I just picked up an early 95 ( build date 10/94) XJR, Canadian Spec. It was sold because it randomly stalled, and several Jaguar mechanics couldn't figure it out.

I have a lot of bugs worked out of it, but I was driving it tonight and it stalls on me. I usually drive with the instant fuel economy showing on the display, I find it a useful diagnosic tool. I was coasting down a hill at 50 km/h, throttle closed and the instant display showing fuel economy in the 50's mpg - as I would expect. Then it goes to 8 and the car feels sluggish, but it recovers and keeps going. Later at a traffic light it again goes super rich and stalls. Took about 5 attempts at restarting, put into drive, move a few feet and it stalls, repeat to make it across the intersection. There were big clouds of black smoke as I'm cranking and it starts and runs for a bit. By the smell, obviously very rich. No codes were set after this, and the CEL stayed off.

As I was driving with the OBDII reader plugged in, I noticed that will feel fine, but it will switch to open loop from closed loop. There was also a display of open loop-drive. When it's in closed loop the instant economy is about where I would expect it to be, and at idle the O2 sensor readings are swinging from high to low voltage as they should.

So far I have replaced the spark plugs with new Champions, checked the air injection check valve and put a new fuse in to get the air pump working ( at least I assume it works, I've never heard it work). The P0411 code that the car was showing when I bought it has gone away. I've replaced the crank position sensor as well.

As I was cruising along at 80 km/h on a level road the coolant temp was showing 80C. I exited onto a slightly slower road for a few km, then came to the light where it stalled. After I was able to get it running again the temp was showing 87 and stayed that way until I was home and that was driving at 80 km/h. Is it possible it's just a bad coolant sensor? Would that cause it to go open loop? I didn't see any flutter in the OBDII temp display, but I also couldn't watch it all the time.

Long term fuel trims are showing 44%, short term around 12-16%. At idle in a dark garage there is no light show from the coils, all are original. Car has 136,000 km on it. Previous mechanics have done a lot of work to this car, such as swapping coils, ECU, supercharger, and doing the headgasket ( twice), with no change in the stalling. I'd hate to know what the PO paid for all that. I know the mechanic who did most of the work, and he's a competent guy. Factory trained on X300's, used to work at the dealer before setting up his own shop and a Jaguar Master Mechanic. He had another XJR in the shop for a transmission that ran well engine-wise, so that's when all the parts were swapped to try, but to no avail.

When it runs closed loop, it runs very well, I'd like to figure out what is causing it to go rich occasionally.

 

Are the STFT really +12 to +16? That would indicate the sensors read a lean condition.

And YES, the engine temperature is used in the fuel mapping. Maybe you do not have a thermostat installed?

 

I'm pretty sure the STFT was positive, yes. Perhaps influenced by the occasional excessive richness?

I think my next plan of attack is doing a thermostat and coolant temp sensor, as well as checking the wiring to the CTS. Perhaps a broken wire in the insulation?

I do have new O2 sensors I can put in, the ones in the car look original, and very rusted in place.

 

There is another known "anomoly" with these. Your rich condition could be due to a lean measurement on the O2s due to cracked exhaust manifolds.
BTW, the LTFT high readings are not valid- A number of us have had those without a clear undertanding of why. In my case, I THINK the high LTFT reaings went away after a TS calibration and a "O2 orientation" done with dealer level software, but it did not fix another car after the same procedures. So, who knows? Certainly not me or Doug.

 

The car is quite rusty, and I have all the bolts out of the exhaust manifold heat shield except one. I broke one, and I don't want to break the remaining one, which is the lowest one next to the air injection pipes. So I have not checked the manifolds for cracking yet. It's on my "to do" list.

As I mentioned, though, most of the time it runs very well and as it should. It's just this random richness that stalls it that is baffling.

I do have the IDS software and a genuine Mongoose tool, I could do the O2 orientation and TPS just to see what happens. I've not connected the software yet.

 

take a look at the rochester valve, i posted info on this site. only 20 dollars u.s. and it may be the cause of your over fueling problem. read the post. Also change the temp sending unit just because it goes into open loop does not mean sensor is good. I have had similiar problems with my 95 xjr, on different occasions, and these were the culprits. Where did you get the IDS software and a genuine Mongoose tool and where do I get it. It looks like you have done everything else, my 2 cents.

 

There is a link to the IDS download on the Cloned Dealer Scan Tool thread, and I bought the Mongoose directly from Drew Technologies.

I do have a P0443 code for the purge valve, however I can't find it. I've seen in the manual how it is supposed to be behind the left headlight and under the air cleaner, but I have nothing there. No vacuum lines either. According to the parts listing I shouldn't have one either, but it does throw the code occasionally.

 

Hi mate the purge valve is in front of your left rear wheel if rhd or right rear wheel if lhd its under the carbon canister witch is a black plastic box with pipes going to it

 

on my 95 xjr the purge valve is under the air box just behind the headlight and the rochester valve is behind the drivers side rear wheel underneath the car below the passengers seat. If you have no purge valve you may have a vaccum leak as the purge valve has a vaccum line that goes directly to the tb adapter. the tb adabter has three vaccum lines, one to the sc bypass valve, the second goes to the purgevalve and the third comes from the rochester valve back into the tb adapter. line comes from the tb adapter into purge valve then from purge valve to rochester valve the from rochester valve back into tb adapter. maybe they removed purge valve and your car just doesnt do well without it.

