Fuel Pump Not Priming

[furinax]

When I turned the car on this weekend I heard the click of the fuel relay but no actual priming of the fuel pump. Main relay and starter were just replaced so I know they shouldn't be the source of the problem. Suggestions on which I should check first?

 

[goldNSX]

The relay might be clicking but still makes no contact to the fuel pump. I'd check the voltage at the fuel pump.

 

[Old Guy]

The fuel pump 'control' relay is inside the main FI relay case. It is a two for one special! What Honda calls the fuel pump relay is actually the relay that by-passes the fuel pump resistor.

With a voltage tester, back-probe the black/red wire at the fuel pump resistor (just because it is easy to get to) and check for voltage between that point and ground while somebody turns the ignition key to the run position to initiate the prime pulse on the pump. If you get a short (approx. 2 sec) indication of 12 v and the pump does not run, sad to say that it looks like your pump might have died. If you don't get that 12 v pulse, then the first place I would look would be the main FI relay (go figure!). The relay may be actuating; but, as goldNSX states, you could have one of the infamous solder fractures on the connection to the pump. The main FI relay is not the only possible cause; but, if it is not the pump its the first place I would look.

Make sure that when you turn your key to run, that the dash lights including the all important MIL light are going through the test cycle. If the MIL light does not light up that indicates that there could be an ECU related problem (not likely; but, good to rule out).

 

[emac]

I went through this not long ago. My car had sat for long time and the fuel pump locked up. The WSM has a very good flow chart to follow to troubleshoot. I had power all the way to the pump. I ended up dropping the tank and replacing the fuel pump and sender unit.

 

[furinax]

Thanks everyone. I'll go check these when I get the chance. No garage means I have to do it on weekends when it's not raining out. Sigh... Would not rule out a bad fuel pump though. That thing has been in there for a while (~15 years).

The last issue I had with the headlights was due to bad ground because of rain corrosion most likely. I wonder if this would be the same case.

 

[furinax]

Well after testing it, it seems like it may have been a faulty main relay. Is the main relay and the fuel pump relay on the same unit?

Reading through the service manual is actually pretty fun. It has a flow chart on diagnosing the issue. First thing I did was jump the fuel pump resistor with a paperclip and it didn't change anything. I didn't measure any voltages. After taking off the back panel I turned the key to on and still heard clicking but no priming. About the third time I heard a short priming of the pump. As I was cycling it off and on the main relay started to spazz out and click multiple times, but eventually it started behaving normally. Now it primes and starts fine.

I just replaced it this past winter... is it possible I got a defective unit? Or could there be an open circuit somewhere causing the relay to not get power? I should leave my old one in the car just in case I guess. The old one was fine I just replaced it when I had a bad starter issue.

Anyway, it works now.

 

[jwmelvin]

The main relay controls power to the fuel pump. There is another relay that controls whether the power to the fuel pump passes through the fuel-pump resistor. That latter relay is small and behind the driver seat, close to the center line. I would certainly keep a spare in the car. I’d probably also fuss with the connections to make sure they are secure.

 

[emac]

I agree, the wsm for the NSX is great! If it was working fine prior to swapping the relay, I would double check all your connections or maybe you have an issue getting power to the relay. I would keep the spare in the car along with enough tools to swap it out.....just in case.

 

[Old Guy]

I just replaced it this past winter... is it possible I got a defective unit? Or could there be an open circuit somewhere causing the relay to not get power?

Both of those are possibilities. Since you replaced the main relay recently, I suggest inspecting the plug that fits into the relay to make sure that there are no loose connections or damaged wiring. Check for the possibility that the female pin in the plug has slipped back in the plug and is not making contact (or intermittent contact) with relay pins. With the car running, you could also try tapping the relay case with a screwdriver handle to see if you can precipitate a failure. Other than that, its wait until it happens again.

 

[furinax]

Ever since I switched it back to the older main relay it has been working fine. I think I may have bought a defective one (the newer one)... I heard these can be rebuilt? Do you just re-solder all the connections? Looks fine to me...

 

[Old Guy]

Yes and no to rebuilding. If the failure is caused by fractures in the solder joints, then reflowing of the solder can fix that. You will likely need a magnifying glass or opti visor to detect fractures. Failure can also be caused by a failed diode in which case replacing the diode can fix that. If the failure is caused by a break in the relay winding, then its typically a throw away.
Edit - failure can also be caused by wearing or other problems with the contacts (switch part) of the relay. On a relay style like that shown in the photo, best to treat it as a throw away.

