Jump to content
Ultimate Subaru Message Board
Sign in to follow this  

Need a quick second opinion on mechanic's diagnosis

Recommended Posts

Hello Subaru experts! I'm in a tight spot and would appreciate any speedy advice - I will be trying to complete the repair tomorrow morning. I will be verbose as usual. The short version should be plenty, however.


Short version:

My '90 Legacy L Wagon (2WD / AT) with 200k miles died on the freeway, started again after cooling, then died for good. The mechanic I went to found a trouble code 11 - crank position sensor - no reference (eventually). The sensor reads correctly on an ociliscope all the way at the ECU connector. Is this a clear indication that the ECU is bad?



I am the proud owner of a 1990 Subaru Legacy L Wagon - 2WD AT. It passed the 200k mark a week ago. I am good about oil changes but have neglected other scheduled maintenance, other than the air & fuel filters. I have had some front end and drivetrain work but never any engine work. After about 175k, it started getting a little cold-blooded in the mornings occasionally, and in the last 10k or so it has started idling rough every ten cold mornings or so.


The problem:

On Sunday morning I was driving back home from a family event. I got gass and drove another 25 miles north, to about 35 miles south of the town my brother lives in, cruising up the freeway at 75mph, when without warning, I felt the power begin to go away and the pedal drop under my foot as I ascended a gentle hill. Looking at the tach, I saw that the engine speed was fluctuating very sharply, on a distinct downward trend. It was fairly warm in the car, as is my wife's preference. A few moments later the engine had stopped and I pulled to the side. It felt much like running out of gas - which I have never done in this car. There was no check engine light, no temp, nothing indicated odd.


What I did:

I waited a few moments and then tried to start the car again. The engine would turn over, and seemed like it wanted to take, but would stutter over a couple times and then die. One time I noticed that the "power" light blinked rapidly - I never noticed this again. I looked at what I could under the hood - no fuel leaks, no lost fluids, nothing obvious. I started to look for a tow truck after a while, but my particular location made a tow truck driver's trip 80 miles long to tow me to the next town, while mine was only 35. Once I had spent a while on the phone, I tried it again, and the car started without complaint.


As I drove on cautiously, I found that the car hesitated a little, but seemed happiest at 3,000 rpm. I made it off the freeway but once I slowed down, it started hesitating and died about a mile down the road. I got my brother to bring his truck and tow me to his mechanic - one he'd recently began seeing with his Mazda.


Monday morning when they looked at it, it started fine and showed no codes. They asked me to drive it a while, and I drove it 45 minutes before it started running very rough. I just made it back to the shop and let the mechanics listen to it before it died - never to start again.


What the mechanics did:

1) The mechanics evaluated the fuel system first, and the fuel pressure test revealed that the pressure was fluxuating between about 20-30 when cranking. They determined that this was due to the pump cycling off due to a lack of the presence of spark.

2) Next they checked out the ignition system, and determined that the ignition coil was out of spec - resistance on the upper secondary pair was about 1.1, and the lower about 0.4 - and apparently were supposed to be at 0.7. He wanted to do more tests, but eager to get on the road, I let him work while I went to a junkyard and bought the coil from a running engine. Its resistances were off as well, and the junkyard guys told me that this is often the case even with working units. I figured for $35 I'd try it. But, it didn't help.

3) Next they replaced the "Igniter" / "Ignition module" because they weren't getting what they expected from it. Still no go.

4) Then he read the codes again. The computer - a Snap On "Modis" - indicated Trouble Code 11 - Crank Position Sensor - no reference.

5) He checked the sensor on a waveform both at the sensor and at the harness connector behind the intake manifold. Both looked good, but he procured a used-but-good sensor to try. When he changed sensor, fine powder of worn timing belt could be observed - I've known for a long time I need one, but have put it off.

6) He checked continuity from the harness connector to the ECU connector, which was good, as was the waveform at the ECU connector. 7) All those present visually examined the ECU and its pins and found nothing.


So, I'm in $600 labor and about to be in a $200 used computer. Does his diagnosis sound reasonable?

Share this post

Link to post
Share on other sites

I can get a used computer for $50.00 here in the Twin Cities, MN., assuming the junkyard I go to regularly has it. I'm off work tomorrow, if needed, I can get and check if they have a 90-92 computer. Let me know.

Share this post

Link to post
Share on other sites

Well, it doesn't sound unreasonable. Diagnosing problems, you have start somewhere!


That said, the engine also has a cam angle sensor, which the ECU should rely on if the crank sensor goes bad. Perhaps your cam angle sensor (driver side of engine) is also covered in timing belt filth?


If your belt is giving off rubber dust, you really should replace it before it snaps.


Still, your problem could stem from the ECU. Especially since the crank signal is arriving at the connector.

Off hand the only other thing I would suspect is the MAF. You mention something about 3000rpm, and a defect MAF means the ECU won't let the engine rev above that.

MAF is cheaper and easier to replace, so it could be worth a try.


I trust you have made sure that all the vacuum and PCV hoses are connected and not blocked?

Share this post

Link to post
Share on other sites

I'm impressed at the amount of diagnostic work done by the mechanic. Most every shop I've been to just likes to swap parts, charging you for all of them, until the problem goes away. Which is why I do all my own work. Of course, they are charging you labor for all that time....


I've only owned a Subaru for 18 months, so I don't have Subie specific advice for you. However, similar troubleshooting methods apply to all cars.


A bad ECU is a reasonable guess, given the evidence. However, I don't think you should pay for guesses when expensive parts are involved.


The only good way I've found to diagnose the ECU is to swap in a known good one. See if you can strike up a deal with the shop. They swap in a known good ECU, if it works, you buy it from them. Otherwise, they reinstall your old one, and you pay them only for thier time.


In my experience (old Volvo 240), the crank sensor failed intermittantly when hot, and the ECU didn't throw a code. After ruling out everything else by following the shop manual diagnostics, I started carrying my DMM in the car (my oscilloscope is too big). The next time it wouldn't start, I was able to confirm that the crank sensor was not providing a signal.


The Subary shop manual should have procedures for testing all of the sensors. If you plan to DIY, I suggest following the prodedures before changing parts. Otherwise, you have to weigh the labor cost vs. parts cost - for some parts, it may be cheaper to swap than test.



Share this post

Link to post
Share on other sites

Thanks for the quick replies!


This morning the mechanic called to say he'd been on the phone with Subaru tech support and they said that if the timing belt jumps a tooth and the cam and crank sensors are not in agreement, that might cause the ECU to read "no reference".


Now, I know if I plan to keep driving this car, I need to do that anyway, and he's willing to do it cheaper - and faster! - than anyone I've talked to about it before. So, I told him to go ahead and do it anyway. But does what he got from Subaru sound reasonable? I'm surprised they didn't put in a code for an angle disagreement.


Thanks for the input.



Share this post

Link to post
Share on other sites

That suggestion from Subaru Tech Support was exactly what I was going to suggest.


Since the timing belt is on its last legs anyway, replace it and go from there. With any luck, that will be the end of your problems.

Share this post

Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
Sign in to follow this