Continue to Site

Welcome to

Welcome to our site! is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

HVDC semiconductor switch

Not open for further replies.


Member level 5
Jul 22, 2014
Reaction score
Trophy points
Activity points

I ideally want a semiconductor/solid state switch capable of a few amp at several (about 5kV) DC.

Best I can come up with so far is putting a few 1.2kV (ish) IGBTs in series with OPTOs driving each device individually.
Looking at it in depth, it appears that I need a resistor across each device to force equal voltage sharing when off. Typical worst case leakage current of the devices is 1mA, to 10mA (rule of thumb) through the resistor chain - doesn't sound ideal.

Possibly need a snubber across each too in order to slow things down so that differences in the OPTO speeds doesn't cause momentary problems with overvoltage on some IGBTs because others have turned on faster.

Doesn't feel like an ideal solution but struggling to come up with anything better. Am I missing something or am I overcomplicating it (i.e. do I need the resistors and snubbers?).


what's about using zener diodes in parallel e.g. four times BZG03C240-M or BZG03C270-M [1], their test current is specified with 2 mA. Something you should keep in mind when using a cascaded approach, that if one of your IGBTs turns off faster (or on slower) than the others, it will see the full voltage of 5 kV. By using the zener diode approach, you are limiting the voltage drop across this IGBT. Of course, you have to limit the current througth the diodes as well.



Research ( as a verb ) papers on cascaded mosfet switches 5kV and higher - there are plenty of goods ones covering this topic - with practical tips,

as well as resistors or zeners for static ( off ) volt sharing - you need snubbers for volt sharing under dynamic conditions too.

Often the top most gates suffer from delay in the gate drive design as their sources are falling past ( pulling more current out of the gate drive in CM & DM fashion ) compared to the lower devices where the source may be much more static - so that is a thing to consider. similarly when turning off - the top most may be last - this is why zeners are such a good idea ... ( as mentioned just above too ).

good luck

Not open for further replies.

Part and Inventory Search

Welcome to