Allpar Forums banner

Emergency - Can a mechanical fuel pump fail intermittently?

23395 Views 119 Replies 5 Participants Last post by  av8r115
Hey guys,

My '87 3.9L Dakota just quit on me and I've got to get it fixed TODAY. I think it's the fuel pump, but let me know if you think it's something else.

Going down the road the engine just died - no sputtering or anything. BUT when I pumped the gas it ran briefly, less than a second. That was while I was still coasting in Drive. It did that the first 2 or 3 times I pumped the gas, but not again.

After sitting about 45 minutes, it started after cranking for a couple of seconds.

I think the carb is being starved of fuel, the brief seconds it ran were when the accelerator pump shot fuel from the bowl (assuming this carb has a bowl), but then ran the bowl dry. I think that's also why, when it did start again, it took a few seconds of cranking.

But I don't understand why it would start again after sitting a few minutes. That's what makes me unsure if its the fuel pump or not.

The fuel filter was replaced in August.

If I was flush with cash I'd just slap a new pump and filter on it. Right now I've got be careful what I spend the $$ on.

I thought about water in the gas, but wouldn't that make it sputter for a while before it just quit completely?
1 - 20 of 120 Posts
Bob Lincoln said:
You really would have to have a fuel pressure gauge spliced in to measure the pump pressure to diagnose it. Or disconnect the fuel line and crank the engine (with primary coil disabled) with the fuel line inside a glass jar, and observe whether it delivers constant fuel supply.

Very possible that it's flooding, and that pumping the gas gave it more air to compensate by opening the throttle plate. And that's why it would take 45 minutes for it all to evaporate before it would start.

Flooding vs fuel starvation should be easy to tell, by removing the air cleaner lid and smelling for gas. If none, starvation. If heavy odor, flooding.
Ok, flooding would explain why it runs again after a while. What would cause it to suddenly start flooding out? What should I check/adjust? I know the carb is getting fuel now because it will run fine after it sits.
So is it pretty unlikely that the fuel pump would be failing intermittently?
The more I think about the flooding, the less sure I am about that. If it flooded enough to kill the engine, I should have been able to smell through the vents. Plus it didn't sputter or surge or anything at all - it just died.

Haynes doesn't say anything in the troubleshooting about flooding, so no help there. I guess I'll put a new pump and filter on it and see where that gets me. I've just never seen a fuel pump or clogged filter act this way.
Guys, could it be an ignition coil going bad? That would explain why it just up and died w/out warning, and then starts again after it sits a while.

It just doesn't act like a fuel pump or filter problem. It's starts fine and revs up fine. And while it was running I shook the truck good and hard to see if there was any junk in the fuel tank. Engine didn't miss a beat.

I know I'm rambling fellas, but I'm barely able to rub two nickles together right now, so I'm sweating putting the $$ in the right place.
Bob Lincoln said:
I'll take a look. I think the ENGINE light doesn't work.

I noticed that the tail pipe is coated with soot, inside as well as about a 1/4 around the outisde.. That means the mixture is too rich, right? Also, this time when I was checking the revs, when I idled back down there was a puff of black exhaust and the engine died. There's a lot of moisture and black on the ground by the exhaust where I've been revving the engine.
Bob Lincoln said:
That's running way too rich. I suspect a carb rebuild is necessary.

Autozone and other parts stores can read your fault codes for free, with a code scanner; so the Check Engine light doesn't have to be working.
Where do they plug in the scanner? I haven't seen anything that looked like an OSBII plug. (Oops - OBDII)

All right. I'm just going to have to try driving it. I'll go to AutoZone first and see if they can pull the engine codes.
I just drove 20 miles of city streets and interstate - no hiccups at all. I guess I blew out all the crud while I was revving the engine earlier.

The O2 sensor on the left manifold looks like the original. I need to replace the manifold (cracked) and I just figured I would replace the O2 sensor at the same time. Found a manifold in a local pull-a-part, but can't get it to come loose from the cylinder head.

How hard is it to rebuild the carbureter? And any ideas on how to break a manifold loose?
valiant67 said:
Sometime you've got to tap the manifold with a hammer to break it loose after all the bolts are out. They can stick stubbornly to the engine.
You can say that again! I even tried prying gently on it at the three places it connects to the cylinder head. When I have the $$ I'm going to head back with a foot long piece of 2x4 and my small sledge to see if I can knock it loose. My engine has a small coolant leak in the head gasket on the right front end, so I had planned to just do all of it at the same time. I'm going to have to rethink that since it looks like rebuilding the carb and replacing the O2 sensor are now a bigger priority.

