Diagnosing Intermittent Drivability Issues

We as automotive technicians all get the intermittent problems or symptoms from a customer’s car that we know can take a lot of diagnostic time to duplicate the symptom and a lot of diagnostic tests. Even so, there are times in all of our encounters when we are forced to make a judgment call as to what component we may feel is causing the intermittent problem.

A former trainer was quoted as saying that he tests and does not guess. In my opinion, that isn’t a real world strategy in this industry. We all have been forced to make a judgment call simply because we could not narrow down the cause or duplicate the symptom. If a judgment call is in order it’s important to communicate the cost and odds of being successful to the car owner. Some technicians seem to lean on the Silver Bullet Syndrome.

Communication with the car owner to get a good description of the symptom is paramount. Having said that, we know the owner may not offer an accurate description. All too often we have to get the vehicle to act up ourselves.

One important thing I would like to point out is that your diagnostics fees should be explained to the car owner first and foremost. I actually have a sign in my office that establishes our diagnostics. See Fig. 1.

Fig. 1 Bill Fulton has a sign in his shop alerting customers for labor time that may be involved in diagnosing an issue.

Over the years when we test drive a vehicle that has a drivability symptom we cannot say enough of what we feel through the seat of our pants when the symptom occurs. A surge or a hesitation can easily be fuel related while a jerk or buck can easily be caused by an ignition misfire. On our test drives we should always take our scan tool and set it up to the record mode.

Let’s take a look at Fig. 2 (above) from a General Motors Co. (GM) vehicle that exhibited a misfire-type symptom only during a steady cruise condition. A clue here is that the symptom occurred only during a steady cruise with the torque converter engaged. Notice the STFT values in Fig. 2 look normal while the Oxygen sensor data graphed out looks normal.

If a lean density misfire occurred the STFT numbers would have increased as well as the Oxygen signal going low and lean from the unburnt Oxygen molecules. Notice the transmission data indicates we are in fourth gear with the TCC commanded on. Now notice the TCC slippage of 258 RPM. If you could see this data live you would see the live TCC slippage needle fluctuate. The torque converter was the cause.

When it comes to misfire-type symptoms, I like to scope check the secondary or primary ignition circuit (if accessible) to verify good spark duration periods during park idle no load conditions. Car owners typically ignore spark plug change intervals. Spark plug gap erosion and carbon impregnation are still a concern even on the premium spark plugs.

Fig. 3 Here is a Toyota system that exhibited a misfire symptom under light load conditions before the spark plugs were replaced. Notice that the spark duration period is barely 1 millisecond.

Take a look at Fig. 3 from a Toyota system that exhibited a misfire symptom under light load conditions before the spark plugs were replaced. Notice that the spark duration period is barely 1 millisecond. Now note Fig. 4 after new spark plugs were installed indicating a 1.5 millisecond spark duration period.

Once the mechanical integrity of the engine is established as good, chasing a misfire with worn spark plugs can be futile. I am painfully aware that access to the secondary circuit on some modern day engines is difficult if not impossible. Whenever I have a vehicle in my shop where I can easily check the secondary ignition spark duration periods I always do it on misfire-type symptoms.

Fig. 4 After new spark plugs were installed on this Toyota we see a 1.5 millisecond spark duration period. Once the mechanical integrity of the engine is established as good, chasing a misfire with worn spark plugs can be futile.

Fuel trim values during a misfire can help us separate a fuel-related cause or an ignition system failure. Lean density misfires will be indicated by double-digit positive fuel trim values while a rich density misfire will indicate double-digit negative fuel trim values. A loss of spark from a faulty coil or worn spark plugs will have little effect on fuel trim values. When noting the STFT values keep in mind that the engine must be in closed loop. On type A misfires the PCM will force the engine back into open loop and shut off the injector on the misfiring cylinder on modern day systems to protect the converter.

