Engine will fire but not run.

Tiny
VIKES0419
  • MEMBER
  • 1994 PLYMOUTH VOYAGER
  • 165,000 MILES
I drove home from a long trip and parked. Van was running just fine. Went to start up the next day and would fire but not stay running. Since then, I have replaced the pump, filter, fuel pressure regulator, and checked all of the fuses and still the same problem. I had air come out a handful off times at the pressure test port when I depressed the pin inside it until it was purely fuel and now it will crank for a while then fire, then die right after firing. I have no clue what to do next but maybe the PCM, or crank and cam sensors. Help!
Tuesday, October 1st, 2013 AT 7:10 PM

20 Replies

Tiny
MADMIKE1735
  • MECHANIC
  • 951 POSTS
Really dumb question. Have you verified fuel level? The gauge may say 1/2 tank or whatever, but are you 100% certain there is fuel? I only ask because you had air comming from the test port. There should be NO air. Once you know you have fuel, then its possible its a spark issue. Have you verified spark yet?
Was this
answer
helpful?
Yes
No
Tuesday, October 1st, 2013 AT 7:21 PM
Tiny
VIKES0419
  • MEMBER
  • 13 POSTS
Definitely 1/4 tank, as I fought with thatch fuel in it as I reinstalled the tank. I actually haven't verified yet. I will tonight.
Was this
answer
helpful?
Yes
No
Tuesday, October 1st, 2013 AT 7:45 PM
Tiny
VIKES0419
  • MEMBER
  • 13 POSTS
That much* not sure how that happened.
Was this
answer
helpful?
Yes
No
Tuesday, October 1st, 2013 AT 7:46 PM
Tiny
CARADIODOC
  • MECHANIC
  • 33,833 POSTS
You're jumping all over the place without actually diagnosing anything. First of all, except for diesel engines, you will never solve a running problem on a Chrysler product by replacing the fuel filter. Next, listen for the one-second hum of the fuel pump when you turn on the ignition switch. If you hear that, the pump is working. The fuel pressure regulator is another waste of money. I've only heard of one that was leaking on a Chrysler product in the last 23 years.

The place to start is by reading the diagnostic fault codes. Chrysler makes that real easy. Cycle the ignition switch three times from "off" to " run" within five seconds without cranking the engine, leave it in "run", then count the flashes of the Check Engine light. There will be one or more flashes representing the first digit, a short pause, then a series of flashes for the second digit. The next code will flash the same way after a longer pause. The last code will be "55" which just means the "end of the message".

Your description of the symptom matches exactly how MAP sensors used to fail. They have the biggest say in how much fuel goes into the engine. After having the vehicles towed to the shop, many customers were amazed when we were able to start the engines and drive the vehicles into the shop. The secret is the engine will stay running as long as the accelerator pedal is moving. As soon as you hold it steady, the engine will stall.

The fact the engine ran at all, even for a few seconds, tells us you have spark. That means the crankshaft position sensor and camshaft position sensor are okay. If you find that you do NOT have spark, that will lead us down a different path.

Also, I've been reading lately about a lot of people checking fuses by pulling them out and looking at them. That's the time-wasting way to do it, but if you did, one of them would have erased the Engine Computer's memory. That is going to introduce a whole new running problem with a real easy fix. The idle speed will be too low and the engine may not start or run unless you hold the accelerator pedal down 1/4". You also will not get the normal idle flare-up to 1500 rpm when you start the engine, and the engine will tend to stall when you approach a stop sign. The additional observation is you will get code 12 for the first one when you read the fault codes. That means constant battery power to the Engine Computer, to maintain the memory, was lost within the last 50 engine starts. If you have that code but you haven't disconnected the battery, let it run dead, or pulled any fuses, that code can only set if there was an intermittent electrical connection. Nothing you said so far points to that, but the common cause would be at the fuse box. Follow the smaller red positive battery wire to the fuse box and be sure that connection is tight. Also follow the smaller black negative battery wire to the body and be sure it is tight.

When you have an engine running problem, please tell us which engine you have. In this case, since you have a fuel test port, and a cam and crank sensor, we know it's a 3.3L or the 2.5L. The 3.3L was much more common, so you'll have to correct me if you have the four-cylinder engine. You may want to borrow a fuel pressure gauge from an auto parts store that borrows or rents tools to see exactly what you have for pressure. I've been running with one tied to my radio antenna for over a year because I was chasing an intermittent problem. I have the 3.0L engine in my '88 Grand Caravan. That one runs close to 50 pounds of fuel pressure, and I can tell you from experience, it runs fine as low as 20 pounds. The engine first starts to hesitate and stall at 15 pounds. Many GM engines won't start or run when pressure is just a few pounds low. I don't have the same experience with the 3.3L but I suspect you're not looking for pressure that's just a little low. You'll either have good pressure or no pressure.

