Updated! 2003 E 350 fuel pump stays on

2003 f 350 5.4 cargo van with a random no start. I am trying to help a church out with this van which seems to have had the parts cannon launched at it.... Including a pcm.

After a alternater replacement it would not start. Then after pulling a coil it did, but randomly.

When i first got to it i crank it. All lights on the dash stayed on while cranking. However the fuel pump stays on while the ignition is on. Also no com with any modules. I check all the fuses and relays relevant to the pcm. Next i went to verify that there is no spark or injector pulse. I unplugged a coil to test, and when i turned the key to the on position, i had both power to the coil, and the pump turned off. So of course it runs now and all the modules are talking just fine. I traced all the wires to find a fault with no success!

I know i watched a video from eric o, or paul danner, on this kind of issue, just can't remeber what direction they went.

Any help is greatly appreciated!

UPDATE!!!!

So next time it wouldn't start i could pull up live stream. It look like no rpm data but i forgot to check the engine light to see if it went out during cranking... sorry but i rarely work on fords!!! Fuel pump prime was normal. Not wanting to unplug anything i gave a injector a quick ground to fire it. The thought was to spray some gas and see if that cylinder had spark..... well it started.

Next time the battery was a brick even though it's suppose to be rather new. Wouldn't start, no com, check engine light stayed on while cranking and no fuel pump prime. Had 5v ref at the tps and i had a crank signal from the sensor. Didn't have anymore time so i came back today to finish.

Everything is the same. Checked the 5v ref at the coolant sensor, it's 10.6 volts which is below battery voltage. Not sure what to make of that! But pressing on I checked the crank signal at the pcm which is good. It also has a single wire from the pcm which the wiring diagram reads" rmp siganal". I assume this is for the tach which it doesnt't have. It reads near battery voltage and lowers when cranking.... 4 volts i think.

Next I hit every power and ground on the pcm with a 3amp headlight bulb and they all passed. So I'm at the horrible point of calling the pcm which i truly hate. I was looking at flagship one for a replacement. The pcm comes flashed with a lifetime warranty. Has anyone dealt with them?

hi I took a quik look the fuel pump is ground side switchrd from the pcm and it goes through an inertia fuel cutoff switch after fuel pump relay maybe relay is sticking or fuel cutoff swith is in position 2 on diagram I am speaking from inexperience here but it looks like a fairly simple circuit I wonder if there is a short to power somehow or pcm is staying on for that driver for some reason

If it was just the pump staying on it would be a simple diagnosis. My issue is thag when this happens it won't start. Perhaps a bad signal from the crank could fool the pcm into thinking it's running but i would think there would be intermittent pulse to the coils or injectors.

Unfortunately i don't have a wiring diagram because it seems like unplugging a coil momentarily seems to fix it. But from what I've been told that doesn't always work!

Thanks for the response!

Gotcha the wiring diagram from BBB Industries



The fuel pump stuck on is actually a helpful symptom, because it proves that the PCM Power Relay is working, too. That's usually my go-to with no comm Fords of this generation, but not the case here.

Do you recall why the PCM was replaced initially? Was it in response to this same condition? Or a different issue?

My initial suspicion is a bad replacement PCM. If possible, get it to not start again and check PCM powers/grounds/5V reference. This is the best way to make 100% sure the wacky PCM behavior isn't being caused by something external.

Pcm was replaced a year or so ago by a dealer, no idea why. I went after the coil to check controll from the computer. When i unplugged the coil it was fine. Which seems to be a common theme. I am kicking myself for not trying the 5v since that is a no brainer!

I keep thinking that something is causing the pcm to wig out. Like a frequency feading into the pcm thru the coil. But it's a coil on plug so i would have to unplug all of them. Plus the coil i unplugged is not the one that had been unplugged before when it would start.

Found some wiring diagrams so i guess I'll do some studying. Since it's a church van i hate to see them use it and this happen again!

Thanks for your responce!

Bucher wrote: Pcm was replaced a year or so ago by a dealer, no idea why. I went after the coil to check controll from the computer. When i unplugged the coil it was fine. Which seems to be a common theme. I am kicking myself for not trying the 5v since that is a no brainer!

Gotcha, thanks for the info. Don't sweat the 5V reference check - you were following good no start diagnostic procedure by checking for spark, IMO.

I keep thinking that something is causing the pcm to wig out. Like a frequency feading into the pcm thru the coil. But it's a coil on plug so i would have to unplug all of them. Plus the coil i unplugged is not the one that had been unplugged before when it would start.

This was gonna be my next question. If it's the same coil getting unplugged to make it start, then you might be onto something. If you can unplug ANY coil and make it start, then you're just seeing a symptom of the problem. It would seem that the act of changing a status flag for a coil driver (which typically sets a P035X code) is somehow snapping the internals back into life. :silly:

Let us know what you find!

