i was dickin around the other day trying to wire up a switch to choose between two transistors.
i'd read that ya tie the e's and c's together, and switch the bases.
i'd also read ya tie the bases together, and switch the e's and c's.
so i tried both. switching e's and c's with a common base connection definitely seemed to work better. each transistor pretty much sounded like itself by itself..
but...
tieing e's and c's together and switching the base was a much different kind of sound, and surprisingly, neither transistor sounded the same as they did on their own.
so i got to thinking about it, and realized for all intents what i had was something like this:
(https://i.postimg.cc/8J0ZcrFq/wtfizzit.png) (https://postimg.cc/8J0ZcrFq)
it seems to have a bit more balls and a nicer tone than either of the transistors alone..
both q's are germanium, too, i should have made that clear from the beginning. i dunno how this would work in anything else, but it's interesting in a fuzz circuit.
what the hell izzit, what did i do, and why would or how can it work? i'd think it was just two parallel transistors, but there's only connection to one of the b's?
i'm almost tempted to install both transistors wired like this and be done with it.
Looks kinda like a long tailed pair...sorta...
It is kinda like a LTP (or as some would vaguely call Super Symmetry) if you're only switching bases I would say perhaps it could be perhaps if it's left floating after switching the electromagnetic field around it is being conducted as a NFB :icon_lol:
I'm making wild guesses but a half floating pseudo LTP like that makes for interesting characteristics
Sort of close to what ampeg did here near the end.
(http://fuzzcentral.ssguitar.com/scrambler/scramblerschematic.gif)
I don't know any academic term, although that don't mean there isn't one. "Half-floating LTP" does seem to fit.
Probably is a Ge thing - because of leakage is my expectation and that leakage x the beta.
You maybe have a kind of voltage dependent resistor across the driven device? Getting a fast real-time compression effect.
WOW, so i reinvented the ltp? like the driver part of a tube amp kinda LTP?
i WAS switching bases, but found one side had a way better fuzztone than the other.
also messed with combining different substrates, like, npn and pnp both forward and reverse biased. found some really interesting tones,
but the "best" one seemed to be two ge's, one so leaky it reads on my meter as two diodes with a center tap.
the combo of two transistors sounded best when just using one base connection. if ya pull the other q, the tone changes substantially, but it still passes signal.
in this case i was using a 2n1302 npn ge and an ancient 073 ge.. thats the one that shows up as two diodes.
the 2n1302 has a decent fuzztone, tho a mite gated. add the 073 and suddenly it screams.
the 073 by itself is more like a warm, brown overdrive by itself. but about half the output of the 1302.
1302 is reading 36hFE @ 226mv on my crappy chinese tester, the 073 reads ike...
1-----------I<---------------2--------------->I-----------3 on my meter, and shows
1 reading 1.1v
and 3 reading 419 mv
rob strand said it looks like a darlington maybe, i dunno
it probably is something to do with ge and leakage i'm imagining.
with the switch, if both q's are properly installed, you get a kinda overdriven tone, and a fuzzier one, roughly the same volume.
but if ya invert the 073, it gets weird.. ya still get the same kinda overdrivey brown tone, but when ya flick the switch, you get a boost about 3x as loud with a decent sounding fuzz.
if you invert the 1302, it won't work, but if ya don't, it gets a rather rude gated trebly fuzztone thats pretty cool.
by itself, the 073 sounds warm and brown with not much output... i'm wondering if maybe the signal going to be is just leaking thru the diode junction between c and c of the transistor stage?
by itself, the 1302 makes a fairly loud, classic sounding fuzztone.
together tho, its weird... i like real versatile fuzzes that clean up well and interact with the guitar, but this sseems to wanna make that thinnish brittle 60's sounding fuzz more.
thanks for the ideas guys and the explanations. i'm still kinda stumped.
In every darlington or quasi complementary setup you have the emitter feeding into the bases of the lower transistor so yea um, that's not darlington by any definition
But it if works fantastic by accident!
LTP is used in transistor setups as well. Very common in hifi speaker amps.
Quote1-----------I<---------------2--------------->I-----------3 on my meter, and shows
1 reading 1.1v
and 3 reading 419 mv
rob strand said it looks like a darlington maybe, i dunno
I was treating it as a black box (with unknown internal connections) which gave those voltages.
