At 03:02 AM 5/20/2003, Mark Dorm wrote:
>david, you don't say what your opinion is of what and whether to do anything
>about this... should we move it from this oscillating position at idle,
>couldn't we have problems if we just arbitrarily move it around...

Oops, almost missed this one -- I'm passing over a lot of the messages
these days, so if you want me specifically it's a good idea to cc: me directly.

The oscillation is from some sort of resonant coupling between the intake
air column and the AFM vane, or else the pulsation in the air column is so
strong that it simply drags the vane along with it.  I suspect a resonance
*somewhere* because of the way it stops so rapidly as you begin to open the
throttle.  I'm sure it's not good for the AFM, but I have no idea whether
it's typical engine behavior or specific to the Vanagon engine.  If it's
not a case of "all four-bangers do that" then I suspect it could be
alleviated by tuning/damping the intake system -- but that's the sort of
thing that could have all sorts of unintended consequences at other
speeds.  Induction-system experts would know about that sort of thing.  I
don't think that changing the effective mass or damping of the vane would
be a good idea.

You can't change the angular position of the wiper without disturbing the
results that the AFM is passing back to the ECU -- but if you move the
ceramic substrate so that the new path of the wiper tracks straddles the
old one you can probably get more useful life from it.  I don't know how
much correlation there is between visible wear and noisy operation -- the
real test is to monitor the voltage between ECU ground and the AFM wiper
while moving the AFM vane with ignition on.  Should be a smooth rise from
one end to the other, from something over 0 to something under 5
volts.  The two ordinary failures would be either noisy (spiky) operation
or an actual dropout in one or more places.  The dropouts are easy enough
to determine, just move the vane very slowly and watch for your meter to
drop -- much easier with analog meter.  Noise is harder because even fast
meters won't respond to it very well. One thing to try would be to take a
DMM and set it to a low AC range.  Moving the wiper slowly and steadily,
say five seconds end-to-end, should give you a fairly constant small
reading -- if there are lots of spikes I'd expect the reading to rise
noticeably.  Good method is oscilloscope, but they're not in everyone's
pocket.  Ah -- if you've got something like a small tape recorder with a
microphone input, make a probe with shielded wire, run a ground wire from
the shield end to go to ECU ground, put a small capacitor -- say 0.1 to 0.5
microfarad in series in the line somewhere and maybe a 10k resistor
outboard of the cap -- I bet you would get a nice crackling sound on noisy
spots (with ignition on).  I'm thinking about an op-amp circuit that would
detect spikes and light up an LED, but just listening to the audio might be
a lot more informative.

Either the VW active-filter cable pigtail or the cheap "capacitor fix" will
considerably cut down the spikes that the ECU sees; that can be a good
answer to getting more life from a somewhat noisy one with no real flat spots.

david

--
David Beierl -- dbeierl@attglobal.net