I'm trying to design an IEC61000-4-4 compliant EFT generator for in-house use, which is based on following schematic: basic schematic

I've been able to determine all values for all components, And I've been able to generate a wave which meets the specifications

rise time 5ns +- 30%; duration time (to 50%) = 50ns +- 30%; Vpeak = 1000V +- 10%

I've accomplished this using 4 avalanche transistors in series, with base and emitter connected, and the last transistor triggered with a 1:1 transformer: enter image description here

The problem with this method is that I cannot vary the voltage from 0V to 4kV, as the avalanche circuit only works for a certain voltage range.

As I would like to make the output voltage variable from 0v to 4kV, I am looking for another way to implement the switch. It needs to switch within 5ns.

If anyone has an idea on how to do this, thanks in advance.

  • \$\begingroup\$ How about thru mechanically tuned gap with a programmable charge pump \$\endgroup\$ – Tony Stewart EE75 Mar 5 '18 at 17:02
  • \$\begingroup\$ I would make it variable to 8kV with 30A pk with 1ns rise time. \$\endgroup\$ – Tony Stewart EE75 Mar 5 '18 at 17:20
  • \$\begingroup\$ There is probably a reason why a basic EFT generator as EFT 500N5 is priced at $24,000. Not exactly the home-use material. :D Did you try to research service manuals for industrial equipment of this sort, to get ideas how it is done? \$\endgroup\$ – Ale..chenski Mar 5 '18 at 20:09
  • \$\begingroup\$ No idea about 5ns, but HV transistor is done like this: upload.wikimedia.org/wikipedia/commons/d/dc/… \$\endgroup\$ – τεκ Mar 6 '18 at 4:09
  • \$\begingroup\$ @TonyStewart.EEsince'75 The charge pump is a very good idea! \$\endgroup\$ – BertVano Mar 6 '18 at 8:10

You might want to examine the following publication, LOW COST IEC61000-4-4 COMPATIBLE PULSE GENERATOR, which has some practical elements.

enter image description here

The developed switch stack uses four STP4N150 power MOSFETs.

EDIT: some theory behind the switch design can be found here.

  • \$\begingroup\$ Thanks for your reply! I had already read those publications, but this circuit is not ideal because at higher voltages, the fast rise time gets compromised due to the Miller effect in the MOSFETs. Otherwise these are very interesting publications though... \$\endgroup\$ – BertVano Mar 6 '18 at 8:04

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.