car in question is '79 BMW 528i with an '86 3.5L bottom end. still has the '79 cylinder head, intake, and L-jet EFI. external fuel pump and external filter. i don't remember if the filter is before or after the pump. rockauto says there 'might be' an in-tank pump...
problem: car starts and runs OK for 5 - 6 seconds, sometimes a little longer (up to 12 - 15 seconds), then begins losing RPM and stalls.
My armchair disgnosis: in-tank fuel pickup is sucking up crap, which limits / blocks fuel flow, which causes engine to die. when engine dies, fuel pump stops running, and crap falls away from in-tank pickup, allowing pump to suck up enough fuel to restart.
what is your armchair disgnosis of this situation?
Yep, that's what I would assume is going on. Sounds like it's getting starved.
thanks ECM. anyone else care to chime in?
Any chance it could be a faulty O2 sensor? Is it throwing any codes?
Another possibility could be the car is only providing fuel in a "start" cycle.
Air, fuel, spark. It is one of these. The odd part is that it starts every time.
What would change after the start?
I question MAP sensor. Once it starts then maybe it reads wrong and cuts fuel or thinks there is no air.
Completely, just a guess but also may provoke some thought.
Will it refire immediately after it stalls, or do you have to wait?
I'm not much of a mechanic, but like John said it's air/fuel/spark for combustion. I wouldn't guess spark as the symptoms don't sound like anything to do with that. So that leaves fuel or air. Guess would be fuel. Getting enough to start it but not to sustain.
I would try to read the fuel pressure before starting and while running. When you turn the key to on, you usually hear the fuel pump run then stop as it primes the lines.
How about a faulty oil pressure switch (or actual low oil pressure) shutting it down after X seconds? Are we talking GM?
In reply to Danny Shields:
Doge......my '92 sho did this when the pump went bad.....it was just strong enough to prime the system for startup then couldn't push enough to keep it running. I crushed the car in reprisal.
I've had this issue on my old RX-7 with a flapper-door MAF (Mass Air Flow Sensor). Like oldeskewltoysaid above, it only provided ~3-5 seconds of fuel pump "on" time, then wouldn't read that it needed to have the fuel pump on still. You might try jumpering the fuel pump permanently on to diagnose that one.
car in question is '79 BMW 528i with an '86 3.5L bottom end. still has the '79 cylinder head, intake, and L-jet EFI.
A '79? On something that vintage, there is at least some possibility that there may be rust in the gas tank unless it has been replaced. I would try to place a large speaker magnet on the tank under the pump inlet and see if that makes any difference. I am imagining rust flakes clogging the pump inlet.
AngryCorvair wrote:
car in question is '79 BMW 528i with an '86 3.5L bottom end. still has the '79 cylinder head, intake, and L-jet EFI.
Does it have a crank position sensor? I'm not sure about the L-jet but the early 5 series big six circa '83 did and that would do exactly what you are seeing.
oldtin
PowerDork
3/28/14 12:46 p.m.
rusty fuel tank, fuel pump turning off after key is released - not getting the run signal from the DME (jumper main and fuel pump relay to figure this one out). Big vacuum leak - will start on the start cycle. Leans out once key is released. Low fuel pressure - pump, filter, fuel pressure regulator.
BigD
HalfDork
3/28/14 1:16 p.m.
Definitely sounds like a fuel flow issue. I like the idea above about jumping the pump and forcing it to run. If that cures it then it's something with the metering/pump control. If it makes no difference, try a new filter, and if that fails get a new pump.
Also worth while checking the bottom of the fuel tank. My E30 would die after driving for a minute or two after swapping the S52. It was enough to run an eta M20 and enough to idle an S52 but once the fuel demands increased, the problem surfaced. The previous owner must have driven it off a curb or something. I put a straight edge on the bottom of the tank and the area around the put where the pump pickup sits was caved in a little bit. Hammered it back out and cured.
BigD
HalfDork
3/28/14 1:19 p.m.
Another thing is that the early cars often had pickup pumps inside the tank and a high pressure pump externally where the filter is. The high pressure pump often gives up the ghost while the pickup pump has a much longer life due to an easier existence. But it's designed for high flow not pressure. So it may build enough pressure to start and idle for a minute but not keep running.
When we fired up our challenge spider the first time, we had a similar problem- which is when I learned that there was a tank pump.
Once i plugged it in, all was well....
Assuming your set up is similar- taking the unit out of the tank was really easy, so you should be able to take it out- chean and make sure it's working ok.
(no need to check o2 sensor- if it had one, it would be a non heated one, and largely ignored for a long time)
EastCoastMojo wrote:
I would try to place a large speaker magnet on the tank under the pump inlet and see if that makes any difference. I am imagining rust flakes clogging the pump inlet.
i heart ECM. that's a friggin' brilliant tech tip right there.
evildky
SuperDork
3/28/14 4:32 p.m.
Is the fuel pump still running when it stalls? The AFM can have a dead spot and when it's dead it shuts off the fuel pump. If the pump is still running check the plugs, the CHTS can cause the engine to run so rich it fuel fouls. Air valve and TB adjustment might be worth looking at as well.
I second the plugged filter, after it stalls some junk falls back down and lets it start again. If you hit them a few times and they run longer you know for sure it is plugged.
If not try hitting the tank with a 2x4 as you start it, if it starts and runs it is a in tank pump gone bad.