Further to my last post, I'm investigating why I appear to be getting discrepancies between reported and measured rpm.
So I have a scope hooked up the the car, and at approx 3000 i can measure the spark fire duration at 3.7ms and the gap between sparks as 17.2ms (its a 4 cycllinder,4 stroke with 2 coils, i'm only reading one)
The output from 1 sample of inputcapture at the same time looks like this:-
32458
5963
which i work out to 17.59 m/s and 3.23 m/s accordingly i.e. both a little long compared to the scope.
With timer1 speed set to 2 the count is of 542 nanosecond ticks I believe. Does this mean that there is resolution to 542 nanoseconds which would be ample or is there a minimum count interval ?
This is running on an old ZX24A as a task with little else going, Is there merit in changing it for a native mode device or putting the count in the main body of the code ?
thanks