7,800rpm rev limit?

I have put around 1000 miles on my 91 NSX since I bought it earlier this year and I noticed that the rev limiter kicks in systematically at 7,800rpm (could be 7,900 but surely never touches 8,000). The engine speed always seemed more than fast enough to my ears so I never bothered to enquire, I even thought it was the tach that was reading engine speed low, but now I'm wondering whether this could indicate an issue with the car, although I can spot no performance issue compared to other NSXs I drove. And yes, I never use revs before the engine is comfortably warmed up, several miles after coolant temperature has settled to its operating mark. I guess it could be the ECU that was programmed to cut at a lower rpm but tuning usually involves moving the limit upward?

Any thoughts appreciated! The car is now in storage, if the ECU needs to be looked at, it's the best time to do it. Thanks!

is it a standard ecu? does it have an AM chip on it? is your car an auto or manual gearbox.
the engine doesn't really make any power above that so you're not missing out on anything

The service manual says that the OEM ECU 0n the 1991 applies fuel cut at 8300 RPM. Fuel cut implies a hard cut as opposed to a soft cut that some systems do such as retarding ignition timing prior to doing a hard fuel cut. I have never seen any mention in the service manual of the ECU retarding timing prior to the fuel cut.

If you have the stock ECU, is what you are experiencing a hard cut? If it is, and assuming that somebody has not re chipped the ECU, then I would put a higher probability on your tach being out of calibration than there being a problem with the setting of the fuel cut. The fuel cut is implemented in firmware. Unless somebody rewrote the settings, that kind of setting doesn't change with age. There is not really anything to 'look at' or check in the ECU, it pretty much works or its time for a replacement.

Unfortunately the 1991 pre OBDII ECU doesn't give you access to the engine parameters to check engine speed. There is a 'service loop' on the igniter module that is used to drive an inductive pick up light for checking the ignition timing. With a little creativity you could hook a magnetic pick up on there and use it to drive a temporary tach to compare the calibration of your dash tach or just hook up a tach which accepts the same pulse rate (100 RPM per 300 pulses/min) as the dash tach.

Old Guy said:

The service manual says that the OEM ECU 0n the 1991 applies fuel cut at 8300 RPM.

Click to expand...

I fully agree but despite that, the rev limiter in my 1991 kicks in just after an indicated 8000 rpm on the tachometer - way below 8300.

Maybe an external rev counter could be hooked up to the engine while driving, running the pickup wire back to the engine compartment and using the cigarette lighter as a power source. Alternatively, the engine speed could be measured while the car is on a dyno. If the facility can't capture the signal from the pickup loop in the engine compartment, they could pop one of the coil packs off of a spark plug, connect the two with a short jumper wire, and capture a signal from that.

NSX Prime member sr5guy graciously posted a "definition" of the data on the OBD1 PGM-FI's chip - see here. He said the next version of that definition will include settings for the rev limiter and the VVIS changeover point. I look forward to the day. It will be interesting to see where the factory really set the rev limiter. Increasing it to 8300 rpm should be OK since that's the limit referred to in the Service Manual. Or at least 8000.

Manual or auto?

My car has the manual transmission and while I'm not sure I understand the distinction between hard and soft cut, I can only say it is the rev limiter that kicks in with fast paced pulses of power interruption preventing the RPMs from going any higher. I only assume this is a hard cut, as the car is designed to do. I drove other NSXs and they behaved the same way except at a higher indicated RPM, though my ear couldn't pick up the difference in pitch that could tell whether the motors were revving higher than mine when the limiter kicks in.

I remember reading somewhere that the ECU can limit the revs lower than redline under certain conditions but since my car hits the limiter at the same RPM each and every time, I'm thinking it's either the tachometer that is off or the ECU that was programmed to limit revs 500 RPM lower. Regarding the latter, why would anyone do this? I have pretty extensive documentation about the car and I can't see any evidence of a supercharger or other performance enhancing devices being installed, but are there any such mods that require lowering the rev limit?

This probably is completely unrelated but I can tell someone's removed the rear bulkhead panels before I went in there to take out my faulty fan control unit, could be to reach the engine control module. To rule out the ECM hypothesis, could I simply swap the module with another one and see whether there is a change?

Maybe I'm being fussy here, the car runs great, but being such a great machine I'm fanatical about it running optimally.

A fuel cut typically shuts of fuel to the engine which results in the engine shutting down - quickly. The fuel cut software typically has an RPM dead band, say 200 RPM or perhaps more. This means that fuel cut could occur at 8300 RPM; but, would resume when the RPM drops to 8100 RPM. A small dead band can result in repeated rapid on/off engine operation if the driver leaves there foot planted on the gas pedal. A larger dead band will result is slower on/off; but, more dramatic pulsations. A soft cut typically involves retarding the ignition like crazy just before the RPM limit so that there is a drastic loss of power which hopefully prevents the RPM limit from being reached; but, the engine does not actually shut down. Soft cuts usually have a hard cut as a back up. What you describe definitely sounds like a hard cut to the fuel.

Since your car is a 1991, chances are that the main FI relay has been replaced which requires access behind those panels.

The ECU has limp modes that occur when sensors or other things fail. These would typically limit RPM drastically. Anything that would trigger a limp mode would almost certainly trigger an error code.

If the knock sensors were detecting detonation they would typically retard the ignition which would result in a loss of power. That would not be a hard cut where the engine shut down. I don't know how much the NSX ECU can retard timing and what the effective power loss would be. I would expect that if the ECU was retarding timing because of detonation that you would have been aware of the detonation.

