Hi all, very happy new owner of a 1999 Sovereign 4.0 V8 and I am slowly working through her few little faults and I have come across one which has me puzzled.

My OBD (Gendan EngineCheck) reader reports P1646 and P1647 error messages which lead me to think I need to do some work on the upstream O2 sensors and that is fine but the OBD reader reports that the Check Engine Light is on BUT strangely the CEL is NOT on. Clearing the errors causes my OBD reader to report that the CEL is off but there is no change of course.

The Check Engine Light and the red/orange warning lights display correctly during the 'ignition-on' system check.

The OBD reader performs flawlessly and exactly as expected with regard to the CEL when I clear errors on my Volvo C70.

Has anyone experienced this and can shed some light on it? Is this some strange Jaguar-specific behaviour on the OBD implementation?

James

 

I wouldn't do a thing to the car except plug another scanner in it.

 

Fair point - I don't have any doubts about the scanner as it has worked flawlessly on my Volvo C70 for some years so is it that I need to get access to something which speaks 'Jaguar' or is this just a process of elimination?

James

 

Why not acceopt the first answer as a course of action and obviate the need for other ones?

I don't know what scanners are incompatable, I know that a WDS works fine and an Actron will throw a car like yours into limp home mode if I drive around with it hooked up for ten minutes. Your mileage may vary. The point is that the only thing that is wrong with your car is the reading on the scanner. That wouldn't lead me to a first assumption that there is something wrong with the car.

 

Well, if you want to play, then unhook a sensor, like the MAF, and see if you get a light. And if you don't? What are you gonna do, change the ECU? Take JTOs advice and "don't worry, be happy!"

 

Well, I would certainly have done that had I another scanner, I have no reason to doubt the one I have and no offence meant but you haven't really given me any real reason to.

You may well be right but I feel that remains to be seen.

I agree, and I'm not assuming it but it seemed a reasonable question to ask as I've done everything I am able to so far. Thanks for your time and help.

James

 

Well, if it comes to it...

Sure, if that's what it takes (after I've checked the basics of course).

I will, if I can get hands on another reader and I'm certainly not worried but I'll be happy when I know what's going on.

Thanks for the help!

James

 

I've had scanners not read the Jag, but work just fine on other cars. They can be particular. Where are you located Jima?

BTW, tradition dictates a formal introduction in the New Member area if you get a chance.

 

James:
Well, like you, my first inclination would be to change any part not working exactly right. However, buying a new ECU will set you back $1200 plus plus . A used one must be an exact match to the engine or you have to flash it, I believe. Either way, you are at three or four C notes. So, if the only problem is no CEL, I would probably just live with it and scan the car from time to time. I will tell you I have never seen a report of that anomoly, and ECU issues are something I follow on the forums.
BTW, if you are near Atlanta, give me a shout and I will be happy to let you scan with my AutoEnginuity, ELM with ProScan, or Ease scanner. I mentioned them all, not to brag, but to qualify my statement that they will all behave differently. Reason is, all but the WDS that JTO mentioned are not optimized for a particular model of car, or a particular bus timing. They talk on three different protocals, and many different cars. So, how the scanner shares the communication bus with the other nodes on the bus can change the scanner, or the car's performance or reliabably in certain circumstances.
For instance, there is common reports that codes p-1637 to P-1642 are read. I am pretty sure that SOME of these occurances (if not most) are due to connection of the scanner. Why? Because those codes cause a CEL and I have connected my ELM scannner, gotten a light Asome time AFTER connection, then found a CAN bus fault code.
As to the O2 sensors. I would rest the codes, see what happens in a few days, watch the waveforms from the sensors to see thm switching reliably high and low faster than every 2 seconds, check the LTFT, STFT, then decide if I wanted to change them just due to mileage. I probably would not. Thats because I have done that on a number of cars and compared the pain of my skinned knuckles to what little performanc increase I could convince myself had occurred. Others report different results. Again, I am talking only about changing sensors that seem to perform right on the diagnostics!

 

Ross,

I think your ELM scanner results are similar to my Actron-induced road test failure, that it cased a CAN bus interruption all by isself. Never pursued it, I just started looking for an upgrade.

 

I would bet on it.

