Problems we have actually had to solve.
Not product promotions. Each note takes one real design problem, shows what was measured, explains why it happens, and ends with the datasheet check that prevents it next time.
No notes match that. Try a different term.
Latch-up: the protection device that never turned off
The board passes every functional test. Then it goes through ESD testing, and one port stops working. The protection device is warm. Replace it, power-cycle, and the board is fine again, until the next discharge. The part is not too weak. It is still switched on.
Read the noteNegative-voltage EOS: the supply that went below its own ground
Units come back from the field with the touch controller destroyed. The obvious suspect is a discharge on the touch lines, so the argument becomes which protection device to fit. Then somebody scopes a returned unit while it is running, with nobody touching it, and finds -3.26 V on the 3.3 V rail.
Read the noteSurge is not ESD: the port that passed 8 kV and died on a cable
The qualification report is clean. Contact discharge to every exposed pin, up to the required level, no failures, criteria A throughout. Then units start coming back with the interface pin shorted, and the common factor is not a person touching the connector. It is a long cable run.
Read the noteDamage that builds up: the failure with no event attached to it
Full IEC 61000-4-2 qualification, every level, every pin, criteria A throughout. The product shipped. Months later units come back with the interface pin reading a few hundred kilohms to ground, and nobody can name a single thing that happened to them. No drop, no strike, no complaint about handling. They simply stopped.
Read the noteAir is not contact: the pass that only held at 40 percent humidity
The product has a plastic bezel and no exposed metal, so there is nothing to put a probe tip on and the whole thing is qualified by air discharge. Ten discharges per point, every seam, every button, at the required level. No failures, signed off. Then the field reports lock-ups, and they cluster in winter and in the dry regions.
Read the noteCrosstalk: two pairs that were both fine on their own
Each high speed pair was tested by itself with protection fitted. Both met their eye mask with margin and the ESD qualification passed on every pin. Run both pairs at the same time and one of them starts showing bit errors. Nothing is damaged, nothing runs hot, and the link comes back on a retrain.
Read the noteToo much protection: the port that worked until it was protected
A 12 Gbps link passed compliance on the bare board. The team then upgraded the protection: higher peak pulse rating, higher kilovolt number, better clamping. On the selection table the new part won every column anyone was looking at. The link stopped meeting its eye mask, and nothing else on the board had changed.
Read the noteHave a board doing something you cannot explain?
Send us the schematic or the protection BOM. Our FAE team will go through it with you.