Engine Mysteriously Loses Power, Dies in Heavy Traffic, Intermittent Problem, No Engine Code

[dirty_mech]

Hi guys,

My 2000 EJ251 Legacy L Wagon has been losing power and dying in heavy stop-and-go traffic.  The engine losing power and dying has happened to me four times.  The car seems to stop behaving this way if it is allowed to cool and heavy traffic is avoided. The engine power loss has never tripped an engine code.

I suspected the spark plugs first; they have been replaced.  They were at least 4 years old, and one had melted to the point that it must've been triple the gap.  I also replaced the fuel filter for good measure, and ensured the air filter was clean.  At first, this seemed to solve the problem.  However, it happened again in heavy traffic today, and in a much more sudden manner compared to previous two failures which were preceded by a gradual loss of power.  After five minutes of rest and several starting attempts, I was able to get the engine running.  It had so little power and RPMs, until I somehow got it into higher RPMs, which seemed to correlate with a momentary return to power.  I was able to limp it off the freeway before it failed again on a side street.

I took out the new plugs and checked them; they seemed dirtier than when I put them in a few days ago but not damaged or out of gap.

My next suspect is the blue and black spark plug wires.  They could be as old as the car; I've never replaced them.  I also noticed that after replacing the plugs, I forgot to put the run of the battery cables into the plastic hook on the left, engine-facing side of the wiper fluid reservoir.  This left them touching and pressing up against a couple of spark plug wires, which I am guessing could exacerbate spark leak.

If that doesn't solve the issue, I might also guess that the ignition coil could be bad.  Since this problem is intermittent, it might be difficult to pin down the offending part.   My only other guess after that would be the ECM or wiring harness, or possibly and intermittent failure in the fuel pump.   The car runs fine the vast majority of the time, so this doesn't seem to be an engine mechanical issue.  It seems like it must be an intermittent ignition, electronic, or fuel delivery failure.

If you have any ideas, please let me know!  I don't want to have this car randomly die on me again in heavy traffic for a fifth time.  
 

[el_freddo]

It could be a number of things:

- this would throw a CEL usually - the idle air control valve. It could be faulty or have a poor seal, allowing air to bleed past the seal when it’s not needed. 

- ignition coil pack could be breaking down under heat stress. If the 2000 has an igniter like the earlier models this could also be an issue.

- fuel pump could certainly be the issue. I’d probably start here if you don’t know the history of the pump. But that said, my sister’s 2000 Liberty has over 400k km and we don’t know the history of the pump on that one. 

If you can swap ignition coil packs with a mate to test that it’s the issue go that way. Same for the igniter (again, if your car has one). 

Cheers 

Bennie

[1 Lucky Texan]

scanned the ECU for pending codes?

[mikec03]

I have a 251 engine in an 2002 so I'm familiar with the engine.  If that engine had a MAF, it would be the first thing to replace, but it doesn't.

First, you have tighten all the supply and ground wire connections

Second, you should replace the ignition wires.  Why didn't you do this already?  OEM only.

Third, I had a serious hesitation problem which took me months to figure out.  Of course this is not the same as stalling.  The cause turned out to be the upstream O2 sensor [which I changed after I changed everything else].  So I would replace this O2 sensor as a general principle.  At that age of the car it's a no brainer anyway   

 

[dirty_mech]

Hello everyone!  Thank you for your input.  Here are individual responses:

el_freddo: 

I will inspect the Idle Air Control valve and keep it in mind and on the lower half of my list for if the bigger suspects don't shake out.  I looked it up and can't find it for cheaper than $150 for a new one on a manual transmission; ouch!  Might have to get a junk yard valve in that case; they are $31 themselves at Pick N Pull which is pricey for a junk yard part.  I will get a coil pack from my junk yard for sure; $13 instead of 70.  On RockAuto it looks like they are calling ignitors an Ignition Control Module (ICM).  I looked up my 2000 and a 1994 Legacy's ignition parts section, and see that is an additional part only on the '94.  So that is off my list! 

I'll also replace the fuel pump as well, with a Delphi unit, because replacing it is relatively cheap but also a big project.  The gas tank needs to be cleaned prior to a fresh install.  I can't do that job properly on the side of the road, and so don't want to risk it failing.

I wish I could swap packs with a mate's but I've got no mate and I've broken down twice on a ten-lane bridge.  Private companies aren't even allowed to tow in that situation, and I've heard it's like a $500 charge from the state transit agency to get you off the bridge.  I'm better off just buying a bunch of parts, hoping one of them is the issue, and chalk up the rest to preventative maintenance.  But I do love the idea of swapping parts on an identical vehicle to identify the problem; that's bloody brilliant.  Makes me want to buy a second Legacy. 

