Yes, this is an old post, but I can add a few comments of value. First, when there's no intake air temperature sensor, they use the "battery temperature sensor". It is indeed inside the Engine Computer where fresh intake air flows through.
If you need to replace the Engine Computer, forget trying to have one rebuilt. There's golly-zillions of them in the salvage yards, and they're all the same for a given engine size. Unlike today where software can be updated to solve running or emissions problems, back then it was done by replacing the computer. The old one got sent back to be updated at the factory, then they got a new part number on a new sticker. If they only needed some minor update, the seven-digit part number ended with an "AB" or "AC" instead of the original "AA". As long as the engine size and model year are the same, you can pop that used computer in and buzz off into the sunset. No silly programming required.
Second, forget about replacing injectors. When Chrysler buys their injectors, they get them from Bosch in flow-matched sets, similar to buying a pair of matched audio output transistors for high-end amplifiers. Injector failures in Chrysler products are unheard of. In ten years at the dealership, and 22 years after that, I only replaced one set of six under a recall program, but even those weren't causing a problem. One of my Caravans, an '88, has over 440,000 miles with the original injectors. My '94 has close to 250,000 miles, and original injectors. My 2014 model has 167,000 miles. Original injectors. You get the point.
In comparison, GM grabs a handful of injectors out of a huge bin, then throws them in the engine as it comes down the assembly line, with no regard to flow-matching. By around 100,000 miles, one or two flow a little less than the others and can cause an elusive lean misfire that you can't feel, but it will keep setting misfire fault codes. When Jim Linder was still around, he owned an injector rebuilding company by the Indianapolis Speedway. On his web site he listed the top ten models with injector problems. Eight were GM models and two were Asian imports. No Chryslers. No Fords. All the GMs were repaired by installing a set of rebuilt / flow-matched injectors. I don't remember what the problems were on the two imports.
I was interested in the comment about the timing chain, but that doesn't sound right for this engine. The Neons had a problem where the key between the sprocket and camshaft would shear off, the shaft would rotate slightly on the sprocket, then the computer would shut the engine down, same as if the timing belt had jumped two teeth. The timing marks look perfect, but the camshaft is late, and the camshaft position sensor is on the back end so its signals are late relative to the signals from the crankshaft position sensor. That was done because the 2.0L engine was an interference engine. It would shut down to protect the valves. That's not the case with the 3.3L. Unlike on newer models that may keep running when one of those sensors fails, the earlier models needed both signals to run.
Two other things come to mind for this problem. First, I would drive around with a fuel pressure gauge clipped under the right wiper arm so you can watch it. This happened to me four times on three vehicles, two with carburetors, so those don't count. On my '88 with a 3.0L engine, it stalled on me multiple times on the two hottest days of summer. It ran fine on the highway, but stalled as soon as I lifted the accelerator to coast. Took me four hours to nurse it through Minneapolis with all three interstate bypasses down to one lane for road construction. Once back on the highway, it ran fine for over two hours until I turned off near home. Ran fine for the next six months once the temperature dropped about ten degrees. And turns out temperature had nothing to do with this. It turned out to be a collapsing fuel strainer inside the gas tank. Twelve bucks for a new one and no more stalling for another ten years.
The second time this happened to that van, I was using it to drag a huge tandem-axle enclosed trailer to an old car show swap meet. I had the fuel pressure gauge tied to the radio antenna for over a year at that point, searching for the cause of this intermittent problem.
Keep in mind a lot of GM engines won't start and may not run if fuel pressure is just five to ten pounds lower than specs. My van called for 45 - 50 psi. As I would tug this trailer down the road, fuel pressure would slowly drop over about 15 to 20 seconds. The engine still ran fine down to 20 psi. It would start to sputter at 15 psi. I found that by lifting the accelerator for just a brief instant, the pressure would pop back up, then slowly drop again. By doing that, I was able to nurse it home 55 miles without any stalling or sputtering.
That reminds me, don't waste your money replacing the fuel filter on a Chrysler product. Unlike on most other brands, they last the life of the vehicle unless they rust out at the seam and start to leak. That sock in the tank is a different story. This problem never occurred on my '94 or '95 Caravans, and those socks are not available separately for those model years. You have to buy a new fuel pump with the housing to get a new sock. Regardless, monitor fuel pressure to see if that's where we need to look.
I've also never seen a defective fuel pressure regulator on a Chrysler product. That was a big problem on GMs, and to my knowledge, just on truck engines. They'd develop a leak in the diaphragm, then fuel would pour out of the vacuum hose and flood the engine.
Another way to look for fuel pressure problems is to unplug the vacuum hose at the regulator and plug that hose. That will make fuel pressure go to maximum. There's a little more to that story if it comes to that.
The second issue you should be aware of, although I don't think it applies here, is what happens to all Chrysler products after you disconnect the battery or let it run dead, or if you unplug the Engine Computer. As soon as you restart the engine, sensor personalities and fuel trim tables start to be rebuilt without you even noticing. The one big exception is "minimum throttle". That takes a specific set of conditions for the relearn to take place. Until that is done, the engine may not start unless you hold the accelerator down 1/4". Idle speed will be too low, you won't get the normal "idle flare-up" to 1500 rpm at start-up, and it will tend to stall when coming to a stop. For the relearn to take place, the Engine Computer has to know your foot is off the accelerator pedal, then it takes a reading from the throttle position sensor and puts that into memory. From then on, any time it sees that same voltage, roughly 0.5 volts, it knows it has to be in control of idle speed. To meet those conditions for the relearn to take place, drive at highway speed with the engine warmed up, then coast for at least seven seconds without touching the pedals. The relearn will be aborted if sees the TPS voltage change, or if you tap the brake pedal, (same as if the brake light switch is out of adjustment and you drive over a bumpy road). Hold the brake pedal up with your foot to solve that.