 

Thanks, enough evidence here from the guys to warrant trying to get hold of another scanner and see what it shows up before I go pulling O2 sensors off.

Profile updated with location and I'll stop by the NM area shortly - please excuse my bad manners

James

 

The codes keep coming back, sensors haven't been disconnected as far as I know.

Thanks for the tip, I'm researching for a scan tool now, the leader at the moment is the Memoscan U581.

I believe that it the car talks ISO9141 but I do want something which can talk CAN and doesn't nuke the ECU

Jim

 

Hi,
I recently purchased a Autel MS509, also known as Auteltech. It has got live data recording and specs include Jaguar. BUT... If i fire up the car with the scanner connected it immediately indicates ABS and ACS error. After a few minutes of driving it lights up warning light for brakes, besides what's alredy is being displayed as MIL. And the speedometer stops working...! Otherwise the car works, but I can't be sure I'll ever be able to capture the intermittent failure that I am looking for since it will be difficult driving around with the tool ready for recording. As soon as I leave diagnostic mode in the tool and restart the car, everything is back to normal. Now the supplier tells me it is only compatible with cars from 2001 and newer.

 

Bo:
The fault codes are recorded, so that should not be a problem. Your data might very well be available too so that you can see trends, which is what live3 data does for you. Poke around in the configuration for the scanner. Anything that SLOWS the scanner might help, since your scanner is polling the network, using up too much of resources needed by the car's modules.

 

Thanks, Bo, great useful information. The MS509 was #2 on the list but it may go to the bottom of it (or off it) now!

I decided to try again with some very useful information received from Gendan about EngineCheck but #1 laptop had just been been nuked by a Vista SP2 upgrade (in fairness although I have a very low opinion of M$ products especially Windoze I cannot lay the blame on them - it was Norton interfering with the upgrade as usual as it interferes with everything) so I had to use #2 laptop and it only has a short battery life but what I saw from the O2 sensors output I didn't like the look of at all however it hibernated before I could save the recorded traces made during a test drive.

I may try again now I have resolved laptop #1 upgrade woes but what I really think I'm going to do is bypass the code scanner route altogether, dust off the Tek oscilloscope and look at the raw voltages coming from the sensors themselves and take the ECU and code scanners out of the equation altogether.

Jim

P.S. Bo, as I understand it Jags talk CAN amongst the modules but ISO OBDII for common diagnostics as required by the standard - if you have a CAN capable reader in like the MS509 and it is in Auto Connect mode (which tends to be the default) it is perhaps possible as sparkenzap has suggested that it may be flooding the CAN bus with connect polls at a critical time which may be disturbing the regular module communication (this communication is prioritised so a low priority message about e.g. bulb failures doesn't block a high priority e.g. ABS or engine management message). If your reader offers an option to set the communications mode manually (try ISO or ISO9141) you might give that a go and see if it improves matters.

 

Jima-
Correct about the communication to the OBDII being ISO9141, but the messages are apparently routed to CAN since regularly there are reports of CAN faults when you connect. One f these days I am going to sniff the CAN and see if I can figure out whats happening.

 

Well, I dusted off the old 'scope today and hooked it up to H2OS 1 bank A (upstream sensor, right hand bank) and what I saw was a nice flat line at about 4 volts, not the pulsing voltage between about 0.3 volts and 0.9 volts I was expecting to see. Conclusion: either the O2 sensor has failed/worn out/contaminated or the EMU is faulty and the input circuits are toast. The scanner reports OBDII diagnostics as saying that the O2 heater circuit has passed so I'm reasonably happy that the O2 heater switching/sensing is OK and not fooling the EMU into shutting down the upstream sensors.

I'd like to test the O2 heater themselves but I have seen warnings published about not doing that i.e. "When testing do not test the resistance between sensor pins 3 and 4 as the generated current can damage the platinum electrodes". Does anyone have any knowledge about this?

 

Jima- Terminals 3 and 4 refers to the zirconia oxygen element. Why they called it platinum, I am not sure of, but connecting a resistance meter to that circuit will damage it because the resistance meter passes a current throught he test leads to measure resistance.
The heater circuit is failrly low resistance (about 10 ohms??) on terminals 1 to 2, and I promise that it cannot be damaged by a standard ohmmeter