Fuel Pump Speed Controls Add to Diagnostic Complexity, John Thornton, Underhood Service, May 2001

Diagnosing fuel supply systems tends to be straightforward. Maybe we check fuel pressure under load, or we test for fuel volume in the bay. Many of us are also using low amp current probes for analyzing fuel pump current patterns. The low amp current probe method works well on most vehicles, but there are some exceptions. Take a look at Figure 1.

This pattern is from a 1996 Olds LSS with a supercharged 3800. The pattern doesn’t look too good, does it? What do you think about Figure 2? That pattern is from a 1999 Ford Mustang with a 3.8L.

Yes, my scope and current probe are working properly. Well, if the test equipment is okay, what does that say about the patterns? Neither one of those two patterns looks typical. Don’t you agree that both of those fuel pump current patterns seem to indicate something unusual?

As you probably suspect, these two vehicles have something in common. Both have fuel pump driver modules which control fuel pump speed. Fuel pump driver modules add a little more complexity to the diagnosis of the fuel pump. If fuel pressure is dropping under load, a fuel pump should be suspected. What if fuel pump feed voltage or ground side voltage were incorrect? Could a feed/ground issue impact fuel pump performance? As you well know, they can! This month we are going to investigate two different types of fuel pump driver modules. The GM design is easy to understand and test. The Ford design, which is used for returnless fuel, is a little bit more involved. Let’s tackle the GM design first.

GM Two-Speed Fuel Pump Driver Module
In 1996, GM introduced a fuel pump driver module for the supercharged 3800 (vin 1) engines. (Note: A resistor is used instead of the module on "W" body-equipped supercharged 3800s starting in 1997.) The driver module is capable of running the fuel pump at only two speeds. The driver module receives its input or command from the PCM. The best place to start our analysis is with a schematic. Please study Figure 3.

The fuel pump driver module has five wires going to it.

• Pin #8 - Yellow wire: This is power in from the fuel pump relay.

• Pin #4 - Black wire: This is ground for the module and the fuel pump.

• Pin #6 - Grey wire: This is the feed voltage to the pump from the module.

• Pin #5 - Pink wire: This is the ground side control of the pump by the module.

• Pin #7 - Brown wire: This is the control line from the PCM to the fuel pump driver module.

The driver module is located in the trunk, behind the left side and rear trim panels. It is pretty easy to access. See Figure 4.

If you want to drop the module for easier testing, you have to pull the rear seat back (on this 1996 Olds LSS) to access the nut that holds the module in place.

Let’s check out how this module is supposed to work using our lab scope and current probe.

Figure 5 shows pump activity during an ignition key-on cycle. Channel 2 is connected to a current probe clamped around the grey feed wire to the pump from the driver module. Channel 1 is on the brown wire from the PCM to the driver module. This is the wire in which the PCM commands the speed of the fuel pump. Again, this pattern was captured during a key-on cycle. Note the pump was turned on for about two seconds. When the PCM driver signal went high (approximately 9 volts), the driver module allowed the pump to run.

So, when the PCM control line is high, the driver module allows the pump to run. Figure 6 shows the cranking-to-running transition. Channel 2 is still pump current, and Channel 1 is on the control line. For the first 2.5 seconds, the control line is high. Then it appears to be modulating very quickly.

Figure 7 shows more detail with a faster timebase. Here we see the PCM’s command to the driver module. The command is a 128 hertz, 32% duty cycle. What is interesting is this 32% duty cycle is not what the pump will run at. The 32% only represents a low speed command to the driver module.

The fuel pump driver module controls fuel pump speed by varying the ground-side duty cycle. Figure 8 compares the PCM command (on the brown wire) to the driver module’s control of the fuel pump (on the pink wire).

Look closely at Figure 8. Channel 1 is the PCM command, and Channel 2 is the driver module control of the fuel pump’s ground side. Channel 2’s signal is squeezed tightly because of the relatively slow timebase. Channel 2 frequency shows 10,000 hertz, which is actually incorrect. At this timebase, my scope can’t make an accurate measurement of that very fast signal.

Figure 9 zooms in on the ground side control of the fuel pump driver module.

