Software Dyno...

[Patrick G. Moore]

Hi Dave,

Been thinking about the same thing.  If you could just pull
timing for traction control, say make it a closed loop for 10%
tire spin, launches would be automatic!  Someone in England has
a web page where they sell a traction control which acts like
a rev limiter. ie. it cuts spark in alternating cyls.  They have a graph
which shows friction vs wheel spin and usually around 10% is optimum.
But this is for road course or F1 cars, I don't know about a drag
race and slicks, but hey .. just try different settings.  Thier system
uses ABS also, they have a sensor kit in case you don't have ABS.

I think pulling timing would provide a much finer control without
dumping raw fuel, and is more immediate than the dumb throttle control.
I don't like the idea of shutting off injectors either, as I may
be using N2O/alcohol also, and quickly melt the engine.

Problem with VSS is that you can't detect tire slip unless you can tell
the difference between the front and rear tire rotation.  You also
have a problem with ABS sensors if the front wheels are off the ground!
In this case you would need an accelerometer and compare acceleration
to that you would get if there was no slip at the driving wheels.
ie. x.x tire rotations = 5ft accelerometer integrates to 4.5ft.  The
VSS would work fine in that case, though it might not be fine enough
resolution.  Could also use some sort of ground looking radar for
true speed (instead of accelerometer).

I was thinking of this because of the problem of my front wheel
drive car and getting optimum launches.  I don't have a problem with
wheels coming off the ground! So if I put abs sensors on, and could read
an extra analog input in my ECM. An external circuit that converted
% wheelspin to an analog voltage could feed a little piece of code that
pulled timing to maintain a certain voltage:)  Whew...

Let me get my car on the road first ;)


Dave Zug wrote:
> 
> This tradeoff I'd be willing to live with, others may not so our
> development time may not be best suited for this task.
> 
> Mine wont catch second and with the 2.73 rear is hard to mess up a
> launch also. An algorithm can detect and flag or account for
> slippage.. maybe its a worthwhile compromise. Tire slippage would
> also affect RPM based readings, as well as ABS sensor based readings.
> VSS pulses would give rear wheel performance figures, another bonus
> to me. Hey if I can measure and tweak to eliminate rear wheelspin on
> a WOT takeoff.. I've implemented a sort of static traction control!
> 
> Okay so what pin / color wire for the VSS on a 165? whats the signal?
> +- 5v? wouldnt that be convenient.
> 
> > Date:          Thu, 19 Aug 1999 15:02:31 -0400
> > From:          Ken Kelly <kenkelly at lucent.com>
> > Organization:  Lucent Technologies
> > To:            gmecm at efi332.eng.ohio-state.edu
> > Subject:       Re: Software Dyno...
> > Reply-to:      gmecm at efi332.eng.ohio-state.edu
> 
> > Can't use VSS if you have any tire slippage. VSS will work
> > with small motors, but not with a fast car.
> >
> >               Ken
> >
> > Dave Zug wrote:
> > >
> > > If we replace ABS with VSS as the input then it may be more
> > > universal. I have no ABS so I'm SOL if this is the input.
> > >
> > > If your software can simply turn parallel input into a binary or text
> > > file of "times between pulses" or "pulses per second" values I can
> > > take it from there and do the UI in windoze. If you have a schematic
> > > I can test it with the scope.
> > >
> > > > The ABS is right behind the drivers seat and you can use the signal from
> > > > it to give you a reading on tire rotation.  The corvette has 43 pulses
> > > > per wheel turn which is fed to the ABS system at 250mv per pulse, I have
> > > > a preliminary schematic using radio shack parts to store and forward 4
> > > > channels through a parallel port but don't have the expertise in
> > > > electronics to put it together.  A busy computer engineer friend did the
> > > > circuit design but has too many  projects going on to help.  I have the
> > > > software in a draft using clipper if I could get the interface worked
> > > > out.  It could be used for any sensing device using the software to tell
> > > > it what the channel represents.
> > > > Anyone interested in helping make it happen?
> > > >
> > > > Tyler
> > > >
> > > >
> > > >
> > > >
> > > ~~~
> > > Dave Z. www.delanet.com/~tgp
> >
> >
> ~~~
> Dave Z. www.delanet.com/~tgp