• Protip: Profile posts are public! Use Conversations to message other members privately. Everyone can see the content of a profile post.

91 NSX CEL + Loss of Power Issue, could use "Prime" Help!

17 August 2011
Alexandria, VA
Update #3: Replaced fuel pump, fuel filter, fuel tank check valves (just because) and the car is running much better, but the issue is not completely solved. The CEL issue seems to be resolved (still test driving, but so far, so good). My mechanic pulled a fuel sample for me to examine and I was shocked. The 93 Octane fuel looked like green pond scum, and there was residue at the bottom of the clear container. I am going to run the tank almost to empty, then refill with the 93 or 94 Octane fuel and continue to test.

Update #2: Car back in shop, replacing fuel filter to see if that can fix the issue. If not, Square 1 here we come...

Update #1: I am sad to report that replacing the fuel pump resistor did not solve the problem. The car is still indicating O2 sensor codes are the issue, but since we just replaced them, we (the Acura tech and myself) think it may be something else. The car goes back in the shop Monday. Just for grins, we are going to replace the fuel filter (not sure when it was last replaced) and then we will go from there. Current symptoms = engine "flutters" with normal on-throttle application, usually at any rpm below 4k, although on rare occasions it may occur slightly above 4k rpm, but I don't think it has ever happened at 5k or above. On occasion, the engine will still "hunt" for idle, maybe 2x in 5 minutes. Normal idle appears to be 800 rpm, and it may fall to 500 and then quickly go back to 1k, settling in at 800 rpm all in the space of less than 5 seconds. Check engine light may still come on every now and then, resets when ignition turned off.

Sidebar: When I removed the fuel pump resistor and got to the mounting screws on the back side of the resistor heat sink, the phillips head screws had either been secured with nuclear locktite, or they had corroded a bit, or a combination of both. No amount of swearing, WD40, hair dryer heat, and taping the screw heads with a screwdriver and hammer could budge them. I went to my local gas station mechanic and begged help. 30 seconds later, problem solved. The mechanic had a pair of mini-vice grips with tapered jaws. He secured the vice grips to opposite sides of the screw head, parallel to the mounting bracket, and broke both screws loose with twisting torque...easily. Thought that I would mention this so others might be able to try this in similar situations.

Update number two to follow some time next week...

I have searched the forum and found some discussion about CEL issues, but I also have a Loss of Power (LoP) issue too, so I'm hoping the collective owner knowledge at Prime can help me. I have my suspicions about the cause, but I don't want to share that until I get unbiased feedback here first. I'm hoping to share what you suggest with my Acura mechanic to help shorten the trouble-shooting process.

History: I purchased my 91 NSX about 7 years ago, and I am at least the third owner. I'm sure the original owner loved the car, but the person who had it before me...not so much. So, this has been a labor of love to get the car back to perfect condition.

The CEL issue has been occuring even before the following mods were installed: Downforce air induction scoop; Downforce air induction system; DC Sports headers, SoS exhaust. SoS red powder coated intake manifold and valve covers are also installed, but I don't see how they could be creating any of the issues below.

1. On occasion (maybe once every 5 min or so, sometime less), engine will "hunt" for idle rpm. Normal idle appears to be about 800 rpm. Sometimes, this slowly drops to what appears to be 500 rpm, then the engine auto adjusts to 1000 rpm, and falls back to normal 800 rpm.

2. Check engine light still comes on (after my mechanic replaced O2 sensors, adjusted valves, checked coils and plugs), but less often. Turning off the engine and restarting resets the CEL. Typically while driving in 2,500 - 3,500 rpm in 3rd or 4th gear, the CEL will trigger under sustained load or if I gradually accelerate (perhaps on the Interstate to maintain speed going up a slight rise or hill).

3. Engine "flutters" (no backfire or other strange noise that I can hear), typically while driving in 1,000 - 4,000 rpm range and accelerating, any gear, and exhaust note changes (lower). Normally, when rpms reach >3,500 rpm, the issue clears itself and the car takes off as if VTEC had kicked in, only a bit more pronounced.

4. None of the above appears to occur once engine rpm reaches 4,000 or higher. The main "trouble spot" appears to be between 2,000 - 3,000 rpm, but that may just be because the engine spends more time in that rev range. Once above 4,000 rpm, the engine appears to run fine, accelarate fine, etc.

FWIW, I always run 93 Octane or a bit higher if I can find it, but nothing lower than 93 Octane, and no additives.

Anything that you could suggest would be helpful at this point. If I can just solve this issue, the car will be perfect! :smile:
Last edited:
Your #1 item may not be related...sounds like the AC turning on which adds more load to the engine.

What may be the single most-important piece of information in troubleshooting this you have failed to disclose: What code(s) is(are) associated with the CEL?


Without a code: If you had indicated a cut-off between trouble and no trouble at 4500 RPM I would suspect a bad fuel pump resistor or bad fuel pump (or some other issue that is masked by more fuel). However, you indicated 3500 or 4000 RPM...so this it probably a red herring.
Air filter brand new, will check fuel filter. Will also see if mechanic can supply some of the codes pulled...will post as soon as I have them.
Your problem sounds like the fuel pump resistor (it sits right above the fuel filter). It regulates the fuel pressure to run at a lower pressure untill the engine reaches above 4000rpm and then it increases the pressure to facilitate added fuel usage. You can override the resistor by shorting the connector to it (that runs the fuel pump at max pressure all the time) and see if that clears up your problem.
Good luck!
Last edited:
Wow, fuel pump resistor...never would have thought of that one. Will pass on to mechanic. Still waiting for some codes, will advise.

