I volunteered to take a look at a convertible XKR that had been sitting for a little while at our local classic car business. It’s in a pretty sorry state as it’s previously been standing for over a year due to a no-crank fault that the dealership apparently couldn’t fix.

It’s travelled around 115,000 miles, has a cracked windscreen, scruffy interior and several other minor problems. VIN is A43863, hence mid-to-late MY05.

Today, I mostly spent some more time on the no crank 2004 XKR, chasing an apparent open-circuit on the CAN.

The W/S manual 3rd edition says to lower steering wheel and extend to max, then "Using a thin plastic lever, remove major gauge module veneer panel".

So, I gently levered, and off came the veneer trim.
"Oh dear" - that trim panel is attached to a plastic instrument shroud (by metal tabs) which is secured to the Instrument Cluster by two screws hidden at the bottom, and two tongues/slots hiding at the top.

Ah well, the tabs will survive one re-attachment before they weaken and break off.

Onwards to the CAN issue and pinpoint tests. Across CAN at the data link connector, 120Ω - one terminator missing.
Disconnect ABS module and measure again - 120Ω so it's not that - check.
J-gate Illumination module CAN in - 120Ω so IC termination and connectivity good so far - check.
J-gate Illumination module CAN out - 120Ω - check
Linear switch in and out 120Ω - continuity still good so far so check.

Next stop, according to the 2003 EG, is the Adaptive Speed module. Oh - wait - there's only 5 wires on the plug so can't be CAN in and out there, despite what the EG shows - perhaps its just a tee, but the wire colours don't match.

The ABS module termination read 125Ω - a little high but OK for now, so I reconnected it.
So - where's the break?

When I first looked at the car, there was a module under the ECM marked 4R83-14B588-AC, which appears to be something to do with SRS. I can find no information in the documentation and it's not shown in the EG. It was water damaged, so I just removed it - SRS isn't on the CAN right?

So I dismantled its plug, found 2 sets of green and yellow wires, and measured across each pair. 120Ω across one, and 125Ω across the other. I plugged the module back inline and, sho' 'nuff, the CAN is now measuring 60Ω. FFS...

So I'm back out of the duff information rabbit-hole and back to more testing. Still can't connect to the ECM, and the message centre is giving, among others, 'DSCCM fault' and 'Engine system fault'.

Keep calm and carry on. I'm probably not going to get too much further until I can put some proper diagnostics together.

 

Having chanced my arm on a clone mongoose, I got a brief play with SDD (V130) on the no-crank XKR. Thankfully, unlike my cheap 'n cheesy ELM327 it communicated with the car and read the VIN:


So, something adrift with the K-line. Shouldn't the IP show up somewhere?

There's definitely a KTM fitted as someone* left it hanging out of the dash.Also got the expected P1260, an unwelcome P1638 and flight recorder data available. At least some progress

*The local dealership has had its grubby mitts on this car so heaven knows what 'improvements' have been added (think motorbike ashtray, submarine catflap)

 

That's a really interesting update (or maybe I'm just really sad and nerdy) - I don't suppose you have a scope you can put on the k line to see if it is doing anything?

If I were you I would add these updates to your original thread on this car, as I'm sure someone else in the future will have the same problem.

 

I'd previously repaired both seat modules (usual corroded print tracks on the pcb edges) as the seats were stuck in silly positions, and fixed the only fob, although the car doesn't respond to it

I do, although it's an analogue Kikusui and the car lives outside. I have found a pinpoint ISO checklist, so I'll run through that the next time I'm on site. If that doesn't throw anything up I'll scope it when we next get a dry Saturday.

One of the obstacles I keep hitting is that there are a few deviations from the 3rd edition workshop manual and 2003 electrical guide, e.g. the airbag module is under the ECM in the false bulkhead and I can find no documentation.


Fingers crossed it can be made driveable so it will have some value. Otherwise it will be destined for the crusher as there are no scrapyards here.

 

.These are the 'related events' from the above session:



The LED on the J-gate flashes rapidly and dimly when the ignition is turned on. That may or may not mean something.

 

I think the fact that SDD can't communicate over the ISO network is a really good clue as to what is going on and now you just have the drudgery of finding out why.

If you do get the scope on it look at the signal with and without the Mongoose/SDD connected.

Don't ignore the L line while testing and remember, you can't make things worse at this point.

 

Hi, and thanks for the encouragement.

Sorry, the pinpoint tests also include the 'L' line - I was being lazy with the description, and there's always scope for me to mess up, lol.

