Date:         Fri, 2 Feb 1996 16:21:28 CST6CDT
Sender:       Vanagon Mailing List <vanagon@vanagon.com>
From:         "Dan Houg" <fairwind@northernnet.com>
Subject:      Air Flow Meters.  -or-

The Dead Return....

This post will be nothing new, well maybe a little new, but several
people have restored the Dead with respect to air flow meters (AFM)

This past week has seen us in several mornings of -47F weather.  The
van started running lousy with intermittent cutting out.  Did the
cold weather 'fix' by adding several cans of fuel anti-freeze to no
avail.  Got worse.  Van jerked and shuddered, no power, ran smooth at
idle.  hmmm....... thinking worst case scenario, the ECU, I started
digging into it. outside.  evening temp of -27F.

Fuel Pressure.  29 w/ vac  34 without.  textbook
O2 sensor voltage.  flucuating when connected.  pegged rich when 
disconnected.
Exhaust odor.  no anti-freeze.

I was begining to think I needed an ECU.  Prepared to commence to 
think about removing it and resoldering everything.  hmmm..... it 
idles smooth....  but the cut outs aren't at a repeatable rpm, 
throttle setting or anything satisfying like that.  not typical AFM 
high-jinx.

I'm not one to quickly condem air flow meters, now.  Replaced one 
unnecessarily in my '85.  Had one go bad in the Toyota that i was 
able to clean up (thus saving $550).  But i took it off and checked 
it out.  

THe resistance from the wiped output jumped all over the place.  So 
it has on 3 others i've tested including a brand new one.  You'd 
think they would give a nice linear increase in resistance but that 
isn't the way they work out.  "Need to test them on a scope" I was 
told from a German specialty garage.

After severely gouging a hunk of near frozen skin off the underside
of my finger I pried the black cover off.  Cleaned up the carbon
track with a new pencil eraser and slide a business card under the
wiper arm to clean the contacts there.  LOTS of gunk came off the
wiper points. Resistance still jumps around.   Re-installed the meter.

Lacking a scope but having in possesion a new digital multimeter 
(gift from Barb, she understands me) i hooked up the voltmeter leads 
to the wiped output lead which is #2 of the 4 pin AFM's.  Cover still 
off so i could see everything.  Turned the key on.  2.5v.  slowly 
moved the wiper upscale-- the voltage increased in a *very* linear 
fashion upto about 4.5v.  I set my meter to MIN/MAX which can catch 
transient spikes of 1 millisec or greater.  No spikes,  meaning no 
dropouts on the carbon track.  I felt i either had fixed the AFM or 
it was good to start with.  I didn't measure voltages prior to 
cleaning it up.

Started the van.  Idle smooth.  Revs nicely.  I start to laugh.  

While the cover was off, i wanted to see if i could lean out my
mixture when the running in open loop mode.  By just ever so
slightly pushing the wiper arm the readings off the O2 sensor swung
from rich to lean.  I kept repeating this like a kid wiggling a
loose tooth.  it was fun.  Now, for the really fun part.  I
increased the spring tension 3 clicks to lean out my idle.  it
worked.  got an open loop O2 sensor reading drop from .92v to about
.6 to .7v   Good enough, I'll leave it there.  NOW my mixture
control screw worked!!! Before i could turn that thing in or out all
the way with no effect.  I really wonder if *alot* of AFM's are set
with too much spring tension because many people have said the
mixture screw has no effect, including our dealer here in town.

Assemble the whole thing back together, noting posts of the Idiot 
Factor with respect to rags, wrenches, and what-not.  Temp now -31F.

Test drive reveals a wonderfully running engine!

I *think* it WAS the AFM but i'm not convinced.  It is the only 
'fixing' i did that night but i need more miles to be sure the 
problem is gone.  At least my van didn't burn up when i was done.  :)

-dan