Blogs Car Info Our Show Deals Mechanics Files Vehicle Donation

94 Subaru Legacy - dies when in drive

1994 Subaru Legacy, 2.2 L 4-cylinder, front wheel drive, automatic transmission
Trouble started when I let the gas tank get very low, drove the car 3 less than one mile to the gas station from a cold start and filled the tank to full. As I pulled out of the station onto the road the car died as I hit the gas. When I tried to start it again it died right away. Got the car into a safe spot and couldn’t get it to start until I primed the fuel pump about a dozen times. Managed to get the car home. After this, the car drove around town for a few days without any issues. Thinking back it may have been a little hesitant on the gas, but didn’t start stalling again for a few days. This time it would stall while idling as well as when accelerating in drive. The next serious episode occurred on the highway. The engine was choking as I drove over 60mph. I pulled off into a rest area and the car died as I parked. Rather than try to start it again I just called a tow truck and called it a day. Now, even after the replaced parts, it will start just fine and idle ok until I hit the gas pedal then it will idle a much lower rpm and sputter. If I shift into drive and hit the gas it dies immediately. If I let it idle long enough it usually dies eventually. Replaced the fuel filter. No change. Replaced the fuel pump. No change. Replaced the ignition coil. No change. Spark plugs and cables replaced less than a year ago. I have driven this car from Eugene, Oregon to Denver, Colorado and back about three times in the last 6 months, so the long distances could be a factor. Other than that I drive it occasionally (not every day) for short distances in town. Help?

Has anyone pulled any diagnostic trouble codes?

Is your car turbocharged?

The CEL hasn’t come one so I haven’t pulled codes. It is not turbocharged.

Codes can be set without the CEL being illuminated.

A clogged catalytic converter can cause symptoms like this but if that were the case it should run poorly all of the time. This would also be very noticeable on the uphills you travel.

Is there a pickup sock ie filter in the tank? Bad gas? a seafoam treatment cannot hurt.

Without knowing if the engine is losing spark or fuel pressure it can be near impossible to make an accurate guess. Failing ignition switch, EGR stuck open, leaking fuel pressure regulator, you name it.

If the car were mine what I would probably do is hook up a vacuum gauge to get a feel for what’s going on with it. It’s cheap and easy to do that and could provide a lead anyway.

The fuel pressure regulator would be easy enough to check in case the diaphragm is leaking and causing the engine to run to rich. That would cause a loss of power and could also cause an engine to load up and die after idling for a few minutes.
Just pull the vacuum hose loose and check for gasoline in the hose.

A clogged catalytic converter can cause symptoms like this but if that were the case it should run poorly all of the time. This would also be very noticeable on the uphills you travel.
-I recently had another car do this to me and the symptoms were a bit different, like you said, so I am not sure that’s the answer. I will try to find a code reader to use and see if any codes come up.

If the car were mine what I would probably do is hook up a vacuum gauge to get a feel for what’s going on with it. It’s cheap and easy to do that and could provide a lead anyway.
-I’ll need to research how I would do this, haven’t attempted it yet. Any suggestions?

Is there a pickup sock ie filter in the tank? Bad gas? a seafoam treatment cannot hurt.
-Yes, there is. Brand new one went in with the new fuel pump about a week ago. Someone in town mentioned that it could be bad gas. I will look into this.

A vacuum gauge is a cheap tool. It is simply connected by hose to any vacuum source from the intake manifold. Those gauges come with an instruction pamphlet.
It takes a little putzing around to learn what it all means but it’s not difficult.

If you decide to go this route and need any advice on what it all means post back and I’ll at least make a stab at trying to help.
I’m a huge fan of vacuum gauges simply because it takes seconds to connect one and it will let you know in seconds whether there’s an issue or not. It’s a great preliminary check tool.

Check the throttle position sensor voltage. You should see voltage changes in the 5 volt sense lead as you press on the gas pedal. If that is good then check the MAF sensor for a problem.

@ok4450 Lucky for me there is a Harbor Freight in town. I will post back with what I find.

@Cougar This would require an ohm meter, correct?

Mander, no a voltmeter, or a multimeter in the volts DC mode.

You could use an ohmmeter with no power applied to the circuit but it is easier to have power applied and monitor the voltage changes on the sensor output lead as you move the throttle. Depending on how your system is set up idle voltage may be near 5 volts and go less with more throttle applied or it may start out below 1 volt at idle and go higher.

