Fuel Pressure Woes

[NSX-Files]

My 1991 NSX is NA now with CT headers, CT exhaust, and Dali big mouth cold air intake. 78k miles. I dyno'd the car a couple weeks ago to get a baseline. The power was probably 15 hp down, and it was somewhat lean, suggesting to me the 19 year old fuel pump wasn't keeping up. I already put in a new fuel filter only 7-8 months and 2,000 miles ago so it should be clean.

I'm planning on installing a twin turbo kit in the near future. So in preparation for that, and considering the lean run on the dyno, I just changed out the (I assume tired) OEM fuel pump for a new Walbro 255 HP. I also installed an AEM 1:1 Fuel Pressure Regulator, which comes with 3 different size return orifices. I initially tried the largest orifice as I figured the Walbro would be putting out a lot of pressure and volume and I wanted the FPR to be able to be adjusted to the right pressures.

The '91 manual says the pressure at idle should be 46~53 psi with the vacuum disconnected from the FPR and 36~44 with the vacuum hose connected.

I've tried all three orifices, and even put the stock FPR back on. I could not get higher than 44 psi with the hose disconnected, where it should be about 50 psi. Plus when I reconnected the vacuum hose, I was not seeing a dramatic change in pressure, if any.

So I then jumpered the fuel pump resistor so the pump would see full voltage all the time. I also put back the AEM FPR with smallest orifice. This enabled me to raise the pressure up to 60 psi. I adjusted it back down to a steady 43 psi with the vacuum hose connected. I'm still not getting an increase in pressure when the vacuum hose is disconnected, even though it should increase to between 46~53 psi.

I don't know what else could be causing the low unresponsive pressure. I've been running the tank low because I kept anticipating changing out the pump and I wanted it near empty. Could I have picked up a bunch of junk and clogged the filter? The sock on the old pump didn't look bad and that should have prevented any large particles getting to the filter.

I figure I must have a problem since I should be able to have stock pressures with the fuel pump resistor connected, but I don't. I might just install another new filter and see if that helps, but I hate wasting money on something that isn't the problem. Can tired injectors cause this, if they are perhaps sticking and leaking fuel into the cylinders? I sort of sense it has an intermittent "miss" at idle sometimes which might point to injectors.

I don't know what else to do. I'm reluctant to even toy with forced induction until I'm confident of my fuel system. I'm looking for suggestions.

 

[EdinKali]

Check the voltage coming out of the resistor(these do go bad). If it is bad you might not be getting full voltage in the lower RPM range. Also, check to make sure you are actually getting full vacuum at that hose(ie. no vacuum leaks).

 

[Daedalus]

The mistake here was not doing proper diagnosis from the start. The stock pump may have been fine. Now you've added more than 1 aftermarket variable and are trying to figure out symptoms to a problem. The return orifice on the AEM is not meant to be used to set steady-state fuel pressure. The adjustment screw does that. Why not put the stock FPR back on? They are very reliable and designed for your car. I would not suspect injectors as a cause of fuel pressure problems. If your pump really had trouble supplying adequate fuel at idle, no load, you would almost certainly have drivability problems on the road. Try pinching off the return hose for just a moment and see if your pressure doesn't shoot up. Don't let it go past 80psi.

What makes you say the car was running lean on the dyno?

 

[NSX-Files]

The mistake here was not doing proper diagnosis from the start. The stock pump may have been fine. Now you've added more than 1 aftermarket variable and are trying to figure out symptoms to a problem. The return orifice on the AEM is not meant to be used to set steady-state fuel pressure. The adjustment screw does that. Why not put the stock FPR back on? They are very reliable and designed for your car. I would not suspect injectors as a cause of fuel pressure problems. If your pump really had trouble supplying adequate fuel at idle, no load, you would almost certainly have drivability problems on the road. Try pinching off the return hose for just a moment and see if your pressure doesn't shoot up. Don't let it go past 80psi.

What makes you say the car was running lean on the dyno?

I appreciate the advice. The A/F reading from the wideband o2 told me it was slightly lean on the dyno.

The stock pump did not have supply problems at idle or noticeable drivability problems. I replaced the stock fuel pump because it was 19 years old with 78k miles so I didn't trust it to be reliably good for future forced induction and 550 cc injectors.

