2004 Dodge Intrepid runs erratically

Help! We inherited a 2004 Dodge Intrepid SE 2.7L V6 with about 110K mileage. Basic problem: sometimes on startup the car runs terribly. It’s like a cartoon car–it wheezes, bucks, snorts, revs up, drops back down again, etc., and cannot be driven. You know as soon as you turn the key whether or not it will run correctly. Sometimes when you turn the key, it starts up fine and runs fine. We’ve taken it to two different repair shops–our local guy and the dealership-- several times, and $1,700 later, it still happens. These are the repairs they have done: replace the map sensor, replace the camshaft position sensor, replace the crankshaft position sensor, replace the powertrain control module, replace the throttle position censor and, finally, replace the pigtail for the map sensor. The codes usually indicate a problem with the map sensor. As a matter of fact, the last time we took it in while it was running terribly, our mechanic unplugged the map sensor while it was running and the engine ran smoothly while it was unplugged. This is what caused him to put in a new pigtail for the sensor. Anyway, after that last repair, a few uses later, the car acted up again. We’ve learned that if you let the car sit for a few minutes after a bad start up, it will often start up fine the next time you try. One note is that the problem has not occurred as often this summer as it did in the cooler months. My son uses this car to go to his two jobs and to his college classes and it’s very frustrating for him. Plus, my husband lost his job so we can’t afford to pour more money into the car!! When it starts fine, it runs fine. But it’s a crapshoot. Any thoughts??

PCV valve is cheap, as long as we are throwing parts at it. Spark plug wires are another thought if it is a beginning of the day problem.


Does that map sensor have a hose attached to it?

Do you have any idea if the mechanics have been using factory parts?

Or at least high quality aftermarket parts?

Has anyone made sure the engine coolant temperature sensor is working correctly?

I believe a crankshaft position sensor replacement warrants a relearn procedure.

Was this performed?

I don’t think there’s any relearn for a CKP sensor on this car. It either generates a pulse or it doesn’t, and there’s only one way to install it.

If you keep getting MAP sensor codes and you’re sure a good MAP sensor is being used, then the problem has to be some sort of major vacuum leak on the same ‘circuit’ that the MAP sensor is on, or bad wiring or PCM. There’s really nothing else. I don’t know where the MAP sensor is located on an Intrepid. On my 1994 Chrysler it is screwed directly into the intake. The nice thing about a MAP sensor system vs. the MAF sensor system that GM uses is that it’s mostly immune from all but the worst air leaks in the intake system. Has anyone checked the EGR valve for a bad leak? Another possibility is a leaking intake manifold gasket, though I doubt this is the problem if the car runs fine with the MAP sensor disconnected.

I’d try a different MAP sensor first, probably a factory part or premium aftermarket part as db4690 has said. If a cheap part is used, it’s easy to get a bad one and end up chasing phantoms.

Vacuum leak would sound good to me if it was not intermittent. Crankshaft position, air intake temp sensor feeds the MAP sensor along with the AC pressure transducer. Crank and air use the same ground. The AC uses the same 5vdc. It should be easy enough to see with a meter if you are loosing the ground or the 5vdc. You could unplug the AC pressure and see if that corrects the intermittent problem. The ground could be bad and that would cause the problems too.


On some GM vehicles, if you don’t perform the crank sensor relearn procedure, you may very well regret it.

I’ve personally seen my colleagues scratch their heads wondering why the vehicle runs like s . . . t
after replacing the sensor, without doing the relearn. After borrowing my scanner and doing the relearn, the engine was running perfectly again.

Don’t get me wrong, the engines started up without doing the relearn, but they ran terribly.

Even on some of the Benz vehicles, the engines did not run their best if the relearn wasn’t performed.

This kind of drivability problem may require a shop experience w/fixing Dodges and who has a Dodge scan tool. A dealership shop or a well equipped inde. The MAP sensor measures the vacuum in the intake manifold and is a vital component to the ECM’s calculation of the required amount of gas to inject to achieve the correct fuel/air mixture. If the MAP sensor isn’t working, it could produce all kinds of drivability problems. It may be for the OP’s case there’s a problem in the wiring between the MAP sensor and the ECM, or the ECM is bad, or the replacement MAP sensor is bad too. Especially consider that the replacement MAP is bad if it is an aftermarket version. A new OEM version from a dealer is probably best to use for something like this.

I’d be inclined at this point to trust the DTC code saying there is something wrong w/the MAP sensor, and focus on resolving that issue first, before moving on to other possibilities. It should be possible to test the MAP sensor independent of the car. I mean in principal you could remove it from the car and test it on the bench. On the bench, it’s just a barometer/altimeter, who’s resistance varies with the local barometric pressure and altitude. The shop manual probably has a method to do this test in a snap. Either it produces the correct resistance or not for the altitude and barometric pressure, and resistance is easy enough for a mechanic to measure. In practice this could be done with the MAP left on the car and the engine cool and off, like before the car is started in the morning. If it tests ok before starting the car in the AM, but the problem remains, the same test could be repeated at a time when the car is acting up. Just turn the engine off and test the MAP again. It should be close to the same reading as before, since the barometric pressure doesn’t change much day to day.

It could fail only when under vacuum too though. There may be a method to test a MAP sensor under vacuum, using a vacuum pump to apply a test vacuum. Anybody ever done that?