This vehicle is a MY18 Chevrolet Camaro that is still in dealer stock, it has 60km on it and was delivered to the service department with MIL on and “Reduced engine power” message on the instrument display. These vehicles are LHD to RHD conversions which may or may not be relevant but suffice to say they’ve had major surgery before even being delivered to the customer. We scanned the vehicle for DTC’s and found the following.
P0090 and P00C8 both not current and relating to both circuits feeding the HPFP (High Pressure Fuel Pump) pressure regulator being open. At this point I reviewed the description of these circuits and the schematic of the solenoid. The description stated that when activated the ECM high side driver energizes the solenoid and the low side driver pulse-width modulates the low control circuit to ground. The ECM monitors the voltage on the circuits to detect a failure. After getting an idea on how the circuit works, I wanted to see how easy this fault would be to reproduce. We connected the scan tool and monitored the circuit data with engine idling. The engine idled from cold through to operating temp without failure but after a while we noticed the data values shown below change from OK to Malfunction which indicated the circuit was currently faulting.
We noticed the failure was sporadic and would only last a few seconds at a time then would return to OK. We tried gentle manipulation of the related wiring from the solenoid all the way to the ECM but didn’t find it to change the fault. We searched for any bulletins that might apply to this fault and found something promising referring to a couple of possible rub points in the harness that our circuits run through. This joy was short lived as we inspected the harness and found it to be very well insulated along the entire length. Due to the fault only being present for short periods I wasn’t confident with carrying out traditional intrusive circuit testing as once the circuit is dismantled there is no guarantee the fault is present while testing. The picoscope was connected in an attempt to catch the fault when it occurs without disconnecting anything. Referring to the schematic shown, channel A was backprobed into the high control circuit at the solenoid. Channel B backprobed into the low control at the solenoid and Channel C was the current of this circuit.
Once the scope and scan tool were hooked up we started testing and true to form the vehicle didn’t play up for over an hour. I was watching the scan tool the whole time so I could stop the scope straight away after the fault just in case it wasn’t obvious in the waveform. Finally we captured an event and then began the hard part (for me anyway) dissecting the trace and developing a plan. I first looked at the parts of the trace when the circuit was working properly and noticed the service information regarding the operation seems to be a little off (not an uncommon story).
It seemed the high side driver was PWM not the low driver. Once I worked that out the trace made a lot more sense. Just before an ECM PWM event the voltage rises to alternator voltage on both sides of the solenoid then the ground driver switches on and the current starts to rise. The high side driver is then switched off and on to maintain the desired solenoid position. We then looked at a bad example and immediately noticed no current flow through the circuit and very similar voltage readings on both sides of the solenoid.
This reduced the possible fault causes, ruling out the high side driver and associated wiring to the solenoid. It also eliminated the solenoid as if this was faulty you would expect to see the low side voltage at close to zero during the entire event. So at this point we worked out the remaining possible causes were intermittent open/high resistance in the low side driver wiring between solenoid and ECM or the ECM itself (yeah I know I also second guess myself when it ends up at ECM). We decided it was time to get intrusive and test pin fit and continuity of the low side driver circuit. Terminal tension tested good at all points but when continuity testing I noticed the resistance would change if I moved the test terminal at the ECM slightly which was odd as the terminal fit felt really good. I decided to remove the terminal to inspect and found this. The moisture seal was holding enough tension on the wire to keep the circuit together for the most part.
We soldered a new terminated lead, returned the vehicle and then submitted the warranty claim to inevitably get short changed for the time spent.
Bravo,great catch,usually the usual scenario with warranty you only get paid for the time taken for the repair.
Not all the time in research and testing
Many thanks Ben for taking the time out to create and share this case study.
They are so time consuming and often longer than the diagnosis itself!
I have no doubt this study will help others based on the fact that when the scope has captured a momentary event, you can pursue the circuit with confidence because you have proof.
Yes that's a good question, one I did ask myself after finishing it. It looked like a clean break at the thinnest part of the terminal just after the crimp. Not sure how though. The powertrain and engine harness would've been removed during the conversion process so the ecm connector would've been disconnected reconnected but not much else. The harness was secured to the connector body with ziptie so I don't think pulling on the harness would have caused it. Maybe the terminal was left weak after the crimp and broke on fitment into the connector would be my guess.