Relay operation is easy to test. Below is the diagram for the relay with pin assignment out of the service manual.

You can check the relay operation as follows.

1st - apply +12v to pin 4 and ground to pin 8. The fuel pump section should operate and a continuity tester should show a closed circuit between pin 5 and 7.
2ndt - apply +12v to pin 3 and ground to pin 8. The fuel pump section should operate and a continuity tester should show a closed circuit between pin 5 and 7.
3rd - apply +12v to pin 6 and ground to pin 2. The ECU section should operate and a continuity tester should show a closed circuit between pin 1 and 3.

 

[liftnot]

i re-soldered mine and it has been happy for 5yrs+. Tiny fractures r hard to see. If that doesn't cure, then buy a new switch.

 

[furinax]

OK well it failed to start again even with the old main relay. I think next to check is the fuel pump + sender unit. Any idea what the part number is for the sender?

I would just like to note I'm on AEM 1 and have a walbro 255lph fuel pump. Did some research and it seems like the lifespan of the pump is not too long if you run often on less than half a tank of gas. I guess I can't tell the quality of the wires until I drop the fuel tank and see them, but it would be nice to have the items on hand. I should verify it with the 12V pulse at the resistor like you guys stated...

 

[Old Guy]

I have no experience with Walbro pumps; but, my limited experience with Bosch pumps is at the end of life they would enter a sometimes they will start sometimes they won't start phase. One pump did this for about 10 days before quitting for good, another pump did this for 3 days before it checked out for good. That here again gone again kind of performance can complicate diagnosis.

 

[furinax]

I talked to my mechanic this morning. He told me these are low failure items and it may likely be a bad ignition switch. Since we already ordered an ignition switch that is on the way, we can try replacing that and see if it fixes it. Start with the cheapest solution first.

 

[Mac Attack]

If you say the ignition switch was just replaced in the first post, it is very unlikely changing it again will fix your problem.

That 15 year-old Walbro pump on the other hand is most likely your problem. That thing has got to be near the end of its design life. Sorry, I've dropped the fuel tank twice on the NSX and it's not a lot of fun.

 

[furinax]

Sorry I had to correct my mistake. I meant to say the main relay was recently changed.

Ignition switch was cleaned recently as it was super corroded.... however I never did replace it, so I am not sure how old it is.

Yes I've heard dropping the fuel tank is a pain from the forum searches but thankfully I have a day job and a competent mechanic. :smile:

 

[Old Guy]

If you have a new switch on order, replacing the ignition switch on an older car won't hurt and almost falls into the category of preventative maintenance. If the check engine light was failing to illuminate during the key on dash sequence or the other dash lights were not illuminating during the key on sequence, then I would say yes to the ignition switch causing your pump problem. If everything is lighting up correctly and your pump is not priming, then I think your switch is OK and the problem is elsewhere.

 

[emac]

Have you (or your competent mechanic) run through the WSM flow chart for testing the fuel pump? I dont recall the exact procedure, but it is pretty 1,2,3. The NSX wsm is one of the best manuals I have ever used. If it is the fuel pump then you will have power all the way to the fuel pump connector. Not that hard to check. Checking for power seems easier than swapping an ignition switch.

As for dropping the tank, it wasnt too bad a job, but I have a lift. I dont think it would be fun laying on your back.

 

[Jinks]

If you do end up changing the fuel pump, go with an AEM. These are highly recommended.

 

[furinax]

Got the car back last week. It drove fine on the way back from the shop. I wanted to take it out on Saturday, turned the car on, and again click but no prime.

I was going to ask [MENTION=12656]blodi[/MENTION] if he had some insights, judging from his experience: http://www.nsxprime.com/forum/showt...ne-crank-no-start-no-power-to-fuel-pump/page3

My mechanic had the ignition switch replaced. The old one had a loose solder joint, and he showed me. I could post a picture here but it didn't fix the problem anyway. Just to clarify, I am on AEM v1.

I have not jumped the fuel pump resistor yet to test. The only thing I know for sure is on a cold start it does not prime, only clicks. I let the temp warm up to about 45F and try again when it's sunny, and boom it works. My mechanic said it is probably not the fuel pump as he has not seen many of those fail in his days. I could also order new relays but again it works when warm so it seems like the relays are just fine. My best bet is a grounding issue with the AEM that exhibits when the car is cold. It just sucks to think that at any point you could be stranded somewhere.

 

[blodi]

From my quick look over your thread it looks like it may be the pump itself. But just my guess. It's not a fun job to replace one, but if you do it and it doesn't fix the issue, at least you'll know it's not that.