One last comment; don't even think about replacing the Engine Computer unless all other potential causes are eliminated. Chrysler had an extremely low failure rate through the '95 model year.
Was this
answer
helpful?
Yes
No
+1
Tuesday, October 1st, 2013 AT 7:54 PM
Tiny
VIKES0419
  • MEMBER
  • 13 POSTS
Ok. I have a 3.3, sorry I forgot to post that. I will update about the codes and spark soon. I had a dead battery today from a light left on so I may not get codes.
Was this
answer
helpful?
Yes
No
Tuesday, October 1st, 2013 AT 8:26 PM
Tiny
VIKES0419
  • MEMBER
  • 13 POSTS
I received code 12 as you said would be expected from a dead battery. Second code was 24. I know my '04 Wrangler did this type of thing in the digital display for mileage, but didn't know my '94 dinosaur did this. What does the 24 mean? I have yet to verify spark. As you said, caradiodoc, it is firing so that is less of suspicion.
Was this
answer
helpful?
Yes
No
Tuesday, October 1st, 2013 AT 8:49 PM
Tiny
VIKES0419
  • MEMBER
  • 13 POSTS
Caradiodoc I take back the less of suspicion thing, you never said that. I googled the code and got throttle body as the code. I look forward to solving this with you. Thank you, and I have donated 20 dollars because your quick response and help already.
Was this
answer
helpful?
Yes
No
Tuesday, October 1st, 2013 AT 9:20 PM
Tiny
CARADIODOC
  • MECHANIC
  • 33,833 POSTS
You're a wonderful human being. Thank you.

Here's a link to the fault codes:

https://www.2carpros.com/articles/retrieve-trouble-codes-for-chrysler-dodge-plymouth-odb1-1995-and-earlier-car-mini-van-and-light-trucks

A problem with the throttle position sensor itself will not cause the engine to stall. Also, they have a very low failure rate. It is more likely the wires going to it have a problem, and some of those wires have other sensors in common that COULD cause a no-start. Also, just to point something out, there is always a long list of conditions that must be met to set any fault code, and one of them is that certain other codes can't already be set. That is because the Engine Computer compares the sensors to each other and various operating conditions. If a code is set for a sensor it uses for comparison, it knows it can't rely on that so it suspends some self-tests.

A perfect example is the coolant temperature sensor and the intake air temperature sensor. The computer knows those two had better be reading the same temperature after the engine has been off for more than six hours. When they read wildly different temperatures at engine start-up, it may be able to figure out which one is wrong. However, if either one is disconnected, there is a circuit that forces the signal voltage to go high or low enough to trigger a fault code. At that point, the computer can't use it as a reference for the second sensor, so a code will not be set for that one if it is just out of range. The second one can only set a code if it is unplugged too.

I wanted to mention that because sometimes those codes are all a mechanic has to go on when looking for the cause of intermittent problems, and once one is fixed and the code is erased, another code will suddenly appear. The computer is able to resume the self-tests and that's when the next code shows up.

The good news is the throttle position sensor circuit is real easy to diagnose. The hope is the fault is there all the time so it can be found. The computer will detect intermittent glitches too but those make the diagnosis more difficult.

The TPs has three wires. You must take the readings with the plug still connected. You can poke a pin through the rubber seals around each wire and use that with your voltmeter. Start with the middle wire. The ignition switch must be on. The voltage must be within approximately 0.5 to 4.5 volts. 0.5 volts is at idle / closed throttle. At wide-open-throttle you'll find around 4.2 to 4.5 volts. It's when that voltage goes outside that range that tells the computer to set a code. As you can see, the codes for '95 and older vehicles don't get real specific. Code 24 doesn't specify if the voltage is high or low, just that it's wrong. The voltage reading will tell you which.

If you do find around 0.5 volts at closed throttle, the other two wires have to be okay, and the sensor is working properly, at least at this moment. The other two wires are the 5.0 volt supply and the ground wire which will actually have 0.2 volts. If you find the full 5.0 volts on that signal wire, the ground wire is broken, (or the sensor is broken internally). If you find 0.2 volts, the 5.0 volt supply wire is broken, (or the sensor is broken internally).

Before we get too involved, measure that voltage, then we'll figure out where to go next. The 5.0 volt feed wire and the ground wire are in common with a few other sensors that CAN all set fault codes at the same time, and if either of those wires breaks, you'll have multiple codes at the same time. Since there aren't multiple codes, I'm betting the wires are okay.

