Help, please. 90 Leg, no-start... "open" between relay and fuel pump

[NorthWet]

Pretty sure this is ECU/connector/sensor related.

 

Wife's 90 Legacy LS, 2.2L (ODBI), 4EAT. Failed to restart twice after picking up gas (from the same station, but I think is just a coincidence). First time was after <1mile drive from cold start, second time was after 6mile (15 minute) drive from cold start. First time by the time I had gotten there she had already restarted it. (She had retried after about 10-15 minutes of letting it sit.) Second time it refused to start without supplemental gas dribbled into TB; w/o extra gas, when first activating the starter, the engine would "catch" for maybe 1 revolution and then crank without further feeling of combustion. With supp. gas, it would run for a second or two before dying. I left it and returned in the morning, when it fired up first turn of the key; son drove it 8 miles home (diff route) without incident.

 

Printed Legacy777's directions for pulling codes, and with black connectors got a 22 (knock sensor; should still let me start?). Reseated K.S.'s connector, still getting ECU code 22. Tried green connector, and failure reoccurred; unplugging green conectors, still no start.

 

Currently, I have the battery disconnected after cleaning and reseating engine harness to chassis harness connections. Decided that while the battery was out I would let the ECU reset/clear.

 

As far as I can tell, the fuel pump was cycling off/on properly while green connectors were together, so I do not think that it is fuel pump or its control. However, I have not checked for pressure or flow yet, but this is low on my list at the moment.

 

My thought is that the ECU is sensing something on startup after the first revolution or two and then shutting down the injectors. Since the engine runs with supplemental gas, I am pretty sure that spark, compression, and valve/ignition timing are still happening/adequate. The fact that the engine (so far) has not quit once it is started makes me think it is something during startup only.

 

Anybody run across this type of failure?

[Manarius]

Assuming it turns over every time, I would say either the crank or cam position sensor. Both are easy swaps.

[mjreilly]

If it's only happening when the engine is warm it could also be the coolant temp. sensor.

[NorthWet]

Assuming it turns over every time, I would say either the crank or cam position sensor. Both are easy swaps.

I would expect that these would throw codes and also not fire the spark plugs... though maybe unreasonable expectations.

[NorthWet]

If it's only happening when the engine is warm it could also be the coolant temp. sensor.

Only 3 total occurances, so hard to say. But the first restart was after only a mile of town driving.

[Manarius]

I would expect that these would throw codes and also not fire the spark plugs... though maybe unreasonable expectations.

My toasted MAF didn't throw a code until I broke it.

[ron917]

I would expect that these would throw codes and also not fire the spark plugs... though maybe unreasonable expectations.

 

I've had crank and cam sensors go bad on my Subaru and on other cars. Classic symptom is a no start after a hot soak - letting a hot engine sit while you run into a store or whatever, then no start afterward. Let it cool down for 20 minutes, and then it starts right up and runs fine.

 

I never got a code for the problem. Apparently, it sorta kinda works, so the ECU thinks it's OK, but it's not working well enough to provide a proper timing signal.

[Richard1296]

Just for kicks try listening for the fuel pump when you start

the car normally to see if you can hear it. Then listen for it

when it won't start. I'd replace the fuel filter before replacinng

the pump and even replace it if the pump proves to be bad.

 

Richard

[nipper]

I would expect that these would throw codes and also not fire the spark plugs... though maybe unreasonable expectations.

 

They wont throw codes on a no start. There is only a code when they malfunction when the car is running. Thats what drives people nuts. Next time it happens tell us if you have spark.

 

nipper

[NorthWet]

They wont throw codes on a no start. There is only a code when they malfunction when the car is running. Thats what drives people nuts. Next time it happens tell us if you have spark.

 

nipper

Thanks nipper. Pretty sure I have spark. I haven't checked with "Mark-I eyeball" yet, but as stated in my monologue supplemental gas allows the engine to start and run until supp. gas is used up. BTW, when I checked several hours after last no-start it started with a turn of the key.

[Bill90Loyale]

Pat-

1. MAF sensor dirty/failing?

2. CTS pooped?

3. If coil pack equipped, bad pack(s). If not, check high tension lead for short/failure (wire from coil to dist).

[NorthWet]

My initial thoughts were crank/cam sensors, except lack of codes (wouldn't occur per nipper), and I would expect spark cut in addition to fuel cut (will run with added gas, so spark exists).

 

Looks like it is time to do some swap-and-see... just that it may not "prove" anything with an intermittent. :-\

[nipper]

Thanks nipper. Pretty sure I have spark. I haven't checked with "Mark-I eyeball" yet, but as stated in my monologue supplemental gas allows the engine to start and run until supp. gas is used up. BTW, when I checked several hours after last no-start it started with a turn of the key.

 

Spark will help us determine if its the temp sensor or the carnk/cam sensor. Since you say you have spark, look at the engine temp sensor.

 

nipper

[idosubaru]

i've seen faulty sensors (bad connections/wiring actually) cause a no-start but not give an engine code before as well.

