Vox Repeater 2N2646 replacement? (kinda lost)

[richon]

Hi,

a friend of mine want's the old VOX REPEATER circuit in his guitar (alla teardrop Vox guitars)...  and it seems that the 2N2646 can't be replaced very easy with a PNP or NPN transistor

coudl someone explain me what type of transistor is, and how could i replace it?

[doug deeper]

small bear has them.
they are under the diode category.
they are unijunction transistors.

[R.G.]

The 2N2646 can be replaced with the much more available 2N6027 PUT (Programmable Unijunction Transistor). The circuit is a little different, but has the same number of resistors and capacitors. You're undoubtedly doing this to make a Vox Repeat Percussion; I've done these with the PUT in both the pedal form and the version that's in the old Thomas Vox Beatle amp. They work.

[R.G.]

Upon thinking about it, I realized I needed to provide more info.

The UniJunction Transistor has two emitters, E1 and E2, and a base, B. The base B is connected to the junction of a resistor pullup to the power supply and a cap go ground. The emitters have resistors in series with them. How it works is that the base junction is off, reverse biased, until the base voltage ramps up above the voltage in the middle of the channel between the two emitters. The resistances between the base and E1 and the base and E2 and the resistors in series with the emitters make the voltage where the base will start conducting be a voltage-divider fraction of the power supply. When the base gets high enough, it conducts, and this current causes the lower emitter (E2) resistance to drop, so the base conducts until it almost empties the capacitor. When it quits conducting, the channel resistance comes back up, the resistance to E2 is reestablished, and the cap starts charging again.

The PUT does the same thing, but in a different way. It is a latching thryristor like an SCR, but it starts conducting when its gate is more negative than the anode. So to make this act like a UJT, you tie the base to a fixed voltage divider to set the trigger voltage. The anode is tied to a resistor/capacitor like the base of the UJT, and the anode to ground, or nearly so. The anode ramps up as the capacitor charges until it gets high enough to conduct into the gate voltage. The gate current turns on the thyristor/SCR action, and the cap is discharged through the anode/cathode until the current drops below the holding current, which is quite small.

So if you connect the anode of a PUT where the base of a UJT was, the cathode of the PUT where the E2 used to be, and connect the gate of a PUT to resistor divider to ground, the resulting circuit acts very much like the UJT used to. The gate is not the exact equal of the UJT's E1, so the circuit is different there.

[ch0p_ch0p]

Upon thinking about it, I realized I needed to provide more info.

The UniJunction Transistor has two emitters, E1 and E2, and a base, B. The base B is connected to the junction of a resistor pullup to the power supply and a cap go ground. The emitters have resistors in series with them. How it works is that the base junction is off, reverse biased, until the base voltage ramps up above the voltage in the middle of the channel between the two emitters. The resistances between the base and E1 and the base and E2 and the resistors in series with the emitters make the voltage where the base will start conducting be a voltage-divider fraction of the power supply. When the base gets high enough, it conducts, and this current causes the lower emitter (E2) resistance to drop, so the base conducts until it almost empties the capacitor. When it quits conducting, the channel resistance comes back up, the resistance to E2 is reestablished, and the cap starts charging again.

The PUT does the same thing, but in a different way. It is a latching thryristor like an SCR, but it starts conducting when its gate is more negative than the anode. So to make this act like a UJT, you tie the base to a fixed voltage divider to set the trigger voltage. The anode is tied to a resistor/capacitor like the base of the UJT, and the anode to ground, or nearly so. The anode ramps up as the capacitor charges until it gets high enough to conduct into the gate voltage. The gate current turns on the thyristor/SCR action, and the cap is discharged through the anode/cathode until the current drops below the holding current, which is quite small.

So if you connect the anode of a PUT where the base of a UJT was, the cathode of the PUT where the E2 used to be, and connect the gate of a PUT to resistor divider to ground, the resulting circuit acts very much like the UJT used to. The gate is not the exact equal of the UJT's E1, so the circuit is different there.

