Herbert Graf farcite.net> writes: > And still you've violate specs. Diodes don't turn on in zero time, they > need time to switch from forward conduction to reverse conduction. > During that time the voltage on your PIC pin will be out of spec, by > quite a bit depending on the turnaround time of the diode. One of the Schottkys I analyzed for this turns around in 5 nsec from +10 mA to - 10mA (data sheet data). I assume that it turns around in 1-2 nsec or less from I~zero. By that time the circuit board wiring and lead capacitance count more. 1 nsec is 1 Ghz. If the PIC can switch on/off that fast (it can't) then it should also be able to cope with the self-generated du/dt. Else it would be a harakiri pic by definition. Iow, the Toff of the output mosfet is so slow that the diode will be on a long time before it finishes switching. > Shottky's are MUCH faster then PN junction diodes, but they still have a > finite time to switch. Yes, please see above. > I just wouldn't be personally comfortable connecting a relay directly to > a PIC (or any other MCU) pin, it's just asking for trouble IMHO (heck, I > even go as far as putting the relays on a seperate power supply if > possible). Try that when running from a single cell supply boosted by a low power booster to 3V, or straight off a Lithium cell, or from a 'derived' power supply that can take no load and no noise (like USB and so on). Peter P. -- http://www.piclist.com PIC/SX FAQ & list archive View/change your membership options at http://mailman.mit.edu/mailman/listinfo/piclist