You can switch out the ECU if you can find someone who is prepared to donate theirs for an experiment or you are prepared to spend the $ for a new one. I would check my tach calibration first to confirm that the problem is real. Anytime you go around changing things that do not need to be changed you run the risk of damaging things (wiring and terminal plugs) and making things worse.

I had a similar situation during the first couple of winters I had my X. After 10-15 miles of driving the temperature gauge would register in the middle, but the rev limiter would kick in at lower RPMs than it was supposed to. I found that if I kept driving that symptom would eventually go away and the limiter would kick in at 8K, and the rev limit gradually went up with time. I'm wondering if the engine isn't really up to temp until the oil has reached operating temperature, which would take longer than the water temp. Not sure if there are any engine temp sensors that would feed this info to the ECU? I also found that when I went to 5W-30 from the original 10W-30 the issue has disappeared, which could make sense with the opil temp theory. But you might try driving it longer than you think you need to and see if that makes a difference. In any case, I could be way off with my analysis :->

After a long drive on the highway, does the water temp remain in the middle and is the rev limiter still at 7,800rpm?

Could be a bad temperature sending unit preventing the engine from reaching its proper redline..

When I first saw this thread, I thought OP was using the AT ECU because 7,800rpm fuel cut is spot on.
However, unless it was converted from AT, the OP is running the MT ECU.

Temperature sender unit (temp gauge) is nothing to do with the EFI control.
There is another one called TW sensor that is used as the input for the ECU.
In fact, on our NSX, there are 3 coolant temperature related sensors.
The sender unit for the temp gauge, TW sensor for the ECU and rad fan sensor for the cooling fan control.
If TW is low (IIRC, below 60degc), VTEC will be disabled and you will get fuel cut at 7,000rpm.

For your reference, offset of 500rpm on the rpm gauge at high rpm region is nothing new.
Seen it on many NSX.
Please note that I’m not saying this is the reason for your phenomenon but one possibility.

If your gauge calibration is not accurate enough, you probably have about 100rpm offset even at much lower rpm.
If you have oscilloscope, it's easy but even without it, you can still measure the frequency using the multimeter with frequency counter.
There is small connector at the right edge of the engine bay for rpm pickup.
It's 6cyl so 3 TDC/rpm.
If you hold 2,000rpm on the dash gauge, you should see 100Hz but you are probably at around 105Hz.

The higher the rpm, the larger the offset and after years of usage, the gauge could show lower as well as higher than the actual rpm.

Kaz

Kaz-kzukNA1 said:

When I first saw this thread, I thought OP was using the AT ECU because 7,800rpm fuel cut is spot on.
However, unless it was converted from AT, the OP is running the MT ECU.

Kaz

Click to expand...

Now THAT'S interesting! I thought the AT fuel cut was closer to 7,000rpm. Would an AT ECU even work in a MT car?

bruckner said:

Now THAT'S interesting! I thought the AT fuel cut was closer to 7,000rpm. Would an AT ECU even work in a MT car?

Click to expand...

Certainly, and a AT ecu can be modified to MT.
The RPM gauge not being accurate like Kaz pointed out is a big factor, non of the gauges are there to be very accurate, they just give you a idea of what's going on.
Perhaps the most accurate would be the voltage, the speedo is calibrated but we all know that has a margin.
Oil pressure is indicative too it's going to be accurate if your cruising but not whilst going up and down the revs.


A OBD-II NSX would give you the easiest comparison as you can check the sensor feeds digitally.

That said these margins are present in most cars if not all, if for example I'm working in a new Maserati 250-500 rpm off set isn't odd, it's so common I can calibrate and program each needle in the dash.

I must add that I've seen the RPM limit change quite a bit prior vs post maintenance work being done, often prior the car may have a error present in ether current or history.
Such as a 02 sensor or other fuel/ignition related sensors, and after servicing the cars ecu is rest.
Also something we see with new customers is that timing belts especially rear exhaust weren't set correctly this obviously directly effects performance and rpm capabilities.


Has your cluster been dissembled before? bending or altering the needle or it's stops is very easy they are super-delicate

A.S. Motorsport said:

Has your cluster been dissembled before? bending or altering the needle or it's stops is very easy they are super-delicate

Click to expand...

The car has been in my possession for less than a year, I can't see any evidence of the cluster ever being disassembled.

Fuel cut at 7,800rpm is consistent, even after (fairly) long drives, in any case I never use revs until 10-15 minutes of driving.

My car is now in storage but I do have access to it. In the unlikely case that my ECU is a modified AT unit, would it show on the case? I have pretty extensive maintenance history on the car but it doesn't go all the way to 1991, and the car has done 90,000 miles since new, I guess anything is possible!

For kicks, come the Spring 2017, I may put the car on a dyno and find a way to monitor RPM, this should put the matter to rest?

From ECU point of view, it doesn’t matter whether it’s AT or MT. You don’t even need the real engine to run the ECU. All you need is the power supply, TDC/CRK timing sensor signal, required multiple sensor inputs, feedback loop for certain load, etc and the ECU will run without triggering the CEL.
That’s how the ECU software/hardware engineers create/test the system using the model or simulator.

So, yes you can run AT spec ECU with MT gbox.
Several owners converted their AT NSX into MT without replacing the original ECU.

You can tell it by the parts no. written on the ECU housing (unless someone replaced the ROM).

If using AT spec ECU on MT model, you need to tell the ECU that you are in N/P range or not.
Otherwise the idle control won't work properly and unless you tell that you are out of the N/P range, you won't get VTEC and the rev limit will be reduced to 7,000rpm as the ECU thinks that you are just opening the TH in neutral.

Kaz