Correction to my own email: ground strap is on the LEFT side of the engine
(driver's side in the US) ;)

Jon

On Thu, 15 Feb 2007, jon wrote:

>
> Good work! As a true keyboard mechanic, I agree that it is electrical.
> Those grounds will get ya. Look at the ground strap on the right side of
> the engine, and go from there. It doesn't hurt to add some extra ground
> connections.
>
> One tip on the idle stab. valve: no need to disconnect the vacuum lines.
> Simply unplug the electrical connection to the ISV. No muss, no fuss, and
> no vacuum leaks.
>
> Good luck,
>
> Jon
>
> On Thu, 15 Feb 2007, Mike Frost wrote:
>
> > It's looking more and more like this is an electrical problem. I will be
> > looking at all the grounds related to the O2 and AFM systems.
> >
> > One thing I discovered was that you can't just pull the hose from the idle
> > stabilizer valve, plug it, and leave the end of the valve open. That
> > condition effectively creates a huge vacuum leak, and was the reason why I
> > couldn't get the idle low enough without killing the engine. Now when I go
> > to adjust the idle I pull the hose, put a surgical glove across the mouth of
> > the valve, and put the hose back on. This blocks the air flow nicely.
> >
> > I adjusted the AFM mixture screw and the idle ajdust on the throttle body
> > with the above-mentioned air flow block in place. I was able to set the idle
> > to 850 and found that the AFM mixture screw was 3 turns from fully CW. I
> > experimented with various settings of the AFM mixture screw but the results
> > were inconclusive. The readings from the O2 sensor were too erratic. So I
> > put the screw back to its original position, installed a monitoring cable so
> > I could watch the O2 sensor voltage while driving, and called it a day.
> >
> > This morning I drove to work with no problems, although I was surprised to
> > see the O2 sensor voltage oscillating between zero and some value between
> > 0.17VDC and 0.84VDC. It struck me as odd that the voltage should swing
> > rhythmically up and down like that when the engine is at a constant speed
> > with a constant load. If anyone has an O2 sensor voltage monitor installed,
> > you can tell me if that is normal.
> >
> > Later in the morning I went to drive over to the smog test and there was a
> > fairly heavy rain. When I first started out, the O2 sensor voltage was above
> > 3.0VDC! The van was running very poorly with hesitation and no power. After
> > a few minutes, the voltage reading dropped down into sub-one volt range
> > again and the van started running reasonably well again. This is when I
> > decided that I have some kind of electrical problem, one that is exacerbated
> > by humidity. I have always noticed a stuttering in wet weather, although not
> > nearly this bad.
> >
> > BTW, I still failed the smog test although the HC almost passed this time.
> > No significant change to the CO reading. All of this makes perfect sense if
> > the ECU is not able to understand what the O2 sensor is trying to tell it.
> >
> > Mike Frost
> > 86 Vanagon GL "WasserLeaker" aka "The Polluter"
> >
>