No voltage to fuel pump and no reverse

You didn't damage the other fuel pumps. There isn't anything you can do to hurt them. There IS a cause for repeat failures, but those failures always occur weeks apart. What you have is some other intermittent problem that we haven't identified yet.

I got confused when trying to follow your relay testing. It sounds like you have an idea of what is supposed to happen, but so I don't steer you in the wrong direction, I'm going to stick with my standard troubleshooting procedure. Instead of testing each little part of each circuit, which is too time-consuming, it should make us some progress much quicker.

First of all, you mentioned the one-second burst from the fuel pump when you turn the ignition switch to "run". If you hear that, you're done with the fuel system. The pump is working. The automatic shutdown, (ASD) relay and fuel pump relay are working, and the Engine Computer has control over them.

Second, since the engine runs on injector cleaner, we know it has spark. We're done with that system. Next is the circuit that causes 99 percent of the problems, and that's the one in common with both the ignition and fuel supply systems, the crankshaft position sensor and the camshaft position sensor. The clinker is though, those both have to be working for the Engine Computer to turn on the ASD and fuel pump relays, and to have spark.

Based on that, we're back to the fuel pump which I only guessed was working. If you don't hear the hum for one second when you turn on the ignition switch, start with that weak battery you observed. I think at this point, instead of covering all the possibilities, do some basic testing first before I wear out my fingerprints on this keyboard.

Measure the battery voltage. It will be 12.6 volts if it's fully-charged. If it's near 12.2 volts, it's okay but discharged. Charge it at a slow rate for an hour, then measure it again. Remeasure the voltage while a helper is cranking the engine. It must not drop below 9.6 volts. A good battery won't have any trouble maintaining at least 11.0 volts.

Measure the voltage to the fuel pump. You will see 12 volts for one second after turning on the ignition switch, then it will go back to 0 volts. What is important is it must come back during cranking, but the value is important too. If the voltage is too low, the pump may not start up. You will still have pressure in the fuel line from that one-second priming burst, but the engine will only run on that for a few seconds. Sometimes the pressure bleeds down and one second isn't enough time to get it up for starting. If the voltage to the pump remains too low during cranking, even if the pump runs, it will be too slowly to build the needed pressure. The clue there is the pump would quickly start up AFTER you stop cranking the engine and it's running on the injector cleaner. You may want to borrow a fuel pressure gauge from an auto parts store that rents or borrows tools so we know exactly what is happening in the fuel system. I've had mine hanging from the radio antenna on my daily driver '88 Grand Caravan for almost two years. I was chasing an intermittent problem, and now that it's solved, I like having it to look at.

To go in a different direction, you can bypass the fuel pump relay or the ASD relay. Sorry I can't remember if your van has both. I have a '95 Grand Caravan that also hasn't been started in years, but I really don't care to tunnel through the snow to look at it. Some things you can try are first, to swap the fuel pump relay with another one like it. The starter relay is a good one to use. That will tell us the contacts are okay. Next, you can pop the fuel pump relay's cover off, reinstall it that way, the squeeze the contact to make the pump run. Watch that movable contact to be sure it's clicking when you turn on the ignition switch. Instead of removing the cover from the relay, you can also bypass it by jumping terminals 30 and 87 to each other in the socket with a piece of wire.

I started going all over the place again. Do some of those tests, then holler back with your observations and we'll continue.

One thing that I think is going to give more accuracy is if you can use a test light instead of a voltmeter. When you get back to the pump connector, I never paid any attention to the voltage, just whether it was there or not. If there is a high-resistance break in that wire, all it takes is a tiny spot of corrosion or a little carbon track to conduct enough current that the voltmeter will detect the voltage, but obviously that's not enough current to run the pump. A test light needs a lot more current to operate and will not light up if there's a break in that wire.

The alternative is to measure the voltage on that wire while it's still plugged in and the pump is drawing current, but that's not easy to do unless you can fashion a piece of wire for a test point.

The pump still should have run with an external battery.

Oops. Let me back up for a minute. When you turn on the ignition switch, you already heard those relays get powered for one second. That is done to run the fuel pump in case the pressure bled off overnight. I don't know about the 3.3L, but I do know from chasing an intermittent problem on my '88 with the 3.0L, that both engines run around 50 pounds of pressure, but my 3.0L still runs fine down to 20 pounds. (Some GM engines won't start if pressure is just two pounds low).

