Fuel Pump Resistor Weirdness

[vnelson]

Hi Guys:

I have a mystery going that I was hoping someone could solve or at least point me in the right direction. I got my timing belt service, which included fixing all the parts that should be fixed at the time (coils, water pump, etc.). Subsequent to that, the car would idle like a dream but hesitate slightly off the line unless I eased up to speed. In fairness, the car sits most of the time, but I use a marine fuel treatment and am reasonably sure that the gas is good, although the injectors might be gummed up somewhat from lack of use (I know, bad me). After reading other some threads of possible causes, I decided to test my fuel pump resistor by short-circuiting the plug. When I short-circuited the resistor plug, the car would start but quickly die. So, I connected the resistor back up and now the car will not run. I let it sit overnight and then started it. It ran for about 10 seconds, then died and gave me a CEL. I don't know what's wrong and would like to tap the collective intelligence for next steps. Any help would be greatly appreciated!

 

[jwmelvin]

What is the CEL code?
What year is the car?
Seems possible the fuel pump is dying. I’d try to check fuel pressure.

 

[goldNSX]

Did the phenomena occur right after the TB job?

Check the fuel pressure at idle. If it checks out ok clean your injectors and replace the seals.

 

[NSXGB]

http://www.nsxcb.co.uk/entry.php?2361-Eng-Refresh-Cooling-Sys-etc-01

 

[Old Guy]

I second the link provided by [MENTION=8378]NSXGB[/MENTION].

Normally applying a short to the fuel pump resistor connector should take the fuel pump resistor out of the equation. However, you have to make sure that the short / shunt is actually working. Depending on how you inserted the paper clip, it may not have made contact; but, in doing this you may have disturbed the wiring on the resistor. As described by Kaz, if the wiring has deteriorated you may have created a bad connection which is resulting in really low voltage to the fuel pump.

When you shorted the fuel pump electrical connector, even if the short was faulty the engine will start because the ECU by-passes the fuel pump resistor during cranking (learned that from Kaz). Once the engine catches and engine RPM rises above the running threshold the ECU switches off the by-pass. If your short is faulty, the engine will then immediately die. A faulty fuel pump resistor can mimic a worn ignition switch in this sense.

The definitive test would be to remove the resistor from the car and connect an ohmmeter to the resistor leads to measure the resistance. First is the resistance within spec for your model year (it changed sometime during the production run)? If it is out of spec, especially if it is high, its time to replace. If its in spec, with the ohmmeter still connected wiggle or move the wires where they exit from the resistor body. If you get any material fluctuation in the resistance its time to replace the resistor. If the resistor passes the test, its time to look elsewhere.

 

[vnelson]

Thanks for the suggestions, guys. Sorry about the delay in responding; I stupidly posted the question and then went out of town. I'm so appreciative of the explanations that help me learn more about the inner workings of my NSXs (92 & 93!). Anyway, I have a lot more information (and I think it's leading to a new fuel pump). :frown:

1. I took off the resistor and tested it at .7 ohms. The car will not run with the resistor attached. It starts and dies in 5 seconds.

2. I sprayed electrical cleaner on the leads and jumped the resistor connector again and this time the engine started and ran, although idle was very rough. Once I got up to about 1500 RPM, the engine ran better. I have a sloped driveway and the car died every time I tried to drive it up the driveway. I finally got the car up the driveway and on to the street. If I gave too much throttle too fast, the car would stumble terribly. if I slowly increased throttle, I could almost avoid all stumbling (except starting off from a stop). Once the car was up around 3000 RPM, it ran fine and did not stumble (again, without the the resistor attached).

3. While driving, I got a CEL, but when I stopped to top off the gas tank, it was gone and did not come back for the 3 miles back to my house.

The only other thoughts of note is that the car sits most of the time, I use marine fuel stabilizer, and I recently put Techron in the tank thinking there injectors might be gummed up.

Should I test the fuel pressure and do you know where I can get a gauge that will fit the bolt on top of the fuel filter?

Again, thanks so much.

 

[Old Guy]

0.7 ohms is in the correct range for an na1 NSX. If the wiring looks good then it is probably not the resistor.

Use the service check connector to retrieve the error code. Even though the light does not remain on, the code may be stored. Some errors require more than one event before the light stays on permanently. However, based upon your description it may just be a misfire code which may not be particularly helpful.

