Car starts but will not stay running

[Petros]

I recently have been having trouble with our 1990 Corolla All-Trac, which has the transverse 4AFE engine with fuel injection. I posted this problem to another Toyota forum but they were not any help, so I thought I would try it here for those of you with EFI experience.

Here is the background (I do not think it is relevant, but it might be), it was a daily driver and has been reliable, but I accidentally overheated the All-Trac last winter (leaky hose) and than it needed a head gasket, it was still running but I parked it anyway. A few months ago I finally replaced the head gasket, all went routinely, I replaced some hoses as I went. It went back together normally and it started roughly and I found a bad vacuum leak around the number 4 injector gromet. I took it out and wiped all the grommets with RTV, put it all back carefully and it ran fine. I drove for about 40 miles, with multiple starts and stops, and all seemed normal. I parked it for a few weeks and arranged to have the windshield replaced (it is full of cracks), but that day it would start but would run weakly and than die out. I could not keep it running, even with pumping the pedal.

I check for spark and got a strong yellow/white spark, I squirted starting fluid into the intake and it started and ran strong for about 10 seconds. I repeated this several times, it appears only to run when there was starting fluid present in the intake. So it appeared no or weak fuel supply.

I replaced the fuel filter, no change. I replaced the MAP sensor (I had a spare one, and it is easy to change), same results. When I replaced the fuel filter the line had good fuel pressure, and when I turned the key I could hear the tank mounted pump running. And the fuel pressure returned to the injector manifold.

So I figured it could be the signal to the injectors, I checked all the system fuses, connectors,and even swapped out the system relays (I also had some spares).

I got a spare ECU from the local Pull-A-Part (only $10 so I thought it was worth a try), but it behaved the same with the replacement ECU. Since ECU failures are rare I figured that it was unlikely both were bad, even though both were used units.

I know sometimes the fuel pump can work but not develop enough pressure, and I had a spare pump from a Corolla tank I bought some time ago for my Tercel EFI conversion. I was hoping to avoid it because you have to drop the gas tank to replaced the pump. After everything else failed to get it to start I went ahead and replaced the in-tank fuel pump. I bench tested the replacement pump first in a can of gas, and a fuel pressure gauge on the outlet. With 12v it pumped to 70 PSI, so I was good. After I put the new pump in the car started, ran good for a few seconds, went weak and than died out again.

I than replaced the fuel pressure regulartor from one off the 4age engine that will eventually go into my Tercel (they are slightly different design, but it fit and should work the same). And I got the same results.

So at this point I have replaced the fuel filter, the ECU, the system relays, the fuel pump, the pressure regulator, and the MAP sensor. I have doubled checked all the system connections, relays and system fuses, and all the vacuum lines. And it still acts the same, starts weakly and than dies out, runs with starting fluid in the intake.

I can not think of anything left to check or replace, and it is not like I am not familiar with the engine system. Are there any sensors or other components that will not allow it run? AFAIK, even with failed sensors it will still run, just poorly.

Are there any other parts I should test or check or replace? I am stumped.

The service manual is all but useless.

[Mattel]

Hi Petros, here is a forum I found for the all-trac http://www.alltracwagon.net/forum/ unfortunately it doesn't seem to be as well attended as this forum though. It's interesting you had a bad head gasket. My mates all-trac did the same thing. Good luck with that fancy 'fuel injection' technology - I'm not sure if it will ever take on.

[splatterdog]

Since it happened after the head job, I would recheck everything you touched or even got near(if you haven't already). Could be a single ground messin with you or possibly a brittle wire that didn't like the operation.

I would be testing fuel pressure/volume on the car.

Another important sensor is the airflow meter.

Any codes?

Power and injector pulse at the injectors?

Coolant temp sensor in range?

[Petros]

Mattle, thanks for the link. The only reason I damage the head gasket was a little hose leaked and I was not paying attention to the temp, ran it until the needled pegged.

