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

Wideband O2 simulating stock O2 signal?

Joined
30 August 2014
Messages
150
Location
Colorado
I am interested in installing a pair of LC-2 wide-band controllers and simulate the stock narrow-band O2 sensors using the stock ECU. I know the narrow-band simulation works on other brands (Mazda) but I don't know enough about the NSX ecu. I know it monitors O2 heater current, but I don't know how it uses that information.

Has anyone done this?
'91 (naturally aspirated)
headers & exhaust,
un-confirmed RM chip.
 
I don't have the LC-2; but, I have the LC-1 on another car with an MS II fuel injection system.

First problem. On the LC-1 and likely on the LC-2 the heater current and the controller current supply are all in one so you won't be able to monitor the status of the heater circuit which is going to cause an error code.

Most OEM ECUs use the O2 signal to try and keep the AFR as close to 14.7:1 as possible which is where the catalytic converters do the best job on post combustion clean up. That is why cars usually come with narrow band O2 sensors. The only thing that matters is being close to stoich! The LC-1 and I expect the LC-2 can be programmed to emulate a narrow band controller which should satisfy the control objectives within the NSX ECU (notwithstanding the fact that your MIL will likely be lit up).

The obvious question is why bother. The OEM ECU needs a narrow band signal. If you want a wide band O2 sensor to give you a signal confirming that your AFR is always around stoich, then weld in a second bung for a wideband O2 sensor and don't mess with the narrow band signal going to the OEM ECU. The moment that you decide to install an aftermarket ECU and want to mess with the AFRs, then its time to install a wide band O2 sensor.
 
Thanks for replying, all good information about the heater circuit, if true, could be a roadblock.

But why bother?
First, my headers are ceramic coated, have you ever tried to weld one that has been coated? It seems like you can't grind enough material to get good metal.
When you have, headers, rear exhaust , unknown chip and your functioning fuel pump resistor is bypassed you really don't know where your engine is. It would be nice to monitor the afr while you test repairs without inducing a fault/code. Does this unknown chip need the FP resistor bypassed? Because the fuel pressure is roughly 3/4ths at idle with the resistor connected, that is a significant change in pulse width, is the chip tuned for this? I don't know. There are a lot of questions that a wide band could answer. I have been getting a random code 44 & 2, rear fuel supply/o2 code. Sensor has been replaced and the wiring examined. A wideband could help diagnose this.

maybe I should just do the RDX injector mod and tune, and get rid of this chip altogether and start over?
 
The ceramic coating does provide a good reason for not wanting to weld in a second O2 bung.

As an observation, the consensus of NSX owners with aftermarket parts seems to be that the changes associated with retrofitting headers and exhaust is small enough that the trim function in the OEM ECU can bring the AFR back to its target value without any problems. By-passing the fuel pump resistor will increase the fuel flow rate through the injector at lower throttle openings. Given a stock ECU, this will mess up the fueling equation; however, the ECU will attempt to bring the AFR back to its target value by applying lots of trim to reduce the injector opening time. It may or may not be successful in doing this. One of the problems with doing this is that the really short injector opening times associated with the higher flow rates can make for lousy operation at idle. I don't know about OBDI cars; but, on OBDII cars if the long term trim exceeds a certain value (I seem to remember 20%; but, I am not sure), the ECU generates an error code.

The unknown chip presents a whole set of unknown problems. I can agree that sticking in wide band O2 sensors will tell you whether the AFRs are straying away from 14.7:1. The fact that you have an altered chip with a presumably altered fuel map suggests that you should probably expect that the engine is no longer running at 14.7:1. The next question will be what are you going to do if the AFRs are no longer close to 14.7:1? You have no ability to go in and modify the fuel map in the stock ECU and even if you did, do you know what you would want to change the fuel map values to?

The RDX injector modification has mixed reviews. Just check the Prime forums. If you want trouble free operation, take the money associated with installing two LC-2 controllers and apply it to purchasing a stock ECU (there are probably a number of used ones available at a reasonable price) and restore the operation of the by-pass resistor. The stock ECU should manage your aftermarket exhaust system without problem. If you plan to go big in the future, purchase the LC-2s and an aftermarket fuel management system and get ready to start spending a lot of $$$. The up-side to going with an aftermarket EMS is that there are probably lots of tuners out there who can reprogram the EMS (including yourself if you choose to take the time to learn). With the RDX modification, I believe that there is only one vendor supporting the product.

My NSX is stock because I like to drive it rather than play with it. I have another car that I fiddle with and which seems to spend more of its time in a non-operational state!
 
Last edited:
The stock narrow band O2 sensors measure to stoich, but that is not the performance ideal for the engine, which is where the tuning comes in. Honda tuned the NSX ECU to meet specific fuel efficiency and emissions targets. They also tuned it to be reliable in all driving conditions, weather and fuel grades. All of this means something other than 14.7:1, but it is not ideal for power. Most naturally-aspirated tuners assert that a ratio between 13.5 and 13.9 yields max power and a safe AFR. This is their secret sauce though, so you won't find out what it is. But, your narrow-band O2 is really a reference point for the ECU, since even Honda tuned it to a different AFR.