Testing pressure would be a good idea. Honda makes a gauge with a fitting which goes where the center part of the banjo bolt fits - part # 07406-0040001. I expect it comes with a suitable price tag. However, with this thing you should be able to connect up a test gauge.

https://www.glowshiftdirect.com/fuel-pressure-banjo-bolt-thread-adapter-for-honda-vehicles/

Your pump may be dying; but, change the filter first to make sure it isn't the problem (unless it is relatively new). Filter is easier to change. Pump, not so easy to change.

 

[vnelson]

Thanks! I changed the filter about 2 months ago. I will test fuel pressure next.

 

[goldNSX]

The only other thoughts of note is that the car sits most of the time, I use marine fuel stabilizer, and I recently put Techron in the tank thinking there injectors might be gummed up.

These additives might help a little bit but fuel goes bad over time. How old is the gas in the fuel tank? How long has the car been sitting since the last storage?

If the injectors are the culprit an additive won't help much in that case. You have to dismount them and let them clean by a shop. Do you think it's running rich (strong smell of gas off the exhaust?) Dripping injectors can induce the problems you describe.

But as Old Guy mentioned it could be the other way around. Your exhaust will give you a hint about what's going on...
A fuel pressure test is little bit easier than taking out the injectors but I'd check both anyway.

 

[vnelson]

Thanks. No gas smell, but I wouldn't rule out the injectors needing a cleanup. I will do the pressure test first and see where it goes. I priced the fuel pump job and it's coming in around $550.00. I think that's low enough to outsource this time!

 

[vnelson]

So I did a pressure test with the fuel regulator disconnected and at first, the gauge read 50 psi and dropped a little when I put the vacuum hose back on the regulator. Then I noticed that there was a leak and shut it down, retightened everything, and started the car. No leak, but the pressure reading was zero. I messed around with the gauge, blew compressed air through it (it moved), took out the Schrader valve, and re-tested. I let it run for a while, which was stressful because the engine was loping and sounded kind of oil-starved (it's not). Net effect, zero psi. Can the car run with no fuel pressure registering on the gauge? It just doesn't make sense to me...

 

[vnelson]

Oh, one other thing, I do hear the fuel pump run to prime when the keys is turned.

 

[Old Guy]

Fuel pressure regulator disconnected? Based upon your later comment I am assuming you meant the vacuum reference line from the manifold disconnected - not complete disconnection of the regulator which would be a very bad thing to do.

Most fuel injection pressure gauges have reduced accuracy at the lower end of the scale on the gauge and less expensive ones probably have quite a large error below about 15 psi. So, even though your gauge was reading 0 psi, it might not have been 0 psi. Also, remember that the gauge reads 'gauge' pressure relative to atmospheric pressure. At idle the NSX might have a manifold pressure of -10 psi so even if the pump is close to clapped out and can't make much pressure or the FPR has a problem there might be enough pressure differential across the injectors to allow the car to idle - badly. Although, never having tried it I can't be definitive in that statement.

Take a clamp or pliers with the teeth covered by a rag and use it to gradually clamp off the fuel return line on the fuel pressure regulator with the engine at idle. As you constrict the fuel return line you should see the fuel pressure rise. If the fuel pressure rises then your pump might be OK and you may have a faulty regulator. If you can completely shut off the return line and the fuel pressure does not rise, then the fuel pump is worn out or your fuel pressure gauge is faulty. Its also possible that you have some mega fuel blockage someplace; but, you have to rule out the pump / FPR / filter as culprits before moving onto that particularly nasty diagnosis. I would confirm that the gauge is working before jumping to pump replacement.

In the very first post you mentioned that you had a CEL; but, you never did provide the error code. The error code may provide some useful info, particularly if it is a mixture out of range error code.

 

[goldNSX]

You've read 50 psi at first which is a good sign. 0 psi at idle means having a serious problem in the fuel system.

Check the following scenario: Let it cool down overnight, turn the key to IGN but Do NOT start the car and read the fuel pressure within the first 2 seconds after doing so (or a long you hear the fuel pump priming). Value? Again around 50 psi?

Restricting the return line is also a good test to follow as Old Guy mentioned. You have a serious leak within the fuel system if the fuel pump is working correctly. It could be a bad FPR (aftermarket?) or dripping injectors.