By Glen Beanard, Technical Contributor
Here we go again. A simple electric pump pushing fuel up to the engine, with its pressure being controlled by a little valve, and returning the unused fuel to the tank was not good enough. Now, we need to make that all fancy by adding a new controller, adding a new sensor and speeding up and slowing down the fuel pump to control pressure.What sense does that make? Are they just trying to sell new parts and tools or what?
Well, actually, the main idea is to reduce fuel vaporization back in the fuel tank. With the older return type fuel system, the fuel would be pumped up into a hot fuel rail overtop of a hot engine, and the fuel would return to the tank full of heat energy. Naturally, the heated fuel would vaporize in the tank at a higher rate than if it were kept cool. That’s the idea of the new system; no return line to send extra heat energy to the tank.
As an additional benefit, a returnless system, having one less fuel line, reduces the chances of a fuel leak. Also, making it electronically pressure controlled gives the PCM more flexibility over fuel control by adding another option if needed. With an electronic returnless system, the PCM can now raise or lower fuel pressure at will.
Theory and Operation
This system consists of a rotary vane fuel pump (same as before), fuel line (pressure only), fuel rail pressure sensor (that also contains a temperature sensor), fuel pump control module and, of course, the PCM. This much should remain the same for any electronic returnless system. A couple of things that become optional across other makes that Ford still includes are the fuel pump relay and the in-line fuel filter. One Ford-specific component that still remains after more than two decades is the inertia fuel cut-off switch. One notable change however, is the PCM no longer controls the fuel pump relay. The fuel pump relay is now toggled by the ignition switch.
It’s a system with a simple goal: maintain pressure in the fuel rail. It just has a high-tech twist on achieving that goal, when compared to its older brother. The PCM uses the fuel pump control module to electronically maintain the desired fuel pressure in the fuel rail. Fuel pump “on” command is still, overall, the same as before: one- to two-second “burst-on” during the initial key on event, and again after a crank position sensor signal is received by the PCM, as during start-up, cranking and engine running. While running, the PCM is attempting to maintain 40 psi of pressure at the injector nozzles. The PCM will raise its target above that pressure if it “sees” the fuel temperature rise high enough to cause vapor pockets in the rail. The higher pressure counteracts vaporization.
To achieve its goal, the PCM communicates a duty cycle to the fuel pump driver module (FPDM). Between 5% and 50% duty cycle is the normal range of demands from the PCM to the FPDM. In this range, the fuel pump output is directly proportional to the fuel pump’s “on” time. Whatever the PCM’s duty cycle is, multiply that by “2” and that is the fuel pump’s on percentage. For example, a duty cycle from the PCM of 20% will equal 40% of fuel pump full-on, and a duty cycle of 50% would equal 100% of full-on at the pump. (See Figure 1).
If the PCM wants the fuel pump turned off, it will send a 75% duty cycle to the FPDM. However, the FPDM will accept 67.5% to 82% as a valid “off” command. As for the duty cycles that fall between the cracks 0% to 4%, 51% to 67%, and from 82.5% to 100% they are invalid. The PCM will not send those duty cycles unless something has gone terribly wrong inside the PCM, or unless a technician commands the PCM to send those duty cycles via a suitable scan tool. Any of those signals from the PCM will result in zero fuel pump operation. (See Figure 2.)
The FPDM controls the fuel pump according to the PCM’s command, and provides diagnostic feedback to the PCM. On the Thunderbird and the Lincoln LS6 and LS8 however, there is no actual FPDM. Instead, the role of the FPDM is performed by the rear electronic module (REM). The REM receives commands from the PCM and controls the fuel pump in the same manner as the FPDM would, except that it receives commands through the data bus. The Ford GT40 uses two FPDMs due to its dual injection fuel delivery system. The dual FPDMs share a single output from the PCM, yet have two separate diagnostic feedback circuits to the PCM so that it can monitor them separately.
In accordance to the command from the PCM, the FPDM modulates the fuel pump ground to rapidly turn “on” and “off” the fuel pump. The rapid “on” and “off” switching, speeds the pump up or slows it down to maintain fuel pressure demands. The FPDM communicates system conditions back to the PCM in the form of a duty cycle signal. The FPDM monitors its commands from the PCM and the fuel pump secondary circuit. It continuously sends a signal back to the PCM that defines the state of health in the circuits between the PCM and FPDM, as well as the circuit for the fuel pump.
