When doing mobile diagnostic work, no-code stalling complaints are a major part of your agenda. In most cases, the client shop is simply too busy to duplicate the failure or, in some cases, a long test drive will yield nothing in the way of useful scan tool data. With many no-code stalls, all you’re going to see on a scan tool snapshot is that the engine’s running perfectly one second and then isn’t the next second.
With those issues in mind, this month’s Diagnostic Dilemma is about solving a no-code engine stalling complaint on a 2003 Chevy Tahoe equipped with the 5.3L V8 engine.
The Tahoe’s owner complained that it would stall at highway speed, but it would always restart and usually proceed without further incident until it reached home. The shop retrieved a P1518 trouble code and, following service data diagnostics, elected to replace the ignition switch to repair an intermittent low-voltage condition. After the ignition switch was replaced, the Tahoe kept stalling but without storing the P1518 or any other code. In addition to the stalling complaint, the starter was engaging and disengaging noisily and erratically. The engine would consistently stall 30 minutes after a cold-soak start but not produce a diagnostic trouble code (DTC).
The P1518 DTC
As a matter of concern, let’s discuss the P1518 DTC. To paraphrase service data, the throttle actuator control (TAC) module and powertrain control module (PCM) communicate through a dedicated serial data circuit. The accurate transmitting and receiving of serial data obviously requires good circuit integrity and adequate system voltage.
My experience is that excessive voltage drops through General Motors ignition switches are fairly common, so the switch replacement was indeed justified — but a TAC failure doesn’t create a stalling condition. In most cases, a TAC failure defaults to a 1,500 rpm fast-idle condition with no throttle control. Nevertheless, I took note of the P1518 and the starter’s erratic behavior.
The Missing Code
Obviously, the P1518 DTC didn’t explain why the engine would suddenly stall then instantly restart. The P1518, the behavior of the starter and the fact that the ignition switch on this Tahoe commands the PCM to engage the starter, made it seem that the stalling and starter control problems might be related. Since there was a theoretical relationship, I kept the PCM option on the table.
The real Diagnostic Dilemma was the stalling, no-code condition. Remember that any electrical/electronic glitch must be of sufficient duration to store a trouble code. In many cases, an electronic failure can be so brief that the diagnostic monitor simply won’t detect it. The trouble code must also meet a very application-specific set of enable criteria — a set of parameters under which the various PCM test monitors actually begin running.
So, is the PCM losing a ground connection? This would simultaneously cause an engine stall and erase any existing DTCs from the diagnostic memory. I’ve also seen a few bad batteries erase DTCs during cranking. Fortunately, the client shop had replaced the Tahoe’s battery and cleaned the terminals in the process. I also instructed the client technician to inspect and clean the ground connections located at the driver’s side oil pan rail and at the rear of the passenger side cylinder head.
A Diagnostic Scenario
Thanks to many years of solving intermittent stalling problems, building a hypothesis for engine stalling first includes factually confirming or denying a crankshaft position (CKP) failure. I focus on the CKP because the ignition and fuel delivery functions depend solely upon the CKP sensor functioning as designed. Some of the time, on some vehicle models, a CKP failure occurs so suddenly that it fails to store a trouble code.
At this point of our dilemma, I tried to understand why the CKP might not be storing a DTC by researching the enable criteria. The results were surprising.
Finding The Missing Code
Looking at the available CKP trouble codes listed in my latest trouble code manual, I found P0335 (CKP circuit malfunction), P0336 (CKP range/performance problem), P0337 (CKP low input), P0338 (CKP high input) and P0339 (CKP circuit intermittent).
Looking at a schematic, the Tahoe’s CKP appears to be a three-wire Hall Effect sensor that’s powered by a 12-volt reference from the PCM. The CKP should therefore produce a square waveform that should pull a 12-volt reference down to zero volts. In some diagnostic cases, a Hall Effect signal can cause a stall by “floating” upward from the zero-volt ground as the sensor warms up. Since the signal is no longer pulling to ground, the PCM can’t read the signal, which causes a no-spark, no-fuel stalling condition. In other cases, the signal can intermittently fail due to a short-to-ground or a fail open circuit due to an internal break inside the wire. Looking at the code list, I decided that P0336, P0337 and P0338 best described the instant stall/restart problem. So, I began with the P0336.
