Two days ago I got a help call from a mechanic, he runs its own garage, I've been there once to help him out.
This time it was about an Opel Astra J, 2012, diesel 2.0, engine code A20DTH, automatic. Engine has been replaced, because original one was running in two cylinders (no further details here, but that would not matter anymore). After engine replacement, engine did not start, start motor did not crank, lots of warning indicators in the dashboard instruments and no communication with the engine ecu. That's all info I had over the phone. And some pictures.
I doubted at the beginning if I would accept the challenge, I normally don't play on cars >2008, due to lack of proper wiring diagrams and limited programming possibilities for various modules. Side note, I only play with Opels. However, garage being relatively close to my place, I went ahead, giving him the warning I might not be able to do much.
First meet with the car, few bits and plastics loose (I don't like this at all, I prefer the car to be in pristine status and work on plugs, connectors myself). First diagnostic try, had no communication with any module on high speed canbus. On low speed canbus, no problems.
Got some diagrams out and went to check the engine ecu power supplies, all fuses found in order, all supplies and ground to the ecu in order. Odd enough, with contact on, I could hear what seemed to be a throttle body motor being actuated prepared for starting. So it seemed that ecu was doing its job. I had less hopes about the happy ending of this diagnostic (did you ever had that 'feeling' of doubting own skills?).
Still had more ideas and went further and dig more for communication diagrams and learned that Diagnostic plug high speed canbus is connected to Body Control Module, then further on to ABS, power steering, headlamp leveling ecu. From ABS it went to automatic transmission and finally engine ECU. Where could the problem be?
To get a better idea where I stand, I went for the Pico CAN bus break out box, on the very first real use for me. Plugged it in and there you have it, CAN High LED pin 6 went blinking but CAN low LED pin 14 was always off. I managed to capture that in a picture.
Pico oscilloscope connected and that's what I saw. It didn't fit what I expected from a high speed canbus.
Obviously a problem on can low but can high didn't look any better. Next thought, can it be that BCM is defect? Does it act as a gateway? A quick resistance measurement over can showed 119 ohms, that must be spot on, right? This is where I made the first diagnostic mistake but I come to that later. To get to the BCM is not such a big problem and soon after, I was all over the plugs.
Continuity measurement for wiring can BCM to diagnostic plug done, all ok. Found that BCM is not a gateway and can lines continue further towards other modules. Can lines come in BCM at the light blue plug and go straight out at the white plug.
How to confirm that the BCM is bad or good? Taking him out of the circuit would be a bit difficult, after all, this is the central brain of the ignition, locks, doors, etc. With the plugs out, I measured further the resistance of the can lines towards the engine ecu and the good news, no good results. Kilo ohms, mega ohms all over the place.
Time to dig further. Power steering plug out, no change. ABS plug out, no change. ABS can lines to engine ecu measured, 120 ohms, very good. ABS can lines towards BCM, nothing. This is where the problem lies. Why and where the lines are cut? This is where I had the idea to do a measurement of abs can lines against + and -. And what do you know, one of the lines was shorted to ground. I could of measure it from the beginning at the diagnostic plug. But it gets more confusing, a short to ground would explain the can low waveform, but 60 ohms equivalent resistance should be more or less the same. Time to follow the wiring, what is going on here? Is there a headlamp leveling ecu? Lets find out.
Here is where my mechanic went diving in front of the left front wheel, under the bumper. He knew some plugs there and while the ohmmeter was still on, measuring the short to ground, after he disconnected two plugs, short was gone. Now its time for me to go diving, went straight on to discover the color of the wiring behind each plug.
And there it was. The magic mistake of connecting 4 plugs wrong. The canbus plug was connected to what it looked like ABS left wheel sensor (not confirmed, does it really have 7 wires?). This is what I could capture:
All connected in a correct manner and nice, some blue and red lights going on. Excited to see the can lines this time. Too bad in the picture I could not capture the lights on.
Before cranking, I always like to go ahead and delete all faultcodes from all modules. Then I start clean. This is one example of what I found, it just makes sense:
Finally engine started after 3rd try, it had to prime the fuel system. One faultcode remaining, cylinder 3 glow plug, and that was later confirmed to be defective. It was time for another cup of coffee, milk and suggar. I was relieved.
I was proud to take this picture, cannot be more happy with the Pico tools. Thank you Pico.
Looking back now, could of been done better, faster, easier? Maybe, but I do this for fun and not as a profession, for some, this might be peanuts, for me, this was a full war and nice taste of victory.
Liviu 2004 Very much enjoyed your well documented efforts with the Astra. Great to see PicoScope being used so effectively, and not just by professionals. It makes such a difference to include good photos, too. Well done, great job.