Max,
I bought a Dataman scope for some work I was doing - it plugs into the USB port on my laptop and cost <$100. It is a hugely useful tool for so many things - not just Jagwaars - so why not get one?
Picoscope is more car oriented but (IMHO) not as good a general tool.

 

Maybe the return springs on injectors 2 & 8 are cockeyed, weak, broken?

An injector can be taken out of service by disconnecting it. If the engine runs lumpy but clean ... there you go. It will be one of them, but then the question is injectors or control of injectors? But, it isolates it to two injectors, 8 wires, and the ECU.

But, then again, it could be coils 2 & 8. Is the insulation cracked anywhere along the way. It likes to crack at that bend near the rear of the coil cover. The wires crack too. People will tend to wiggle test in the open runs, not the runs that are hidden away.

BTW, which is it? XJR, XKR?

 

Don't know.

Coolant temperature sensor seems like a likely candidate. Higher resistance = lower temperature. Disconnected is infinite resistance = winter in Kansas.

MAF is another one. The ECM falls back to throttle position sensor and engine speed as a failsafe spark and fuel map.

Thankfully those are two of the easiest connectors to access

Expect either one to trigger a code that needs to be reset.

 

In the category of things NOT to do: (See attachment)

So, Tuesday night, after installing a brand new set of spark plugs (desperation move) I started the engine and from the right side of the engine compartment, while it was running, I leaned over the engine to disconnect the O2 sensor. As I braced with my right hand to look behind the throttle body I DIDNT EVEN THINK that I had a shop rag wrapped around my hand.....

Well, guess where the free end of the rag went? Ummmmm, directly into the supercharger belt and pulley and my hand went with it!

BAM!!!!! That was quick! I didn't even realize what had happened for about two seconds - not until I started seeing pieces of shredded belt flying up from the engine. I knew my hand felt kind of like it had been hit with a hammer - but I HAD to get that thing turned off !! I ran around the back of the car to reach into the window and hit the ignition key, and as I brought my right hand up toward the window I saw blood splattering all over the door. I paused for a moment and realized I had left a trail of blood as I rushed around the car.

Drip Drip Drip on the garage floor and the visual confirmation of severely mangled fingers.....

I got the car turned off with my left hand as I held my draining fingers of my right hand as far away from the Recaros as I could.

So, about an hour later I'm getting stitched up by a very attractive lady Dr. at the hospital. And wondering where I had been keeping that extra goodyear belt!

Man, One trip through that pulley happens in about .02 of a second and what an efficient meat slicer it can be!

I actually fee very lucky - it could have been way worse.

Moral of the story: Some cars try to kill you when you try to drive them - Apparently Jags try to kill you when you try to FIX them.

 

Glad it wasn't any worse. Hopefully it heals well.

Gotta say you are a true car guy. Who else with a painful, bleeding appendage would worry about getting blood on the seats of his car. My kinda guy!

 

Yikes...it seems like you are being tested. Glad to see you still fighting!!

 

You were lucky - a friend had two fingers taken off by a Vee belt - and too mangled to be replaceable.
I'd get a Priest in to exorcise the damned thing - and top the rad up with Holy Water.

 

I'm sorry this post is so long - If you are so inclined to stick with me, thank you!

Since my finger was mangled and I found it a bit painful to work on a car, I decided to take the XKR to a Jag specialist hoping to get her back running in no time.

Wishful thinking!

Since it ran so bad in "Closed Loop" I was afraid to attempt to drive it the 25 miles to the shop, but finally I figured out how to get it to stay in "Open Loop" all the time... Disconnect either (or both) O2 sensors and it will run off of the fuel maps Denso programmed into the ECM. It runs a bit rich and kind of notchy (as it goes from program to program based upon what the TPS, RPM, and temp sensor is telling it to be at.) But it doesn't misfire! I was able to drive it to the shop.

The car was there for three days - stumping the technician that was trying the best that he could - Even reading all the posts on this thread as he was attempting everything he could think of that wasn't already done by me. He put the car on a computer using IDS. He noted that the injector pulse width and ignition pattern was looking great. He, like many on this forum have suggested, did not think the problem laid in the ECU. The new O2 sensor was showing some unsteady readings so he took a leap and replaced (again, for the second time in 4 days) the bank 2 primary O2 sensor.

That seemed to smooth the idle out, so he drove the car to his home Thursday night. He drove it back Friday morning - a total of nearly 100 miles.

He says it ran perfectly, in and out of closed loop, fast, slow, at idle and at full throttle. He called me that morning and proclaimed it FIXED! When I arrived to pick it up at at 5 pm in the evening. He had taken it for a final test drive about 2 hours before I arrived, just to make double certain it was performing well. He said it was. So I get in the car and turn the key and immediately THE MISFIRING WAS BACK!!!!! He was incredulous! And I didn't want to believe it! But there it was: running like sh*t again.... We even pulled the plug on the O2 sensor and it continued to misfire even in "Open Loop" operation. We test drove it and, as usual, the only time it would run correctly was when I floored it and it went to the full throttle fuel map in 'Open Loop". Wow! Were we both ever discouraged.