Just spoke with mechanic, who said codes 42 and 43 initially, still trying to see if any more were indicated. O2 sensors were replaced with SoS ones (tyvm, yet again, SoS)!

I am very intrigued by the fuel pump resistor business, because it seems to match the symptoms...but then I wonder why a 47 code wasn't thrown...
Last edited:
codes 42 and 43
43 would be the best evidence we have so far of a fuel supply issue. 42 could be a red herring (sensor could be bad but not related to the symptoms you see, or it could be good but getting funny readings because of the fuel issue - or some other reason). I am reminded of some issues people have had when the protective covering on their O2 sensors' wires melts on aftermarket exhausts (those issues are generally worse than what you describe, but you still may want to double-check that the wires are all good).
I am very intrigued by the fuel pump resistor business, because it seems to match the symptoms...but then I wonder why a 47 code wasn't thrown...
Unless the pump fails in a way that the ECU is able to distinguish, you wouldn't get a 47. If, say, it hasn't failed but is under-performing this could certainly be the case. I would also expect 47 to be specific to fuel pump issues and not include fuel pump resistor issues (else you would have a code telling someone the fuel pump is bad, and they'd replace it, when really another part is bad). I would certainly consider the resister first as it is more likely to fail than the fuel pump (a very low failure-rate item, despite perceptions to the contrary) and much cheaper to replace. Both of these (and the O2 sensor for that matter) should be able to be diagnosed/tested (hook up a fuel pressure gauge to test pump, I don't know how to test regulator or O2 sensor but the service manual and others here probably do) to be sure you're not just throwing money/parts at the problem without knowing they'll help.
Last edited:
Both the O2 sensors and the fuel pump resistor can be easily tested before any money is spent. Just disconnect the O2 sensor and if the car runs OK then it is bad. As I said above, to test the resistor just strap/short the connector and if the car runs OK then it's bad.
Based on all I have read recently, here and other threads, I would suspect it may be time for a new fuel pump.

You really need to get a fuel pressure gauge on it, and set it up in a way you can read it while the car is running and acting up.

The thing about reisistors is typically they are good, or open. If it was open the car would not really run.

My $.02


Thanks for the input. I'll keep everyone posted about the results. Even if I have to pay for fuel pump and (gulp) the labor to go with it...it will be worth it...if it solves the problem! Checking resistor first though... :smile:
If you go the fuel pump route, make sure your tank is low to empty. More often then not, a pump ALWAYS fails on a full tank!! Make it a real PITA, since you need to drain the tank;).

Thanks, mechanic already mentioned that. Keeping fingers crossed that the pump is OK and only the resistor needs to be replaced. Can't get to car 'til this weekend...due to work.
Based on all I have read recently, here and other threads, I would suspect it may be time for a new fuel pump.

You really need to get a fuel pressure gauge on it, and set it up in a way you can read it while the car is running and acting up. The thing about reisistors is typically they are good, or open. If it was open the car would not really run.

My $.02


I am sticking with my original suggestion, get a fuel pressure gauge on it.....
Larry is suggesting that you put on a fuel pressure guage to see if you are having a fuel pump failure problem.

I had the same symptoms on my car that you are experiencing. It turned out to be a worn fuel pump problem. When I put a fuel pressure guage on the Fuel Filter and had a low presssure reading, this further supported the worned fuel pump theory.

After replacing the fuel pump with a new one, my CEL wen away and my car performed well. No more hesistation, acceleration suge, at 3500 RPM.
OK, that is it. The fuel pump gets replaced. I will keep everyone posted. If it isn't the fuel pump, at least I will have eliminated that variable. I can't thank Prime enough for all the suggestions you have give to me.
Ok, now if she is still stumbling a little, disconnect one 02 sensor, once you get the CEL, take the car out and see the result. If your stumble goes away, replace them.

Just closing the loop on this thread as OP, in case any info contained herein might help someone else.

I never did find out what was causing the problem(s). Replacing the OEM EMS with the AEM EMS unit solved the issue entirely with custom mapping on a dyno. I now also have two AEM digital A/F gauges mounted on the driver A pillar as well. I don't have the before/after dyno charts available, but the car picked up about 40 rwhp mid-range, and 5 rwhp up top. None of the other replaced parts or things done that were edited into the original post seemed to do much of anything. Obviously, I am enjoying an entirely different (as in very good) driving experience.

Since I was at least the third owner of this "rescued" NSX, I had no baseline to compare "normal" with what I found on a used car lot, but I tend to believe that I had been driving a "wounded" NSX for about seven years, all the while thinking that the surge in power that I felt at just over 4.5K rpm was the VTEC "kicking in", and I was just unlucky with O2 sensors.

Since the OEM EMS is no longer installed, maybe that had something to do with the issues the car was experiencing, but who knows.

I greatly appreciate everyone's suggestions in trying to help me solve this issue.
Last edited: