Car starting issue

[Dalamar]

This one isn't a problem as such but a bit of a characteristic of the car. It's a 4L from 1995.

If I go to start the car from cold then it usually takes 2 goes. Turn the key to ignition on (immobiliser off), wait until fuel pump primes and stops, turn over and no start. Turn the key back, repeat and more often than not she will start.

Instead, if I turn the key to ignition on, wait for the fuel pump to prime, stop and then wait 10 secs approx. then turn over the engine and the car will generally start first time. Strange.

Over recent times the car has had new battery, leads, coil, ignition amp, spark plugs distributor cap etc. Also, since I redid the manifold gasket this year, the injectors are clean, don't leak and I know the fuel rail and fuel pressure regulator are OK too. New injector seals were used and the stepper motor works fine.

I haven't changed the fuel pump although I plan to do that when I finish off the last bit of fuel hose to the tank this winter. Electrical connections to the pump are OK as I cleaned these up when I changed the hoses to the engine bay last winter.

I was thinking that maybe the one-way valve in the pump is a bit aged and/or the fuel pump is a bit weak and the double prime is what helps the engine start. Not sure why the extra time after a standard prime also helps the car start.

Any thoughts please? Like I say, not really a fault but I'd like to improve the situation.

[citizen smith]

Have you changed the Fuel filter?

[Belle427]

Maybe try a fuel pressure check first save you chasing your tail, fairly easy to do.

[Steve_D]

Not an answer to the problem but just to say there is no need to wait for the pump prime to finish you should be able to just turn on and start.

Steve

[Dalamar]

citizen smith said:

Have you changed the Fuel filter?

Filter changed around 3 years ago. Fuel flowing through it OK as I flushed the new fuel lines into a jam jar before hooking them up to the cleaned fuel rail.

[Dalamar]

Belle427 said:

Maybe try a fuel pressure check first save you chasing your tail, fairly easy to do.

I did this (of sorts) when I cleaned the injectors and fuel rail. I connected up the fuel rail with fitted injectors and clips off the car to a bike pump. I can't remember the correct pressure (2.5 bar?) but I remember it held OK for > 3mins before every so slightly dropping. I thought that was pretty good for air rather than petrol so went ahead and put everything back into the car.

[Steve_D]

Dalamar said:

Belle427 said:

Maybe try a fuel pressure check first save you chasing your tail, fairly easy to do.

I did this (of sorts) when I cleaned the injectors and fuel rail. I connected up the fuel rail with fitted injectors and clips off the car to a bike pump. I can't remember the correct pressure (2.5 bar?) but I remember it held OK for > 3mins before every so slightly dropping. I thought that was pretty good for air rather than petrol so went ahead and put everything back into the car.

That is telling you it can hold pressure but not telling you if the pump is producing pressure or just noise.

Steve

[Dalamar]

Steve_D said:

That is telling you it can hold pressure but not telling you if the pump is producing pressure or just noise.

Steve

Thanks Steve, I'll pick up a cheap gauge and test the output pressure. I suspect it is just a very old and tired pump. I've tidied up and repaired the electrical connections so I'm happy the power feed is OK.

[QBee]

My Mercedes ML used to do the same as your Chim.
Are you sure you haven't put glow plugs in, instead of spark plugs, and filled it up with diesel?

[blitzracing]

The fuel line should hold pressure after you stop the engine so technically its at fault if the pressure drops to zero. Having said that, mines has always dropped to zero within 30 seconds after running, but starting has not be an issue even with 0 psi to start with.

[Steve_D]

Yes, holding pressure after the pump has stopped is irrelevant. As I said earlier you don't need to wait for prime just start the car. Even after prime the pump will turn on as soon as the ECU detects the first spark of the ignition.

Steve

[Steve_D]

Should also have said the ECU opens all the injectors when it detects the first spark.
Found this out when testing for spark by (gently) shorting the dizzy trigger to the spiked rotor.

Steve

[davep]

Steve_D said:

Should also have said the ECU opens all the injectors when it detects the first spark.
Found this out when testing for spark by (gently) shorting the dizzy trigger to the spiked rotor.

Steve

That's interesting Steve. The way the ECU is coded seems to indicate that it expects the very first spark to issue a non-fuelling interrupt, so a bank of injectors will only open on the second spark. I can't see how the code will allow the ECU to fire all eight injectors at once, but it does employ a double injector firing rate during cold start/cranking for approx 3 seconds.

