Fuel Issue - Pump/Pressure/Regulator/Electrical?

[Nickjon]

I have been banging my head on the wall about my fuel issue.   In the last week I have twice replaced the fuel pump with two new ones and both times got the car to fire up and drive perfectly for about a mile. After that it choked and died. Then I went back and put the OEM pump back in and it did the same thing. It seems that it is only after replacing the pump that I get the car to start (and when I spray some starter fluid in the air intake). 

I am only getting about 5-10psi of fuel pressure at the rail which leads me to think either fuel pump or pressure regulator. I am leaning towards the pump/assembly. I just bought the whole pump assembly for about $300 because 1) every time I replace just the pump the car works so it makes me think it is something related to that piece, 2) my fuel gauge does not keep accurate reading & the sending unit was tested bad, 3) my low fuel warning light does not turn on (it is also wired through the sending assembly). I hope this is the right move. I ordered it online but haven't received or installed yet. Maybe you have suggestions? 

My one concerns is my voltage at the pump. I did already replace my relay with new OEM and am getting about 11.57 volts to the pump.   Is that enough to fully power the pump and get full pressure? My pump turns on (audibly) and runs (fuel is moving to the motor). But, I have read that the volts need to be at 12v at the pump. Is 11.57v too low? Or sufficient? Any help would be appreciated! Thanks!

 

Edited May 26, 2015 by Nickjon

[forester2002s]

Fuel-filter clogged (or partially clogged)?

[naru]

Model? Year?

 

Sufficient,as long as you are measuring w/the pump plugged in.

I suspect you are not.

 

It is not the pumps(not all 3)

[Fairtax4me]

Sending units on these are notoriously inaccurate. Even with a brand new unit it probably still won't read correctly. And that low fuel light will only come on when it wants to. It has a mind of its own. Not worth worrying about.

 

Where are you measuring to get 11.57v?

Is that with the pump running?

How does that compare to the battery voltage?

If your battery is only at 11.8, then 11.5 at the pump isn't too terribly far off.

[1 Lucky Texan]

cycle the key to ON/OFF - ON/Off a few times without starting. then pull plugs to look for a wet one. If you find one wet with fuel, maybe that injector is leaking?

[Nickjon]

It is a 1998 Legacy L 2.2

Fuel filter is new, I have cleaned all the lines and I don't have a clog anywhere.

 

My battery is at about 12.6v. I have (in the process of trying the fix) charged the battery to full charge a couple times.

I am reading the 11.57 volts with the key in (engine not running but power on) and the fuel pump disconnected at the terminal. I am reading the terminal coming from the fuel pump relay. 

 

When I read the volts the pump is not plugged in.

[Nickjon]

Lucky Texan, I have checked the injectors and power going to them. I have not checked for leaks. I will look into that. Thanks.

[1 Lucky Texan]

volts with no load won't really tell you if the circuit is free from resistance. maybe look for where the FP is grounded and make sure the ground is good?

 

the ECU has the FP on a timer when you first turn the key ON so, there is pump voltage from the ECU. maybe the other voltage yor'er reading is for the sender unit?

 

I'm kinda out of my knowledge/experience zone here so , I'll let fairtax and others carry-on.

Edited May 26, 2015 by 1 Lucky Texan

[Cougar]

While the voltage level at the pump is a little lower than it should be it isn't anywhere near low enough to cause a problem. That is, as long as it stays at that level. The slightly low voltage may be an indication there is a power connection problem between the fuse panel and the pump motor. If there is a connection problem, it will get hot due to the current draw and the heat is causing the resistance of the bad connection to go higher, which will reduce the voltage getting to the pump even further and finally the motor will not have enough power to work. From what you say about the trouble this very well could be what is causing your headaches and frustration. There is a simple way to prove this if you want to check it out. Simply run a separate power connection directly to the pump. If it was me I would run a switched power connection to the pump in parallel with the original power connection. So as you start out the test drive the switch is OFF. Then when the engine starts to die you turn the switch ON. If the engine then continues to run okay you have proven there is a connection problem. You could also just run some wires from the pump connection and monitor the voltage with a voltmeter to see if the volt drops as you continue on the test run.

 

Check the charging voltage at the battery while the engine is running around 1,500 RPM. You should have at least 13.5 volts getting to the battery. Clean your battery connections and make sure the main alternator output lead is in tight and in good condition. Make sure the fan belt isn't slipping.

