Re: MOSFET driver killed with no particular reason

On Apr 19, 3:52 am, tony.sunny...@xxxxxxxxx wrote:
All Bridge drivers have the potential for short-circuit failure unless
there is a safety circuit to prevent this. With separate drive
circuits for simultaneous complementary switching , it is cruicial to
control the switch delay to allow some dead-time in the switches and
have each one protected for reverse pulses. Asymmetric delays for
rising and falling pulses and reactive loads all contribute to this
risk. Try to design the pulse delays so that the turn-off delay is
always shorter than the complementary switch's turn-on delay. This can
be controlled by adding turn-on delay only in the pre-driver signal to
compensate for the natural longer off delay of the switch. Examine
worst case time delays with temperature effects and best case to
calculate the dead-time needed to avoid short-circuiting the high-side
and low side switches thru each other to power. A power transistor
will have a safe operating area for voltage and current but also look
for deratings from thermal effects and pulse speed. Once these limits
are all observed, your MOSFETs will last a long happy life.

Tony
EE'75

Thanks for your reply. I understand the necessity of deadtime to
prevent shoot through in bridge, and as said in driver's data***
such prevention is already built in. Moreover, it's not MOSFETs, but
the driver itself get destroyed by just playing with input rails H/L,
even haven't connected them to high switching frequency yet.

This picture shows what I did:
http://www.geocities.com/w2kwong/MOSFET.jpg
(NOTE: bootstrap diode is built in)

.

Quantcast