Mysterious start-and-stop cylinder misfires on a 2001 Forester
December 17, 2014 4:46 PM   Subscribe

Car Talkers of MeFi, help me solve a mystery! My 2001 Subaru Forester had a cylinder misfiring: car shook, check engine light flashed. Then it healed itself and the mechanic couldn't tell what was wrong, or that anything was wrong at all, so I brought it home. Then it started misfiring again tonight. Then it healed itself partway, but the check engine light is still on. What should I think and what should I do? Details below the fold.

Two nights ago, on the way home from work, my 2001 Subaru Forester started to judder and shake while I was stopped at a light. A few seconds later, the check engine light started flashing -- something I'd never seen before -- and the car was clearly laboring, not driving at full power.

When I called the garage the next morning, they told me that a cylinder was misfiring, and that driving the car was likely dumping raw fuel into the catalytic converter, potentially destroying it, and that I should have AAA tow the car to the garage. So I did -- and when they started it and drove it around, it ran completely smoothly! He said misfires like this are usually caused by bad spark plugs, but my spark plugs are new. He also said that the car's computer didn't have any record of any codes being activated the previous night. He said sometimes a car just "loses its mind" and the odds were there was nothing wrong with it. So I brought it home. Drove it around on several errands today, ran completely smoothly, but coming home tonight, same thing happened: car shakes, check engine light flashes, engine feels underpowered. When I parked at my son's school, I could smell a kind of burned rubber smell in the back of the car. Leaving the school, though, the engine smoothed out, the check engine light went from flashing to steady, and the car ran, as far as I could tell, normally.

What is going on here? Should I have the mechanic change the spark plugs even though they're new? Is it OK to drive it to the garage tomorrow so I don't have to wait for the tow truck again? Is it normal for a misfire to come and go like this? Is my car gaslighting me? Am I right that it would be immensely stupid to try to drive this car 90 miles to the airport on Saturday?
posted by escabeche to Travel & Transportation (22 answers total)
It could be one of a number of things, but in general you are looking for something that will disrupt a single cylinder.

A bad ignition wire -- or one that is only partially connected to its spark plug or coil or distributor cap -- could be easy to spot...but then, one would hope that a professional shop would do a quick check of the plug wires to make sure they're all tight. This would typically register an error code, though. Burt the wires themselves do go bad, especially if they are the stock wires with 13 years of use on them -- they get hard and brittle, and their carbon core can break or even shatter..

A bad spark plug is less likely, but spark plugs do have a connector at the end the plug wire connects to and this could be a source of trouble. On some plugs this end is threaded and the terminal screws on to the top of the plug. This can come loose and lead to a spotty electrical connection.

A bad injector, a bad electrical lead, or a bad injector electrical connector could cause this, but it would likely leave an error code.

A leaky head gasket -- pin hole leak, not yet a complete HG failure -- could cause something odd like this. I've had this happen myself, so it's possible, but not common.

I dunno -- those are some ideas off the top of my head. I'd start with plugs and injectors and verify those before diving any deeper.
posted by mosk at 5:02 PM on December 17, 2014 [2 favorites]

I'm far from a car-talker, but I had a similar problem years ago, in college, with my '87 Toyota MR2. I replaced the spark plugs, didn't help. Finally got around to driving it (while misfiring some of the way) to the nearby dealer. By the time I got there, the problem had gone away. After keeping my car for several days, they diagnosed the spark plug wires as the issue. Replaced them, and I didn't have that problem any more.

I can't say that's the only potential cause of the problem, but it's worth looking into. Changing the wires yourself is a fraction of the cost of paying someone else to do it.
posted by WasabiFlux at 5:17 PM on December 17, 2014

I'd start by replacing the spark plug wires, they're cheap and can cause intermittent issues.

I'd pull the plugs and check them to see if they're fouled or something, and double check the wiring connection.