 

EDIT:

There has been some further information while I was typing this. Since you say that the pump voltage is lower than the battery voltage with no load this is really pointing to a bad power connection to the pump. 

[Nickjon]

Thanks Cougar.  That is helpful info and a good direction to go. Only problem is I cant get the car started. It has started and run for more than 10 secs ONLY  those couple times right after I replaced the fuel pump. Maybe it is a bad ground or connection that gets moved around while I am replacing the pump? How would I remedy bad power connection to the pump? Run new wires? How do I find out where the connection problem is?

 

Battery connections are all clean.

[Nickjon]

Another thought, is there a specific fuse for the fuel pump? I couldn't find it. Obviously an ignorant newbie question. I looked in main fuse box and the alternate fuse box under the dash. Nothing was labeled for the fuel pump or anything similar as far as I could tell. I did however locate the fuel pump relay and replace it. 

Thanks Again!

[Cougar]

Looking at some service data that I hope matches your model it looks like there is a fusible link that supplies power to the fuel pump. The links should be in a small box close the battery. In the Loyal model the box was mounted on the coolant reservoir, yours may be the same way. When find the links (3) make sure they are making a good snug connection and are clean. They sometimes get loose contact connections. Just the kind of trouble you seem to have and is most likely where the trouble is at. If it isn't we'll will look further down the line.

 

To see if there is a power problem to the pump I suggest you run a direct connection from the battery to the pump motor. Place a in line fuse with the power lead close to the battery, for safety sake.

[Nickjon]

Thanks Cougar. I will take a look at that.

I am an electrical idiot so running a direct connection to the battery is even beyond my ability. But, I hope I can figure it out. If I do it would I leave the pump installed or remove the pump assembly?

[Cougar]

Leave the pump installed. My info shows there is a black/red wire going to the pump for power. The return side of the pump is grounded. Tie a wire from the positive battery post and run it back to the wiring going to the pump. The easiest way to break into the power wire may be to use a pin and stick it into the power wire. Then connect the jumper wire to the pin. Be sure to not short the power to ground or anything else. An even better thing may be to just monitor the voltage on the wire while trying to start the engine. If the voltage is less than 11 volts you need to check the wiring for a problem. Be sure to check the fusible links as that is most likely where the trouble is at. 

[Nickjon]

Cougar,

Props to you for your kindness in assisting community members here. Much appreciated you sharing your knowledge and insight.

 

I just checked the fusible link (found only one in the main fuse box in engine compartment) Looks in tack, strong/sturdy connection on both sides, no corrosion or rust. 

I double checked both fuse boxes for anything related to the fuel pump. Pulled and examined quite a few fuses just to be safe. Saw nothing out of the ordinary. 

 

I will test the voltage while trying to start the car this afternoon. 

 

Thanks again!

[Cougar]

You're welcome for the help.

 

I didn't look at things carefully before. The fuse I was referring to earlier is SBF-2 and is located in the panel under the hood. That fuse ties to a black/white wire out of the fuse panel and goes on to the fuel pump circuit along with some other things. Make sure power is at least good there. If so, then check it at the relay and see if that it is okay there. This where having a meter and knowing how to use it helps pin down problems.

[Nickjon]

Here is the electrical schematic from my Haynes repair manual. Apparently there is a different colored wire that exits they relay from the wire that enters the terminal at the fuel pump. Blk/Red and Blu/Yel respectively. And the fuse is #16 saying it a 15A? 

 

I am going to go check SBF-2 Right now as well as test the volts at the terminal as I try to start it. 

We will see how it goes.

 

I hope I can fix this. I love my little subaru and if I don't get it I am afraid my wife will make me get rid of it. YIKES!

[Nickjon]

Checked SBF-2 looks fine. Strong sturdy connection, no rust or corrosion. How do I test it with a multimeter?

 

I tested the volts at the fuel pump terminal (unplugged) while starting the engine. They dropped drastically. I have two other hot wires at the pump too. Both read about 5v. One drops when the engine starts, the other maintains its volts. 

 

Unfortunately, the colors mentioned in the Haynes manual do not match the colors found on the wiring in my car. But I am following the hot wire (11.57v) backwards toward the relay. Found a connection under the rear seat. It looks a little blackened near the hot wire. Which seems to indicate it is overheating? But the connector piece itself is in good condition. I still notice no loose wires, cuts in the wire, or anything. At this connection my fuel pump wire is still 11.57v. This connection also has a thicker (smaller gauge) wire colored blue and yellow. This wire is reading 12v exactly. . (Both were tested with the connection unplugged) I will upload a pic later.

