Low Engine Power before 4k RPM

[jsullyboy]

This was my first summer with my 91 NSX. It had 36.5k on it when I purchased it. I did about 500 miles thus far.

I have done a number of repairs over the summer. A lot of deferred maintenance and replaced items that have failed or were failing due to age.

The latest project is to address the low power at low RPMs.

The car seems to lack torque and power in low RPMs until it gets to about 4,000 RPM. Also it does not fire up right away. It seems to crank longer than I would expect it to need to start. (not extreme, but my S2000 fires up super fast) Since I never had the car from new and have never driven one other than my own I have no frame of reference. It could be that this is how it normally is but, it does not seem right to me.

Reading other posts on this subject I see that the usual suspects are:

- Fuel Filter
- Fuel resistor
- Fuel pump
- O2 Sensors

I looked through the service records I got with the car and I see no evidence that that fuel filter or fuel pump have ever been changed.

The first thing I did after purchasing the car was to change the water pump, timing belt, spark plugs. [I have done many other things mostly unrelated to the engine. Clutch master, slave, ABS flush and unstick solenoids, gauge cluster repair (almost caught on fire), AC restoration and interior restoration like windows, broken trim etc.]

My plan is to start with some injector cleaner first. Also I will order a fuel filter. I don't see anywhere in the records it has ever been done.

In other threads I see references to by-passing the fuel pump resistor. I could use some advice on this. I assume that by-passing means putting a jumper on the following connector?




Doing this means that the fuel pump will run full power at all times? Then if I notice a boost in power. It could be the fuel pump or the resistor itself? The fuel pump does not have a lot of milage on it. But its old and has probably sat for long periods in between use.

The oxygen sensors

If they were failing wouldn't the check engine light come on? Or could they be failing but just not enough to trigger the code? Reading though other threads I have seen a debate on if they would cause this issue or not. In some threads I have seen replacing them fixing a similarly described problem.

The suggested test I see for this is unplugging them and running without them. By doing this the engine would default to a base program. I assume unplugging them will trigger the check engine light. I have a few questions on this:

- What is the easiest way to reach them to unplug them? Under the car or can it be done from the top?
- Is putting a jumper on the service connector and counting the flashes on the dash the only way to read the codes from the ECU?

If I needed to replace these sensors how hard are they to remove? I have seen examples where they needed to be heated up with a torch and a sensor socket to remove them. Is this the norm? My car has never been driven in the elements. The bolts connecting the muffler are very rusted though. I can't see the manifolds. I had no problem getting the sensors off my S2000. But they are in a super easy to access location. Seems like replacing the back O2 requires removing the muffler to access it.

Fuel pump

I ordered the fuel pressure gauge from SOS. I will follow the procedure to test in the service manual.

Any advice or suggestions would be much appreciated!

 

[Old Guy]

The NSX normally starts quickly for a car of its vintage. Modern cars can typically sense TDC on any cylinder and can start the ignition timing calculations almost immediately; but, the NSX may need close to a complete crank rotation before it can start the ignition timing sequence so it is not instantaneous start like my 2021 Audi. My 2000 will typically fire up with less than 2 seconds of cranking.

The biggest contributor to slow fire up is loss of fuel pressure in the fuel rail while parked. This means that the pump prime cycle is not long enough to fill and bring the fuel rail up to specified pressure. The easy way to check for this is let the prime cycle complete without engaging the starter, turn the ignition off and then let the prime cycle complete a second time (or third time) to make sure the rail is pressurized and then engage the starter motor. If this gives you a fast start then you have a leakage problem somewhere that is causing loss of pressure in the fuel rail during shut down. This could be:
- leaky no return valve in the fuel pump
- fuel pressure regulator allowing leakage out the return line
- fuel injector(s) that leak
1 you can live with. 2 you can live with providing your test confirms that the regulator operating pressure is not compromised. 3 you should probably do something about because its a symptom of injector problems.

I would not bother with add-in-the-tank injector cleaners. The cows will come home before they have a material effect. That is more of a continuous maintenance thing if you are not using top tier gas. If the injectors are leaking or questionable I would remove them and send them out to somebody like RCInjector services for cleaning.

