The pinpoint tests for the HO2S sensors are found in the "electronic engine controls" pinpoint tests section of JTIS and not in "emissions controls" as most would expect. The codes against them do not include the codes mentioned, but the tests are there if you want to pursue them.

However ...

consider that the engine runs open loop when cold, and that in open loop the HO2S sensors are not in the equation, therefore good or not they will not affect the running quality. Fueling is when cold is determined by the default fuel map + LTFT.

The suggestion is to do a full hard reset. Some obscure reference in Jaguar materials say that the "hard" reset involves a 15 minute disconnect. It is mentioned in the context of conventional resets not achieving the desired result.

The hard reset will reset your fuel maps to default and zero out LTFT stored in KAM.

If you then start the car from *cold*, and *if* there is a HO2S sensor fault, it should still run acceptably. If the engine does not run acceptably, it is unlikely to be a HO2S sensor fault and the problem lies elsewhere. You will only have a couple of minutes or so from bone cold to make your observations before the H02S sensor lights up.

How are the grounds on the COPS and injectors? The grounds go to a common point and can be a suspect when an entire bank is involved.

This is just speculation ... and is worth what you paid for it

 

As I have no idea how to test the sensor by which I mean the accuracy (the ECU would report faults like high/low voltages etc), I would swap it with another one (assuming you have verified the heating/relay of the unit). And if it turns out to be something else, you have a spare unit available.

Another thought could be the camshaft sensor, as from documentation that one determines the ignition/injector timing, and thus can have an effect of course.

It’s odd that you say it has driven fine even for 20 minutes, so at cold the defective parts/cables might be fine, but once warmed up they start acting up, sorry have no direct diagnose for you, so eliminating possibilities might be the only way for now.

 

Plums & AVOS. Please keep the ideas coming. I really appreciate your help.

I have thought about buying another O2 sensor, knowing that I'd have an extra if I that didn't solve the problem. I didn't test the heater on the sensor yet - thanks for that suggestion. Also, I need to find the relay that powers the sensor heaters - or I suppose I could just turn on the ignition and see if I have power on the proper sensor leads, correct?

Plums, I have read several things on-line about the 02 sensor not being consulted by the ECU when the engine is cold - I even read somebody's account that the O2 sensor was not used except during full throttle or hard acceleration...??? Perhaps I read THAT one wrong - because I found it hard to believe.

Anyhow, regarding the "cold" operation - since the car is not being intermittent anymore, it doesn't matter if it's hot or cold it pretty much misses constantly - on the road tests I let the engine come up to operating temp for several miles before I pull the vibrating and sickly beast back into my shop. (Those that followed my earlier thread about removing the supercharger to repair the fuel leak will remember that when I first put it back together, it ran decent cold, but after a few minutes of running, it began to miss-fire. After a couple of heat-related ideas were addressed it began to run fine for a day or two and then suddenly it began to misfire constantly)

Like Plums suggested, the above described situation seems to sound like a possible grounding or other wiring problem. I think I'll look harder for a problem like that today. But, I'm still am confused about the super fouled plugs on cylinders 2 and 8 - but not on 4 and 6??? Hmmmmm...

AVOS, I have thought about cam timing but put that out of mind when the car ran so well for those few days... You have reminded me to check that sensor and it's wiring for possible problems. That IS the pigtail harness that goes into the back of the valve covers or heads, correct?

By the way, I'm confused about precisely how to do a HARD reset of the computer. I've read many accounts here in the forum and several conflict with each other... Battery disconnected for 15 minutes?, Batt disconnected overnight?, Battery disconnected and the cables touched together to dump the capacitors? Should the ignition be switched ON while I do any of the above? Or Ignition off??

On the last operation attempt, before I began to disassemble the left exhaust system to check the O2 sensor and convertor - I attempted to reset the 'puter by disconnecting the battery. Then, within seconds, I shorted the positive and negative cables from the car together for a couple of seconds to drain the capacitors in the system. That had no effect at all with the way the car ran - but did result in 16 trouble codes - All were low voltages - so I reset the codes and resumed looking for problems.

And so the search goes on.....

 

Just read that the ECU is only during start checking the camshaft position to know which cylinder needs fuel/ignition (on the AJ27 engines), so once running its relaying on the crankshaft sensor again, or in other words, the issue you have doesn’t seem to be related to the camshaft positioning (/sensor), but am not sure for AJ34 engines like you have.