The P0336 DTC
As described in the OE service data, P0336 is a “crankshaft reference sensor performance” trouble code. Enable criteria specify that the engine is cranking or running and that the PCM detects an inconsistent signal for a 3-second duration. P0336 is also a two-trip code that will illuminate the malfunction indicator light (MIL) only during the second trip. So, why isn’t the PCM storing a P0336 or other CKP-related codes? Let’s read further:
“To enable the CKP test monitor, the MIL should be off, no emissions-related DTCs should be present, accessories should be turned off including A/C, fan motor, headlamps and rear window defroster. The battery voltage should range between 10-18 volts, and the engine should be running at idle or cruise speed. The test procedure should use a scan tool to clear any stored DTCs and to reset the diagnostic monitors. The engine should be started and run for five minutes. Accelerate to cruise speed for two to three minutes. Decelerate to zero mph and allow to idle. Do not turn off the key.”
Since P0336 won’t illuminate the MIL until the second drive cycle, the tech will need to find the P0336 in the pending codes instead of the history codes. A pending DTC, in General Motors terminology, would be labeled “failed last ignition” or “failed this ignition.”
According to the enable criteria, all emissions-related DTCS must be erased before the P0336 will be stored. So, if the P1518 DTC was stored, the CKP monitor presumably wouldn’t store the P0336 trouble code, and the P0336 wouldn’t be stored if any or all of the accessories including the HVAC were turned on. Even without spending a lot of time analyzing the enable criteria, there are many reasons for the P0336 to not be stored in history codes.
Testing The CKP Sensor
Keep in mind that I was looking for the P0336 DTC on a strictly hypothetical or speculative basis. Dozens of other DTCs might have appeared as well, but the P0336 would most likely explain the instant stall, instant restart complaint.
As part of my routine, I attached a fuel pressure gauge to monitor fuel pressure. Unless this old Vortec system was converted from poppet valve fuel injectors, it requires a minimum 55 psi to start the engine. Lesser fuel pressures would usually indicate a bad fuel pump. Next, I wanted to monitor PCM powers and grounds. On older GM vehicles, the PCM usually has a B+ and key-on power source and two redundant grounds. But in this application, the Tahoe’s PCM had multiple power and ground connections, which made them difficult to monitor. Just to keep the procedure simple, I monitored the 12-volt power source and ground at the MAF sensor by connecting two DVOMs set to the min/max position. This wasn’t very precise, but better than nothing at this stage of the diagnosis.
Next, I connected my scan tool to monitor engine data, which included a PID indicating the presence of a DTC. After warming the engine for 30 minutes, it stalled as predicted, and, just as predictably, it started and ran for another five minutes. Keeping in mind the Component Monitor Drive Cycle Instructions, I did not turn the key to “off.” Instead, I kept restarting the engine until my scan tool indicated that a DTC was indeed stored in the diagnostic memory. Scrolling back through the diagnostic memory, no history codes were stored, but I found my long-suspected DTC P0336 stored in “failed this ignition.” Two instances were actually recorded: “failed last ignition” and “failed this ignition.”
Due to access issues and the black-and-white nature of the P0336, I recommended simply replacing the CKP sensor, which is located behind the starter assembly. As soon as the technician accessed the starter, he found several wires loose on the starter solenoid, which explained the erratic starter operation. So, in less than 45 minutes, including warm-up time, this month’s Diagnostic Dilemma was solved.
Engine oil temperature was a key part of this diagnosis. During the initial warm-up, I was careful to monitor engine coolant temperature (ECT) to make sure it was reaching the rated temperature of the thermostat, which is 194º F. The engine warmed up to the specified ECT in about 10 minutes, and the engine didn’t stall, but after another 20 minutes of run time, the engine stalled. Since the CKP sensor fits into the engine block behind the starter, the engine oil temperature — not the coolant temperature — was the critical factor in making the CKP sensor fail.
A new crankshaft position sensor solved the Case of the Missing Code.