Why does my fuel pump shut itself off **8/10 update**

[VegasKyle]

It first happen a couple of weeks ago. I was driving home from work on a hot day after a couple miles my car (91 318is) died. It started back up and went a few more miles then died, I came back later that night to change the fuel filter but it ran fine. I replaced the fuel filter and 1 week later it did the same thing. I left it for a few hours and then it ran fine. Sunday I replaced the fuel pump, it ran fen untill today. Driving home from work its 107* and it dies.

With the back seat out I can start the car, hear the fuel pump run then it shuts off after a few seconds 10-60 and the car dies.

Anyone have a clue why? Is the heat a factor?

[DesktopDave]

I'd guess the DME is losing signal from the crank sensor and the FP relay is unlatching.  Could also be a loose pump connector, that happens sometimes.  How old is the pump & relay?

[VegasKyle]

Dave,

The pump is NEW I put it in on Sunday.  The crank sensor you are talking about is it the one in the front of the timing case just above the water pump?  

I checked and cleaned the connections to the fuel pump when I replaced it. I'll have to check the fuel pump relay, it seems like that is suspect #1.

[monty23psk]

crank is the bottom one at the front of the engine by the crank pulley.

[VegasKyle]

crank is the bottom one at the front of the engine by the crank pulley.

Thanks, I think the heat fried my brain, I was thinking of the cam sensor. How do I test to see if it's good?

[VegasKyle]

Where is the fuel pump relay?  I cant seem to find a straight answer anywhere on the internet.  It seems the 91 318 is different than the rest of the e30s

[VegasKyle]

One last question for the night: Would a bad relay, crank sensor or bad connection explain why it works fine when it's less than 100* out or the car hans't been sitting outside but has the problem only under the exact same conditions (sitting in a parking lot all day then after being driven ~10mi in temps over 105*)?

[bmwman91]

One item to check is the fusible link on the fueling system power wire. You'll notice 2 wires coming from the positive battery terminal...a really big one, and a smaller one. The smaller one is dedicated to the fuel pump/injectors & various injection-related sensors.

When the car cuts out, the first place to check is the positive terminal block on the firewall. There is the one that the huge wire goes to, and there's the one that the smaller wire goes to. They are electrically separate. Test for 12V on the smaller wire's block. If it isn't giving proper voltage, then the fusible link is the likely culprit. Sometimes the connections corrode & lead to all sorts of havoc. I was plagued with weird injection issues for years until I realized that I had made a bad repair on it. It is located about 24" from the battery, under some shrink tubing.

If it isn't that, then it could be a faulty relay. That is fairly uncommon though. The FP relay is under the plastic cover on the firewall, sitting with the main & oxygen sensor relays. I can't remember which one is the FP relay (the blue one, I think). I know the orange one is for the O2 sensor.

[DesktopDave]

The relays are on the top of the firewall toward the fusebox under a plastic cover.  From left to right if you're looking at the car they are the main relay (usually white), fuel pump (usually blue), and then o2 sensor heater (orange).  Here's a great picture, post #5.

If the crank sensor isn't sending a good signal, the car will cut the fuel pump relay.   Test that CPS before you go out again...pins 1 & 2 should have about 640 ohms resistance.  It's not too hard to get at that connector, it's on the front of the wiring plenum by the oil filter housing under the intake manifold.  The cam and crank sensors are both there, crank is the lower plug.

Take a jumper wire with you next time.  If you get stuck again, pull that fuel pump relay and jump sockets 30 & 87 to see if the pump turns back on. If that works, I'll bet your crank sensor is on the way out.  If that doesn't work, test the relay, wire harness and fuel pump power connector.

[kcaldwel]

We just went through this with my brother's E36.

The fuel pump stops when the fuel pressure gets high enough. We got tricked by that too, and messed around with the fuel pump and relay, but that wasn't the problem.

The crank position sensor was intermittent. Replacing that fixed the problem.

Kevin

[VegasKyle]

**7/5 Update**

I drove the car around a little hoping it would die so I could test the fuel pump relay and see if it would run when I jumped it.  The car ran fine and I never got the chance to test it out.  I decided to move on and try some other potential causes.  I un-hooked the crank position sensor and tested the resistance on terminals #1 and #2.  The Bentley calls for 1200 +/- 10% ohms but I was getting a reading of 560-580.  

So I think I found the problem.  I'm going to order up a new CPS and hope that fixes it.  I guess I'll chalk the new fuel pump up to experience and preventive maintenance.

[VegasKyle]

Test that CPS before you go out again...pins 1 & 2 should have about 640 ohms resistance.  

Hmmm.  I missed that part of your post dave.  I have the e36 Bentley and was looking at the pre-OBD II data.  Is the 1991 318is going to be different?  

Damn, I thought I had it figured out too.

Edit: This thread seems to back up the lower ohm number http://www.bimmerfest.com/forums/showthread.php?t=274774

[DesktopDave]

Mr Bentley is wrong.  IIRC the cam sensor is 1200 give-or-take on all M42s; the crank sensor is half that resistance.  BMW spec is 520 ohms +/- 5% (pins 1 & 2) and very high insulation test >100 KOhms.  I'm not sure how to do that one yet, maybe pin 3 is a floating ground?

[VegasKyle]

Mr Bentley is wrong.  The cam sensor is 1200 give-or-take on all M42s; the crank sensor is half that resistance.  BMW spec is 640 ohms +/- 10%

Well shit, I'm back to the drawing board again. Could the CPS still be the problem or is it safe to assume its good?

[DesktopDave]

LOL, caught me editing that last post...

I'm wondering if the jacket of your CPS is broken...if the DME loses that signal it shuts the pump off.  Annoyingly, the stomp test will give a 1444 (all clear).  Maybe the heat is stretching a jacket break?  Or the CPS is really dirty?  Mine stopped once just because it was coated with oil & rust.