Car starts then engine shuts down

[tbm3fan]

The car in question is a 144,000 mile 1990 Mazda 626 automatic. Up until yesterday the car was fine. Then while my wife was driving back to our office, on the freeway, the engine just shut off and she needed to coast off the freeway.

An hour later I was able to get to the car and take a look. Found that the car will start immediately and normally every time. With a normal start the engine will go to the typical idle speed and then slow down to a dead stop in 5 seconds. If started and then given gas up to 2000 rpm the engine runs fine for about 6-12 seconds and then quickly shuts down at 2000 rpm.

So now where to look. After the car was brought home the AAA driver went to steer it out of the middle of the street by starting the car. Had to do it twice, before engine shut down, to get enough inertia to start rolling. He then made a comment about the IAC. My take on the IAC is that it is design to allow small amounts of air to flow into the manifold at start and control idle speed. Consequently, when the engine is running at speed, and the throttle plate is open, then the IAC is not as important. Yet I have total shut down at increased rpms.

Initially and still now, I wonder if the original fuel pump is starting to fail but hasn't completely yet. From there I have fuel pump and fuel relay. Of course, there is ignition in the sense that there is a complete and sudden shut down of power as though you pulled out the coil wire. Yet what could be responsible for shut a shut down electrically? Will be able to look more at the car Saturday and want to pinpoint my efforts at diagnosis.

 

[Ramblejam]

Have you checked for codes?

Run a wire from the green connector to battery negative.

Turn key on, and watch for flashing check engine light. Count the codes.

 

[SwampSurvivor]

Timing belt jumped?

 

[440Magnum]

Don't know much about 1990 Mazdas, but if it were an American EFI car of that era I would look at a weak fuel pump, or fuel pump that runs during the initial igntion-on "prime" cycle but then doesn't run during normal engine operation.

 

[Mr_Joe]

I thought my 1985 626 had a fuel filter? Do you have one, has it been changed ever?

 

[3800Series]

Personally if it was me I would replace the Fuel Filter. They get pretty well clogged up and if that's the case a weak pump isn't doing you any favor. If it's the pump then a new filter will probably help it run and if it stays running you know it's the pump.

Filters are a lot cheaper and easier to replace then a pump. Plug it's always a great idea to replace both at the same time.

 

[bullwinkle]

I had a '90 Civic do the exact same thing-it would run, then die at random times, no OBD I codes. It was a fuel pump. If yours is original @ 25 years old, you're way overdue for one anyway.

 

[Chris Meutsch]

*shouts as loud as possible*

DISTRIBUTOR!

I had a 1989 MX-6 with these exact symptoms. It failed soon after to run at all and the tach stopped bouncing during start-ups so I knew that was the issue immediately.

Same fuel pump and filter, never changed those.

 

[Pontual]

You mean the magnet that reads the revolution on the cap of dist or the gear?

 

[AKhan87]

Originally Posted By: bullwinkle
If yours is original @ 25 years old, you're way overdue for one anyway.
how can something like this be overdue? I don't see how a fuel pump is a maintenance item, just like oxygen sensors, its not like there is a set maintenance schedule for them. It just goes bad, it doesn't need to be replaced at certain intervals.

 

[Chris Meutsch]

Originally Posted By: Pontual
You mean the magnet that reads the revolution on the cap of dist or the gear?


The actual distributor/igniter failed, not just the gear. But yes, when the assembly failed, it wouldn't bounce the tach and that solved the problem for me.

 

[Vikas]

Fuel pump is being told to shutdown. Who knows, may be alternator is not charging anymore? Some manufacturer used spinning alternator to confirm engine running and without that signal, shut down the fuel pump. The symptoms would be exactly what is being seen here. The car will start fine because the pump prime signal is different than pump run signal. If the car does not provide the pump run signal, it will die within few seconds of starting.

 

[tbm3fan]

Originally Posted By: Ramblejam
Have you checked for codes?

Run a wire from the green connector to battery negative.



Not yet since it was 8:00 pm last night and need daylight on Saturday.

Now I don't think the fuel filter has ever been replaced and since I have only had the car 3 weeks it is quite possible. That's why I feel the same about the pump. When I had an 86 626 that car experienced a pump failure and an alternator failure in it's life. Symptoms then much different from these but also 4 years older.

One interesting thing I noted. After every time I started the car and before a restart I could have sworn I detected the faint smell of gas and actually went nosing around the engine compartment and rear seat area.

 

[Mr_Joe]

The fuel filter is a lot easier to change than a fuel pump. If it hasn't been done in a long time, or ever, it needs it anyway, nothing to lose.

 

[TFB1]

Lots of guessing...

I'd be sure it was maintaining spark after key returns to run, if that's OK check fuel pressure... Spark loss will cause INSTANT die, would drop RPMs and stall with loss if fuel pressure)...

As mentioned if the PCM isn't getting a tach or whatever type pulse it's expecting to see, it will shut off pump after initial start...

If fuel filter were plugged so badly it won't idle, engine would not restart without several seconds of cranking...

 

[Pontual]

Fuel pump relsy or Tfi hall effect.

 

[tbm3fan]

First thing today I put a long line on the fuel filter outlet to see if I pumped fuel with the key in start which it did.

So I started the car and it ran 5 seconds before shutdown. I then started the car and moved up to 2000 rpm quickly and the car ran 30 seconds before two brief sputters and then shut down. Finally started the car and moved up to 3000-3500 rpm and the car ran long enough to warm up the coolant and open the thermostat so about 4 minutes. It then had two brief sputters and shut down. After that the car would no longer start under any condition. Only sputter on and then right off.

Running that long at 3000+ rpm seems to tell me the fuel pump is supplying enough fuel at least for the time I ran it at 3000 rpm. The different running times depending on the initial rpm is puzzling so now somewhat stumped. No wonder I like my plain jane Ford 289, 302, 390 and 410 engines with key, coil, distributor and basic fuel pump.

 

[mk378]

The first thing I think of here is the ignition switch. After it stalls out, do the oil and battery lights come on? Those lights should be on any time the key is on but engine not turning.

Have you checked the ECU codes yet?

Now that it generally does not start, check for spark at one of the plug wires.

Fuel pressure should be checked with a gauge. If you don't have a gauge, the second best way is to disconnect the return line (goes from the regulator back to the tank) A good flow of fuel here proves that the pump can make enough pressure and force it through the filter to reach the regulator setpoint.

 

[BreakerBreaker]

It starts without starting fluid?

 

[tbm3fan]

Codes pulled today which didn't seem to follow procedure with lowest numbered service code first.

First flash was long and only one = 10?

Second group was 6 quick flashes = vehicle speed sensor

Third group was 8 quick flashes = air flow sensor

Swapped in the air flow meter box from my 91 626 and got the same problem. However, the code could also be due to a failure of some sort in the connection and not the box.