I'm beginning to question if the codes actually match the offending cylinders. I'm going to scan the codes again to make sure nothing has changed. One thing, when the car is running, roughly, it has a slight low in the idle which happens on a fairly regular schedule. It drops about 50 rpms and then goes back to a steady idle.

 

Hi,
So 1A is the cylinder nearest the radiator on the RHS, looking from back to front, I believe.

I've found a copy of the manual for the meter you're using, and the specs show that it should be able to detect the injector pulse.

I guess you could pull the injector plug and check if that makes the idle even worse?

 

Yes it did.

 

Well I've tried everything and I can't get a Hz reading on either the injectors or coils. I've checked all coils and injectors with key on engine off and I get 12V. Engine on I get 12v, but no Hz reading on wire #3 on the coil or either of the injector wires.

 

It sounds like you're doing everything right. It's just not playing ball.

I've dropped you a PM

 

I'm like a bad penny, back again I had a great talk with Michaelh about the car. Because of our discussion I decided to see what I could do about testing the injector circuit. I finally figured out how to separate the injector plugs from the plastic rail that connects each bank of 4. I then was able to use my noid light to check the signal from each injector. This is what I got.
Cylinder
1. lights normally
2. lights normally
3. no light
4. lights normally
5. intermittent light, Blinks for about 6 seconds, then stops for about 3 seconds and then repeat
6. lights normally
7. lights normally
8. no light.

So this test shows Michaelh and Gus are on the right path. The 3 cylinders that I have codes for are not getting signal from the computer. I have made a short video showing the noid light working on cylinder 4.

 

OK, so you've proved the thrown codes are valid and causing the ECM to take the 'default action' described in the DTC list. The behaviour of 5 is odd...

I'd expect the injectors to fire for the first 5 seconds or so since the ECM allows a short period for allow everything to settle down after engine start, No matter - we know the injectors aren't firing and it's progress.

So we're left with:

1. faulty coil packs
2. 'spark OK' feedback outut from coil packs for the affected cylinders not reaching the ECM
3. ECM not functioning correctly

You've confirmed spark to all cylinders. We need to check (2), which is more difficult as the outputs are spliced into two groups of four in the wiring loom, so a simple lamp won't work here (the pulses are too frequent to count anyway).

Straw clutching here - is there anything obviously different about the coilpacks on the affected cylinders?

 

If you clear the codes, the ECM ought to energize all 8 injectors until it finds a reason to disable cylinders. I often clear the codes over and over again and verify even the pending codes are cleared, then fire the engine up and check fault codes continuously to see if & when the code reappears. You may want to clear the codes and then retest each injectors that seem to be funky.

I understand, all coils were replaced and there were significant resistance variations among the 2 sets, but if coils are swapped, does the misfire persist on the same or different cylinders?

 

We have a '99 XK8 convertible with ECM LNE1410BB/003 which is very close to your LNE1410BC. If you wish to send it, I'll test run your ECM in our Jag and we can figure out, once and for all, if the problem is in the ECM or not.

 

I would love to take you up on this offer but I think you wouldn't be able to start the car with my ECM since it has to be coded for anti-theft.

 

Take your test leads and tap into the correct lead for the problem injector at the ECM and see what you come up with. That would eliminate all the other leads and connections.

Gus
JagRepair.com - Jaguar Repair Information Resource

 

That is my theorising (and it is just that) as noted previously:- the DTC implies that it will wait 5 seconds before taking the default actions. This for P0351, which all the P035x reference:



I'd expect the ECM to continue to try to fire the coil so recovery is automatic in the event that the fault is transient/intermittent (as appears to be the case for cylinder #5).

The OP reported that tests showed all coils were sparking. I guess a double check would be to use a noid lamp across pins 3 & 4 of the coil pack to verify the the ECM is actually issuing a 'fire coil' command.

The fact that swopping coils between good & bad cylinders suggests that they are likely OK.

Gus' test above will confirm whether the ECM is issuing the injector pulses.

