Daro we all sympathize. It must be incredibly frustrating to be able to start w/o any problem then instant off. Over and over. Give the considerable level of technical expertise (!!!)) in this forum, maybe just fly one of us up there? London looks like great place to visit. Airfare from Cleveland or Detroit isn't that much (500?). There's also the idea that someone in here might want to drive up there. I think the idea of repairing the original ECM is interesting. Unfortunately sounds like more reprogramming would be needed. Also curious about the corrosion on the OBD. Why? I'm sure you've already looked but any water leaks or moisture under the dash? Can the problems somehow be water related?

 

OK, sniping at what is past isn't constructive, and I do feel your frustration.


Diagnosis is essentially a process of eliminating what isn't causing a problem to narrow down the scope of the ongoing investigation. It's even harder when you can't just swop in a part to confirm. Hope this doesn't sound condescending.

You have been given assurances that the coilpacks, fuel pump and injectors themselves aren't at fault. I suspect this is correct (especially since you have confirmed fuel pressure is maintained) but, given that compression doesn't mysteriously disappear after two seconds, the abrupt stop must be due to either non-delivery of spark and/or fuel to the cylinders.

The ECM controls both of these, therefore it could be faulty. Or, it could be proceeding correctly from a mistaken premise because it's missing feedback signals from somewhere.

Does any of the diagnostic work you've had done explain what happens at the point where the motor dies yet is still rotating?

Is there still a spark and injector fire? Spark can be checked with a spark tester, and a noid lamp on pin 1* of any of the injectors will tell you if it's receiving a 'fire' signal.


I can't get near to my car for a few weeks or so (endless DIY), but I'll try to put some tests together that you can perform. It's a different MY, but the way it works is broadly the same.

It would be really helpful if you can enlist the help of someone with a 'scope.



*According to the EG, it's the black/not red wire on the injector plug. The black/red is B+ with ignition on.

 

Read a bunch of this and my head started to hurt. I'm sorry you are having so much trouble. This may sound a bit too simple but have you considered whether the ignition switch itself might be faulty? Will start but when the key is released to the run position the switch internals are bad?

 

I had a word with my friend Mike who builds race engines and programmes the ECUs to tune them and after going thro all the things you have eliminated suggested you go to a Ford specialist who works on fords of that era. This engine and ECU combination was used worldwide so the right specialist should have the "old system".
I re read post no 1 and you said it was re programmed ..?? Surely the brand new unit would have no VIN in it and needed your vin programmed into it ...from the windscreen ?? Sounds to me like the people who programmed the new unit did not get it right or corrupted it . There are certainly shops in the UK that do have the old programming tools.
I am making enquiries at a specialist near Stratford upon Avon who If I remember right has the old gear. Will let you know .

 

I had a reply from the guy in Stratford upon avon..He seems to know what he is talking about because he said on the X100 its not a VIN that goes into the new ECU but a security code ...I assume that comes from the old unit . I will send you his details tomorrow in a PM and you can give him a ring . Looks like the end may be in site thanks to DHL !

 

For those in the UK needing the old technology look here


Web is www.independentjaguarspecialists.co.uk

 

I want to thank everyone again who has jumped in here with suggestions, it is much appreciated and seems like a team effort.

I have not been on here or working on the car as I had a call back out of retirement for a 3-month stint at my old job, so I took it. After what I have spent on the car with no luck it seemed like a good opportunity to put some dollars back in the account.
I am not sure where I left off in the story so I will go back to after I got it back from Jaguar with nothing but a $4000 bill, I towed it off to Toronto to Berkshire Automotive, a Jag and British car specialist who was recommended by some folks on here. As soon as it got there, they called me and asked for $2000 to get started on it. They said if there was no rush, they would work on it as they had time. I said that was fine, still winter here. After about 6 weeks they called and said they had no luck even though they did put a used Key Transponder Module in it as it seemed to be related to that. They asked me to send more money to keep working on it. Once again, the same as Jaguar I paid $145 and hour for them to tell me what I already knew, fuel pump good, ignition good, injectors good. At this point I have $8000 into her and can't just keep pouring money into her so I floated her home, another $650.
So now she sits outside covered up and I have finished up the job so in a couple of weeks I will get into her again. I have been given the name of a local Jag guy who is willing to talk to me about it, so maybe I can bring a coffee in exchange for an idea.
It does continue to bother me that the Jaguar dealership would give up, but I do know that we older passionate Jaguar owners who remember when a Jag made you drool are not their target market. When I get into her in a couple of weeks,

I will keep everyone up to date.