Hello, Im really sorry to bother you by digging up some 11 year old thread and making possibly dumb questions, but just now I saw this and I didnt understand much, so I have some questions if that´s ok . In the beggining you said that the UJT has 2 emmiters and 1 base, but all I´m seeing online, and more specifically in this scheamtic (http://effectslayouts.blogspot.com/2015/10/earthquaker-devices-hummingbird.html) points to it having 2 bases and only one emmiter. You also said the PUT´s anode is tied to a resistor/capacitor, and then again that the anode is tied to ground. Which is right? Im really confused about the UJT pinout thing because you said we should connect the PUT´s anode to where the UJT´s base was, and etc., but if it turns out the UJT actually has 2 bases, how does the wiring turn out then? Do I assume that E1 and E2 are just B1 and B2 in reality?

You also said to connect the base of the PUT to a fixed voltage divider to set the trigger voltage. Is "base" the same as "gate" for the PUT? This brings us to the next point, which is where I really got confused. If our voltage divider will be sending in the current to the gate, and the voltage after the voltage divider being the tirgger voltage, why does the gate of the PUT go "to resistor divider to ground"? shouldnt it be receiving the current that passes through the voltage divider to teh gate, making there be less voltage at the gate than at the anode, so it starts conducting?

Im also really lost about how to get this voltage divider to work. Should I just add it to the circuit and let the rest inatct, or should I (Im really making guesses here) replace it for R12 in the schamtic I linked above/R10 in the Vox repeat Percussion schematic?

Also, sorry if Im pushing it here, but what results am I expecting with the voltage divider? should the voltage at the gate just become any number less than the voltage at the anode, or do I need a specific number? I have no multimeter yet and my breadboard is not with, so I dont really have a way right now of checking this out (even if I knew how lol)... so I´d  really appreciate any info you could help me with!!! ))

Thanks in advance and again sorry to be bothering you with these kinds of questions so many years after this post

[Rob Strand]

I'll let RG fend off the question but if you can find a copy of the GE Transistor Handbook Manual it does a good job of explaining UJT's.   There's different editions of the handbook and I'm not sure which edition the UJTs first appeared. [Seems like Ed 3 or later has UJT's]

FYI, from an electronics industry perspective the 2N2646 was largely replaced by the NE555 timer.  Not a plug-in replacement but more a functional replacement.

[iainpunk]

i have wanted to build a tremolo like the repeat percussion for a while now, but with my own weird and wacky design features, does anyone know the amplitude of the oscillator?

FYI, from an electronics industry perspective the 2N2646 was largely replaced by the NE555 timer.  Not plug-in replacement but more a functional replacement.

is that a chalenge to build an UglyFace based on the 2N2646 ?

cheers, Iain

[Rob Strand]

i have wanted to build a tremolo like the repeat percussion for a while now, but with my own weird and wacky design features, does anyone know the amplitude of the oscillator?

The emitter swings from VE_min to VE_max.

VE_min = (1/2) VEsat
VE_max = n VBB + VD

where  device parameters are

VEsat = 3.5V, n = 0.65, RBB = 7k ohm, VD = 0.7V

VBB is more or less the supply voltage but we might do a bit better.   On the RP ckt there is 1k + 470 to the 9V supply, that forms a divider with RBB so you end-up with VBB = 9 * 7 / (7 + 1 + 0.47) = 0.826 * 9 = 7.43V.

That gives a humble estimate for the emitter voltage of,

VE_min = 3.5 / 2 = 1.75V
VE_max = 0.65*7.43 + 0.7 = 5.53V

I suspect the 82K + 330k divider on the base of the transistor is about getting the base voltage below 0.55 to 0.6V when the UJT emitter is at  1.75V.   1.75 * 82/(82 + 330) = 0.2 * 1.75 = 0.35V.

It's worth someone else having a stab at an estimate.

