ABS light

[onrails]

ABS light comes on in a few meters of driving. All sensors meter out the same. About 900ohms. I've re-sanded all 8 grounds. Is there more in the fender that I need to be looking at? I do get the flutter in the peddle before the light goes on. check valve? Any ideas?

[onrails]

How do you post on the "what did you do to your Porsche today"

[onrails]

im a dork! LOL just updated y email

[924RACR]

That's when the pump motor and valves are cycled - same speed every time, I'm sure? I wanna say 20kph, a pretty common speed... but it's only a parameter, and can change for different projects (cars/OEMs)...

Check or replace the relays, is where I'd start, if you've already been over the grounds...

[onrails]

Are there more than one? The only one I know is in the fuse box. I've heard of others having relays around or on the pump in the fender. Maybe there?

[924RACR]

I seem to recall there are. Unfortunately I don't know that really old hardware; my area's the software, which hasn't changed nearly as much. (I work for Bosch in ABS/TCS/ESC)

[Lapkritis]

Start at the wheel and methodically work your way in. The gap might be too large between tooth and sensor. If there was recent work done near the wheel then you might have damaged teeth.

The flutter is the pump kicking on due to one wheel counting fewer teeth indicating a slip. If the controller supports OBD then you can monitor the input/count that is converted to wheel speed. On dry pavement the counts should all be the same. With a condition you describe there will be at least one lagger or 0 count.

[924RACR]

Sorry, not quite correct.

EDIT: however, I do agree with looking directly at wheel speed data, if available; I've used that successfully on GM trucks with bad WSS, that recall not so long ago...

Every ignition cycle on drive-off the pump, relays and valves go through what we call a BITE check. I forget what it stands for, but it's a self-diagnostic function of the ABS unit needed to ensure proper function. So if you have a light at standstill on key-up, it usually means a very basic electrical fault like power, ground, or connectivity to a wheel speed sensor (WSS).

OP is describing a symptom that very much sounds tied to BITE check at drive-off. BITE check speed is a fixed value, will occur at the same speed every time; only variance will be if the foot's on the brake or not. If BLS (brake light switch) is high (on), it'll occur at a higher speed most likely. Speed is determined by NVH considerations, usually.

Dynamic WSS faults, like plausibility etc. (as mentioned by the previous poster) will more likely take a bit longer to set than an error associated with the BITE check.

If you remove the relays - and maybe someone else here can point out where they all live on these cars - you can test 'em on the bench by applying +12VDC to the actuating coil and verifying that the other, controlled circuit does indeed close (using a ohmmeter/multimeter). I recommend the OP does this...

[Lapkritis]

Originally Posted by 924RACR

Sorry, not quite correct.

EDIT: however, I do agree with looking directly at wheel speed data, if available; I've used that successfully on GM trucks with bad WSS, that recall not so long ago...

Every ignition cycle on drive-off the pump, relays and valves go through what we call a BITE check. I forget what it stands for, but it's a self-diagnostic function of the ABS unit needed to ensure proper function. So if you have a light at standstill on key-up, it usually means a very basic electrical fault like power, ground, or connectivity to a wheel speed sensor (WSS).

OP is describing a symptom that very much sounds tied to BITE check at drive-off. BITE check speed is a fixed value, will occur at the same speed every time; only variance will be if the foot's on the brake or not. If BLS (brake light switch) is high (on), it'll occur at a higher speed most likely. Speed is determined by NVH considerations, usually.

Dynamic WSS faults, like plausibility etc. (as mentioned by the previous poster) will more likely take a bit longer to set than an error associated with the BITE check.

If you remove the relays - and maybe someone else here can point out where they all live on these cars - you can test 'em on the bench by applying +12VDC to the actuating coil and verifying that the other, controlled circuit does indeed close (using a ohmmeter/multimeter). I recommend the OP does this...

No need to apologize but maybe explain where. Where was I incorrect?

The abs indicator will illuminate if a sensor is in open loop with 0mph/0kmh. This is the layer 1 circuit throughput test. When you power the controller this runs this within the first few seconds. Go power a controller with a sensor unplugged and observe. You only need to key-on check.

The drive-off test is data acquisition on the second layer of the circuit. This is when the controller looks at wheel speed data that is returned on the circuit. If the count is off (broken teeth on the tone wheel or broken sensor that passes layer 1 testing) then the light will illuminate when driven.

[924RACR]

What you missed was the pump valve and relay check that occurs in drive-off and will typically be completed before a plausibility fault etc is set on a wheel speed sensor or the like (what you refer to as a drive-off test).

Your labeling of the second test as a drive-off test is not quite correct, as that signal monitoring is active at all times, not merely once on drive-off then forget. It's not a test that is run once per key cycle; like the circuit tests (open loops, or shorts, etc) those diagnostics are active at all times that power is on to the ECU. These are necessary of course to catch a failure that occurs while driving, to prevent inappropriate ABS control.

Generally your statements are correct, just incomplete. Hope that helps, cheers...

[Lapkritis]

Ah, I see what you're saying. Yeah, if something breaks during operation then the system notices and alerts. It doesn't wait for the next power cycle... That's fairly obvious if you thought I meant otherwise.

[924RACR]

Yeah, sorry - but it is true that the drive-off test of valves, pump and relays does only happen once per key cycle. If ABS is activated while driving, of course there's feedback controls and monitoring on those circuits as well to detect faults... but that's the only check that's not always running...

Cheers,

[Lapkritis]

We're describing the same functions. One from behind the steering wheel and one from behind the software. Perspective of driver perception is more simplified for people without C++ training.

[924RACR]

Yep - actually, it's simplified when you don't actually have access to the code... just wanted to all all the gory detail, since sometimes the lack of those details - which are usually unavailable - can lead to wasted effort/money...