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

Back to NA after 10 years of FI

Thanks John!
I went for a 100 km shake down ride this afternoon and it felt like I had just bought a new car?
Quite amazing to drive an OEM NSX again!
The engine is so quiet (too much maybe), no fumes whatsoever and a smooth power delivery that compells you to rev the engine.
The OEM ECU also does marvels to keep the AFR at an indicated 14.8 on my AEM AFR display at all times.
For some reason the car felt downscaled in size with less weight and an improved steering feedback?
Could it be due to me raising the front end by 5mm or so?
Long live the NSX!

The original ECU is a lot more capable than people give it credit for :wink:
 
The original ECU is a lot more capable than people give it credit for :wink:
I've seen your thread on using the OEM ECU with ITB's.
What about adapting it to E85 fuel?
Given the current prices in France it would be most interesting!
 
You could certainly tune the car to run ethanol if you made the proper fuel system upgrades. Ethanol has a much lower stoichiometric AFR so you'd probably need bigger injectors as well since the stock ones are very close to maxed out.

That said, true flex fuel support is probably not a realistic possibility so you'd be locked into whatever fuel mix you tuned for. You could reflash different tunes fairly easily or even install a switch to change between regular gasoline and ethanol tunes but I don't think there's enough capacity in this ECU to support on the fly adjustments for fuel mix. To really do that properly you need at least 2 full sets of fuel and timing maps, then interpolate based on the fuel mixture and there's just not enough ROM space for that and the code paths that access the fuel and timing tables are hard coded in the MCU.
 
Last edited:
You would need to flow ~50% more fuel to run the same power with E85. The OEM injectors definitely wouldn't be capable of this as they are pretty marginally sized just for the OEM power. The RDX injectors would probably be ok, but you'd be back to roughly the same margin as the OEM injectors on gasoline (~90%+ duty cycle). I doubt the OEM fuel pump would be a good idea, and you'd have to upgrade everything in the fuel supply to be ethanol compatible.
[MENTION=33247]MotorMouth93[/MENTION] You could run a "2-timer" board just like on other chipped ecus, or I believe you could do this with your programming knowledge with a stock ECU. The OEM ECU chip has both auto and manual maps saved and one pin is given a 5V signal to determine which one is used. You could run a gasoline setup on the manual location of the chip and an E85 map on the auto side and switch between them with a 5V signal. You'd only have the 2 options as you mentioned, no fancy blending with a fuel sensor like what many ECUs can do now. I thought about doing this with a "standard" and a "power" tune, but this would actually be a much better use of the switching ability.

Check out this post by SR5guy regarding the Legend ECU (which is very similar to ours):

https://www.acura-legend.com/threads/ecu-chip-info.79837/page-43#post-923968

See post #854
 
You would need to flow ~50% more fuel to run the same power with E85. The OEM injectors definitely wouldn't be capable of this as they are pretty marginally sized just for the OEM power. The RDX injectors would probably be ok, but you'd be back to roughly the same margin as the OEM injectors on gasoline (~90%+ duty cycle). I doubt the OEM fuel pump would be a good idea, and you'd have to upgrade everything in the fuel supply to be ethanol compatible.
[MENTION=33247]MotorMouth93[/MENTION] You could run a "2-timer" board just like on other chipped ecus, or I believe you could do this with your programming knowledge with a stock ECU. The OEM ECU chip has both auto and manual maps saved and one pin is given a 5V signal to determine which one is used. You could run a gasoline setup on the manual location of the chip and an E85 map on the auto side and switch between them with a 5V signal. You'd only have the 2 options as you mentioned, no fancy blending with a fuel sensor like what many ECUs can do now. I thought about doing this with a "standard" and a "power" tune, but this would actually be a much better use of the switching ability.

Check out this post by SR5guy regarding the Legend ECU (which is very similar to ours):

https://www.acura-legend.com/threads/ecu-chip-info.79837/page-43#post-923968

See post #854
My question is really rethorical as I moved back to OEM and like it as is (for the moment...).
However, the two map option that is standard in the ECU could be a solution forward?
Suzuki on their GSXR 1000 motorbike did the same to adapt to the the old french market regulation of 100 hp maximum.
Simply bypassing a resistor let the engine develop it's full power.
One could imagine an NSX "racing" setup with 12/1 pistons, upsized injectors and fuel pump.
I wonder what the theoretical power delivery would be?
 
