Diagnostic Strategies on Broken Cars
'It Ain't Easy'
My best friend, the late David Winge, who was a world-class auto service technician, had a favorite quote: “It ain’t no damn easy thing.”
That was in reference to diagnosing engine performance issues.
I have spent many hours with David in the bay dealing with broken cars.
Those of you who are in this industry as an automotive technician, know very well the truth in this quote and the challenges we face.
It was in the early 1990s that the U.S. Department of Labor rated Automotive Technology third in technical demands for skilled trades.
I wonder where we are now.
A CASE STUDY
The first case study that I want to share with you is a 2001 Lincoln Navigator with an intermittent no start and a misfire under road load conditions. My shop was the sixth and final shop challenged with this problem.
The engine has been replaced first with a salvage yard engine with no improvement. Shop No. 5 replaced the engine with a rebuilt 5.4L engine; still with no improvement. Virtually all sensors, including all coils, all injectors, the converters, the PCM and fuel pump had been replaced before I took the vehicle in.
The scan tool is our No. 1 choice of diagnostic weapons in the effort to scan for clues as to a trouble code or deranged data that might yield a clue. In the case of the Navigator, there were no codes and no conclusive data other than double digit negative fuel trim values before the PCM forced the engine back into open loop during the engine misfires under road loaded conditions.
When faced with a diagnostic dilemma, we make a decision on what testing we need to do and what diagnostic tester we are going to use. What you will see here are many tests conducted with a DSO, a current probe and a pressure transducer. You can easily come to a conclusion that I pyramided this vehicle with too many tests. Please keep in mind that five previous shops basically struck out in fixing this vehicle. I did not want to miss anything or leave a rock unturned. Please don’t get the impression that I bat 1.000 or that I walk on water because I don’t. I have had my share of trials and tribulations and have made my share of mistakes that come from working in this industry.
My first encounter with the crank and intermittent no start occurred in the parking lot of my shop on a cold winter day. Using an ST 125 spark tester indicated a good consistent spark. Removing a spark plug indicated wet and fouled plugs. This by itself proved to be very misleading.
As you will see later, if I had weak or inconsistent spark my diagnostics would have been much easier. But keep in mind, the word intermittent plays here.
While many intermittent no starts occurred or the engine missing under loaded conditions resulted in fouled spark plugs, there were a number of fouled plugs that were replaced. My first thought was that the engine might be starving for air. In as much as the MAF sensor was new and the scan tool showed more than 5 grams per second during cranking seemed normal but I felt it alone was not conclusive.
I decided to conduct an exhaust back pressure test. My findings indicated a slight value more than 2 PSI during a power brake condition. Thinking that catalyst substrate may have broken apart from the original converters and ended up in the muffler, I removed the muffler and what I found is indicated in (Figure 1).
Feeling pretty confident at this point, I decided to take a test drive if I were fortunate enough with a good startup. Much to my demise, the engine misfires were still obvious under road loaded conditions resulting in more fouled plugs. The initial assumption that the engine was starving for air was subjective. I simply needed to do more testing. Shop No. 5 replaced the salvage yard engine due to discovering two cylinders low on compression. However the rebuilt replacement engine exhibited the exact same symptoms.
I was still not convinced that the engine was breathing properly, still entertaining the possibility that the fouled plug issue was due to the engine starving for air. The diagnostic weapons I chose at this point were the DSO and a pressure transducer (See Figure 2). The conventional compression gauge was only designed to conduct a WOT cranking compression test. My objective here is to verify good compression values during cranking, idle and WOT snap conditions.
More important, I wanted to verify good cam timing as I was still convinced that the engine was not breathing properly and maybe due to retarded cam timing. Figure 3 indicates a good cranking compression value of 175 PSI. Figure 4 shows a 50 PSI value captured during idle conditions. The idle compression values will only be about 40% to 50% compared to the WOT cranking compression value. This is because the throttle plates are mostly closed during the idle compression test thus impeding the engines ability to breath.
The WOT snap compression values matched the cranking compression values. Since the idle compression values seemed to be on the low side, this created the need to verify proper cam timing. This is another test we can do with the pressure transducer. As in the cranking compression test, we fit the electronic transducer to a conventional compression gauge hose. Make certain that you remove the Schraeder valve located in the conventional compression gauge hose, so that we can view all four strokes occurring and time the exact opening of the exhaust valve and the intake valves.
The first step is to capture TDC events and then using the cursor measure the time between the two TDC events. Note in Figure 5, the value indicates 170 milliseconds. We know that the internal combustion engine has four strokes — intake, compression, power and exhaust. We simply divide 170 by 4 and find that a stroke will occur about every 42 milliseconds. Note Figure 6 where we have placed the 180 degree cursor at the 42 milli-second mark. The point marked EVO is the point of the exhaust valve opening.
If you count the number of minor divisions between the TDC mark and the 180 degree mark, you will come up with 21 minor divisions. 180 divided by 21 will tell us that the crank turned 5 degrees per each minor division. Now count the number of minor divisions between the EVO point and the 180 degree cursor and you will see 8 minor divisions. Five degrees of crank rotation per minor division times the eight minor divisions tell us that the exhaust valve opening occurred 40 degrees before the BDC or 180 degree mark. Normally on most engines with fixed valve timing this point occurs very close to 45 degrees before BDC. The wave form tells us that the cam timing is proper and not the fault.
