Blogs Car Info Our Show Deals Mechanics Files Vehicle Donation

Need troubleshooting advice for a 95 Caravan that won't start

It is possible your engine has flooded, Hold the gas pedal to the floor while cranking see if it starts, or pull a plug and see if it smells like gas.

If the engine cranks robustly, the battery is ok. For cranks but won’t start the quickest way to a solution is to first determine whether the problem is no spark, or no fuel. It’s almost always one or the other.

Do the obvious stuff first of course. Make sure there’s enough gas in the tank. Nothing leaking under the vehicle. Hoses and connections appear to be the same as they always were, etc etc. Next, decide which you want to test for first. Usually spark is the easiest. Connect up a spare spark plug, figure out a way to ground it to an engine ground away from anything that might explode or catch fire, and watch for a blueish-white spark to appear at the working end while a helper cranks the engine. Report back on the results.

1 Like

Update: I tried starting it again today when I got home from work. Today it was warmer outside (71° right now, versus 48° on Sunday). It started on the first try, but misfired horribly for several minutes, then suddenly smoothed out.

I let it idle in my driveway as the engine warmed up. When the temperature gauge reached approximately 33%, the CHECK ENGINE light came on. The tachometer showed approximately 800 RPM, the oil pressure gauge showed approximately 60%, the battery gauge showed approximately 13 volts. I turned it off and did the “key dance” to retrieve the codes, and repeated this several times to make sure I got the codes correctly. The result is 12, 52, 55.

Based on my Internet research, code 52 is an internal fault in the PCM, is that correct? Time to look for a used PCM? 4-cylinder Caravans from this generation are rare, even in junkyards.

Here is one website that shows them “sold out”.

Fault 52 is Oxygen sensor stays above center (rich). Probably caused by the misfire. If it is still misfiring clean the spark plugs and erase the fault.

It’s interesting that you mention that, because back in August I had a problem where I drove this van to work, and then could not get it started again if my life depended on it. That’s actually what made me buy another car.

Every day for a week, when I got off of work, I tried to start the van, but it never did.
Then finally one day it did start, and was misfiring a lot, but once it warmed up it ran ok. The next day, I replaced the spark plugs, which were all blackened even though they were not old, and then it ran ok until the battery went dead a few weeks ago.

If the spark plugs are carbon fouled again the oxygen sensor may be the problem.

Any way to test the oxygen sensor, if I remove it, or just buy another one? I already replaced it once, because the engine was burning coolant (cracked head, which I replaced) and I bought the Denso First Time Fit brand.

The start-up misfiring symptom could be the engine was flooded. After it runs for a few minutes it clears the flooded condition and then runs smoothly. A leaky fuel injector could cause that. The fluid rail is supposed to remain pressurized when you turn off the engine, and the injectors are supposed to not leak when the engine is off. Also double check the fuel pressure regulator is holding vacuum also. If its diaphragm springs a leak, that can result in a flooded engine. If this happens again remove a spark plug and see if the electrodes are wet w/fuel. If so remove all the spark plugs and let the cylinders air out for a couple days.

How do I do that? Remove the vacuum hose while the engine is running? I replaced the fuel injector with new (Standard Motor Products brand) approximately 2 years ago. The fuel pump was also replaced at that time (Denso First Time Fit brand). The fuel pressure regulator has not been changed.

If it’s the type that has a vacuum hose connected to it, the way I do it, remember I’m no expert, but with the engine off I first check to see if there’s any gasoline in that vacuum hose. There shouldn’t be. Then I connect up my hand-held vacuum checker gadget (mighty vac, something like that) and verify the fpr’s vacuum holds steady to 20 inches. Everything done w/engine off.

I had a professional mechanic look at the Caravan. They had it for a month, and quickly discovered that the problem is that something causes the computer to supply way more fuel than necessary at startup, and while the engine is warming up. Also, at idle it has difficulty providing the correct amount of fuel, which causes a slight misfire and “pulsating” RPMs. However, they could not figure out what the root cause is.

