times.

@#$%!

You're the man.

[stallik]

Well, I had no idea that the getter kept on working. Always thought it to be a single use element activated during manufacture. Goes some way to explain why I see getter descriptions in some tube ads.

[Banjan73]

Hi Jimi /pjp or whatever.
I do understand mostly of the stuff you are saying. Have been watching some U.S. movies during the years, ya know! But when you really start to freak out, it could be hard to follow, hehe.

[stallik]

Actually, I find Jimi's writings quite eloquent, conveying both meaning and intention without giving secrets away. Of course, my understanding of it may have been tainted with fermented grape juice but nonetheless, down the rabbit hole I went. Trouble is, there's more than one rabbit hole and mine apparently went in a different direction and I got to thinking.

Submini, military tubes are meant to withstand extremes for a short time - maybe minutes. How extreme is extreme? Filaments as bright as my light bulb, sparks, arcs and banishment to the shed for blowing the circuit breakers while the missus was working is what I've found out so far.

Must have sounded good at some point cause the wife asked "what the @*$%ing hell is that noise?"

[Banjan73]

Must have sounded good at some point cause the wife asked "what the @*$%ing hell is that noise?"
[/quote]

[amptramp]

There is an interesting blurb on the Steve Bench audio site here:

http://diyaudioprojects.com/mirror/members.aol.com/sbench102/dht.html

on directly heated triodes showing a reduction in harmonic content when the tube filaments are run at reduced voltage.  He shows that with reduced heater voltage, the µ, transconductance and plate resistance remain constant over a large range of plate voltage and the results are shown for voltages from 10 to 50 in ten-volt steps.  I would assume the same thing applies to indirectly heated cathodes like those on a 12AX7.

This may be why your attempts to get fuzz from an underheated cathode seem to stubbornly stay in the linear region with a minimum of distortion.  It seems reasonable that a little more cathode emission may get you out of the range where the THD is low.  He measures the THD and operation at 10 volts on the plate seems to give the lowest distortion with reduced filament voltage.

He has several other pages showing harmonic content at various voltages and it seems to be minimized at low filament voltages.  This one on the 6B4:

http://diyaudioprojects.com/mirror/members.aol.com/sbench102/6b4.gif

shows minimum distortion on a 6.3 volt filament operating at 3.2 volts.  Maybe if you want a fuzz to fuzz properly, you need to run the cathode at the design temperature.

[pinkjimiphoton]

hi guys

its definitely working and boxed and even fuzzing, tho tube fuzz is much more transparent that transistors. its a "clean" fuzz, for lack of a better word.

the "getter" is the silver shit inside the glass. its purpose is, indeed, to take care of any oxygen in the tube.  but the key here is in the term vacuum tube. there's videos on youtube of how to make your own.

when a tube is manufactured,  the elements are placed in a glass tube and heated over a flame. this eventually drives out most of the air, and the end of the tube will be pinched and twisted... thats where the little nippile at the top of a 12ax7 comes from, in fact.

the getter's ONLY job, as i recall from back when i actually was doing tubes all the time and reading everything i could, is to remove any residual ATMOSPHERE from the tube. atmosphere will make the elements burn up; so if any air is left in there, its the getter's job to burn it off. this is the "sound of static" that tubes sometimes make while warming up. a tube that develops a leak, you'll actually see the getter oxidize, and turn from silver/black to white and milky.

generally, the only time you need to worry about the cathode being hot enough are with nixie tubes, or tubes nearing the general end of their life cycle. while emissions and transconductance <what we call hFE> are slightly reduced, the overall life of the tube is largely unaffected.

remember, line voltage. when most stuff for tubes was invented,  line voltage was significantly lower than it is today, used to be 110, but usually more like 100 when i was younger. now its usually closer to 125-130 coming out of the wall. now, this shouldn't matter, with transformers... they're gonna still produce that 6.3vac for the heaters just fine. but the rest of the circuit will still be affected somewhat. voltage could be high, could be low. hell, you could run your heaters at 5 volts, running off a rectifier tap, and they'll still work.

lowered heater voltage will make a weak tube die slightly quicker. but a normal tube will often last much longer, and they tend to distort MORE slightly low. i did set up a 12v heater supply <dc>and tried it to see if it were worth messing with... was thinking an lt1054 to bump it up to 18v and then regulate it back to 12 to run just the heaters on... i know, too much current requirement... well, almost. i looked at the data sheet, and that jellybean can handle up to 150ma current, so it would work at least for a while... but the 12v supply i used didn't really make much difference tonally. no more fuzz. somewhat less, in fact. can't think linear when making fuzz. shit don't work!!

so anyways, when all's said and done, i DID use a charge pump. the heaters run straight off the 9v supply. the charge pump is set up to deliver ludicrous and possibly dangerous voltage levels to run the plate. get a mess of 5817's and 10uF caps and have at it til ya find the sweet spot its there for the taking with some experimentation. don't try a 7660 or 1044, all it will do is whine.