Since you *know* the connection that's a completely different story!
Something like that with an unconnected base won't be a "named" configuration.
I suspect the key is leakage. The leakage is making the transistor with the unconnected base present a funny load to the first transistor. The change in sound could be a very simple explanation: it changes the DC bias point on the collector and that's it! More complicated effects would be it presents a non-linear load, or, the second transistor's capacitances creates a load which is non-linear and frequency dependent.
You could test the idea:
1) measure the collector voltage with one transistor present.
2) add the second transistor, re-measure the collector voltage. Has it changed?
If it changed then you could try adding a resistor across C and E of the first transistor. Trim the resistor so the collector voltage is the same as in (2). Does it sound the same/similar?
(Oh, if you make that connection with silicon devices I've got doubts it will do anything.)
I just tried this with *silicon* transistors.
Start off with a CE amp. Something like this,
http://www.brianmay.co.kr/image/ScreamingBirdTB.jpg
Experiments:
1) Added the second transistor. Result = No significant effect.
2) Added second transistor except second transistor has a 1MEG resistor between C and B.
That dropped the collector voltage from about 4.2V to 2.5V
3) Removed second transistor and replaced it with a resistor across C and E of the first transistor.
Tweaked value to get a 2.5V collector voltage.
Result: pretty much same gain and same non-linearity of the output waveform underclipping.
So *looking* at the waveforms there is little difference. There maybe some subtle differences when listening however compared to the original circuit I suspect case (3) is a good approximation to (2).
i gotta re-read all this when i'm awake bro lol
but i see what you mean, i think that may well be it. maybe the leakage thru the super-shitty transistor is just acting like a diode so it changes it from a fuzz stage to an almost bazz fuss stage. i am gonna mess with it.
but the key is having ONE of the ge's, the weird one, reverse beta'd... the way so that by itself, its higher gain and more distorted and the other way more muted and quieter. then when ya kick the other q on, it changes pretty drastically.
if i can tomorrow, i will try and get a little video. things are kinda crazy still.
thinking about it for a minute, i think THIS is actually more like what i should have posted earlier... for the loudest, fuzziest tone
(https://i.postimg.cc/F1c7kR18/wtfizzit-2.png) (https://postimg.cc/F1c7kR18)
maybe this way makes more sense... maybe it creates a feedback loop of some kind? if replacing a single transistor, the upper common would go to + and output, the lower common - and ground(ish) like a single transistor.
i'm gonna have to play with this a little bit and try and get some voltages so we can see what's going on.
like everything in my life, its all weird ;)
ever feel like you're caught in a dystopian nightmare? ;)
sorry .... felt we needed an alice cooper moment ;)
that weird "transistor" i tried using... think it may be a shotke diode of some sort?
(https://i.postimg.cc/w1H15dn2/Screenshot-2019-02-21-Sourcingmap-30-A-100-V-3-Terminal-Schottky.png) (https://postimg.cc/w1H15dn2)
it looks like the opposite of this, what i ascii'd earlier
1-----I<-----2------>I------3
the markings on the picture are identical to the black things on the tremolo board in that ovation preamp me and rob have been working on... but its two of them. the markings, i mean
did i mention life is weird lately?
Haha Jimi.
You cant just mate them and expect baby jelly beans pal.
Although it would be nice.
Quote from: pinkjimiphoton on February 21, 2019, 12:36:49 AM
that weird "transistor" i tried using... think it may be a shotke diode of some sort?
(https://i.postimg.cc/w1H15dn2/Screenshot-2019-02-21-Sourcingmap-30-A-100-V-3-Terminal-Schottky.png) (https://postimg.cc/w1H15dn2)
it looks like the opposite of this, what i ascii'd earlier
1-----I<-----2------>I------3
the markings on the picture are identical to the black things on the tremolo board in that ovation preamp me and rob have been working on... but its two of them. the markings, i mean
did i mention life is weird lately?
It's definitely a dual diode but it isn't always a schkotty diode. MBR series are definitely schkotty diodes however
<wtf did i do, and wtf would ya call it?>
Both with the same adjective but are you sure you want to know it..?? :icon_redface:
P.S.
