Weird dead-on-road problem 2002 Impreza

[jdemaris]

I've been doing mechanical work for 50 years - but this is a new one for me. If I hadn't seen it myself, I would NOT have believed it.

 

This is on my newest Subaru. A 2002 Impreza Outback with a 2.5. I just had the engine out a few weeks ago and put a new clutch in, along with a timing belt.

 

Our family went for a drive about 25 miles from home. On the way back, on a unplowed snowy dirt road we hit a big bump and the car bottom out for a second. Just after that, the tach and speedometer started going crazy and the ABS light came on. Then the car started losing power. I kept it going and when we starting coming down a hill, I turned off the engine in an attempt to "reboot" the electrical system. Well . . . it was dead and would not turn over. Good thing I was coming down a hill. I roll-started it and limped back home. When I got to my repair shop, I could barely keep it running. It would not idle, and was spitting, breaking up and sounded like it was running on maybe two cylinders.

 

I got it in the shop. Checked the battery and it only had 11 volts - i.e. stone dead. I put another battery in it and it started right up and ran fine. I then tried to pull codes of the computer and there were none. Then, scratching my head a bit - I stuck a voltmeter on the battery while running. 12.6 volts - i.e. not good and going dead. I checked voltage at the alternator post to alternator case ground and got 14.8 volts. This did not seem possible. I then checked voltage between the alternator main post and engine block ground and only got 12.6 volts. I then took a sharp knife, stuck it at the seam where the alternator is bolted against the engine block and sparks flew. Seems the alternator actually lost all electrical contact with the engine although it is bolted up tightly. I would of not believed this was possible. I then ran a jumper-cable from the alternator case (after I filed a clean spot) to the battery ground and battery voltage started climbing right up.

 

I call this bad Subaru engineering. The "no charge" light only comes on when the alternator output lead falls below proper charge voltage. Voltage to the remaining car's electrical system is NOT monitored. So, what happened here is - the alternator made proper voltage BUT the battery was not charging and went stone dead. We almost did not make it home yet the "no charge" light never came on.

 

I've had Subarus so rusty the drive axles broke off - yet I've never, ever, seen an alternator lose contact with an engine it was bolted to when tight. I would of said it wasn't possible. Now I've got to take it off and wire-brush all the metal surfaces where it contacts the engine. Funny thing is . . . this is my cleanest almost rust-free Subaru. I will note that when I pulled the engine to put a new clutch in, the aluminum engine was almost "welded" to the transmission with corrosion. I had to use to big chisels to separate. The steel bolts also pulled out the aluminum threads in the engine which I had to fix. Another design flaw as I see it. Steel touching aluminum when wet causes a galvanic reaction. Dis-similar metal parts when touching should at least be treated with an anti-corrosive compound and these surely were not.

[Olnick]

Fascinating problem jdemaris, I'm impressed with your ability to track it down--and thanks for sharing the solution with us. It very well may help someone else down the road.

[nipper]

This isnt as uncommon as you think it is on cars in general. It is just the Aluminum corroding on the surface and causing the higher resistance. You may want to add a ground from the alt to the battery as a backup solution.

[ocei77]

I will note that when I pulled the engine to put a new clutch in, the aluminum engine was almost "welded" to the transmission with corrosion. I had to use to big chisels to separate.

 

Yes to an anticorrosove here. The solution I've found with this and similar is to use a bottle jack against wood on the firewall and some strong point on the engine.

Has never failed me yet.

 

O.

[jdemaris]

This isnt as uncommon as you think it is on cars in general. It is just the Aluminum corroding on the surface and causing the higher resistance. You may want to add a ground from the alt to the battery as a backup solution.

 

Well it's uncommon in my experience. I doubt anything in the USA gets any rustier then it does here. Tons of road salt used all winter and many of our back roads are salted all summer. I've been driving Subarus since the late 70s and have yet ever worn one out. They just rust out 'tilll I can't drive them anymore. Most lose the back spring supports that go right through the body/frame. That being said, as rusty as they get, as well as my other winter fullsize trucks, I've never seen an alternator bolted tight to an engine what was so insulated from the engine that no charge took place. Two bolts holding it with multiple surfaces clamped tight, some of which should of been almost corrosion proof just by contact-sealing.

 

I am REALLY surprised with the low cost of monitoring a multitude of systems in the car via sensors and the ECM, that there is nothing that gives a driver warning when system voltage gets dangerously low. When I think of all the stupid "check engine light" causes that amount to little in real problems -there's really something amiss here in the design when this can happen and the car gives NO warning whatsover. I suspect we've been driving this car for a week with no charging system and were totally unaware. My 70s Chevy Blazer has a amp-gauge that would of told me there was a problem. My 80s Chevys and Fords have voltmeters that also would of told me of such a problem. But this 2002 with sophisticated electronics, sensors all over and a master ECM is not capable of such a thing? Heck, even my 1949 Case farm tractor has better warning systems.

[Fairtax4me]

I've had to pry the engine and trans apart on every Subaru I've pulled an engine or transmission out of.