Splatterdog, thanks for the response. it does not use an air flow meter, it uses the Manifold Absolute Pressure sensor, or MAP sensor which I already swapped out.

It does not appear to be related to the head gasket since I left in the intake manifold, and all the wiring, in place when I pulled the head (I unbolted the intake manifold from the head and left in hanging there). I had already checked all the visible wiring, and the ground (I found one loose ground, but it made no difference) though most of injector wiring is in plastic bundle sleeves. It is possible one or more is broken, the wiring is pretty brittle. I could check for signal next, and the temp sensor is easy enough to check. I need someone to help me check the injector signals since it has to be cranking while I test them, which I have not had (wife and daughter is away at summer camp this week).

I would not get a check engine light when it was running, but it would come on when it died, along with the other dash lights. So I do not know if it is storing a code, but it does not appear that it is, though I could check that too.

There is no easy way to test fuel pressure without cutting the hose (it has swagged fittings on both ends). I could grab one from Pull-A-part to use as a test hose. But even so, I replaced the fuel pump, the fuel filter and the pressure regulator, I have lots of pressure in the fuel manifold. So what would a pressure test show that I should replace? How do you do a volume test?

[takza]

I've had my runins with FI. Some simple things you might check...a vac leak will mess them up real well...and appear to be other things. A loose connector to AFM or TPS can be a problem for sure. Those are 2 recent issues I've dealt with...and found after doing exhaustive tests vs the FSM. Nissan though.

[keith]

Have you done a compression test? It take three things for an engine to run; spark, fuel and compression. It appears that you have two of them anyway. Might try retorquing the head bolts.

[Don Jorgensen]

How old is the gasoline or did you fill up during the 40 miles it did run well after the head gasket repair?

[Neu]

How old is the gasoline or did you fill up during the 40 miles it did run well after the head gasket repair?

I'm gonna say clogged injectors.

[danzo]

I think Splatterdog hit on it - the culprit may be a coolant temp sensor. There are prolly a few of them, but find the one that feeds the computer (not the gauge) and replace it. Shouldn't be more than 15 bucks, it's worth a gamble. A bad one won't trigger a check engine code on a car that old.

Years ago I had an '87 Accord that was extremely difficult to start and would really only start with carb cleaner help. For weeks it was in a similar condition as your car - good fuel pressure, no vaccuum leaks (BTW I believe a car needs to be running to check for bad injector seals), no check codes, plugs firing, everything checked out. On a hot tip from a mechanic friend I slapped in a damn 10 dollar temp sensor and it started on command.

Good luck and keep us posted.

[Petros]

Well it is now running again, and I feel like an idiot, idiot, idiot. Though no totally my fault, it was the fuel tank sender unit, it showed a 1/4th tank on the gauge, but it was just out of GAS!!!

it finally occurred to me the last time I drove it the "low fuel" light was on (which was about 3 weeks before), and it just might be out of gas. I was not the last person to drive it, and I thought perhaps someone put a few gallons into it since the gauge was showing slightly less than 1/4 tank, so it appeared to have plenty of fuel. The clue came when I dropped the tank to swap out the fuel pump, when I inspected the inside of the tank there was only about an inch of gasoline in the bottom of the tank, which looked like the fuel pick-up could hardly reach it. That was when I remembered the low fuel light from weeks before.

I put in a gallon of gas from a can for our lawn mowers, the gauge still showed 1/4 tank, but it started and ran fine. I feel like an idiot. Strangely I drove it to my office this morning and it took about 4-5 miles before the fuel gauge dropped down to just above empty and the low fuel light to come back on. I have a spare sender unit, I might just try swapping that out too, it is easy to reach without dropping the tank.

Of all the stupid reasons for it not to run. ARRG.

[ARCHINSTL]

You feel about how I did when I discovered that my recent Nervous-Nellying about differential or wheel bearing or CV axle mega-problems turned out to be something lodged between the rotor and backing plate.
I think we've all been there.
Tom M.