I've been trying to tune my '84 Westy, and reading about others' attempts to
do the same, for a few months now, and I've come up with a question I can't
answer. It regards the input of the various sensors to the ECU and the way
the ECU determines fuel/air mixture.

Let's say you have two identical, 22 year old Vanagons next to each other,
and you're trying to tune both. The voltage from the sensors is going to be
different for the same conditions, just because of age and manufacturing
variability. So at 950 rpm at 185 degrees F (or whatever) the voltage from
the AFM, the O2 sensor, the Temp II, the intake air temp sensor, and so
forth, are going to be different from van one to van two.

There's no way around this, if only because you can't get "correct"
resistance specs for the potentiometer in the AFM.

In practise, to set the FI mixture, mechanics put a sniffer up the exhaust
pipe and turn the AFM screw, etc., until the readings are within specs.

But this doesn't really tune the engine at all temps, RPMs, loads, etc. It
just sets it for one condition.

The sensors in the two vans are going to be putting out different voltages
for 4150 RPM, 190 degrees F water temp, 85 degrees F air temp, open
throttle, etc.  So the ECU will be sending a different mixture to the FI's,
FOR THE SAME CONDITIONS.

This leads me to believe that the ECU is in reality a relatively crude
device, with huge margins for error from the sensors. Otherwise, the vans
just wouldn't run. Or they would run at one set of condition, but not at
another.

This in turn leads to relatively inefficient operation, in terms of engine
output and gas mileage.

The variability of the sensors's output, take together, also makes these
cars exceedingly difficult to tune accurately. And this isn't even taking
into account the ignition!

Am I on track with my thinking here? Is there something I'm missing about
the fuel management system?

Thanks for reading

Geza