Hi to everyone, I'm trying to generate a 5-10 ns pulse width out of an edge, without the use of any clock, just with internal delays. Unfortunately I can imagine how much this delay will depend on temperature and voltage and how much the width will be affected, but still I have some margins (anyway everything will be tested in a thermal chamber to verify functionalities). So far I developed a sort of a "ripple-counter" with an auto-reset for each FF (Q(0) will asynchronously reset Q(0)) so that I can start an internal oscillation (a clock!). After that a normal counter can be realized on the basis of this clock, allowing to build a pulse. Once the pulse is over a reset is generated to stop asynchronously all the FFs. Does it make sense at all? I think no! :-)
Any ideas?
thanks a lot
Al