1 Lucky Texan:

I haven't scanned it for pending codes.    I hadn't heard much about pending codes until recently.  I need to get a new reader, and might run out to wally world for a replacement tonight since you mention it.  Mine has been stolen twice after leaving it in the glove box all the time.  Unless you live in a very high trust area then be sure to leave your code readers at home, folks.

mikec03:

Correct, it has the MAP sensor, not a MAF sensor.  Is a MAP sensor failure not likely to cause this type of behavior?

I will absolutely make sure to install new wires and make sure they are tight.  The vehicle behavior only started a little over a week ago, and I've been super busy, so haven't had a chance to address the issue.  I will also replace the upstream O2 sensor, but only if the other issues fail to resolve the problem, and even then with a used junk yard sensor.  A new one for me is $100 before shipping.  I'm guessing you didn't include the downstream O2 sensor because it is not an engine-killer when it fails, and is only for emissions purposes?
 

Thank you all again for your input.  I will update you once I've inspected the IAC valve, scanned for pending codes, replaced the plugs and fuel pump, have the spare oxygen sensor and ignition coil in hand, and have tested the vehicle again.

Cheers!

Edited April 25, 2019 by dirty_mech

[FerGloyale]

14 hours ago, dirty_mech said:

 

I'll also replace the fuel pump as well, with a Delphi unit, because replacing it is relatively cheap but also a big project.  The gas tank needs to be cleaned prior to a fresh install.  I can't do that job properly on the side of the road, and so don't want to risk it failing.
 

 

No need to drop the tank and clean.  In CA, you won't have any gunk in that tank.

The problem will likely be the metal cap on the "surge" canister built into the fuel pump pickup assembly.  It's a known issue for these cars.  The ear on the cap will have a split, and will no longer lock onto the plastic tabs.  You;ll see when you get at the pump.  BTW, the pump is accessed from under the rear seat.  panel in the floor come out, then the fuel pickup is pulled out from the top of the tank.

Unfortunately that cap is not available to buy separately that I've found.

I very carefully, after cleaning and away from the tank, put a few tack welds over the cracked part of the cap, and installed it back ont eh unit with a new O-ring (use sili-glyde silicon lube on O-ring) 

[1 Lucky Texan]

this cap for the next generation cars will work;

https://www.amazon.com/dp/B00L2OWO2U/?coliid=I11LNGBWZESKLY&colid=IDEFQ97CY9SH&psc=0&ref_=lv_ov_lig_dp_it

 

throw the strainer/filter away, get a new o-ring for the cap; most of us use a 928 Viton metric o-ring.

[Rooster2]

I had similar problem with my 08 Outback. My car would not give me a code to tell me what the problem was. I changed the fuel pump, knock sensor, plugs and wires, and coil pack, and my car still ran like crap. I finally gave up and went to a Subie dealer. The problem was fixed by replacing the PCV valve. Hope this info helps.

[3Pin]

I just had a very similar issue and it was the fuel pump.

[dirty_mech]

tldr;  It was the oxygen sensor.

The car started losing power again on the freeway, I pulled off and it died.  Did it three times in warm weather street traffic.  I tested the spark plug wires' resistance in ohms at piping hot temperature while waiting for the vehicle to cool down a bit so it would become driveable.  I later tested their resistance while cold, and also tested the resistance of a new set.  Since I couldn't find the information online, perhaps someone will find this a good reference.  Here were the numbers on a set of NGK 8691s:

#        Old Hot    Old Cold   New Cold

1         4,150         4,120         3,820

3         4,210         4,180         4,080

2         4,990         4,950         4,790

4         5,280         5,210         5,120

The spark plug wires' average increase in resistance due to high temperature was close to 1% of total ohms.

Once the car was driveable, I made it to AZ and picked up my Russian parts roulette set of NGK spark plug wires, Delphi fuel pump, AZ brand ignition coil, and a Denso oxygen sensor.  Got a wally world code reader after that; no pending codes.  Couple more sputters on the street. 

Pick N Pull was next, which wasted my time for used parts.  The catalytic converters all get snipped out of the cars, along with the oxygen sensor that screws into it.  Also, at the end of the day, they apparently no longer send the tractor around with a horn to call closing time.  Instead, they just close the register.  I was waiting to hear the closing horn and never heard it.  I got there three or four minutes after the hour, and was told leave the part and come back tomorrow to pay.  Maybe that's what they need to do thanks to some tweaky agro locals who ignore the horn, but FFS at least play one for the people who won't.

