BMW Forum - BimmerFest BMW Forums banner

ABS Trifecta lights- how I would approach

40K views 12 replies 5 participants last post by  rbelton  
#1 · (Edited)
As some of you may know, I have been contributing a fair amount of my personal experiences with BlueBee on how to approach the whole situation of ABS issues. Well, I just got the 3 lights of death back again (I call them the trifecta). History, replaced my module a little over a year ago and still had a bad RF sensor as determined by diode test. All has been well since replacing my module with new. Tried the re-build route and just lost time and got frustrated.

Car had been up on blocks for last 6 weeks as I did major rust repair and then sprayed the car. Got it back on the road again and travelled to and from work twice (about 100 miles total)and trifecta lights came on. WTF??!! New module (relatively) and I think car has maybe two new sensors RF and LR. Well, hooked up Carsoft and Carsoft says my LR speed sensor CABLE is bad. How it distinguished between the cable and the sensor I don't know. I will assume it saw the circuit as "open"- could be the opposite too and it saw a dead short. However, when sensor is bad, they will often read open both ways.

I checked with my VOM in diode position at the Module end of cable (connector). Tested between 13-29 iirc and showed 50 ohms in both directions. Checked a different speed sensor and it read a correct "open" in one direction and 1.707V in the other direction so I knew I really did have a sensor problem. I unplugged the sensor at the LR wheel and measured at the end of the speed sensor pigtail and guess what...it read the same...basically a "soft" short (50 ohms) in both directions at the sensor. I have one on order.This is in my estimation the correct way to try and separate the sensor from a module issue. Had that sensor read open in one direction and 1.7V in the other, I would be seeing if the module was still under warranty. Some have said they don't have a meter with a diode check position or that their meter gives different readings from my Fluke 85. In my opinion, it does not matter. Test all four sensors and it does not matter whether you use the resistance setting (use mega ohms setting) or a diode setting, don't sweat getting readings identical to mine.Look for one senor of the four that reads DIFFERENT from the others and then double and triple check that your paper clips or whatever you are using to insert into the connector are absolutely making good connection. Also, you must test each sensor twice. Once with positive lead of your meter and black or negative lead from you VOM oriented one direction and then reverse the meter leads to the same two pins. You want to check the sensor forward biased and reverse biased. If the readings are the same, your sensor is toast. Three of my four sensors checked fine. One was a bastard child! Hope this helps someone. Nothing here is new. You will find all this same information in BlueBee's comprehensive post. I just walked you through how my mind thinks on this issue.
 
#2 ·
Carsoft says my LR speed sensor CABLE is bad. How it distinguished between the cable and the sensor I don't know.
Great information!

I'm amazed also that Carsoft can tell the cable from the sensor. Especially as the codes thrown are not always reliable (I had a "bad pressure sensor" before I changed out my ABS and a P0500 "bad speed sensor" after I switched out my ABS, and both were bogus errors.

when sensor is bad, they will often read open both ways.
Good information. Especially since some people report strange readings which are hard to decipher.

measured at the end of the speed sensor pigtail
I always wondered WHERE to access the bare metal at the wheels (since my four sensors were good from the engine, I never needed to dig deeper).

Test all four sensors and it does not matter whether you use the resistance setting (use mega ohms setting) or a diode setting, don't sweat getting readings identical to mine.Look for one senor of the four that reads DIFFERENT from the others and then double and triple check that your paper clips or whatever you are using to insert into the connector are absolutely making good connection. Also, you must test each sensor twice. Once with positive lead of your meter and black or negative lead from you VOM oriented one direction and then reverse the meter leads to the same two pins. ... If the readings are the same, your sensor is toast.
This sounds like an excellent logical avenue, especially as people sometimes report they can't get any good readings from their sensor tests.
 