That location in the photo is where you would install a jumper (paper clip) to short out the fuel pump resistor and yes it will cause the pump to run at full power. If that resolves your problem that does point a finger at a problem with the fuel system. However, that could be a fuel filter problem or a fuel pump problem.

Running the car with the O2 sensors should put the car into open loop fuel control and should set an error code immediately. The engine normally runs in open loop until it has warmed up so this will only make a difference on a hot engine. If the heater fails in the O2 sensor that will trigger an immediate O2 sensor error code. However, it is possible for the O2 sensors to deteriorate to the point where they are causing fuel mixture issues; but, have not crossed the threshold where they qualify as completely dead. If you know what you are looking for and have access to a high quality voltmeter with high input impedance you can determine whether the O2 sensors are faulty. The best tool for doing this is an oscilloscope because the voltage swings up and down and on cheaper digital meters with slow up date rates it just looks like the numbers are jumping around. The service manual shows the connector location which is as I recall up top; however, your 1991 is different than my 2000. I have heard from authoritative sources that on the early cars with the cast iron manifolds it is the front sensor that can be a bugger to remove.

Right now your problem is too loosely defined to conclusively say that you have an O2 sensor problem and in fact that would not be high up my probably cause list. I would start with the fuel system first. If the car did not have regular use for a long period of time then the injectors might be a really good place to start. Low mileage; but, with regular use would be less of an issue because at least the injectors are being operated regularly. With clean injectors you can get an accurate fuel pressure test which might indicate a pump, filter or FPR problem. If the fuel system checks out 100% then you can go sniffing around the O2 sensors.

 

[jsullyboy]

I put in fuel injector cleaner and was doing the test described here. Got the check engine light to come on.



Going to attempt to read the code(s).

 

[jsullyboy]

The NSX normally starts quickly for a car of its vintage. Modern cars can typically sense TDC on any cylinder and can start the ignition timing calculations almost immediately; but, the NSX may need close to a complete crank rotation before it can start the ignition timing sequence so it is not instantaneous start like my 2021 Audi. My 2000 will typically fire up with less than 2 seconds of cranking.

The biggest contributor to slow fire up is loss of fuel pressure in the fuel rail while parked. This means that the pump prime cycle is not long enough to fill and bring the fuel rail up to specified pressure. The easy way to check for this is let the prime cycle complete without engaging the starter, turn the ignition off and then let the prime cycle complete a second time (or third time) to make sure the rail is pressurized and then engage the starter motor. If this gives you a fast start then you have a leakage problem somewhere that is causing loss of pressure in the fuel rail during shut down. This could be:
- leaky no return valve in the fuel pump
- fuel pressure regulator allowing leakage out the return line
- fuel injector(s) that leak
1 you can live with. 2 you can live with providing your test confirms that the regulator operating pressure is not compromised. 3 you should probably do something about because its a symptom of injector problems.

I would not bother with add-in-the-tank injector cleaners. The cows will come home before they have a material effect. That is more of a continuous maintenance thing if you are not using top tier gas. If the injectors are leaking or questionable I would remove them and send them out to somebody like RCInjector services for cleaning.

That location in the photo is where you would install a jumper (paper clip) to short out the fuel pump resistor and yes it will cause the pump to run at full power. If that resolves your problem that does point a finger at a problem with the fuel system. However, that could be a fuel filter problem or a fuel pump problem.

Running the car with the O2 sensors should put the car into open loop fuel control and should set an error code immediately. The engine normally runs in open loop until it has warmed up so this will only make a difference on a hot engine. If the heater fails in the O2 sensor that will trigger an immediate O2 sensor error code. However, it is possible for the O2 sensors to deteriorate to the point where they are causing fuel mixture issues; but, have not crossed the threshold where they qualify as completely dead. If you know what you are looking for and have access to a high quality voltmeter with high input impedance you can determine whether the O2 sensors are faulty. The best tool for doing this is an oscilloscope because the voltage swings up and down and on cheaper digital meters with slow up date rates it just looks like the numbers are jumping around. The service manual shows the connector location which is as I recall up top; however, your 1991 is different than my 2000. I have heard from authoritative sources that on the early cars with the cast iron manifolds it is the front sensor that can be a bugger to remove.

