QuestionQUESTION: I have a 1994 Dodge Grand Caravan ES with a 3.8L engine, and it starts but will not run. The Caravan ran well for 100 mile trip, and then I went to work and parked it, came back eight hours later and tried to start the van, and it would turn over, but did not idle without constantly goosing the gas pedal up and down, the moment the gas pedal was released the engine would die. The Caravan went from running perfectly fine to not running at all on the same day, which leads me to suspect an electrical component has failed. I checked for codes with the on /off (three times) and all I get is a 55. I have an Auto Xray scanner and it shows no stored codes. I have unplugged the battery to reset the PCM and this did not fix it. I checked all relevant fuses ( 16 & 20 under dash, and all under hood) I have checked the coil pack (proper primary and secondary resistance), I have strong spark on all plugs (pulled plug and viewed spark), I have fuel (pulled the Schrader valve on the fuel rail and pumped fuel into a jar), I examined fuel for water or contaminants (none found), I bought a fuel pressure gage and checked with KOEO and get 46psi ( Service manual says 48psi spec, suspect gage a little off), fuel pressure with engine on stays at 46psi ( with vacuum to regulator removed) I've checked for any restrictions in the exhaust pipe (none found, even disconnected exhaust pipe before catalytic converter to see if that may have collapsed and restricted air follow - did not fix), checked the air cleaner, and also tried to start the car with the air cleaner tube off (still will not run). When I say that it will start but will not run, it's as if the timing is off or the engine is starving for fuel or air, it will even begin to backfire at the intake manifold, but it will not stay running, and is not anywhere near close to even idling. I spray starter fluid in air intake and start the engine and it does run smoothly very briefly (maybe 1-2 seconds) before it acts starved again. This should indicate good air flow and spark. It raises the question of why the fuel flow is not consistent. I checked the relays by 12V energizing them and checking continuity (ASD and fuel pump and they work properly, even rotated other relays to be sure). I even jumpered the ASD and fuel pump relay to put in energized state - no help. I suspect the PCM, which I have removed and looked at the connectors for corrosion or dirt (I cleaned them anyway). I disconnect sensors (AIS & evap canister) and the scanner picks up the new codes and I can clear them, so that part of the PCM seems to work. I do not know if the camshaft position sensor or the crankshaft position sensor could be a possibility. But I do not understand if the PCM checks those in the first five seconds of cranking. Is the PCM not energizing the fuel injectors? I am baffled what to do next, other than to go buy a PCM and put it in. What do you think?
ANSWER: Hi Greg,
There is one part that I believe you haven't looked at that will possibly solve your problem: the exhaust gas recirculation valve (egr). Look for it in a small diameter pipe that runs from the side of the rear exhaust manifold back to the 'under throttle body' area. It has a valve body bolted into that pipe, a round top, and a flange area in between those two parts. The valve passes exhaust gas up to the intake manifold, only it is supposed to be closed when you idle/start the car. Otherwise if it is ajar it will lean out the mixture and prevent idling. The valve stem is inside the flange area and it has a circumferential slot into which you can insert the tip of a screwdriver and then move the stem back and forth against spring-action in one direction that tries to close it. The action gets gummy over time and the valve than won't seat closed. Spray some WD-40 on the stem where it enter the body and exercise the stem some more to make sure that it seats well to a full stop. Then try to start the engine.
Thanks for the very complete and educational report, it helped me to feel this egr may be the answer.
Roland
---------- FOLLOW-UP ----------
QUESTION: I pulled out the EGR valve and cleaned it. It appeared to be in good shape as I had replaced the original last year. I had replaced it last year due to poor idling especially when the AC was on (idling at 500rpm and then stalling) and replacing it fixed that problem. I put a vacuum gage on the vacuum lines of the EGR and it opened and closed properly. I put it back in the engine and started the car and no improvement. Exactly the same problem: starved for fuel and backfiring and then dying. What should I try next?
ANSWER: Hi Greg,
Another possibility that I experienced (no start/no codes) is a MAP sensor which was off-value but not enough to set a code. See if it reads about 4.7 volts when the engine is off (betweem dark green/red and black/light blue). Also check the resistance of the coolant temp sensor (should be 7,000 to 14,000 ohm when cold). Inaccuracy in either of these sensors could cause a too lean mixture.
Thanks for the continuing technical info. Please keep me informed.
Roland
---------- FOLLOW-UP ----------
QUESTION: I checked the map sensor, engine cold, and the reading was 5.4 volts. My Autoxray scanner displayed 4.9 volts. Is this close enough?
The coolant temperature sensor resistance was 10,400 ohms. My Autoxray scanner displayed coolant temperature 75.2°F, coolant sensor 2.6 volts. Other scanner readings;
map sensor = 31.1 inches of mercury
barometric pressure 31.1 =inches of mercury
I did notice that when I disconnected the map sensor it threw a code 42: "An open or shorted condition detected in the auto shutdown relay circuit." Would this be typical?
Are there other sensors that have to be "correct" before the PCM controls the injectors to fire? If not, am I back to the PCM may be bad?
AnswerHi Greg,
The coolant sensor sounds OK. The MAP sensor could be inaccurate if the barometric pressure locally is not in the vicinity of 31.1, but I am uncertain as to how far the inaccuracy has to be to cause a mixture that will not allow for idle. My experience was in the other direction: the sensor thought we were at 10,000' elevation instead of the sea level elevation we were actually at. That resulted in a too lean mixture. 31.1 sounds high to me but I could be wrong, depending upon your local meterology. If so, you would have too rich a mixture. See if you can call a local weather bureau or check the Weather Channel to see if you can get a current reading in your vicinity and then see what the MAP is showing with the scanner.
The 42 code doesn't seem right but I can't explain why disconnecting the MAP would cause it.
Have you cleaned the throttle body plate, throat, and idle passageway?
About the only other mixture sensor is the TPS which you could check for any discontinuities in the resistance as you measure between the center and either side-pin when changing the throttle.
I would still believe you would get a faulty pcm code if that were the case.
How about pulling the spark plugs after trying unsuccessfully to start to compare the "wetness" of the plugs to one another?
Please let me know... You are sure the egr valve stem is seated in the fully closed position?
Roland
