Trying to diagnose a fault by long distance so I don't have a wasted 600 mile round trip (as the car is in France) and need someone with a 1999 or later car to do a little test for me. The car in question is a 1999 P38 diesel and the fault first appeared autumn last year. When turning the ignition on, only the SRS light came on and no instruments, warning lights or the message centre worked. After the car had been driven for a couple of miles and the interior was starting to warm up, it all came to life and worked normally. At that time, having a sneaking suspicion in the back of my mind that the connector behind the RH kick panel had something to do with data to the instruments so checked that. Sure enough, a bit green and after giving it a spray of contact cleaner, the dash all came back and worked as it should again. For a few days anyway then it started to do it again and was ignored for a while. Over summer when it was nice and warm and dry it all worked fine but now it has stopped working again.

So, while passing a couple of weeks ago, I called in and chopped the connector out, soldering and heat shrinking all connections. Switched the ignition on and all I got was an SRS light..... Checked the connector at the cluster and gave that a squirt of contact cleaner too. Still the same. Consulted RAVE and found that other than connections to various other systems, all the data connections to the instruments come off one connector at the BeCM, C1280, a white 16 way connector. One of the easier ones to get to as it is under the fuse panel and nothing other than the flap has to be removed. It has a locking tab on the bottom that can be lifted with a screwdriver and the plug pulled out by its wires.

Unplugged that to see if it had any moisture in it and to give it a squirt of contact cleaner only to find that every single warning light on the dash came on. This was with the ignition off and the key not even in the ignition! Plugged it back in and all the lights went out. Turning the ignition on and we were back to SRS light only.

My theory at this point was that a wire or two had chaffed the insulation (or a mouse has been in there nibbling at it) and with some damp one of the wires has been getting a high resistance short to ground. But as it was late at night, dark, below freezing and dinner was being dished up, gave up at that point. As I had a ferry to catch the following morning I concluded that a bit more research and thought had to go into this.

So I acquired an instrument cluster from a 2000 car and did some tests. From the ETM, the instruments have permanent power on pins 4 and 14 with permanent ground on 7 and 17. My idea was to power it up and then ground various other connections one at a time to identify which one was causing all the lights to come on. So I powered it up from my bench supply with nothing other than these 4 pins connected and guess what......

Wondering if this is normal behaviour and is some sort of undocumented bulb check feature, I went outside earlier and tried it on both of my cars. With the plug at the BeCM unplugged, the dash just sits there and looks at me and doesn't light up every warning light.

Now both my cars are 1998 and the ETM is different for a 1999. The fault is on a 1999 and I'm using an instrument cluster from a 2000. As far as I was aware, a later instrument cluster can be fitted to an earlier car and has almost certainly been done by someone because they want the posh green faces even though the parts listing shows loads of different part numbers for the instrument cluster. So can someone with a 99 or later unplug C1280 with the ignition off and tell me if all the warning lights come on or not?

I'm determined to get this fixed as it has become a personal challenge but the owner needs to sell the car and a French registered, diesel, P38 will fetch 4-5,000 Euros in France whereas a French registered, diesel, P38 with an electrical problem is virtually worthless.

I tried this on both the 99 and the 02
Both petrol obviously......
In both cases, it took about 3 seconds for the dot on the temperature gauge, the low fuel dot on that gauge
And the suspension high symbol to come slowly up to about 30 percent brightness.
That was it. This was done with the key out of the ignition.
I should note that as I was in a hurry, I didn't turn on the lights inside the shed. It's possible the same thing
was happening to yours but there was too much ambient light for you to see the 3 dim lights? Perhaps trying it in
the dark would shed some light??!

Happy to perform any other tests.
TC

Thanks Tom, that's interesting. I didn't notice if the lights all came on immediately or slowly. As the car is LHD and I was poking about at the BeCM on the RHS, I only noticed them when the plug had been pulled for a few seconds at least. I've just tried another test with the cluster on the bench. 8 of the 16 wires at the connector at the BeCM carry data. Fortunately the BeCM SID also gives the pin outs for the instrument cluster so I could work out which ones. It uses serial data and there's a clock, a direction signal, data and a data ground. These are doubled up so there is a pair of each. If I ground the data ground pins and turn the power on, nothing happens, no lights, no nothing. If I then disconnect the data grounds with the power still on, the lights all come on, slowly lighting up gradually until all of them are on. So similar to what you are seeing although after about 10-15 seconds or so, they were all lit up.

I'll have another go on my car tomorrow and see if it lights up gradually. That is one thing I didn't do. I pulled the plug at the BeCM, looked at the dash, saw no lights so plugged it in again. If all else fails, I'm going to backprobe the connections at the cluster on mine with the ignition off and with it on and engine running. I've got a cheapo Chinese graphical multimeter that should allow me to see the data on the data lines.

An interesting extra is that the owner of the car has sent me a video. While driving, and before the dash wakes up and starts working, there is a whistle coming from the dash which sounds like it is coming from the speaker. My first thought was that the rpm signal from the engine ECU is somehow getting into the speaker driver. But at an assumed 2,500 rpm (and the diesel ECU supplies 3 pulses per revolution to the rev counter not 4 like the petrol), that would be a 125Hz signal, but it is much higher pitched than that.

I've a 2001 dhse, can do that test tomorrow for you.

Tried mine again this evening. Unplugged it at the BeCM, nothing on the dash. Waited a good 3 minutes and nothing changed, still no lights. So it does seem that pre and post 99 are slightly different.

Interesting.
As usual, when you have an inexplicable, intermittent issue with a P38, start with all the earths.
Of course, I don't have to give you that particular advice!, but possibly an earth point that a Bosch has that a Gems doesn't?
Also, pull the RH door sill plate and have a good look at the wires in question.
The first issue I ever had with mine was caused by the wiring loom being slammed in the door opening prior to the rubber door seals being installed. This was very obviously done on the assembly line.
All the idiot did was to wrap a second layer of tape over the 2 affected areas which were about 2ft apart and cut on the diagonal
Since the harness was hanging over the door sill.
There were 8 wires total that had the insulation nicked. Add a bit of moisture and they corroded and broke.
Main clue was the data lines from obd plug were broken in 2 places! Repairing the coax got the radio working as well!

My initial thought was earths but I would expect any iffy earth to be good when damp and not when dry and this fault only occurs when it is cold and damp. The fact that when I first looked at it, pulling the RH kick panel connector apart and reconnecting it cured the fault for a while, suggests I disturbed something in that area. It all seems to point towards the most likely cause being a fault in the wiring under the sill plate and damp is either putting a high resistance earth on a wire or two or it is putting a high resistance short between two (or more) of the wires. That would explain it starting to work again once the car interior has warmed up and it doesn't do it during the summer.

It may well have different earth points, even though the ETM doesn't show that much difference. But of course here I am looking at a third alternative, not Bosch, not GEMS but late diesel. Although I can't see them being that different, they used the same bodyshell so you would expect earth points to be in the same places.

Definitely have a good look under the sill plate. There's been several folks had issues with those looms over the years.

Luckily mine are ok (or seem to be), but I did find an old Motorola tracking system under the carpet & underlay after a heater core disaster many years back.

That's what I intend to do and if it was local would have done it already. It's just that the 600 mile round trip and Dover-Calais ferry make it expensive to go over and not find the cure.....