 

I'd guess you did that with the ignition on. Doing it with ignition off doesn't generate any fault codes.
What you did (disconnect ground, touch ground lead to batt +ve for a few seconds) is a hard reset,
but you see Plums said in post 21'The suggestion is to do a full hard reset. Some obscure reference in Jaguar materials say that the "hard" reset involves a 15 minute disconnect. It is mentioned in the context of conventional resets not achieving the desired result.'
I can't see why but it won't hurt to try it like that.

 

Be very careful to note the exact wiring colours to each one before disconnecting the sensors.

If the XK8 is like the early XJ8, the sensors are connected at the factory and then configured with the Jaguar WDS. Getting them mixed up is a nightmare in the making.

Don't ask how I know that!

Graham

Hahaha Graham, I learned that the hard way too! Seems the previous owner had swapped the plugs and I spent about 1000 in parts and many hours working on mine before a mechanic here in town discovered someone had swapped the upstream and downstream plugs.

 

It was either in a TSB or the engine theory/repair course manual.

 

Steve, I tried to reset the ECU both ways: Key off, touch the cables together = Nothing that affected the mis-firing... Key on, touch the cables together = Nothing affecting the miss-fire and 16 codes thrown.

Since the exhaust is off at the moment, I'll have to wait for reassembly to try it again.

Question: Is is accepted by all that I do the reset by - disconnecting the batt., touch the cables together for a moment and then wait 15 minutes before re-connecting them? Or should I not touch the cables together and just wait the 15 minutes?

Thanks AVOS, I will still check ALL the wiring that I can get to once again. Particularly the ground for the COPS on the left bank.

I haven't had a lot of time to play in my Man-Cave the last few days because a big project at work has left me feeling tired after getting home - but I spent some time yesterday and FINALLY got that left convertor off.

Boy THAT was a hassle! FYI: From the top of the engine bay, under the coolant expansion tank, I had to use an E7 "star" socket with a impact U-joint and three 1/8" extensions attached to a electric impact driver to pull the studs out of the top of the convertor flange (after removing the four 13mm flange nuts holding the convertor to the exhaust manifold - also from above.) All that effort was required to enable me to tilt the convertor far enough forward to pull the lower exhaust studs from the mid pipe flange without removing more of the exhaust system towards the back of the car.... Got that guys?? 'Aint easy to do!

Looks like all the trouble taking the exhaust sensors out & checking them plus removing the convertor looking for a clog was for NOT !!!! The sensors are good, I think??? I still may replace the upstream sensor - but I really have the idea that it is fine..... And the catalytic convertor is certainly not bad.

RE: Convertor... Although it LOOKED fine at first, I THOUGHT I was not getting enough air flow when I first ran air from my compressor through it - but I couldn't seal the air hose to the exhaust input well enough to satisfy my certainty. So I fabricated a flange for the input side that had both an air input nozzle and an air output nozzle - to hook to a pressure gage to see any back pressure that might develop. Well.... I noted ZERO back pressure at full column of air at 120psi from my "body-shop worthy" air supply. I'd say that convertor is flowing just fine.

Please see the attached pics below to see my silly home-brew convertor back pressure testing rig and also some fine shots of my equally contrived injector testor - The very one I used to establish that the professional cleaning and testing done by "Injector-RX" in Houston a month ago was spot on... Hey, look at the spray pattern at 45psi - NICE huh?

And NO I don't own "Chitty-Chitty-Bang-Bang," but my father WAS Caractacus Potts....

Thanks, again for the input ya-all... Now back to reassembly and the continued challenge of my lifetime.

 

Perhaps you could use an old spark plug and make a "test port" to replace the O2 sensors on either side, and then connect it to a fuel pressure gauge that reads up to, say, 5 psi. Then you could run the engine at idle, and then again at around 2000 rpm and note any difference in the exhaust backpressure without removing the cats or exhaust. If the pressure is the same, you need to look elsewhere. As I recall, up to 1.5 psi was considered pretty normal at 2000. Any more was considered problematic. I have no idea what the Jags are! I would also imagine that you can "buy" an O2 test fitting out there someplace!

 

Yeah, good idea "chopr" You made me think.... I was using a compression gauge and they are not real good at displaying really low pressure. A fuel gage would have been better - and your spark plug idea for a "test port" on a running engine is a good one. Thank you.

 

I suggest hooking the car up to a scanner and reading out all the sensors in real time. You will be able to zero in on the O2 sensors and tell what is working and what is not. I also would try to retrace my steps in the fuel repair to see what might have gotten put back wrong after. Get in there with a strong light and look for loose conectors, etc. When an entire bank is missing, it is commonly a cam timing issue. There just aren't too many things that can affect all 4 on one side. This has happened before and the solutions have been: low fuel pressure...replace pump; bad MAF..replace MAF; cam timing....fix tensioner and finally ECM..but.. fixing th efuel leak probably did it, find waht the fix broke.. It is unlikely that the O2 sensors are doing it, I have never heard of them causing misfire on an entire bank at once.

