I'm looking at a potential network issue on my '06, not sure if this is a problem or a red herring. Network termination resistance is spot on 60 ohms. No shorts to battery or ground. Some packets on the CAN bus look perfect, 2.5 to 3.5v CAN-high and 2.5 to 1.5v CAN-low. But when I zoom out, I see some packets are very noticeably above spec. I haven't been able to capture one on the scope in single shot mode, need to play around with the scope settings more to figure that out.

But before I spend more time on this than necessary, is this normal?

Thanks,
Michael

 

To judge the can signal on this level, you typically require a differential probe. The spikes are not unusual for an setup without diff probe.
If you attach a common can interface to monitor can traffic, do you get errors (like CRC) ?
What made you suspect canbus to have problems ?

 

Good point on the CM probes. I did not get any errors while probing. Just to clarify--not concerned about the spikes on the squares, but rather the first packet group is almost 1v above 3.5v. See yellow arrow below.

Issue is an odd comm code. When I start the car cold, the temp gauge goes to the previous known value (190F) and then slowly drops back down to current engine temp (video link below). The IP has set U2199 code for "invalid data for engine coolant". The diagnostic tests for this just have you go down confirming CAN network wiring integrity.

The engine coolant temp is always available when looking at the OBDII data, so it is something about startup, but not related to voltage drop. If I turn the car on but don't crank, gauge works normally. Car had this issue beginning 10 years ago, intermittent at first, then became constant. Been through several batteries, different temp sensors, new cluster, checking grounds, etc. My thought was CAN bus was becoming corrupt during cranking due to voltage drop, but looking at the scope I can see it was completely steady in this test. Only thing that looks odd to me is that some packets are at a higher peak level.

 

I had some trouble with failures related to the canvas, and I did pretty much what you’ve done, and I couldn’t see anything really wrong, but I noticed that the twisted pair of wires coming out of the harness head somewhat long and twisted length going into the connector at the modules, and I thought maybe noise was coupling onto those wires at that untwisted point and causing the errors. so I started twisting them nothing fancy just took my two fingers and rolled them together so it was twisted in one direction for half the distance and then twisted in the other direction the other half of the distance and surprisingly it worked. Give it a try and see if it helps.

 

normal noise this is why twisted pairs exist in the first place.

i've only ever seen that weird cooling gauge behavior on my x206, on a cold start mine sometimes starts in the middle and then corrects itself. must be some sort of issue with the early panpag ecms, check engine temp over CAN in the cluster datalogger. short of that replace ecm or give up and do something better with your life.

 

Digging some more in the service manual, I agree that the issue sounds like the ECM and not the CAN bus or cluster. There are two different definitions for U2199 given, one I had not seen before which says the IP "received a message with default data for >1 sec". So it's receiving a packet, but it knows the data is a placeholder instead of current. The second code definition makes it sound like the IP received a corrupt packet.

I suspect it would take equipment that could read & decipher Jag CAN bus packets in order to diagnose this, and then someone familiar enough with the ECM programming to determine what's going on with that. Basically I'd need a Jaguar engineering team circa 2008 to diagnose and fix this without just throwing parts at the car.

I will not replace the ECM for this nuisance gauge issue.

There are a few other simple things I will check: #1 the ECM ground, #2 clean and reseat the ECM connector, and #3 I want to see if the IAT/ECT/FRT sensors are close to agreeing on a cold key cycle. I know in the past my car had issues with the fuel rail retaining pressure overnight, I wonder if that could be throwing off the fuel rail temp sensor readings. But there are no issues with the car completing drive cycles, so it does have the ability to detect a cold start.

 

I too have noticed on my 2006 that the temp gauge occasional is not "zeroed" at first start up but gradually drops to cold after about 15-20 seconds then resumes normal functionality.
Only happens occasionally - perhaps 1 out of 20 starts currently.

I was thinking that the cluster CPU perhaps was occasionally not fully implementing the reset command at the start of key roll, or in the post shut down sequence was not clearing the outgoing temperature reading ready for next engine start.

Haven't gone looking for that issue....as it doesn't seem to be causing any problems at this stage and I have more pressing jobs to complete on other cars.

 

That is how mine started out back around 2013... I only noticed it would happen once in a while. Then it started doing it every time. Never had any affect on the running of the car, just bothers my OCD. Actually glad to know it's not just my car that does this. It could very well be components in the ECM that are aging (e.g. electrolytic caps), or maybe it's just poor programming. I've found a few other niggling errors from the FEM that I chalked up to bad code on Jaguar's part.

I did check my temp sensors yesterday, all report within 1 degree of each other cold. When I have time I will pull the wiper cowl and check the PCM connections.