Once you start cranking the engine, pulses arrive from the cam and crank sensors. That is how the computer knows the engine is rotating, (cranking or running), and that is when it turns on those two relays. The fuel pump relay feeds only the fuel pump. The ASD relay feeds the ignition coil, or coil pack in your case, injectors, alternator field, oxygen sensor heaters, AND the fuel pump on older vehicles without a separate fuel pump relay. This is a safety circuit and it's real effective and reliable. If the vehicle is in a crash that ruptures a fuel line, raw fuel will get pumped onto the ground, and some people seem to think that's a fire hazard. The injectors can't spray fuel with no fuel pressure in that ruptured hose, so the engine stalls. When it stalls, there's no more pulses from the two sensors so the computer turns the ASD relay, (and the fuel pump relay) off. That stops the gas from being dumped on the ground.

I don't recall the computer monitoring the coil part of those two relays. It just grounds one terminal and assumes they turned on. There IS, however, another circuit coming off the ASD relay that, along with feeding the injectors and coil pack, goes back to a terminal at the computer. If no voltage shows up there, the computer can set a fault code related to the ASD relay not turning on, or that voltage is missing, or something like that. That circuit might supply other things inside the computer but it has nothing to do with the engine not running.

On the newer engines, as a point of interest, the engine will continue to run if one of the two sensors fails, but once it is stopped, it will crank but not restart. (There is a way to trick it into starting and running), but on this one the engine will stall if either signal is missing. The fact that the engine runs proves there's spark and the injectors are firing. That proves they're getting current from the ASD relay, so we know the computer is turning it on and the two sensors are working.

That brings us back to the fuel pump. GM pumps commonly fail while you're driving, leaving you sitting in the middle of nowhere on the side of the road. Chrysler pumps rarely if ever fail once they've started to run. They typically fail to start up. There's two causes for that. Most commonly, at high mileage, the brushes in the motor wear down and don't make good contact. The clues are that often a bang on the gas tank gets them going, you don't hear the hum for that one-second burst, and if you jump the relay or squeeze its contact, you won't see the normal tiny spark when you open the circuit. That's because there's no current flow, hence, no spark when you turn it off.

The second cause, which I eluded to earlier, has to do with repeat failures, usually from a week to a month apart. Many people buy a replacement pump from an auto parts store, and it works fine for a few weeks, then fails to start up. After four or five failures, out of frustration, they buy one from the dealer and don't have any more problems. In reality, the pumps weren't the problem. Unlike GM truck fuel pumps that you can hear a mile away, Chrysler pumps are built to very close tolerances to make them really quiet. Microscopic debris collects in the tank, (I suspect mold growth from using ethanol), and that debris clogs the impeller which locks up the motor. Each replacement pump collects a little more of that debris until finally it's all gone and the next pump won't fail. That one just happens to be from the dealer. As another point of interest, NAPA buys their fuel pumps for Chrysler products from the same supplier that makes them for Chrysler, so you're getting the same pump as the dealer's pump, and they're just as quiet. That may be true of other auto parts stores too but I only know that for sure about NAPA. The clue to this problem is you WILL see a spark when you jump the fuel pump relay but it won't blow a fuse. There's obviously current flow but the motor doesn't run. The fix for this was to remove the tank and have it steam-cleaned at a radiator repair shop. I haven't heard of this happening lately, possibly due to better additives in the gas today.

Getting back to your pump, jump the relay so you don't have to turn on the ignition switch or crank the engine to do the troubleshooting. Check for voltage at the pump on the dark green / black wire. If it's missing, there's two connectors between there and the relay that are good suspects. If you DO have voltage there, check on the black ground wire. If you have any voltage on that one, there's a break in it. You may be able to get a clue from watching the fuel gauge on the dash. The sending unit shares the pump's ground wire. If the gauge reads correctly with the ignition switch on, but changes during cranking, it has to be due to voltage on the ground wire.

42 - Auto shutdown (ASD) relay voltage not sensed at controller.

This is right out of the '94 service manual. The generic description from this site covers multiple years. That's why it includes the fuel pump relay with the ASD relay in the description.

