INJ signal signal with only one int
 

It would be useful for my Mpguino version if I could get one external interrupt free. So I thought about how INJ time could be measured with only int0. My idea: trigger the interrupt on every signal change and meter time between int calls. Then compare actual measurement with last one. Shorter one should be INJ time? What is your experience with this signal, is injection time realy always shorter then time between them? I think it should be true but i am not shure.

Pulse time is much shorter than the wait time.

Consider this: you are only watching one cylinder out of 4 (or 6, or 8)... and, that cylinder only fires once per 2 revolutions, so at a minimum, the wait period will be 7 times longer than the injection time.

Do you need a specific edge trigger (i.e. rising or falling?) If not there are several interrupt pins available that will trigger going both high and low.

Yes, thats the theory. But I'm asking because I'm not shure if this is also true for most new engines. For example, I know a diesel engine (direct injection) that injects a smal amount of fuel at the beginning and the main part a short time later. Or, my motorcycle has no camshaft sensor so it can not differentiate between top dead center of compression and gas exchange -> it injects every revolition, but ~1/2 amount of needed fuel.

I think, if the diesel example is not usual in other engines it should work. I will give it a try..

Edit: dcb, yes I found the PCINT today. As I read they trigger always on SWITCH (but this would be ok I think). But for standby wake up ATMega only acepts INT0 or 1 on LOW.. I'm not shure what interrupts I need at the moment. I have one prototype MPGuino for testing (finished but untested) and thinking about the final one (a general on-board computer for motorcycle with grapical touch screen).

Hmm, I wonder if this thread would help with the diesel
http://ecomodder.com/forum/showthrea...html#post56562