the hardest part, again, is the goddamned biasing. the "sweet spot" to dial in v2 is ridiculously tight. and every miniscule, picayune adjustment made means you have to wait 2-3 minutes for the bias to settle and the tube to operate. it wants to see very specific voltages... all quite doable, i mean, my dumb ass figured it out.... but if ya go above or below, the tube will go into cutoff... cuz its being run in a VERY unlinear region. but once its right... you get fuzz. and tone. together, ya get fuzztone!

at this point, the only thing left i'm likely gonna do is add a hi pass filter to the input; it can get quite woofy like a real ff does.  rolling off the bass from my guitar makes a huge impact on the tone.

its a fair bit above unity gain, not as much as the average transistor ff, but enough to drive your amp.

to get ample volume out of it required a slight modification. look up my "fuzzface +" circuit and you'll see what to do. moving ONE component to a different node gains an easy 6db output or so. if ya think about it, you'll figure it out. you'll be like... @#$%! this can't work! and then ya try and go...cool!

so its got volume, fuzz, tone, and a "voice" control. all the way up on that last one is fairly "clean"... as ya turn it down, it gets more gated til ya get the classic velcro ripping sound of a misbiased 60's fuzz. incense and peppermints ? we got it. spirit in the sky? nope. nothing sounds that crappy

for the voice, add a variable resistance to the feedback loop between v1 and v2. again, wolf computer mods work pretty well.

i DID add a diode clipper, as well.  i wasn't gonna say where, but @#$% it. there's two places in the circuit where you can put an antiparallel to ground diode clipper. one is off the output cap. there's only one other place in the circuit you can do it. you gonna make me say where?

you're thinking... pink, that won't work, its improbable, its not done. but it DOES work. old radio trick. increase compression a little, and removes some noise by clipping the signal. make it assym.  i'd reccomend led's cuz they have the highest "compression" in this circumstance <for the newbies, yes, i am indeed talking about the clipper going to ground being located off the circuit side of the INPUT cap.>..... weird trick, but works. and ya don't lose the output volume you'd get at the end of the circuit. brings a little of the "woodiness" of the ge sound back to the circuit, and helps drive it a little better.

still sustains like a fuzzface.. almost like a compressor
LOTTA second harmonic. you can starve it into a bit of an octave up with the voicing.
cleans up like a fuzzface.

my advice, if ya follow me down this rabbit hole, is to do the biasing with knobs full up, and guitar rolled down to where you want it to sound "clean" or "cleanish" ...say, 2-3 on the guitar knob. bias to get max volume and best tone, and ya can't go wrong when ya crank the guitar up.

again, its not as saturate as a fuzzface.  more like an improbable cross between a boogie, fender, and marshall kinda tube saturation.  all the way up, is a nice compressy brown tube preamp sound. start starving it with the voice, and the distortion goes up. its kinda cool. the fuzz control and voice interact quite a bit, letting ya get a lot of different fuzz tones.

is it perfect? hell NO. was it a pain in the ass? well, i worked on it for a freeking week, lol... but was it worth it? hard to tell. ymmv. i like it.

can it be done?

hell yes.

[Ben N]

so anyways, when all's said and done, i DID use a charge pump.

Ha! I knew it!

still sustains like a fuzzface.. almost like a compressor
LOTTA second harmonic. you can starve it into a bit of an octave up with the voicing.
cleans up like a fuzzface.

my advice, if ya follow me down this rabbit hole, is to do the biasing with knobs full up, and guitar rolled down to where you want it to sound "clean" or "cleanish" ...say, 2-3 on the guitar knob. bias to get max volume and best tone, and ya can't go wrong when ya crank the guitar up.

again, its not as saturate as a fuzzface.  more like an improbable cross between a boogie, fender, and marshall kinda tube saturation.  all the way up, is a nice compressy brown tube preamp sound. start starving it with the voice, and the distortion goes up. its kinda cool. the fuzz control and voice interact quite a bit, letting ya get a lot of different fuzz tones.

Yeah, I'm psyched to get going on this. Just gotta clear some stuff off the bench. (You know how that goes.)

[PRR]

> something to do with the Getter element

Tantalum and a few other Transmitter Tube getters must be run hot to absorb gas.

Receiving Tube getters, the micro-thin silvery flash inside the glass, do not have to run hot. However if they were assembled in air they would be used-up before they got to the vacuum pump. They are made in a lump so most of the getter-stuff does not get contaminated. Yes they commonly use a burst of RF to vaporize the lump into a flash. Making the getter-stuff holder ring-shape helps the RF heat.

[pinkjimiphoton]

thanks, paul!

