ABS and related issue?

[BrendanM]

I have a 1996 series 2 XM 2 litre petrol injection estate with Bendix ABS.

The ABS warning light and message comes up intermittently. It is most likely to come up when I am moving very slowly in a traffic queue or when I brake to a standstill at traffic lights.

As usual the warnings goes out when the car is restarted. The brakes feel fine.
I measured the resistance of all four ABS sensors from the plug of the ECU and they were all OK. The front sensors are marked 'Siemens'.

I am going to clean the connectors under the wheel arch.

I have noticed another issue which may be related. Occasionally when the car has almost come to a stop under braking there is a very short 'jerk' for a fraction of a second.
It feels a bit like my car has been 'bumped' from behind.
I imagine it is probably either wear in a suspension bush or a momentary loss of braking pressure.
Could it be an automatic gearbox issue?

Are the two problems related?

[Sharo]

Hi Brendan,
check your Auto GB accelerator cable - this could be the reason for the bump. Probably it is downshift from 2nd to 1st.
As for the ABS - common fault is the relay box, which is placed on the ABS hydraulic body. You can check the errors using LED and resistor - and then sort out where is the problem.

[xmexclusive]

Hi Brendan

From your description the ABS is Bendix type.
Surprised to see this fitted to a 1996 car as only V6's were fitted with BENDIX ABS after ORGA 6741.
It seems that the ABS start-up routine completes correctly with the ABS light blinking and going out.
This proves the power supplies are correct, the safety relay functions and the resistance values of all 4 wheel sensors are good.
As has been suggested the next stage is to read the ABS faults from the ECU.
As Sharo says this can be done with a led and resistor if you do not have a Lexia.
Ask if you need a link into simple ECU reading.
Make a note of all the fault numbers read from the ECU then clear the ECU of all faults.
Run the car for a while then re-read the ECU to see what new faults are now recorded.
This approach is much more important than most people realise.
The ECU collects and stores faults over time.
It ranks them in importance, quantity and also combines them.
This is a set of safety routines that determine when to lock out the ABS.
If the ECU is not cleaned after repairing a fault it may still contain enough safety reports to be just one new one short of locking out the ABS.
As an example that fault might just be a vibration jolt on worn contacts in the ignition switch. This shows up as a loss of power to the ABS ECU.
In a cleared ECU this fault has to happen a large number of times before the ECU locks out.
By waiting to see the new faults and comparing them with the previous ones you get a much more accurate picture of the current faults with the ABS.
Can also save replacing expensive components that are actually not faulty.

John

This post has been edited by xmexclusive on October 14, 2011 04:01 pm

[BrendanM]

Thanks John and Sharo.

Please give me that link to the LED and resistor method for reading the ECU faults.

Many thanks

[BrendanM]

If its the 14 page pdf file dont worry, I've found it.

[BrendanM]

I used the blinking LED method to read the ABS fault codes. The results are;-

21 - Electrovalves Relais

25 - RH front wheel sensor
31 - RH Rear wheel sensor
32 - LH front wheel sensor

I have previously checked the resistance of the wheel sensors from the ECU plug and they were all 2.66 k Ohm. Of course there could be intermittent breakages inside the flexible cables.
But all 3? Thats an expensive repair. Could it be an ECU problem?

I noticed a crust of rust on the teeth of the right front reluctor ring which I rubbed off with a flat file.
I will clean the others if I can get to them.

My MOT has run out so I cant go for test drives now that I have cleared the stored errors.

I would like to investigate the Electrovalves error 21. Does anyone know the difference between code 21 and codes 15 and 22 on the instruction sheet which are also electrovalve errors?


P.S. I tried to read the Hydractive error codes on pin E2 but it wont respond to my switch.
Any ideas?

[BrendanM]

I found that code 21 is related to the 'hydraulic control block - front body side relay'.

My manual on an ebay CD suggests checking the resistance of the coils from the ECU connections/plug.

However the next page ( 456-0/2 [pdf page 915]) is confusing me.
It seems to suggest the resistance is either between 2 and 4 Ohms, or between 50 and 60 Ohms or infinity?

I'll try to make sense of it after a nights sleep.

[xmexclusive]

Hi Brendan

Code 21 is a fault with the ABS safety relay.
As Sharo has stated this relay is on a Green circuit board bolted to the top of the ABS block.
The ABS block is bolted on top of the front subframe under the battery box.
You can get to it by removing the plastic shields in the front wheel arch.
The relay is not a seperate item but part of the board components.
The relay board has a number of wiring connectors.
Checking, cleaning and refitting these may clear the faults.
The inability to run the car on the road may well make it difficult to check further.

John

This post has been edited by xmexclusive on October 21, 2011 02:01 pm

[xmexclusive]

Hi Brendan

Action for codes 15, 21 & 22 is listed as the same in the AutoData ABS book.
ECU Multiplug pins 8 & 19 resistance value should read between 50 and 60 Ohms.
If not as specified then disconnect modulator 5 pin plug and measure the resistance between pins 2 & 3 again looking for 50 to 60 Ohms.
If resistance now correct check wiring back to ECU.
If resistance still wrong suspect faulty ABS relay on green board.

John

[BrendanM]

I thought I would post the results of my tests so that others can make use of it.

After disconnecting the battery I unclipped the plug from the ABS ECU and tested almost everything wired to it.

The relay coil - pins 8 & 19 measure 60.9 Ohms which is probably ok.

I then measured between pin 3 (which is chassis) to pins 4,5,21,22 & 24.
This measured the Electrovalve coils via the relay contact and all read 2.6 Ohm which is good.

Pin 3 to 26 is shown as a single diode on the schematic and reads infinity Ohms. However on the diode test it reads 0.563 and 0.548v when I swopped the polarity of the meter probes.
This seems to indicate parallel back to back diodes or some strange fault!
Maybe it was repaired incorrectly? This needs further investigation.

Pin 4 to 5 & pin 21 to 22 (through two coils) both measure 5 Ohms which makes sense.

Measuring pin 3 to 20, 27 & 28 tests all chassis connections. Very low resistance measured.

I had previously measured the wheel sensors and they were all 2.6K Ohms but now the rear right hand side is open circuit! I will concentrate on that tomorrow.

[BrendanM]

Just thought of something regarding the back to back diodes issue in the Hydraulic block relay board.

There is a wire going from the relay board to the ABS warning bulb in the instrument cluster.
It has a diode in the circuit and that might be what my test meter is reading.

[BrendanM]

Back in October I got my car through the MOT!

I jacked up the rear of the car and eventually managed to disconnect the ABS sensor cable from the wiring loom. It is very difficult to reach into that space but at least I did not have to take the car apart.
The sensor tested OK for continuity even after stretching and pulling the flexible cable.

I wiped the plug pins even though they looked clean and sprayed them with 'Servisol' contact cleaner and lubricant.

Reconnecting them again in that inaccessible place was very awkward but it did clear the fault.

Many thanks for all your help.

[Jan-hendrik]

QUOTE

The sensor tested OK for continuity even after stretching and pulling the flexible cable.

I wiped the plug pins even though they looked clean and sprayed them with 'Servisol' contact cleaner and lubricant.

Reconnecting them again in that inaccessible place was very awkward but it did clear the fault.

Had the same problem with my '95 XM in 1995. Blamed it on an overdose of factory grease - the car was (almost) new.