I ran into a similar situation a while back on an f250 and found a poor / corroded connection to pcm power diode at bottom side of under hood fuse box. It was intermittently causing a power loss to pcm. Worth a look

Diode makes sense. Since i didn't have a diagram i was thinking maybe the pcm was powered by multiple sources. I saw a case study like this where only half the pcm powered and acted strange.

I pulled the diode with the ignition on to verify a response. The pump turned off and when the diode was pulled back in, the pump turned back on. I did that several times. The terminals looked like the were 15 years old but not that bad. Having said that i wouldn't be suprised if something like that is the issue!

If you don't mind me asking..... What were the symptoms of the 250 you mentioned?

Intermittent no start & intermittent module communication. Customer had said that vehicle had shutoff driving down road after hitting a hard bump when symptoms got worse. I was actually checking pcm relay & found terminals to be minor corroded when I removed from under hood fuse block and I decided to check under side of fuse block because they are known corrosion spots for NY & discovered that the corrosion / poor connection to be pcm diode next to relay. Hope this helps good luck.

Thinking i will check under the box.... Pa isn't much better when it comes to corrosion and rust!

Thanks again.

WIRES to diode connectors in fuse block was the problem for me not the diode itself

Was just thinking about your point about setting a code.

I unplugged the coil when the ignition was off, then turned it to the on position. That's when the pump returned to it's normal 3 second prime. Wouldn't the engine need to be cranking to turn on the driver and set the code?

I did unplug the coil while running and it did flag a circuit failure for that cylinder. I would think there would be a bias voltage that would change when the driver is on. At least that's my guess as to how the pcm sets a code.

Yeah.... I've been thinking i will put that on the list of things to test. If it's cheap i may just replace it based on the age of the van. Unplugging the coil may have nothing to do with the problem, especially since it doesn't always fix the problem!

Bucher wrote: I unplugged the coil when the ignition was off, then turned it to the on position. That's when the pump returned to it's normal 3 second prime. Wouldn't the engine need to be cranking to turn on the driver and set the code?

Nope! The PCM will flag a code based solely on the voltage present at the PCM pin, even with the driver off. It's what Paul calls an output state monitor. The PCM is watching that pin any time the key is on, and flags the P035X whenever the voltage doesn't correspond to the driver state. Driver off, voltage should be high. If it isn't (like with the coil unplugged), the PCM knows there's a circuit issue.

I did unplug the coil while running and it did flag a circuit failure for that cylinder. I would think there would be a bias voltage that would change when the driver is on. At least that's my guess as to how the pcm sets a code.

Actually, there's no bias voltage on these coils. It'd be helpful if there was! :silly:

Makes sense to not have it though. If the pcm messures voltage inside the pcm, just before the driver, it would have battery volatge. When the driver turns on the voltage drops.

So if that theory is true, then unplugging the coil with the ignition on could flag a code. Maybe it's looking for voltage with ignition on and then a drop when it sees the rpm signal and the driver is working.

If so then unplugging something like the coil will flag a code and thus waking the pcm up from it's weird behavior.

UPDATE

So next time it wouldn't start i could pull up live stream. It look like no rpm data but i forgot to check the engine light to see if it went out during cranking... sorry but i rarely work on fords!!! Fuel pump prime was normal. Not wanting to unplug anything i gave a injector a quick ground to fire it. The thought was to spray some gas and see if that cylinder had spark..... well it started.

Next time the battery was a brick even though it's suppose to be rather new. Wouldn't start, no com, check engine light stayed on while cranking and no fuel pump prime. Had 5v ref at the tps and i had a crank signal from the sensor. Didn't have anymore time so i came back today to finish.

Everything is the same. Checked the 5v ref at the coolant sensor, it's 10.6 volts which is below battery voltage. Not sure what to make of that! But pressing on I checked the crank signal at the pcm which is good. It also has a single wire from the pcm which the wiring diagram reads" rmp siganal". I assume this is for the tach which it doesnt't have. It reads near battery voltage and lowers when cranking.... 4 volts i think.

Next I hit every power and ground on the pcm with a 3amp headlight bulb and they all passed. So I'm at the horrible point of calling the pcm which i truly hate. I was looking at flagship one for a replacement. The pcm comes flashed with a lifetime warranty. Has anyone dealt with them?

The fuel pump has power from the ignition on the control side. The ground side of the control comes from the pcm and is missing.

Originally it was staying on. But as i stated it my update it won't come on at all.

10.6V at your coolant temperature sensor? :huh: Were you measuring a single wire sensor, or the two wire sensor beneath the intake on the drivers cylinder head?

Anyway, you had crank signal and 5V reference, so I'm inclined to agree that the PCM is intermittently failing.