I've been having issues with surge testing for a product of mine and am looking at replacing my current front-end with your eFuse. I'm attaching a copy of my schematic and layout (2 layer board). I'm using the TPS26621 and have the resistors setup for ~7v for uvlo, ~26v for ovp and ~0.66a for Iol. I haven't dialed in the dvdt cap yet. We're trying to pass IEC 61000-4-5 w/ +/- 1Kv pulses across the power lines...
1) does the design look reasonable,
2) your design calls out a bidirectional TVS, I'm curious why we wouldn't use a unidirectional TVS. Isn't there just a diode in the reverse direction in that case?
3) I don't have a lot of room for this, so the copper pour for the RTN/PAD is reasonably small, I assume we're dumping heat into that copper? Is that there mostly for in-rush current on startup or is it required during surge (I would imagine the TVS would dump the bulk of the surge energy).
4) I'm using a very beefy TVS, I need to operate normally at 24v or less, could I get away with a smaller, less expensive part here and still have some margin?
5) I like the auto-restart feature of the TPS26621 and the ability to set the OVP voltage (my downstream circuitry doesn't want to see more than 30v (absolute maximum) with 26v as the rated maximum.
6) We're looking at many thousands of these boards so I'd like to have something reliable that's cost-effective too.
Anything else I should consider?
Thanks,
David Pariseau.