I vote you for declaration of the forum most risky person for other members brain cells damage..!! :icon_lol:
Quotethat weird "transistor" i tried using... think it may be a shotke diode of some sort?
Well the Schottky would only have an effect if it was very leaky.
Quote from: antonis on February 21, 2019, 06:10:26 AM
<wtf did i do, and wtf would ya call it?>
Both with the same adjective but are you sure you want to know it..?? :icon_redface:
P.S.
I vote you for declaration of the forum most risky person for other members brain cells damage..!! :icon_lol:
hahah yeah, i'm like, the mutant bastidge offspring of keef richard and david crosby on this forum, i think....lol
if you're ever here, bro, look me up, i'll show ya how we swamp yankees like to party ;)
(https://i.postimg.cc/fWqXFSsz/wtfizzit-2.png)
It's close to Arsenio Novo tube sound overdrive, a kind of Bazz Fuss. See my gallery and posts about it.
The original schematic was in ascii :)
The first setup, 2 NPN, what about using a Silicon and a leaky germ?
mac
Not the same hookup, but related-enough that it deserves a link here: Piggybacking transistors (https://www.diystompboxes.com/smfforum/index.php?topic=21801.msg136280#msg136280).
Quote from: mac on February 21, 2019, 08:11:54 PM
(https://i.postimg.cc/fWqXFSsz/wtfizzit-2.png)
It's close to Arsenio Novo tube sound overdrive, a kind of Bazz Fuss. See my gallery and posts about it.
The original schematic was in ascii :)
The first setup, 2 NPN, what about using a Silicon and a leaky germ?
mac
hi mac,
i DID build that circuit before, years ago. it was nice sounding, i forget what i called it but i think i still have the verified vero for it.
yeah, si and leaky germ may work. i gotta get a chance to play with it, so many projects all at once lol...
but i DID try it with si too... in the circuit i'm messing with <that ovation fuzz> anything over about 60hFE sounds like poo, and the lowest SI i have is maybe 90ish so hard to tell. i may have to breadboard and mess with it more.
it DOES sound like a cranked up marshall kinda distortion more than fuzz, so maybe it's closer to arsenio's circuit?
all will be revealed eventually ;)
QuoteIt's close to Arsenio Novo tube sound overdrive, a kind of Bazz Fuss. See my gallery and posts about it.
The original schematic was in ascii :)
IMHO, not connecting the base changes things a lot. But an even bigger difference is the 2x NPN vs NPN+PNP.
For the unconnected base the things I see having an effect are:
- leakage of the right side transistor
- breakdown of the BE junction of the right side transistor.
(sorry I left off increased capacitance)
Both of these will show-up when comparing the collector voltage bias point to the ckt without the right-side transistor.
QuoteIMHO, not connecting the base changes things a lot. But an even bigger difference is the 2x NPN vs NPN+PNP.
Novo's circuit has a trimmer to set the distortion DC sweet spot:
Vc - Vb -> 0, Ib(pnp) -> 0
This is exactly what happens in a well tuned Bazz Fuss.
I have to see my notes, I remember to have breadboarded a single NPN with a diode or zener from C to E, maybe some bias pot.
In Arsenio Novo's design or Bazz Fuss, the diode restricts the motion of the collector to one diode drop above the base.
In Jimi's wtf, IF the transistor at the right is acting like a diode, the collector and the base of the one at the left must be equal because it is on and the base to emitter voltage is a diode drop.
Wtf, just guessing.
mac
QuoteNovo's circuit has a trimmer to set the distortion DC sweet spot:
Vc - Vb -> 0, Ib(pnp) -> 0
This is exactly what happens in a well tuned Bazz Fuss.
I have to see my notes, I remember to have breadboarded a single NPN with a diode or zener from C to E, maybe some bias pot.
In Arsenio Novo's design or Bazz Fuss, the diode restricts the motion of the collector to one diode drop above the base.
I can see that could work like you say.
With the base unconnected it changes things a lot. A *silicon* transistor with the base open, look like two back to back diodes so it doesn't have any obvious function.
E ---|<--- B --->|--- C
If the BE voltage increases to 8 to 10V then you might break-down the BE junction, that will do something.
