Another shop called me a while back. They had a PCM that was generating high voltage codes for both of the APP's, one of the TPS's, the iAT (plus a few more). I had them test and the PCM wasn't supplying a ground for all of these inputs. By temporarily bypassing the PCM's ground the vehicle functioned normally. Once they proved that the PCM was bad they had me program the replacement and set up the security. I stressed concern that something didn't feel right and this may well have been caused and didn't just happen. They assured me that they looked and didn't find anything.

A week later they called back, same problem.

Upon arriving I retested and confirmed that the ground circuit had failed inside the PCM. Opening it up, it was easy to see the trace that was burnt off of the board. It was also easy to repair with a jumper soldered into place.

That left the question as to "why"? Worse yet how can we find it without tearing the whole car apart. It turns out this ground path affected five ground circuits. Somehow power was being applied to one of them. To find it efficiently was going to take some creativity.

All five grounds involved were labeled and removed from the PCM connector. They were then jumped to ground through a fused jumper. Now this would work with one current probe but it would be more tedious and would require first testing a group of three and then the other two in hopes of capturing a current spike. Having three current probes allowed for the following routine.

The grounds were numbered 1-5.

Grounds 1,2,3 were captured with one current probe connected to channel A.

Grounds 3,4,5 were captured with one current probe connected to channel B.

Grounds 2,3,4 were captured with one current probe connected to channel C.

Making a logic table a spike on only channel "A" was ground #1, on channel "B" was #5.

A spike on two channels "A & C" was ground #2, and "B & C" was #4.

A spike on all three channels was ground #3.

The spike occurred on B and C indicating ground circuit #4 which was the AC pressure transducer. The idea of course at this point was to identify the affected circuit so that only it had to be closely examined. It was notable that the spike only occurred when hitting bumps, under load with the headlights turned on. What was found was a parking light feed wire that was sticking out of its harness and was touching the pressure transducer harness where it exited the engine harness. These two wires, the parking light and the AC Pressure sensor ground both had a tiny spot abraded where one strand of the wire was visible and when they touched was putting 12v into the sensor ground circuit for the PCM.

I hope you never need this routine, but if you do I hope it helps.

John from Pennsylvania