Approach intermittent faults with logic, diagnostic curiosity

Dec. 27, 2016
I don’t get scared anymore when the word “intermittent” is included in the problem description. I approach these situations as logically as possible.

This article could be aptly named “Compounded Human Error, part 2” (see”, August 2014). In this case though, I was one of the people that erred while diagnosing an intermittent charging system warning indicator on a 2009 Chevy Cobalt LT.  Yes, as hard as it is to believe, I make mistakes (ahem). I am indeed human after all!

What’s typical in a repair shop is not necessarily what a mobile technician sees. In fact, don’t you “stationary” techs and shop owners handle the typical situations? I’m usually not called to help until the shop’s exhausted most — if not all — of the ways to resolve their customer’s vehicle complaints. In some cases though, a shop will call me in because they are so busy with “gravy” jobs — and they also have a job that can potentially tie up a technician’s time in ways that aren’t as productive. This car’s problem happened “every so often” and the shop was very busy already, so I got the call.

2009 Chevy Cobalt LT; customer complaint: battery indicator illuminates intermittently

The nature of intermittents

The very nature of an intermittent fault implies that a considerable amount of time may be spent waiting for — or trying to recreate — the conditions in which the fault is present (and can be diagnosed). To an inexperienced or an impatient technician, they can bring a type of anxiety that may cause that person to jump to conclusions or to misdiagnose. I know this from personal experience. I don’t get scared anymore when the word “intermittent” is included in the problem description. I approach these situations as logically as possible.

I can’t recite an exact number, but I can say a majority of intermittent faults I’ve dealt with were related to a previous repair. If the customer is in your shop for the first time, do you ask questions about what may have been done previously? Do you strive to have repeat customers? Isn’t it great when you can look up the repair history on a vehicle when that customer returns for more work? Don’t you just love when you can verify which services and previous repairs had been done using your service history? Believe it or not, this is one of the most powerful tools a shop has when it comes to diagnosing problems.

This car had no history. That is, it had been sold at an auto auction a few months prior to being sold again by the used car dealer who called me in to look at the charging system problem it had.  The Carfax the dealer provided the customer was spotless. The dealer’s technician notes about the pre-delivery inspection performed prior to the most recent sale were minimal. Basically, I was without one diagnostic tool already.

When the repair/maintenance history isn’t available, I try to put all my senses to work. In addition to all the other diagnostic tools I may need, I have to be very observant. I have to listen like I’m in a quiet room, smell like I’m near a bakery, see things (not just look at them), etc., etc. I must be The Car Whisperer (also my company name, for those of you not in the know)!

AC pressure sensor location

In addition to getting what little history was available, I asked all the questions I could think of about the problem before approaching the vehicle. The tech had verified the alternator continued to charge when the “battery” indicator was illuminated. The battery had passed tests — both a conductance test and a load test. It was unknown whether the indicator would illuminate only after the engine warmed up or if it was lit when first starting as well. I found out it didn’t matter if the car was moving or not and it sometimes happened on turns, other times not. The tech did little more than this because he was eager to tackle the many jobs the shop had taken in. I was on my own from this point going forward.

Where would you begin?

You now have the information I had. Where would you begin your diagnosis? Do you approach all your diagnostic dilemmas with a logical, systematic strategy? I believe we must use a logical approach to properly perform diagnostics. My first step was to make sure I understood exactly how and why the car battery indicator would be illuminated.

I’ve gotten in trouble before by assuming the car I’m working on today is diagnosed identically as the same model car I worked on in the past. I could have started my diagnosis assuming the charging system on this car is controlled the same way as the one on the last 2.2 liter GM I worked on before — but I know better. The slightest engineering difference could lead to a complete misdiagnosis, which benefits no one. That type of short cut sometimes costs you in wasted time and money!

I read the “Description and Operation” of the charging system and found on this vehicle, like many, the alternator is controlled by the Engine Control Module (ECM) and the Instrument Panel Cluster (IPC) controls the charging system malfunction (“battery”) indicator. The two modules communicate with each other — and with other modules — through the vehicle’s networks. The ECM commands how much the alternator should charge based on the demand calculated by the Body Control Module (BCM).  The BCM receives information from the Battery Current Sensor located on the battery negative cable. The sensor measures how much amperage is flowing into or out of the vehicle’s battery.

Chafed wiring insulation Engine harness pigtail break-outs

I like to look at the wiring diagram(s) too when doing my initial analysis. In this case my research was done using Mitchell 1 wiring diagrams. I prefer these because they’re in color and are what I call “interactive” in that when you hover your mouse over a circuit displayed, that wire is made bold, which makes it easier to follow through the diagram. I made mental notes when I reviewed the diagram. Now armed with knowledge that applied to this vehicle, it was time for me to actually get near it!

