SRR2
12-15-2005, 08:26 AM
I posted about this in the E30 forum and didn't get any response, but thought that someone else might profit from the problem resolution.
The problem was that the idle in my E30/M20 was so unstable that the car was undriveable with surging and stalling. Normal speed operation was fine, except that mileage had fallen off quite a bit, but power seemed normal -- probably ruling out a clogged cat. The ICV was clean and the intake rubber was good. I dumped a bottle of Techron in the last tank full without any effect. The next choice was the AFM, and this turned out to be right.
Now Bentley doesn't have much in the way of troubleshooting for this device. There is a spec for the air temp sensing thermistor, and that checked out OK. I put my ohmmeter across the two terminals of the AFM element and found that the resistance jumped all over the place (from ~50 ohms to over 2K) over the first ~15 degrees of flap movement. Above that it was smooth, rising from ~250 ohms to ~500 at full open. The jumpiness at small openings looked like the problem.
I didn't have a new AFM, and at this point there's nothing to lose by opening this one up, so I removed it from the engine and separated it from the airbox. The black plastic cover is held on with RTV that you can cut with a razor blade under the cover's lip. Then it pries off.
Using the probes of the multimeter directly on the element I found that the element itself looked reasonable, so the ceramic board was probably intact. Looking at the element with a low power microscope revealed that it appeared to be in good condition. So, the problem turned out to be contact resistance variation over the wiper travel.
NOTE: I recall seeing a web page somewhere where the guy mentioned a similar problem and solved it by nudging the circuit board a few thousands of an inch to allow the wiper to use a virgin part of the element. I DO NOT RECOMMEND THIS! In my AFM the screws that hold the circuit board in are bonded into the housing and will BREAK before they will loosen. I know this because, predictably, I broke one. Therefore, I highly recommend that you don't attempt to loosen the ceramic circuit board.
Anyway, I swabbed the element with a Q-tip and 95% isopropyl alcohol over its full length, and while it was wet, moved the flap back and forth 20-30 times. Before the alcohol dried, I swabbed it up with a clean Q-tip. There was noticeable dark material removed from the board.
Immediately the CRV dropped to about nothing. The resistance increased smoothly from ~20 to ~50 ohms over the first few degrees of travel, then jumped to about 200 and increased smoothly above that. I believe that the jump is deliberate, since there's a visible trace (of conductor?) across the element where the jump occurs. I believe that this is approximately the correct behavior for the AFM, but there aren't any specs available, so this is mostly educated guess. Regardless, it's behaving MUCH better than it was when I took it out.
At this point, all that was left was to clean the remaining RTV off the cover and re-glue it to the body. It's curing now and I'll have it reinstalled in the car by the end of the day. I'm also cleaning the ICV again, since I had it out. Also, I pulled off the negative battery lead so that the ECU would re-adapt to the 'new' AFM.
The problem was that the idle in my E30/M20 was so unstable that the car was undriveable with surging and stalling. Normal speed operation was fine, except that mileage had fallen off quite a bit, but power seemed normal -- probably ruling out a clogged cat. The ICV was clean and the intake rubber was good. I dumped a bottle of Techron in the last tank full without any effect. The next choice was the AFM, and this turned out to be right.
Now Bentley doesn't have much in the way of troubleshooting for this device. There is a spec for the air temp sensing thermistor, and that checked out OK. I put my ohmmeter across the two terminals of the AFM element and found that the resistance jumped all over the place (from ~50 ohms to over 2K) over the first ~15 degrees of flap movement. Above that it was smooth, rising from ~250 ohms to ~500 at full open. The jumpiness at small openings looked like the problem.
I didn't have a new AFM, and at this point there's nothing to lose by opening this one up, so I removed it from the engine and separated it from the airbox. The black plastic cover is held on with RTV that you can cut with a razor blade under the cover's lip. Then it pries off.
Using the probes of the multimeter directly on the element I found that the element itself looked reasonable, so the ceramic board was probably intact. Looking at the element with a low power microscope revealed that it appeared to be in good condition. So, the problem turned out to be contact resistance variation over the wiper travel.
NOTE: I recall seeing a web page somewhere where the guy mentioned a similar problem and solved it by nudging the circuit board a few thousands of an inch to allow the wiper to use a virgin part of the element. I DO NOT RECOMMEND THIS! In my AFM the screws that hold the circuit board in are bonded into the housing and will BREAK before they will loosen. I know this because, predictably, I broke one. Therefore, I highly recommend that you don't attempt to loosen the ceramic circuit board.
Anyway, I swabbed the element with a Q-tip and 95% isopropyl alcohol over its full length, and while it was wet, moved the flap back and forth 20-30 times. Before the alcohol dried, I swabbed it up with a clean Q-tip. There was noticeable dark material removed from the board.
Immediately the CRV dropped to about nothing. The resistance increased smoothly from ~20 to ~50 ohms over the first few degrees of travel, then jumped to about 200 and increased smoothly above that. I believe that the jump is deliberate, since there's a visible trace (of conductor?) across the element where the jump occurs. I believe that this is approximately the correct behavior for the AFM, but there aren't any specs available, so this is mostly educated guess. Regardless, it's behaving MUCH better than it was when I took it out.
At this point, all that was left was to clean the remaining RTV off the cover and re-glue it to the body. It's curing now and I'll have it reinstalled in the car by the end of the day. I'm also cleaning the ICV again, since I had it out. Also, I pulled off the negative battery lead so that the ECU would re-adapt to the 'new' AFM.