The fuel rail pressure (FRP) sensor is a diaphragm-type strain gauge device that measures the difference between the fuel rail pressure and the internal intake manifold pressure. It is referenced to MAP pressures, instead of atmospheric pressures, so that its value reflects the fuel pressure at the injector nozzles. The PCM measures the voltage drop across the terminals of the FRP sensor. The internal resistance of the FRP sensor increases as the pressure increases. As the resistance climbs, so does the voltage drop, and then so does the fuel pressure PID. The FRP sensor also houses an engine fuel temperature sensor (EFT). The PCM also is looking at voltage drop at the EFT as well. Keep this in mind since the majority of circuit wiring problems (such as broken wires, unplugged connectors, loose or corroded pins, etc.) result in high resistance. If your fuel pressure PID and/or fuel temp PID are too high, don’t forget to include sources of high resistance in your diagnostic checks.
However, not all of these systems use an actual EFT sensor yet still have a PID for it. The reason is that the newer versions are dropping the EFT sensor in favor of the PCM software simply inferring the fuel temp value based on ECT, AAT, fuel flow rate and time from start.
On some vehicles, you may notice that there is no Schrader valve pressure test point. That is one feature that is being omitted from these vehicles. Some have them, and some will not. In order to get a mechanical pressure reading on vehicles that don’t have a pressure test point built in, a special fuel pressure adapter tool will have to be added inline between the fuel rail and the fuel line that has its own Schrader point to attach a fuel pressure gauge. However, most of the time, you will discover that the mechanical pressure gauge is not needed. A suitable scan tool provides a fuel pressure PID.
Now you can monitor fuel pressure without the risk of adding a fuel leak over top of a hot engine. However, there may come a time when you feel the need to check the FRP sensor’s accuracy. If you do, there is something you need to know: The mechanical reading and the data PID reading will not match on a running engine, so don’t go replacing the FRP sensor for that.
Since the FRP sensor is referenced to manifold pressure, its readings will be different from that of your fuel pressure gauge, which is referenced to atmospheric pressure. There can be a bit of math involved in determining whether the FRP sensor is accurate. The PCM is trying to get about 40 psi at the injector nozzles, which is what the PID reflects. That is the “Injector Delta Fuel Pressure” (IP). The IP equals Rail Absolute Pressure (RAP) minus Manifold Absolute Pressure (MAP). RAP is the fuel pressure in rail referenced to absolute pressure.
What you are measuring on your fuel pressure gauge is not RAP if it is a standard gauge referenced to atmospheric air. At sea level, your gauge will read about 15 lbs. less than RAP. So assuming your engine is idling with 20 hg of vacuum, convert that to the absolute pressure in pounds per square inch (5 psi).
For practicality reasons, we will need to work the above formula backward, since the fuel pressure PID should stay about 40 psi at idle. Take the 40 psi on the PID, add to that the 5 psi of absolute pressure for a sum of 45 psi. Then subtract from that sum the 15 psi that your gauge is referenced to (at sea level). You should be left with 30 psi. If the FRP sensor and your gauge are both properly calibrated, they will match that 30 psi. Simply put, at sea level, your gauge will read approximately 10 psi lower than the PID.
Now that you have seen the complicated way, put away your calculators and conversion charts. I am about to make your life much easier. The easiest and fastest thing to do is simply compare your mechanical reading and PID readings to each other with the vacuum hose removed from the FRP sensor, or at key on engine off. With the hose removed from the pressure sensor, or at KOEO, the pneumatic pressure referencing side of the sensor is filled with the same barometric pressure as what your fuel pressure gauge is referencing to. The mechanical reading and data PID should match, assuming of course, that your mechanical gauge is properly calibrated, and the FRP sensor is accurate.
The nice thing about diagnostics on this system is that you have two control modules that are both reporting diagnostic information. So you not only get to look at the system from the PCM’s “eyes,” you also get to see it through the FPDMs “eyes” too. There are only three diagnostic monitor signals that the FPDM will return to the PCM. However, as simple as that may sound, reading them requires knowledge of your own tool. See the chart below.
The yellow column represents the duty cycle that you may see if you tapped into the FPDM monitor circuit to the PCM with a duty cycle meter or an oscilloscope. Be mindful of the trigger setting on your tool. Depending on your setting, your readings could get reversed. For example, a 75% duty signal could be misread as 25%. As you can see by the chart, that would send you in completely the wrong direction. The green column reflects the fuel pump monitor PID on a scanner. Beware, some scanners may display the readings in the yellow column, so know your tool.
