When a Mosfet goes bad.

[Ray-NM]

Just an observation from my own mistakes.
If by chance your fet on the HH runs too hot for too long or your wires to the coil shorted momentarily then check the drive signal on the gate if it has a step when the fet is turned off then best replace it. I have done this a couple of times and find that the turn off time increases for the fet. A known good fet has a very quick turn off, I think in the ns range. A bad one is on the order of 1/2 usec.
RayNM

[1843]

Hi Ray-NM,
My IRF840 (or 740) fall time is about 1.5us with LM556 gate driver.

Regards

[PhilippeM]

DATASHEET IRF 840

What's mean exactly for you "1853": FALL TIME?

here is the datasheet URL:
http://www.ortodoxism.ro/datasheets/...ild/IRF840.pdf

TOFF is 74ns but very depedent of the temperature .
Regards
PhilippeM

[Ray-NM]

1.5 usec is not good. Measure at the gate to be sure. You can disconnect the coil I think to do the test.

[1843]

1.5 usec is not good. Measure at the gate to be sure. You can disconnect the coil I think to do the test.

When I measured it, Coil was disconnected...

If you mean the gate driver fall time, I think it is about 1us or so...

[1843]

What's mean exactly for you "1853": FALL TIME?

The time that MOSFET turns off...

[Max]

The time that MOSFET turns off...

Hi,
generally speaking... the time considered here is 'the time required for signal to go from 90% to 10% of its maximum value' but you need to compare with gate signal shut-off as reference...then use it as reference e.g. in middle point of gate signal and drain signal.

I think 1us could be even better ...try to reduce e.g. cable capacitance ...a good coax (e.g. some rg-1xx)

Best regards,
Max

[woody]

Fets

Try replacing your IRF840 fets with ST W20NK50Z this is the same power fet Minelab use in their P.I detectors. They are bullet proof, fast and have low output capacitance that helps in the dvdt realm.