Typically the pulses on UJTs are quite narrow.

is that a chalenge to build an UglyFace based on the 2N2646 ?

The replacement path is often only one way but in the bad old days you will find people doing innovative stuff with the parts they had.

[ch0p_ch0p]

I'll let RG fend off the question but if you can find a copy of the GE Transistor Handbook Manual it does a good job of explaining UJT's.   There's different editions of the handbook and I'm not sure which edition the UJTs first appeared. [Seems like Ed 3 or later has UJT's]

FYI, from an electronics industry perspective the 2N2646 was largely replaced by the NE555 timer.  Not a plug-in replacement but more a functional replacement.

Thank you so much!!!!! I´ll definitley look for that manual, really lookinf forward to it!! Also thank you for the heads up about the NE555, I´m not too good with the soldering iron so I´ve been trying to stay away from ICs of any kind (also beacause I havent done any research on them so I don´t really know what I´m dealing with lol) but it´s definitely something to think about.

Thanks again!!!

[iainpunk]

so the saw teeth ate 3.8v pkpk, thank you so much for that!!!
i think ill implement a triangle oscillator that has 2 controls, rise time and fall time. maybe even a gain and bias control so i can clip the wave from and dial in the asymmetry of that clipping

cheers, Iain

[PRR]

....if you can find a copy of the GE Transistor Manual it does a good job of explaining UJT's.

http://www.introni.it/pdf/GE%20-%20Transistor%20Manual%201964.pdf   13MB PDF
Page 300

This is pure original UJT. The PUT came later after G.E. quit being useful.

[R.G.]

The now-obsolete-too TIS43 in a TO92 package can be substituted for the 2N2646 pin for pin. It may be more available. There is at least one commercial pedal maker using the TIS43.

The 2N6027 PUT works, but it does require some wrangling to set the gate threshold right. Somewhere in the archives I have example circuits for how to apply the 6027 to the Vox repeater circuit. Not too tough - adjusted resistor values and one additional resistor.

All PUT/UJT circuits rely on the slow discharge of a capacitor, then the sudden dumping of that cap through the device. The capacitor voltage is an quasi-sawtooth (that is, a resistor-capacitor charging waveform) and a sudden, sharp return to nearly zero when the control electrode trips.

[Rob Strand]

The now-obsolete-too TIS43 in a TO92 package can be substituted for the 2N2646 pin for pin. It may be more available. There is at least one commercial pedal maker using the TIS43.

I pulled quite a few TIS43's out of industrial equipment as a kid.  I can vouch that they are pretty much plug-in replacements.

In passing I've seen some criminal prices for the 2N2646's.    The TIS43's plastic TO92 package might fend off some vintage dollars.

[iainpunk]

plastic TO92 package might fend off some vintage dollars.

i have noticed that! just before covid, i was designing someone a custom fuzz/distortion, and we were swapping a BJT and its collector resistor in and out of sockets to try some flavors, and he kept liking metal can transistors better than plastics, despite both being all over the map sound wise. he even liked the metal can 2n2222 's better than the plastic 2n2222 's despite them sounding nearly identical to me... people seem to listen with their eyes

cheers

[ch0p_ch0p]

....if you can find a copy of the GE Transistor Manual it does a good job of explaining UJT's.

http://www.introni.it/pdf/GE%20-%20Transistor%20Manual%201964.pdf   13MB PDF
Page 300

This is pure original UJT. The PUT came later after G.E. quit being useful.

Thank you!!!!! I was actually reading the 5th edition so this probably has some more stuff.

The now-obsolete-too TIS43 in a TO92 package can be substituted for the 2N2646 pin for pin. It may be more available. There is at least one commercial pedal maker using the TIS43.

The 2N6027 PUT works, but it does require some wrangling to set the gate threshold right. Somewhere in the archives I have example circuits for how to apply the 6027 to the Vox repeater circuit. Not too tough - adjusted resistor values and one additional resistor.

