ABS-II Rant

[Danny caddyshack Noonan]

I bought this bike 2.5 years ago for far below blue as the owner thought it was messed up. Fixed the ABS faults by making foil shims for the front sensors. Very random faults since. Did the re-wire about a year ago so all my faults are roll-off.

 

Last summer I replaced all of the brake pads with a cheap ebay batch and did an end to end bleed. Not long after, I had ABS faults increasing in frequency. They continued until they were constant.

 

A month plus ago I did a hydro unit bleed and appeared to solve the problem with no faults for a week. Then over the course of a couple days, back to constant faults. This is where I started going all engineer on this thing.

 

I replaced the front pads with beemerboneyard and no change (not surprised). Checked and adjusted gap in front.

 

Couple of weeks ago I pulled apart the top end of the hydro unit for a looksee per this. ABSII Repair Nothing obviously wrong but, I did not pull the front circuit piston while I was in there. This isn't very difficult BTW. It seems a bit daunting because you're in the middle of the ABS system. You just have to make, and keep, the workspace clean.

 

Last week I pulled the rear sensor to inspect the ring and re-gapped it. Instantly the faults cleared for two test cycles. Then back to all bad.

 

Replaced the rear pads and rotor (rotor was shot due to a rear only braker). Initially I got no faults and than all faults.

 

Today I pulled apart the front side of the ABS and inspected. Nothing obviously wrong...seals looked good and no real wear under magnification. The front circuit piston was wet on the back side of the seal though. Not sure whether that is bad or not though. Getting to the relay is almost more work than the hydraulic side of things. Relay doesn't show to be the culprit on logic check. 7 dips.

 

Did a few checks and no joy. Still all faults.

 

So, this bike has had enough fluid pushed through it that its good for a colonscopy and it is still jacked up. Must be some polyps.

 

I think I can say for close to a certainty that some things may not be easily, or even difficultly, fixed. I'm pretty sure now that the no-fault roll-offs were just random stuff.

 

Time to get a ebay hydro.

 

rant over.

[Boffin]

I would do a resistance check on the sensors whilst bending/twisting the wires. You may just have an intermittent open circuit in a wheel sensor harness - not uncommon in wires that are constantly flexed. The copper work-hardens then cracks, with the ends touching some times and not others. Disturbing the sensors to re-gap them bends the wires bringing the ends into contact for a few miles.

 

Andy

[dirtrider]

Morning Peter

 

It sounds like you are slowly eliminating things.

 

A couple of questions-

 

What stored failure codes are you getting when testing after a roll-off malfunction?

 

Have you looked CLOSELY at the (entire) wheel tone rings for any signs of damage? Just one bent or damaged ring petal can cause a roll off issue.

 

Any looseness in either wheel like a loose wheel bearing or failed rear crown bearing?

 

Definitely do a wheel sensor resistance test as well as a rolling wheel A/C voltage test.

 

 

 

[Danny caddyshack Noonan]

Andy and Dirtrider,

Thanks for the responses.

 

The stored code shows 7 spikes and dips on the analog meter. Hydro.

 

Interesting thing about the wheel sensors and continuity for troubleshooting. The ABS would not either reset or initialize (I forget) with an open on the rear sensor circuit. I unhooked the rear and couldn't get a frequency or AC on the DMM I have. This could always be a scale issue (20k scale on Hz and 200mA AC). I unhooked the rear, fired it up and used the gearbox in first gear to roll the rear tire and tried to sense the pins (to get the Hz up higher than hand spin). That's when I noticed the continuity requirement of the ABS controller.

Those pins are very very small gauge and tough to find with probes because they are submerged in the socket.

 

Splicing into the wires themselves is undesirable since it shows one line to be shielded and, therefore, a PITA. This kind of goes with low strength signal on those sensors.

 

Ring in front is solid. Ring in rear has about 30 percent of one tooth missing on the ID side. The rear was dirty so I cleaned it pretty well when doing the rotor. Rear ring runout is near .003 inch (max) based on feeler gauges. The min and max are within the tolerance band.

 

No movement that I could sense in the rear wheel for crown bearing, front feels good too.

 

I've got today to play with it more so I'll try again on the rear sensor. The scale on the DMM goes down to 2mA for DC so, might try that too.

 

It's always possible I wasn't even touching the pins. May have to make some pin out adapters to make sure.

[NonComp]

Ring in rear has about 30 percent of one tooth missing on the ID side.

As the Myth Busters like to say, "there's your problem."

[cruisin]

Ring in rear has about 30 percent of one tooth missing on the ID side.

As the Myth Busters like to say, "there's your problem."

 

I still have both rings from the '98 I am parting out. Click the link in my signature lines or PM me if you want to discuss.

[Danny caddyshack Noonan]

Ring in rear has about 30 percent of one tooth missing on the ID side.

As the Myth Busters like to say, "there's your problem."

 

I would agree if the failure mode was different. That defect will not give a progressively more frequent failure. It would either be 1) random or 2) constant bad or good. I did some head scratching on this when I found it.

 

Clarification, the ABS will not initialize when a sensor is open. It needs to see the resistance on power up, in this case with the ABS delay relay installed. In that instance, fault code 4 results (rear is just much easier to get to than front).

 

Still no joy on getting either a current or voltage out of the rear sensor on spinning the rear.

[NonComp]

Still no joy on getting either a current or voltage out of the rear sensor on spinning the rear.

You've invested a large amount of time and effort into diagnosing this problem. If it was me, I'd fix the obvious. Replace the rear ring and the rear sensor and wiring to the rear wheel.

[dirtrider]

Afternoon Peter

 

That rear tone ring damage is not good but probably not your Code7 problem. A code 7 is usually internal to the ABS controller.

 

You can play with the sensors if you like but with no wheel sensor code set they are probably working OK. Those sensors are really reluctors so you can measure their static (disconnected) resistance but unless your have a meter with a very high impedance maybe not the voltage output.

 

If your meter is loading the circuit about your only means of measurement is to make up a jumper to allow the wheel sensor to be connected in the ABS controller circuit but still allow meter probe hookup.

 

If you can still get your ABS controller to re-set and shut the failure lights off you might try riding (cautiously) on a gravel or sandy road and doing a number of back to back ABS stops. That will exercise the pistons and shaft motor as well as purge the unit. Kind of a long shot but I had an 1100 years ago that had ABS issues with a reoccurring code 7 that eventually went away after some spirited ABS stops. I bought the bike with the ABS lights on knowing there was a problem there.

Do be careful as when I was forcing ABS operation I had a couple of ABS failures “during” the ABS event that allowed wheel lockup. Just be ready to manually release the brakes if wheel lock up occurs.

 

 

[NonComp]

Dirtrider, in the PC world, it's not unusal to get an erroneous error code (largely because the current error was not specifically identified by the designer). Does the ABS II system have a catch-all or default error code?

[dirtrider]

Dirtrider, in the PC world, it's not unusal to get an erroneous error code (largely because the current error was not specifically identified by the designer). Does the ABS II system have a catch-all or default error code?

 

Evening NonComp

 

Not that I know of but I only have the ability to read the stored error codes that set in memory.

 

The BMW ABS-2 system is fairly antique, slower than the dickens, and lacks much in the way of self diagnostics. As far as I have ever been able to tell it is only set up to warn about things that actually hinder the basic ABS function like front and rear wheel speed, front and rear pressures, front and rear sensor gaps, obviously battery voltage, motor relay, the basic controller function itself.