Injector Cycling

[garfield at pilgrimhouse.com]

On Wed, 3 Jun 98 14:27:52 -0400, "frederic.breitwieser"
<frederic.breitwieser at xephic.dynip.com> wrote:

>That's easy to fix Gar.  Run a junkyard in-tank fuel pump into a bucket of 
>some non-explosive liquid similar to gasoline (solvent?), feed a fuel rail, 
>and have the injectors on a metal plate hanging over the bucket.  By using the 
>real injectors rather than just a resistive load, you have a "real model" to 
>experiment with.  Unlike resistive loads, inductors create a backwash of 
>electricity when the internal coil collapses when the current is removed.  
>Maybe this is important, maybe not.  But its not too difficult to rig.

Yup, understood. I was just saying that you don't really need to fire
the actual injectors, to figure out to a pretty close approx. the fuel
delivery, but rather look at the scope & measure the pulses. Actually,
Bruce said he has a frequency counter, so he can use THAT to measure the
pulses even more precisely. That's SOP. Thus, the only diff. between
estimated and actual flow would be from turn-on/off inertial dynamics in
the injector, which you could "measure" at the end of your experiments,
instead of running with all this "apparatus" of battery and liquids
squirting every which way, all the time.

Remember, we were tryna suggest a way for Bruce to do his testing whilst
avoiding the hassle of a real battery & charger; the suggestion of using
a PC power supply was made, except that with 8 injectors and an 8-cyl
IGN going, it might not have enough current. (Actually, you wouldn't
need to run the IGN coil, either, which would save even MORE current,
cuz on the GM IGN's the signal that goes back to the ECM is from the
MODULE, not the coil; it (the coil) doesn't actually have to be there
for the IGN module to completely function 100% normally.

I mentioned the need for cooling at high rep rates not because it was
difficult to solve, but as an additional reason why for most of the
testing, looking at the injector driver waveform, instead of actually
running the injector, might be a better way to go.

Thas all.

P.S. I've run a setup like you describe to measure injector flow; man
what a mess! And try and find a solvent that's REALLY non-combustible,
nor gives off fumes. Especially for Bruce with his lungs on perpetual
overtime, I wouldn't wanna have to do THAT everytime I wanted to run my
"simulator". I'm not real sure, but I'd guess using water would be
asking for trouble.

Also, NOPE, the presence of the inductive noise from the injectors isn't
essential or even important to modelling the correct behavior of the
ECU, unless you were checking for noise immunity. B)

Plus, once you built one of these acoustically sensing injector cleaner
gizmos, with his scope, he could look at the injector driver waveform,
up against the acoustic sensor output, and get a REAL good idea of
inertial effects in his injectors, and probably be able to get a VERY
close correction/estimate of REAL fuel delivery, without squirting a
drop of solvent/fluid around anywheres. Woohee, whadaconcept!

Gar