ABS fails after driving for a while

[Mr. Toasty]

I have what seems to be an unusual ABS issue in my ‘87 944 S. I can’t find a reference to this specific issue anywhere on the great intergooglesnet. I have more diagnosis planned, but figured I’d get the thread going for anyone’s future reference.

The startup procedure is perfect: key on, warning lights on, engine starts, lights all turn off, drive 4+ mph and ABS pump runs for a few seconds. However, the pump continues to run multiple times for 1-5 seconds at a time while driving at 5-25 mph while the warning lights stay off. If I drive gently over a speed bump, the pump runs again but the warning lights remain off. Once I accelerate past 25 mph, the ABS and main warning lights turn on and stay on until the engine is off. Occasionally, the ABS will fail shortly after the car starts moving, but that is rare. The other day, I rolled around the neighborhood for a solid 7 minutes and the warning lights stayed off, but the ABS failed the instant I crossed the 25 mph mark.

I’ve cleaned all the sensors and tone rings, verified correct relay function (x3), cleaned all grounds, and tweaked sensor connections so the pins press harder against each other. No change. I plan to measure the resistance between pins on the ABS computer plug to verify continuity from sensors to computer and also basic sensor condition, and then measure AC voltage at the plug while driving to verify sensor function.

My theories:
-bad front wheel bearing (passenger front) is allowing the hub/tone ring to move away from the ABS speed sensor, especially when I hit a bump. Bearings are getting replaced now
-bad connection/wire somewhere
-bad wheel speed sensor

Place your bets and/or share your thoughts! I’m hoping it’s all down to that one bad wheel bearing or a connection because those sensors are way too expensive for what they are

[johnb]

I have what seems to be an unusual ABS issue in my ‘87 944 S. I can’t find a reference to this specific issue anywhere on the great intergooglesnet. I have more diagnosis planned, but figured I’d get the thread going for anyone’s future reference.

The startup procedure is perfect: key on, warning lights on, engine starts, lights all turn off, drive 4+ mph and ABS pump runs for a few seconds. However, the pump continues to run multiple times for 1-5 seconds at a time while driving at 5-25 mph while the warning lights stay off. If I drive gently over a speed bump, the pump runs again but the warning lights remain off. Once I accelerate past 25 mph, the ABS and main warning lights turn on and stay on until the engine is off. Occasionally, the ABS will fail shortly after the car starts moving, but that is rare. The other day, I rolled around the neighborhood for a solid 7 minutes and the warning lights stayed off, but the ABS failed the instant I crossed the 25 mph mark.

I’ve cleaned all the sensors and tone rings, verified correct relay function (x3), cleaned all grounds, and tweaked sensor connections so the pins press harder against each other. No change. I plan to measure the resistance between pins on the ABS computer plug to verify continuity from sensors to computer and also basic sensor condition, and then measure AC voltage at the plug while driving to verify sensor function.

My theories:
-bad front wheel bearing (passenger front) is allowing the hub/tone ring to move away from the ABS speed sensor, especially when I hit a bump. Bearings are getting replaced now
-bad connection/wire somewhere
-bad wheel speed sensor

Place your bets and/or share your thoughts! I’m hoping it’s all down to that one bad wheel bearing or a connection because those sensors are way too expensive for what they are

Interesting. I don't know of any situation where the pump would run without the antilock mechanisms engaging, except the self-test when you first start moving. So when the pump runs again between 5 and 25 mph, that sounds to me like it's resetting. Like maybe the computer lost power briefly due to a bad connection. But if that's all it is, then why does the ABS light come on above 25 mph? My best guess is there's something wrong with the computer.

[Mr. Toasty]

Interesting. I don't know of any situation where the pump would run without the antilock mechanisms engaging, except the self-test when you first start moving. So when the pump runs again between 5 and 25 mph, that sounds to me like it's resetting. Like maybe the computer lost power briefly due to a bad connection. But if that's all it is, then why does the ABS light come on above 25 mph? My best guess is there's something wrong with the computer.

