ECM fault?

[Akselrud, Boris]

Yesterday I finally completed fuel injection device I've been working on for
the past three weeks. It allows me to start the engine, cut out ECM from
injectors and substitute the device to provide consistent timing which can
be varied with high degree of accuracy for experimental purposes.
Well, hunting idle behavior is completely gone when I drive the injectors
with the device. There is a little bit of roughness left which I attribute
to ignition or other reasons.
I have yet to scope MAP sensor signal, Ignition signal and disconnected ECM
injector signals to see if they still vary despite constant fuel flow at
injectors and stable idle rpm.

This experiment proves that the engine can run steadily at idle and there is
no mechanical reasons which cause idle to fluctuate. But what would be a
permanent measure to fix it? One of the things I can do is to detect idle
condition myself, then modify my device to cut ECM control and substitute
itself to drive the injectors. Then, when Idle condition is gone, put ECM
back in control. But that's extreme.

Someone suggested playing with MAP sensor by putting a receiver tank (pipe)
and restrictions. Would it be the same as running MAP sensor signal through
a low frequency filter and then to ECM? Receiver tank is effectively a
mechanical low freq. filter.

Did I loose anyone on the subjects discussed here?

For those who not familiar with the engine, it's 93 5.7 TBI in Cadillac
Fleetwood.

Boris.