Next I performed a visual inspection, which is where I sometimes find the cause of the customer’s complaints. Visual inspection is often a skipped step in lieu of something more technical. Many times I’ve found something obvious when called to a shop after the technician had spent countless hours analyzing scan data, digital oscilloscope patterns and sometimes had pinpoint tested ECU connectors. Has this ever happened to you? We have all been guilty of over analyzing problems and vow to never overlook the obvious things again afterwards.

Ground on a painted surface

During my visual inspection, I looked at the underhood grounds, the ones I’d made mental notes of while looking at the wiring diagrams. I noticed some had been painted; one looked as if it was attached to the painted surface of the inner fender, and I saw paint overspray on the headlamp bezel. “Hmmm, this car’s been wrecked,” I thought to myself. One more item was added to my collection of analytical data. Note that this is the stage of diagnostics I was in: data collection. I wasn’t anywhere near reaching for a tool yet.

What’s my next step? Should I get my DVOM and start checking for a voltage drop on those grounds? Should I load them while testing? Should I be taking a wrench to them to see if they were tight enough? STOP! No, these are NOT the most efficient steps to take at this point (and may actually impede an accurate diagnosis)! Resist the temptation to fiddle with something yet. I continued my visual inspections looking for oil leaking on the alternator or its belt, aftermarket electrical devices installed and for anything that just didn’t look right. I didn’t find anything else.

Graphed data - fault captured Graphed data - fault not present

I then wanted to duplicate the customer’s complaint. I turned the key to the RUN position and observed the IPC carefully, watching for bulb prove-out(s). I confirmed all the indicators, including the battery, worked as designed so I started the engine. Of course, as it usually happens with an intermittent fault, the battery light went out and stayed out, as it was supposed to when everything is working correctly. Since the problem was not occurring at this time, I left the engine running while I started gathering the tools I might need.

Time to ask the patient a few questions

After shutting the car off and installing a scan tool, I restarted the engine and scanned the vehicle systems. I perform network tests in situations like these because there are so many modules involved. It saves time when doing a Diagnostic Trouble Code (DTC) check of all modules, especially when multiple ECUs report a similar failure. In this case, only the ECM reported any alternator-related codes and that was a History DTC. The Supplemental Inflatable Restraints (SIR) confirmed my suspicions based on my observations of the painted surfaces under the hood. I also found a code in the radio, but neither it nor the SIR DTC was important to my diagnosis of the battery light, or so I thought.

Okay, I have a History P0621 in the engine controller so I looked up the information about the code while I’m waiting for the indicator to illuminate. Once again, I want to be sure the diagnostics I use are applicable to this car. I read the DTC Descriptor section in Mitchell 1 to better understand why the DTC set and what I will need to do to diagnose it. Basically, the ECM sends a signal to the alternator to control its output and on another circuit the ECM monitors a voltage. If the engine’s running, there should be high voltage on the circuit and if not detected by the ECM, it will signal the network modules to turn the charging system fault indicator light on and it stores the P0621 DTC. Please tell me, what is considered “high” voltage?

The troubleshooting chart is relatively simple for the code; there are only four steps. Unfortunately, where do you think the chart will lead if the fault isn’t present while I test? While the charging system operates as designed, I wiggle electrical connectors, tap-test the ECM, revved the engine, let the engine idle, and I drive the vehicle while monitoring the voltage on the circuit. Never did the battery light illuminate. I checked the ECM calibrations to ensure they were the latest available (they were). So, what would you do next?

I suggested we let the problem become more apparent, happen more frequently and test it while the light was on. The shop felt otherwise. Since their tech mentioned he thought the ECM was at fault, they picked up a used module “in case” I found it faulty. Remember what I’d mentioned earlier in this article about jumping to conclusions? They asked me to replace the ECM, reprogram it for use in this car and to perform whatever functions had to be done when the ECM is replaced (there were four, in addition to Vehicle Theft Deterrent (VTD) relearn). When I was done, everything worked properly.

The next week, they called me back. Yes, the light was back on and now it’s happening more frequently. When I looked at it this time, in addition to the complaint about the “battery” light being illuminated, there was a new situation: The air conditioner was not blowing cold.

A simple mistake?

Did you ever get the feeling there was just something you were missing? That there was something you should have picked up on?  These were the feelings I was having when they called me! In addition to the alternator DTC, there was another ECM History DTC, P0641, for the A/C pressure sensor voltage reading low. When I researched that code, I saw there was a problem detected in a 5 Volt Reference circuit, which included the A/C pressure sensor. It seemed the car was getting “sicker” the longer this went on!