 

i have little to add that hasn't been mentioned but i would look at:

crank/cam sensors

coolant temp sensor

coil pack (unlikely though if it's running with supp gas).

 

where are you adding the gas anyway?

[NorthWet]

I still have to arrange to verify that spark is occurring instead of just inferring (I hate non-factual ''facts" ), but work schedules have made this difficult. Kind of a pain to have a system that doesn't tell you why it won't start.

 

Thanks nipper. Still seems a little silly to have spark and no injection due to any sensor failure, but there must have been a reason.

 

Gary, dribbling gas into TB airhorn, reseating duct, and starting it for a second or two before gas is exhausted.

[nipper]

I still have to arrange to verify that spark is occurring instead of just inferring (I hate non-factual ''facts" ), but work schedules have made this difficult. Kind of a pain to have a system that doesn't tell you why it won't start.

 

Thanks nipper. Still seems a little silly to have spark and no injection due to any sensor failure, but there must have been a reason.

 

Gary, dribbling gas into TB airhorn, reseating duct, and starting it for a second or two before gas is exhausted.

 

the cam or crank position sensor have to do with spark, the engine temp with fuel delivery. A no start means no codes are stored since a lot of sensors will give bad data till the car starts, hence why they are ignored.

 

nipper

[Cougar]

NorthWet, after reading through these posts it sounds to me that the injectors are not opening due to a electrical problem. Have you checked for voltage getting to the injectors while cranking the engine?

[NorthWet]

NorthWet, after reading through these posts it sounds to me that the injectors are not opening due to a electrical problem. Have you checked for voltage getting to the injectors while cranking the engine?

Not yet... worked 13 hours of overtime at my second job last night, wife's second job is in retail at a mall, so diagnostic time and "testing with help" time has been almost nonexistant so far. I plan to do more testing and parts swapping tomorrow. Doubt that a DMM will show much, as the general feel is that injectors work for a cycle or two (i.e. - it tries to "catch" first revolution or two each time a start attempt is made).

[mjreilly]

I'm still pulling for the coolant temp sensor. My 95 Outback would often start right up when cold, I would drive a short distance for gas and it wouldn't start up at the gas station. Like Nippler said, if it's getting spark, it may be the coolant temp sensor telling the ECU it's dead cold when its not and flooding the engine.

 

good luck,

 

matt

[NorthWet]

... it may be the coolant temp sensor telling the ECU it's dead cold when its not and flooding the engine...

Its not flooding: Dribbling gas into the TB airhorn and resealing allows engine to start immediately(dies once dribbled gas runs out). Indications are that it is a fuel cut after a revolution or two.

[Cougar]

Not yet... worked 13 hours of overtime at my second job last night, wife's second job is in retail at a mall, so diagnostic time and "testing with help" time has been almost nonexistant so far. I plan to do more testing and parts swapping tomorrow. Doubt that a DMM will show much, as the general feel is that injectors work for a cycle or two (i.e. - it tries to "catch" first revolution or two each time a start attempt is made).

 

If the injectors are turning off as you think they are then I would have to think the problem is with the ECU since the spark is working ok. Using a analog meter or a noid light to watch the pulses to the injectors should show the story. I don't know if there are external current limiting resistors for the injectors on this model or not.

 

I don't think a bad coolant sensor will cause this trouble. It will run rich but it still should run.

[mjreilly]

I don't think a bad coolant sensor will cause this trouble. It will run rich but it still should run.

 

Since his is starting with gas dribbled into the TB maybe it is something else but a bad coolant temp sensor can cause a no start condition when the engine is hot. I had it happend to me and it drove me crazy for weeks until I replaced the sensor.

 

Maybe the time it takes to undo the air intake and having the air intake off is allowing the extra gas to evaporate. Of course, why would it then start with more gas dribbled in. Maybe the extra time it takes to do the above is allowing the engine to cool down and the engine temp is getting in line with what the sensor is reporting to the ECU. Hmmmm.

[NorthWet]

Maybe the time it takes to undo the air intake and having the air intake off is allowing the extra gas to evaporate. Of course, why would it then start with more gas dribbled in. Maybe the extra time it takes to do the above is allowing the engine to cool down and the engine temp is getting in line with what the sensor is reporting to the ECU. Hmmmm.

10-30 seconds. Runs out of added gas and dies, giving same issues with no-start.

[Cougar]

I checked my service info and saw there is a yellow, common return wire, for the injectors that goes to pins 5 and 9 of the ECU. The other side of the injector coils are pins 13, 14, 15, and 16. I would check voltages at those pins with reference to the return wire.

 

Since there are no codes that would cause the ECU to turn the injectors off then it would seem that if the injectors are cutting out, the problem would have to be inside the ECU itself.

[nipper]

Its not flooding: Dribbling gas into the TB airhorn and resealing allows engine to start immediately(dies once dribbled gas runs out). Indications are that it is a fuel cut after a revolution or two.

 

Next step is to put a test light on a fule injector and find out if the injectors are cutting out or the fuel pump is cutting out.

 

nipper