A Chevy Express Van came into the shop with an intermittent jerk buck and stall symptom with a good crank and intermittent no start. The PCM flagged a crank sensor code. The crank sensor had been replaced and a visual at the tip of the crank sensor did not indicate a clearance problem between the tip of the crank sensor and the reluctor. This is a common failure which is caused by worn crankshaft main bearings. GM actually has a fix for this by selling you a shim to re-establish the air gap. On the Express Van this was simply not the case. So much for the Silver Bullet or TSB ideas.

Fig. 5 Take a look at the secondary waveform on a GM engine. Notice that the point of primary turn on is missing.

Take a look at the secondary waveform in Fig. 5. Did you notice that the point of primary turn on is missing? Now let’s look at the CKP signal in Fig. 6.

Fig. 6 Notice the CKP signal. Notice the signal dropouts. The cause was from a faulty wiring harness. We simply abandoned the signal wire from the crank sensor to the PCM and ran a new wire from the crank sensor to the PCM.

Notice the signal dropouts. The cause was from a faulty wiring harness. We simply abandoned the signal wire from the crank sensor to the PCM and ran a new wire from the crank sensor to the PCM.

CAN compliant systems will indicate which cylinder misfired from the last 10 drive cycles as indicated in Fig. 7. Notice the good test results. GM has given us per cylinder history misfire data on the enhanced side of the scan tool for years.

Fig. 7 CAN compliant systems will indicate which cylinder misfired from the last 10 drive cycles as indicated here. Notice the good test results. GM has given us per cylinder history misfire data on the enhanced side of the scan tool for years.

Some technicians replace the spark plugs and the coils from a misfiring cylinder. A good idea would be to stress the coil with a spark tester. Note Fig. 8. The adjustable spark test can be adjusted to a three-quarter-inch air gap which requires a 30KV demand. If a good consistent blue spark is established it is likely that the coil is good. The ST125 spark tester also stresses the coil but in this unit a 25KV demand is established.

Fig. 8 Some technicians replace the spark plugs and the coils from a misfiring cylinder. A good idea would be to stress the coil with a spark tester. The adjustable spark test can be adjusted to a three-quarter-inch air gap which requires a 30KV demand. If a good consistent blue spark is established it is likely that the coil is good. The ST125 spark tester also stresses the coil but in this unit a 25KV demand is established.

Ford Motor Co. has issued a bulletin to note more than 50% of their coils returned under warranty tested OK.

A 2009 VW came in with an intermittent idle stall followed by a crank and intermittent no start. Our info from the customer said that occurred only after an engine warm up. After the engine stall, if the car sat for a while and cooled off it would start again. The previous shop had determined a loss of spark occurred after the stall and thus replaced the DIS and ignition module assembly. I was called to look into the problem.

The Variable Reluctance CKP sensors are known as a pattern failure on these cars. When I arrived at the shop we could not get the symptom to repeat. I removed the CKP sensor and did an ohmmeter check of its resistance which tested within limits.

When heating it up with my heat gun on the bench the ohmmeter went into open. The two clues here came from the car owner who told us that the stall only occurred at an idle condition and with the engine fully warmed up. A variable reluctance sensor’s amplitude increases with RPM and decreases as RPMs decrease.

Fig. 9 Access to the CKP for a circuit check is difficult, at best. But it can be done at the PCM.

Access to the CKP for a circuit check is difficult at best. Access the circuit at the PCM as in Fig. 9. Bench testing the sensor’s resistance change as we heated it up can be seen in Fig. 10.

Fig. 10 Bench testing the sensor's resistance change as we heated it up can be seen in this example.

Now let’s take a look at an intermittent no crank condition from a Cadillac system. Initially we conducted a buss circuit sweep test on this car that utilizes the GM class 2 protocol. A laundry list of U series codes appeared on the scan tool including from the theft deterrent module. This system uses the Pass Key theft deterrent system which will lock out the starter in the event of a fault or tamper. The security light was not on.

Fig. 11 The best way to find a corrupted buss signal is to use a DSO. Note the initial signal in this example indicating a good 7 volt to .2 volt toggle.