Channel 2 is the ground side control of this pump. I am backprobing the pink wire of the fuel pump driver module. Can you believe that frequency? 18,660 hertz is remarkable!

The driver module is controlling pump speed by toggling ground. The pump receives ground about 75% of the time, 18,660 times per second.

That is why the pump current pattern looks the way it does. The pump is being turned on and off over 18,000 times a second when running in low speed. Average current in low speed is about 6 amps.

During cranking or under load, the PCM requests high speed by keeping the command line close to 12 volts. Figure 10 shows this.

Channel 1 is the command line from the PCM, and Channel 2 is fuel pump current. The PCM sensed a load and requested the pump to run at full speed. Fuel pump current increased from about 6 amps to about 8 amps. If we were to jumper 12 volts from the yellow wire (fuel pump relay feed) to the brown wire (PCM command line), the pump would run at full speed. Now we can examine the pump’s current pattern. That was done to acquire the current pattern shown in Figure 11. Compare Figure 11 to Figure 1. There is quite a difference, isn’t there?

At high speed, the pump draws about 8 amps. Pump speed is approximately 6,666 rpm. Determining pump speed is usually a three-step process.

1. Determine the period of the pattern. In other words, measure how long it took for the armature of the pump to make one revolution. Look for a repeating pattern. I estimate the period to be about 9 milliseconds or 0.009 seconds.

2. Determine pump speed or frequency. Frequency is equal to 1 divided by the period. 1/.009 seconds = 111.1 hertz.

3. To find rpm, multiply the frequency by 60 since there are 60 seconds in one minute. 111.1 hertz x 60 = 6,666 rpm.

The only way to use the current probe on this system is to force the fuel pump driver module into high speed. Again, that can be done by jumpering 12 volts to the PCM command line. Or we could bypass the driver module completely by jumpering the yellow wire (pump relay feed) to the grey wire (fuel pump feed). Then jumper the pink wire (fuel pump ground side) to the black wire (ground). The pump will run at high speed, and the module has been temporarily bypassed. This is also a good method that allows us to evaluate the fuel pump independent of the fuel pump driver module.

Ford Fuel Pump Driver Module
Ford uses both a mechanical and electronic returnless fuel system. The one that we are interested in is the electronic returnless fuel system that was introduced in 1998. In this system, there is no fuel pressure regulator. There is a device that looks like a regulator mounted on the fuel rail. This device has a vacuum hose going to it and a three-wire connector on it. See Figure 12. That device is the fuel rail pressure sensor, which informs the PCM of fuel pressure. Finally, a fuel pressure PID can be seen on the scan tool.

Fuel pressure is regulated through the speed/current control of the fuel pump. This is what we are interested in. How does the PCM control pump output? How should we test it? What does good look like? I hope to provide some answers to those questions.

Our investigation into this system starts with understanding the relationship between the components. In my opinion, this is best done using a schematic. Figure 13 shows a simplified block diagram of this system that we can use for our discussion. Since wire colors do differ on different engine families, I will identify the circuits by pin numbers on the fuel pump driver module. Most of my patterns in this article are from a 1999 Mustang with a 3.8L vin 4 engine. I will reference the signals by name and pin number at the driver module.

• Pin #9 - This wire brings power into the driver module from the fuel pump relay and inertia switch. Note: On some of these engines, the fuel pump relay is on all the time KOEO, like the power relay.

• Pin #2 - This is ground for the driver module and the fuel pump.

• Pin #10 - This is feed to the fuel pump from the driver module. This is not modulated. This voltage remains steady with the engine running.

• Pin #3 - This is the ground side control of the fuel pump. The driver module controls pump output by duty cycling this line.

• Pin #1 - This is the command line from the PCM to the driver module. On this line we can see what duty cycle control the PCM is requesting.

• Pin #7 - This is the feedback line from the driver module to the PCM. This signal will tell the PCM if there are any faults with the fuel pump driver module.

I am doing most of my testing of this driver module in the trunk since that is where it is located. Figure 14 shows the driver module on the left side of the trunk with the trim cover removed. Figure 15 shows an up-close look at the driver module removed from its bracket.

