Diagostic Dilemmas: Lost in ‘The Diagnostic Woods’

By Gary Goms. This real-world case study of a 2002 Chevy S-10 pickup, a 1995 Buick and a 1995 Lincoln Town Car illustrates why “chasing” trouble codes can get you lost in the Diagnostic Woods.

Working as a part-time mobile diagnostic technician, I find that many techs find themselves “lost in the woods” while diagnosing engine management system failures.

Although not perfect, the “lost in the woods” analogy describes the dilemma of a technician through a forest of information without a good sense of direction. Unfortunately, most of us have found ourselves in that situation more often than we would like to admit.

To extend the “lost in the woods” analogy a bit further, let’s think of our test equipment as our compass and our technical information as our map. In most cases, my first step is to write down as much information as possible. Carefully interviewing the customer is the cornerstone of any successful diagnosis because the seemingly most insignificant scrap of information can occasionally prove to be the most critical.

Because scan tool data can easily be erased, my second step is to download and record trouble codes and their related freeze-frame data in a diagnostic notebook. While I’m at it, I poll all other modules for trouble codes and to verify communications.

For quick reference, I keep a written record of the vehicle’s VIN, mileage, license plate number, the repair order number, and the owner’s name and phone number in a slender reporter’s notebook, which I buy from a local office supply. Even in the age of laptop computers, having hand-written information immediately at hand for current and future reference is a diagnostic plus.

The Freeze-Frame Compass

Our first case study is a 2002 Chevy S-10 pickup with a California emission calibration that, according to my client shop’s customer, intermittently lost power during the long, 3,000-foot climb from Denver, CO, to one of its surrounding mountain communities. The 4.3L Chevy was in average condition with nearly 190,000 miles on the odometer. Fortunately, my client shop owner had recorded several P0300 misfire DTCs, fuel trim numbers and freeze-frame data in his notebook.

This real-world case study illustrates why “chasing” trouble codes can get you lost in the Diagnostic Woods. While the P0300 misfire codes were the only available DTCs, it was easy to see how cylinder misfires could be attributed to a bad ignition coil or distributor cap.

It didn’t take long to find my way out of the Diagnostic Woods when I discovered in the freeze-frame data that the calculated load value recorded only 60% at WOT and that the short-term fuel trim number temporarily spiked at plus 50%. The high fuel trim numbers suggested to me, at least, that the misfires were caused by the cylinders starving for fuel rather than by the ignition system breaking down. Of course, misfiring cylinders can dump a lot of raw oxygen into the exhaust stream, so I left the misfire scenario open.

The low 60% calculated load under WOT, high-rpm driving conditions was confirmed during a follow-up road test. The short-term fuel trims, on the other hand, hovered safely in the plus or minus 5% range. The owner reported that the fuel pump had been replaced several years before and the tech reported that the fuel pressure ran at specification during a previous test drive, so the road test results weren’t surprising.

But, fuel pressure readings can be misleading because they don’t exclude the possibility of an intermittent fuel pump or fuel filter failure. When analyzing data, it’s important to look at the Big Picture. Given the high spike in fuel trim, I could speculate that the S-10 had temporarily suffered from a fuel delivery problem during its long climb home.

Keep in mind also that the PCM basically calculates engine load data by analyzing input data such as mass air flow, throttle position and engine speed. At a very basic level, if the air flow rate doesn’t match the engine speed and throttle opening, the calculated load data will tend to be low. With that said, calculated load at WOT varies considerably among different applications. In this particular application, a calculated load value below 80% generally indicates that the physical volume of air flowing through the engine is insufficient or that the MAF itself isn’t measuring air flow correctly.

For that reason, I recommended removing the MAF sensor for inspection and cleaning and testing for exhaust back pressure. While I don’t want to get into the arguments for or against cleaning MAF sensors, let me say that I hesitated to recommend replacing the MAF at this point because I didn’t know where this diagnostic trail was leading. Given the high mileage of the vehicle, the 60% WOT load calculation might also be indicating that the extremely expensive catalytic converters were partially clogged.

Last, fuel filters are always suspect in my mind simply because they can intermittently clog from debris accumulated inside the filter shell. The plus 50% short-term fuel trim indicated that, at least for a very short time, the engine had starved for fuel. Although the truck had a relatively new fuel pump, I also know that many shops replace the fuel pump without changing the fuel filter. And, of course, a tank full of dirty gasoline can also clog a new fuel filter in a single trip.

See Photo 1.

Considering the Chevy’s high mileage, its low market value and its current state of disrepair, it would be easy to spend a lot of money only to discover even more expensive problems in the engine or powertrain. While very few, if any, MAF sensors can be restored to 100% efficiency by cleaning, cleaning the MAF allowed the PCM to record an 80% load at WOT, high-rpm conditions.

Because the catalytic converters tested a maximum of 6 psi at road speeds, a minor catalyst degradation and loss of power was indicated. But, without a P0420 catalyst efficiency code being present, 6 psi certainly doesn’t suggest an imminent catastrophic converter failure. Last, the fuel filter was a no-brainer because it was definitely clogged.

Whereas we could have gotten bogged down in chasing misfire problems, the path I took out of the Diagnostic Woods was a scenario of a truck low on power due to a neglected maintenance schedule suffering from a dirty MAF and a clogged fuel filter that temporarily leaned out the engine while climbing a high Colorado mountain pass.