Clearing the U series codes and retesting showed a repeat of the U series codes. Thinking that we had a corrupted class 2 network we decided to look at the class 2 signal at pin 2 of the DLC. The best way to find a corrupted buss signal is to use a DSO. Note the initial signal in Fig. 11 indicating a good 7 volt to .2 volt toggle. Now look what happens to this signal during cranking in Fig. 12. Do you see the signal rise well above ground when we went to the cranking mode? This tells us that there must be a ground problem.

Fig. 12 Look what happens to this signal during cranking. Do you see the signal rise well above ground when we went to the cranking mode? This tells us that there must be a ground problem.

Fig. 13 Looking at the starting circuit schematic, the PCM sees the crank request off the ignition switch on the yellow wire. If the PCM gets the enable signal from the theft deterrent module on the class 2 network the PCM will then supply the ground for the pull in windings of the starter relay which then closes the relay contacts and sends B+ to the S terminal of the starter solenoid. All three DVOMs showed proper voltages. The starter still did not engage. We now know that the starter gets its ground through the engine block.

Looking at the starting circuit schematic in Fig. 13 the PCM sees the crank request off the ignition switch on the yellow wire. If the PCM gets the enable signal from the theft deterrent module on the class 2 network, the PCM will then supply the ground for the pull in windings of the starter relay which then closes the relay contacts and sends B+ to the S terminal of the starter solenoid. All three DVOMs showed proper voltages. The starter still did not engage. We now know that the starter gets its ground through the engine block.

Fig. 14 Notice the battery is located under the back seat and that the battery ground is connected to the body. This tells us that there must be a ground cable between the engine block and the body.

Notice in Fig. 14 that the battery is located under the back seat and that the battery ground is connected to the body. This tells us there must be a ground cable between the engine block and the body. Fig. 15 indicates a faulty block to body ground cable. Whenever dealing with an intermittent no-crank complaint on systems that are equipped with a theft deterrent system it is always a good idea to get both keys from the car owner. Keep in mind fob batteries are known to go bad. The extra key is not needed on the GM Pass Lock System since there is no IC chip in those keys.

Fig. 15 This indicates a faulty block to body ground cable. Whenever dealing with an intermittent no crank complaint on systems that are equipped with a theft deterrent system it is always a good idea to get both keys from the car owner. Keep in mind that fob batteries are known to go bad. The extra key is not needed on the GM Pass Lock System since there is no IC chip in those keys.

Many technicians are aware of the diagnostic value of the low inductive current probe. In addition we are aware that the fuel pressure test ports have disappeared. In order to test fuel pressure on Asian European and domestic systems an array of expensive adaptors are needed to connect into the fuel pressure line with a T fitting. A low inductive amp probe would be a better solution coupled to a DSO. See Fig. 16. Based on the schematic we can simply jumper across the power side of the fuel pump relay or jumper across the fuel pump fuse with a jumper wire and clamp our amp probe around the jumper wire.

Fig. 16 In order to test fuel pressure on Asian European and domestic systems, an array of expensive adaptors are needed to connect into the fuel pressure line with a T fitting. A low inductive amp probe would be a better solution coupled to a DSO. Based on the schematic we can simply jumper across the power side of the fuel pump relay or jumper across the fuel pump fuse with a jumper wire and clamp our amp probe around the jumper wire.

On the Ford systems simply locate the inertial switch and clamp the amp probe around one of the two wires. The amperage values displayed on the DSO are directly proportional to fuel pressure values. In addition, electrical connection problems and weak fuel pump grounds can affect the amperage values, so be sure to check for good dynamic voltage to the fuel pump and voltage drop the ground circuit to ensure a good fuel pump ground.

Fig. 17 Electrical connection problems and weak fuel pump grounds can affect the amperage values so be sure to check for good dynamic voltage to the fuel pump and voltage drop the ground circuit to ensure a good fuel pump ground. Here is an example of some good specifications. Notice that the amperage values are directly proportional to fuel pressure.

Let’s look at some good specifications first in Fig. 17. Notice that the amperage values are directly proportional to fuel pressure. Now using the amp probe let’s look at some intermittent problems caused by a faulty electric fuel pump. In Fig. 18 notice the lack of oscillations indicating the pump is mechanically frozen. The symptom was an intermittent no start.