We will use the lab scope to examine the signals first, then we will discuss what the scan tool can provide. Figure 16 shows the command from the PCM to the driver module at idle. This was taken from Pin #1 at the driver module. Signal frequency is about 150 hertz, and the all-important duty cycle command is 26% (this command can be seen on the scan tool). Additionally, the 5 volts on this line are supplied by the driver module, and the PCM toggles it low for pressure control. Even though the PCM is commanding 26%, the fuel pump’s actual duty cycle is twice that. I know this seems strange, but it is true. A 75% duty cycle command tells the driver module to keep the pump off. This is seen KOEO. So, the 26% command equates to a 52% pump duty cycle. This duty cycle command will increase if the PCM wants to increase fuel pressure.

For example, 25% duty cycle command equates to 30 psi (measured with a pressure gauge); and 34% duty cycle command equates to 52 psi (measured with a pressure gauge).

Figure 16 shows the PCM command to the driver module. Figure 17 shows the driver module controlling the ground side (Pin #3) of the pump. Check out the frequency of about 9,500 hertz. The ground side duty cycle is almost 50%.

Figure 18 allows us to study this important comparison between the PCM command (Pin #1) and the driver module control of the fuel pump’s ground side (Pin #3).

Channel 1 is the PCM command line with a duty cycle of about 26%. Channel 2 is the driver module ground side control with a duty cycle of about 50%. Obviously, the frequencies are very different. This says we must check the driver module ground side control ourselves. There is no guarantee that the module did what the PCM commanded.

Now let’s compare fuel pump current to the driver module ground side control (Pin #3). We will start with a typical timebase of 2 milliseconds/div. for fuel pump current analysis.

The pattern shown in Figure 19 is from a 2000 Continental with a 4.6L vin V. I know, I know, it doesn’t look like much. Compare that to Figure 20 on page 51 where the timebase has been set to 20 microseconds/div. Now, we can see a relationship. Channel 1 is on the driver module ground side control line (Pin #3), and Channel 2 is fuel pump current. You can see the ringing in the current pattern from the pump’s windings.

So far we have looked at the PCM command line (Pin #1), and the driver module ground side control line (Pin #3). There is still one more line/wire to study. That is the feedback line. Ford calls this line the FPM (fuel pump monitor). This is Pin #7. Figure 21 shows what this line looks like most of the time.

This is a 1 hertz, 50% duty cycle signal that is produced by the driver module and sent to the PCM informing it that there are no problems with the driver module. This is shown on the NGS scan tool as an alternating ON and OFF display. If there is a problem, the FPM signal duty cycle will change. Figure 22 shows this.

I was purposely creating a problem in the PCM command line (Pin #1) on Channel 1. Note the duty cycle change on the FPM line (Pin #7) shown on Channel 2. There are four pulses with a 25% duty cycle. This indicates a fault to the PCM.

Figure 23 shows the PID is available for diagnostics. This recording was taken from KOEO to running. Fuel pump duty cycle is the top PID seen. Initially, we see a 75% command, which tells the driver module to keep the pump off. Once the engine starts, the duty cycle command drops to about 26%, which equals a 52% driver module ground side control of the pump.

The second PID from the top is the fuel pump monitor. This is the feedback from the driver module to the PCM. The third PID is fuel rail pressure. This pressure does not take manifold vacuum into account. Therefore, the 39.9 psi shown is actually close to 30 psi when measured with a gauge. The next PID is the voltage output of the fuel rail pressure transducer.

The first step in diagnosis may be to monitor the fuel rail pressure PID and the fuel pump command duty cycle. Under load, pressure should stay high. If the fuel pump is weak, pressure would drop but the duty cycle should go high (approaching 50% on the scan tool). We may have to check actual pump duty cycle at the driver module ourselves just to make sure. What becomes important is our ability to separate the fuel pump from the PCM and fuel pump driver module when diagnosing.

Thanks for reading and I hope you enjoyed this month’s dilemmas. I know this article was a little on the long side accompanied with many figures, but I hope it will provide you the information you need when one of these fuel pump speed control systems finds its way into your shop.