Also, if you have a scanner, under sensor data, look at the number of "steps" the computer has placed the idle speed motor at. For a properly-running engine, step 32 is typical out of a possible 256 steps. For comparison, with a single misfiring cylinder on a V-6 engine, you can expect to find it at around step 50 to keep idle speed up to the desired speed. If you find it is at step 0, minimum throttle hasn't been relearned yet.
I read a reference to adjusting the base idle speed at the throttle body. That is not correct, not necessary, and won't solve anything. Most throttle bodies don't have an adjusting screw. For those few that do, that was set at the factory and should have been sealed.
The Engine Computer has no way of monitoring the idle speed motor, often called the
"automatic idle speed (AIS) motor". It is a stepper motor. Those don't have brushes or a spinning armature like normal motors. They have four electromagnetic coils of wire that are pulsed with varying voltages and polarities to slowly rotate the armature. It's attached to a threaded shaft that extends or retracts a pintle valve. That valve opens a controlled air passage around the throttle blade. As the computer retracts that valve to let in more air, it lengthens the amount of time the injectors are pulsed open. Those two together control idle speed. Every time the engine is stopped, the computer runs that valve all the way out to close off the air passage, and a little more to be sure it's fully closed, then it opens it a certain amount to be ready for the next start, to provide that idle flare-up for a few seconds. After that the only thing the computer knows is to leave that valve alone, or to adjust its position to achieve the desired idle speed. If you could make that motor sluggish, the computer would just keep on pulsing it to a higher step number until it got to the speed it wants.
There was a problem in the late '80s with carbon build-up in that air bypass passage. It caused the same stalling and low idle speed. Once the AIS motor was removed, it was a quick, simple task to clean that passage with carburetor cleaner. I never saw that on any engine other than the Mitsubishi-built 3.0L., and we no longer saw it by the mid '90s, I assume due to better additives in the gas.
By the way, the AIS motor it totally out of the picture once you're off idle with the accelerator pedal, so don't look at the circuit if the problem occurs while you're driving. It only affects idle speed, and only once minimum throttle has been learned.
If you're using a scanner to view live data, watch if the throttle position sensor shows a nice smooth change from roughly 0.5 volts at idle to roughly 4.5 volts at wide-open-throttle, with no glitches or dropouts. I had a very unusual problem with that sensor on my '94 that I'm almost certain I'll never see again. For now, also watch that the voltage comes back smoothly to 0.5 volts after you open the throttle part way to different positions. It shouldn't stick at any voltage when you're moving the accelerator pedal.
Most scanners have a "record" feature where you press the "record" switch when the problem occurs. It takes a few second snapshot of what the sensors are reporting that you can play back slowly, later, to see what changed. Because that data flows through the scanner's memory, the recording actually begins a couple of seconds before you pressed the switch. I have a Chrysler DRB3 for all of my older vehicles, but I really haven't used that function very much. For my newer vehicles I have a Snapon Solus Edge. That can record four readings at once for a longer period of time. I've played with that but really haven't needed it. The goal here is to press the switch when the stalling or sputtering occurs, then see what changed. If everything looks normal except maybe the oxygen sensor went a little more lean, suspect a problem with low fuel pressure. That entire supply system is not monitored by the computer, so you won't see any reference to it in the sensor readings or fault codes.
If you have any temperature sensor that's reading 205 degrees, that sounds like a default value when it is maxed out. Typically that happens when the wire to that sensor is grounded, or in some cases, "open", meaning cut. All of the sensors are fed with 5.0 volts and ground, then the acceptable range for their signal voltages is 0.5 to 4.5 volts, give or take a little. Reaching 0.0 or 5.0 volts is only possible when there's a circuit defect. Those are the voltages that trigger a fault code. If there's no fault code related to a sensor, but it shows an unusual signal voltage on the scanner, the best suspect is corrosion between the terminals in the connector. The sensor itself is the last thing to suspect. Temperature sensors have just one component inside them, so failures are very rare.
For fuel metering calculations, the MAP sensor is the most important. The TPS is very unimportant. For modifying those fuel calculations, the coolant temperature sensor is the most important. The intake air or the battery temperature sensor just tweak those calculations a little.
We need to concern ourselves with that 205 degrees as that is not correct and it affects fuel delivery. Look at fuel pressure too so we don't overlook something. If this still doesn't help, we can go further, but you may want to consider starting a new question specific to your van. Unlike other forums where anyone can chime in to confuse the issue, here this became a private conversation between just two people. When you added to the thread, I was the only person to get an automated e-mail directing me back here. As such, none of the other experts saw your addition or had a chance to reply. That may not get you the help you need. If you need to start a new question, here's where to do it:
https://www.2carpros.com/questions/new
New questions also allow them to be categorized by model and problem for the benefit of others, such as yourself, who are researching a similar problem. Please be sure to list the engine size and mileage. We use that to find the right diagrams for making judgement calls as to best suspects.
Let me know what you find, even if you start a new question that I might not see.
Thursday, January 12th, 2023 AT 11:08 PM