My failure was never fully understood. Cody ended up rewiring the pump so that it runs on full pressure all the time. A bit noisy, but you don't notice it once the car is fired up.

 

[emac]

Identify the wires going to the fuel pump behind the seat where they enter the firewall. When it doesnt prime, check for power at the wire. If it has power then, it is the fuel pump that is bad.....or possibly the very last part of the wire going to the actual pump. This is the exact way I determined that my pump was bad. Pretty simple electrical testing. Power at the pump and no pump running is a bad pump.

Once I had the old pump on the bench it wouldn't run, I tapped it a few times and it fired up and ran fine, but I replaced it anyway.

 

[furinax]

Now that it's cold again, the issue is back. Granted I don't drive it much in the winter, but winter time is also when I get all my maintenance work done. I will have to fix this issue if I want to drive it to the shop. This problem is almost purely electrical and it only happens when the car sits overnight in under 40F degrees weather.

I measured voltage at the resistor, and it is not getting 12V pulse. So it has to be an electrical issue somewhere. Jumping the resistor makes no difference. Since the ignition switch and main relay were both replaced recently, I would rule those out. I could swap in my extra main relay and see if a warm one helps, but I have a feeling I tried that already and it didn't work. Testing the wires behind the seat may be the next step, though I may have my mechanic figure it out. Again- I hear the relay click (although it is a bit of a weak click) but no pulse and no fuel pump action. The battery is fine as the starter still works and I have had instances where the temp warmed up and the car worked again.

The funny thing is, the fuel pump primes whether or not the resistor is plugged in. Is that supposed to be the case?

 

[Old Guy]

Asking for advice and giving advice with respect to a modified car (AEM V1 ECU) is a bit of a crap-shoot. An unmodified car is much easier because everybody has a common understanding of what the hardware arrangement is.

In the attached diagram of the factory wiring arrangement, you can see that if the fuel pump resistor is disconnected the fuel pump can't run at start up (the black/red wire from the bottom of the relay goes to the fuel pump). The fuel pump can run if the fuel pump relay closes and by-passes the resistor. This normally only occurs when the engine is approaching VTEC operation and requires additional fuel delivery. It can also occur if the fuel pump relay contacts weld themselves closed. The fact that the fuel pump primes with the resistor disconnected may be the reason why you didn't measure any voltage at the resistor connection.

Does the AEM switch the fuel pump resistor in and out? I have the impression that most of the aftermarket ECU arrangements just switch the pump on or off with no intermediate voltage operation associated with the resistor. If the AEM completely eliminates the fuel pump resistor function then you may have revised wiring for the fuel pump supply. You would need to trace the wiring to find out what is going on and whether you are still supposed to have a functioning fuel pump resistor. If the AEM does control the fuel pump relay then the relay is likely damaged (or the wiring associated with it is damaged) because an operating relay would prevent the fuel pump from starting up when the resistor is disconnected.

The fuel pump relay is normally located on the inside back bulkhead. I think it moved around a bit so check the service manual to find the location specific for your year. If it is not there or not connected then you have a little wire sleuthing to do to figure out what is going on. If it is there and is connected, examine it carefully to determine why it is allowing the pump to run with the resistor disconnected. You probably need to test the relay as per the FSM.

Electrical contact failures can be temperature sensitive, reconnecting or opening up as the metal contracts and expands. On one of the cold days when the pump has failed, try taking a hair dryer and warming up individual components to see if the temperature change allows the pump to prime. I suggest you test:
- the main EFI relay
- the fuel pump relay (if it is still connected)
- the plug terminal on the AEM

Final suggestion. On the back bulkhead there is a removable round panel that provides access to the top of the tank. A cable with a 5 pin connector enters through the center of that panel. That cable goes to the pump and level sender. Open up the connector and slip a thin wire into the female side of the connector on the black/red wire position. The wire has to be thin enough to allow you to close the connector back up again. The black/red wire is the + supply to the fuel pump. Connect a voltmeter between the black/red test wire and chassis ground. Turn the ignition key to start. If you measure 12 volts for the 2 second prime (or whatever is programmed into the AEM) and the pump does not start, then the pump is faulty or you have a wiring problem between that connector and the pump. In either case the pump has to come out. If you don't get 12 volts for 2 seconds, then you have an electrical problem somewhere back up the food chain.

An AEM V1 usually means that you have a supercharger or a turbo. If so, by any chance do you have one of those boost a pump things. If you do, that would be a real good place to check a a cause of flakey operation.