They tested fuel pressure, inspected the throttle body assembly and fuel injector, and observed the diagnostic data on their scan tool. They also checked to make sure nothing is wrong with the engine itself (poor compression, etc). It should be noted that I already replaced the fuel injector twice (first with a GB Remanufacturing one, then with a new one from Standard Motor Parts) and that made no difference. The fuel pump and fuel pressure regulator were also replaced with quality brand-name parts.

The suspicion at this point is a defective PCM. They told me that only a dealer with the factory scan tool can definitively prove that the PCM is bad. Is there anything else to try besides a used PCM?

When I’ve had problems with my Chrysler minivans, I have found help at They have a minivan forum. Good luck.

It could be the PCM. But it could be a case of GIGO too. Garbage in, garbage out; i.e. if the computer’s inputs are incorrect it will produce an incorrect output. So first make a list of all the sensors used for the algorithm that the PCM uses to determine the amount of fuel to inject.

  • Engine coolant temperature
  • Ambient air temperature
  • Start mode vs running mode

Of those I’d suspect the coolant temperature and the start/running mode as the most likely causes of incorrect computer inputs. It could be something very simple like a dirty electrical connector or faulty ignition switch. Ask your shop to measure the inputs from those sensors right at the computer’s connector.

When I had trouble with an 87 Reliant 2.5, a mechanic diagnosed it as a faulty computer. I was skeptical but I owned an 87 Caravelle 2.2. I checked and the computers had the same numbers on them. I swapped and the problem did not move with the computer. Turned out the problem was carb icing.

My point is, I would check any Mopars with the 2.2 or 2.5 engine not just minivans for identical computer numbers.

When you mentioned there is a temperature difference and the mechanic said something is causing the pcm to provide too much fuel, I wonder if you have a defective engine temperature sender. That’s what tells the computer what temp the engine is so that the correct fuel mixture can be provide. The mechanic should have been able to read what temp was being sensed but the sensor when he did the computer diagnostics.

The very first thing I would try is the ASD relay. (Auto Shutdown Relay)… its in the fuse paneI. It would cause the symptoms you have…and it would happen all of a sudden. Look into it on the net… … Im sure you will see plenty about it. Your comment on the tach needle not moving at all made that pop into my head…

A failing automatic shut down relay will cause the spark plugs to become carbon fouled? I think the oxygen sensor has failed.

No @Nevada_545 they became fouled during the drive time he had on the plugs just prior to the batt dying…and then this new issue becoming the main problem… I think you might be under the impression that these new plugs had zero run time?

The Op stated. The next day, I replaced the spark plugs, which were all blackened even though they were not old, and then it ran ok until the battery went dead a few weeks ago.

I would be all over that ASD… its as simple as checking a fuse, well sort of, just be sure it functions…why not?

P.S… You can also be correct as well in this scenario… I just wanted to mention the ASD…as a failed one would kill ignition suddenly…etc

1 Like

That’s simply not true. However, since the 1996 model year, automotive interface has been standardized by OBD2 requirements. That means:

  1. Things that set fault codes on a modern vehicle may not set codes in yours.
  2. Scan tool data that is available on a modern vehicle may not be available in yours.
  3. It may be difficult to find a shop that has maintained a now-obsolete scan tool to access and diagnose your ECM with live data and bi-directional controls. The shop would also need a tech with enough gray hair to know what he is looking at.

With the proper equipment and experience this should be easy to figure out. If the problem is in fact that the engine is overfueling at start up, prove it. Scan tool data will show injector pulse width in milliseconds, a labscope trace at the injector can be recorded and measured to verify. If the injector pw is indeed high, then a step by step test of all sensor inputs and outputs needs to be done. For instance, does the ECM think the temperature is -40* when the problem is happening? Is the MAP sensor reading from the scan tool the same as the reading at the sensor?

But since none of this equipment seems to be available, I think you are at the mercy of replacing parts.

What is the reason for suspecting over fueling? How wet are your plugs? A sleeping ignition coil will cause a fuel smell…and make the plugs wet.

I keep thinking of the mention of the tach not moving…when normally it did move even during crank. That makes me suspect the coils arent firing at all.

Has the OP checked the ASD relay… swap it with an identical one…just for giggles.

I wouldn’t immediately jump to the most severe condition possible here…PCM failure… imho those are rather rare, possible but rare and shouldn’t be the square to jump to so easily. Just saying.