[Steve_D]

davep said:

Steve_D said:

Should also have said the ECU opens all the injectors when it detects the first spark.
Found this out when testing for spark by (gently) shorting the dizzy trigger to the spiked rotor.

Steve

That's interesting Steve. The way the ECU is coded seems to indicate that it expects the very first spark to issue a non-fuelling interrupt, so a bank of injectors will only open on the second spark. I can't see how the code will allow the ECU to fire all eight injectors at once, but it does employ a double injector firing rate during cold start/cranking for approx 3 seconds.

I may have been mistaken or more likely actually gave a 'double tap' when inducing the spark. It also sounded like both banks fired but again my hearing may have let me down.

Your understanding of the code trumps my casual experimentation.

Steve

[Dalamar]

Just an update on this one. Bought a simple fuel pressure gauge and hooked it in line at the fuel rail. Max pressure was 32psi which I think is too low. Should be ~45psi? Held for a few secs and then started to drop quickly. Recycling the key didn't increase the pressure and if anything pressure dropped quicker with each cycle. I thought there was a non-return valve in these pumps.

I can also hear the rush of fuel in the lines when I first switch the pump on in the mornings plus a kazoo like sound towards the end of the 2.5 sec priming cycle. Electrical connections are still clean and well fitted.

Anyway it was only a quick test as I had to get the car ready to do the school run and go to work the next day. Trying to use the car as much as I can before the end of the month before I SORN.

Sounds like I need a new pump. Not a big deal as I need to drain the tank to change those pieces of 1/2" tubing between the pump and tank this winter.

[Steve_D]

Engine off the pressure should be 36psi (2.5bar).
Can't see from your picture if you have the gauge connected in the supply line or return. (return is wrong).
Some pumps do not include the non return valve but still work fine.

With the gauge in the supply line (or direct on the rail) crimp the return hose and try the pump again. If the pressure is high then the pump is not the issue.

Steve

[Dalamar]

Steve_D said:

Engine off the pressure should be 36psi (2.5bar).
Can't see from your picture if you have the gauge connected in the supply line or return. (return is wrong).
Some pumps do not include the non return valve but still work fine.

With the gauge in the supply line (or direct on the rail) crimp the return hose and try the pump again. If the pressure is high then the pump is not the issue.

Steve

The gauge was definitely in the supply line. Will try again soon.

[blitzracing]

Don't forget the pressure depends on the vacuum in the plenum, so you need to disconnect the thin pipe from the plenum take off to the regulator to make sure the pressure regulator is running at its default setting during cranking. The engine will likely run without the pipe fitted, just not as well as the fuel pressure will be too high and you will have a small air leak if you don't block the pipe off. Best using Rovergauge to turn the pump on if you have it and leave the engine stationary for a stable reading. Its very unlikely to be low at cranking speeds as fuel demand is very low, problems normally occur at WOT.

[Dalamar]

blitzracing said:

Don't forget the pressure depends on the vacuum in the plenum, so you need to disconnect the thin pipe from the plenum take off to the regulator to make sure the pressure regulator is running at its default setting during cranking. The engine will likely run without the pipe fitted, just not as well as the fuel pressure will be too high and you will have a small air leak if you don't block the pipe off. Best using Rovergauge to turn the pump on if you have it and leave the engine stationary for a stable reading. Its very unlikely to be low at cranking speeds as fuel demand is very low, problems normally occur at WOT.

Thanks, I forgot about that Rovergauge functionality. I'll do any testing with the engine off. I'm happy the pressure regulator was working fine earlier in the year and no leaks or leaky injectors on the fuel rail.

[Dalamar]

Just to close this one off.

Fitted the new fuel pump (Bosch 0580464070) and it made a world of difference. When priming it settled around 40psi and when the engine was running around 30 psi if I remember correctly. Sorry didn't take any pics of the gauge in line.

I did check the fuel rail for leaks beforehand using a bike pump and it held 3 bar comfortably as before.

My thinking is that the oomph of the pump had been failing over time and/or any internal one-way valve too.

Glad I changed those bits of 12mm tubing and it was good to drain the fuel tank, give it a good rinse out and remove any water at the bottom.

Changing the filler neck hose and the breather hoses (that go to roll over valve and the charcoal canister) I think helped reduce the overall petrol smell in the boot.

Next job is to finish off the boot carpets. Going to use LED number plate lights as the standard incandescent ones are a bit hot for my liking that close to the carpet in that area.