Electronic fuel pressure controller

[2000StreetRod]

Determined correct connections

I wired together a test hookup using my trickle charger to power the fuel pressure controller (FPC) and my battery charger to power the fuel pump via the removed fuel pump relay contacts. I confirmed the following connections for the fuel pressure/temperature sensor:

Pin Signal
1 Pressure output
2 +5 vdc input power
3 Temp output
4 Ground

I powered the pump long enough to fully pressurize the fuel rails and then let it bleed down. I used the Banshee M90 kit supplied fuel pressure gauge and my voltmeter to get data points to compare to the FPC provided table:

0.50 volts – 0 PSI
1.2 volts – 10 PSI
1.65 volts – 20 PSI
2.2 volts – 30 PSI
2.75 volts – 40 PSI
2.89 volts - 43.5 PSI
3.45 volts – 50 PSI
3.9 volts – 60 PSI
4.6 volts – 70 PSI

Identified by black "x" below

My data points identified by red dots above

0.45 volts - 0 psi
1.93 volts - 25 psi
2.30 volts - 40 psi
3.12 volts - 50 psi
4.28 volts - 67 psi

I removed the instrument cluster to install a boost gauge and oil pressure gauge. I now have access to a wiring bundle that probably contains the pink/black wire from the inertia switch shutoff.

I may decide to interrupt it behind the instrument cluster.

 

[2000StreetRod]

G204

I used G204 located behind the driver side kick panel to ground the wire to the fuel pressure controller (FPC) supply side. I cleaned the stock ground terminal and attaching bolt and then added another ground terminal under the head of the bolt.

I snaked it and the other three wires up next to the main bundle coming out of the kick panel behind the central juncton box.

There wasn't room for a loom around the wires but the kick panel is flexible and non-metallic so abrasion should not be a problem. I installed a female 4 pin connector to the FPC wires to the fuel pressure sensor. I installed a male 4 pin connector to the FPC power wires with loom.

I used cable ties to attach the FPC to the underside of the turn/hazard flasher.

In that position I'll be able to access the up and down pressure adjust buttons and observe the LED with my head next to the accelerator. It won't be convenient but I shouldn't need to adjust the pressure since I plan to use the default 43.5 psi setting and the LED just indicates functionality.
I had to loop back the pink/black wire splice from the inertia fuel shutoff (spliced to red wire).

All of the fuel pump power wires (black, green & red) are routed upward next to the main loom.

 

[2000StreetRod]

43.5 psi fuel pressure!

Yesterday I finished wiring all of the connectors and connected everything together. Then I removed the fuel pump relay from the battery junction box and powered up the fuel controller using my battery charger connected to a jumper to the appropriate relay pin socket. I had high expectations of seeing the current meter on the battery charger jump and hearing the fuel pump run. Instead, absolutely nothing happened. No current flow and no fuel pump running. Last night I considered all of the potential wiring errors I might have made in the multiple wire tracing, splicing, and connector wiring. I went back thru all of the wiring diagrams for my vehicle and the fuel pressure controller (FPC). That's when I was reminded of the note that the FPC defaults to a preset 43.5 psi output which can be adjusted by depressing the up or down buttons. Since I purchased an assembled and tested unit instead of wiring it myself I wondered if the testing could have left the default pressure at 0 psi. So today I powered up the FPC and depressed the up button for 3 seconds to reset the default pressure to 43.5 psi. After 3 seconds the LED flashed twice and I released the button as the fuel pump energized for a couple seconds. I climbed out from under the dash and checked the fuel pressure gauge under the hood and it registered 43.5 psi. The pump pulsed for a fraction of a second every 10 seconds to maintain the pressure. The photo below shows the final FPC installation.

The only thing I have left to do is modify the SCT X3 analog cable from one channel to two so I can datalog the air/fuel ratio and fuel pressure and the mating connector for it under the dash. I'm anxious to see if the FPC maintains the default pressure at WOT on the dyno. So now I can perform a datalog for my custom tune progression.

Edit: When James Henson suggested that I purchase the Aeromotive Stealth 340 fuel pump with a capacity of 340 liters per hour I agreed because it's associated literature indicated its design was compatible with pulse width modulation control. With my FPC very little (if any) fuel will pass thru the intank fuel pressure regulator bypass. Fuel vapor emissions and fuel pump energy consumption should significantly decrease and fuel pump longevity should increase.

 

[2000StreetRod]

air/fuel ratio impact

Lowering the fuel pressure at the fuel rail from the standard 60 to 65 psi to 43.5 psi should result in a leaner air/fuel ratio when the PCM is in open loop. I have not determined the impact because (hopefully) it will be compensated for in my M90 custom tune. My system should be comparable to the 1997 and 1998 models with a fuel pressure regulator on the fuel rail and a return fuel line. However, unlike the typical 2nd generation fuel system, my fuel pressure should not need compensation for changes in battery voltage. The power to the fuel pressure sensor is a regulated 5 vdc. The fuel pump controller will adjust the duty cycle of the pump supply circuit to compensate for any output change due to supply voltage variation. I could have set my fuel rail pressure to 62.5 psi but the lower 43.5 psi should improve the accuracy of the high flow injectors at idle. My Siemans Deka 4 injectors can flow 60 lbs/hr @ 43.5 psi (about 3 times my stock injectors).