 

Well, oldmots, I agree - at this point it looks as though a fix may require a hook-up to a diagnostic scope with an experienced diagnostician at the helm. But first, I need to get this thing back together and try starting and running it again. I've checked on so many individual problems since the last attempt to start the car that I've certainly violated the first rule of problem detection. If it runs correctly, I will have no clue which thing I attended to that "fixed" it. But I suppose at this writing I've lost fully 1/6th year of use of my favorite car and I just want to drive it again.

I have stripped the entire top of the engine off twice - first to get to the fuel rail and repair the leak, and then a second time to get to the injectors and their harness for testing. I am near as humanly possible 100% positive everything was done properly - I took greater care and attention to detail than I have ever done in the past - stopping just short of actually developing a checklist and checking each and every move off the list.

I haven't pulled the valve cover and checked the valve timing because the car ran perfect for a couple of days just after the original fuel leak repair - so the assumption would be that the valve timing is still correct.

Man, I wish I knew what "the fix" broke....

 

Bet you guys thought (or wished) I'd gone away.

Nope. I'm still pullin' my hair out chasing a real mystery - Family Labor Day vacation put XKR on the back burner for about 10 days...

Well, I got it put back together after checking many, many things + a hard reset of the ECU and unfortunatly the same problems exist. (-25 STFT on bank 2 at idle and just off idle, bad missing, Restricted Performance goes into effect when driving for a few minutes, Throws codes p300, 302, 304, 306 & 308)

In the last major tear down I eliminated these possible causes:

-Bank 2 Catalytic Convertor is NOT clogged.

-All Four O2 sensors are working fine. The signal or voltages are right on point and the heaters are functioning. I also swapped the primary O2 sensors from left to right banks - as an added check.

-Checked grounding on Bank 1 COPS - Also checked continuity on wires going to ignition relay and parallel sync wires on Cyl 2&8 and 4&6. I don't think ignition wiring harness is the problem.

-Checked wiring harness for injectors - It's OK

I DO HAVE NEW CLUES THOUGH!

Car was still missing at first start up - but I decided to follow the EXACT procedure called for by JTIS on battery reconnection after my ECU Hard reset. This included a full warm up and a "drive" sequence. So I took the car out for a drive while monitoring the STFT & LTFT on both cyl banks, MAF sensor and TPS.

With gentle driving, the car ran horrible with lots of missing on left bank and -20 to -25% STFT on bank 2 - but perfect on bank one. It went into R.P. several times and threw the usual left bank "missing" codes. MAF and TPS looked normal to me.

Once, I got frustrated and mad so I floored it after it came out of R.P. and OOOOOHHH MY GOD! It ran like a striped *** ape!!! Fuel Trims went to normal across the board!!!!

I did this probably 15 times - and each time I gave it anything more than 50-60% throttle, the car went from running like crap to the beast that an XKR can be!

Each run, the trims stayed at Zero even after letting out of it, until I gave it any throttle at all - or rolled to a stop. At any throttle input less than hammering it, the left bank STFT went crazy (-25%), the "miss-firing" started again and R.P. alert was sure to follow. I watched the instrument panel, and as soon as it would go out of Restricted Performance, if I nailed the throttle the car ran perfect....

Okay, diagnostic dudes. What's this telling me? I've GOT to be close to the answer.

 

No comment related to your codes because the definition of the codes was not included in the post.

However, the reason the car runs well under heavy throttle is that it runs open loop under those conditions. Fueling is completely driven by the fuel tables.

 

He did say codes p300, 302, 304, 306 & 308.
So, if he's running OK on the fuel tables there is nothing mechanically wrong.??
As I understand it in open loop only the O2 sensors are ignored, MAF etc are still used.
So, if the engine's OK mechanically the problem must be that the ECM is miscalculating the injection time in closed loop, putting too much fuel in, so low O2 so -ve fuel trim.
Does this point to an ECM fault?

 

You're right, Plums - I should have defined what the codes I mentioned pertained to. My posts are wordy anyhow, so I omitted those details.

All codes are related to miss-fires: P300=Random Missfire and the other codes all indicate miss firing of cylinders on the Left or "Bank 2."

Thinking out loud here....

I was beginning to come to the same conclusion that both Plums and Steve are theorizing. When I nail it the 'puter goes by the programmed fuel maps - or "tables" and all is good for a few moments.