[pinkjimiphoton]

ok, after @#$%ing with this thing for wayyyyyyyyyyyyyyyyyyy too long, and playing thru it for mark at mla last nite, we decided its simply not worth @#$%ing with. too much of a pain in the ass to deal with.
in true fuzz face fashion, literally EVERYTIME ya plug it in, it sounds DIFFERENT. @#$%ING MADDENING!!

input audio? changes the biasing. move a knob? change the bias. unlike transistors, ya gotta wait for the bias to settle. again. its insanely interactive, just like the real deal. but to get a decent fuzztone out of it is crazy hard. and its unreliable. and it sounds different everytime. and its @#$%in' impossible to bias.

but it was still kinda fun to take on, and may well be a worthy "circuit snippet" to add to a fuzz or dirt pedal project, but on its own its just not really worth it imho. your mileage may vary.

anyways, since we're NOT gonna pursue it, i can release it to you guys. too @#$%in lazy to scan the schematic at this time, but i can tell ya what i did.

go to tagboardeffects and check IvIark's layout for a charge pump power supply. ya wanna shoot for around 30v output. i used an lt1054 and 5817's to ultimately get about 26v to the plates of the 12ax7

run the heaters on straight 9v. pin 4 to ground, pin 5 to +

input cap bump up to 22uF. you MAY wanna add a hi pass filter inline with the input to roll off some bass. it NEEDS the big cap to get enough bass to fuzz that first stage. you can use other caps, obviously.

on the other side of the input cap, the circuit side, make a diode clipper to ground off the circuit side of the input cap... this adds a little compression and smooths the circuit out a little bit. i used a white 5mm led, band to ground on one side, and a red 5mm and 1n270ge in series for the other side, bands pointing to the signal.

v1 plate load resistor i used an a500k 9mm pot as a trimmer, you're gonna wanna shoot for about 15v on the plate.

v2 plate load resistor, i used an a1m 9mm pot as a trimmer <the actual unit i just used little blue n white trimmers like they sell at tayda...>  shoot for about 21v on this plate.

decoupling resistor between the plate loads goes down from 470r to 47r.

feedback resistor between grid of v1 and cathode of v2 wants to see roughly 1.5meg.  i used a 470k fixed resistor off the v1 grid, and an a1m pot to vary the resistance. i call this a "voice" control... all the way clockwise, ya gotta fairly standard fuzzfaceish sound... as ya rotate it left, it will get more distorted, jagged, and gated. in fact, it makes a fairly decent noise gate at some settings. go figure.

for the fuzz control, a1m again, cathode bypass cap taken from 22u to 220u.

v1 cathode directly to ground.

output cap moved from plate/plate load resistance to node of 47r and plate load resistor of v2. this gets ya enough output to be able to use the circuit... hooked up normal, just plain not enough output. this MAY be a little noisy. a smallish cap, say 1000p-3300p to ground should eradicate it. i didn't bother.

output cap, i went ultimately 3.3uF. output pot, i went a1m again.

for power supply, i used 2200uF electro with standard 1n400x protection, and a 3.3n snubber cap to nuke noise. ya need the big cap so ya have enough reserve to stop the tube from sagging too much under brutal abuse. brutally abusing the input can rebias it to the point it goes completely unlinear and freaks out. then it will gate real bad and kinda swell back in. neat. not super useful, but neat, and definitely exploitable.

you can get a lot of standard fuzz face weirdness with this. everything but the full-on fuzz.

if ya dare follow down the rabbit hole and @#$% around with this piece of shit, let me know how ya make out!!

at some settings, ya can get a decent octave up fuzz, too. in fact, the thing it reminds me most of is an anemic version of gus smalley's OUSB or tim escobedo's jawari.

anyways... i DO have a vero and schematic for it, if ya REALLY need it, let me know and i'll post it... but odds are its unnecessary if ya read this here post.

so that's it. no more cloak and dagger. i may still use it as a snippet in another circuit <it sounds @#$%ing great being driven by, or driving, a wide assortment of dirt pedals> but standalone, its just not good enough to @#$% with anymore imho.

hope this helps some future daredevil out.

love to all... wash your hands, please... despite the @#$%in morons claiming this is a liberal hoax going on, yeah, @#$% that, my @#$%ing BROTHER is currently in an icu with it, and is likely gonna @#$%ing die.

[Scruffie]

Have you considered a hybrid face? A tube in the 1st spot seems to work just about in simulation with some fiddling.

[pinkjimiphoton]

too lazy

[amptramp]

If you had been able to use p-channel tubes, it might have worked like a PNP germanium Fuzz Face and been the new darling of the industry.  I could picture one using a 6D4 gas triode (thyratron or noise generator) and offering a great new sound.  Somebody should do this.  Somebody who is not me.

[bluebunny]

love to all... wash your hands, please... despite the @#$%in morons claiming this is a liberal hoax going on, yeah, @#$% that, my @#$%ing BROTHER is currently in an icu with it, and is likely gonna @#$%ing die.

Love right back at ya, Pink One, and your bro.