[3.2 Motronic] Air Flow Meter wiper arm re-tracked - WOW!
 

First off, a big shout out to Steve Wong, who suggested I look into this, rather than letting me send him money for a custom chip (that probably wouldn't have helped anyway).

I contacted Steve asking about something to reduce the very abrupt on-off throttle transitions my car experienced at low RPM, thinking that perhaps disabling or reconfiguring the decel fuel cut off might help. Steve said he could do that, but asked if I had checked the resistor track wear in my air flow meter? He said that was common cause of low RPM drivability problems.

I hadn't heard of this before, but it took very little searching to find out what he was referring to. I won't re-post it all here, as this writeup on the944.com does a great job of explaining it.

Over the weekend I removed my AFM and opened it up, and this is what I saw:

http://forums.pelicanparts.com/uploa...1396927270.jpg

Looked bad to me, so I did the re-index job (moved the arm down) so that the wiper contacts rub on fresh surface, and put it all back together and in the car.

Holy cow! What a difference! Virtually every aspect of low RPM running is improved. No hunting on cold start. Idle is smoother and more steady. No more harsh lurching during on/off throttle transitions in traffic. Easier to shift smoothly due to smoother, more consistent throttle response. Maybe more, it's so improved I can't remember all the ways.

I'm pumped up about this. This free fix has made the car I love to drive even better. Incredible.

Short version - anyone with a 3.2 that has more than 100k on the original AFM should probably take a look at this.

"Looked bad to me, so I did the re-index job (moved the arm down) so that the wiper contacts rub on fresh surface, and put it all back together and in the car."

Most are not as bad as they look. A little WD40 sprayed on the track usually resolves
any potential problems, i.e. allows for a smoother travel of the wiper. The problem
with repositioning the wiper is the tension of the wiper can cause more wear to the
track in one method and in the other method the wiper needs to be bent. Most all
AFMs last in excess of 200K miles when maintained properly not requiring any
modifications.

Also adding oil to the two pivot points of the flap improves its overall operation.

Interesting - going to look at this this weekend.

What looks bad, I dont see anything on the track?

Those "tracks" are wear marks.

When I bought my 88 with 80k, the car was shuddering and cutting out at 2k RPM.

A rebuilt AFM did the trick. I would have done the arm adjustment, but the prior owner paid for the rebuilt AFM as part of the deal.

ooooh the black stripes are supposed to be the same color as the rest of the gray area.

Interesting. I bet this will smooth out my cold starts as well.

I have that low speed on/off condition at part throttle as well. Timely thread. I'll be cracking my meter soon to see what is going on.

Exactly! Open even very low mileage AFMs and you'll see the same.
The AFMs were used for over 20yrs by VW, Porsche, Alfa, Audi,
and others with a very very low failure rate or functionality issues.

Interesting for sure. My car's AFM is 26 years old with >250k miles on it. It certainly makes me suspect it as the potential source for my crappy, hunting idle (after checking nearly everything else...)

It is not that there are wear marks per se, but that the areas that get the most arm friction begin to wear through to the point that connectivity is lost.

That is why cars will often idle OK, but seem to stumble at ~2000-3000 RPM range, and then do fine at high RPMs.

When running, that arm bounces around 2-3k more often than 4k plus -- at least the way most people drive. ;)

I wonder what the cost is to just have the whole thing cleaned up and a new board, needle, spring etc. installed?

I can't comment on the reliability of these units, this is the first one like this I've owned.

But what I can say is that - in my case - there must have been a problem, considering the vast improvement in drivability in my car, in numerous ways.

So I still believe that any 3.2 owners who are experiencing hard-to-cure low RPM drivability issues should take a look at this.

Rebuilt unit from Porsche is $1100 exchange from our host.

I've seen it at $375 elsewhere online (you can google it), with a $300 core charge -- which is about what the mechanic paid for mine.

$1100 is pretty steep.

Racing Mazda Miata's in SCCA we learned how to "clock" or index the AFM which will affect response on the dyno and the track. Essentially, you move the spring pointer a few notches in either direction and see how the car responds. Miata AFM's were easy to come by so we had multiples that had been "clocked" and had a code marked on the AFM to tell us which AFM we were dealing with.

I highly discourage purchasing any of the 'rebuilt' AFMs from various sources. Every car I've seen use these aftermarket 'rebuilt' AFM causes the car to run outrageously rich, and not even reclocking the internal spring tension corrects them. Instead, renewing your existing AFM what is much preferred to restore original function and calibration. I usually readjust them by moving the wiper upward on the pivot shaft, bringing in the wiper tip to sweep inward about a mm on a new arc. I also re-angle the copper wiper to retain the same contact pressure and angle as was originally made to prevent accelerated wear on the sensor. Do this by removing the four screws holding the harness plug on the housing, so you can remove the wiper assembly off the shaft. Bend the wiper on the inside slightly downward, and halfway out on the wiper, bend the wiper slightly upward maintaining the same angle and pressure back on the track.

On reassembly, sweep the arm towards both ends of the arc, both full open and closed and make sure the contact tip stays equally centered within the ends of track, and adjust if needed. A little light oil helps keep everything working smoothly.

Below are are some example of worn AFMs we renewed by repositioning. White is very obvious wear that will cause a complete signal dropout, signaling a closed throttle cutting off fuel and shifting the ignition timing load curves out somewhere unintended:



http://forums.pelicanparts.com/uploa...1396993604.jpg


http://forums.pelicanparts.com/uploa...1396993630.jpg


http://forums.pelicanparts.com/uploa...1396993668.jpg

FR Wilk, a poster in the 944 section of this board, way back in the stone age of 2000, did an excellent job of creating a site that explains how the AFM works, how to test, and, if necessary, refurbish your AFM. While the 944 AFM is a bit different from the 911, they all work the same. They even look quite similar. So for all of you DIY'ers, here is the link: Air Flow Meter (AFM) - from "The 944 Motronic DME" by FR Wilk* ©2001 www.the944.com

Same link I included in my original post. But it bears posting twice.

Did anyone do a bench test prior to opening up the AFM? My car had always run very smooth but I did do a bench test with a 9V battery and it tested out fine, linear voltage increase.

thats a good point. I am going to take mine out and remove the lid and look at the path. Then I will probably check it to see if its Ok prior to bending things.

Probably not a bad idea to bench test, but really, if you are experiencing less than great drivability in low RPM situations, what's it going to hurt to try this?

In my case, the proof is the results. Car drove bad. I re-indexed the AFM wiper arm. No other changes of any type. Car drives much better. I'm convinced.