Measure that voltage and holler back with what you find. Don't use a test light because it will load the circuit down resulting in a false reading.

Excuse me, madmike1735, for butting into the conversation with my first reply. As you can see, I type long replies, and I don't see when someone else has already posted a reply while I'm still typing.
Was this
answer
helpful?
Yes
No
Tuesday, October 1st, 2013 AT 10:57 PM
Tiny
MADMIKE1735
  • MECHANIC
  • 951 POSTS
Doc, I love reading your replies. Never feel like your "butting in" we are on the same team. I have learned a thing or two from you. Keep it up.
Was this
answer
helpful?
Yes
No
Friday, October 4th, 2013 AT 4:24 PM
Tiny
VIKES0419
  • MEMBER
  • 13 POSTS
Now since I have no time and seem to have somehow misplaced my dmm, a note I wanna make is that when turning the key on but not starting the gas light comes on and the pump hums for barely half of a second. Sending unit?
Was this
answer
helpful?
Yes
No
Tuesday, October 8th, 2013 AT 7:57 PM
Tiny
CARADIODOC
  • MECHANIC
  • 33,833 POSTS
Do you mean the "low fuel" light? There was a common complaint of the gas gauge reading correctly when the tank was between "full" and "1/2", and between "1/4" and "empty, but from "1/2" to "1/4" the gauge would drop to below empty. That was due to a weak movable contact on the float arm on the sending unit that caused intermittent contact. Most people replaced the sending unit which just snapped onto the fuel pump housing, but that one could develop the same problem. The better fix was to pull the pump assembly part-way out of the tank, press a little retention finger to allow you to slide the arm off, then bend that contact to make a more solid connection, and pop it back on.

That arm had two contacts for the '91 models, one for the gauge and one for the "low fuel" light. For '92, they used a Body Computer to read the resistance of the sending unit with a single movable contact, then the computer decided when to turn on the light. If the light turns on, the gauge should be close to "empty". If that's not the case, that computer is the likely suspect. It lives right in front of your right knee cap under the radio. Remove the "knee blocker". That's the panel under the steering column, then the computer is slid in to the right. It has two connectors on the left side that you can see and use to pull the module out.

The fuel pump is only supposed to run for one second when you turn on the ignition switch. That is to insure fuel pressure is up for starting in case it bled down overnight. After that the Engine Computer will only turn the fuel pump relay and the automatic shutdown relay on when it sees engine rotation, (cranking or running). That is their safety system in case a fuel line is ruptured in a crash. With no fuel pressure the engine can't run. When the engine stalls there will be no pulses from the camshaft position sensor or crankshaft position sensor. With no pulses, the computer turns those relays off. That stops the fuel pump from dumping raw fuel on the ground creating a fire hazard, and the ASD relay removes power to the ignition coils and injectors.

When you hear that hum, you know the fuel pump, relay, and wiring are okay, and the Engine Computer has control of it. The coil part of the relay is wired in parallel with the ASD relay so you would know that one is turning on too. What you don't know is if they're turning on during cranking because there's too much noise to hear the pump. That's where it becomes necessary to take a voltage reading at the coil pack or any injector. If I didn't describe it already, look for the same color wire, usually a dark green / orange, at the coil pack and at every injector. You can back-probe through the rubber seal on any of those wires to see if the voltage comes back during cranking. If it does, you know the cam and crank sensors are working.
Was this
answer
helpful?
Yes
No
Wednesday, October 9th, 2013 AT 2:29 AM
Tiny
VIKES0419
  • MEMBER
  • 13 POSTS
I see. I need to find my DMM to get to the next step. The fuel light only comes on when I turn the key on and the hum of the pump is present. It turns off after the pump shuts off. It hums again after it dies as if to prime again for another start. I will get those readings and update soon. Hopefully I can pinpoint this soon.
Was this
answer
helpful?
Yes
No
Wednesday, October 9th, 2013 AT 7:43 AM
Tiny
VIKES0419
  • MEMBER
  • 13 POSTS
Ok, so to re-open this thread. It is cold out and I have short amounts of time to do anything. I put a used throttle body on and now she will run for 5-30 seconds and die. Sputtering the whole time, voltage readings on tps are within range. Old gas? Too cold? I'm in St. Paul area of minnesota. Help? I might have to salvage her if I can't get a fix soon. Also, coil packs? Seed it was running on 3 cylinders when I got it to.
Was this
answer
helpful?
Yes
No
Sunday, December 15th, 2013 AT 2:02 PM
Tiny
CARADIODOC
  • MECHANIC
  • 33,833 POSTS
Sorry about your climate. I'm three hours straight east of you and I really, really hate winter.