Right now your problem is too loosely defined to conclusively say that you have an O2 sensor problem and in fact that would not be high up my probably cause list. I would start with the fuel system first. If the car did not have regular use for a long period of time then the injectors might be a really good place to start. Low mileage; but, with regular use would be less of an issue because at least the injectors are being operated regularly. With clean injectors you can get an accurate fuel pressure test which might indicate a pump, filter or FPR problem. If the fuel system checks out 100% then you can go sniffing around the O2 sensors.

Thank you for your reply "Old Guy". After putting in the injection cleaner I did a test drive. In 3rd gear from 2500rpm I floored it. Very slow response but the check engine light did come on. Once the car hits 4500rpm its like turbo kicks in there is such a difference.

The accelerator pedal increases resistance the further you press it. Is this normal? Sorry for these simple questions but I never drove the car factory new.



I jumped the check service connector. When fishing it out I noticed the grey connector that was not plugged into anything. Another newbie question. Is this normal? Looked around and could not see a place to plug it into.

With the service connector jumped I see 2 quick flashes. Which should be code 2 if I am interpreting this correctly.


View attachment Check Engine Code.mp4

According to the diagnostics it seems this is the rear oxygen sensor.



Please let me know if I am interpreting this correctly.

Thanks!

 

[TANTO]

I can only comment on the "simple" questions you asked. The pedal pressure on my car stays consistent throughout it's range. Since i have no rear O2 sensor on my 92 (header install) i would be very surprised if it's malfunction caused a big lapse in torque. I'm sure Old Guy and others will comment on your code (you got a check engine light before your test?). IMO you really need to find a ride in another NSX for a baseline. Regarding torque - in the 90's when i drove a Tercel to work the torque of the X felt pretty good, but daily drivers today have so much more torque. I equate our ride to a high end 600cc sport bike where for a short "sortie", you just always keep the R's up a bit to stay in the torque band. Similar to the Ferrari 348 of the time. My V8 torque monger friends who have driven my car are not impressed with the torque. Their loss of course as it's my contention that if i let them drive it for a month they wouldn't want to give it back! So perhaps a slight difference in driving style will keep that smile on your face! The engine is certainly built for it.

 

[jsullyboy]

I can only comment on the "simple" questions you asked. The pedal pressure on my car stays consistent throughout it's range. Since i have no rear O2 sensor on my 92 (header install) i would be very surprised if it's malfunction caused a big lapse in torque. I'm sure Old Guy and others will comment on your code (you got a check engine light before your test?). IMO you really need to find a ride in another NSX for a baseline. Regarding torque - in the 90's when i drove a Tercel to work the torque of the X felt pretty good, but daily drivers today have so much more torque. I equate our ride to a high end 600cc sport bike where for a short "sortie", you just always keep the R's up a bit to stay in the torque band. Similar to the Ferrari 348 of the time. My V8 torque monger friends who have driven my car are not impressed with the torque. Their loss of course as it's my contention that if i let them drive it for a month they wouldn't want to give it back! So perhaps a slight difference in driving style will keep that smile on your face! The engine is certainly built for it.

Hi Tanto, thanks for the response. I will look into if there is a way to lubricate the throttle mechanisms. It definitely gets stiff when pushing it down.

It would be great to have the chance to test another car and compare the difference. I know what you mean by staying in the power band. I also have a 2001 AP1 S2000. It only has 153lb-ft of torque @7500rpm. It feels way more powerful right now than the NSX especially when I drive them back to back. I know the gearing is totally different but it seems like too much of a gap. The NSX should have 210lb-ft @5300rpm.

