Just another fuel pump issue
 

While attempting to finish the new drivetrain on my turbo project, I ran into an electrical issue with my fuel pump. Here is the sequence of events:

1. Turn key, battery (walmart special) is dead. plug in smart charger for a while.
2. Try again. Fuel pump primes and engine struggles to turn over. Leave charging for longer.
3. Try again. Fuel pump primes and engine caches. I shift into gear and verify the new drivetrain is working in the air at idle then shut off engine.
4. Try again a minute later. engine struggles to turn over and does not catch.
5. Immediately try again and realize the fuel pump is not priming at key on.

Here are the troubleshooting observations I've made so far after charging:

1. Fuel pump (FP) fuse looks good.
2. Main relay looks good and CEL appears at key on.
3. I am running the DW upgraded FP hardwire/relay kit. Fuse is good.
4. With key on, I can jumper the FP from the diagnostic box successfully.

It's been a while since I've touched the fuel system, so I can't think what to check next. Any ideas?

SOLUTION EDIT: Low power cranking caused ECU to corrupt. Firmware reflash fixed issue.

Just to confirm, step 4 in the second paragraph, the fuel pump *will turn on* when jumped from the diagnostic box? Have you charged the battery since noticing the fuel pump doesn't prime with the key on, and does the engine still have a slow crank?

I'd cover the absolute basics first, bust out a voltmeter and check battery voltage at the battery terminals. If it's holding a charge and you have at least 12v on tap, unplug fuel pump connector and confirm what voltage is at the connector with the key on. You could also turn the key on and turn on the fuel pump output in tunerstudio so you don't have to turn the key and look back at the voltmeter reading simultaneously. I'd confirm there's no resistance on the ground circuit while the connector is unplugged as well.

With most cars I've worked on (haven't confirmed this on my Miata, though), ~10v is enough to turn on the instrument cluster and dash lights, but won't be enough to do much else.

Honestly, based on all the info so far though, it sounds like the battery is probably toast.

Yes, that's correct. The fuel pump turns on normally when using the diagnostic box, but does not prime when the key is turned to on.

It originally did prime for the 1st few start attempts, but then stopped priming after the 3rd or 4th.

The battery is now charged and reads 13.1V and the car cranks as expected, but no fuel pressure.

Given that it will turn on from the diagnostic box, the battery looks good, and all the fuses I can think of look good, Im fairly lost as to where in the circuit I am having an issue. I'll have time to dig deeper into the car over the weekend when I have time.

Well if there's good power and ground at the pump itself, that'd rule out the pump hard wiring and point to something in the trigger circuit to DW relay being the concern. What's weird is that a wiring concern would develop like that directly after the battery dying.

The DW hardwire kit you have is the generic one to hardwire the FP to the battery, using the factory FP power to trigger the relay, yeah? Like this one: https://deatschwerks.com/collections...products/fphwk

I'd remove the power wire that triggers your DW FP relay (the OEM wire that used to power your fuel pump) and confirm voltage supplied to it with the key on. If there isn't voltage there, something with the OEM fuel pump wiring, or the OEM fuel pump relay itself has failed. I'm assuming the ground side of the relay trigger is good, since the pump runs with the diagnostic pin jumped.

Battery.

Time to spring for a new one.

What is the reasoning? With 13V and all other systems operating as expected, It seems fair to rule out the battery, no?

Yes, that's the kit i used.

I'm also assuming its something in that trigger circuit as you mentioned, but i've completely forgotten what the potential failure points are. I'm unfortunately one of those cases of "knows enough to do damage" when it comes to wiring, but i'll dig into those wires sunday and narrow down the fault.

Your ecu primes the fuel pump with key on. Sounds like you’ve tested the fuel pump and it’s working as expected. What ecu? Can you connect to it if standalone?

Assuming MS3 by the signature. If the battery has enough juice to crank the engine at a reasonable speed, I was gonna suggest turning the fuel pump output on in Tunerstudio and tapping the FP output wire at the ECU to see if it's providing ground signal, then look at the rest of the trigger circuit/OEM relay after confirming that’s good.

Have you load tested the wires with a headlamp bulb? I'd disconnect the pump and connect a bulb with leads then key on the car. See if it's bright. Just cause you are seeing 13v on the battery or at the pump doesn't mean the amperage is there to run it.

Got hit with 10* and snow (essentially end times for Dallas) so I'm shelving this until I can feel my hands in the garage.

I am running MS3 and thanks for the pointers. Ill see if the amperage might be affecting it, then start testing from the ECU.

Food for thought, but could something like a clutch switch failure in the ignition circuit cause this behavior? I'm thinking not since it cranks.

The clutch switch should only interrupt the starter circuit not the fuel pump circuit. If the car cranks with the clutch in but not out then it should be working correctly.

Pull the battery and have it load tested

check all your grounds

if you did the kpower PPF delete you have the ground that attaches to the rear of the ppf that you need to relocate.

Thanks for the reminders. I previously did the PPF delete and happened to be missing the head -> chassis gnd + I improperly re-grounded the PPF -> chassis gnd, resulting in a disconnected starter circuit. Just about tore my hair out trying to solve that one.

I remembered the clutch switch was somewhere in there, but glad to hear they're separate.

Finally warmed up enough for me to hit the garage and I've fixed the issue.

After tracing wires and just becoming more confused, I tried to plug in to the ECU for the testmode and got a "no firmware loaded" warning.

It would appear that somehow during my half-charged battery crank attempts, the ECU corrupted and turned into a non-useful brick. Luckily a firmware reflash seems to have solved the problem as it starts and idles like normal now.

Thanks for the follow-up. That was not my first guess but glad to hear you got it sorted semi-painlessly.

I had a similar instance a month or so ago while trying to set up the wiring for my flex fuel sensor. Was sitting in my driver seat looking at TunerStudio and placed my phone (which has a magnetic case) on the ECU. The ECU didn't like that. Immediately got a 20 page difference report in TunerStudio saying the ECU settings didn't match what was showing up on my laptop. Luckily, I was able to resend my tune from TunerStudio right there and it wasn't an issue. Would've possibly gone down a similar rabbit hole if I didn't have the computer connected, though.

Glad you mentioned that because that's something i absolutely would do :rofl:

All the input is appreciated, I guess ill update my build thread after a year as a treat.