Thread: ABS sensor replacement - Long Lead

Hey guys. Having just completed a rather stressful ABS sensor change I figured it may be of some use to others. First up, I have probably made some assumptions about the working of the system, or causes and effects that may not be correct, so please feel free to add to this as it will only help the next guy.

I have had the Amigos showing up intermittently for about a year now. The fault was 'low output front right sensor'. My assumption has been that the signal was not consistently getting through to the SLABS ECU. Up until now, a quick removal of the sensor, clean of the grease on the end of it, reset with Nanocom, and all was well for a couple of months. This was bugging me though and so, over the course of the year, I had purchased a replacement short sensor lead, and also, a long sensor lead.

For any one new to this, the D2 has an ABS sensor located in the hub, the cable for which, runs through the wheel arch and up to a plug in the engine bay.


Here you can see the long end of the plug disconnected. This is the lead that runs back to the ECU. The female part of the plug is on the short lead to the sensor.

This plug simply serves as easy removal point for the sensor plus 1.5 meters of cable, as opposed to removing all the rest of the cable back to the ECU. The SLABS ECU being located in the passenger footwell.

The down side of the above arrangement is that you have two potential sources of fault. The short cable from plug to sensor, and the rest from socket back to ECU. It is well documented that people have spent years chasing faults in this system.

Anyhow, a couple of months ago, I tried replacing the sensor with a new short lead. Issues with this. While the factory lead was resulting in occasional faults, the new short lead showed no output at all.

When you hook up a Nanocom to the system and jump into the SLABS inputs, (ie the signals it receives from various sensors) you can see two values for the ABS sensors.

One is voltage, so you would assume there is a current being passed through the sensors. 2.21v.


The other is wheel speed. If you have the car jacked up, the Nanocom displaying the wheel speed page, and spin a particular wheel, you will see the wheel speed reading change. My assumption was that there had to be a connection for volts, and if this was the case then you would think that the measurement of the wheel speed would be no issue. For some reason the wheel speed at rest is 1.7kmph.

So back to the new short lead - volts there, but no change of reading for wheel speed. WTF? In this case the original sensor was better!

A leap of logic later, and I have a new long lead, but nervous as hell because using this to replace the original sensor, means having to run the lead all the way through the car to the SLABS ECU, then cutting the (correct) wires for the front right wheel, and hooking the new lead into the loom. I was worried about all sorts of things with this job. The nature of the wires, and whether simply cutting and soldering was suitable for this situation, would this effect the delicate nature of the sensor signal and so on. Plenty of anxiety.

So, wanting to give my self as many backing out options as possible, I used the following process.

First up, confirm as much as I could about the problem, and possible solutions before the point of no return, (cutting).
Hooked up the Nanocom, so it was always reading the inputs from the sensor, and started by trying the replacement short lead again.
Wheel, and disc off so i can get to the sensor, undo the allen screw and pop out the original sensor, insert the replacement short sensor, and for speed, simply run the cable up past the outer guard into the engine bay and plug in there, (it will reach without following its normal route), and test by spinning the wheel. No speed change.


Here you can see the sensor in place. To remove the sensor, the allen screw is remnoved and the sensor pulls out othe hub. To test the other leads I just put the sensor in without running the lead through the hub, unlike this picture where the lead is properly run. There has been a heap of debate about whether you can just push the lead with its plugs and grommets through this gap shown. I never could so I had to undo the 4 bolts holding the hub to the swivel and pull it out a cm or two. You can see one of these bolts removed in the pic. Its not a big job, just 15mm socket and turning the steering to assist in getting on the bolts.

Next, I figured I could test the long lead in a similar way, by installing the sensor, and again with a temporary run through the drivers door, across the cabin to the passengers footwell, and simply insert the bare ends of the two wires into the back of the appropriate pins on the relevant plug. (See Rave, Disco 99-03 Electrical Library ,- Anti-Lock-Braking System. )


The ECU end of the long lead, as it come from Britpart.



OK, we got voltage, and wheel speed change. Brilliant! So assumption is the fault was in the rest of the cable, not the short lead. Most likely the plug connection between the two. Sorry about the crappy pic here, but I was taking the photo with one hand, and spinning the hub with the other.

Now though I have to face up to cutting into the wires. The type of wire in the replacement long leads does not look particularly special, so I have assumed that a solder job will be ok.


This is the plug in question. If you assume the securing clip on the left is the top of the plug, it numbers 1 in the top right corner then down the side, then top middle and so on. So the two wires we want for the front right wheel are 4 & 5, which are the two green wires in the middle row.


El snipo and we are ready to join up the lead and plug.


To run the lead properly I ran it through the firewall at the grommet, drivers side top, then used a thin rod to find a way through the extremely cluttered space behind the centre console, pulling the lead behind it, and across to the passenger footwell. Plenty of length left.



Solder it up, apparently it does not matter which way you join the two wires, ie no pos or neg etc. Finish and tidy, reset with the Nanocom, and done. Fingers crossed.

Update.

The aftermarket sensor was , in the end, around 1mm longer than original, hence the problem above. If i can find it anywhere, I will post photo for reference. Suffice to say, if you are changing sensors, get the current, and replacement along side, and CHECK THE LENGTHS.

So, after another trip through the interweb I had a new hub/bearing assembly, complete with short lead already installed. I looked for a long lead as it would mean just rerunning it back to the current solder join at the ECU, but was not able to find one anywhere near the price I had already uncovered.

The swap is obviously pretty straight forward from a process point of view, but I could not get the axle nut undone. So off to the mechanic, who swapped it out, and took the initiative to solder the new sensor lead into the one I had installed already. So two solder joins in line now, and its good as gold. Shows there is a bit of margin for error in the line.

cheers
Nick