I sold a 2021 Mach 1 to make room for the NSX. Naturally it was a torque monster and super fast. I am glad I had a chance to experience owning it. What I discovered is that is not what I was looking for. I feel way more connected to the NSX when driving it. There are no places where I could use the full potential of the Mach 1 unless I went to the track. Which I did a few times. I am at the stage in my life where it is more about the back road drives and carving a corner than ripping the tires off.

I don't expect the NSX to be the Mach 1. My goal is to get it working correctly like it would have in 1991.

I think this is probably a combination of problems. Like Old Guy said the injectors are likely not clean. The car did sit for long periods before I owned it. When driving the NSX it really feels like it is being starved of fuel in those low RPMs. Over the winter I will remove the injectors and get them cleaned and serviced.

I have babied the NSX except for the odd pull here and there when I was higher in the RPM range. So it was not as noticeable when I first got it. It seems to be getting worse. I did about a 2 hour trip all highway at about 130-140km/70-80mph for most of the trip with the cruise set. Its after this trip that I really noticed a big difference.

I am going to test by-passing the fuel resistor. See what happens. Also Need to take my S2000 out and enjoy it today. I am running out of driving days for this year.

I intend to keep the NSX for a long time. It was a long time goal to own one. Things like the fuel filter, sensors, fuel pump I am going to have to replace over time anyway. Over the summer I have done a lot. I didn't expect to being so much work all at once.

I really appreciate all the suggestions and advice of the community here. I am not an expert by any means on auto mechanics. But, I want to learn. Getting the NSX back to its original glory is part of the fun of owning to for me.

 

[Old Guy]

I have a 2000 with a drive by wire throttle so completely different system and can't comment on the early system.

Yes, you have correctly diagnosed the error code as 2 which is a rear O2 sensor issue. The error code 2 is low voltage on the O2 sensor. I expect that grey connector might be for the light inside the glove box or the valet switch for the trunk (does your trunk open switch in the door work?).

The diagnosis for error code 2 in the 1991 service manual just forwards you to the diagnosis for code 44 in the service manual so I am not exactly sure what triggers a 2 versus a 44. The 1997+ service manual has a little more detail in it and specifies that you should check the sensor voltage during wide open throttle 4th gear acceleration. This is easy to do with an OBDII car because you connect a scan tool which gives you direct access to the O2 sensor voltages. Not so easy on a pre OBDII car. You could run some wires back probed into the sensor connector to a voltmeter in the cab and have someone watch the voltmeter while you do the acceleration test. If the voltages stays at 0.1 volt or lower then you can have
- short in the wiring
- O2 sensor has crapped out and is reading lean all the time
- you really have a lean fuel mixture problem

The service manual says that the first thing to check is confirm that the fuel system is OK, then you look at the wiring or a dead sensor. Given that things get better when the ECU by-passes the fuel pump resistor I am thinking its a fuel system problem rather than an O2 sensor problem. You can confirm this with the paper clip short at the fuel pump resistor connector. One thing to check is examine the actual fuel pump resistor connector and the resistor itself. On older cars the heat can cause these to deteriorate which results in bad connections and even lower voltage. You can measure the resistance and if you search on Prime you can find what the measured value should be (the SM does not provide an actual test value for the resistor). Your car is low miles so I don't expect this is the problem; but, good to rule out.

 

[docjohn]

Are you in touch with the nsxcc ie the nsx club of Canada, there are many helpful owners in your neck o woods, pm Bram..

 

[asylum]

Bypass fuel pump resistor and try. The fact that the problem is only under 4k is really pointing toward the voltage switch of the fuel pump.

 

[jsullyboy]

Thanks everyone for the replies. I will test the resistor and let you know what happens.

 

[jsullyboy]

I did a few tests with the fuel resistor by-pass and without. I was not able to reproduce the issue. The car is up to temperature but it was evening and cooler at 19c or 66.2f. I think it is more noticeable after about 30-40min of driving, which I did not have time for tonight. Here is what is looks like. Let me know if this seems like normal acceleration for a 1991 3.0L. I am in 3rd starting at 3k rpm. [i know if we were trying for acceleration one would downshift to 2nd] I did not get a check engine light this time. So that is a good thing.