 

Just throwing this out there... but back when I had the original jag motor in my car I distinctly remember that if I had the air intake removed from my car it would do the exact same thing... start for two seconds and then shut off... If I put the intake setup back on the throttle body and tightened it all up then the car would start and stay running. I don't know what caused this issue (MAF or something else) but it was your exact symptoms. If I still had that motor in the car I would go test it again, but I can't...

 

I'm not sure if it is applicable to this make/model so please grant me some elbow room.
Could possibly the alternator be causing interference resulting in the ECM getting scrambled & shutting down car. Again, possibly 2 secs before alt triggers into charge state.
Easy check, just disconnect alternator & give it a whirl.
Timing of 2 sec consistently of good running seems to be good at ruling out a lot, and should help diagnose.
A top notch auto electrical geek may be worth tracking down regardless of marque specialty.
Hoping you resolve this nightmare quickly & enjoy a great many years with your Jag.
Cheers

 

Strange you have it next to a British phone box in the photo above ..Maybe its giving you a hint to phone the the guy with the gear /experience and brain in the UK. Its cheap to talk on a phone. If you sent him the ECU he charges $90 per hour so $200 would easy fix it. You have the number.

 

This is the worst horror story I've ever read about on here. Goes to show what good some of the "experts" are - I mean the ones that were paid the thousands of bucks basically to tell you what you already knew. I really hope there is a happy ending. It has to be either loss of fuel or loss of spark, for reasons unknown at the present time.

 

I have been able to get back into my 2004 XK8 with the runs for 2 seconds and quits problem in the last couple of weeks. It was a good thing I got a call back to work out of retirement to top up my back account after the $8000.00 I have wasted since last fall. I have al of the responses here and appreciate all of the knowledge that is available in this group.

Some History and Updates.
1. After running well on a trip to Owen Sound and back, about 400 miles I parked the car in the garage like any other time.
2. Next morning, started the car to head to Niagara Falls. It coughed and ran terribly, with misfires and knocking. I shut it off and tried again, same thing.
3. Error code showed memory capacitors had failed in the ECM.
4. After checking on this group I saw that capacitor failure leading memory loss in the ECM was a common failure mode.
5. I sent my ECM to a rebuilder in Florida and he confirmed that the failure was the capacitors, but he could not repair it. $500.00 charge for nothing
5. I ordered a replacement ECM from an identical car that had just become available online from a 2004 wreck. $400.00
6. Replacement module would not work, as it needed to be reprogrammed to match my car info.
7. I sent the car to my local JLR dealer. They said they could not reprogram the used ECU and I would need the original back to download the car info.
8. JLR says I need a new ECU as they could not work with the used one. $2400.00
9 Jaguar retrieved the cars info from the old ECU and uploaded it to the new one.
10. Jaguar discovered some corrosion, on the OBDII connector and tried to clean it up. They could not get the car to run, just do the 2 second start and then shut down.
11. Jaguar contacted Jag in England but they had no solution. I retrieved the car with a total bill from Jag of $4200.00.

Update:
1. I could read the Systems with the Mongoose I got from Jaguar Diagnostics, and it showed some intermittent connections to different modules.
2. I put in a new OBDII connection, uncertain as to the effects of the corrosion and now I get a consistent report showing all the modules as OK except for the KTM which has a question mark.
3. Someone mentioned this earlier, and it is happening, I can do the 2 second start and shut down about 6 times and then it will not even start. I do hard reset with the battery terminal's, and I can get back the 2 second start.
4. I have read the Data which was uploaded to the new ECM and it is complete for my car, most have probably never seen this but it is correct with all of the options, VIN and car information. I know that Jag got that uploaded correctly.
5. When the car went to Toronto in the winter to Berkshire Automotive they saw the KTM fault readout and ordered a used KTM off teh web and put it in, with no effect. I got the car back in the winter and just parked it outside as I was working and knew I wouldn't have time to work on it. I didn't even look at the invoice, I just knew that they used the fist $2000 I sent them with no fix. I called to ask them which KTM was in the car and they told me the one they had gotten used, and they did not return my original KTM and when I asked for it he got real snarky and said his customers don't want the old parts in their cars. So the car does no have the original KTM in it but it does not seem to have made any change in the car. They also billed me 10 hours at $145.00 an hour to change the KTM.

