Smog Test - O2 Pass - Air Fuel Fail

[NSX_NA1]

Hi,

As the title.

My car just failed emissions due to running too rich. It did however pass O2.

To me that would suggest my O2 sensors and cats are fine?

As for running to rich, my mods are as follows.

Induction kit
Full exhaust system including headers and sports cats
Vapor crankcase return delete
Intake coolant bypass
upgraded fuel pump, psi taken from filter within scope, between 47 and 53.

I have fairly new O2 sensors
Recently replaced the intake manifold sensor

What i did however do is recently reset the ecu, i believe i followed the correct procedure.

However upon googling, i found it can take 100's miles for the ecu to calibrate and settle down the fuel trims.

I believe this might be my cause?

The catch is, i'm not officially allowed to drive the car on the road without a valid MOT certificate, so... that doesn't really help me much, especially if i need to put loads of miles on it.

Please help, i have 14 days to correct the issue until my retest expires.

Thanks.

 

[NSX_NA1]

The manual and the info that use to be online for resetting the ecu doesn't exist like it use to.

There was a good guide but can no longer find it.

The steps i followed to reset the ecu was:

run the car up to temp (i didn't drive the car, it was only at idle, i believe ideally you would drive it up to temp)
i pulled the clock fuse under the front bonnet and waited 15 min (this i got from a old guide but the manual does say 10 seconds)
i then started the car up and ran for 15 min, did not touch the throttle (this time i found online as a suggested time that should allow for everything to calibrate)
I then turned the car off and turned the ignition back on straight away (this i also found online)

However, after reading deeper i started finding posts saying it can take loads and loads of miles for the calibration to complete. Not sure how i get round this if i'm not suppose to be driving the car.

 

[NSX_NA1]

Also more info, the car has sat still for a couple years and i've also recently changed the oil. Doubt oil makes any difference, but worth mentioning.

 

[NSXGB]

You will most likely find that you just need to blast the car up the bypass and get the cats really hot then get it tested as soon as you get to the MOT station. I do the same with my 29 year old cats.

 

[NSX_NA1]

awesome thanks, i will try that.

The cats are brand spankers.

Not sure how many retests they will give me before they start getting fed up at me lol.

 

[Honcho]

Is this an OBD-I or II NSX? On the OBD-I cars, the ECU will immediately begin to learn and adjust fuel trims once the car enters closed loop mode, which occurs once the water temperature crosses a certain threshold (can't remember what that is at this moment). This is why the guides say to warm up the car at idle. However, the car will only learn and adjust fuel trims based on the idle AFR. So, part two is to do as [MENTION=8378]NSXGB[/MENTION] says- drive it around. You don't need to mash on it since that will move the car out of closed loop, but you should drive it around town and on the highway for 30 minutes or so, trying to get the engine into as much of the low-midrange as you can (even just on the way to the MOT). The car will adjust the LTFT to be more accurate for non-idle conditions- it corrects much faster than 100's of miles. The cats will get nice and hot during this time too. IIRC, the base fuel state for a ECU reset is on the rich side (for safety) and this will affect your smog. After some driving, the car will learn and you'll be a lot closer to 14.7, where those brand spankers can do their job properly.

 

[Old Guy]

Hi,

As the title.

My car just failed emissions due to running too rich. It did however pass O2.

I am guessing you are based in GB so I am not at all familiar with the MOT test procedures. However, running rich versus pass O2 is confusing? O2 is not a controlled tail pipe emission anywhere that I am aware of. Do you mean that the car met its NOx limits; but, exceeded its HC limits (running rich)? That would make sense since excess fuel generally drops combustion temperatures which can reduce NOx.

Passing the NOx limits does not necessarily mean that the O2 sensors are OK. The fact that you are running with high HC values indicates that there could be a problem with your O2 sensors (but not necessarily). However, there are a number of other things that can cause high HCs. When you pulled the constant power fuse for the ECU, you wiped out the stored LTFT and any stored error codes. The ECU generates LTFT to correct the air / fuel ratio back to 14.7 to compensate for deterioration in other components in the fuel system. If your engine is perfect and everything is in tune the LTFT should be small. If stuff is less than perfect, as Honcho notes, you now need to drive around so that the ECU can start calculating whether a long term fuel trim correction is required to bring the AFR to 14.7. The other thing you did when you wiped the ECU's volatile memory is that you removed any stored error codes. If you have a pre OBD II ECU, checking for stored error codes may give you and indication as to whether you have a problem with the O sensors (voltages out of range). If you have an OBD II ECU, using a code reader can unlock a lot more information including the actual long term fuel trim values. I think the LTFT error limit is +/- 20%. However, if you use the code reader and discover that your LTFT is -19.5%, you know that something is up that might be worth checking out.

First thing would be to do as Honcho says and drive around moderately to try and re establish LTFT values and accumulate any ECU errors (not all errors cause the MIL to light instantly). If you have a pre OBDII ECU, use the service check connector to see if their are any stored error codes which may give you a clew as to what is going on. If there are stored codes, don't bother with the MOT until you get them sorted.

Intake coolant bypass

- what is that?

If you have a time crunch and restricted ability to drive around, by any chance does your replacement header have provision for two O2 sensors upstream of the cat? If so, I would be inclined to add a wideband sensor in to monitor AFRs while you are driving. If your AFRs are wandering all over the place then you have a problem that needs to be addressed. A lot of wideband sensors have two outputs. One is a linear output which provides wideband info and the other is a non linear output which emulates a narrow band sensor. You replace your existing narrow band with the wideband sensor and then use the narrow band output to drive the ECU and you can read the wideband output to see what your AFRs are actually running at. The simulated narrowband output can be flakey so I would never recommend that somebody try an emission compliance test using a simulated narrow band; but, it might be a good way of diagnosing a problem if you have a time crunch.

 

[NSXGB]

Being new is a worry. