From Ireland. Diagnosing a TCM issue.

Come here to introduce yourself, who you are, what you do, and tell us your
experience with our automotive products.
Post Reply
Posts: 1
Joined: Thu Apr 29, 2021 2:11 pm

From Ireland. Diagnosing a TCM issue.

Post by gijoe50000 »

So, this is my first post here. I'm from Ireland, and I like messing with cars, electronics, and generally messing with stuff to figure out how it works. I was one of those kids who wasn't happy until he had disassembled every electronic device in the house just to have a nosey inside.

I bought a 2010 Jaguar X-Type a month or two ago for cheap, because it had a "gearbox fault" which just turned out to be a sticky turbo actuator, and an easy fix, but it was only after this that I found out that it also had a TCM issue.
After connecting up a diagnostic unit to it it said there was a TCM comms problem with the high CANBUS, and the TCM was gradually racking up errors, one every few seconds. The car is running OK but it has a MIL on.
I decided to go looking at the CAN network so bought myself a Picoscope. (Well, I originally bought a Hantek, but quickly sent it back when I saw how bad the software was!)

When I connected up I found that the network was perfect when the TCM was unplugged, but when it was plugged back in I was getting waves/glitches on the CAN line. I assume this is the cause of the faults.
These are some screenshots of the captures I got, with the purple being A-B and the pink is A+B, and red and blue are CAN high and low.
I think the A+B probably shows the problem most clearly.

This is the main gist of pretty much all the captures, some noise, and two different oscillations that seem to be always exactly the same size and shape, and about 3ms apart. I'm assuming it's some kind of short inside the TCM, or maybe a broken low pass filter/capacitor. But I'm not an expert in this area by any means.
I did have a quick look inside the TCM last week and it all looked fine, nothing clearly burned out, no water ingress, etc.
The TCM is the 7g9n-14c336-db type, it's on a lot of Fords and Volvos, it mainly controls the solenoids in the gearbox, and has an oil temp sensor in the gearbox harness (connected under the TCM to the gearbox), and it does all the other TCMey stuff too. 6-speed automatic.

And as I said, the TCM is actually working fine most of the time, gear changes are OK, except that I think it maybe changes gear at odd times sometimes, like it might downshift a second or two too soon, or this might be totally normal for this car, I don't have a frame of reference.

Or perhaps these waves are the TCM trying, and failing, to send signals, but I'm not sure that that's the case, since the diagnostic unit I hooked up seemed to show most of the TCM data is fine, solenoid current, shaft speed, etc. So the unit is probably working OK for the most part.
Termination resistors are in the instrument cluster and the ECM from what I've read and they're fine (60ohms on the CAN).

So, does anybody recognise these waves in the images, and know what they might represent?
Or maybe seen a similar problem in the past?
Or have any ideas of where to go next that doesn't include a Jag dealer?

I wouldn't be opposed to whipping the soldering iron out, but I'm a bit of a butcher with it, and it would be hit and miss whether I repair something or melt something else by mistake!
I do also, very rarely, get errors in the AC unit, or ABS with the diagnostic reader, but they seem to be minor issues, and most likely those units just resend their CAN messages if something gets messed up. At least that's my thinking as far as I understand how CAN works.

User avatar
Posts: 381
Joined: Thu Dec 13, 2012 11:26 am
Location: London, UK

Re: From Ireland. Diagnosing a TCM issue.

Post by FioranoCars »

Welcome to the world of Pico, had a Hantek too, groan, big mistake!

Sorry that no-one has chimed in to help you on this, which is a shame, as the data you've posted is pretty good and informative. I don't have any previous with this vehicle or gearbox, so this is just general thoughts, but my curiosity is peaked by the voltage variation in the can packets, and their relative stability per packet. The "noise" oscillations concern me less, although interested to see where that is being induced, as I think that is likely being picked up somewhere in the harness. It's net net zero to the ECU's understanding and is precisely why the CAN is twisted wire and opposing voltages, to allow noise correction.

First, do you have the PSData files to share?
and do you have captures with the TCM removed?
Do you have the network topology?

My suspicion is the varying voltages might be coming from different ECU (their internal generation chips are stable but offset verse others) ... and/or this might be internal to the ECU's or possibly poor chassis earthing?

However I am not seeing anything in the captures to shout that the signals are corrupt, you'd need to decode them to prove the CRC's etc are all correct and message format intact.

What is the Fault code in the TCM and are other ECU reporting lack of comms with the TCM? ie Engine or ABS, and do the PID's within those ECU reflect the gearbox state plausibly?
Does the TCM report PIDs from the ABS etc correctly?

Please post an update if you found/fixed this, as interested in all CAN issues especially in how different defects can be seen and identified from the signal, and or traced using the voltage offsets/current flows (need accurate topology and ideally several branch elements to split the current for it to be useful) ...

I think there is a fair bit out there in little nuggets of useful info but little that provides a comprehensive review and comparison ... but if anyone has recommended reading then be great for you to share!

Good luck and hope the reply is not too late.

Post Reply