I'll post up the tests once I've finished discovering/annotating where the various connectors & splices are actually located in this car.
The MGB is eating up most of my spare time at the moment, but I should get on-site during the week.

 

I had a brief session with the XKR this morning.



I noticed that there was some water around the OBD connector when I plugged in the patch lead (it had been raining recently). It appears that the door seal has a small leak above the end of the dash, which is unfortunate as the 'someone' had left the KTM wedged in the fascia fusebox. Another water-damaged module, sigh. Thankfully, there was a new one left in the car by whoever had been in there previously. Not the immediate cause of the problem, however.

I did some preliminary resistance & voltage pinpoint tests, and there's an issue with the 'L' line. It's showing close to (but not quite) B+, and the test is a fail if it exceeds 3 volts with respect to ground.

The battery then went flat as it had only had a short charge, apparently. Now on a CTEK.

The test says to 'repair the short circuit'. Where to start? I propose to disconnect each module in turn hanging off this line in the hope that it's one of them rather than a harness issue.

Stay tuned.

 

Well that doesn't sound right. You should get 12v if you test against B+, 0V against GND and the resistance should be in the 10s of K ohms to GND.

I've just had a look at the wiring diagram for the 4.2 cars and it is different to the 4.0 cars. It shows the L line is only connected to the Adaptive Damping Module and the Air Con Module. On the 4.0 cars it also goes to the ECM - whether this is a real difference or just in the documentation, I don't know, but yes, I would start by unplugging them in turn (and leave unplugged if possible until you have completed the test).

 

Agreed it may not be relevant to the main issue, although something's obviously wrong there. It's not a hard short, as there's about a volt difference from B+.

On a plus side, there were so many DTCs reported in the 'unrelated events' that I did a clear & re-read, and the P1638 hasn't reoccurred. I have a copy of the originals anyway in a previous SDD session.

 

Other commitments and the lousy weather have kept me away from the car, but I have revisited the pinpoint checks as it finally clicked that there was something odd about the text..

From the 2003-2006 workshop manual:

"
D4: CHECK THE ISO "L" LINE FOR SHORT CIRCUIT TO BATTERY (ADCM AND A/CCM ONLY)
Measure the voltage between pins 15 (W) and 16 (NW) of the diagnostic connector.

Is the voltage greater than 3 volts?

Yes

REPAIR the short circuit. For additional information, refer to the wiring diagrams. Test the system for normal operation.
"

The test for short to switched B+ is similar.

Now, a short or low-resistance path to B+ (pin 16 of the DLC is B+ always) would surely give zero or little above? I am seeing ~11.5V, which, according to the test, is a fail. The test would make more sense if pin 16 was at ground potential - as dibbit pointed out. Another rabbit hole.

So, I performed the same tests on my car, and got the same results:- ~11.5 volts between the two. It's not exactly like-for-like, but I would expect the results to be similar.
I took the opportunity to put a scope on all three network lines. CAN & SCP both showing activity & nothing on ISO - pretty much as expected. I also now have some voltages from my car to use as a sanity check.

I've made up an extension OBD lead that will allow me to tap into the dialogue between IDS/SDD which I'll try out on my car first.

 

Yep, I think the manual is wrong, but then maybe the 4.2 cars are different?

Seeing nothing at all on the scope for the ISO confirms a short or break somewhere, rather than a bad or noisy connection/module - I'm looking forward to the next instalment in this murder mystery challenge - what killed the car.

 

I took a break from the MG and paid another visit to the stricken car. Having been down another rabbit-hole with the ambiguous pinpoint tests, and armed with some measurements taken from mine, plus a read-up on the ISO9141 spec, in we go. I've used the Jaguar acronyms for brevity.

I did a K-line continuity check between the DLC and the KTM socket - open circuit, yet I saw continuity between the KTM and the RCM.

OK - i checked the DLC for a broken/loose pin, and I the orange (K-Line) wire didn't seem to be where it should be (pin #7). After finagling the socket out, I could see that 'someone' had decided to move that connection to pin #3*.

You can see the marks on the DLC body where 'someone' has opened the little flap on the side that retains the pins. I despair.



Orange wire really shouldn't be there...


Where it should be


So now we have K-line.




OK, missing the LHS headlamp, but more importantly, it's not reporting the KTM. I was running out of daylight by this time so I retreated to regroup and give it some more thought.

The LED on the J-gate flashes quickly with the ignition on. I'm sure it's trying to tell me something - perhaps because the car is immobilised?

