MOV time domain voltage graph

[coppercone2]

Does anyone have a graph that shows what a MOV does when it clamps a voltage? Say like if there is a AC or DC relatively high voltage and a spike is regularly pulsed on it.

Does it just flat out like a clipped rail in a op-amp, or does it have some kind of hysterisis where it takes the rail down under its regulated voltage during the end of its conduction phase when the transient decayed? I don't think I ever noticed a subsequent dip with op-amps before, but I think they are not operating correctly as they leave saturation. Does it have a dependence on supply impedance?

Preferably a before and after.

I can't find a good picture on line and I don't have a simulator here. I mean if its trigger voltage is really close to the rail voltage. When its way above it seems to be pretty flat but I wonder if there is a slope to the line (there are only conceptual pictures I can find in datasheets, not measured ones). 

[Rerouter]

It behaves like a zener clamp, it starts conducting at a knee voltage. and shunts far more current as it continues

this knee voltage also degrades with each conduction event and with the length of a conduction event. for mains protection this can lead to it reacting to normal mains, and collapsing down to a short / flare / thermal explosion

[coppercone2]

so say it starts conducting at 155 volts. What would it typically stop conducting at?

could you make a special mov that is almost broke by design so you can trigger it with a HV pulse to act as a 1 shot crowbar?

So its used as a over current triggered shunt that blows itself up or gets red hot or whatever it does.

can you reliability degrade a commercial MOV to function like this (one of the big ones that is like a puck).

[T3sl4co1l]

Offhand I found a waveform here, https://www.bourns.com/data/global/pdfs/ACTP250J1BJ.pdf

The V-I curves are in the MOV datasheet from which you can calculate maximum peak voltage.  You can also run a simulation with whatever SPICE models you can find, though all the ones I've seen are written terribly; maybe they still run on LTSpice.

Tim

[BravoV]

Not sure the model's accuracy, claimed to be Littelfuse's MOV part no V20E175, the LTSpice model & simulation originated from -> HERE

Voltage clamped at 480V and the MOV's dissipation attached below.

[coppercone2]

hmm it seems like a regular MOV is really flat.

I am going to assume trying to damage it precisely is not possible?

[T3sl4co1l]

Yeah that right there is a model that ain't working worth shit... or its parameters are completely wrong.

Tim