WHY DOES THIS REMAIN?
We don’t have to monitor the point of intake valve opening since this is a single cam engine. The questions still begs us as to why we have the intermittent no start and the fouled spark plugs. The spark plug fouling problem occurred during the no start and also during test drives.
Is the engine starving for air with a restricted exhaust or retarded valve timing?
Since we are still using the pressure transducer let’s check for proper spark timing. Notice Figure 7. The number of minor divisions between the two TDC marks measure 56. 720 degrees of crank rotation is needed to complete all four strokes. We are using a secondary KV wand on top of the number 1 COP coil hooked to a spark tester indicating the firing event. The firing event occurred three minor divisions before TDC. Each minor division represents 12 degrees of crank rotation meaning the actual firing event occurred 36 degrees before TDC. Spark timing is good. At this point in my diagnostics I have eliminated both possibilities.
When faced with these diagnostic dilemmas, have you ever had to walk away from the vehicle and spend some time planning your next diagnostic plan? I’ve done that many times at home in my easy chair where I can think clearly without interruptions. So where are we now? We know the exhaust is not restricted, the engine’s valve timing is correct and the spark timing is also correct.
We are still stuck on the fact that the engine is either getting too much fuel or not enough air. I failed to mention earlier that the fuel pressure is within specs. In addition on a warm startup the injector on-time measured a normal value of 5 milliseconds. Could we be getting multiple injector pulses causing the extreme rich condition?
We know that the PCM uses the CKP & CMP sensor input to determine the timing and sequence of injector pulses. Let’s take a dual trace DSO with the Pico scope to look at both the CKP and CMP signals captured during idle conditions. Notice the good CKP and CMP signals. The missing tooth on the reluctor wheel appears 45 degrees before TDC which is indicated by the CMP signal. A CKP pulse occurs every 10 degrees of crank rotation. There were no erratic pulses from the CKP and the CMP sensors.
Over the years, I have always been bullish on scoping the secondary firing event with specific focus on the spark line duration and angle. Let’s take a look of a good secondary ignition waveform, not from the Lincoln Navigator in Figure 8, but rather from a secondary lead from a GM DIS system. We are using the COP wand. We simply lay the probe on the secondary lead.
Notice that at 1 millisecond per division the spark line duration measures 1.3 milliseconds. Now using the same COP wand laid on top of the coil, let’s look at a good secondary waveform from the Lincoln Navigator.
The Ford systems will multi-fire the coils when engine RPM is below 1000 RPM. Notice the three good firing events in Figure 9. This waveform was captured from the Lincoln Navigator during idle no load conditions without a misfire occurring. The time base on the scope is 1 millisecond per division. The spark duration on the first firing event measures exactly 1 millisecond. The second and third firing event spark durations measure about .5 milliseconds each. All three spark durations added together become 2 milliseconds. Ford uses this strategy below 1000 RPM to ensure good spark propagation and a consistent flame front. Above 1000 RPM, the PCM shifts to one firing event per cylinder. Now let’s take a look at secondary with the engine missing in Figure 10. Notice the sloping downward sparklines, which is an indication of a rich condition and a fouled spark plug.
Could it be possible that we are intermittently losing ignition and not the injector pulses which could cause the plugs to foul?
Have you ever heard about the cardinal rule of electrical troubleshooting when it comes to intermittent electrical problems? It says,” Initially do not disturb the circuit!” Meaning ... don’t disconnect a connector and reconnect it. You may have eliminated the problem.
The proper procedure when doing dynamic voltage checks is to back probe the voltage supply circuit with the circuit loaded and active in order to detect the loss of proper voltage supply. At this point, I violated that rule. Take a look at the fuse box. There is a 30 amp fuse that supplies voltage to the coils. Using my fused jumper lead I removed the fuse and jumped across the female terminals at the fuse box. My diagnostic weapon of choice at this point was to current ramp the coils with my low inductive current probes clamped around my jumper lead.
At first, I noticed the engine started immediately and ran perfectly with no road loaded misfires. I’m ashamed to tell you that after removing the jumper lead and reinstalling the fuse the no start symptom returned, and if and when it did start, it still missed under road loaded conditions. With the jumper lead reinstalled, the engine started immediately and ran perfectly. The male terminals of the fuse and the female terminals of the fuse box were too loose, causing loss of B+ to the coils. You know they say hindsight is always 20/20. Looking back, I think of what I could have done, might have done or could have done to find the problem faster. Has that not happened to all of us?
I could have taken two different approaches. First and foremost, I could have used a DVOM, and then back probed the plus side of a coil. By using the Min/Max mode of the meter, I would have discovered the loss of voltage supply to the coils when the engine began to miss or no start. In addition, I could have used my current probe and clamped around the red wire with the green tracer conveniently located on the fire wall, which is the feed wire for B+ to the coils.
This would have prevented me from disturbing the circuit noting what happened when I used my jumper lead across the bad connection between male terminals of the fuse and the female terminals of the fuse box. Both choices would have verified the loss of coil saturation or loss of voltage supply to the coils resulting in no spark. However the injector pulses continued resulting in no spark and fuel fouled spark plugs. In closing, thank you for your commitment to this very resilient industry that more and more requires our dedication.