All PUT/UJT circuits rely on the slow discharge of a capacitor, then the sudden dumping of that cap through the device. The capacitor voltage is an quasi-sawtooth (that is, a resistor-capacitor charging waveform) and a sudden, sharp return to nearly zero when the control electrode trips.

Thanks so much for answering!!! Saddly TIS43 doesnt seem availiable around me either ://

Could you point me in the direction of said archives please?? I´d be very interesting in reading them ahah, epsecially since my knowledge is yet too small to fully understand what Im reading and how it would apply to making a PUT behave like a UJT. Your remark is really helpful though, I can now see it works like an oscillator of sorts, where the voltage is constantly rising and dropping, as the capacitor slowly fills up and when the voltage at the emitter becomes high enough for it to open, it discharges everything very quickly (is this right?)! If these archives arent something that you published and you just have it stored and cant find it then dont worry, I´ll just try to make do with the help up to this point. Thank you bery much!!!

[PRR]

> said archives please??

R.G. has much material at  www.geofex.com  Some of the linkage is broken. It will keep you busy a while.

[ch0p_ch0p]

> said archives please??

R.G. has much material at  www.geofex.com  Some of the linkage is broken. It will keep you busy a while.

Thank you!!! Will definitely take a look


While I was thinking about this I did try to whip up something on paper, and just maybe I might have kinda got it right:



Can someone please check if this even makes any sense?? Seems kinda too simple to be right lol. R.G. did mention the resistor values would change and I have no idea what would be the calcaulations for the new one, so I would appreciate any help with that as well, if there´s some formula or something. (This was copied form this schematic - http://effectslayouts.blogspot.com/search?q=tremolo , it´s the bottom left part).

Thanks!!

[Rob Strand]

For the VOX RP, you might be able to build a circuit using normal transistors,



The min VE voltage seems a bit low.  I'd have to look into changing that.

Surely someone has already done this in the last 50 years?

[Rob Strand]

Can someone please check if this even makes any sense?? Seems kinda too simple to be right lol. R.G. did mention the resistor values would change and I have no idea what would be the calcaulations for the new one, so I would appreciate any help with that as well, if there´s some formula or something. (This was copied form this schematic - http://effectslayouts.blogspot.com/search?q=tremolo , it´s the bottom left part).

The gate is used as the sense and conduction goes though A to K.

You can see a pic halfway down this page,

https://www.technocrazed.com/7-8-the-unijunction-transistor-ujt

The external resistors R1 and R2 are what programs the trip point.    The resistors can be chosen by the user, so that's why the PUT is called programmable.   For the UJT the trip point is built into the device and is specified through the parameter 'n', which is normally about 0.65.   So to match the UJT you would choose R1+R2 = 7k ohm and you want R1 = 0.35*7k and R2 = 0.65 & 7k.   More or less, as there's always a heap of evil finer points in reality.

[ch0p_ch0p]

The gate is used as the sense and conduction goes though A to K.

You can see a pic halfway down this page,

https://www.technocrazed.com/7-8-the-unijunction-transistor-ujt

The external resistors R1 and R2 are what program the trip point.    The resistors can be chosen by the user so that's why the PUT is call programmable.   For the UJT the trip point is built into the device and is specified through the parameter 'n', which is normally about 0.65.   So to match the UJT you would choose R1+R2 = 7k ohm and you want R1 = 0.35*7k and R2 = 0.65 & 7k.   More or less, as there's always a heap of evil finer points in reality.

THANK YOU SO MUCH!!!! This completely clears all my doubts!! Damn, I completely butchered that schematic ahah, I had just assumed from before that the gate would replace the emmiter in the UJT and went from there, so the whole thing is backwards.

This will make it all a lot easier, thanks!! That other schematic you showed with the bipolar transistors is also useful, something to think about, I see you replaced the PUT with the NPN and PNP which are basically what's inside of the PUT. This makes it very clear for me!!