How to chop straight DC?

[jsmith12]

I have a high voltage DC signal I would like to convert to pulsed DC, or a sine wave for that matter. The point is to charge a cap bank which when reaching 1200V is triggered by the gas discharged tube to create a higher voltage pulse out T2 as shown. The problem is straight DC will only allow the circuit to pulse once and then short out. The way I think to solve this problem is to convert the straight DC into a fast pulse or sine wave so the gas discharge tube can see the voltage has dropped and reset to allow the recharged cap bank to pulse again.

What is the best way to chop this straight DC signal? An n-channel mosfet as a low side switch (gas discharge tube between the 1200V+ and drain) does not work because the gas discharge tube still never sees the voltage drop to reset. They don't make p-channel mosfets above 600V so I can't use one to create a repeating high side switch (with GDT after the source of the mosfet).

Would an IGBT work for this? And if so how would it be wired?
http://ca.mouser.com/ProductDetail/...=sGAEpiMZZMv4z0HnGdrLjs5sdeiuc7sc6o7mxtpM2uI=

Thanks for any suggestions.

 

[kpatz]

Where are you getting 1200V from?

It might be easier to "chop" a lower voltage and then step up through a transformer.

Why wouldn't a low side switch work? The voltage only needs to drop across the tube for it to stop conducting, but it can still be 1200V relative to the low side. The high side and switched-off part of low side would both be 1200V during the off cycle, which translates to 0V across the tube.

A series resistor may do the trick too, if it drops enough voltage to get the tube to stop conducting. It'll slow down the rate the cap charges though. What's the desired timing cycle?

EDIT: I just googled "gas discharge relaxation oscillators" which is basically what you're trying to make and most of the schematics show a series resistor and parallel capacitor. So the resistor is probably the way to go.

 

[jsmith12]

1200V comes from an adjustable DC power supply. I don't think the GDT can be 1200V relative to the low side, I find it doesn't reset which allows it to sparkover again. It just stay open conducting. So I thought I could use a IGBT to put the circuit/GDT below the IGBT's source which I assumed would not show 1200V to the GDT when off. I just don't have much experience with IGBT's and thought there might be a simple schematic someone had which shows this, preferably being powered/driven by a waveform generator like I do with a mosfet.

 

[kpatz]

An IGBT could work if you can find one that can withstand a 1200V C-E voltage. But my thought is that a series resistor before the capacitor would also work. When the GDT fires, it essentially closes the circuit between the capacitor and the discharge transformer. The resistor would then drop the voltage coming from the 1200V supply as well as limit the current so the GDT can reset after the cap discharges.

Doing it that way, it would automatically "chop" when the GDT fires, making a relaxation oscillator.

What's the desired repetition rate? How many times per minute or second does the circuit need to fire?

 

[jsmith12]

The frequency of the pulse needs to reach about 250Hz, a resistor is not practical I think.

I'm looking for an easy way to create a high side switch with an n-channel mosfet or IGBT but its difficult because I guess I have to float the voltage for the gate up to 1200V first.