Well I got lucky when I played parts roulette.  I was hoping to install one part at a time and test for resolution of the problem, but at this point I was really tired and it was late.  I opted to install two parts at a time and then test.  I went with the spark plugs and oxygen sensor together for my first attempt.  The old plugs didn't seem bad, but they didn't make a great solid connection on the ignition coil, either, and I didn't want to install expensive parts before cheap ones.  The oxygen sensor I decided to install after first searching how often to change them, since it's not listed in the maintenance section of the Chilton.  Most said 100k miles or so.  I read deeper and came across this gem of a sentence:
 

Quote

"Other signs [beside a CEL] that you need a new oxygen sensor include a rough idle, misfiring spark plugs, lack of power, stalling or a significant increase in fuel consumption."

Boom.  That's basically all my symptoms, with the caveat that they happen when the car is hot and go away when it is cool.  Loss of power was followed by rough idle, then sputtering and stalling.  Misfires happened while attempting to restart the engine.  This cycle would most often happen under engine load at full temperature, or at low RPMs, and especially when trying to accelerate hard under engine load at low RPMs.   Also I can't say that I noticed it being crazy with the gas.  It was doing 25 mpg highway and almost 20 in urban traffic, which I figured was good for an old 2.5, but maybe it's improved a lot now.  I'll keep an eye on it.

The sensor went in, and I test drove the car last night and today.  No sign of the problem.  I swear I've heard before that the oxygen sensor doesn't work reliably until it is at operating temperature, and thus doesn't factor into ECU engine management until the car is warmed up.  Maybe that would explain why it performed better when cool.  Or perhaps the failure is due to the repeated heating and extra resistance present when hot.  The sensor pigtail was covered in greasy engine crud as well.  I have no clue how old it was at the time.  The inside of the connector was bright orange; perhaps factory?

I did a resistance test on the old and new oxygen sensors before installing them.  Looking inside the electrical connector for the sensors, with the side with two plastic grooves facing up, I numbered the terminals as follows:
    ___|_|___
__|   1    2  |__
    |   3    4  |

They had identical readings; only 1.5 ohms of resistance could be detected between terminals 1 and 3.  No other resistance measurement registered on my meter's maximum reading range of 2 million ohms.  Perhaps it would have been different had I read the old one while it was hot instead of cold.  The only visible difference between them was the old one had grime on the wire lead and old one looked frosty white too:

I don't know how to test these oxygen sensors any better than I tested the spark plug wires.  Nothing in the Chilton about how to test them.  Couldn't find a guide online for free; Subaru wants you to pay a $35 subscription to access their service manual.

All said and done I got lucky.  This could have just as easily have been the fuel pump like you guys were saying.  I am guessing the differentiating factor was how it happened only when the vehicle was hot.  I will probably do the fuel pump and that strainer cap's O-ring while I'm at it too, just to avoid another problem.  Cheapest I can find a Viton 928 ring for is $8 for a ten pack on eBay; maybe there's a place I could find just one but I'm not sure.

Well that's my report.  Thanks again for all the help guys.
 

 

[dirty_mech]

Hey guys,

I'm grabbing a sensor from RockAuto to swap with AZ since they charge over $30 more for the part (they are actually cheaper on the fuel pump though), and I saw this part

https://www.rockauto.com/en/moreinfo.php?pk=10353284&cc=1358210

There's your replacement surge filter cap, complete with O-ring and even a fresh gasket, for 30 bucks.  The part description on Amazon even says it has a stronger cap due to the original ones breaking:

https://www.amazon.com/Dorman-902-438-Housing-Repair-Select/dp/B07KG7K8GB
 

Think I'll scoop that part with the sensor as well. 

[heartless]

Quote

Couldn't find a guide online for free

here ya go.. pretty much any year, any model.. (aside from the newest)

http://jdmfsm.info/Auto/Japan/Subaru/

[1 Lucky Texan]

15 hours ago, dirty_mech said:

Hey guys,

I'm grabbing a sensor from RockAuto to swap with AZ since they charge over $30 more for the part (they are actually cheaper on the fuel pump though), and I saw this part

https://www.rockauto.com/en/moreinfo.php?pk=10353284&cc=1358210

There's your replacement surge filter cap, complete with O-ring and even a fresh gasket, for 30 bucks.  The part description on Amazon even says it has a stronger cap due to the original ones breaking:

https://www.amazon.com/Dorman-902-438-Housing-Repair-Select/dp/B07KG7K8GB
 

Think I'll scoop that part with the sensor as well. 

Nice find!