You need to be careful about k-line on any T5, because they use a hybrid system that changed with time. I think that the driveline components may use a virtual k-line right from the start of production. If that's the case then they may not even have bothered to connect the k-line between those components. The driveline CANbus network is a different matter - you could expect to get serious problems if that were faulty.

Why do you think that a problem with the k-line would cause the ECU to drop the engine into limp mode?

 

Hi, Thanks for replying.
I think a sensor problem is probably causing the ECU to stay in limp-home mode.
Sensors are wired to the ECU.
The ECU has been tested and no-fault-found.
Diagnostics simply say 'Comms error' for ECU module.
The upshot is I cannot find out what sensor is faulty.
Most modules have CAN and K-line.
The BCM (On Board Power Supply CU in VW-speak) has no K-line connection but has CAN to the Dash panel insert. It all tests OK.
I have the full VW current-flow diagram with what looks like every variant by month and model (2859 pages of pdf)!!
Relevant pages show a hard wire connection between the diagnostic socket and the ECU - grey/white wire.
I think DTCs will be reported via K-line.
Using Picoscope on a breakout for the diagnostic connector shows that queries for each module are sent via K-line.
The ECU doesn't respond.
Using a camera probe I can see a grey/white wire going into orange T10h/2. I cannot manage so far to get a meter probe down into it.
I've tried making a probe with a needle and a bit of stiff wire but the spaghetti of other wires makes positioning the camera and aiming the probe difficult.
It looks as if the connector tray is meant to slide upwards for access.
Maybe a previous owner had a go and the tangle of wires makes that look risky.
I don't want to use brute force.
I'll have another try when the weather improves.
It's a bad idea to try vehicle repairs in the open on the west coast of Scotland.
Cheers.

 

I think DTCs will be reported via K-line.

It's a difficult one that has been debated here without any firm conclusion.

The main problem with seeing activity on the k-line and trying to interpreting what is happening is as follows: The K-line on diagnostic port T16/7 is connected directly to T32/25 on the instrument cluster, and also connects to a common point in the wiring loom (known as Connection B444 in main wiring harness). As you probably know, all of the diagnostic data that passes through the diagnostic port travels via the k-line.

And that's the problem - the k-line is a single data bus that connects all modules. You can look at it anywhere and you will see the same data being sent. But unless you have a data analyser capable of reading the address data, you don't know where it's coming from. That means that as far as I know, you have no way to distinguish between traffic sent to/from the ECU and the k-line traffic that's passing between your scanning tool and the CAN gateway - it's all just traffic on the k-line.

My gut feeling is that once the driveline components moved to CANbus for inter-module comms, the k-line traffic was transferred over to become "virtual k-line" over CAN - the CAN gateway being responsible for translating that into k-line on the diagnostic connector. There's some evidence of that in VW's SSP311 - it shows a significant number of control modules as having no physical k-line at all.

 

Thanks triffic,
I've now got hold of SSP311, pages 84/85 show the hardwired and virtual wire setup.
I'll also look at SSP186 and SSP238.
My model is very basic; manual windows, manual gearbox, simple steering wheel etc.
You could be right that I'm concentrating on K-line T16/7 >>> T94/72 and it isn't the problem ... but they should be connected.
So I will try to find why they aren't ... just so I can re install the battery etc.
I can then re-run VCDS and look at some waveforms.
Attached is a pic of the harness just before the E-box ... gives an idea of why I'm chasing this.

Solid rain forecast for next 2 days so it will have to wait.
Cheers.

 

pages 84/85 show the hardwired and virtual wire setup.

It does, but unfortunately the description that covers the various combinations of connection doesn't really make sense (or at least not to me). It probably did in German, before it was translated...

Do you only get "Communications error" with the ECU, and not the ABS controller? If you get problems with both, that would suggest a potential break in or disruption of the driveline CANbus wiring - though of course it's not a simple matter to pin down exactly where that problem might be.

Is that rodent damage to the loom? Whatever caused it, that's a frustrating position to be in. Good luck - I hope the weather cheers up for you soon.