[Diy_efi] 60-2 crank sensor algorythms - Detecting missing pu

[Sid.Young at qml.com.au]

Markus,

I am working on an algorithm at present that says if the timer elapsed is >
1 3/4 of the last time it is the missing tooth.

Does this sound feasable to you ?


Sid Young
Brisbane
Australia



-----Original Message-----
From: maier_markus at t-online.de [mailto:maier_markus at t-online.de]
Sent: Thursday, 3 April 2003 9:08 PM
To: Brian at lightprocessor.co.uk; List for general do-it-yourself EFI talk
Subject: Re: [Diy_efi] 60-2 crank sensor algorythms


Hi,

the algorithm is quite straight foward:

at every positive flange of the tooth, you get interrupted (by the uC
pin input). Then you're determining the time for that tooth (by a timer
or whatever is appropriate on your micro) and store the result in an
array (with the size of three). Before storing it, you need to shift
the current array contents by one.

The algorithm then to detrmine, if the missing teeth have been passed
is very easy:

your array contains the times for the last three theeth t1, t2, t3

if t2 > t1 + t3 then the missing teeth is the overlast one, which is
ok.
if t1 > t2 + t3 then the gap was detected one tooth to early and you
can run a compensation algorithm
if t3 > t1 + t2 then the gap is detected one tooth too late and you can
run another compensation algo.

Once you detected the gap, you can reset your main timer (that
represents the crank angle) to zero and reload the timers, that control
the inginition and injection pulses.

I have to admit, that I'm working with Infineons TriCore, that has
plenty of hardware timers linked to each other, which makes things
easier. But implementing that in software is also not a big deal.

Ciao
Markus



Brian Sidebotham schrieb:
> Hi all,
> 
>  I'm currently worked on a few bits that require me to
> work with a 
> bosch 60-2 crank angle sensor. Now the EFI332 project
> uses the 
> TPU section of the processor which looks like a fine
> idea, but the 
> code does not seem to be available from the site anymore
> :(
> 
>  Anyway, I am using a different make of processor, and I
> would like 
> to get crank angle information without the need for a
> TPU, but 
> using standard timers, and external interrupts, etc.
> 
>  I wouldn't even mind doing a dedicated processor that
> finds the 
> two missing teeth, and outputs them so that they could
> cause an 
> interrupt on my main processor, and the CAS signal can be
> 
> connected to a timer input on my main processor to keep
> track of 
> engine angle.
> 
>  What has everyone else done?
> 
> Cheers, Brian Sidebotham.
> 
> 
> _______________________________________________
> Diy_efi mailing list
> Diy_efi at diy-efi.org
> http://www.diy-efi.org/mailman/listinfo/diy_efi
>

_______________________________________________
Diy_efi mailing list
Diy_efi at diy-efi.org
http://www.diy-efi.org/mailman/listinfo/diy_efi

_______________________________________________
Diy_efi mailing list
Diy_efi at diy-efi.org
http://www.diy-efi.org/mailman/listinfo/diy_efi