Pulse width limiter with Schmitt trigger

IIRC if the voltage on one of a CMOS chip’s pins drops sufficiently below ground or above the positive rail it can trigger large currents that can only be stopped by removing the power temporarily. This is because the N and P type silicon is often arranged in such a way that it creates accidental thyristors between the two power rails and these thyristors can be triggered by voltages exceeding the rails, whether it be below 0V or above 5V. If it’s discrete MOSFETs then it’s not an issue, but it’s something to be aware of for CMOS chips. If the voltage is enough to flow significant current through a silicon diode, then no doubt it’ll trigger the silicon thyristors. Perhaps a germanium or Schottky diode would work, but it’s still a little risky. A possible solution is if the ground side of the diode was kept at 0.7V, perhaps with the help of another diode plus a capacitor to keep the voltage constant. That way the voltage won’t drop much below 0V.

@jason9: Thanks! I didn't know that. Got me thinking about the 0.7v rail.

Hmm... Wouldn't it actually take at least a diode's turn-on voltage to trigger a thyristor? So a Schotty diode would help, in that case.

Yeah. I recommended the 0.7V rail because a Schottky might still be a little risky. I don’t know just how much it takes to trigger the thyristor, so it could be lower than expected, which combined with an initial voltage spike as the input jumps from 5V to 0V might cause trouble.