Programmable Unijunction Transistor PUT as clipping device

[Vivek]

Anyone use a PUT as a clipping device

And implemented adjustable clipping radians/symmetrical/asymmetrical/sag/wave wrapping by altering the gate voltage ?

[antonis]

Peak point Emitter current might be within acceptable levels but Valey point one might load op-amp output..

[Vivek]

Please help me understand how a PUT in feedback loop could tax the Opamp ?

[antonis]

Consider V_BB above as op-amp output <-> inverting input..

Better use PUT as relaxation oscillator (sawtooth) triggered from op-amp output..

[iainpunk]

i have experimented with a saw-tooth generator where the generator was activated when the signal became above a certain DC point, and the bias could be shifted to determine the sensitivity. it gave a cool effect because the voltage on a certain resistor determines the oscillation frequency, and if you feed it a sine wave or audio signal, its frequency changes during the cycle and goes totally bezerk. biasing it just right or wrong gave me a wide array of oscillating noise chaos. if biased too low, it was just a fuzz, if biased too high, it was oscillating constantly and playing just modulated the frequency. i don't have the schematic anymore, but ill draw up a rough approximation in falstad, to be posted later today.

i have also had some experience with using triacs to clip, it gives a totally nasty tone and too high input made it almost quiet.

https://en.wikipedia.org/wiki/Thyristor
https://en.wikipedia.org/wiki/TRIAC

in this example its used for gating instead of clipping, so the trailing edge is left in tact, i used it as clipping and the leading edge was left intact. also, i used a ""DIY"" low voltage triac so my experiment wasn't full wave rectified like the image.

now i am inspired to use this for wavefolding instead of clipping, total chaos will ensue

cheers, Iain

[duck_arse]

i have experimented with a saw-tooth generator where the generator ......  ill draw up a rough approximation in falstad, to be posted later today.

total chaos will ensue

cheers, Iain

I'm interested in this chaos you mention. perhaps start a new thread?

[iainpunk]

that total chaos was with the triac wave folding idea, im not totally figuted out yet how to do it since im still working on my regular wavefolder, but the sawtooth oscilator thing was less chaotic and more chirpy and weird: here it a proof of concept for the sawtooth clipping:


the bottom opamp is a simple schmitt trigger sawtooth oscilator driven by the input signal

cheers, Iain

ps. the schmitt trigger is also a negative resistance, but its opamp controlled, so its not off topic

edit: output voltage wave forms. with and without distortion added

[Vivek]

What I was trying to do was


A) have user settable DC voltages to the Gate, so that it clips at different levels, emulating different diodes ( Almost like the variable Rubber diode concept)

and/or

B) Feed signal envelope to gate, to get a envelope driven clipping, ie clipping varies from start of note to end of note


How to do ?

[iainpunk]

a PUT isn't like a variable diode, i wish...
a PUT is like two BJT's that if they get triggered, they pull each other open until the current drops below a certain point. the behavior of a PUT is independent of the programming voltage after its triggered. (thyristor and PUT are like the NPN and PNP versions of the same thing)

its more like a switch controlled by the voltage over one side (anode) and a separate contact (gate)
its normally a near infinite ohm resistance, and it looks if the voltage on the anode is above the gate voltage, and if it is, it switches a diode in to place, and it only turns back to an open contact if the current drops below the threshold.

cheers, Iain

[Vivek]

UPDATE : ERRORS FOUND IN THIS POST !!!!


I tried to SPICE a PUT as a shunt clipping device (for positive swings at the moment)



Vref is 4.5 so IC output will swing around that.

Suppose we apply 4 Volts to the gate of the PUT, it will start conducting when Opamp Output exceeds 4.68V





Suppose we change the Voltage of the gate (With a voltage divider pot for example), the PUT will start acting like a diode with externally programmable Vf. This is what I meant by "act as a variable diode".

Suppose we feed Signal Envelope to the gate, clipping will be modulated by Envelope (Possibly a method to emulate time dependent harmonics, sag, swirl etc etc)

Suppose we feed a square wave to the gate, we have a tremofuzz

[antonis]

Suppose you grasp a breadbord, you verify all above..

[iainpunk]

variable clipping threshold? check this thread:
https://www.diystompboxes.com/smfforum/index.php?topic=121391.0

if your PUT has a strong DC bias like in your schematic, it doesn't even matter that you have a ground connection, since you are just using one of the PN junctions as a diode, instead of utilizing the PUT's true functionality.

if you remove the ground connection and place a diode pointing towards the Vb instead of the PUT, its exactly the same as you have now, give or take the difference between types of silicon diodes.

in your configuration you have now, place a 10k resistor between the Vgate and the actual Gate. it gives quite a different output.

i took the liberty to simulate this with the equivalent circuits with BJT's and a complementary version of it ''''clipping'''' the singal at 4V and -4V. (those transistor pairs function exactly the same as PUT or Thyristor)
but trust me, it sounds quite nasty and thin and broken.


cheers, Iain

[Vivek]

Thanks !

I will check my SPICE file again

[Vivek]

you are just using one of the PN junctions as a diode, instead of utilizing the PUT's true functionality.

cheers, Iain

Iain, you are right !!!


Back to the drawing board for me !!!

[iainpunk]

its Iain, not ian...

good luck with your ideas,

cheers, IaIn