Ah well, at least a step forward for now.


*Used to interrogate the airbag system on the older cars. Not connected on the 4.2L

 

Ive enjoyed reading this even though i havent a clue what most of it means. Hope you manage to resolve it.

 

Sounds like the sort of thing someone would do to an apprentice before telling him to fault find the car, lol.

 

I sure feel like the apprentice. I ought to fire myself as I would have found that much sooner if I'd been more methodical.

My only excuse, m'lud, was getting sidetracked by the dodgy pinpoints.

 

Michael . . . you champion . . . how you managed this without warm slippers, comfy club chair, cigar and glass of crisp pinot noir, is beyond belief.

I dips me lid to your perseverance, especially when navigating the pin point tests. I came away in a cold sweat when recalling some of my own experiences that involved missing or, worse still, misleading documentation. Your patience in negotiating rabbit holes was remarkable . . . I often resort to swearing and tears! As for the diabolical soul who transposed the K-Line from pin7 to pin3 . . . may you rot in hell! No apprentice deserves that time bomb a-tickin'.

For us mere mortals, and especially those still learning the arcane world of comms networks used on our Jaguars, your methodology, explanations and, above all, your displays and pics, are a great case study in logical analysis.

Really appreciated your journey and how you have presented it. I liked it so much that I have run out of :thanks; clicks. This rates as one of the best threads on such a poorly understood topic. You may just have saved this car (forever dubbed "Lazarus") from the scrappies!

Arise, Sir Michael. We salute you . . . (now that I've embarrassed the hell outa you)
Cheers and best wishes, mate,

 

Don't be too hard on yourself. I was asking ChatGPT the other day for the pinout of the OBD plug from the front side (the mirror image of the socket). It gave me the OBD socket pinout. I pointed out it should be mirrored, so it gave me the top and bottom rows swapped. Finally on the third attempt it gave me the mirror image. In other words you, Michaelh, are the future.

 

Thanks, guys, for the approbation

It seems to get worse towards the late models. The 4.2L is a lot more complex in so many areas, and the documentation for the end-of-runs cars like this one is scant and not entirely accurate (or it may be that I've just not found it). I often resort to swearing too..


Anyway, I had a spare hour or so yesterday, so I took the opportunity to do some checks around the KTM connector. B+ and ground, plus the two ignition-switched grounds all check out OK.
I cleaned both its and the LH headlamp connectors to no avail. Still neither respond to SDD. I was going to try IDS 125 but it wasn't playing ball yesterday and couldn't see the Mongoose

The replacement KTM is marked new, and shows no sign of tampering, yet doesn't respond. So, I'm at a hiatus while I determine whether the gremlins have got at it or there's a residual car fault.


K and L lines resistance to ground (battery disconnected) was 18.3KΩ and 49.5KΩ respectively. I need to double-check K as it's outside the ISO spec of 50KΩ for both.

K and L line voltages with respect to ground, ignition off, (nothing plugged in the OBD connector) are floating around zero - completely at odds with the pinpoint. These values tally with those taken from my car.

Ignition on, and the K-line rises to 3V, L-line still zero. On my car, both remain at zero.

Attaching either an ELM327 or the Mongoose pulls both ISO lines up to around 11 volts - logic '1' - (on my car also). See attached part of the ISO 9141 spec.

At least my ELM readers can now talk to the car .

I've now got a contact # for the owner, so I'll give him a call next week. It won't fix the car but I'd like to know the backstory.



Separately, the 4R83-14B588-AC is the 'Occupancy sensing control module' in Jaguar parlance. According to the EG, it's located A’ POST / RH SIDE OF FASCIA, and doesn't appear on the CAN interconnection network diagrams - neither of which is correct for this car.

Also, I read ('03-'06 Workshop Manual) that the rapid flashing on the J-gate is apparently a signal that the car is immobilised - no sh*t Sherlock...


I realise this thread is getting long and rambly. I'll do a synopsis which should be easier on the eyes if I ever get there.

 

I did type a longer reply but it seems to have been eaten by the forum software. My K and L lines show 18K and 26K respectively, so yours is strangely a lot more - whether that's important is anyone's guess. On mine both are 0v with ignition off, with only a tiny value when ign on.

We don't have a huge amount of information on the 4.2 cars, so we have to rely on the wiring diagram, which as you've found may or may not be accurate. I suppose they never bothered to revise it with the range coming towards the end of production, too busy with the new version of the car (and sourcing the interior from the Argos electronics catalogue).