 

If the coils were swapped with no change, then the coils primaries should all be OK. It would also suggest, the coil secondaries are good, but the plugs are an inherent link in the secondary side.

Modern plugs have a specified resistance. The ECM and coils expect that value they left the factory with. Are the plugs the original make & number? Along with resistance, gap is important. A wider gap has to fire at a higher voltage and conversely, smaller gap breaks down at lower volts. On two of the bad cylinders, try a different plug with a smaller gap in one and different plug with wider gap in another and see what happens. Fortunately, we don't have plugs wire in the equation, but you may check that the coil snaps down onto the plug securely when you have them out.

I'm a bit confused with the "Possible Causes" referring to ignition "module." When these engines first came out, there were two ignition amplifier modules like in our '97 XK8 but they went to coils with internal amplifiers soon after and our '99 does not have them. I have to wonder if the reference to the modules is an oversight in not removing them from the "causes." As I recall, the ECM powers the coils directly, half from one pin and half off another. Each coil has an individual wire to a unique ECM pin that grounds them. Maybe you tried this- unplug the ECM and check resistance from the ECM common feed pins to each individual ground pin. This should test each coil primary circuit, from start to finish- wires, primaries and all connectors in between. I understand the wiring harnesses have been changed, but the meter will either eliminate another variable or point us in a different direction. .

 

Checking resistance in the ECM plug described previously may or may not yield usable info. Looking at the wiring diagram one might get the impression the terminals lead to the coils but they run through the ignition modules built into the coils.

The attachment is taken from the Jaguar XK8 1998 update and reveals details on the ignition system introduced in 1999 MY XK8s. It describes how the module monitors the amperage to determine misfire. Rather interesting. It doesn't mention firing a high voltage spike through the spark.

It also mentions they came with double platinum tipped plugs. Tweaking the gap could waste the platinum button on the ground electrode if one is not careful.

By the way, I may be able to check out the suspect ECM. The immobilizer circuit was more complicated starting in 2003 MY but the '09 - '02 might only store the chip code in the ECM and not the Body Control and Key Transponder Modules. I'll need to locate a detailed explanation of that system, but if the code is only stored in the ECM, then I can check out a different ECM with its key taped in my detached immobilizer ring.

 

According to the description in the SERVICE TRAINING COURSE 682 ADVANCED JAGUAR ELECTRICAL SYSTEMS, dated 2/15/2002, the Key Transponder Module (KTM) stored the key codes. When the transponder ring transmit a code stored in the KTM, the KTM outputs an OK TO START signal to the ECM via a serial data link.

That would suggest, in that '09-'02 window of model years, the ECM can be swapped out without any reprogramming.

.

 

Hey everyone. I thought I'd upload this video just showing the noid light for the injector on cylinder 5. It shows signal for a while and then stops, then continues.

 

This video is the noid light and spark tester on cylinder 8 which is non functional . It shows spark but no injector.

 

#5 cylinder reminds me of one engine that utilized a controlled, alternating fire/shut down for an overheated engine limp mode.

As sporadic as #8 is misfiring, I can see why the computer deleted that cylinder's injector.

Would clearing the codes and restarting bring either back to normal, at least temporarily?

 

I've cleared the codes many times and it continues to do the same.

 

I think it is just that. The text is unchanged from the early DTC lists until the 2003 one which is slightly different:

"
ECM to ignition module / coil drive circuit:
open circuit, short circuit to ground,
high resistance
Ignition module / coil ground circuit:
open circuit, high resistance
Ignition module / coil B+ voltage supply circuit:
open circuit (including relay, if fitted)
Ignition module / coil failure
"
All the coils have a common B+ feed from the ignition relay via an 8-way loom splice.

It will do no harm to check the injector feeds at the ECM end of the wire, although this circuit is monitored by the ECM and I'd expect to see P020x codes if there was an issue there:



I'd really like to know what's happening on the two confirmation pulse inputs to the ECM (slate/green wire from each coil pack on one group & slate/light green on the other). That's a little harder as it's not a go/no go noid test due to the grouping.