Ok here’s a quick update: I enlisted the help of a friend of a friend and he helped me do some diagnostics.
When checking the fuel system it appeared to be reverse pressurized, but both the filter and pump are installed correctly. Not too sure what this means?
He cleaned the throttle body and MAF. Also did a seafoam treatment. This appeared to have worked at first. He drove the car to work and back, took it out for 20+ miles on the highway and the car ran without any hiccups or stalling.
I went to pick the car up today and it died less than a block away with the exact same symptoms. (Died when I hit the gas) I fiddled with the for a while trying to get it to restart and here are the behaviors it showed:
First time I restarted the engine sputtered for a few seconds, then jumped up to 3000 rpm for a few seconds, and then dropped below 1000 rpm and then died after a few second. All this happened without me doing anything.
Second time I restarted I primed the pump first and it started with no sputtering. I hit the gas and held it at 3000 rpm for a few seconds and the engine died as soon as I released the gas.
Third time the engine died as soon as I hit the gas. I left it at that and called it a day.

My friend did pull some codes for me: 21, 24, 31, 41 - I will be researching these later today

Here are the code definitions:

21 Coolant temperature sensor or circuit
24 Air control valve or circuit
31 Throttle position sensor or circuit
41 Air/fuel adaptive control

There may be a intermittent power problem to the sensors. I assume 5 volts is used for all of them. I would start with the throttle sensor.

The air control valve is probably what is commonly called the IAC or the idle air control valve, and if that failed, you’d get irregular idle, with the idle rpm going up and down without you stepping on the gas pedal. I’d start with testing that first. Sometimes it will get gummed up with gunk, and good cleaning will fix it. Often the throttle body insides need to be cleaned out too. that’s probably where I’d start if I had this problem on my early 90’s Corolla, clean and test the IAC, and clean the throttle body internals.

If that doesn’t fix the problem, next, test the throttle position sensor. Still doing it? Finally, test the coolant temp sensor. If that failed it could cause weird drivability problems. That’s usually just a thermistor and is easy for a shop to test. Don’t worry about the air/fuel control code until the first three have been tested and any corrections necessary are complete, as it is probably a result of some problem or another w/ the first three.

If nothing above seems to help, when you ran out of gas, I know it doesn’t seem possible, but that could have damaged the cat. There’s a test called “cat backpressure”, and it probably worthwhile in your case. If the cat remains in doubt, it can always be temporarily disconnected to see if it fixes the problem. That’s a pretty definite test for a bad cat.

The other thing worth checking is the EGR valve. It might be sticking open or opening when it shouldn’t.

I’m not sure what you mean by the fuel pressure comment above. One thing is certain with electronic fuel injection engines, which yours is: If the fuel pressure isn’t spot on, the engine will run very poorly, if at all. Suggest to double check the fuel pressure.

@GeorgeSanJose Cleaned the IAC and put in a new gasket (cleaned the throttle body last week). When I started the car the first time after reinstall the rpm jumped to 3000 and held for a few seconds before dropping to normal cold start idle, running for about 30 seconds and stalling. Every time I tried starting after that the engine would ‘chug’ a few times before stalling.
I will check back after I’ve tested the sensors and fuel pressure. Thanks!

Definitely check the rail fuel pressure with those symptoms. There’s a check valve which is supposed to prevent the pressure from dropping and gas leaking back into the fuel tank when the engine is turned off. That check valve may have stopped working. That will show up if you pressurize the fuel rail, then turn the car off and watch the fuel pressure for a couple of hours. If the check valve is bad, the fuel pressure will drop significantly.

There’s a second way the fuel rail can leak, into the intake manifold due to a leaky injector. This could cause higher than normal rpm when first starting the engine. If the fuel pressure drops significantly in the above test, that’s an alternative explanation.

And there’s a third way the fuel rail can leak, through the tube which connects the fuel pressure regulator to the intake manifold. Usually that tube can be removed and inspected for signs of gasoline inside the tube. There shouldn’t be even a drop of gasoline in that tube.

Not sure about the Legacy, but on my Corolla the EGR valve is easy enough to check. Just connect a hand-held vacuum pump to it, pump it up, and with about 18 inches of vacuum, it should open and stall the engine.

@GeorgeSanJose I did some tests with the multimeter but I am not sure if I am doing the tests correctly and unsure how to interpret the resistance readings. But it looks like I will run out of time. I am away from home for graduate school and I will have to head back home in two weeks. Was planning on driving, but I guess not! She’s just going to have to wait until the fall to solve this problem.

Forgot to mention: The car started just fine today and drove home with no issues. Took it around the neighborhood and it did just fine.