#4 · (Edited)
The explanation is simple if you think about it a minute. If the sensor is still connected to the module (any sensor...pressure, steer angle, speed, etc.) Carsoft or any of even the best readers like a GT-1 can not tell whether the short/open/failed diode/no pulse getting back et al. is before the module or in the input circuitry in the module itself. The only test that a carsoft reader can correctly diagnose that quickly comes to mind is a "communication fault" meaning that the Carsoft software recognizes that the module is unable to communicate with the buss that it is on.
Let's see if I can come-up with another example....Oh, ok... let's say you are listening to your stereo and the left rear speaker quits working. How would you know if the speaker is bad or the output for that channel is shot? You can't until you break the connection between the output terminals of your amp and the speaker. Then you can test the speaker itself. The Carsoft is upstream and would be like your ears. You just detect that a channel is not working, but as long as the speakers (read-speed sensor, pressure sensor, steer angle sensor, etc) are still connected to the amp you can't figure out which is to blame. I am not sure why Carsoft goes out on a limb and even tries to pinpoint whether the sensor is bad or the module. At best, it an only tell you which channel is not drawing any current, but it can't tell you why. Does that make sense? Dealer uses the GT-1 to just point them in the right direction and then they start throwing parts at it until they find the correct culprit. The stealers will charge you for everything they replace until they find it. A dealer will put the speed sensor back on the shelf and continuity test the wiring to the speed sensor. If good, they will replace the module.
 
#6 · (Edited)
Wish I had a dollar for every ABS/Brakes/DSC question

I started this thread for a couple reasons. If I had a dollar for every sole that posts something like "I have Brake/DSC/ABS lights all on. WTF??!!" I could retire early. Part of me feels sorry and recalls my newbie/Provie (NCIS inside joke) days. Then I get pissed and ask why do people like Bluebee and many others bother to do EXHAUSTIVE posts in tremendous detail in the first place as people are too lazy to read..so here is my summary of some of the more salient points in no random order. BTW, the new speed sensor cured my problem. That is about a 1 in 10 long shot BTW.
* It's fine to drive your car with the module out. You just won't have anti-lock brakes, stability control, speedometer, working odometer, etc. Car will drive as they did in the 70s
* No tester I know of will tell whether a speed sensor is bad or the module is bad. They will say the speed sensor is bad, but they simply can't tell for sure. None of them can. You must disconnect the sensor from the module and test with a VOM. A scanner- even the lowly Carsoft, a GT-1, Autologic, Peake, et al can't tell a bad speed sensor from a bad module. They can be right sometimes by luck, but if you understand electricity you will understand how it is impossible to test a sensor from a module unless your scanner plugs in BETWEEN them! Now, any of these scanners can detect a bad module (sometimes). Like if the scanner says there is a communication error then that is pretty conclusive that the module is bad as no bad sensor will cause a module to stop communicating.
* So, how do I tell a bad sensor from a bad module? How does BMW stealership do it them smart guy? Even the BMW scanner will only send them in the right direction. From there they just use experience or they throw parts at the problem till they get it. Example: Scanner says bad LR speed sensor as mine did. I disconnected the sensor (unplugged it) and then used diode test portion of VOM to test. Details are in bluebees post. READ IT!!
If I did not have a VOM what could I then do? Swap the LR with the RR sensor and re-scan. If sensor is bad, scanner will show now the RR sensor is bad. If the scanner still says it is the LR sensor, it is either the wiring leading to the sensor or the module. The VOM can eliminate the wiring, but it is a detailed procedure and not going into it here. If you do not have ASC or DSC, you may only have one rear sensor and one front sensor. You then can't do the swap test. Fronts are interchangeable with each other and rears interchange, but fronts don't interchange with rears and vice verse. Not sure why, but that is what I am told anyway.
*about 80-90% of the time your module is bad rather than a sensor. They are heat sensitive and the solder connections break with time and age. If your system works fine, you might consider fabricating a heat shield to keep exhaust manifold heat from getting to module. May help save years. Also remove and clean sensor and re-scan. Spray contact cleaner in that filthy connector box! To clarify, separate the connector halves and spray into each to clean actual contact points. Clean the box with a shop vac or similar.
*Bad module? Options are to get new or get re-built if money is a real concern. Not all modules can be re-built. MY RECOMENDATION is to use BBA as the re-builder of choice. This is my opinion only, but based upon experience. In my opinion, they do the best job for the money and C/S is outstanding. They will refund $$ if not rebuildable. A new module will set you back around $475-$500 plus it will need to be re-coded at a dealer or an INDY with very good tools such as an Autologic, GT-1, etc. None of the cheap (Carsoft) scanners will re-code a module.
* If you scan the system and get an error other than a speed sensor, it increases that 80-90% chance that it is a module. One exception is if you get "bad steer angle sensor". These steer angle sensors can come loose as they are just held on with a clip similar to a hose clamp. Look at your sensor and from the dirt you can usually tell if it has gotten whacked out of place.
* Last...NEVER,NEVER,NEVER trust a scan to determine a bad speed sensor from a bad module. They can get "lucky" and get it right, but it is far from a perfect world. Test the sensor disconnected from the module and you will be doing yourself a huge favor!