One thing to keep in mind is that, currently, the same battery positive feed used to supply the fuel pump is the same power feed used to operate the FPDM. So, if you get one of these vehicles with no fuel pressure, and you see the fault codes P1233 and/or P1234 (Fuel Pump Driver Module Disabled or Offline), don’t be so quick to assume the FPDM has failed. If the fuel pump fuse blew, or the inertia switch tripped, the FPDM will not power up and the PCM will store those codes.
An oscilloscope tapped into the FRP signal lead can provide some fast and useful information about the fuel pump’s performance. See Photo 1. This is the voltage signal from the FRP sensor on a 2005 Expedition. The fuel line was completely bled down to zero psi. The key was switched on, but the engine was not started. What useful info can be gathered here? Let’s look at it a minute. At point 1, there is nothing happening, this was key off. At point 2, the key was switched on. Here we can see that there is about a quarter of a second time span from the time the key is switched on, until the fuel pump starts to affect the fuel rail.
Point 3 holds some of the most useful information. That entire upward ramp shows how fast a healthy pump should be able to bring the fuel rail up to peak fuel pressure from zero. It looks like this pump was able to peak out in about 1/8th of a second. That looks like a quick, no-mess test for checking for a slow-to-prime pump. Points between 4 and 5 could be used to inspect the bleed down rate and residual pressure in the rail. This might be helpful in spotting a leaking injector or a leaking check valve in the fuel pump.
A four channel scope can be used to view nearly the entire pump control system at one time. In Screen 1 you can see the diagnostic monitor (feedback) from the FPDM (blue), command from the PCM (red), voltage supply to the fuel pump (green) measured on the ground side of the pump, and the amperage/current ramp of the fuel pump (brown). This is a perfectly normal operating system, except that the time bases have been changed in the capture for display reasons.
Let’s zoom in on the bottom a minute (see Screen 2). This shot shows the pump to driver module’s relationship clearly. The voltage is steadily supplied to the pump and the module is pulsing the ground to turn the pump on. In this capture, the voltage probe for the scope is on the ground lead from the pump. Every time the module completes the ground, the supply voltage (green) is pulled low by the ground path through the FPDM to frame ground. Current begins to flow in the pump as you can see by the current ramp (brown). This duty cycle of “on” and “off” is calculated by the PCM, and executed by the control module.
Also, in Screen 3, I would like to draw your attention to the numbers at the bottom. If you can make them out, you will notice that the PCM’s command (channel B in red) is at 34.84% duty cycle. However, the duty cycle from the FPDM to the pump is at only 30.15%. How can this be? Earlier we discussed that the duty cycle from the FPDM would be double that of the duty cycle from the PCM. What happened in this capture? Was the information earlier incorrect? No, it is still very much in effect. In this case, when the scope sees positive voltage present, then the fuel pump is not energized. The scope is set to display the positive duty cycle. Since it is showing the positive duty cycle, the 30% duty cycle at the bottom of the screen actually means that it is “off” 30% of the time. That of course, will make the pump become “on” 70% of the time, which is double the approximately 35% duty cycle from the PCM. This is a perfect example to show why you should know your tool, as was mentioned earlier.
Screen 3 shows a 100% duty cycle command (red) from the PCM to the FPDM. As you can see, the FPDM didn’t like that very much. It responded by shutting down the fuel pump (evident by the missing current ramping and the constantly high volts on the supply). It then sent a 25% duty cycle back to the PCM (blue) that informs the PCM that the signal it was receiving was not a valid signal.
Screen 4 shows the fuel pump also commanded off. We can see by the 25% duty cycle on the diagnostic monitor circuit, that the FPDM still isn’t happy. As you can see, that is because there is a 0% duty cycle signal from the PCM.
In Screen 5, the pump is yet again turned off. The FPDM has been sent a valid pump off command of about 75% (74.33%). The FPDM module shut the pump down, and continued to return a 50% duty cycle back to the PCM (although we don’t get the greatest of view of that at this time base), since it is receiving a valid command.
Although the outcome means the same for the pump in all of these situations, the driver module’s reasons for each are completely different. In captures 3 and 4, the driver module is acting on its own “decision” to deny the pump. However, in capture 5, the FPDM is acting as a “slave” to the PCM’s “decision.” The FPDM is trying to tell us the reason with its diagnostic monitor signal. Having the proper tools to monitor these signals, and knowing what they are pointing to, can obviously be a huge help in diagnosing these systems. I hope that I have been of help in taking some of the mystery out of this system.
Special thanks to Tom Roberts of AutoNerdz.com for donating the oscilloscope captures and for technical assistance.