Date: Mon, 6 Jan 1997 16:35:27 -0800
Sender: Vanagon Mailing List <vanagon@vanagon.com>
From: bayer@sybase.com (David Bayer)
Subject: oxygen sensor test
Hello-
While I was trying to track down another problem, I found some
behavior which seems a little odd based on my understanding of how the
ECU should respond to changes in voltage from the lambda sensor. This is
on a 87 syncro (your clamps may vary). First, my test...
With the ignition off, I unplugged the O2 sensor and hooked the
signal wire up to my multimeter, and the other lead of the multimeter
to my usual testing ground (the metal clamp on top of the air filter box
that secures the air box to the frame). Then I started the engine, and
read about .8 volts across the oxygen sensor, then I stuck a AA with the
positive side toward the ECU connection in line between another wire to my
ground and the signal wire to the ECU from the oxy sensor while the engine
was running to simulate a really rich mixture - no change after a few
minutes on my multimeter (still hooked across the oxy sensor itself). So
I put the ECU wire to directly to ground (false lean reading, should make
the engine get richer right?) - still .8 volts across the oxy sensor... So
I stop the engine, plug the oxy sensor back into the ECU but leave the
multimeter reading from the oxy sensor - ECU connection to ground... Start
the engine and watch the voltage increase to .5 volts - about right...
Now, I am thinking about this and I know the .5 reading was made
in a half arsed way cuz I now have the internal resistance of the O2
sensor (which is old and probably should be replaced) and the resistance
of the ECU making a little voltage divider, so maybe I should forget
about that reading...
At any rate, I am wondering if the ECU has some sort of programming
to filter out real rapid changes in O2 sensor voltage and thus ignore my
leaning test (the insertion of the battery) and the test straight to
ground or is this a sign that my ECU is shot? Bentley makes no mention of
testing voltages which makes me a little suspicious that the ECU might have
some sort of built in testing which causes it to run open loop (I checked
the throttle switch for continuity to make sure I wasn't causing the
ECU to run open loop inadvertently. I also checked all connections to the
ECU and they were within specs in Bentley. I let the engine run for a
couple minutes while testing the ECU for response to the O2 signal so
it should have been out of the start and warm up phases. Are they any other
conditions which might cause the ECU to run open loop that I have missed?).
Anyone know where I can borrow a properly working Digifant ECU to compare
against in the Bay Area?
dave
|