If the engine runs on starting fluid, the ASD relay is working. That only leaves the wire that splices off and goes to the Engine Computer so it can verify that relay turned on. Put your test light on either smaller terminal on the back of the alternator, or the dark green / orange wire at the coil pack or any injector. You can back-probe those terminals through the rubber seals. You should see 12 volts there for one second after turning on the ignition switch, then again during engine cranking. If it's missing, try a different relay. If it's only missing during cranking, we have to look at the cam and crank sensors.

If you have voltage there, it may be not reaching the computer. That comes in on pin 57, also a dark green / orange wire. The connector is a 60-pin low-insertion plug meaning it takes no pressure to install it. You remove it by unscrewing the bolt in the middle, and you install it by just holding it in place, and running that bolt down with a small hand ratchet. Pin 57 is in the bottom row of pins, fourth one from the firewall end. You don't have to remove the plug. There's a plastic cover that can be snapped off to get to the terminals.

First of all, I'm not an expert on the electronic part of the transmission, so I'll just think on that for now. Second, I get the feeling you may not have a clear understanding of electrical theory. If I'm wrong, forgive me, but to be safe, allow me to clarify a few points that ARE confusing to most people. Voltage is electrical pressure and it can be compared to water or air pressure in a pipe. Current is electrical flow, just like, ... Uhm, ... Flow in a water pipe or compressed air line. Direction of that flow depends on where the higher and lower pressures are. "Power" is a legitimate term, and is simply the volts times the amps in a circuit, but it is always used incorrectly in conversations. And for the biggee, there is no "juice" in electrical circuits. I always told my students they could use that term on only the first day of class until they knew better, then we would stick with correct terminology. No matter how hard we squeeze a wire, there will be no juice coming out.

You have a much better grasp of electrical theory than most people. The only thing that made me want to mention these things for the benefit of other people researching this problem and reading this was your comment about:

"I was able to get voltage at the injector, but not at the computer or its connector (not knowing which way the power flowed, I tested both of them).

Normally I would say you just need to check for voltage, and if it's there, the current flow will take care of itself, but there are too many things here pointing to a POOR connection which acts totally differently from a broken connection. That's where people with decent knowledge of electrical theory get stuck. That's also why I ask a lot of questions that would seem to have obvious answers. Besides just analyzing the results of tests I tell you to take, and arriving at a fix, my mind works best when I explain how the circuit works along the way so next time you have a better chance of finding the problem yourself.

Getting back to the important issue, you found no voltage at pin 57 of the computer. I know from working with these connectors that 99 percent of the time that is because the meter or test light probe did not make good contact with the terminal. The first thing an inexperienced mechanic or do-it-yourselfer might do is bare the wire near the connector to take a reading. That is another thing I never allowed my kids to do because we don't damage a customer's car that way, and most of the time moisture will get in there and you'll have a corroded connection in a few weeks. In this case, the best test is to remove the entire plug from the computer, the probe on the exposed terminal end to measure the resistance between that terminal and that same color wire at one of the injectors or the coil pack. Your meter leads will have a few ohms of resistance so overall, you can expect to find, ... Oh, ... Perhaps two to five ohms total. If you find considerably more than that, I'd start to be concerned there's a corroded splice. I can look in the service manual for the location of that splice.

You said that code isn't setting anymore so we might be spending too much time on something we don't need to, but if you suspect there is an intermittent problem here, a better way to test that wire is with it trying to conduct some current, as in running the test light. To do that you'll have to bypass the ASD relay since the computer won't be there to turn it on, then use the grounded test light to test for voltage at pin 57 in the connector. If the test light is dim or it flickers, wiggle around on the harness to see if anything makes it change. Your test light will draw a lot more current than the computer will, so if that wire can run the test light at full brightness, it will be working properly for the computer.

As for the wires at the fuel pump connector, current has to go through the fuel gauge circuitry in the gauge cluster, then to the sending unit in the tank. That circuitry has some resistance, just like standing on a garden hose, so the voltage at the tank will be lower. Just like that partial restriction in the hose though, you WILL have full voltage at the tank when no current is trying to flow, as in when using a voltmeter. If your nozzle on the hose is turned off, you'll have full pressure there even when you're standing on the hose. The test light allows current to flow which causes some of the 12 volts to be "dropped" across the gauge circuitry. The remaining voltage is dropped across the test light, and since that is only part of the 12 volts, the light is dim.