If the transistor is high-leakage like a GE then it *will * pass current from C to E, that will also do something.
I guess my beef with all the "similar" circuits in the thread is they don't cover the open base case.
its gotta be leakage. or what i thought is a transistor, is, in fact, two diodes.
lotta wtf to go around lately. man!
i gotta sit down and hook it up again and measure voltages, you guys are way above my humble paygrade.
monkey with a breadboard. making shit fuzzy is easy. the rest... not so much.
but making a GOOD fuzz is a challenge.
cuz they all are good for different things ;)
anyways, i'm thinking leakage thru the other transistor.
maybe i'll do it "live" on youtube ;)
Might be possible to work with silicon IF the floater is low hfe in the first place - like a medium power type such as 2N3053 and also used in reverse beta. Because it won't leak on like a Ge, it won't actually work floating base, but would need a resistor b-c, maybe large like 100k.
I don't know if it actually will work, but it would be nice if it did since we don't have much use for those TO-39 types these days and there's bags of them laying around.
Quote from: anotherjim on February 23, 2019, 11:00:27 AM
Might be possible to work with silicon IF the floater is low hfe in the first place - like a medium power type such as 2N3053 and also used in reverse beta. Because it won't leak on like a Ge, it won't actually work floating base, but would need a resistor b-c, maybe large like 100k.
I don't know if it actually will work, but it would be nice if it did since we don't have much use for those TO-39 types these days and there's bags of them laying around.
2N3053 is actually pretty pricey compared to your average transistor! Perhaps use BD139s as they are pretty low gain :icon_razz: Or TIP41... I can get 5 of them for like 60 cents
WTF, I better breadboard this thing :)
Right now I have a BC550C, input and output caps 100n, 10k collector resistor, 1k emitter resistor with 22uf bypass cap.
Bias -> 1M resistor from base to lug 2 of a 50kB pot, lug 3 to vcc, lug 1 to gnd.
- If I place a npn silicon transistor with open base it does nothing, as expected.
Germs are a different story. Some 2N1114 and 2N388 work, but really leaky AC187 can't be biased except by adding a leakage shunt resistor from base to gnd, say 100k or so.
- Silicons can be forced to work too with a similar bias circuit, a 1M and 50kb pot to feed the base, no guitar signal. I am using a MPSA42 and another BC550C.
- If I remove the second Si or Ge, and I put a 47k to 10k resistor from C to E I get almost the same result (small SS amp, late night :) )
So maybe the second transistor is just a resistor as said before.
- Diodes from C to E are a something different, more synth like sounds.
mac
What you first posted wasn't so different from: (http://www.muzique.com/schem/diffuzz.gif)
But this new thing, that's ... something ... else.
QuoteWTF, I better breadboard this thing
You get the gold star!
Quote from: mac on February 24, 2019, 02:05:55 PM
WTF, I better breadboard this thing :)
Right now I have a BC550C, input and output caps 100n, 10k collector resistor, 1k emitter resistor with 22uf bypass cap.
Bias -> 1M resistor from base to lug 2 of a 50kB pot, lug 3 to vcc, lug 1 to gnd.
- If I place a npn silicon transistor with open base it does nothing, as expected.
Germs are a different story. Some 2N1114 and 2N388 work, but really leaky AC187 can't be biased except by adding a leakage shunt resistor from base to gnd, say 100k or so.
- Silicons can be forced to work too with a similar bias circuit, a 1M and 50kb pot to feed the base, no guitar signal. I am using a MPSA42 and another BC550C.
- If I remove the second Si or Ge, and I put a 47k to 10k resistor from C to E I get almost the same result (small SS amp, late night :) )
So maybe the second transistor is just a resistor as said before.
- Diodes from C to E are a something different, more synth like sounds.
mac
try the leaky ge reverse beta and watch what happens. it SHOULD make the other transistor's gain go way up
definitely louder and fuzzier with the second transistor backwards. again, not all will work. but when it does... its like... holy shit.
g will be back to work this week, so i should be able to video some experiments and the original circuit in question so you guys can see and hear.. and i'll do voltage measurements live so you guys that are wayyyyyyyyyyyyyy smarter than me can hopefully figure it out!
;)