The importance of asking 'why?' during automotive diagnostics

Dec. 3, 2018
Sometimes we get lucky and the cause of the customer's concern is readily apparent after a test drive and a review of some fundamental PIDs. And then, other times — well, you know!

Content brought to you by Motor Age. To subscribe, click here.

What You Will Learn:

Computers aren't smart; they simply respond to inputs

• Always reference wiring diagrams as part of your diagnostic approach

• Ask "Why" at least five times, to ensure you locate and repair the root-cause fault

OK, So I live near the Philly area, born and raised. True, my mom is from Jersey City and dad, from Brooklyn. My friends across the country poke fun at the way I talk but hey, let’s face it, I am a “WHY’s GUY,” through and through (Yep, I spelled that correctly). It’s not what you think, though. When it comes to troubleshooting drivability faults, sometimes we got it made. We simply drive the vehicle and analyze the data we capture, looking for the root cause of the fault presented at the time. We invest time learning how to understand the relationships between the PIDs and how they reflect underlying faults. We all know, it ain’t that easy all the time.

Why ask why (or when or how)?

Sometimes, It’s a bit more involved than that. Sometimes we get that vehicle that requires us to dig way deeper than we are accustomed to. Because I spend a lot of my day troubleshooting drivability faults, to remain efficient I had to devise a plan of attack, a strategy. This strategy involves analyzing preliminary data (the low-hanging fruit). I use this data to ask questions of the vehicle, like:

  • Why are you performing poorly? (Are you lacking sufficient fuel supply, can you breathe?)
  • When do you perform poorly? (Is it when you are idling, or when I force you to work hard?)
  • How can I get you to reveal your fault to me? (Are there certain weather conditions you don’t like?)

Questions like the ones above help me decide which road I will head down. More importantly, it gives me a sense of direction and justifies the next test I will perform. Otherwise, we’d just be shooting from the hip. I do get lucky every now and then, and diagnosis falls in my lap, but I’d take a rock-solid game plan over luck any day of the week.

The bum

The subject vehicle of today’s topic is a 2002 Buick Park Avenue with a 3.8L (K) engine and just shy of 130k miles on the odometer. The customer is concerned with the vehicle’s lack of power and a “knocking” noise from underneath the hood. The vehicle was in the shop in the recent past for replacement of spark plugs and ignition cables and had been without fault for at least a few months.

I began the analysis with a scan for DTCs and, to my surprise, none had been stored. I drove the vehicle, monitoring some basic PIDs and within a very short distance, the vehicle began to ping hard and lacked power. This is where the questioning began. You can see that I had placed the vehicle under heavy acceleration (Figure 1).

Now, at this point, the vehicle wasn’t pinging, but the goal was to see if the vehicle had adequate fuel supply. Ruling out what is “good” with the vehicle is equally as valuable as discovering what is faulty. As you can see, both the pre and post CAT H2O sensors reflected high voltage output, indicating a lack of O2 (or otherwise stated, plenty of fuel). The demand for fuel is much greater under heavy load than it is under less of a load. So, the fault doesn’t appear to be related to a lack of fuel delivery. But why the lack of power output?

The next screen capture answers that question (Figure 2). It’s clear to see that as this vehicle begins to ping, the PCM compensates (through the ears of the knock sensor) and generates a command to retard spark. You are all likely aware that a spark occurring too early initiates a combustion process that hinders the piston’s ascent towards top dead center of the compression stroke. This in turn, places a tremendous force on the piston and can damage the internal engine components, due to the violent collision. So, the question then becomes why is the engine pinging?

To battle the production of the harmful gas, NOx, an EGR valve is utilized in this application. NOx is produced in abundance in temperatures exceeding 2500 deg F. The EGR valve is set to reintroduce exhaust gas back to the cylinder. The idea is to fill the cylinder with the inert gas to make the cylinders’ effective area smaller, reducing the intensity of the combustion event. This in turn cools the combustion chamber and reduces the potential for NOx production. My thought process is, if the EGR valve doesn’t open or fails to deliver EGR, it may be the root cause of the “ping” the engine is suffering from. Let’s have a look (Figure 3).

It shows the PCM’s intent to introduce EGR at the expected engine load levels. The EGR position feedback is reporting about 90 percent open, indicating that the PCM is hearing what I’m hearing and commanding a spark-retard of 20 degrees. So now the question is why is the engine still pinging, if the EGR is opening as intended? Is there a restriction of some sort, within the EGR system?

A stroll through the bi-directional control function of my scan tool can answer that question right from the driver’s seat. I simply commanded the EGR valve open at idle, and the engine struggled to maintain idle. It's clear to me the EGR ports were not restricted…Time to roll up my sleeves and dig in deep.

