QuestionWe have a 1997 Voyager 3.3 LE that occasionally starts but cuts out after 4-5 seconds. This has happened about 6-7 times this year but with no apparent pattern or reason that we can detect. If the Voyager is left for a period of time (maybe hours or the next day) it will start as if nothing was ever wrong and run smoothly. Otherwise the Voyager starts and runs perfectly without any engine problems at all.
We have noticed that when the Voyager has this problem the fuel gauge does not move when the ignition is swithed on but when the fault clears the gauge works normally.
Chrysler mechanics with all their diagnostic equipment are unable to find out what the problem is let alone rectify it. Can you help?
AnswerHi Colin,
That is the sort of gremlin that you probably have to catch just at the time it is present. If the engine is working normally at the time the shop looked at it they may not be able to detect the reason for the failure. Of course getting it to fail at the shop is nearly impossible. People often have their vehicles towed to the shop and find them to start right up!
The observation about the fuel gauge may be relevant and so I looked at my Haynes manual that covers the '97 van, though it doesn't have complete wiring diagrams. The interesting aspect is that all the signals for the instrument cluster are digitized and arrive there on a common data bus that connects together a plethora of engine operational sensors with the body computer and the collosion detection system, etc. But if the gas gauge was the only non-working unit on the cluster when this happens it would imply that it is not the bus that was involved. Do you recall whether the odometer, the transmission position indicator, and maybe any of the warning lights which you would have expected to illuminate were also impacted during the gremlin? The other possibility is the voltage supply to the cluster was impacted and if that supply was also involved with the engine operation that would give us a clue as to which power circuit being absent would shut down both the cluster and the engine. Then when it happens again you could check out the bus and the power circuit to see which one was bad and where.
The usual reason for not starting is the failure of either the crank or the cam sensor to put out its signal, or the autoshutdown relay fails to sustain power to the spark coil and the fuel pump. That relay also depends upon the sensors' signals and the engine controller which tells it to stay activated as long as it 'sees' pulses from the sensors. So one thing to do when this happens next is to see if the relay is activated or not, and whether there is spark being generated by the spark plugs or not, when you crank the engine and it won't start. Also relevant would be whether you hear the fuel pump run for 1-2 seconds when you first turn the ignition switch to run (before trying the starter position). The 4-5 second run and then shut down phenomenon is not the right time frame for a failure via the engine controller (it is programmed to shut off the spark and fuel at 1-2 seconds if it doesn't 'see' the sensor outputs). So that would make me wonder if the relay is weak. On the other hand, the sensors can fail in this manner, and then recover hours later and appear normal. Ideally there would be a trouble code recorded in the engine controller when this happens. Did the shop say they found nothing in the memory?
So I would suggest that you familiarize yourself with the wires that might be specifically in question and see how they behave voltage-wise when things are running normally, and then the next time it fails be ready to test them again so as to identify what isn't right and from that infer what component is bad. Do you have the Haynes manual? And do you have a voltmeter?
We can discuss this some more if you like.
Roland