I did try it with the stock FPR back on and it did not improve the symptom of low pressure. Only with the fuel pump resistor jumpered and the adjustability of the AEM FPR could I get the fuel pressure up to the 50 psi spec with the vacuum hose disconnected. Ultimately, I had to turn the adjustment screw down some though because the pressure with the hose connected was still 50, I needed it to be below 44 per the manual. So now, I have 43 psi whether the vacuum hose is connected or not.

That's the main concern I have - the fact that the pressure does not change with the vacuum hose being connected versus disconnected. With my finger, I can feel pretty good suction on that hose when disconnected, but it doesn't have any effect in varying the fuel pressure. It should.

I need to get a vacuum gauge that I can connect to the vacuum line to see what the reading is. What should it be? Can my Mityvac pump read this someway?

 

[Daedalus]

Can my Mityvac pump read this someway?

Absolutely. Switch it to "Vacuum" and you can use a hose with a conical connector to connect to the FPR vacuum hose or some other manifold vacuum source. You should get about 18psi of vacuum @ idle.

EDIT: Just FYI, my car runs "slightly lean" in closed loop too. By slightly lean I mean AF around low 15s, sometimes a little higher. WOT is a different story.

 

[NSX-Files]

Okay, so I used my Mityvac. The vacuum line that runs to the FPR reads around 20" at idle. I also used the Mityvac to pull vacuum onto the FPR at idle and still that has no effect on the fuel pressure.

I also just took it for a quick test drive (1st one since putting in the new pump and FPR). It doesn't seem to run any different on the butt dyno. On one WOT accel in 3rd gear, it threw a CEL, which has now gone away after restarting the car.

I'm fairly pissed off right now. The old "if it ain't broke, don't fix it" keeps sounding off in my head.

Here's another possible clue. Immediately after shutting off the engine, the fuel pressure drops from 43 psi down to 20. If I recall correctly, there's supposed to be a check valve that keeps pressure up, so that restarts are easier. In the past, it seems like after engine shutdown the pressure would drop very gradually. Like it would take 15-20 minutes to go from 40 psi down to 20.

I'm really dreading having to pull the tank again. But I don't know what the problem(s) is/are.

 

[Daedalus]

Try pinching the return line and see if the fuel pressure shoots up or not. If not, then I would suspect the pump. If it does, then I would suspect the FPR. If the pump is in question then certainly don't drive the car hard. Check out the stored error code and see if it is related to fuel delivery.

 

[EdinKali]

Here's another possible clue. Immediately after shutting off the engine, the fuel pressure drops from 43 psi down to 20. If I recall correctly, there's supposed to be a check valve that keeps pressure up, so that restarts are easier. In the past, it seems like after engine shutdown the pressure would drop very gradually. Like it would take 15-20 minutes to go from 40 psi down to 20.

I'm really dreading having to pull the tank again. But I don't know what the problem(s) is/are.

You are correct. Fuel pressure should gradually drop so it sounds like you have an issue in that respect. I would suspect you may have a defective FPR since the pressure doesn't change when vacuum is applied and the pressure drops too fast after engine shutoff.

 

[NSX-Files]

I tried to pull the CEL code, but there was none stored when I jumpered the blue service connector under the dash. The CEL has not come back on.

I had a stock FPR from a 2000 sitting around so I installed that. I also reconnected the fuel pump resistor. I installed my AEM digital fuel pressure gauge inside the car so I could monitor fuel pressure while driving. Here's what I have now:

At idle, with the 2000 stock FPR, I've got a steady 42-43 psi with vacuum hose connected. When I unplug the vacuum hose, pressure spikes up to 50-51 psi then settles down to about 47 psi. This is all good and tells me the original stock FPR I had wasn't working right and neither was my brand new AEM adjustable FPR (I have taken it apart several times to change the orifice, and everything looks good inside, but who knows?).

I went for a drive and I get these symptoms: Under full throttle load but below 4000 rpm, the car surges and the fuel pressure is oscillating between 25 and 45 psi. The surges are in time with the pressure oscillations. Once the rpms get past 4000 (where the fuel relay clicks behind me), the fuel pressure jumps to 57-58 psi and is rock steady. At that time, the acceleration surging goes away and the engine pulls strong. Anything below 4000 rpm and I get surging and pressure oscillations when under load. In constate state cruising, the pressure is steady at around 41.

I wasn't getting this surging last night when I had the fuel pump resistor jumpered and with the AEM FPR set to static 43 psi at idle.

