I have had a 2015 Camaro in the shop for about a week now with an intermittent problem. The customer purchased the car from another dealer who can't repair it and told the customer that the failure would have to become more obvious to be diagnosed. She called Chevrolet to complain and the zone told her to bring it to us.

The car has 4000 miles with a problem that is very intermittent. To duplicate it, the customer states that the vehicle has to be left overnight in cold weather and, directly after starting, needs to be driven very rapidly across railroad tracks. When these conditions are met, the red brake light on the dash will flash on and off 1 time and will not re-occur again until sitting overnight again in cold weather. We're in Florida and cold weather is not easy to come by.

We gave her a loaner and told her that we needed the car for two weeks. The first 4 days were warm but I tried driving the car anyway. We have a railroad crossing about 1/2 mile from the shop. The problem didn't occur and no codes were set in any module. (They may have been cleared by the selling dealer)

The next day turned cold and I drove the car over the tracks almost fast enough to launch it. The red brake light did flash 1 time....Wow. Rescanning the modules, the BCM set a history code for lost communication with the EBCM...A CLUE. Well, that takes a lot of possibilities out of the picture. It sounds like a data transmission problem. I hooked up a laptop to the DLC to monitor the LAN with the GM testing tool and let the car sit for 4 hours then tried the tracks again. This is the main high speed LAN on terminals 6 and 14.

[2015 Chevrolet Camaro LT, ECM/Inputs/Outputs Waveform]

The light flashed again and this waveform shows no obvious problem. The EBCM resides on two LANs, the main LAN and the chassis expansion LAN. A com problem on either LAN can cause communication losses with the BCM. I let it sit overnight and tried it again with the LAN tool watching the chassis expansion bus on terminals 12 and 13. The light actually flashed twice this time and I see why.

[2015 Chevrolet Camaro LT, ECM/Inputs/Outputs Waveform]

I would expect to find a loose wire or terminal. Temperature can cause expansion or contraction as well making the condition worse. The chassis expansion bus leads from the EBCM through the firewall to the steering angle sensor and the multi-axis acceleration sensor. How do we decide what to test? The LAN is a two wire twisted pair. Looking at the failure, both LAN+ and LAN- dropped out and resumed exactly at the same time. What's the odds of that? Not very good.

The EBCM could be losing power or ground, but in that case, I would expect the multi-axis sensor to have stored a communications code as well. What's next? Well, all high speed modules need a independent communications enable 12 volt circuit for wake up. The EBCM com enable circuit is shared with other modules through a splice connector, J202 but is fed independently through one branch that goes through connector X104, terminal 1 under the hood. I back-probed terminal 1 with a voltmeter in min/max mode and drove around the lot hitting speed bumps as hard as I could and I saw drops from battery voltage on my meter over the bumps. Good - CLUE 2.

The J202 splice connector is in the dash wiring harness just below the inside fuse block. I was able to reach up under the dash and squeeze and wiggle the harness in this area while watching the waveforms on terminals 12 and 13 and saw drop-outs occur. I cut open the harness to expose the splice connector, circuit 5986 and cut off the splice tape. I found bare wires lying side by side without a crimp splice installed, just splice tape. Nice factory wiring job. I soldered the splice and retaped the harness. The next two days, I drove the car under similar conditions and saw this....

[2015 Chevrolet Camaro LT, ECM/Inputs/Outputs Waveform]

The vehicle is repaired and gone and the customer is overjoyed. The threat of a buy back is eliminated and we grew a new customer.

James from Florida

Files Referenced: