97 LEgacy Outback Fuel Problem

[97LegacyOB]

Ok, so I have a 97 Subaru Legacy Outback with the 2.5. My wife drove it up to the gas station the other day, and when she came out it wouldn't start. I brought it home and started testing. Just replaced the fuel pump relay today with no effect. So here's what I have:

 

- Battery voltage at the red wire on the relay connector at all times

- Battery voltage at the yellow wire with the key on

- Continuity to ground at the black wire below the blue-yellow wire on fuel pump connector

- Continuity between the black-red wire on the relay connector and the blue-yellow wire on the pump connector, however with the relay installed, I can't get a voltage reading from the blue-yellow wire with the key on or while cranking

- Pulled the pump, supplied power and ground and it started instantly

- Pulled plug wire off the coil and cranked engine, has spark

- I did not test anything on the purple wire at the relay connector since it runs directly to the ECU...

 

With everything hooked up I can turn the key on, crank the engine, but the pump never turns on...

 

Hoping someone on here might have some sage advice for me. I really need to fix my car before Tuesday so I don't get dropped from my classes.

[Cougar]

Verify you have voltage on the red relay wire while the pump is supposed to be running. Like when you first turn the key on for the first couple of seconds. If that is ok then check the blk/r wire on the other side of the relay and make sure voltage is getting to that point. If those points are ok then it would seem there is a connection problem between conector B97 and R57 for the blk/r wire.

[97LegacyOB]

Did some more testing. Traced the purple wire from the relay to the computer. Continuity on the purple wire at both relay and ecu plugs. No short to ground. When I check voltage at that wire with everything hooked up, I get very low voltage with the key off and battery voltage with the key on... Seems like this is backwards, doesn't the relay need a ground on the switch side (fuel pump) and one on the control side (through the ecu)?

[hankosolder2]

I suggest jumpering the contacts of the fuel pump relay and seeing what happens. Jumper Red to Black-red (pin 2 to pin 4.) and the fuel pump should run with the key on.

 

If that works, you've got an issue on the ECM side of things. But, you can probably start and drive the car by pulling the jumper when you turn the car off!

 

If it doesn't work, get your voltmeter out. Unloaded voltages and continuity readings taken with an ohmmeter are almost useless. A high resistance due to a corroded connector will not show up unloaded. You want to check voltages under load with a complete circuit. Verify the voltage at the relay jumper wire to ground. Low? Problem is "Upstream" of that. OK? Check voltage at hot wire of fuel pump (Black-red) Low? Issue is wire or connectors between relay and fuel pump. If voltage at hot terminal of fuel pump is OK, measure voltage between fuel pump ground (black) and chassis ground. If you measure any significant voltage, you've got a high resistance connection between the two.

 

Let us know what you find!

Edited July 24, 2012 by hankosolder2
Wrong color wires listed

[Cougar]

My info doesn't show any purple wire in the circuit. Are you sure you are on the correct relay? The info also shows that one side of the pump motor ties directly to ground so your model is controlled differently than some others. From your previous info it looked like there is a bad wire connection on the power side of the wiring to the pump.

[97LegacyOB]

I'm positive I have the right relay. As for voltages, I get very low and unsteady voltage at the blue-yellow wire on the fuel pump (supply). Same goes for the connector under the seat where the black-red wire from the relay becomes the blue-yellow wire going to the pump.

 

As for the pump motor, it has two wires in the connector. One is power supply, the other is a ground. I checked the pump to see if it grounded itself, and it does not. Only ground is through the connector. I though that might be the issue, so I provided an extra ground with a probe and some heavy gauge wire, but that didn't do anything.

 

Whatever is going on, it's causing the pump relay to not close, keeping voltage from getting to the pump.

 

Also, I keep seeing things online about an inertia switch. Do these cars have inertia switches? And if so, where would it be located?

[Cougar]

Ok, my bad. The Cat hasn't been getting enough rest at night and is worn down. The hunting skills aren't what the should be.

 

After looking at the data again I see the violet and yellow wires you are taking about. They tie to the pump relay coil. The yellow wire supplies power from fuse 16 and should be hot when the ignition is turned on. There is a connector in between the fuse and the relay connection so that may be where the trouble is if the fuse isn't causing the trouble. The connector is B62 and F45 according to my data. If the fuse checks ok then you need to check that connector for a problem. The violet wire is the return side of the circuit that makes the ground connection for the relay coil through the ECU.

Edited July 24, 2012 by Cougar

[porcupine73]

There have been a few cases where the transistor in the ECU that controls the fuel pump relay goes bad. Someone had a nice thread with photos on here about how they repaired it some years back. I think others have used a toggle switch to then control the relay instead of having to repair the ECU.