Fig. 18 Notice the lack of oscillations indicating the pump is mechanically frozen. The symptom was an intermittent no start. Banging on the fuel tank caused the pump to start spinning.

Banging on the fuel tank caused the pump to start spinning. In Fig. 19 notice that the signature amperage waveform indicates the amperage drop offs from an open commutator bar. The intermittent symptom was a loss of power and an intermittent no start.

Fig. 19 Notice that the signature amperage waveform indicates the amperage drop offs from an open commutator bar. The intermittent symptom was a loss of power and an intermittent no start.

Fig. 20 waveform was from a Ford Ranger with a good crank and intermittent no start only after a cold soak condition. The waveform was captured during an initial key cycle. We simply clamped around a wire at the inertial switch. The attenuation setting on the amp probe is 100 mV equal to 1 amp. The voltage per division on the DSO is 500 mV per vertical division which means that every vertical division equals 5 amps. Notice the initial current surge of 25 amps followed by a loss of conductivity. The pump is finally on line at about 7 amps.

Fig. 20 This waveform was from a Ford Ranger with a good crank and intermittent no start only after a cold soak condition. The waveform was captured during an initial key cycle. We simply clamped around a wire at the inertial switch. The attenuation setting on the amp probe is 100 mV equal to 1 amp. The voltage per division on the DSO is 500 mV per vertical division which means that every vertical division equals 5 amps, Notice the initial current surge of 25 amps followed by a loss of conductivity. The pump is finally on line at about 7 amps.

Now let’s look at the signature waveform with the pump on line in Fig. 21. Notice the erratic waveform. During a cold start condition the injector on time can exceed 100 milliseconds which can affect the pump’s ability to generate enough fuel pressure. In reference to fuel pump relays, most systems use the standard ISO type relay. Chrysler began using a solid state relay integrated into the TIPM module. This solid state device is known to cause an intermittent loss of voltage to the fuel pump. Chrysler sells a retro kit to replace the solid state relay with the conventional ISO type relay.

Fig. 21 Shown here is the signature waveform with the pump on line. Notice the erratic waveform. During a cold start condition the injector on time can exceed 100 milliseconds which can affect the pump’s ability to generate enough fuel pressure.

Another example is a GM Vortec engine with a loss of power complaint and a MIL along with a P0300 series misfire code. A previous shop had replaced all the spark plugs, coils and secondary leads. The problem was very intermittent. We finally got the symptom to occur on the truck. Scope checking the secondary circuits on cylinders 1, 3, 5, and 7 showed no secondary events on bank 1. All secondary events on bank 2 indicated good, consistent secondary events. So we asked ourselves what could cause the loss of all secondary events on the cylinders on bank 1. Let’s look at the schematic in Fig. 22.

Fig. 22 Using the Min/Max mode of our DVOM we probed the pink wire at circuit point 4. The ignition relay supplies power through the two pink wires to supply B+ voltage to the coils on both banks. This connector is located on top of the left valve cover. Using a Fluke 87 meter in the Min/Max peak detect record mode, we wiggled the connector and the meter beeped alerting us to a connector problem. The left bank cylinders came back on line. Pulling the record values from the DVOM indicated a Max voltage of 13.7 volts. Hitting the Min/Max button again indicated a Min voltage of .4 volts. A new pink wire was spliced in to fix the bad power feed connection to the coils on bank 1.

Using the Min/Max mode of our DVOM we probed the pink wire at circuit point 4. The ignition relay supplies power through the two pink wires to supply B+ voltage to the coils on both banks. This connector is located on top of the left valve cover. Using a Fluke 87 meter in the Min/Max peak detect record mode, we wiggled the connector and the meter beeped alerting us to a connector problem. The left bank cylinders came back on line. Pulling the record values from the DVOM indicated a Max voltage of 13.7 volts. Hitting the Min/Max button again indicated a Min voltage of .4 volts. A new pink wire was spliced in to fix the bad power feed connection to the coils on bank 1.

