Can a Square Wave at 500 kHz and 2.5 kV Be Achieved?

[grae313]

I've been told that a square wave at this speed and > 1 kV P-P won't happen, but I need one, and our lab has plenty of funding. Is this possible? Commercially available? If 500 kHz is out of the question, what about 100 kHz? Any information is appreciated--I don't know **** about electronics :(

 

[grae313]

I've been told that a square wave at this speed and > 1 kV P-P won't happen, but I need one, and our lab has plenty of funding. Is this possible? Commercially available? If 500 kHz is out of the question, what about 100 kHz? Any information is appreciated--I don't know **** about electronics :(

I guess I need to talk about rise times... it just needs to be an improvement over a sine wave, so I guess maximum rise time would be 0.5 us to produce a square-ish wave at 500 kHz.

 

[es1]

That is very likely going to be tough. What's the load?

 

[Okefenokee]

A 1,000 Volts square wave?! What kind of application needs a signal like that?

Have your engineers tried simulating an ideal square wave connected to a high frequency transformer? Some transformers can go up to about 3MHz or more. That would allow plenty of harmonics to slip through and rebuild a reasonable approximation of a square wave on the high voltage side.

The transformer will, of course, have some frequency dependent phase distortion. On the generator side, you can deliberately create harmonics and phase shift them in a way that cancels the effect of the transformer. That way, the main 500MHz oscillation and all it's harmonics will add up to a square wave just like a Fourier transform on the high voltage side.

Anyway, that was just me brainstorming. I saw an 800V thyristor on Digi-key. That's almost strong enough for your app. I don't track trends in power semiconductors so who knows what you can get if you know the right people to call. There may be a 1000V switch you can buy right off the shelf.

Good luck.

 

[grae313]

That is very likely going to be tough. What's the load?

Uh... like I said I don't know **** about electronics :shy: But the voltage I need is headed for plates on either side of a crystal (an electro-optic deflector) so I think the load would just be a result of the capacitance of the configuration? The capacitance is around 50 picofarads.

 

[grae313]

A 1,000 Volts square wave?! What kind of application needs a signal like that?

I want to modulate the position of a laser beam using an electro-optic deflector, and the deflection sensitivities of EODs are in microradians per volt. I'll need at least 1.5 kV for the design to be feasible at all, 2.5 would be better, and I want a square wave because there should be some dwell time to the beam at its peaks.

Have your engineers...

Heh. I'm a graduate student in physics. I have no engineers I also have no idea what you're talking about! Anyways, it's sounding like the answer to 500 kHz, 2.5 kV square wave is heck no.

A sine wave with these specs is easy to achieve. Is there any way to modify it somehow to get it a bit squarer? Sorry for my ignorance and thanks for the suggestions...

 

[f95toli]

Heh. I'm a graduate student in physics. I have no engineers ...

Have you tried contacting any of the companies that make similar products (e.g. amplifiers for piezo drivers)?

This one comes pretty close to what you need if you use two

http://www.lab-systems.com/products/amplifier/a303.html

Also, even if this is not exactly what you need they might be able to come up with a custom design (At a price of course), at 450kHz BW you should be able to get a decent square wave at 100 kHz.

 

[grae313]

Have you tried contacting any of the companies that make similar products (e.g. amplifiers for piezo drivers)?

I sent off a few emails so I'll see what kind of responses I get. Thanks for the suggestion.

 

[dmtr]

It's possible. You are trying to build high voltage pulsed power supply (pulse generator). High voltage/high frequency switches are commercially available. For example: http://www.behlke.de/

And here is an article in which a similar square impulse generator (2Mhz/several kV) was designed: http://accelconf.web.cern.ch/Accelconf/p99/PAPERS/THDR4.PDF

 

[dmtr]

Here is ready to use lab unit: GHTS 30 (3kV, 4MHz (is it ?) ).
http://www.behlke.de/pdf/ghts.pdf
http://www.behlke.de/separations/separation_d.htm

 

[dmtr]

BTW, is it a beam chopper? Something along these lines?
http://accelconf.web.cern.ch/Accelconf/e08/papers/thpp094.pdf