0
\$\begingroup\$

This is the current waveform of the switch and the Vds of the MOSFET.

Ideally, the current should fall off in a normal fashion with the switch turns off.

Here, the current fall stops and becomes flat for some time and then falls again.

Unfortuinately, the flat current here is flowing when the switch is operating in the active region and resulting in huge power losses.

Any views here as to why this is happening?

enter image description here

\$\endgroup\$
5
  • \$\begingroup\$ Does the power supply have a load attached to it? \$\endgroup\$ – David Mikeska Apr 15 at 9:52
  • \$\begingroup\$ yes, it does have the load. \$\endgroup\$ – Durgaprasad Apr 15 at 11:14
  • \$\begingroup\$ Need to work on a pulse shaping, The slight negative voltage is from the rectification effect. In crt televisions, the unloading effect is tampered by what we called "saftey capacitors " (usually, 0.000033 and additional reverse diode) But it is the same loss in those crt horizontal/power sections when the signal is present. I think it has to do with the class C operation the device is put in. A gate stabilization resistor might help, but the main thing is it seems like you need to do some wave shaping before and after the MOSFET. \$\endgroup\$ – David Mikeska Apr 15 at 11:37
  • \$\begingroup\$ You might even have luck using a 0.3 ohm resistor from source to ground. But it is hard to tell that without a schematic. \$\endgroup\$ – David Mikeska Apr 15 at 11:45
  • \$\begingroup\$ Where do you measure the current, in the drain, in the source, across the sense resistance? Can you please add the driving voltage \$v_{GS}(t)\$ as that plateau is likely to be the Miller effect at turn-off. \$\endgroup\$ – Verbal Kint Apr 15 at 18:28
0
\$\begingroup\$

Try adding a 2.2 nF capacitor in parallel to the MOSFET, between Drain and Source.

Then make a new measurement and see if that odd behavior gets amplified.

It if does, the answer might be the Drain to Soruce parasitic capacitance of the MOSFET.

What MOSFET is that?

\$\endgroup\$
7
  • \$\begingroup\$ OK. C2M1000170J is the MOSFET part. \$\endgroup\$ – Durgaprasad Apr 15 at 8:22
  • \$\begingroup\$ That MOSFET has a huge TAB connected at the Drain terminal. It's a SiC MOSFET. Do you really need 1700 Vds-max? \$\endgroup\$ – Enrico Migliore Apr 15 at 8:27
  • \$\begingroup\$ Yes, my operating voltage is 850V, I hit voltages near to 1.4kV. Its needed. \$\endgroup\$ – Durgaprasad Apr 15 at 8:29
  • \$\begingroup\$ at those voltage and at those switching speeds, it doesn't take much stray (L or C) to cause this. What is the layout like. What is the gatedrive like? it could be poor layout to the gate and as such miller charge injection is partially turning the device back on \$\endgroup\$ – JonRB Apr 15 at 8:30
  • \$\begingroup\$ If you hit 1.4 kV I guests your transformer has got a big leakage inductance. Can you confirm that your voltage spikes are due to the leakage inductance? \$\endgroup\$ – Enrico Migliore Apr 15 at 8:39

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.