Cars don't just "lose their mind". That's not a confidence inspiring statement by the mechanic.
posted by TheAdamist at 5:18 PM on December 17, 2014 [3 favorites]

So it sounds like the car is doing it about the same time of day, is it also maybe raining at that time? Wetness tends to make wires not work as well also so that would add to that diagnosis. In addition to the wires this car has the coil on the spark plugs themselves and that could definately be the the problem and coil on plug arrangements usually have a little clamp mechanism on the plug connection that can cause this problem if it either gets broken during service or is not put on properly so all things to check-and they weren't kidding about the misfire problem-it can cause really, really expensive damage so when it happens the best thing to do is pull over to a safe location (like a parking lot) and stop the car and either let it idle or kill it and restart it.
posted by bartonlong at 5:56 PM on December 17, 2014

Total shot in the dark, but my RX-8 once went mental and shook like you describe even to the point of stalling. Stop the car, start it problem solved!? Long story short, my alternator had been failing for a while and my battery was on the verge of death. New battery, new alternator problem fixed permanently. I never would have connected the missing/stalling and a bad battery.

So check your charging system!
posted by Zedcaster at 7:23 PM on December 17, 2014

Have your smartphone ready and take a video the next time it happens. This has helped me a great deal by being able to show the mechanic exactly what the issue is.
posted by mlis at 7:53 PM on December 17, 2014

My Honda CRV did exactly this (and it's in the shop for the second time). The first time, they couldn't find anything wrong with it. This time, they figured it out and I'm getting a new computer.
posted by mmmbacon at 8:00 PM on December 17, 2014

>coming home tonight, same thing happened: car shakes, check engine light flashes, engine feels underpowered.

Same thing in the same circumstances? That is, was the car stationary, idling?

What did the car do that made you think to say "feels underpowered?" Was your car moving at the time? Did it feel unresponsive to your commands to the throttle?
posted by the Real Dan at 8:38 PM on December 17, 2014

Same thing in the same circumstances? That is, was the car stationary, idling?

Yep -- both times, the shaking started when the car was stationary and idling, and was definitely worse when the car was stationary and idling.

What did the car do that made you think to say "feels underpowered?" Was your car moving at the time? Did it feel unresponsive to your commands to the throttle?

I felt I had to press the gas pedal harder than I expected in order to get the car to move at the speed I wanted. For lack of a better word, it felt kind of sludgy.
posted by escabeche at 8:42 PM on December 17, 2014

I just replaced an engine in a 2005 Forester that had the same problem. Has your car been leaking oil? How's the oil level now?
posted by lester at 9:18 PM on December 17, 2014

>A few seconds later, the check engine light started flashing
> He also said that the car's computer didn't have any record of any codes being activated
> the previous night. He said sometimes a car just "loses its mind" ...

These guys don't sound trustworthy. Consider changing "garages".
There should be persistent codes from your ODBII system from this.

There are a number of causes for the behavior you saw, and a number of subsequent ODBII reportable conditions as a result of those causes. Autozone will "rent" you an ODBII tool that will allow you to read the codes yourself, after the next occurrence. After that, you'll be a little better informed when you talk to a mechanic that says "it just went crazy..."

IAC sticking is my favorite for running rough at idle, but thermal or intermittant failure of high-voltage ignition components can do this.

Where's Brockles in this discussion? He's got good insights on these kinds of questions.
posted by the Real Dan at 9:27 PM on December 17, 2014

Are you handy at all? i'd drive the car to an oreilly's(or similar $BIGCHAINAUTOPARTS store that does free testing) and test the battery, then test the alternator(which obviously you'd have to pull, which means you might need to release the serp belt tensioner).

If both those pass, new spark plug wires.

A bad battery caused similar issues on my dads 99 forester, bad plugs and wires caused similar problems on my 60s plymouth, and i've heard of bad plug wires causing problems like this on lots of other cars both offline and on.

I'll note that my dads forester has had several electrical issues, which were finally resolved when it was discovered wires going in to the computer had just kind of... frayed out of the socket because ??? sorted that several years back and it was solid until the battery issue, which was just an old tired battery.