He suggested that I drive it home and bring it back on Saturday to do some further investigation- In the meantime, he was going to order the newest version of Jaguar's IDS for that engine to "re-flash" the computer.

I drive it home with the O2 sensors connected. It was running horribly, going in and out of "restricted performance" and throwing the usual codes.
P0300,302,304,306,308 & 0316.

Saturday morning, I disconnected the bank 2 O2sensor and drove it back to his shop. Like Wednesday morning, It ran fine in full time "Open Loop" while I was driving the 25 miles.

While I was at his shop, we began taking more "stabs in the dark." We pulled the cam position sensors and verified that they both worked and I traced all four wires back to the ECU - they were fine. Discouraged, I drove it back home to wait for the newer IDS programing to be installed on his computer.

Since I am "old school" and comfortable with reading spark plugs. I have continued to be intrigued about plugs number 2 and number 8 always showing fuel fouled black while the other six are totally clean with white insulators. Saturday night, I changed position with some of the plugs and noted that the even when the car is driven 10-15 miles in forced full time "Open Loop" and not misfiring, plugs 2 and 8 come out super black - while the other six are nice and clean. I put the car to bed on Saturday night, having just cleaned the plugs, reconnected all O2 sensors, cleared all codes and with a full reset of the ECU... Basically cleared it's brain of all bad things.

Sunday morning, with the car stone cold, I fired it up and to my surprise it was not misfiring at all ! I hooked up my scanner and monitored live data for a test drive of 35 miles. The car ran absolutely perfectly, clean and strong in all modes and all driving conditions. The "short term fuel trims always hovering around 0% in all modes of driving and no codes were thrown... It was like having my beautiful Jag back again!!!

I drove home, shut it off and pulled the plugs. Again, to my surprise, plugs number 2 and 8 were badly fuel fouled but as usual, the rest were clean. I dont get THAT???? I cleaned the plugs and let the car cool off for about FOUR hours, making NO other changes.

When I restarted the car it fired up nice with no misfire - so I took it out again for a test drive. Within, two miles it was back to its old bad deeds. Misfiring badly and going into "restricted performance." I took the car home and using fans and 3 hours of time, brought it down to room temperature and reset the computer. When I fired it up, I expected it to run good again... but NO, it was misfiring from the start.

I am wasting my time aren't I? I am waiting for a call from the Jag repair shop.

Depressed and tired.....

 

It's too bad you could not move the whole circuit.

But, how about moving plug+coil from holes 2+8 to some other holes? Move the matched pair to other holes.

The fan cooling is actually a good clue. The temperature sensitivity is somewhere that was not cooled by the fan. The pattern is different when cooling over time.

 

Sorry the car's still winning.
If plugs 2 + 8 were missing the O2 sensor would see too much O2 so richen the mixture.
You're at hard lean so it's trying to correct a rich mixture.
If injectors 2 + 8 were hanging open too long then that would blacken your plugs and force the trims lean.
Logically I can't see it can be anything else - but what's causing a random fault on two injectors?
If the injectors were mechanically bad then you'd get the same over-rich running in open loop -
which you suggest you are doing although the car runs much better in open loop.
Getting to the injectors to re-test is another supercharger off, of course.
Did you follow the other Steve's idea and look at fuel rail pressure ?

You say the mechanic looked at the injector pulse - was that at the injector itself or inside the ECM ?
Was it in open or closed loop?
Was the fault showing at the time?

Flashing (may, I don't know) re-write the fuel maps and maybe the algorithm for STFT/LTFTs but it won't cure a temperature sensitive ECM fault - but I have to ask where the fault would be that runs a bit rich in open loop and very rich in closed loop.
Plums says
The fan cooling is actually a good clue. The temperature sensitivity is somewhere that was not cooled by the fan. The pattern is different when cooling over time.
and that 'somewhere' could be the ECM.
Does it go away if you stick a cool block on it?

 

The open loop test shows at least that the focus should not be on the O2 sensors anymore.
Swaping coils and plugs would be a good start, reset the ecu do it will clear the learned values, and run in open loop again.
That Wil then determine where the culprit lies, although it seems intermitted. You said earlier that you had swapped the injectors from 1 bank to the other, was this also the case for the coils and plugs?
If not the latter, it seems either related to cabling or the ecu.
I can't see how a ecu reflash would work, it would mean you have now some data corruption, but as its easy to, why not.

 

Thanks guys, for hanging with me and continue to come up with new suggestions.

Point of note: All pattern of trading out COPS and plugs has been done. In fact EVERY time I change the plugs I put them in different holes and reposition the coils, just by habit... NEW plugs were put in last Sunday. Nothing has EVER changed.