Brought in for questioning

As I mentioned earlier, sometimes we get lucky and we can get the vehicle to tell us everything we want to know with little effort. Other times, we must push to get the answers we need. In situations like this, maintaining a structured game plan is even more crucial to prevent going down a rabbit hole. Rather than trying to find out what is broken, I chase the symptom. I do this because I know what the symptom is. I’ve felt it and I can easily recreate it. I want to see the ping. I want to see inside the combustion chamber while the engine is running to determine the health of that combustion event. I can’t think of an easier way to do this than to view it through the eyes of an ignition scope.

This vehicle utilizes a waste-spark system, using three coils provide the energy to initiate combustion for six cylinders. This system tethers the coils to the spark plugs with ignition cables. The good news is that I can (unobtrusively) acquire the waveforms capacitively, right from under the hood in seconds. With the help of an assistant in the driver’s seat (to place the vehicle under the fault-conditions), the testing was carried out for all cylinders under heavy brake-torque conditions. Displayed is a Bank #1 ignition event in yellow and a Bank #2 ignition event in red (Figure 4)

The waveform displayed (indicated by the red trace) demonstrates an increase in cylinder resistance as the duration of the spark burn-line carries on. We can see this because the waveform slopes upward very sharply. A cylinder that is adequately fueled has less resistance and less energy is expended, trying to maintain the plasma channel (as displayed in yellow). The significance of this event tells me that the bank #2 cylinder is, indeed, under-fueled. [all the cylinders were tested, and each shared this similar characteristic].

So, I now understand why the engine was pinging. But must now ask why is the cylinder under-fueled? A quick test of injector current (using an amp probe and lab scope) ruled out any differences between the injectors ability to flow amperage. An injector balance test was also carried out and showed the ability to deliver fuel was equal among all six injectors. Gaining the answers to these tests justified my need to dig even further. I’m not dealing with a flow issue; I’m dealing with a control issue.

Being in control of fuel delivery also means being in control of the fuel injectors on-time. I will have to monitor the suspect bank’s injectors on-time during the fault and compare it to the known good — dynamically! To gather all that data is as simple as sampling current from a single common point in the under-hood fuse box. I positioned my amp probe to acquire the current flow from fuse #12 of the under-hood fuse block. This fuse’s only purpose is to supply current to all six injectors (this was very convenient as there were no other circuits that could skew my results).

The results of the test exhibit all six injectors yielding the same 1-amp current ramp. Hmmm — the ignition waveform clearly exhibited a lean condition on the rear bank of cylinders, yet the injector flow test and amperage waveforms yielded no difference between Bank #1 and Bank #2. My mistake was the acquisition was not acquired during the fault conditions. I performed the testing at idle. I assumed the fault would be present because I assumed the injectors were restricted or lacked enough current flow. It goes to prove that we learn something new every day and knowledge will continue to beget new knowledge.

Just the facts

After a few moments to gather my thoughts, I had another idea. I would place the vehicle under fault conditions while capturing the injector current. The current is the result. It represents the work performed. I know the bank #2 cylinders are under-fueled so I’m confident I will see the fault reflected in the current waveform.

After recreating the fault conditions, the symptom was exhibited, and the injector current trace revealed the cause (Figure 5). it’s clear to see the that one injector ramp dropped out and another mis-triggered. This created a lack of injector on-time, which explains the lack of fuel delivery exhibited in the ignition trace, as well as the ping. This only proves that the injector failed to open properly. We have yet to decipher the cause.

Remember, current is the output generated by the PCM’s reaction (or processing) of an input. This vehicle uses a sequential fuel injection strategy. It uses a CKP 18x signal and sync signal, referenced from the crankshaft balancer’s reluctor. It also monitors a CMP pulse referenced from the nose of the camshaft. These same inputs effect ignition timing as well.

The Ignition events were also being affected but I’m just chasing the symptom though. Regarding injector control only, these signals are processed by the PCM to determine injector timing and TDC of the number one cylinder, so that it may synchronize the correct injector to #1 cylinder. A PCM just does what it is programmed to do. In this case, drive an injector when it sees the CKP 18x, sync pulse and CMP correlate in a certain manner.

Because the failure was reflecting a fault pertaining to injector control, if the PCM receives a bad input, it’s going to create a bad output (unless the bad input is recognized as such). Shortly, you can see exactly what my next plan of attack is. The strategy was to capture the fault occurring and use that as my point of reference for the other correlating signals. Here, I monitored the fault (injector current) and I correlated that to the responsible inputs that the PCM relies upon for fuel injection calculation (CKP 18x, sync, CMP). It answers a question... When the fault occurred, did the inputs to the PCM show any kind of deficiencies (Figure 6)?