I would never have believed I had a bad speed sensor except that mine tested open in both directions when it should show open in one direction and about 1.7V in the other. Test all four and look for the outlier. If they all test the same reverse leads and test again. They should test different but all alike again. You can use the resistance (ohms) setting, but you are reading like 4-8 mega-ohms so only a decent meter can read this high accurately. Sorry for long rant/post, but this is one complicated issue to troubleshoot and so many people have it. Most toss way more money at problem than they need to. I helped one individual who trusted his scanner and replaced all four speed sensors before he finally gave up and got a new module. The chances of having four bad sensors is astronomical!
Good Luck.
 
#7 ·
No tester I know of will tell whether a speed sensor is bad or the module is bad. They will say the speed sensor is bad, but they simply can't tell for sure.
I can't count the number of times folks have 'trusted' their scanners - only to belatedly find out they were mislead (including myself).

Here's more information why the scanners fail to adequately test, even when/if the ABS control module is apparently still good!

Image
 
#8 ·
Thanks Bee for validating what I have been saying, but this is even ANOTHER way a scanner can be fooled in addition to some of the more obvious. Anyone who says this tester or that can get it right 100% of the time is just incorrect, but they will still take shots at what I have said. I don't know where you got this, but likely a sensor manufacturer who wants you to know WHY your scanner's solution was simply not correct and did not cure your error codesand that their sensor is not defective. I have read, and I'm sure everyone else has of sensors bought on Ebay that were reportedly "junk" and I have wondered if they all were bad or just didn't solve their problem that their scanner said it was. Sure there are swome of both. It is because of those nay-sayers who have tried to egg-me-on with "I am putting down the GT-1 or I just don't know what I'm talking about" that I have stopped trying to help anyone anymore on this topic. Also, because people don't read what you have assembled, but will just want someone to give them the answer that SOMETIMES can't be given by anyone or any scanner (I've never seen the autologic product miss even once because of its lightening fast scan rate, but suspect it can be fooled too). What I can say with certainty is that the auto-logic can clear errors that the GT-1 can't, AT TIMES! This presumes that the real error has been dealt with. Certain combinations of good modules with horribly out-of-range steer angle sensors with other legacy (historic) codes somehow fool the GT-1 at times and I've seen it personally. My only disclaimer is that possibly the makers of the GT-1 have resolved in later revisions the problem the GT-1 had, but to my knowledge the GT-1 software and firmware have not been supported by the makers in some years as the GT-1 has gone the way of Windows XP :-(

Bee, you're the best, but watch the nay-sayers come outta their caves. There is so much bad information in these forums that I just hit "delete". This is one of those topics. Note I have never said the GT-1 is garbage or that it doesn't work. I have just touted that which I know has the greatest success with all situations.