Glitch detection becomes very easy when using a DVOM such as the Fluke 87 series DVOM as seen in Fig. 23. Pressing the Min/Max button puts the meter in the Peak detection mode and records any voltage changes in the D/C voltage selection. A 100 millisecond peak detect mode has been entered. The meter will emit an audio beep when a 100 millisecond voltage change has been detected. The Max/Min values are recorded in the meters buffer. Pressing the Min/Max button once indicates the maximum voltage values. Pressing the Min/Max button again the meter will display the minimum voltage values. A third press on the Min/Max button will display the average voltage values. Pressing the button just below the Min/Max button puts the meter into a 1 millisecond peak detect record mode. This is a valuable test when conducting a wiggle test across a connector. The Min/Max Peak Detect record function is available on all meter functions.

Fig. 23 Glitch detection becomes easy when using a DVOM such as the Fluke 87 series DVOM as seen here. Pressing the Min/Max button puts the meter in the Peak detection mode and records any voltage changes in the D/C voltage selection.

Freeze Frame data can be very beneficial when a DTC is set. Mode 2 on the global side of the scan tool should always be investigated. Note the Freeze Frame data in Fig. 24 from a P0302 misfire code. Notice that the misfire occurred 40 seconds after startup. The engine RPM was flagged at 1,498 RPM. The engine temperature was at 77 degrees. The MAF reading of 7.64 GPS seems normal for this engine. A shop had previously replaced the spark plugs and the No. 2 coil. We can easily assume the engine was in closed loop since the Toyota Prius system uses heated Oxygen sensors and also stores heated coolant in a thermos device to speed up closed loop operation. The 2.0L engine was a high mileage engine with normal compression values on all four cylinders.

Fig. 24 Note the Freeze Frame data from a P0302 misfire code. Notice that the misfire occurred 40 seconds after startup. The engine RPM was flagged at 1,498 RPM. The engine temperature was at 77 degrees. The MAF reading of 7.64 GPS seems normal for this engine. A shop had previously replaced the spark plugs and the No. 2 coil. We can easily assume that the engine was in closed loop since the Toyota Prius system uses heated Oxygen sensors and also stores heated coolant in a thermos device to speed up closed loop operation.

We were forced to make a judgment call on this vehicle. We communicated to the car owner that we felt that a high flow rate on an injector may have caused a rich condition causing the PCM to lean out all of the injectors causing a lean density misfire. Ideally we would have wanted to set the scan tool up for a record and record data when the misfire occurred.

This problem was so intermittent that we simply could not duplicate the problem. We got the OK from the car owner to flow test the injectors. Three of the four injectors had excessively high flow rates. The one good injector proved to be the No. 2 injector. There are times on intermittent issues that we are forced to make a judgment call and this is perfectly normal once we communicated our assumptions to the car owner.

Electrical connection problems are a common cause of intermittent problems. There are many TSBs addressing this issue. There is a contact enhancer chemical known as Stabilant 22 that greatly enhances conductivity across connections. See Fig. 25.

Fig. 25 A contact enhancer chemical known as Stabilant 22 greatly enhances conductivity across connections.

I try to abide by what I call the 45 minute rule as shown in Fig. 26. Even so, intermittent problems can become a major problem and there is nothing wrong when forced to make a judgment call once you have done some good testing strategies. All too often we spend more diagnostic time and test drives on these intermittent problems than we can honestly charge for. Our best hope is that we retain a customer.

Fig. 26 I try to abide by what I call the 45 minute rule. Even so, intermittent problems can become a major problem and there is nothing wrong when forced to make a judgment call once you have done some good testing strategies.

To try to limit my time on test drives on intermittent problems I communicate to the car owner that I can drive their car home after work to try and experience the symptom. My friendly neighbor once came over to my house and asked if I had a used car lot since he saw a different car in my driveway so often.

I know that my primary focus was on intermittent misfires and intermittent no starts but these two problems are at the top of the list on intermittent symptoms. More often than not we are forced to perform as detectives trying to get as many clues as possible about the intermittent problem. The industry is better because of your commitment.