CarTalk.com Blogs Car Info Our Show Deals Mechanics Files Vehicle Donation

1996 subaru outback not starting consistently

The Subaru as 260k miles on it and I live in Georgia.

When I attempt to start the car, it turns over and sounds like the engine catches and then dies out without getting over 200 rpms or so.

Its a manual and I made sure the floor mat isn’t impeding the clutchsensor.

The car has been doing this the past three days and finally today after driving around for a few hours and multiple times shutting the engine off with at least two start attempts before it fires, I get home and shut off the car. I then attempt to restart it about two hours later and after 7 tries, all with the car sounding like the the engine has fired but not yet fully running , then the engine dies.

Any ideas. I would push start it but am parked on an uphill slope.

Thanks
-s

Can you fill us in on the car’s maintenance record–in detail–over the past 3 years?
Is the Check Engine Light lit up?

After repeated and successive tries, can you tell us whether it eventually starts?

I ask because Subarus of that vintage have this weird mode they get into when the thermo sensor (basically the computer’s thermometer) fails. It kinda does what you describe: starting, catching and immediately dying.
Start it repeatedly and see what it does. I seem to remember that, with a bad thermosensor, after 10 times, it is designed to get into ‘limp mode’ where it will start but will run rich. It will not run well but it is designed to at least get you driving.

It is actually a fairly common failure. I’ve had to do it twice on two separate cars and have heard of many other instances as well. The thermosensor is mounted on top of the engine and can be swapped in a couple of minutes.

Edit: now that I’m thinking about it, do the electric fans come on as well? I seem to remember that being another indication of that ailment.

Service in the last three years:
1.) new plugs/wires about 10k ago
2.) Fuel filler pipe (little over three years ago)
3.) Regular oil changes
4.) Vehicle tuneup (new fluids, flush brake fluid, etc.) about 5 months ago

I will head out to the car now and see if i can record an audio sample of what’s going on.

I have not checked to see if the electric fans were spinning, but i will bring a notebook with me and write down any information i can see/hear/smell.

I’ll also see if i smell unburnt fuel from the tailpipe in case something has been flooded.

Thanks so much for the quick response,
-s

Ok you guys must be mechanics…cus this always happens when i bring my car in to check out a noise.

I went outside, got my recorder all set up, hood up, door open…And she fired up on the first shot.

SO first i turned the key to ON, heard the buzz of the fuel pump (might be diagnostically useful)

To give a better timeline…

Time of Day: Noes

1100 - 1400 : Driving around, multiple stops and starts (usually took two times to fully fire the engine)
1600 : Tried starting car 7 times. All times car fired but did not run
1900 : Car fired up on first shot

Don’t know if the duration between engine stop/start cycles matters, but figured i would give all information i can think of.

If you have any other diagnostic tests or voodoo magic that i can try, feel free to suggest.

Thanks again,
-s

So–the check engine light is not illuminated?

Correct. Check engine line is not illuminated.

I had a light one about 3 weeks ago when I had a massive hole in the exhaust. Chain mechanic said I had to replace the cats cus of the bad catalyst error code.

The failure was at the flange rearward of the cats. Cut out pre and post flange piping and had flex pipe welded in its place. No exhaust leak now. I then reset the ecu and completed the driving cycle, all monitors are running after checking scantool last weekend.

Another update…

I went out this morning and the car fired up perfectly first shot…sounded like it took about 7 cranks before the fully fired.

Drove for 7 miles in 90 degree weather, then turned car off at my destination. 30 minutes later, started car (first time again with about 10 cranks before firing).

I have heard suggestions of fuel pressure being a possible issue as well.

Any ideas?

-s