See Photo 2.

The Data Stream Compass

Let’s look at a 1995 Buick with DTC 45 stored in the ECM’s diagnostic memory and an owner complaint of a foul-smelling exhaust pipe.

It’s easy to get lost in the woods over a DTC 45 “rich fuel” code. To illustrate, a number of failures can cause a DTC 45 in an OBD I General Motors vehicle. Although MAP sensor failures are relatively rare in these vehicles, the MAP can go out of calibration, or the vacuum line to the MAP might be loose or leaking.

I’ve had several cases in which the oxygen sensor had enough authority to cause the system to run rich due to low oxygen sensor voltage being reported to the ECM. Keep in mind that an open circuit in the throttle sensor ground can drive throttle sensor voltages high, which may also cause the engine to run intermittently rich. A sticking EGR valve can also cause the oxygen sensor to report a rich condition. The last and most common cause is a leaking fuel pressure regulator.

See Photo 3.

The key to finding my way out of this particular Diagnostic Woods was to perform a non-invasive test by recording the integrator (short-term fuel trim) and block learn (long-term fuel trim) numbers as the engine warmed up.

Starting from a cold soak condition, the integrator hung around 128 (center) after the oxygen sensor warmed up to operating temperature and began toggling. As the engine block warmed up, the integrator gradually began subtracting fuel. Most significantly, the integrator dropped from a normal 128 to nearly 120 as the engine reached operating temperature.

Based on scan tool data and simple logic, I suspected that the crankcase oil was diluted with gasoline, which was confirmed by sniffing the oil on the dipstick and observing a slightly high oil level. I might also add that if removing the PCV brings the integrator back to 128, then we know we have fuel contamination in the oil.

The next step was to pull the vacuum hose off the pressure regulator to check for raw gasoline. Instead of raw gasoline, I found that the vacuum hose had rotted away, which is another indication of exposure to raw gasoline.

My recommendation to my client shop was to change the engine oil, replace the vacuum hose, and check for leakage from the fuel pressure regulator and retest the integrator for negative fuel trim compensations. If there’s a lesson to be learned here, it’s to not make the simple more complex than it really is.

The Wiring Map

Due to the Great Recession, many owners are resurrecting some older vehicles by installing used engines. In this case, a local shop referred a 1995 Lincoln Town Car with a used 4.6L V8 engine that would crank but not start.

When the vehicle was delivered, the battery was dead and the condition could not be confirmed. Because the customer interview had yielded little of value, finding my way out of the Diagnostic Woods on this case was a little tougher than usual.

I’m normally very cautious when any vehicle comes through the door with a recently installed engine that won’t start. If the engine turns out to be defective, a lot of my own time and money is invested in a project that could lead to a diagnostic dead-end. But, proceeding with due caution, I decided to invest one hour in the initial diagnosis.

I installed a spare shop battery and tested the fuel injectors and ignition coils for battery voltage. With B+ established, I found that neither activated when the engine was cranked.

Next, I verified the fuel pressure to ensure that the fuel pump was activated when the ignition was turned on. Because the engine was foreign to the chassis, I also checked the crank position (CKP) and camshaft position (CMP) sensor resistance and made sure they were properly connected.

As I said above, test equipment is our compass and service information is our map. So, to understand how the Lincoln’s engine electronics were configured, I spent some time examining the wiring schematics.

The lack of coil triggering and injection pulse suggested that the CKP sensor was defective, which was a real possibility on a used engine. Fortunately, the most accessible test point for the CKP signal was at the ignition control module (ICM) located on the rear driver’s side fender panel.

In Ford terminology, the signal being sent from the ICM to the PCM is called the Profile Ignition Pickup or PIP. The timed signal returned from the PCM to the ICM is called the Spark-Out or Spout signal. The normal PIP signal is a 12-volt digital signal.

The lab scope produced the waveform shown in Photo 4.

As you can see, the CKP waveforms were erratic and floating far above the zero scale.

The erratic waveforms suggested to me that the ignition module wasn’t correctly grounded to battery negative. Testing even further, I recorded 5 volts on the ICM ground wire that ran through the main harness to the G 101 ground connection on the passenger-side fender.

Here again, I would have found myself wandering deeper in the woods had I assumed that the CKP or the ICM was defective. Instead, I checked the CKP pattern with a scope and confirmed my suspicions with a digital multimeter.

See Photo 5.

The existence of 5 volts in the ground circuit indicated a splice failure in the engine compartment wiring harness. Another part of our diagnostic dilemma, a splice chart for the Lincoln, indicated that the ignition ground was spliced to other ground wires in the engine compartment harness.

To avoid the time-consuming process of stripping the harness down for inspection and repair, I ran a jumper wire from the ICM ground to the connection point on the passenger fender. Another test revealed zero voltage on the ICM ground wire. As expected, the engine now started at the turn of the ignition key.

The way through this particular diagnostic forest was the simple act of looking at a wiring schematic and testing each wire for correct voltage or scope waveform. While a scope is handy, a simple digital volt-ohm meter could have been used to detect the bad ICM ground.

The diagnostic scenario I put together afterward was a ground splice that separated when the engine harness was initially removed. All too often, removing and replacing an engine explains many “mysterious” wiring harness failures.

Without the aid of a wiring schematic and splice chart, it’s entirely conceivable that a technician might have spent hundreds of dollars replacing the CKP, CMP and ICM — all to no avail.