UCC37322DGN

you must be doing something very, very wrong. can you post a copy of your schematic?

have you tried placing an spst switch in series with +10V, apply it gently then toggle the switch?

are you plugging in both +10V and 0V at the same time? make sure 0V connects FIRST.

do you have a long dangly wire to the 10V psu, which is an inductor, and may give rise to inductive overshoot that triggers the internal SCR across the chip supply rails? that'll break it, for sure. either pulling a pin above Vcc or below 0V can do this (if enough current flows).

stick a 1k 1W resistor in series with each output (giant Rg) and see if it still blows up. If so, then its what you are doing to the supply rails. If not, then its your FETs.

HTH

Cheers Terry

Probably....

It bounces just the same way as the connector does. As to the +10V and 0V plugged in at the same time... UCC37322DGN input is MOSFET gate (or so the datasheet says). So, I apply only 10V rail for few milliseconds without applying ground (if I am plugging 10V first) . The gate (UCC37322DGN input pin) is either at 0V or 5V (logic) - I do not see how this may be damaging. The 10V rail is very well decoupled - it is very unlikely that the 10V rail would go anywhere (because of ~1' of twisted pair connecting the board to the external voltage source.... The external voltage source is ~7F capacitor - fairly stable (should not bounce). Copper pad at the IC bottom is not soldered (the board is hand-assembled) - it is not needed from the thermal point of view, I do not see anything "electrical" about it in the specsheet . ????????????????????????????? :o(

so the chip blows up using the +10V switch?

try connecting +10V, then putting the switch in series with 0V.

do it the other way around - apply 0V first.

directly short the input to 0V. If the chip still breaks, that'll tell you its not the input causing the problem.

can you measure it?

try sticking a resistor in series with the 10V rail (eg 10R or 100R) and see if the chip still breaks. If you are triggering an internal SCR, a

its connected to the substrate, no connection to any pins - see last sentence in Thermal Information paragraph on datasheet p.11

HTH

Cheers Terry

The switch would bounce same way (I do not use it). I shortened 10V pins of the connector (to make sure that GND gets connected first. The IC just blew up. :o( I noticed is that the logic PS was off, I.e. input signal was at 0V. So.... UCC37322DGN input is at 0V, GND connected and I bounced 10V supply up and down bunch of times.... It shouls not blow it up...

External cap is 7F, onboard cap is 2200uf and plenty of ceramics scatterd around..... Oh... I took a better look at the schematic block diagram - I have couple EMI chokes that might be messing things up (too many noise sources on this board :o( ) Thanks for the postings - made me think

Need to take a closer look at the 10V rail voltage transients....

go read linear tech's AN47, it'll show you how to make sensible measurements.

Cheers Terry