Volkswagen Golf 2005 - Communication fault

Hillsautoelectrical · Post by **Hillsautoelectrical** » Tue Mar 31, 2020 2:56 am

Hi Everyone - first time user of Pico and first post....

05/2005 Volkswagen Golf 1.6L BGU 91000km

I'm posting this to see if anyone has some input on this intermittent fault we are having with the high speed CAN line.
Vehicle was driven in to us from another workshop to which vehicle was towed too.
Went to bring into Workshop that night and vehicle would crank but not run.
Next day the same - connected scan tool (VCDS) - found not able to communicate at all with any module on the high speed CAN bus(Engine, Transmission, ABS, Power Steering, SRS) but convenience CAN bus OK (Gateway, Doors, Sound system etc..)

Tested fuses OK - then went to check scan again and communications had returned - vehicle started and ran OK
Checked codes logged in Modules - See Scan VCDS TXT File

Then Connected Pico to both CAN bus lines at the Steering column Module - Test Drove and captured this waveform when vehicle faulted - Will go into Limp Home...
Was at one stage able to make this fault fairly consistently going around a hard left hand turn or heavy undulating ground, but now i have found this intermittent and can drive for a week with no fault.

: CAN

Not sure if it is possible to export an interactive section or one frame of the buffer from Pico 6???
If someone can point me in the right direction for that? Thanks

Now to me it looks like both CAN H&L are being dragged down to ground at the same time.
Now my thoughts are, where to start looking?
Is it a module dragging down the CAN line?
Or is it a wiring loom fault - I am currently leaning towards this, but would find it unusual to have both data lines shorting to grounding together - I would suppose that would be possible though..
Have done a wriggle test on any accessible wiring looms, but not able to replicate fault in the workshop

Any thoughts or ideas on this are appreciated ...

liviu2004 · Post by **liviu2004** » Wed Apr 01, 2020 5:34 am

If you ground can low it will pull can high also, as in each node, there is only one voltage regulator! Try on bench on simulated network and see for yourself. Which module sits under water?

Hillsautoelectrical · Thu Apr 02, 2020 12:16 am

Hi and Thanks for reply.

I have done the experiment ... see image
CAN H short to ground is on the left.
CAN L short to ground on the right.

Also for the purposes of education i experimented with other scenarios
Here is image of both CAN H and L shorted together and as expected both flat line at 2.5V

Then also Both CAN H and L shorted together and grounded.

So from this it is plain to see that we have an intermittent grounding of the CAN High wire..... somewhere!
I'd suppose it could still be a module..
Will isolate and check each module and check for water damage

Thanks

liviu2004 · Post by **liviu2004** » Thu Apr 02, 2020 6:51 am

Well done. On my experiments, grounding can low pulled can high also, bit more then you have.
Your simulations are gold, keep in mind you’re doing full ground connection, while in the vehicle, there will be a bit of resistance playing up.

Hillsautoelectrical · Post by **Hillsautoelectrical** » Fri Apr 03, 2020 3:46 am

YES SUCCESS!!!
It pays to be persistent even though it can be frustrating......
Well I started with exposing all related modules and disconnecting some and inspecting connectors - No obvious faults found..

While doing this i was monitoring the CAN signal with PICO...Setting the trigger to single and monitoring Channel "A" with the trigger set to about 0.5V. This way i would be able to see if something had happened while I wasn't looking at the computer
Then i started to inspect the wiring loom and examine any areas that may be in any way close to metal grounding edges or points...
As always you generally start at the extreme end of where the eventual problem is found!....
Checked engine bay loom then moved inside to passenger side loom, pulling more and more of the car apart as I went...

Then moved to passenger side to gain more access to SRS module, when suddenly realised that PICO had recorded a trigger.
Reset PICO trigger to auto and started to manipulate the area I had been working on... Finally!.... was able to replicate fault in the workshop.

It was extremely touchy.. I only had to move the main wiring loom that ran behind the main dash mounting frame the slightest amount to recreate the problem...
I thought at first that i would have to remove the whole dash board to gain access to this area, when we realised that behind the wiper motor assembly was a large plastic plate with a wiring loom from the nearby Engine ECU running though and inside the vehicle. See photo.

As you can see in the next photo, you can just see the Orange/Black CAN High wire slightly exposed running right past a sharp edged dash frame bracket.

After exposing this wire, here is the best picture i could get of the slight cut in the wire

So there you have it.
Obviously there is an oversight on VW's part in protecting the wiring loom in this area. Every where else the wiring loom is well protected...

Keep in mind that this is a right hand drive vehicle (Australia)

This has been the first paying job for PICO and it has helped. So great work PICO team!

Now to put it back together!

PicoScopes
& kits

For vehicle manufacturers

EV kits and
accessories

Noise, Vibration and Balancing

Milliohm tester

Pressure sensors

Current clamps

Breakout
leads/boxes

Accessories

Automotive Forum