Hi,
If rise time is a problem then this should occur even before at slower
rpms, Rise time would nto be different at higher rpm.
What is the timer clock rate. Is it possible that your interrupt
processing time is coming into picture and you may be missing an
interrupt. However this should show a lower rpm and not a higher rpm.

Why not do an averaging. This is what is done on all rpm meters anyway.
You may like to use integration type of measurement (i.e count pusles in
a known period) rather than counting pulse width  and inferring speed.
rgds
ramana



-----Original Message-----
From: Heinz Czychun [mailto:hgczyc@REACH.NET]
Sent: Saturday, September 30, 2000 4:59 AM
To: PICLIST@MITVMA.MIT.EDU
Subject: Re: [EE]: Problem with PIC Project in a Car


At 12:01 AM 9/30/2000, Christian Dorner wrote:
>....
>It seems like the interrupt will be fired correct. I'm really confused
about
>the fact that the result is sometimes higher and sometimes lower than
the
>expected result.
>When the Port-Change interrupt is the problem than i think the PRM must
be
>everytime to low. (the interrupt occures to late and the messured time
is to
>long)
        If the trigger time is wobbling all over the place and there is
no
apparent noise, then I would concentrate on the fall/risetime of the
signal. Is it clean and fast? If not change to the schmitt trigger input
on
the PIC, or add a schmitt trigger gate. I would expect this is
necessary,
as a minimum, to condition the signal.

Just my 2 cents CND,
Heinz


>
>Thanks for your idea.
>
>Christian Dorner
>aka Doc, ...
>
>--
>http://www.piclist.com hint: PICList Posts must start with ONE topic:
>"[PIC]:" PIC only "[EE]:" engineering "[OT]:" off topic "[AD]:" ad's

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
"[PIC]:" PIC only "[EE]:" engineering "[OT]:" off topic "[AD]:" ad's

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
"[PIC]:" PIC only "[EE]:" engineering "[OT]:" off topic "[AD]:" ad's