Since this new development wasn’t present before the ECM replacement, coupled with the fact it was a used ECM that was installed, once again the shop owner jumped to the conclusion that it must be a bad ECM and ordered another. I convinced the shop owner to let me diagnose where the faults reside prior to module replacement. Begrudgingly, he allowed me to do so. This is where I admittedly, screwed up. No, the mistake was not in wanting to diagnose, but in how I did it.

When I focused on the original problem, the Alternator “L” terminal Voltage DTC, the P0621, and ignored the A/C problem, I was perplexed with how much voltage was too much, or how much was enough voltage, when the test was to check if the voltage was “high.” I decided to look at the OE service information to see if that would help. Indeed, I was to measure the key on, engine off voltage on an orange wire, the “Generator Turn-On Signal” circuit, and see if it was between 4.5 and 5.5 VDC. Would you, dear reader, have interpreted in the original instruction 5 volts to be the deciding point between “high” voltage or not? I have learned to not settle for ambiguity when diagnosing, and I find the extra expense of the additional subscription is more than worth the cost!

So, I needed to measure voltage at the alternator, and if not within specification, measure at the PCM. My mistake was in how I back-probed the PCM Micro-64 connector. I’d improperly placed my back probe in the ECM connector, made contact with the wrong circuit, and I read a different voltage at the ECM than what I read at the alternator. It made no sense to me that there was a different voltage, especially since there was nothing else on that circuit. It was past closing time by then and based on the results of my tests, the procedure indicated the illumination of the battery light was caused by the ECM. I asked the shop owner to have his tech replace it. Can you guess what happened when they test drove the car after installing the replacement module?  Everything was fine! I cleared codes and sent the car down the road. Uh, that was mistake No. 2, right? I mean, shame on me for not including diagnostics of the Air Conditioning complaint at the same time!

Five DTCs now

Yes, the car returned again. This time there were five ECM DTCs stored, all having to do with the same 5 volt reference circuit!  I graphed the 5 volt reference data PIDs and saw an intermittent “drop” to ground, only when the air conditioner was on. Once I took my blinders off and started looking at the whole picture, I noticed things in common! Instead of focusing, I took a step back and looked at the whole picture; I looked at the problems with a different perspective. I remembered seeing grounds painted over, grounds installed on painted surfaces and the SIR DTC that indicated the vehicle had been in a front-end collision. I also heard from the driver after the last repair that the light stayed off until he accelerated out of a turn. All these clues made me take a closer look at what might have been affected either during or after a frontal collision.

I traced the alternator wiring to a juncture where about three other sub-harnesses split off the main engine harness. One of those was the A/C pressure sensor! Another was one of the sensors with a 5 volt reference “low” code, etc. I looked at the whole harness where it was routed along the front of the engine until I saw where someone had placed a heater hose on top of the harness between the oil dipstick and the block. That hose didn’t belong there and was being stretched in that location. Once the hose was properly located, I saw the convoluted tubing that protects the engine harness had been worn through by a sharp edge of the block, just rear of the engine oil dipstick! By lifting the harness, I saw bare copper wire where the insulation had worn through on two of the wires in that harness, one being the 5 volt reference wire, the other being the orange alternator wire!

Harness modification PCM comparison (used module)

Yes, compounded human error played a part in how this car was not only resisting being fixed, but instead was developing more problems the harder I tried to repair it! The last clue is what led me to look closely at harness routing. When the driver said he made a turn while accelerating, I suspected the harness was rubbing against something and the insulation had worn through to the conductors. The body shop tech had misrouted the heater hose, which put pressure on the engine harness, which eventually rubbed through causing circuits to short to the block. Their human error, and mine, made this job a bit more difficult than it needed to be (but it’s fixed now).

Oh, and yes, I did address the grounds that were questionable, “just in case!”

PCM post-programming configuration options

Sponsored Recommendations

Best Body Shop and the 360-Degree-Concept

Spanesi ‘360-Degree-Concept’ Enables Kansas Body Shop to Complete High-Quality Repairs

ADAS Applications: What They Are & What They Do

Learn how ADAS utilizes sensors such as radar, sonar, lidar and cameras to perceive the world around the vehicle, and either provide critical information to the driver or take...

Banking on Bigger Profits with a Heavy-Duty Truck Paint Booth

The addition of a heavy-duty paint booth for oversized trucks & vehicles can open the door to new or expanded service opportunities.

Boosting Your Shop's Bottom Line with an Extended Height Paint Booths

Discover how the investment in an extended-height paint booth is a game-changer for most collision shops with this Free Guide.