There are a number of engine horsepower calculation algorithms around. The best ones you have to pay for and require a fair amount of input data and user knowledge if you don't want 'garbage in - garbage out'. The bgsoflex.com calculator is on-line, free; but, incredibly simplified and uses a lot of average factors to calculate horsepower. For the C30B with 2.98 litres, 10.2 CR and peak hp at 7300 RPM the calculator estimates 246 hp. The average Ve values or BMEPs in the calculator are probably not reflective of the C30A. Increase the static CR to 12:1 and the calculator yields 290 hp. Increase the CR to 12:1 and raise the engine RPM to 8000 RPM and the calculator yields 318 hp. The presumption in the calculator is that there is no change in volumetric efficiency between 7300 and 8000 RPM or any other factors which would limit the power increase (detonation, manifold IAT, ....). The absolute numbers generated by the calculator are clearly not valid. The relative numbers generated by the calculator are probably representative in terms of identifying the potential for increase in power. I am a little suspicious of the hp increase associated with the 10.2:1 to 12:1 change in compression ratio. It is a little more than I expected. The horsepower increase associated with the RPM change is straight forward - purely a function of RPM reflecting no drop in Ve.

The more sophisticated calculators will do a better job; but, you need more input data to get a better result.
 
There are a number of engine horsepower calculation algorithms around. The best ones you have to pay for and require a fair amount of input data and user knowledge if you don't want 'garbage in - garbage out'. The bgsoflex.com calculator is on-line, free; but, incredibly simplified and uses a lot of average factors to calculate horsepower. For the C30B with 2.98 litres, 10.2 CR and peak hp at 7300 RPM the calculator estimates 246 hp. The average Ve values or BMEPs in the calculator are probably not reflective of the C30A. Increase the static CR to 12:1 and the calculator yields 290 hp. Increase the CR to 12:1 and raise the engine RPM to 8000 RPM and the calculator yields 318 hp. The presumption in the calculator is that there is no change in volumetric efficiency between 7300 and 8000 RPM or any other factors which would limit the power increase (detonation, manifold IAT, ....). The absolute numbers generated by the calculator are clearly not valid. The relative numbers generated by the calculator are probably representative in terms of identifying the potential for increase in power. I am a little suspicious of the hp increase associated with the 10.2:1 to 12:1 change in compression ratio. It is a little more than I expected. The horsepower increase associated with the RPM change is straight forward - purely a function of RPM reflecting no drop in Ve.

The more sophisticated calculators will do a better job; but, you need more input data to get a better result.
I'm not too surprised by the gain due to CR increase.
My KTM 1290 Super Duke R with 1301 cc develops 180 hp at 9500 rpm with a CR of 13.5/1!
!
 
The KTM horsepower may be more reflective of the elevated RPM that the Super Duke operates at. As a comparison, the Honda F20C develops 120 hp per l at 8300 RPM. The KTM develops 138 hp per l @ 9500 RPM. Assuming a flat torque curve the KTM would be right at 120 hp per l if the RPM is adjusted to 8300 RPM. The F20C provides that number with an 11.5:1 CR.

Compression ratio helps; but, not as much as cranking up the RPM (assuming the head and induction system can keep up).

I am a little surprised that KTM can manage the 13.5:1 CR without direct injection. Perhaps their twin plug head is what they use to achieve fast combustion so that end gasses don't have time to heat up and detonate. If you are going to try 12:1 on a C30A I think you would at least want to have a set of cams with longer duration so that you have a lower dynamic compression ratio and also probably ditch the VTEC - or be on the look out for really high octane gasoline.
 
Last edited:
I took the car out for about 200 km so far and I'm very pleased with the result overall.
In fact having a turbo obliterates the intrinsic qualities of the car.
The NSX is much easier to drive fast and in fact it pushes you to do so.
With the turbo setup you know it can go much faster on a given road but you don't feel the need to do so while in NA every bend challenges you to get closer to the limit.
The only negative is that you need to rev the engine harder and the sound with the OEM exhaust is disappointing.
I also noticed that the TCS (Traction Control) is inoperative despite having the OEM ECU back in the car?
Could it be due to the NA2 ABS?
Maybe [MENTION=25737]Kaz-kzukNA1[/MENTION] has an answer?
In the meantime, I will check that the 18 pin connector mentionned in the instructions booklet is well connected!
 