Most tuners concede there is about 20 whp available in the stock ECU through a good tune, but the cost/benefit is high and most opt instead for a supercharger or standalone system. Chip-by-mail won't cut it- you need your car tuned on a load-bearing dyno so each load cell can be tuned to that magic AFR. That's why we're working on getting Prospeed out here.
 
The stock narrow band O2 sensors measure to stoich, but that is not the performance ideal for the engine, which is where the tuning comes in. Honda tuned the NSX ECU to meet specific fuel efficiency and emissions targets. They also tuned it to be reliable in all driving conditions, weather and fuel grades. All of this means something other than 14.7:1, but it is not ideal for power. Most naturally-aspirated tuners assert that a ratio between 13.5 and 13.9 yields max power and a safe AFR. This is their secret sauce though, so you won't find out what it is. But, your narrow-band O2 is really a reference point for the ECU, since even Honda tuned it to a different AFR.

Most tuners concede there is about 20 whp available in the stock ECU through a good tune, but the cost/benefit is high and most opt instead for a supercharger or standalone system. Chip-by-mail won't cut it- you need your car tuned on a load-bearing dyno so each load cell can be tuned to that magic AFR. That's why we're working on getting Prospeed out here.

You are correct that it is generally accepted that you can extract some more horsepower from an engine by running it with a lower AFR for the engine speed and MAP values that correspond to peak horsepower. I have seen some engines running with AFRs in the mid to high 12 range for those bins in the fuel map that are in the range of the horsepower peak.

However, I would be surprised if Honda tuned the engine to run at something other than 14.7. First, the narrow band sensor is absolutely useless at providing a useful signal from AFRs that deviate much from stoichiometric. As you can see from the typical narrowband curve attached, there is probably only a 70 mV change in output when the AFR changes from 14.5 to 12 and essentially miniscule change in output once the AFR is less than 12. Its response curve is even worse on the lean side. Narrowband sensors are really good at sending a strong signal to the ECU that they are above 14.7 (which would be about 450 mV) or below 14.7 which allows the ECU to trim the injector pulse width to bring the AFR back to stoich. The reason stoichiometric is so important is that catalytic converters perform best when the exhaust gas is at stoich and tend to struggle as it deviates from stoich. Honda might want to maximize horsepower by running at something other than stoich; but, they have to meet emission standards and they have to meet those emission standards without scheduled maintenance for the period determined by legislation. You do that by sticking to stoichiometric ratios.

That said, the O2 sensors drop out of the equation at peak horsepower because every engine management system that I am aware of ignores the O2 sensor output at high MAP values and throttle openings (typically greater than 90%). The reason is that the transient response of the O2 sensors is too slow to be reliable and the ECUs go into a straight feed forward mode (open loop) running off the fuel maps and don't try to correct to a target AFR. If the new vehicle emission certification does not check for compliance at wide open throttle, then Honda could adjust the fuel map to give a slightly rich mix in the engine speed / MAP bins that correspond to peak horsepower and still have the ECU correct to 14.7:1 once it comes back into closed loop operation. If that loophole exists in the certification process, then it would be possible to have a factory tune to something other than 14.7:1 for the horsepower peak. However, I am suspicious. The fact that diddling the target AFRs on an ECU with open code can yield material horsepower gains at peak with no other material engine modifications says to me that Honda is probably sticking pretty close to stoich. The upside is that it leaves relatively easy horsepower pickings for somebody with a dyno, an ECU with open code and a laptop.

Update:

After a little more consideration on the matter, OEM manufacturers might be tempted to set the wide open throttle portions of their fuel maps to deliver an AFR slightly on the low side of 14.7:1 when they are running in open loop operation. This would give them some safety margin in the event that something like low fuel pressure or some other factor was messing things up.
 
Last edited:
To save fuel, I have my wideband analog output configured like this (.002V = 15.3:1, 1.002V = 14.8:1). Using the factory computer this yields a closed loop operation around 15.2:1. You can play with the numbers to give the "swing" you desire.

Last couple of years however, I have found open loop to be more desirable than any closed loop. Perhaps it is because I've spent about 10 years behind the laptop tuning various combinations. That said, what I've come up with is sort a checkerboard hypothesis, where the fuel map dips up and down from 14.6 -> 15.5 around the cruise area. This was so you do not spent all of your cruise time at 15.5 or 14.6, but rather, back and forth between them as road conditions (throttle position, climbing hills, turns, etc...) change. It mimics closed loop operation, looking at the wideband while you drive you wouldn't be able to guess it was in open loop. I discovered that after a 250~ miles drive (or rather about halfway through it) conditions change enough that I no longer wished to keep the engine steady in the 15's the entire time, the motor wanted a little more fuel once in a while to keep the EGT down and cool off. The last thing I want to do after driving 150+ miles in a lean condition is decide to make a 5th gear WOT pass, tipping into a hot lean engine.
 
Last edited:
Back
Top