Narwhal:
dc42:
What's the application? If you need pulses of 50kV at a few mA, it's easier to generate pulses at a lower voltage and step them up using a transformer. If you need to interrupt an otherwise continuous 50kV supply, you could derive the supply from an inverter and switch the inverter off to interrupt the supply, if you don't mind the supply taking a few milliseconds to turn on and off.I am doing an experiment that involves the timed switching of ground terminals before the charge reaches the terminal (making it go in a circle until it runs down the voltage due to resistance). I need the high voltage for other applications, but a big plus to it is the electrostatic field it generates on uninsulated wire allowing me to measure it during the process.
Obviously the charge is going extremely quickly, so a switch with around a 100 ns response time would be good.
25kV on uninsulated wire is going to cause corona discharge and flash-over - I'm confused, electrostatic levels of voltage are not as easy to measure as low voltages, yet you seem to imply they are necessary. If you can avoid 25kV you'll make everything easier (and safer).
For fast switching at low voltages there are transistors and MOSFETs and these are simple to use and cheap. At 25kV you're talking exotic high-voltage valves as far as I know - the sort of device that can drive a large Kerr cell perhaps?