Last edited:
The TCS should work without any issues. If the ABS is okay, but the TCS do not, try to clean the ABS sensors and the teeth gear on the hub. The ABS use a kind of "raw" signal from the sensors while the TCS use a "modified" signal which is converted from the ECU. If the raw signal is not so nice, it is okay for the ABS but the conversion of the signal is not that good and the TCS dont work. I had such an issue a few years back. We tried all and after cleaning the sensors it was fine.
 
Thanks for the mention, CB72.
Received the notification from the server but with lots of addons and ad blocker on my Firefox setup, not easy browsing through this site…..

Using chrome temporarily but lots of Ads and keeps me logging out everytime when I just preview my reply.....

Anyway, hope I can post this eventually.….


If you had turbo setup, most likely you have disabled the TCS.
I presume your NSX is non-DBW so the TCS is done through the TCS controller and not the ECU.

When you turn the IGSW into P2 ON position, does the [ TCS ] light at the gauge cluster stay On?

If it stays Off all the time, most likely, you have disabled it.
Check the multiple connectors at the TCS controller behind the left side seat.

Like on my NSX, for the non-DBW model, the easiest way to disable TCS is to unplug all connectors from the TCS controller.
This will result in CEL #36 all the time but it won't trigger the CEL on the dash because the ECU must see 2 consecutive signal failures within the same drive cycle and since the TCS controller is disconnected, it can be only detected once per cycle.

As you already mentioned, check the 18pin white connector on the ABS upgrade loom if you are going to enable TCS.
You don't even need to connect it if TCS was disabled but if going back to TCS, you must connect it.

Personally, I hate TCS on non-DBW models….


Kaz
 
Thanks for the mention, CB72.
Received the notification from the server but with lots of addons and ad blocker on my Firefox setup, not easy browsing through this site…..

Using chrome temporarily but lots of Ads and keeps me logging out everytime when I just preview my reply.....

Anyway, hope I can post this eventually.….


If you had turbo setup, most likely you have disabled the TCS.
I presume your NSX is non-DBW so the TCS is done through the TCS controller and not the ECU.

When you turn the IGSW into P2 ON position, does the [ TCS ] light at the gauge cluster stay On?

If it stays Off all the time, most likely, you have disabled it.
Check the multiple connectors at the TCS controller behind the left side seat.

Like on my NSX, for the non-DBW model, the easiest way to disable TCS is to unplug all connectors from the TCS controller.
This will result in CEL #36 all the time but it won't trigger the CEL on the dash because the ECU must see 2 consecutive signal failures within the same drive cycle and since the TCS controller is disconnected, it can be only detected once per cycle.

As you already mentioned, check the 18pin white connector on the ABS upgrade loom if you are going to enable TCS.
You don't even need to connect it if TCS was disabled but if going back to TCS, you must connect it.

Personally, I hate TCS on non-DBW models….


Kaz
Thanks Kaz, I will try that as soon as it stops freezing over here.
BTW, I agree with you; the TCS is almost useless and I had to switch it off as soon as I got to any track meeting.
The main reason for getting it to work again is to get the car back to OEM as far as possible...
However as I was studying the NSX manual, I saw that the TCS unit works through a throttle actuator.
Maybe we could use this throttle actuator to act as a throttle blipper to aid downshifts?
 
The TCS should work without any issues. If the ABS is okay, but the TCS do not, try to clean the ABS sensors and the teeth gear on the hub. The ABS use a kind of "raw" signal from the sensors while the TCS use a "modified" signal which is converted from the ECU. If the raw signal is not so nice, it is okay for the ABS but the conversion of the signal is not that good and the TCS dont work. I had such an issue a few years back. We tried all and after cleaning the sensors it was fine.

Thanks Christian, I'll try that after checking the connection of the ALB 18 pin connector as suggested by Kaz.
 
Thanks Kaz, I will try that as soon as it stops freezing over here.
BTW, I agree with you; the TCS is almost useless and I had to switch it off as soon as I got to any track meeting.
The main reason for getting it to work again is to get the car back to OEM as far as possible...
However as I was studying the NSX manual, I saw that the TCS unit works through a throttle actuator.
Maybe we could use this throttle actuator to act as a throttle blipper to aid downshifts?
The TCS has caused me more near crashes than it's saved me from one. Whomever buys my car, I'll recommend the same and just hand him the TCS ECU in a box. It's very easy to unplug and remove from the rear bulkhead and leaves behind no nagging lights or beeps. $0.02.
 