Thank you Bee for another of your many great contributions to this forum! You have the patience I have lost somewhere along the way.. Bill :thumbup:
 
#9 ·
See also:
- An ABS control module autopsy (1) & how to diagnose the BMW amber ABS BRAKE DSC/ASC trifecta or bifecta (1) & what are all known options when your ABS control module is bad (1) & Quick99Si's explanation of why the 10-minute wheel speed sensor diode-action quick test doesn't always work, especially on brand new non-OEM wheel speed sensors (1) (2) & explanations by 540iman as to why the fancy diagnostic tools very often (extremely often, like almost all the time) fail to properly diagnose the trifecta or bifecta (1) (2) (3) (4) (5) (6) & where to get the fancy schmancy diagnostic tools for free once you're aware of these limitations (1) & an ABS BRAKE DSC trifecta brake pressure sensor diagnostic DIY (1) & how to test and replace the steering angle sensor (1)
 
#11 ·
For the record, this was posted today, which, if not already explained here, may provide details on why the fancy schmancy tools often fail to properly test wheel speed sensors (in addition to being hooked up in the wrong place, as described by 540iman many times):
> E39 (1997 - 2003) > Need Guidance on testing Wheel Sensors with DIS
Image

See also:
- Quick99Si's saga testing non-OEM wheel speed sensors (1) (2) & explanations by 540iman as to why the fancy diagnostic tools very often (extremely often, like almost all the time) fail to properly diagnose the trifecta or bifecta (1) (2) (3) (4) (5) (6) & where to get the fancy schmancy diagnostic tools for free once you're aware of these limitations (1)
 
#12 · (Edited)
There is yet another good discussion today, over here, on whether or not INPA can give an accurate diagnostic reading when debugging the trifecta:
> E39 (1997 - 2003) > Trifecta lights + no speedo, again

As I said "in most cases". Once an error has been flagged up sometimes it is transient and sometimes it becomes permanent and needs to be reset using diagnostic equipment. Even if the "trifecta" goes away there will still be error codes set which can be read by INPA. This shows whether the error is present or not, if it is transient/non permanent, and how many times it has occurred since the error was last cleared. The last thing you need is the error codes resetting every time you remove the key. If this happened there would be no history of the errors and how many times they occurred. The lights may go out if the problem is transient, ie when the ABS module cools down, but the error codes still remain stored and can be read by INPA.

INPA also has something called "Shadow Memory" which seems to be a permanent record of errors occurring. I'm still trying to understand what this does.

Regards
RonR
INPA can't tell if the module is bad or not, that is a fact. You can't tell by the error codes either. When the module starts to go bad you usually get the trifecta when it's hot. You can confirm the diagnosis by turning off the car and cooling the module and restarting. If cooling (cold spray etc.) clears the trifecta it's the module. INPA is a powerfull tool, but it can't do everything. The trifecta lights are reset when you turn off the ignition, but the fault codes are stored. To reset codes you need to use a scanner.
So I took a couple videos showing how the heat affects my car, although I have probably restated it half a dozen times now.

This is in midday, 85-90 degrees F.
https://www.dropbox.com/s/19g40eqs2tc6daj/2014-05-12%2017.45.19.mov

This is at night around 65-70 degrees F.
https://www.dropbox.com/s/onzm78znu5m836a/2014-05-12 22.27.43.mov

This is why Im suspecting it is the heat causing the problems but I am not positive. If I were to restart my car in the heat, all the lights would come on.
Another set of videos showing what happens when the car warms up.

Before the lights come on:
https://www.dropbox.com/s/9rih5yb9ojjtpaj/2014-05-13 15.40.30.mov

After the lights come on:
https://www.dropbox.com/s/oenrtyk4iohstsf/2014-05-13 15.48.01.mov

BBA told me I need to try and have the codes reset, as they may be stored still. Trying to do that in California is difficult. Autozone won't do it, Jiffylube won't do it, I am basically left with going to a mechanic and paying $100 to have the codes read and reset.

These are the codes that were popping up before I had the abs module replaced:
Image

I replaced the rear left which is responsible for the speedometer not working.
I replaced the fuse which corresponds with code 14 (at least I think, from the research I have done)
I have replaced the front right with a cheap ebay one - I know, this is BAD, I plan on replacing it with a genuine one very soon. From my knowledge this doesn't affect the speedometer which is the main problem I want to fix.

BBA sent me this piece of paper, which is interesting to me:
Image


This goes back to bluebee's suggestion: reading the abs harness. I do not have access to a multimeter nor do I know anyone who does. The $100 that I would have to pay my local mechanic includes reading the codes and finding what the problem is, meaning they would check the signals from all the abs sensors too.