The first things i think of with intermittent firing on idle are plugs, battery/alternator, and plug wires though. And the good news is no combination of the above could exceed $100 on an older subaru if you're doing the work yourself(and shouldn't be hugely expensive if you're not)
posted by emptythought at 11:03 PM on December 17, 2014

Failing ignition coil? That was my problem when my car was having cylinder misfires, although they always registered a code in the OBDII. I n'th finding a new mechanic. IME any time the check engine or service engine soon light comes on you'll have at least one error code stored. Best of luck to you!
posted by InsertNiftyNameHere at 11:47 PM on December 17, 2014

When this happened to my car, it was a failing ignition coil, and it did indeed happen intermittently. However, it didn't go away. (And one of its friends joined it about three months later, so if it is the ignition coil, it might be worth getting a price on replacing all of them.)

I was told that it was OK to drive slowly for short distances to get to the repair shop in lieu of a tow, but my shop is under a mile away, and also this may not be what's actually wrong with your car, so YMMV.

Mind you, this threw a code, but if you mechanic thinks "cars just be crazy sometimes!" I'd have only limited faith in his ability to use a code reader.

I've also had similar symptoms from a wire that was shorting in the rain, so that's another thing to consider.
posted by pie ninja at 5:28 AM on December 18, 2014

One note: this is a shop I've been using for ten years and that I trust, not a random place I had the car towed to. The mechanic who examined the car is the owner. If this guy says there wasn't a stored code, I'm pretty sure there wasn't a stored code.
posted by escabeche at 5:37 AM on December 18, 2014

Our 2001 Forester has a bad O2 sensor - which we've replaced. Check engine light is on when that sensor is messed up and I would think it could lead to problems. It's a known Forester of that vintage thing. The other major problem with that year is head gasket issues so that might be another place to investigate.
posted by leslies at 7:18 AM on December 18, 2014

Started it this morning, engine running smoothly, but check engine light on solidly. Brought it to the shop, now the code is there, it's complaining about cylinders 3 and 4. So they know where to look, they're going to check spark plugs, wires, coils, etc., hopefully mystery gets solved today.
posted by escabeche at 7:51 AM on December 18, 2014

Going with plug wires and ignition coils in that order. Symptoms match, especially if the coil is deteriorating.
posted by Thistledown at 9:01 AM on December 18, 2014

You can change the spark plug wires yourself in 5 minutes with no tools, and the plugs themselves in 15 (you need a spark plug socket and a socket wrench).
posted by zug at 10:15 AM on December 18, 2014

> You can change the spark plug wires yourself in 5 minutes with no tools, and the plugs themselves in 15 (you need a spark plug socket and a socket wrench).

Only if the 13 year old plug wires don't fall apart as you try to remove them and leave broken pieces deep in the plug wells, where they can be very difficult to remove. Been there, done that, and it wasn't much fun.

But yes, in general it should be easy to change plug wires on most vehicles, but wires on DOHC motors that have seen a lot of miles and heat cycles can sometimes fail in ways that suck for the mechanic trying to replace them.
posted by mosk at 10:54 AM on December 18, 2014

I had a very similar problem with my 2013 Mini Cooper just last month: the culprit was a bad ignition coil. Hopefully the new errors help your mechanic diagnose and fix it quickly (and inexpensively!).
posted by culfinglin at 3:11 PM on December 18, 2014

Resolution! Codes showed TWO cylinders misfiring, #3 and #4. Coils come in packs; each coil pack serves two cylinders. Odds of two cylinders independently getting weird at the same time, near-zero, so coil pack serving 3 and 4 evidently the problem. Mechanic had seen a similar problem with the same brand of coil pack the month prior. Coil pack swapped out, Forester running normally so far.
posted by escabeche at 8:17 PM on December 18, 2014

« Older How can you tell if a job offer - is binding or...   |   Suggestions for Small Group Decision-Making/Voting... Newer »
This thread is closed to new comments.