Fuel mixture oscillation (was "Coasting and pulsating")

["Grimsrud, Knut S" <knut.s.grimsrud@xxxx>]

I've spent a little time examining the behavior of the fuel injection system
on our cars (see various back issues of the DeCO newsletters). The DeLorean
fuel/air mixture is controlled by the injection computer based on a signal
received from the lambda sensor in the exhaust. This forms a closed loop
control system.

By monitoring the mixture (using some of the circuits outlined in the back
issues) it appears to me that the mixture control scheme in the DeLorean is
not very stable and has a tendedncy to easily oscillate between rich and
lean. This oscillation is probably the same thing many cars suffer from when
they are cold and the idle surges until the engine warms up a bit. Although
it's not possible to tell from your description without doing a couple
measurements, a perceived change in the engine's performance could be partly
due to the ideal fuel/air mixture swinging between lean and rich due to the
efforts of the injection computer to "hunt" for the proper mixture. The
tendency for a closed loop feedback control system to oscillate is
excacerbated by a slow response in the feedback path (i.e. the oxygen sensor
responding slowly). I suspect that when the sensor is cold that it responds
more slowly, hence the tendency for idle to surge on some cars when the
engine is cold. I'm not sure what the failure mode of the sensors are when
they reach the end of their useful life (they should be replaced every 30K
miles as the sensors do have a finite life, which is why the lambda counter
and dash indicator are there), but if the degradation in sensor performance
near the end of their life impacts their responsiveness, then the injection
computer will have an even harder time regulating the mixture properly.

The closed loop mixture control system in our cars is actually pretty neat,
but unfortunately the performance is not as good as it could be if the
algorithms were better tuned to the particular application.

Knut