By Jeff Bach
It's amazing to me how many interesting things you can learn about a vehicle by probing around with a scope and current probe in one small part - in this case, the fuel pump relay.
When checking a “cranks OK but won't start” condition on most General Motors cars, a good, quick test to determine whether you have a spark or fuel problem can be made by checking the fuel pump relay drive signal from the PCM. This is as simple as connecting a current probe and a lab scope to the feed wire at the fuel pump relay. The relay is generally readily accessible and the drive signal wire from the ECM is usually green or green with a white stripe.
Setting the scope time base first to 2 mS per division and the current probe to the low scale (probe must have a good low-resolution scale), you want to be able to display a full screen of about 200 mA. The pattern in Figure 1 was taken at “key on.” It tells me the PCM has power and ground and is sending out the fuel pump's relay drive signal. From the current waveform I can tell that the circuit is complete, the ECM is at least capable of sending the relay drive signal, the coil is not open or shorted, and the relay points have closed. Next, I'll switch the time scale on the scope to 200 mS per division. This allows me to scope 4 seconds of time. The ECM sends the drive voltage signal to the fuel pump relay at “key on” for 2 seconds, waiting to see a reference pulse from the ignition module. Without a reference, the ECM drops the fuel pump current. This is to avoid running the fuel pump after the engine has stopped, in the case of an accident where a fuel line may have ruptured. To my knowledge, all manufacturers use some type of logic to avoid the running fuel pump with the key on and the engine not running.
The fuel pump drive current signal in Figure 2 was taken at the fuel pump relay of a 1990 Cavalier with a 2.0 that was towed in as a “no start.”
I set the scope to “wait for trigger” and cranked the engine for about 5 seconds. From here, I can see which diagnostic direction I'm going to take.
I see from the waveform in Figure 2 that the PCM energizes the relay at “key on” and the current dropped about 30 percent as cranking began. This is normal, as the strain on the battery from cranking drops the system voltage.
The depth of this downward spike varies with battery condition. After two seconds (combination of “key on” and cranking), the fuel pump drive current drops to 0. Keep in mind that I didn't stop cranking the engine until after 5 seconds.
This test has me looking next at the crank sensor signal, which was - as you probably guessed - a flat line.
This next signal came from a 1988 Beretta 2.8 engine. Figure 3 is a 200 mS per division shot of the fuel pump relay drive signal at “key on.” Seeing the trace moving across the screen at 200 mS per division is about like watching a dog trot across a field at 100 yards.
Figure 4 shows the same signal again during cranking.
Notice the humps in the pattern after the starter dip - evidence of a crank sensor signal (unlike the Cavalier in Figure 2) and that the module is firing the coils. The problem with this one was the reference signal from the module to the computer was missing. In this system the crank sensor signal goes right into the module, which then sends a reference to the ECM. Barring no wiring problems, the module is the most likely cause for this condition. Figure 5 shows the reference signal, along with the fuel pump drive current.
To get a “behind-the-curtain” look at the events that take place in this single, magical signal, I took some comparison shots and overlaid them in the appropriate places in the fuel pump relay drive waveform. Shown first is the voltage signal from the ECM with its corresponding current during cranking.
Upon first glance you would think the humps have something to do with the fuel pump current. Not so. The first drop in the drive current comes from the lowered system voltage that occurs as the starter solenoid is energized, as seen in Figure 7.
The next (even deeper) drop occurs when the starter solenoid pull-in current drops to hold-in as the solenoid contacts close and the starter motor begins turning over the stopped engine (see Figure 8). Since the fuel pump relay current closely follows the battery voltage (see Figure 6), this dip can be a good indication of the health of the battery.
The starter motor robs the system of voltage for but a fraction of a second before the first ignition reference pulse appears (Figure 9).
Notice in Figure 9 the relationship between the first reference pulse and the first dip in the relay drive signal after the starter dip. This dip is due to a combination of all six fuel injectors and the ignition coil charge current occurring simultaneously.
Figure 10 shows the relay drive signal and the fuel injectors. Notice the first current pulse for the fuel injectors is nearly 20 mS and it was nearly 90 degrees outside. This is a very informative pulse, especially on a cold cranking engine.
The ignition charge current takes place during the reference low and fires the coil on the second reference high pulse, as seen in Figure 11.
The ignition current during this cranking period covers the full width of the reference pulse while the injector pulses are only in the beginning section. A better view of the relationship between the reference pulse and the ignition current during cranking can be seen in Figure 12.
About 1 second after the first reference pulse appears, the engine is running and the charging system starts putting current back into the battery. And finally, Figure 13 shows the alternator begins outputting current and the fuel pump relay drive current rise above the initial “key on” current at about the same time. Not surprising, but interesting to see just the same.
Figure 14 gives a recap of the events. These events can be viewed due to the power distribution of the circuits. Not all systems show this amount of detail.
There are a lot of discernable events that take place during this 1.5 second period. There are an awful lot of mid-'80s to early '90s GM cars that use this system, making the fuel pump drive current signal one of my favorite places to hook the current probe, especially on a “no start” or “hard to start” complaint. Take a look now at the image in Figure 15 and try to form a hypothesis about this '89 Corsica V6 fuel pump relay drive current pattern.