Thanks Kaz, I will try that as soon as it stops freezing over here.
BTW, I agree with you; the TCS is almost useless and I had to switch it off as soon as I got to any track meeting.
The main reason for getting it to work again is to get the car back to OEM as far as possible...
However as I was studying the NSX manual, I saw that the TCS unit works through a throttle actuator.
Maybe we could use this throttle actuator to act as a throttle blipper to aid downshifts?

The design of the actuator is such that it can force the throttle closed when its open but it can't actually open the throttle.

I'm inclined to agree with @RYU, the early TCS system seems to create hazardous situations as often as it prevents them.
 
Last edited:
The TCS "unsafe" aspects has been covered many times on prime but basically it takes away the predictability. For example, often when TCS is active, as the car enters a slide, at some point a decent driver will correct the slide, however as your correcting TCS will step in - it's so slow to respond and is not a brake based TCS, it cuts throttle when in some situations you need the throttle to get you out of trouble. TCS in the car is a liability. I'll happily wrap it in a pretty little box with a bow on it and a note saying NEVER INSTALL THIS.

In fairness, it was probably calibrated for slow factory mushy suspension and skinny not so grippy tires. My car was stock for only 3 months of it's 14 years of ownership with me :)
 
Yep I don't think the original team thought about triple spring rates and 235/275 R compounds in 1989...
 
The TCS "unsafe" aspects has been covered many times on prime but basically it takes away the predictability. For example, often when TCS is active, as the car enters a slide, at some point a decent driver will correct the slide, however as your correcting TCS will step in - it's so slow to respond and is not a brake based TCS, it cuts throttle when in some situations you need the throttle to get you out of trouble. TCS in the car is a liability. I'll happily wrap it in a pretty little box with a bow on it and a note saying NEVER INSTALL THIS.

In fairness, it was probably calibrated for slow factory mushy suspension and skinny not so grippy tires. My car was stock for only 3 months of it's 14 years of ownership with me :)

You guys conviced me to leave it unconnected as while one can disable it manually, it will always rearm whenever you start the engine.
 
The design of the actuator is such that it can force the throttle closed when its open but it can't actually open the throttle.

I'm inclined to agree with @RYU, the early TCS system seems to create hazardous situations as often as it prevents them.

It's most probably possible (according to the schematic on page 19-88) to get the throttle actuator to open the throttle if the TCS throttle lever is positionned on the other side of the cam dog?
The way I see it operate would be to actuate the "blip" through a little switch on the gear lever conditionned by the fact that the brakes and the clutch are applied simultaneously.
Having seen it work in a Porsche Cayman GT4 made me drool getting it in my NSX!
I've been out of EE for too long to be able to design my own circuit but apparently the Arduino route appears a lot easier than when I programmed Motorola's 6800 microprocessor!
 
Last edited:
The TCS "unsafe" aspects has been covered many times on prime but basically it takes away the predictability. For example, often when TCS is active, as the car enters a slide, at some point a decent driver will correct the slide, however as your correcting TCS will step in - it's so slow to respond and is not a brake based TCS, it cuts throttle when in some situations you need the throttle to get you out of trouble. TCS in the car is a liability. I'll happily wrap it in a pretty little box with a bow on it and a note saying NEVER INSTALL THIS.

In fairness, it was probably calibrated for slow factory mushy suspension and skinny not so grippy tires. My car was stock for only 3 months of it's 14 years of ownership with me :)

The only time it ever worked "right" was when I drove my NSX in the snow. In those cases, with proper snow/ice tires, it worked well and I could feel it helping in very slick conditions. It was particularly helpful when starting from a stop at a light. Otherwise, I disabled it too- and always at the track.

It's most probably possible (according to the schematic on page 19-88) to get the throttle actuator to open the throttle if the TCS throttle lever is positionned on the other side of the cam dog?
The way I see it operate would be to actuate the "blip" through a little switch on the gear lever conditionned by the fact that the brakes and the clutch are applied simultaneously.
Having seen it work in a Porsche Cayman GT4 made me drool getting it in my NSX!
I've been out of EE for too long to be able to design my own circuit but apparently the Arduino route appears a lot easier than when I programmed Motorola's 6800 microprocessor!

The Arduino is really neat. I'm using it for a custom oil temp warning light.
 
[MENTION=25108]illwillem[/MENTION] did some fancy Arduino coding to disable TCS on his IS-F. On that car it requires not only a sequence button presses but also brake pedal input or something. He calls it's the pedal dance. Now it's a 1 button process with the Arduino doing the dancing.
 
Back
Top