I think I have the same question that Steve proposed: Computers are powerful and complicated so I'm not sure that I can automatically assume that since it runs great "Open Loop" the ECU is fine (I certainly wish that I could eliminate the computer) But I think I CAN assume that the injectors, COPS and associated wiring are all good, right?

Let's just hope for a moment that the problem doesn't lay within the CALCULATIONS that the ECU is making during "closed loop" operations.

Perhaps there are there exterior input devices (sensors or their associated wiring) that the computer constantly consults with while running in "close loop" operations that would affect cyl bank #2 only?

How about the knock sensor on bank 2? I replaced the knock sensors with new Jaguar parts while I had the supercharger off the engine as preventative maintenance. Could a bad one cause a problem on just ONE bank? Hmmmmm I don't know, but I doubt it. Wouldn't EITHER knock sensor detecting pre-ignition cause a "global" timing retard?

How about the cam sensor on bank 2? AVOS touched on this - but he thought that the cam sensor was consulted only during start up.... Anyone know for certain?

The MAF sensor would cause a "global" problem - as would EGR malfunctioning, right?

I exchanged the O2 sensors left to right and the problem still exists on bank 2 only, so that can't be it, correct?

God, I really don't have the $$$$ to buy another ECU and throw THAT at it so I must cover all bases before I go there.

Please keep the ideas flowing guys, I'm doing the best I can.

 

FWIW - we know there's nothing mechanically wrong - it goes like a train at WOT.

If you had - eg - an inlet manifold leak that would be less evident at WOT but would (surely) throw +ve STFTs (too much O2, pile on the gas).

I don't know what the ECM does in response to a knock sensor signal, maybe reduce fuel injection time, but I understand the knock sensor is still being read open loop so the problem shouldn't go away at WOT

I can only assume your Bank 2 closed loop mapping has gone bad.

Maybe that can be cured by an ECM reflash - the values must be stored in ROM.
Can you post your LTFTs?


@Plums - what kicks the ECM from open to closed loop at startup ?
Is there something he could do to force open loop at all throttle values?

 

The LTFT for Bank1 & 2 - BEFORE the last ECU hard reset were consistently right around Zero and have been consistently Zero throughout my ordeal.

As of yesterday, after the most recent ECU reset, and a bit of time with the engine running, I noticed that the LTFT on Bank 2 (the problem bank) was now reading in the neighborhood of -10 to
-14%. Bank 1 was still at Zero.

Probably goes without saying that when I went to full throttle, All the trims - short and long term, would read Zero.

After a full throttle run with a subsequent release of the throttle while the car slowed down - all I had to do was barely crack the throttle plate and instantly the trims were back to their "closed loop" readings - almost as if the TPS was what triggered the ECU to go from "Open loop" to "Closed Loop."

As far as a manifold leak - those that have followed this convoluted thread for it's entirety may remember that was one of the very first things I suspected. I found none. Additionally, when I took the supercharger and charge air coolers off for the second time, I was uber-careful to make certain that EVERYTHING was sealed up tight.

 

See if there is a PID for the knock sensors. Brutal has mentioned that they must be torqued correctly. Too tight=too sensitive, too loose=not sensitive enough. So, if one sensor is too tight, it will see knock. If it is too loose, it may not see knock, and the ECM may have a strategy of overfueling to protect the engine if it thinks it is "knock blind". Think back to rich fueling strategies on carbureted cars.

The "crack throttle from idle" is a good clue. Check TPS sensors, connections and mechanical operation thoroughly. The first ignition on sequence causes the the throttle plate to calibrate. That is all the buzzing and whining, and idle hunting that happens immediately after a hard reset. When you key on from a hard reset, try letting the system do a full boot at position 2 before doing the actual start at position 3 and keep the foot off the throttle while the calibration happens. The system has fully booted when all the dash lights go off except for the oil light and battery light. Do all of the above with a fully charged battery for optimal results. Also check the PID for system voltage. Not what you see on a DVM, but what the ECM is sensing internally as a reference voltage. The *reference* voltage is very important as the ECM *sees* the reading based on voltage or current offset depending on the type of sensor.

Ummmm .... and from left field ... bad fuel? 87 AKI fuel with 10% ethanol seems to cause LTFT of +10 that can be recovered if replaced with 91 AKI fuel. This was observed using long term logging.

 

True, however should one poster do the work of copying the definitions in, or should every passing reader have to scroll to the top, open the listing, find the codes and read them?

Most won't bother and one of the ones that didn't bother might have known the answer but did not recognise the "symptom" disguised as a letter and 3 digits.