Solve this riddle

[aoist]

My 2002 pathfinder SE 4x4 has a very strange starting problem. So far I have replaced the fuel pump, fuel pressure regulator, fuel filter and installed a check valve.  Here is what happens, I start the car, and it won't fire up, sometimes if I crank it long enough, 10 seconds for longer several times, it'll fire up, and other times it won't, no matter how much it's cranked. And when this happens, I notice that the car has been parked in an incline, with the front of the car uphill. Every time the problem occurs, car is found this inclined position. 

But parking in this front up incline position doesn't always produce the problem, I have experimented. And the degree of incline doesn't have to be much, just a few degrees. I have so far, maybe 10 to 15 times, been able to start the car with starting fluid. But there's a further mystery which I can't figure out. Every time I start it with the starting fluid, it makes a sustained very loud noise, maybe 3 to 5 seconds, of a very high-pitched metallic rubbing sound. This sound never happens when not started with the starting fluid.

Not starting problem can occur as short as 1.5 hours if parked in an incline. It actually happened at a Walmart.

Any hints or suggestions will be appreciated.

Thanks  

 

 

[Slartibartfast]

And here I thought it was Ford that only ran downhill.

 

That is a weird one. I have no idea what in the fuel system would change with inclination like that. If the fuel level was really low, maybe the angle could make the fuel run away from the pickup--but if that was the case, it wouldn't keep running after it used up the ether.

 

If you slowly rev the engine in park, can you find a speed where it makes the noise? I don't know the R50 computer, but I wouldn't be surprised if the computer was demanding a higher idle for a moment after a difficult start. If you can replicate the noise by revving it to a certain speed, might be an idler pulley on its way out, a loose heat shield, something like that.

 

If you haven't pulled codes yet, do that first, on the off chance the computer knows something. If that doesn't help, check the fuel pressure. Try and get it to act up while you've got the gauge on it. Also see how fast the needle falls when you shut it off. Maybe a leaky injector is flooding it? Though again I wouldn't expect nose up/nose down to make a difference there, and I wouldn't expect throwing more fuel (ether) at it would bring it around.

 

Intermittent issues like this can be a real pain to track down, because they tend to hide until you put the tools away. Good luck!

[aoist]

Good suggestion there, Slartibartfast, about the fuel pressure test, although I'm quite certain that it's not starting due to lack of fuel. What other mechanism is there besides the regulator, which I replaced. I doubt a leaky injector is a problem. Even if, a leak wouldn't be so big that it wouldn't build up pressure, which it doesn't when the problem occurs. That's why I have to use a starting fluid. Because of the incline aspect of the problem, I'm starting to suspect loose/worn quick connector fittings inside the tank. It must be such that, when in an incline, the quick connector shifts enough to let air into the line. Only this scenario can explain why it doesn't build up pressure with a new pump, and why only a huge suction from the intake will draw up the fuel past the air leak. As to the weird sound, it could be the dried out injectors because it quits after a few seconds. Also because the sound never occurs without the fuel problem.

I'm going to replace the fuel sending assembly. Thanks for the feedback.

[Slartibartfast]

I would expect a quick-connect that's loose enough to leak from being slightly tilted would be leaking bad enough to cause all kinds of other issues once the truck was driving around, going over bumps, going up hills, whatever.

 

Manifold vacuum at idle is, what, 10 psi? And it would be pulling through tiny little injector orifices. The rail should be between 34 and 43 PSI. I don't think manifold vacuum is doing anything for the fuel system that the pump wouldn't, unless something is badly wrong with the pump--and again I wouldn't expect that to clear up once it was running. 

 

That said, air in the pump, or air pushing through the regulator, might explain the noise.

 

The next time it acts up, try cycling the key a few times. (On, wait for the pump to stop, off, on again.) The pump should run for a few seconds each time. It's a return fuel system, so if the lines have somehow drained back into the tank while parked, this should prime them again. If it fires right up after cycling, or if it makes the weird noise while trying to prime, or if the pump doesn't run at all, then you know you're on the right track.

[aoist]

Ya, I know about the 6-second fuel signal after each start. I've actually measured the fuel pump voltage at the connector and noticed that it was ~10.7v, so believing that it might be due to lower voltage not turning on the pump hard enough, I connected a 100ah lithium battery directly to the pump to no avail. I believe there is enough suction created though the injectors to pull the fuel up past the leak once the rpm gets up there with the starting fluid. I'll know eventually once I change the hose or just put some Teflon tape on the stem.  Do you think the tape will deteriorate on contact with fuel?  Just bought some yellow teflon tape. Easy way first.

[Slartibartfast]

AFAIK Teflon tape is fuel-resistant, but it's meant for tapered pipe thread. I haven't had a fuel pump out of an R50, but looking at pictures of them, I'm not seeing tapered pipe thread connections. I'm seeing hoses clamped to nipples and quick-connects, neither of which take Teflon tape. Generally quick-connects take O-rings. If the O-rings are bad, you could probably pick them out and install new O-rings.

 

I remain skeptical of your vacuum theory.

[Der_Vier]

This happened to me once. 
I did the fuel pump and was stumped.

Turns out I was low on gas and the incline was just enough where it couldn't feed...