Hessitation in Late Model Subarus

[Don.F.Broadus at ucm.com]

Dr. Plecan is once again correct in the optimal knock/misfire detection
system. Labeled as a magnetorestrictive sensor, this device measures the
accle/decel (twist) of the crank. SAE has several papers on this type of
sensor and research continues for incorporation into cars. A ring cast into
the crank similar to a DIS timing ring is used to trigger the
magnetorestictive sensor. Experiments could be conducted using this ring an
possibly a modified mag pick up. So far it is claimed that cycle by cycle
detection is possible for each cylinder.
                                                            Don 
                                      

> -----Original Message-----
> From:	Bruce Plecan [SMTP:nacelp at bright.net]
> Sent:	Saturday, May 08, 1999 11:35 AM
> To:	diy_efi at efi332.eng.ohio-state.edu
> Subject:	Re: Hessitation in Late Model Subarus
> 
> 
> OK, just for giggles, how do you tell a bad A/C compressor bearing that
> radiates a frequency of 25 Hz,  from a knock "noise" of 25 Hz?.  (Feel,
> free
> to change to any freg of noise)
>   Repeat the above, with fan belt "noise", Alternator bearings, throw out
> bearings, rocker arm noise lifter noise etc, etc..
>   If you can do it, more power to you, but, I'll still stick with my
> original statement.
> As for miss detection, I'll rely on crank accleration rate changes thank,
> you.
> Bruce
> 
> 
> > > That's not the problem as I see it.  It's seeing a knock you don't
> want
> it
> > > to see.
> > > That is a problem with all these dumb acoustic sensors, only cure is
> going
> > > to ionization kknock detection, please refer to the archives and
> search
> ION.
> 
> > I don't actually believe there is anything wrong with the idea of a
> > acoustic knock sensor, I really believe that most of the issues are
> > with crude way they are doing it (simple filter).  I believe using a
> > acoustic sensor and DSP and some code using a better algorithm would
> > do alot better in removing other noises.     I would also think you
> > could do miss detection using this method also, ie listen for each
> > cylinder igintion and count the ones you missed.     Now I just need
> > to put a mic on my engine and collected and figure out how to
> > process the data in such a way to get meaningful data.
> > Roger
>