Any tips on rebuilding the carb? I've rebuilt an old motorcycle carb, but nothing as complex as this.
Yeah, I never top off the tank for that reason. I've only had the truck for 4 years, so I don't know about the person who had the truck before me.

The issue you described w/ the pickup plate is exactly what's happening w/ my truck, and it would explain why the engine simply quits w/out sputtering or running rough before.

How do you check for spark at the coil wire? Do you mean hold it close to the distributor and have someone crank the engine?
7T3440CHARGER said:
Hold the coil wire about 3/4" away from a ground, the intake will work if a clean unpainted spot is found.(Keep away from fuel source) Have someone crank the engine and look for a clean blue spark. If a red spark is present there is not ample spark, further diagnosis is required to see if the coil is the culprit as well. If there is a ton of miles on the vehicle a coil is advised with the pickup plate, cheap insurance to avoid back to back failure of a new pickup plate.
If I understand, you mean a weak coil could cause a new pickup plate to fail?
The distributor cap, rotor, plugs, and wires are about 3 years old. What are the symptoms of a failed coil driver in the SBEC?
Looks like I need to go ahead and replace the pickup and coil. Sounds like that would not only fix the problem w/ the engine quiting, but would improve my MPG as well. Money well spent.

The truck just crossed 220K miles on it. Given the condition of the truck when I got it, I'm willing to bet neither the timing chain or distributor bushing have been replaced.
7T3440CHARGER said:
Pickup plate is usually less than $75.00 for OEM.. Use a quality part, not a big box auto parts special.... If the coil is original, look for oil leaks at the coil wire terminal, sealing edge at the housing and at the wire lug terminals. If oil seepage is present replace the coil. Set ignition timing on your V6 at 12 degrees before TDC for better performance and fuel economy.. I've worked on these engines since they came out in 1987 in the dealership and have owned a few as well. The last one I installed had to be obtained aftermarket as the manufacturer no longer serviced the plate. You will have to remove the distributor to properly install the pickup, you can check the drive gear once the distributor is removed for excessive "Feathering" of the gear. This will indicate dist. drive gear wear. This was a common issue in the later fuel injected Magnum motors, not carb equipped 3.9's.
My cash on hand only covers the big box stores - O'Reilly's in this case. Not my first choice, but they are within walking distance so that's where I bought a fuel pump and filter that I'm going to exchange for the pickup. Haynes indicates the pickup can be replaced by removing the distributor cap and rotor. Is that not correct?

On the coil, there's no oil seepage at all from any of the posts, so am I safe to keep using the old coil?
I know there's at least some slop because the last time I checked the timing it tended to jump around a couple of degrees. The question is whether it could cause the engine to quit?

It's stopped raining, so I'm going to check the resistance on the coil.
I can't get a decent reading off either the old or new coil. I'm using a regular Radioshack multimeter, not an automotive tester, but it should still be able to test resistance, right? I do get a continuity tone from both the primary and secondary coils (positive to negative, and secondary to negative).
I can't get any reading at all. Is there something significantly different about an automotive meter as opposed to my household Radioshack meter?
Digital meter w/ auto range. I tried in auto-range mode as well as all the different ranges. The coil is an old cylinder style. When I first attached the leads, the reading bounced all over the place, then went to no reading at all. That's on both primary and secondary coils.

I have to work in 2 hours, so I can't mess with it anymore today. The engine runs for now, so I'm just going to return the pump, filter, and coil. Still don't have a solid answer as to what the problem is, but I'm out of time and I can't afford to stick it in the shop and let an expert figure it out.
It jumps around and then goes to 0.00. The engine runs, so the coil is working. Is there a way to test the tester?
Bob Lincoln said:
Simple household test. Take an incandescent bulb that has cooled off and measure across its contacts (between the threads and the center contact). When cold, should be about 0.5 ohms or so.
I tried the meter on a bulb that I know works, and got the same results. The decimal point marches across the bottom from left to right (which I assume means its changing ranges), and the screen alternates between the letters O.F, a string of dashes, or a string of zeroes. If I select a range, I get the letters OF or a string of zeroes. A new battery did not make any difference.

Sounds to me like something is wrong w/ this meter.
1 - 20 of 120 Posts
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.