 

This is getting more interesting. I connected my laptop to the car, tried to do the oxygen sensor orientation and it came up and said the car is not equipped with oxygen sensors!

It had most of the vehicle configuration where it indicated there was a catalytic converter fitted, and North American emissions. However, it did fail to communicate with the security module and couldn't auto fill in the vehicle configuration.

The car certainly does have 4 O2 sensors, they are plugged into the harness and my code reader displays the swinging voltages each puts out. It also showed the car in closed loop while idling tonight. It's not a smooth idle though, there is occasional stumble and roughness.

I replaced the coolant temp sender, the reading from a laser thermometer agrees exactly with what is reported in the code reader.

 

So the old temperature sensior was out of calibration? That could cause a lot of trouble.
Are you sure you used the correct VIN when you connected? AFAIK, all X-300 models had a t least 2 lambda sensors. I think maybe you should try "best out of three" connects? Try for the TPS calibration.

 

The old sensor was similar to the new one, but a few hundred ohms different at 25C. The wire clip on the connector was missing, so I replaced that to make sure the connector stays on. Idle was still occasionally lumpy afterward, I didn't try driving it.

SDD/IDS wouldn't auto connect to read the VIN, so I entered the North American VIN manually. It did pull out that it was a 95 MY X300. I'll try the TPS calibration and see what happens.

 

Yep, it will not pull the VIN on X-300(or X-308 for that matter) in my experience, but it sure knew the car once Iput in the VIN!

 

most dont know that you must reset adaptive terms "fueling adaptions" when you have done repairs for lean or rich fuel conditions. the ecm will continue as usual unless its told the problem is fixed. given the extensive work done on your car, heads off twice and everything else, this is necessary. take a look at attached info from this site.

 

I've seen that bulliten before, but finding a PDU will be a problem. As I mentioned, I do have SDD/IDS software and the Mongoose interface, and it says the car doesn't have any O2 sensors fitted, thus it failed to do the O2 orientation. My generic scan tool can see the voltages from the 4 sensors, and I can watch them swinging from high to low voltage when in closed loop, so there is clearly something amiss here. The car does go into closed loop, although even at idle it wasn't smooth - occasionally. Sort of like a slight misfire.

IDS is a pain to deal with, it can't communicate with the security module for some reason, so it can't pull the vehicle information/spec and it also won't let me exit the application, it just goes into a loop of trying to connect, says it cant and wants to retry. Clicking exit brings me back to trying to connect again. Quite annoying.

The biggest problem with this car is the symptoms are intermittent. Most of the time it runs fine, then it will load up with fuel and stall. Restart and stall a few times, then it clears itself out and runs fine again.

 

Which VIN did you use? The "USA" VIN on the door sticker, or the "World" VIN stamped onto the body beside the battery?

 

I just started having problems with rich conditions and according to the info on reset of fuel adaptions I should take a look at the fuel pressure regulator. I will change out this weekend and let you know.

 

I dropped in to talk to the mechanic who did most of the work on this car. It filled in a number of missing pieces for me, as he said the problem started when the PO was trying to boost the car after having the battery go dead and he connected the jumper cables backward. Must have been lots of sparks! The reason the head was done is that the PO was racing a BMW, over revved it and bent some valves. That's been fixed at the same time as the other work was done.

So now I'm looking at electrical items that could have been damged, or the wiring harness. I drove it today with the alternator disconnected, to test if the diodes were fried, and it still stalled. However, as it did that, the load reading on my coder reader jumped from 15% to 37%. I've unplugged the air flow meter, drove for about half an hours and it drove perfectly. So now I'm suspecting a bad air flow meter or harness. More investigation is needed!

 

I've done more testing on this car, I drove it around with the AFM unplugged and it ran fine, so I secided to plug it in and see what happened. I didn't get 2 car lengths before the car stalled and died! After unplugging the AFM it was a very difficult restart, but it finally stayed running. Took about 10 attempts at starting.

I talked to the mechanic again and he said he sub'd in a different AFM and the car stalled. Then tried the AFM from my car on another car and that car run fine, so he didn't suspect the AFM. Repeated that test with a different ECU, same result, my ECU would not stall in a different car.

Thus I suspected something in the wiring harness, so I directly wired the AFM to the ECU. It started and ran better than it has to date, and I drove it around for about half an hour with no troubles.

Then when I was merging onto a reasonably major road, the engine hesitated and the load reading on my code reader jumped from 30% to 97% while holding the throttle steady. Since the road ahead was clear I put my foot down to match the "load" and the car accelerated and shook off the richness. Clearly it's not fixed! At least it didn't stall.

So now my question is what determines the load the ECU sees and reports to a handheld scan tool? I'm guessing a combination of the AFM reading, throttle position and RPM. Unfortunately, I didn't notice what either the tach or scan tool said engine RPM was when it loaded up. Crank position sensor has been replaced 3 times, so I'm not suspecting that.

 

I THINK you have all that right. No codes? The inlet air temp and engine temp sensors are all important to the calculation. I would chart these aparameters and the ones you mention vs engine load to find the correlation.