Resetting and starting the self-test over is what I thought, too. So the system would not restart the self-test if it lost a sensor reading for a split second? Any way to test the computer, other than replace it with a known good one? I haven’t thrown parts at a car in years, and I’d hate to break that streak now lol

[johnb]

Interesting. I don't know of any situation where the pump would run without the antilock mechanisms engaging, except the self-test when you first start moving. So when the pump runs again between 5 and 25 mph, that sounds to me like it's resetting. Like maybe the computer lost power briefly due to a bad connection. But if that's all it is, then why does the ABS light come on above 25 mph? My best guess is there's something wrong with the computer.

Resetting and starting the self-test over is what I thought, too. So the system would not restart the self-test if it lost a sensor reading for a split second? Any way to test the computer, other than replace it with a known good one? I haven’t thrown parts at a car in years, and I’d hate to break that streak now lol

Honestly I don't know, it might reset in that case. That probably makes more sense because then a faulty sensor/wiring could explain both the resetting and the light coming on. The only definitive way I know of to test it would be with an oscilloscope. If there's anything you'd like me to check on my car (which is definitely working) let me know. It might take me a couple of days to get to it but I can verify when the self test is run etc. I'm not sure how I could simulate a brief sensor failure while rolling, but I could probably stop, pull the pin from the connector, move again and see if it runs the test again.

[Mr. Toasty]

Honestly I don't know, it might reset in that case. That probably makes more sense because then a faulty sensor/wiring could explain both the resetting and the light coming on. The only definitive way I know of to test it would be with an oscilloscope. If there's anything you'd like me to check on my car (which is definitely working) let me know. It might take me a couple of days to get to it but I can verify when the self test is run etc. I'm not sure how I could simulate a brief sensor failure while rolling, but I could probably stop, pull the pin from the connector, move again and see if it runs the test again.

Thanks for your help! I think I’ll proceed with new bearings, test drive, if no change, check connections, etc. It’ll take longer, but I’m too curious if the bearing might be the cause. If you want and have time, it would be great to know if it restarts the self-test if the sensor connection changes, and also if the sensor/tone ring gap changes

[Tom]

I'd be very tempted to open up the ABS computer and reflow all the solder joints, and maybe replace any electrolytic capacitors. Neither age well, and your symptoms sound like the illogical results you tend to see when ECU's go haywire.

[Mr. Toasty]

I'd be very tempted to open up the ABS computer and reflow all the solder joints, and maybe replace any electrolytic capacitors. Neither age well, and your symptoms sound like the illogical results you tend to see when ECU's go haywire.

Certainly sounds plausible. Soldering is one of my weakest skills, but I can at least take a close look. I’d rather work on my soldering skills than overpay for simple sensors lol

[johnb]

I'd be very tempted to open up the ABS computer and reflow all the solder joints, and maybe replace any electrolytic capacitors. Neither age well, and your symptoms sound like the illogical results you tend to see when ECU's go haywire.

Certainly sounds plausible. Soldering is one of my weakest skills, but I can at least take a close look. I’d rather work on my soldering skills than overpay for simple sensors lol

Hopefully it isn't covered in any of that compound they sometimes use on safety related electronics that make it impossible to work on. I've seen teardowns of airbag computers with that. IIRC I couldn't figure out how to remove the ABS computer from my car when I was getting my battery tray reparied. It's mounted with tamper proof screws, which I removed with a channel lock pliers, but then I couldn't figure out how to get the harness out. So I just hung it out of the way.

[Mr. Toasty]

Hopefully it isn't covered in any of that compound they sometimes use on safety related electronics that make it impossible to work on. I've seen teardowns of airbag computers with that. IIRC I couldn't figure out how to remove the ABS computer from my car when I was getting my battery tray reparied. It's mounted with tamper proof screws, which I removed with a channel lock pliers, but then I couldn't figure out how to get the harness out. So I just hung it out of the way.

Oh yay. Well hopefully I’ll have enough time to dig into it tomorrow, so fingers crossed! Bearings first since it’s half torn open, then further testing/diagnosis

[Mr. Toasty]

Well, seems like it was a bearing issue combined with a computer issue. After the new bearings went in, the ABS light stayed off longer than ever (~30 minutes) and at any speed (up to 70 mph). However, it came back on while driving over a series of smallish bumps. I guess the vibration was enough to shake loose a connection or two inside the computer. There’s no other explanation that makes sense, since I purposefully went over a few other single bumps to try and set it off. I’ll update to confirm the fix, but thanks for the input so far!