Well, the problem has been there since day one: start the car up eg. in the morning and after you drive off it is clearly working ok, you can feel it switching from soft to hard automatically via the accelerator pedal response. After a minute - less than a mile you notice it is permanently in hard mode. Reading the eHydractive ECU fault codes the only one that ever comes up is number 24: vehicle speed sensor.
First investigations were sketchy: tried other speed sensors - no difference, cleared the fault codes - they came back after a mile, reasearch on the internet brought up nothing solid: I found some stuff from the old Andyspares chat rooms from the late 90s that outlined the same problem but it was a case of cleaning up connectors.
Then someone on here tipped me off that there was a thing called a 'speedometer interface' that was common failure on Peugeots, Xantias and XMs that share it (from an Aussie site) after I discovered the speed sensor connected to the system via this interface device. The interface is used on XMs that have the trip computer. Mine has this and it also didn't work: speed information reads zero or maybe a few mph at most. After John sent me a couple of wiring diagrams I found the speed sensor was connected to the speed interface and the speed interface has outputs that connect to the hydractive ECU, the trip computer and other devices.
The guy who'd suggested the speed interface originally said to check it so I looked at its 9 pin block connector. Loking at the XM wiring diagrams John and Chris sent me It was clear that there was a connection at pin number 1 of the 9 pin block of the interface that went to pin number 13 on the black connector of the hydractive ECU. I simply checked the continuity of this connection with a multi meter and it was open circuit. I then checked the rest of the pins on the 9 pin connector to find that the 12v supply was OK, the earth was ok, the connections to the speed sensor were good. Then likewise, the 12 v supply at pin 1 of the black ECU connector, the earth pin and the electro valve were all fine and I even revamped all the earth points anyay.
Yet still the speed sensor was the only fault code of the ECU.
I also was drawn to the accelerator pedal sensor because it was the only input that seemed to actively work the hydractive (that and the opening of the doors) I could have the car running, jab the throttle and hear the electrovalve click underneath. I found that constantly jabbing the throttle caused the same thing to happen as when driving, namely that the electro valve would click for several times as I caused it via the pedal movement but then suddenly stopped, therefore going into hard mode. It looked like the pedal input was the cause so I removed the sensor and checked the resistance values: lever up, down , and in places inbetween - it was all right. I tried a second sensor and there was no difference: as usual it worked for the forst minute of driving then went hard mode.
I then found that by unplugging the pedal sensor that the suspension stayed in soft mode permanently! No matter how long I drove it. No inputs of any sort caused it to go stiff and selecting sport modemade no odds: it stayed soft. Plain logic now: with no input from the pedal, it never goes stiff as that is the only input that causes stiff mode when stationary (all other sensors need the road speed to be above 30km/h) and since the speed sensor was not workng , or at least getting its signal to the ECU then the ECU thought the car was stationary all the time and when an XM is stationary with the engine running it is in soft mode. The reason it was going to stiff mode suggested by John I think, was that with inputs from the pedal sensor but no speed information the ECU decides to go to safe mode - that's hard mode of course..
So in the interim I left the pedal sensor disconnected and ran it in perma soft mode all the time. Trouble was that soft mode on an XM is way too soft. (just as hard mode is too hard for constant use - this is Hydrative 1 mind)
Anyway, John and Chris also sent a few other diagrams that were Hydractive 2 and although nearly every pin of the ECU connectors has a different use, the sensors are seemingly all the same save for the speed sensor which changed but this was a red herring because I soon confirmed the speed sensor was working after a fault finding guide in one XM factory page that said the speed sensor should give an alternate + then - signal of a very small voltage on it's own. Well mine was doing that when I tested the sensor on the workbench. I reckoned the problem really was the speed interface as Dieselman suggested it merely boosted the small stepping signal from the sensor.
The connection between the interface and the ECU was certainly broken somewhere as I'd found so surely this was the cause? the ECU just wasn't getting speed signal from the interface but the interface was getting signal from the sensor.
Then the plot thickened bacause I won a set of XM factory/ service folders on ebay that were only a few miles away ( seller was an ex XM owner who'd paid a LOT of money for them back in the 90s) but the key thing was that they have a lot of revisions of the changes to the XM wiring. John had said he only had the manuals for the early XMs and the late ones but less of the middle models (like mine) but this set of manuals covered the cars from 1991 to 1997 as well and it showed they changed the wiring of the speed sensor from a route that went through no less than three different multi plug connectors to being wired directly from point to point, sensor to interface and interface to ECU. It meant that the casue of the break in the continuity was likely to be a wire break and not a duff or corroded connector.
Skip to the end..... so the manuals also showed the XMs without the speedo interface and they have the speed sensor connected directly to the ECU, again pin 13 but also to pin 6. Pin 6 isn't used on the car's with the interface so I guessed pin 6 must be either a V-out to the sensor to return via pin 13 or it was the earth for the sensor (both these jobs would be done by the speed interface unit when used) so I went about connecting the speed sensor directly to the ECU with an external pair of wires and lo an behold.... it worked!! I drove for miles on Wednesday afternoon with fully functioning suspension and now, 2 days later it is still working. The speed interface being bypassed doesn't come into it now so it further proves the problem is with the interface itself and/or the connection that is found to be broken.
Only query was where was the break in the wire? well it was open circuit like I said from the 9pin connector of the interface under the dashboard to the ECU pin itself. O webt about narrowing it down and found it buzzed out between the 9 pin connector (using a needle through the insulation) and mid way points to the ECU connector but was still open circuit at the ECU pin. I kept reducing the length of the test until I was actually testing both sides of the block connector: the pin and the wire on the other side and wound it was open circuit. So the break was actually inside the block connector - I've now fully opened the connector (quite difficult) and I accidently sliced through some wires on removal ( thankfully after I'd marked them all) but now all 12 wires are reoved, rewired with narrow connectors to each of their pins on the top of the ECU to reconnect everything. Weirdly, I found that the broken wire was still open circuit between the now cut end at the ECU and the 9 pin block at the interface..... it seems there is also bad connection in the interface block as well. This has to be sheer chance that the same pin on both ends of one connection was bad? I'd missed it first time bacause i'd jumped to the middle of the wire where I'd stripped the loom back and it coincidently traced to the break at the ECU plug.
Phew what a screed. Well I haven't had time to replace the wire from the 9 pin plug and the ECU but will do when I get access to the garage again... couple of weeks I'm told
But for now I'll run it with the speed interface bypassed until I can work on it properly.
If your XM is doing a similar thing, is it Hydractive 1 or 2? I think each type behaves differently when faulted out. If it's H1 then it's possibly the same fault. If you can get a fault code reader on the ECU to confirm what the stored fault code(s) are then it's a good start. If it's 24, then it's the speed sensor. but like me, it could actually be another device/ connector that's causing it.
At least connect a multi meter to the various pins of the black ECU plug to see that pin 1 and pin 11 get 12v supplied when the ignition is on. Then check pin 8 of the white plug has a good connection to earth. That's the starting point.
Didin't mean to write so much but it came flooding!