I coat the crap out of the alignment pins with copper paste before I put them back together.

[nipper]

Everyone gives the ECU too much credit. It has only one purpose, and that is for emission controls. It doesnt car about voltage, because low voltage car doesnt run zero emissions.

 

Traditionally the Bat light on cars in general runs from handy to useless. In subarus they do sometimes do thier job, but get a bit confusing since it lights up the brake light too. They are called idiot lights for a reason (because they are idiotic and as a UI).

 

Only way to get around this is to get a voltage gauge.

 

To my industries defense, there is a lot of competition on the dash board real estate. 99% of the public really could care less about a voltage gauge, or even now, sadly, a temp gauge (I hope that makes a comeback, no reason why that can't be digital on the outdoor temp readout on the dasg). There is also the cost of adding one (a tremendous one as something has to be moved or redesigned to make room for it). Even with the earlier outbacks i had to go JDM to make them fit

[Dallasej25]

Hi everyone, sorry to wake up an old thread, but i'm in the middle of the same problem. Literally.

i have a 01 outback automatic. Two days ago while driving home from work batt and brake light appeared simultaneously. I made it home just fine and the next morning the car seemed weak while starting. (Low voltage). I drove to work and 30 min. later gauges went crazy, radio was on and off, and every time i stepped on the brakes the car made all kinds of racket, sputtered until finally died at a parking lot and left it there. After work we jumped it and it started just fine. Batt. Light and brake light still on. Drove four seconds and it died again. Scratching our heads went to my buddys house and grabbed a charged battery so i could make it home. Today on that same battery, my wife drove to the grocery store 3 miles away from our house, she said it died at the stop sign but started right back up. So, thinking it was a bad alt not charging the batt. drove straight to the parts store to pick up a NEW alt. Not reman, but new. Took me longer to find the right size sockets i needed than it did to change the alternator. Ok now, new alt on crank up the car and starts right up. Idiot lights vanish like magic.... wait!!. 20 seconds later they are BACK ON!! NOOO!!! filled with frustration i shut it off closed the hood and came inside the house. Decided to do a little more research before touching anything else on the car. Reading jdemaris post im really hoping i have the same problem, bad grounds.. i will take a look tomorrow morning and post results.

Edited September 24, 2015 by Dallasej25

[1 Lucky Texan]

dallas - is your car a 6 cyl (3.0) ?

 

that alt is different than the one for the 4 cyl - could be a mix-up with the new part.

[Dallasej25]

It's a 4 cyl. Im 99% sure i have the right part. I compared the two alts at the store. Has a two pin connection. Thou there is that slight 1% that tells me something isnt right..

[1 Lucky Texan]

alt. seems OK then. usually,

 

batt cables look OK? good connection to ground on the black cable, etc.?

[idosubaru]

If throwing parts and guessing isn't working you should do some proper testing:

 

1.  test the alternator output.   how many volts is it putting out - check it multiple times/conditions - like when the lights on the dash are on.

 

2.  test the battery

 

3.  battery posts and terminal end connections can be corroded or loose

 

4.  run a jumper cable from positive alternator post output to battery positive terminal - do the lights then go away on the dash?

[idosubaru]

Draw a line across the face of the crank pulleye - chalk/sharpie/etc.  

 

Run the engine.

 

If the line breaks up - the pulley is separating.  The belt isn't turning enough to produce much output from the alternator.

 

This produces - alternator failing symptoms, but replacing the alternator doesn't fix it...because the issue is still present.

 

Testing alternator voltage first - would point to reduced output.  Which tells you to check the pulley or test the alternator off the vehicle in the parts store.

[Dallasej25]

Ok. Battery is still good. Car has been starting n running on pure battery power so far..

 

I did the crank pulley test. Passed.

[Dallasej25]

Marked the crank pulley and ran car a few min. Acelleratin a bit. I have no voltmeter to check for voltage drops. But i figured i could check the grounds by hooking up a wire from the alt. case to the -battery post. No change.

I jumped the +output post from the alt. To the +battery post. No change the lights are still on.

 

The ONLY thing i believe could be wrong is the exiter wire towards the alt. I put a screwdriver on the back of the alt with engine running and nothing. The alt isnt getting exited. Thats why jumping a cable to the possitive batt terminal did not work.

I wonder if someone can tell me how to jump or bypass the two wires on the alt connection... thanks.

[kanurys]

The field coil (exciter, note the spelling) is powered through the battery light indicator bulb on the dash in most Subarus. This is a 100 ohm load, typically. With the key on and a good battery, the disconnected exciter wire from the alternator should read near battery voltage. When connected it should drop a bit.

 

You could jump the exciter terminal on the alternator to the positive side of the battery (MAKE SURE YOU KNOW WHICH TERMINAL TO CONNECT TO!) and test the output. if that works, then trace the bad wire until a fault is found.

 

I've been haunted by a similar problem as the OP and am going to try new battery cables on my 1999 phase II EJ out of an impreza outback sport. It's swapped into a 1987 GL, so I have been reluctant to have anyone else look at it, as I know it best. Good luck on yours!