Once a person understands this, it's easier to see that a voltmeter measures the voltage by measuring electrical pressure, and the test light measures voltage by measuring current flow. It just takes a tiny spot of carbon or corrosion to pass enough voltage to be picked up by a voltmeter. It takes a good, solid connection to pass enough current to run a test light.

See how far you get now. My computer just started acting up again with a problem I've been fighting for two years. I have to shut it down and start over. I'll be back in about 15 minutes.

Dandy. I'm back now for another 15 minutes to 15 hours. Never know when my 'puter will give intermittent "not responding" messages again.

On a related note, there were some problems with the wiring harness right underneath the Engine Computer, or right next to it, on the left inner fender. That harness slides back and forth as the engine rocks during acceleration. A coworker, who was the heating and AC specialist at the dealership where I used to work, had a '94 or '95 Caravan come in with overheating issues and a burned open fuse link wire to the radiator fan motor. Obviously the motor had to be shorted so he replaced both parts and sent the owner on his way. It came back three or four times, always with the same fuse link wire burned open, and on one visit, as it was our custom to watch other mechanics' projects for burned-out lights, I pointed out he had no backup lights too. After repairing the fuse wire again, he DID have backup lights, but as soon as he accelerated to leave the shop, the lights went out, and that's when the diagnosis took a whole new turn. The fuse link wire was burned open again before the radiator fan even had time to get turned on.

My trick is to replace a blown fuse with a 12 volt light bulb, and when we put the van in "park" with the engine not running but the ignition switch on, the light bulb would get bright, indicating lots of current was trying to flow, when we rocked the van and made the engine rock and tug on that harness. Finally found a bunch of those wires laying on the inner fender had rubbed through their insulation, and the paint was rubbed off, and there was the intermittent short.

Those fuse link wires were used because they will not burn open from a nuisance momentary current spike which is typical of a motor starting up. While that "slow-blow" feature is a benefit, a carbon track is left behind where the wire burns apart, and it's that carbon track that lets just enough current trickle through that a digital voltmeter can detect something. Enough current to run a test light can not get through. That's why test lights can be more accurate in those situations. A similar thing happened to another very experienced mechanic who happened to be our transmission specialist, when he also had a tight radiator fan motor on a K-Car. He found voltage to the motor with a voltmeter, then couldn't understand why the new motor wouldn't run when it was connected to that wire. Being the smart fellow I thought I was, and the new guy, I convinced him to use a test light instead of the voltmeter, and then he did not have any voltage there. Just needed a new fuse link wire.

Anyway, now that I'm done tooting my own horn, I'll wait for your next set of test results and we'll continue on.

I have the wiring diagram in the Chrysler service manual but it covers over 200 pages. Your best best to find one is on eBay. The '93 and '95 diagrams will be real close too. You might also check at a nearby community college that has an Automotive program. We had AllData on four computers in the shop and one in the library that was available to taxpayers. You can get a subscription to AllData or Mitchell On Demand on the internet too, but I hate diagrams on computer, and I refuse to rent the service. I only buy paper service manuals to keep that I can put notes in and find things quickly.

If you've never used a Chrysler wiring diagram before, be aware you'll be bouncing around a lot of different pages based on the "sheet" numbers on the bottom, not the page number in the book.

Fuse link wires are smaller in diameter than the wires they protect. They're about 8" long and will be in a bundle going around the left strut tower. You tug on them to test them. If they're good, they'll act like a wire. If they're burned open, they'll act like a rubber band. That break is where a carbon track will form. You can find the two ends of a link that's burned open and splice them together, then seal it with heat-shrink tubing with hot-melt glue inside. That is what was used originally. You can also buy new fuse links from any auto parts store. One piece is long enough to do two or three repairs. You buy new fuse link wires by gauge and color. Don't use regular wire for the replacement. The insulation on fuse link wires is designed to not burn or melt.

Don't use electrical tape either anywhere on a car. It will unravel into a gooey mess on a hot day.

All you're looking for at pin 57 is if you have something or nothing. I suspect the code related to that pin was set earlier when you had the ASD relay out.