The mechanic was probably looking at the injector pulse from the ECU - I don't know, I wasn't there - so I also don't know whether it was in open or closed loop.

I did see him do something "old school" at the injectors that I had already done before. He used a stethoscope to listen to the pattern of each injector very carefully and came to the same conclusion that I had before... That injectors 2 and 8 seemed to be responding at the same exact pattern as the others. Also, when I bench tested the injectors, I built a 555 IC timing circuit that (individually) opened and closed them each very rapidly - I too, during that test, noted that all eight performed the same manner.

Cool packs around the ECU may be my next line of attack - Simple to do, and definitely will test the ECU for a heat sensitivity.

This will be done tonight.

 

I honestly doubt you'd hear the difference if the injector was staying open longer than it should. I've no doubt that you can hear it open and close but the pulse width is a few milliseconds - with only a few more milliseconds between lean and rich. Maybe a trained bat could help!
If you're a dab hand IC man why not knock up a '741' based integrator?
At constant RPM it will give you a Vout proportional to the mark/space interval on your injectors. You could test it with your 555 timer. Maybe even a DVM on AC with a capacitor across the input would work - but a scope would be better !!

 

Steve, that's a good idea! I bet I've got a couple of those old Op-Amps around the shop somewhere. I may try the AC VOM with Cap first, but GREAT idea. Yes, a scope would be better.

Perhaps I'll wait until the shop gets the newer IDS installed and suggest that the tech check the waveform at the injector terminal - both with the injector plugged in and without.

I DID find out that with a fair amount of squeezing my non-ripped up fingers between the bottom of the charge-air cooler and the top edge of the valve cover, I can actually pull the electrical connector off of injector #2. Hopefully I'll be able to pull one on Bank 1 also, then I'll be able to compare a good cylinder with a bad one.

Regarding the chilling of the ECM experiment I plan to try tonight... I'm not convinced that it's absolutely a intermittent problem related to temperature - because when the engine has chosen to run right in the last week, each time it seemed as though the car would have continued to run correctly as long as it was powered up. It was only when it was turned off and then restarted moments (or hours) later that it started misfiring again. And it rarely runs correctly, whether hot or cold... So it's a "crap-shoot." Oh well, it will be another possible problem eliminated even if it has no effect at all.

 

Well THAT was a waste of energy!

I did a hard reset then packed the ECU box with freezer cold packs and attempted to get the engine to start up without missing. NO LUCK!

Nothing I thought to do worked. It just WOULD not idle with all sensors hooked up with out a severe miss and -25.8% STFT on Bank 2. It threw all the aforementioned codes too. I ultimately disconnected one of the O2 sensors and drove it a few miles in "open loop." just to see if it would allow me (after it was fully up to operational temp) to connect the sensor and have it run properly... but NO it was not having anything to do with running in "closed loop" mode tonight.

I tried again disconnecting the water temp sensor and that, like before, caused it to go into constant "restricted performance" So that certainly didn't do any thing positive.

Complete waste of time.

 

Sorry about that - worth a try though.

 

Hi Maxwdg, have been following your threads, and you are definitely being tested. Do you know if the coils on these engines fire in pairs, (one of the techs might clarify this point). It might explain why you have 2 plugs constantly fouled. When you took the supercharger off, did you clean the egr valve?
good luck
Pete

 

Thanks for the sentiment and your suggestions fredd60

The firing order of the 4.2L V8 is <1:2:7:3:4:5:6:8>

So no simultaneously firing plug on this engine.

Yes, I cleaned and checked the the EGR valve for "free" operation when I had her apart.

 

Have you looked at those injector pulses yet? Nag Nag - it's the waiting I can't stand, Sarg.

 

Maxwdg, what a story! I hope you get it figured out soon.


The only things that I can offer to help:

- Wondering if it is a clue that Jaguar only sells the fuel rail with 8 new injectors and a new pressure regulator now? Maybe their techs have run into the same kinds of problems as you have been having and determined that if the fuel rail goes then you have to replace everything to avoid the risk of having to tear it down multiple times?

- At one point in your story you were wondering about air in your intercooler coolant. MrTexasDan has a great guide to refilling the coolant on the XKR and bleeding the intercooler as part of his thermostat replacement instructions. It's linked in the maintenance sticky at the top of the forum under "DIY Replacement Instructions" or right here: http://matt.zenfolio.com/xk-coolant-...tat---v1-1.pdf I have followed those instructions and provided a slight update for the 2003-2006 models at https://www.jaguarforums.com/forum/x...006-xkr-65973/

- You are an EE, as am I, and I think many other forum members. You probably noticed in college that the Mech Eng's are built different from us. I'm not going to start a flame war over better or worse, but I think we can all agree that we're different. My point is, thinking like an EE with 555's and 741's has gotten you to a certain point, but maybe you need to think like an ME to solve it completely? More zen than karma?

- PM me if you want me to send you a couple of the injectors I'm pulling off of mine, if you think it's an injector problem.

All the best,

Marc