You'll have to catch me up. It will take a while to reread everything.
Was this
answer
helpful?
Yes
No
Sunday, December 15th, 2013 AT 2:07 PM
Tiny
VIKES0419
  • MEMBER
  • 13 POSTS
94 grand voyager will fire wont run. Replaced fuel pump, regulator, has spark, pressure, all that jazz. Codes pointed to map sensor. Got a used throttle body from my work (viking auto salvage) and put that on as it came with another map and tps. Codes came up as map sensor so I did that. I think my oil might be too thick to run in this cold. I forgot I put 10W-30 for the summer meaning to change to 5W-30 for winter. Round about of the story.
Was this
answer
helpful?
Yes
No
Sunday, December 15th, 2013 AT 2:40 PM
Tiny
VIKES0419
  • MEMBER
  • 13 POSTS
3.3L engine. Tune up done less than a year ago.
Was this
answer
helpful?
Yes
No
Sunday, December 15th, 2013 AT 2:42 PM
Tiny
CARADIODOC
  • MECHANIC
  • 33,833 POSTS
The oil isn't an issue. My daily driver is a rusty trusty '88 Grand Caravan with well over 400,000 miles on the engine, and I run the cheapest 10W-40 year-round.

Where is your MAP sensor located? I have a '95 Grand Caravan but I'd just as soon not run out into the cold to look at it. As I recall, it's on the passenger side near the alternator, not on the throttle body assembly like on the newer vehicles. That means there's a vacuum hose going to it. Be sure that hose doesn't have any leaks.

The MAP sensor is fed with 5.0 volts, and puts out a signal voltage of between approximately 0.5 to 4.5 volts. It has to go beyond those limits to trigger a fault code. As long as it's within that range, it can be wrong, but it won't set a code. When it's wrong, even by a little, it will cause very poor performance or stalling. It can also set a code for a pneumatic fault. That's when there's no change between ignition switch on and engine running. When the ignition switch is first turned on, it takes a barometric pressure reading. Next, the Engine Computer wants to see the voltage drop a bunch when the engine starts meaning it's reading manifold vacuum. That sensor has the biggest say in how much fuel is commanded to enter the engine. When they fail pneumatically, the common symptom is the engine will start, then stall, but you can keep it running if you constantly move the gas pedal.

What you may need to do is watch the MAP sensor readings on a scanner. Most of them will read the voltage and the inches of mercury. The typical voltage will be around 1.5 to 2.0 volts at idle, and will go up during acceleration when vacuum goes down. You should also see around 17 to 20 Hg at idle.

If you find those values but it still sets a code, it is usually because there isn't the normal rapid changes in the readings. MAP sensors are sensitive enough to measure engine rpm because they can respond to each piston's gulp of air. The sensor is not supposed to smooth out those pulses. The computer expects to see a constantly pulsing signal voltage. One thing that can prevent that is if moisture or fuel condenses in the hose to the sensor. The hose on your engine isn't long enough for that to happen, but it did on the older K-cars, and there was a recall and fix for that.

You might try unplugging the three-wire electrical connector to the MAP sensor. The engine will not run well but it will run if that sensor was causing the running problem. The computer will recognize the unplugged sensor, set a fault code, turn on the Check Engine light, then inject approximate values to run on based on other sensor readings and operating conditions. If that causes an improvement, and the engine runs worse when it's plugged back in, either the sensor is defective, the hose has a leak, there's some other vacuum leak, or there's a problem with one of the wires or terminals. You would see wrong voltages on the scanner.
Was this
answer
helpful?
Yes
No
Sunday, December 15th, 2013 AT 5:33 PM
Tiny
VIKES0419
  • MEMBER
  • 13 POSTS
I correct myself where it was a tps code I got. The map sensor is located on the throttle body. I cannot get it to run really therefore hard to get a reading. Let me give it my best shot on Tuesday and get back to ya.
Was this
answer
helpful?
Yes
No
Sunday, December 15th, 2013 AT 6:28 PM
Tiny
CARADIODOC
  • MECHANIC
  • 33,833 POSTS
I think you're mixing up two sensors. The MAP sensor is not on the throttle body. It's over by the alternator. The throttle position sensor and the automatic idle speed motor are on the throttle body.
Was this
answer
helpful?
Yes
No
Sunday, December 15th, 2013 AT 7:09 PM
Tiny
VIKES0419
  • MEMBER
  • 13 POSTS
This is the reason I came to you. I was blind. Now I wonder if it still is the map sensor. If I could I would share a cold one with you. I will update ASAP.
Was this
answer
helpful?
Yes
No
Sunday, December 15th, 2013 AT 10:03 PM

Please login or register to post a reply.

Sponsored links