 

the problem started when you took it to the dealer that had no experience with these cars. save old VID upload old VID into used module run immo done have a nice day.

they started with a new blank module 99% chance they programmed it with banned tools (bosch vci) and it never worked right

 

The JLR dealer downloaded the VID block from the old module and then uploaded it to the new module, I have read through it line by line and all of the info like options and VIN number and all of those items for 2 pages match my car.
They I had to wait a few days while they called JLR in England to seek assistance, but they were no help.

 

You should DHL it to the guy near stratford upon avon I gave you the address of and let him re programme it ..

 

I am back into this problem again after finally getting the Mongoose cable to work, it shows an orange question mark at the KTM, Key Transponder module.
I also have the CAR iSOFT diagnostic tool for Jaguar Land Rover and it tells me it cannot communicate with the KTM.
Berkshire Automotive told me they changed the KTM module and it made no difference. Still the same stat and run for 2 seconds.
I am now going to try and check the wiring to the KTM as that is what the iSoft diagnostic indicates.
My question has anyone gotten into that module which is supposed to be behind the cluster up in the front corner.
I have not been able to find a way to remove the air duct to allow access to that area and I don't want to break anything. Has anyone done this that can offer some advice.
I have included in an attached photo with an arrow marked #1 a module which I cannot gete out past the air duct number 2. I wonder if that module id the KTM.

 

If that's your last SDD display, I'd be more concerned about the X on the BPM. I recollect using my cellphone oscilloscope to watch the BPM communicate back and forth with different seat modules and whatnot to convince myself it was operating correctly.

I'm trying to wrap my head around the issue that you get the KTM, BPM & ECU to agree to start the vehicle and then a renegade command cuts it off. I couldn't find a fault the shuts the engine down; the worst ones would just go limp mode.

You've said it acts like the ignition just switches right off. This goes back to old testing, but this would eliminate a fuel related problem as the culprit. Can you stick an LED in with the fuel pump relay to visualize electric feed? I'd want to see if the LED dies the same instant the engine quits. Likely not as the fuel rail would have enough pressure to run on a bit. ECU controls the fuel pump once it gets past the security protocol and starts. One other question I think we answered is whether the fuel pump gets the 2 second prime before security, but the LED could indicate if the primer fuel is the only fuel pressure provided to last 2 seconds (again, likely not the case) or is it pressurized when the key turns to start.

 

At Jaguar and then Berkshire Automotive the first thing they checked was the fuel pump and said it was good, I have also jumped the relay to keep it running and the car still stops.
The number of tries is not consistent, sometimes 7 sometimes 10 but then it will not turn over and I do a hard reset and I can do the 2 second start again.

 

I just checked back to a few earlier SSD displays and that X on the BPM is new today.

 

By jumpering the fuel pump relay and forcing the pump to run, you are bypassing the ECM altogether. This might seem to point to the shutdown being caused by disablement of the ignition.

I can't remember whether you have checked for ignition activity up to and beyond the point of the shutdown, but that would seem to be a logical test to conduct. The ECM signal that fires each coil can be monitored with an oscilloscope or a quality DMM with a duty cycle setting. It's also possible that the constant voltage supply to the coils is being interrupted, and this can also be tested with a voltmeter.

The only circumstance I can think of that will cause the ECM to disable the ignition is low voltage. Have you monitored the voltage to the ECM from startup to shutdown? I can't recall the exact voltage below which the ECM will disable the ignition, but I think it's around 10.5 volts.

Are you using OE spark plugs properly gapped? It is theoretically possible that spark plugs with a significantly lower or higher resistance could cause the ECM to interpret a secondary ignition problem, though I don't know if the Jaguar/Denso EMS has the capability of shutting down ignition if, for example, it detects an excessive voltage spike.

Regarding the loss of starter rotation after 7 - 10 starts, the starter motor may be overheating.

Sorry if I'm repeating any previous suggestions. I'm also sorry you are still dealing with this mystery!

Cheers,

Don