 

I would like to test at the relay except it is nearly impossible to get to with a multimeter. Far up under the instrument panel. I have to unplug my entire fuse panel and remove it just to get to the relay. So once I am there it is useless to bring a multimeter up there. Especially since the relay is brand new. 

Would a main relay have anything to do with my fuel pump voltage problem?

 

Last thought, My right rear tailight broke in an minor accident and then it rained and got a carwash. Water in the electrical. I think it is unrealated but maybe relevant with some crazy crazy thing. AND, maybe more relevant, occasionally, I mean VERY rarely, my entire instrument panel (including radio which has some lights already burnt out, but it is original) goes black at night with my headlights on. I turn off and on the headlights and the panel lights back up. Maybe another short somewhere but I haven't worried about it. Again maybe connected.

Edited May 27, 2015 by Nickjon

[Cougar]

Fuse SBF-2 does supply power to the main relay also and there is a suspect connector between the fuse panel and those areas. The main relay does (correction, does NOT) provide power to the pump. My info shows a black wire with a red stripe ties to the pump motor. If that isn't what you have then what color do you have for that connection? The return wire should be black. When making voltage tests and especially when you are looking for a bad connection it is best to leave things connected so the voltage will drop across the bad connection due to the current draw through the resistance. Probe the back side of a connector to take a reading or use a pin to pierce the insulation and probe the voltage that way. I would only do that if necessary as it leaves a hole unless you put something like silicone adhesive on it. To check for power place the common meter probe on a known good ground point and measure the voltage using the other probe.

[Nickjon]

Ok, well the car is at the shop. I gave up. But here is what happened. First of all in my ignorance I was checking the wrong wire to the pump which was throwing off my reading of the voltage. I wasn't checking the black red wire that was hot only when ignition was on. After I finally realized that I started checking the black/red wire and was getting about 12v for a 1/4 second. I wasn't sure if this was long enough but at the same time I wasn't getting my pump to turn on at all. So then I checked the circuit through from the connection under the seat all the way through to the pump for resistance. The circuit was good and showed only a minimal resistance. So I installed the pump assembly again and wouldn't you know it, as with the previous two times I installed the pump the car started right up. I drove around the block, parked it, turned it off, and tried again. Started right up. So I decided to try to drive it to the shop. About a 5 mile trip. Drove there like it had never had a problem in its life. Ran great! I gave it to them in defeat. Still not sure what is wrong.

 

Last note, I checked all the wiring I could, all the fuses I could and finally decided to test if I was getting a voltage drop at the pump. I never successfully backprobed the terminal at the pump but at the connection under the rear seat I was getting less than 0.5v of voltage drop on both ground and hot wires while the car was running. And while the car was running I had over 13v of current running through that same connection. 

 

All that to say I assume it is some sort of wiring with the pump assembly or the terminal connection to the assembly.

 

I currently have the OEM pump and wiring in the tank with a new filter bag on the pump and a new fuel pump relay installed. Other than that I haven't done any changes to the vehicle.

 

Another blog I found mentioned this (see link below). I didn't see any O-Rings on the pump assembly in my car so I am assuming this is for a newer model but maybe one of you have any idea. 

http://legacygt.com/forums/showthread.php/fuel-pump-fuel-pressure-regulator-214505.html

 

 

We will see what the shop says.

Thanks again for all your help.

[Cougar]

Well, whatever was causing the issue seems to have gone away for a while at least. Since the trouble would at least go away for a little bit when you changed out pumps it is very logical to think the trouble would be close that area. A connection problem at the connector to the pump itself makes the most sense. The voltages you read on the wiring this last time are real good numbers. I am assuming that the quick drop in voltage you saw on the pump power lead was due to the safety shut down circuit for the pump. This happens when the engine isn't running and the ignition is ON, the ECU will shut down the pump. Now when the engine is running you should have at least 12 volts on that lead to power the pump.  

[Nickjon]

It ended up being the terminal connection to the fuel pump. Unfortunately Subaru does not sell pigtail connections, only the entire wiring harness, so I made a trip to the junk yard. I found a connection and the entire fuel pump assembly in nearly mint condition. The assembly is still shiny, clean, and looks basically brand new. Got them all for only $40. Had the shop wire up the pigtail terminal, plug it into the fuel pump assembly already installed in my car, and now it works great! And I have a basically new fuel pump assembly sitting in my garage just waiting if I ever need it.

 

Thanks again for all your help!