Ok, that puts a little different light on it.
I'd suggest you google MegaSquirt.
It's an open source EFI engine controller that has gone through
a lot of improvement over the years. Even if you have no intention
of using it, it'll point out a lot of details of how to do these things.

-Denny


On Wed, Nov 21, 2012 at 4:32 PM, nik . <xyzqhtc@gmail.com> wrote:

> Hi Denny, I want to multiply the rpm with map values (continuously update=
d
> too) to derive injector pw. All this is much later. I'm starting with rpm
> ideas and writing some basic code for that.
>
> Rev range 7000rpm with 250rpm or 500rpm res.
> On Nov 22, 2012 11:18 AM, "Denny Esterline" <desterline@gmail.com> wrote:
>
> > On Wed, Nov 21, 2012 at 3:49 PM, Spehro Pefhany <speff@interlog.com>
> > wrote:
> >
> > > At 04:56 PM 21/11/2012, you wrote:
> > > >A 4 stroke/4 cyl engine has 2 ignition pulses per 1 crankshaft
> > revolution.
> > > >If this is so then whats the technique behind counting RPM with a
> PIC. I
> > > >need some ideas only. Please don't discuss HW or other.
> > > >
> > > >I've gone as far as thinking the RPM will have to be fed to a CCP pi=
n
> as
> > > >input. I understand how the CCP will be setup but can't quite get th=
e
> > idea
> > > >of how exactly to count RPM and say if that's sorted out, display it
> > (this
> > > >part is not a concern right now).
> > > >
> > > >Any ideas would be great.
> > >
> >
> > Other's have already given you good answers, but there's another piece =
to
> > the puzzle - what are you doing with the data? In other words, how
> accurate
> > does it need to be, over what operating speed range and how frequently =
do
> > you need new data?
> >
> > The time between pulses method works very well on the low end, but
> > the resolution isn't as good at high speed. But at some (very low) spee=
d,
> > the timer
> > will roll-over before the next pulse arrives. You have to handle that
> > differently.
> > It's also very quick to provide new data, you get new information twice
> per
> > revolution.
> >
> > The pulses over time method has better resolution on the high end. The
> > trade offs
> > include slower update rate, simpler math, among other details.
> >
> > Modern automotive designs use time between pulses - but they use the da=
ta
> > to
> > do some rather interesting tasks (detecting single cylinder misfires) s=
o
> > they
> > need new RPM data many times per revolution (several hundred times per
> > second).
> > Displaying this for a TACH for a human to read would be nutty, even a
> > couple
> > times per second is faster than most people can see/react to.
> >
> > So, what are you doing with the data?
> >
> > -Denny
> > --
> > http://www.piclist.com PIC/SX FAQ & list archive
> > View/change your membership options at
> > http://mailman.mit.edu/mailman/listinfo/piclist
> >
> --
> http://www.piclist.com PIC/SX FAQ & list archive
> View/change your membership options at
> http://mailman.mit.edu/mailman/listinfo/piclist
>
--=20
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
.