 

[2000StreetRod]

fuel injector impulses

Replacing the stock diaphragm type fuel pressure damper with a solid state/no diaphragm fuel pressure sensor results in impulses on the fuel rail each time an injector is energized. On my custom fuel rail the fuel pressure rapidly fluctuated between 56 and 70.5 psi when the fuel pressure controller was bypassed and fuel pressure was controlled by the intank 65.5 psi fuel pressure regulator. Since the Explorer utilizes sequential, evenly time spaced injections, short interval pulses on the rail probably have no effect on injector flow rate. However, the constant rapid oscillation of the fuel pressure made it difficult to determine the average fuel pressure. Also, I wanted to maintain fuel pressure stability comparable to that with a damper installed. So I assembled and installed a simple air chamber on the fuel rail to absorb the shock of the injectors opening and closing. The air chamber is constructed of standard, easily obtainable, 1/8" NPT fittings. The photo below shows a street elbow, street T, 45 degree elbow with Schrader valve, and 4" length pipe with cap assembly installed the fuel block port previously occupied by the Schrader valve.

A bonus was easier access to the Schrader valve when attaching an external pressure gauge.

The photo below shows the Banshee kit supplied small fuel pressure gauge that remains permanently installed.

The next photo illustrates that the oil filler tube cap is still accessible.

The air chamber damper consumes less horizontal space than the stock damper.

Unlike the stock damper with a diaphragm that eventually leaks the air chamber should never fail but is very effective. The injector induced impulses on the rail decreased to +/- 0.1 psi.

 

[Runnin'OnEmpty]

Very good writeup Dale, with great pics.
That's an interesting concept of installing an
air chamber dampener instead of the stock
dampener.

One question about that; won't the air chamber
fill up with fuel when a pressure test gauge is
fitted, making it inoperative?

EDIT: Oops, I now see the permantly installed
gauge. Still, would air eventually leak out of it?

 

[2000StreetRod]

common plumbing solution

I got the idea from the stubs that are commonly installed in house plumbing water supply pipes to prevent pipe rattling when turning on and off a faucet. The stubs are vertical and normally near and above a faucet. Since there is no flow of liquid thru the stub the chamber remains full of air to act as a cushion. In my application the Schrader valve located below the stub is analagous to a faucet. If fuel should accumulate in the chamber there will be a noticeable increase in the impulses indicated on the fuel block pressure gauge. If that happens when I bleed off the fuel pressure by opening the Schrader valve I should also drain the stub. In the event vacuum prevents the stub from draining I can replace the stub cap with a second Schrader to allow air flow from top valve to lower valve.

The unusual kluge pressure damper will probably be a conversation topic for anyone who looks under my hood.

 

[2000StreetRod]

Fuel pump response delay

My selected pulse width modulation (PWM) fuel pump controller was designed to control motorcycle and stock automobile modest flow capacity fuel pumps drawing 5 amps or less current. Older pump types were gerotor (gear drive) and rollervane (metal rollers fitted to a rotor). The newer type pump (and my Stealth 340) is the turbine (turbine blade rotates at a high speed). Because of the high turbine speed and much greater flow rate (effective inertia) the pump has a longer response time to the controller increase/decrease speed commands (duty cycle increase/decrease) than the other standard pumps. The worst situation is probably at engine start when the existing fuel pressure is low or zero and the desired pressure is the set value. The engine start plot below is from a datalog I did for the fuel pressure (blue) vs engine rpm (green) for a desired fuel pressure of 43.5 psi.

The vertical red cursor marks the max fuel pressure of 51.3 psi as the engine speed increased from 0 rpm to 129 rpm when the PCM detected the crankshaft rotating and energized the fuel pump relay.

The response delay causes the fuel rail pressure to vary from the desired setting. The pressure will be less than desired while the pump speed is increasing and greater than desired while the pump speed is decreasing. The response delay can be reduced by changing controller variables (more agressive commands), increasing the current capacity of the pump driver module, reducing pump power wiring length (reduce resistance), and increasing pump power wiring gauge (reduce resistance). The stock pump power wiring appeared to be slightly larger than 14 gauge. I used 12 gauge wire for any power wiring that I added. The controller developer is going to experiment with control parameters to attempt to decrease response lag for high current/flow pumps like mine. Even though the fuel pressure varies several psi from the desired after the engine is running I think it is adequately stable for me to proceed with my custom tune development.