Is this car possessed?
February 11, 2015 9:03 PM   Subscribe

A friend of mine has a 2000 Chevy Impala with a mysterious issue that is baffling the mechanic and the dealer. Any gearheads have any ideas?

My friend's 2000 Chevy Impala started acting up a few weeks ago. He was on his way to work, and was only about 2 minutes away from home on the highway when the engine just died. He pulled over to the shoulder, and found that it would start, but wouldn't stay running for more than about a second.

Towed it to the mechanic, and of course they could not reproduce the problem - it started right up for them (after having to jump it, presumably because friend tried starting it too many times), and they claim it ran for several hours with no issues, albeit sitting in the lot. Friend picks it up, has no problems driving it around town running some errands (no highway speeds), and takes it home.

Next morning, same thing, nearly exact same spot, engine dies. Mechanic doesn't find anything wrong again and sends it to the dealer, who cannot reproduce the problem, but gives him this giant plastic cover thing for his ignition key that's supposed to help with the now infamous Chevy ignition issues; supposedly it causes the key to not be pulled down as hard or in the wrong direction by other stuff on the keychain or something. Mechanic puts in new battery and alternator, says it runs fine running around town.

After a third identical failure, mechanic now thinks it's ignition related. He puts in a new ignition switch / mechanism. He also finds that the car is wired for remote start, and disconnects that system completely since friend doesn't use it. He picks it up last night, I suggest he reproduce his commute to see if it fails again. We get past the previous location of failure, go past a mile or two, and come back with no issues.

Today, friend makes it to work fine (about 40 miles, all highway), thinking the issue is solved, only to have it crap out again halfway home. Thinking it was computer related since the car always seems to start up fine the following day after getting towed, he disconnected the battery for about 20 minutes, but that didn't help. Needless to say he's not happy. What other ideas might he suggest to his mechanic?

tl;dr: Car started to quit after 1-2 miles of highway driving. Will start, but dies right away. After letting it sit overnight, it always starts up again and is fine, driving around city streets is no issue. Only on the highway at 65mph+ does the problem manifest itself. Alternator, battery, and ignition have been replaced. What gnomes are infesting this car?
posted by SquidLips to Travel & Transportation (15 answers total) 1 user marked this as a favorite
When the engine dies, do all the lights on the dashboard go off, or is it just the motor and the lights remain on?

I don't have much to go on but it could be something in the fuel system that's clogging a line. Presumably after resting for a while it unclogs itself due to gravity or settling, and when he drives it for a while the fuel pump eventually finds the thing again and it gets clogged.
posted by hellojed at 9:44 PM on February 11, 2015

Some cursory googling points to maybe a failed mass air flow sensor
posted by hellojed at 9:47 PM on February 11, 2015 [2 favorites]

Maybe replace the fuel filter. It's a cheap thing to rule out anyway.
posted by Mei's lost sandal at 10:26 PM on February 11, 2015 [2 favorites]

Is your friend in Minnesota?

I'm given to understand that it's pretty cold there right now, and two minutes at highway speed with the consequent wind-chill factor might be enough to freeze any water that happens to be hanging around in the fuel lines.

Here is an interesting page from Car Talk discussing issues of fuel line icing and modern gas formulations. "Tester" seems most likely to know what he's talking about to me.
posted by jamjam at 10:28 PM on February 11, 2015 [1 favorite]

Going to go for the opposite of the icing up fuel line suggestion and say it could be flooding. especially since you imply with "(after having to jump it, presumably because friend tried starting it too many times)" that your friend was cranking it a lot after it died. You tow it, and it has enough time for the gas to evaporate off the plugs and... hallelujah! it starts again.

my janky testing method would be to get a clear fuel filter at your local auto zone/oreillys type shop, and put it in line in the engine compartment before the fuel rail and all that jazz. The INSTANT it dies, pop the hood and see if there's fuel in the filter through the little window.