Keeping in mind, each test I perform is justified by the previous test. No time is being wasted. The beauty in approaching drivability faults from this angle is that logic prevents a step from being missed. Every test performed will almost always yield a diagnostic clue. Referring to Figure 6, it displays the fault quite clearly as the blue injector current ramp is once again, deficient. More importantly, there is an anomaly visible within the CMP pattern and the 18x pattern as well (only the green CMP signal visible, for better clarity). So, let’s take a moment to ask why again. Why is the voltage CMP signal dropping low? There could be a few possible causes for a failure of this kind:

  • Faulty CMP sensor
  • Damaged CMP reluctor
  • Poor connection or voltage drop within the signal circuit or the reference voltage circuit
  • shorted/ loaded sensor signal circuit or the reference voltage circuit

This simply requires another test. In this next step, I studied the wiring diagram and saw that the CKP 18x, Sync and CMP signals all shared the same reference voltage source (Figure 7). I will monitor the fault as carried out in the previous step but add some new data to the acquisition. We must now view sensor reference voltage feed, the common feed to all three suspect inputs.

Viewing this piece of data will explain whether the reference voltage has a fault. It allows us the ability to divide the circuit up and determine on which side of the input the fault lays. Now, I want to mention something that I feel is a valuable point to make. I’m asked regularly, if it is necessary to own an 8-trace lab scope like the one I’m using in this case study. It certainly isn’t a necessity, but you will see how having one allows me to save a ton of time. As John Anello (the Auto Tech on Wheels) says, “It’s like fishing with a net instead of a hook.” Having the capturing capability of an 8-trace lab scope allows me to see relationships between multiple inputs, the ECUs response and the actions carried out, all simultaneously.

You will see how this characteristic works to my advantage in this next step. There is one more tool that I will utilize in tandem with the scope, to further nail down the fault to a pinpoint. I will implement the use of a microamp clamp. The microamp clamp is a very sensitive device designed to accurately measure very miniscule amounts of current flow (Figure 8).

The final showdown

The final test will be to monitor the CMP sensor signal under the fault conditions (all inputs reflected the fault so; I just chose to monitor the CMP only). At the same time, I will be monitoring the reference voltage feeding the sensor. The third piece of the puzzle is to monitor current flow through that sensor reference voltage circuit. My thought process is simple.

When the sensor signal is deficient, I will immediately be able to see whether it is due to a deficiency in the reference voltage circuit feeding the sensor. At the same time, the current flow will tell me a story too. If a poor reference voltage feed is due to a voltage drop, the current flow through the sensor will diminish. On the other hand, if the reference voltage circuit is being loaded/partially shorted to ground, the current flow through the reference voltage circuit will INCREASE!

When the vehicle was operated under fault-conditions, the engine began to “ping” hard. This occurred while the injector current ramps showed a deficiency. The inputs responsible for the injector commands were deficient as well. They were fed a reference voltage that was common among all three inputs (CKP 18x, Sync and CMP). With the microamp probe surrounding the reference voltage feed wire, it was quite clear to see the amperage increasing as the fault presented in the CMP, CKP 18x and sync signals (Figures 9 + 10). This tells me that I must pursue a short circuit.

So now, the hunt is on for a rubbed-through harness. After a quick visual inspection, a suspect area was located. Just below the power steering pulley, but above the crankshaft balancer, the CMP sensor harness was unsecured and intermittently touching the crankshaft balancer (Figure 11). This exposed some copper and the wire suffering the damage was the sensor reference voltage circuit, common to all the sensors discussed above.

If you take the time to ask yourself “WHY” at least five times, you typically find yourself face to face with the root-cause of the fault and co-workers looking at you like you are a wizard. So, to sum it all up, lets revisit the chain of events through the questions I asked, to lead me down the path to beat the system:

  1. Why is the engine “pinging”? (Lean condition)
  2. Why is the engine running in a lean state? (Not a fuel delivery issue but a fuel injector control issue)
  3. Why is the PCM failing to drive the injectors correctly? (The PCM operating with skewed inputs)
  4. Why are the CKP 18x, CMP and Sync signals skewed? (A Loaded common reference voltage circuit)
  5. Why is the reference voltage circuit loaded? (Ref. voltage wire feeding CMP is shorting to ground)

As mentioned before, the steps taken were not achieved in record-time but not a step was missed, and this led to an accurate and efficient diagnosis without any parts replaced unnecessarily. Taking the time to interrogate the vehicle will yield you some valuable diagnostic clues that will save you time in the long run. A great side-effect is the developing the understanding of PCM strategy and how different inputs are used in different applications. So, in the end, being a “WHYs GUY” can really make you a Wise Guy.

Author's note: I’d like to say that this was a fast and simple find but that wouldn’t be correct. True, it was not difficult but did require some thought. Logic, studying of circuit topology, system strategy, and lots of practice with the tools I have were a huge part of drawing an accurate diagnosis, but my process would’ve been random without inquiring “why.”

Sponsored Recommendations

Best Body Shop and the 360-Degree-Concept

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

How Fender Bender Operator of the Year, Morrow Collision Center, Achieves Their Spot-On Measurements

Learn how Fender Bender Operator of the Year, Morrison Collision Center, equipped their new collision facility with “sleek and modern” equipment and tools from Spanesi Americas...

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...

Coach Works implements the Spanesi Touch system

Coach Works Uses Spanesi Equipment to Ensure a Safe and Proper Repair for Customers