View attachment Fuel Resistor Test.mp4
No difference that I can tell. On the weekend it was pulling noticeably harder at ~4500rpm. All the examples in the video are fairly consistent.

 

[jsullyboy]

I have a 2000 with a drive by wire throttle so completely different system and can't comment on the early system.

Yes, you have correctly diagnosed the error code as 2 which is a rear O2 sensor issue. The error code 2 is low voltage on the O2 sensor. I expect that grey connector might be for the light inside the glove box or the valet switch for the trunk (does your trunk open switch in the door work?).

The diagnosis for error code 2 in the 1991 service manual just forwards you to the diagnosis for code 44 in the service manual so I am not exactly sure what triggers a 2 versus a 44. The 1997+ service manual has a little more detail in it and specifies that you should check the sensor voltage during wide open throttle 4th gear acceleration. This is easy to do with an OBDII car because you connect a scan tool which gives you direct access to the O2 sensor voltages. Not so easy on a pre OBDII car. You could run some wires back probed into the sensor connector to a voltmeter in the cab and have someone watch the voltmeter while you do the acceleration test. If the voltages stays at 0.1 volt or lower then you can have
- short in the wiring
- O2 sensor has crapped out and is reading lean all the time
- you really have a lean fuel mixture problem

The service manual says that the first thing to check is confirm that the fuel system is OK, then you look at the wiring or a dead sensor. Given that things get better when the ECU by-passes the fuel pump resistor I am thinking its a fuel system problem rather than an O2 sensor problem. You can confirm this with the paper clip short at the fuel pump resistor connector. One thing to check is examine the actual fuel pump resistor connector and the resistor itself. On older cars the heat can cause these to deteriorate which results in bad connections and even lower voltage. You can measure the resistance and if you search on Prime you can find what the measured value should be (the SM does not provide an actual test value for the resistor). Your car is low miles so I don't expect this is the problem; but, good to rule out.

I never noticed there was a light in the glove box. I see the main switch to disable the trunk button. The trunk works so it must be the light. Great suggestion I will hunt around for that.

 

[jsullyboy]

I am fairly confident that its the fuel pump. I am trying to get the gas as low as possible to make it easier to drop the tank. Is there no low fuel warning light on the 1991 model year?

 

[Chris_Lum]

Its pretty easy to drain the tank (just will make a bit of a mess if you don’t have a container big enough to hold all the rest of your fuel and you have to close up the drain bolt prematurely—ask me how I know). In my case, jumping the fuel pump resistor didn’t improve my hesitation and so I know the fuel pump wasn’t likely the issue. I decided to change my pump out and had no improvement. Wasn’t exactly a fun job either. Not terrible, but not fun, especially if you decide to change out all the small diameter hoses (the unique clips they use suck).

 

[jsullyboy]

Good point. I have a couple of empty gas cans I could drain it into. I plan to replace the lines while I am in there too. Will probably be some colourful language emanating from my garage Saturday. What ended up as the fix for your car?

Its pretty easy to drain the tank (just will make a bit of a mess if you don’t have a container big enough to hold all the rest of your fuel and you have to close up the drain bolt prematurely—ask me how I know). In my case, jumping the fuel pump resistor didn’t improve my hesitation and so I know the fuel pump wasn’t likely the issue. I decided to change my pump out and had no improvement. Wasn’t exactly a fun job either. Not terrible, but not fun, especially if you decide to change out all the small diameter hoses (the unique clips they use suck).

 

[Chris_Lum]

Good point. I have a couple of empty gas cans I could drain it into. I plan to replace the lines while I am in there too. Will probably be some colourful language emanating from my garage Saturday. What ended up as the fix for your car?

I wish I could say I fixed mine but can’t. So far I’ve cleaned injectors with RC Engineering, checked spark plugs, installed new oem coil packs, new oem fuel filter, new oem fuel pump, new Denso 02 sensors and new gas cap. I have a new oem fpr ready to go on. I’m just throwing parts at it because I’m lazy to put a fuel pressure gauge on it though. I’m gonna go through the emissions equipment next.