If there IS, then it could be flooding. If there isn't, then you're having icing or some other supply problem. At that point i'd start to wonder if the pump wasn't starting to shit out, or was struggling because of not only an iced up line but a fucked up return or something causing it to not be able to push much since it had been doing it for a while and had started to die, or who knows what else. In stop and go/in town traffic you could have enough time and variation of fuel draw to let it either build up pressure, or relieve pressure before the system starts to shit up.

But really, if there's fuel in the filter, i'd start to wonder about some kind of flooding scenario.

One of the times i've had to troubleshoot this specific problem on a fuel injected car, it was an injection system fuckup. but any other time i've dealt with it there was either something plugging up the fuel line or it was flooding.
posted by emptythought at 12:16 AM on February 12, 2015

I had a couple of failed sensors in my car over Christmas, and one of them was the mass air flow sensor. The result of this was that, depending on some kind of lottery within the engine, it would either start and run perfectly, or it would take about 5 minutes to get started, and then only with an engine warning light, and the car would be rev-limited to 4,000 (i.e. 'limp home mode'). Slightly different situation (and a different car), but it may be worth getting the computer checked to see if it's logged any sensor issues.
posted by pipeski at 3:07 AM on February 12, 2015 [1 favorite]

Does any "security" warning light come on? Some GM products have issues with the Passlock immobilizer chip key system; I don't believe those issues were addressed by the recall.
posted by Seeking Direction at 7:15 AM on February 12, 2015

If your friend is willing to spend a few dollars you can get code readers that blue tooth to your phone for not much money. That would allow your friend to read the conditions when he has this problem hopefully catching the misbehaving component.
posted by Mitheral at 8:36 AM on February 12, 2015 [1 favorite]

I've experienced overheating during the winter when ice clogs the radiator during high-speed highway driving. We would often put a piece of cardboard in front of the radiator to keep it from freezing up.

Is your friend keeping the car in a heated garage overnight, or outside?

As others have mentioned, fuel supply (pump, filter, water in lines) is also a potential source of problems.
posted by clawsoon at 9:01 AM on February 12, 2015

A friend with a GM car had a similar problem, car would run and then mysteriously die. Usually it wouldn't immediately start. It was taken to the dealer multiple times and nothing could be found. They'd start it up and it would run for hours. Then he'd take it home and it would die again.

It wasn't the battery or alternator or anything in the fuel lines or fuel pump. Finally, it died a block uphill of a different dealer and he coasted in and had them look at it (instead of having it towed his original dealer). They got it started and had it hooked up to their diagnostic computer when it died again. The error was basically the throttle position sensor returning a value out of range which causes the engine to immediately shutdown. Not sure why, perhaps it's to prevent the vehicle from suddenly accelerating. For some reason, the code would clear itself so if you checked it afterwards it wouldn't be listed. This is why it hadn't been seen before. The replacement sensor wasn't cheap but it finally fixed the issue.

It's at least worth checking into.
posted by tommasz at 2:16 PM on February 12, 2015

two minutes at highway speed with the consequent wind-chill factor might be enough to freeze any water that happens to be hanging around in the fuel lines.

This is phenomenally unlikely. Red Herring. Fuel doesn't 'hang around in the lines', it is flowing through those lines at a decent flow rate/speed so there is no fuel sitting in the air flow to get wind chill. Besides, if there was enough water in the fuel to freeze and block the line, there is more than enough water in the fuel that the engine would stop when it reached the water. It won't freeze, either way. But water in the fuel is extremely unlikely unless there is an active hole into the tank to keep replenishing it. It would have been evaporated out by now (with the warm fuel returning to the tank from the engine). Also, this is likely to happen once, not randomly.

The error was basically the throttle position sensor returning a value out of range which causes the engine to immediately shutdown.

Thats' what I came in to say. Either fuel filter collapsing, but this would require a similar load case for each stall (if full or mostly full throttle and good engine load or high rpm). Basically conditions that require a sudden increase in fuel to the engine.

Question: Where it was in the commute may not be important because it may not be 'time from start' or miles related, but does something consistent(ish) usually happen at that spot? Maybe accelerating from stop start traffic on an incline? Can they record/pay attention and learn to replicate the exact conditions just before it stalls? If he is just at partial throttle on the highway then this is not an issue. But if it is always when accelerating from a stop sign, or when he lifts off for a red light after prolonged running, this is important info.

So if we can get more info on the fault case (ie consistent behaviour through each fail case) that will help.

My suspicion is that some sensor (air flow, crank, cam, throttle angle etc) that is relatively fundamental to the engine's ability self-regulate fuel etc is suddenly giving a wrong or no signal through a short. A throttle position sensor used to do this and drove me insane on my Jaguar. Randomly just... shut the engine into limp home mode. I suspect the Impala doesn't have a limp home mode so it is just shutting the engine down (or it is something other than the TPS that demands shutdown rather than limp home mode).

I have no idea why the mechanic changed the battery and alternator for this fault. That makes no sense at all. It didn't fail to start or fail to charge, it stopped while running. Neither part can cause that - the voltage regulator (part of the alternator) ... maybe, but that's easy to see in workshop tests. Ignition switch I also can't see unless there are no lights on the dash when the car has stopped - ie when the engine stops and he looks down at the dashboard - are any lights on? Which ones? If ignition and oil pressure lights are on, then it isn't an ignition *switch* issue but more likely to be ignition related.

Unless there is more information, the mechanic sounds clueless and shooting in the dark, to me.
posted by Brockles at 5:10 PM on February 12, 2015

To summarise:

1 - is something common happening when this failure occurs (sudden or strong acceleration, steady state speed or as he slows down)

2 - what lights are on the dash when the engine stops? Warning text? Anything unusual?

After letting it sit overnight, it always starts up again and is fine

This would be consistent with a a faulty sensor, too. Or a wiring short. Things move and flex when they get hot and only reset themselves physically if they cool down properly.
posted by Brockles at 5:13 PM on February 12, 2015

Thanks for the replies. In answer to the questions:

When the engine dies, do all the lights on the dashboard go off, or is it just the motor and the lights remain on?
Lights remain on. Just the engine dies. It will start again but only run for a second or two before stalling again, until it sits for a long time.

Does any "security" warning light come on?
Not to my knowledge. The mechanic did replace the ignition switch, possibly thinking this might be related, but that didn't solve the problem.

is something common happening when this failure occurs
The only consistent thing about the failure is that it happens (so far) only at highway speeds. The first three times it failed, it was only a mile from his home, but in a spot where the car would already be up to speed and would have been driving straight and level for about 1/2 mile. The last time it failed, it was after about 20 miles of solid freeway driving, on a pretty flat stretch of road. Friend says nothing unusual about what he was doing at the time.

At no point were any codes seen; mechanic probed the computer and found nothing. I would question anything fuel related simply because it doesn't idle or run rough at all, or have trouble starting after sitting for a period of time. He can run around town on city streets all day with no issues, it's only when he gets on the highway that it dies. But, a hot-rodder friend of a co-worker also said his first thought was a fuel issue when they talked today, so what do I know. I just don't grasp a fuel flow problem that would suddenly cut off all fuel only at highway speeds, remain 'bad', and mysteriously resolve itself after sitting overnight. Said person also did mention the Throttle Position Sensor as a remote possibility.

Don't believe the cold weather to be a factor, since the first few times it died it was relatively moderate by MN standards, even into the 30's. Last night it was colder, but still around 15 or so.

Friend will mention the TPS and fuel ideas to the mechanic. Thanks for the ideas! Feel free to suggest more.
posted by SquidLips at 7:39 PM on February 12, 2015

I just don't grasp a fuel flow problem that would suddenly cut off all fuel only at highway speeds

It's feasible, or we'd not mention it. It's best not to dismiss possible solutions from people that have seen similar issues unless you have actively eliminated them (rather than just finding them tough to believe) - namely other people having seen really weird things caused by old fuel filters that we had been told had been changed. I've had high speed misfires, sudden fail to start (it'd fire then immediately stop) and all kinds of weird crap just from the filter being blocked in some weird way.

Your misunderstanding is that you assume that fuel is 'cut off' to zero, rather than being 'restricted enough to be unable to maintain pressure in the rail sufficient to open the injectors'. This is effectively 'cut off' but not quite the on/off tap that you are envisioning. It's possible that a restriction in the fuel filter (or a weak fuel pump) is meaning a volume of fuel can be flowed that only just reaches the flow required to maintain the car at highway speeds/rpm. It may be that the car is running lean because flow is ever so slightly restricted. Maybe it is right on the knife edge of being able to supply fuel and over sustained periods of this fuel requirement the feed is slowly emptying the filter because it requires 1% more fuel than the filter can through-flow and so empties the filter slowly instead. When the filter 'reservoir' back up volume is depleted the flow is not enough and the engine stalls. As soon as you lift off for traffic or the rpm drops, the reservoir in the filter can replenish so it is only if the fuel flow requirement is sustained that supply fails. I've seen both pump and filter cause this before, but usually only resulting in misfire rather than stall, but these were usually racing ECU's that don't have 'save the engine' cut off circuitry.

While my personal feeling is this could well be a loose connection/broken wire in the TPS or air flow meter, I am suspicious enough of a restricted filter/weak pump. If he hasn't had the fuel filter replaced at all recently then it needs doing anyway and it'd be nice to discount the possibility. Flow testing the pump should be easy enough but if it's cheap, just change it.

Lights remain on. Just the engine dies. It will start again but only run for a second or two before stalling again

That makes it even more dumb that the guy changed the battery and alternator. Unless they were independently dodgy by sheer coincidence. It's baffling that was their first point of call to me. It would have been cheaper to change the filter and pump as a first port of call and far more logical.

The only consistent thing about the failure is that it happens (so far) only at highway speeds.
YEs, but what was he doing? Steady state throttle? His foot still for several seconds before it happened? Or he'd just lifted off for something and the engine wasn't 'there' when he went back to power?
posted by Brockles at 4:52 AM on February 13, 2015 [1 favorite]

The logical set of what-could-go-wrong, based on only-stalls-at highway-speed:
  • Something is failing when the engine gets hot at highway speed, and working again when the engine cools down.
  • Something is being provided in enough volume to run the car indefinitely at city speeds, but not enough volume to run it indefinitely at highway speeds.
Here are a few suggestions that fall into one of those two categories. From category 1:
We’ve seen certain crank sensors that measure within specifications when tested cold with an ohmmeter, but suddenly go wide open if the sensor gets hotter than 200 to 220° F. One minute it’s working fine producing a good signal, and the next it goes dead and causes the engine to stall. If an overheated crank sensor is causing a stalling problem, the engine may not restart until it sits awhile and the sensor cools down.
Intermittent faults in ignition modules can also cause an engine to suddenly stop running, and are often heat-related. When the module overheats, it quits working — but starts working again after it cools down (say 15 to 30 minutes later). When a bad module is responsible for an intermittent stalling problem, the engine will usually cold start just fine and run OK for short distances, but stalls after driving more than a few miles or when driving at highway speeds
If you suspect a bad [ignition] coil, measure the primary and secondary resistance with an ohmmeter — first with the coil at normal room temperature, then after heating it up with a hot air gun. Primary resistance is usually very low, maybe a couple of ohms or less, while secondary resistance is usually high, say 8,000 to 15,000 ohms or more. If you see a significant difference in the resistance readings hot versus cold, or a reading goes open, replace the coil.
From category 2:
Even if the [fuel] pump is generating normal pressure at idle, it may not deliver adequate flow at higher engine speeds. The engine may start and idle fine, but run out of power or even stall at highway speeds.
Like Brockles said, replacing the alternator and battery was a weird idea, and suggests your friend should be looking for a different mechanic. (Or maybe telling the current mechanic more detail, if he hasn't already: Does the mechanic know that the stalling only happens at highway speeds? That's the most important piece of diagnostic information we've got so far.)
posted by clawsoon at 6:15 AM on February 13, 2015

« Older Getting information into my head faster?   |   What to do with all my boxes? Newer »
This thread is closed to new comments.