did you check the IBIS model?
almost all "modern" CMOS logic explicitly allows 10s of mA injected
old fashion parallel PCI used source termination
did you check the IBIS model?
almost all "modern" CMOS logic explicitly allows 10s of mA injected
old fashion parallel PCI used source termination
Not a ringing endorsement.
It always is, until it misbehaves.
Scoping a working device does exclude the ones that don't work.
It's more difficult to get them absolutely right.
I had precise A/D converter that got a little less precise when the input being digitised got close to the rails. A germanium diode that clipped an in-coming signal to ground before the substrate diodes could catch it solved the problem. Latch-up is a rather more gross failure.
They talked about putting "moats" around sensitive bits of circuitry to prevent charge carriers migrating in from the wrong side of the tracks. Barry Gilbert really did have a stable of impressive experts, and he doesn't seem to have drunk as much as Bob Widlar. I met him once and like him. I doubt if I would have enjoyed a face-to-face encounter with Jim Thompson.
Have something to add? Share your thoughts — no account required.
Ask the community — no account required