So is what I'm looking at a fuel pump resistor gone bad? Is the steady 57 psi at 4k rpm, WOT and full load appropriate? Should it be more or less?

I can't run the stock FPR with the pump resistor jumpered because the static fuel pressure is then too high (49 psi vacuum connected). I'm going to take another look at the AEM FPR and see if I can find something wrong with it. Like I said, it's brand new so I can't see why it would be bad right from the start.

So academically, what are the roles of the pump resistor and the fuel pump relay above 4k rpm? After searching Prime threads, my understanding is that the relay bypasses the resistor above 4k rpm so the pump sees full 12V., right? Most of the searches I did lead to fuel pump replacement being the solution. I've just put in a new pump. Did I get screwed and sent the wrong pump, or get a dud? My static pressure at idle is correct at 43 psi. But is 57 psi max too low? I read one thread where it said 84 psi above 4k rpm. This tempts me to drop the tank and reinstall the oem pump to see what the max pressure is. But it will be the third time dropping the tank in one week. I'm kinda sick of doing it.

One more thing. After driving around for 40 minutes with many WOT runs, watching pressures, I pulled back into the garage. When I turned off the A/C, the idle got really lumpy and the fuel pressure dropped down to 37 psi. The exhaust then had the burbling sound of a slight miss. Is this an injector dumping fuel? The injectors are the original oems from 91.

I feel like I'm on the path to a cure. Hopefully, the advice here will get me to the promise land. TIA.

 

[pgilliam1]

Bad engine performance below 4000rpm due to fuel suggests a bad resistor. My fuel pressure (1991 w/119K miles) stays around 40psi at idle and drops very slowly after I kill the engine. I'm not sure what pressure I have at WOT because my guage sits on top of the fuel filter. I sent my injectors to RC Engineering and it made a difference in easier/faster start-ups, but the car ran well before I had the injectors balanced so it's probably psychological that it seems to run better (they did find 2 were "dripping"). Let us know the final diagnosis.
Happy Motoring!

 

[Daedalus]

You can always break out the voltmeter and probe various points along the pump circuit. E.g., if the voltage to the pump varies with fuel pressure, then suspect something upstream of the pump. If the voltage to the pump is constant during pressure fluctuations, then the pump itself--or a connection between the pump and the meter--is suspect. Probe the voltage to and drop across the resistor to confirm or eliminate it as a suspect.

The resistor is there to reduce the pump voltage during low-demand conditions simply to reduce the wear and tear on the pump. You could run the pump @ 12V all the time, but expect it will fail sooner than usual.

Again, prove me wrong, but I strongly disagree with a leaking injector. If it was leaking that badly to drop fuel pressure I think it would be more obvious in other ways.

 

[EdinKali]

In addition to checking the resistor, check for vacuum leaks. A surge at idle is a dead giveaway for a vacuum leak. I agree with the others that this is likely "not" an injector issue unless an injector is stuck wide open. Might be worth pulling your spark plugs to see if one or more cylinders is running way too rich. If you have one or two that are black I would replace those injectors then retest.

 

[NSX-Files]

Latest Update:
I borrowed a known good fuel pump resistor from NetViper's car. Low and behold the fuel pressure oscillating immediately stopped. I then put mine back in and surge is back. So I have found the problem. Many thanks for those who offered advice.

 

[NSX-Files]

For anyone interested, I had asked what the proper fuel pressure should be at WOT on an NA car. The answer is approximately the same as static pressure at idle with the vacuum hose disconnected (i.e., 46-53 psi). This makes sense because at WOT the intake vacuum has dropped to close to atmospheric pressure.

On mine, below 4000 rpm (with the working fuel pump resistor) it was dead on 53 psi. Once past 4k, the resistor was bypassed, and the voltage to the fuel pump jumped to full 12v., then the pressure rose to 57 psi. I suspect this is just a matter of the stock FPR not being capable of bleeding off the excess pressure from the Walbro. But, it results in a little extra fuel so I'm okay with that for now since I have an SOS chip in the ECU right now.

 

[Larry Bastanza]

If you have an ohmmeter handy it would be great to see the resistance reading comparing the bad one to the good one.

Regards,
LarryB

 

[NSX-Files]

If you have an ohmmeter handy it would be great to see the resistance reading comparing the bad one to the good one.

Regards,
LarryB

Agreed. I thought about that. I don't have a new one yet to measure, but I will post the ohms of the bad one.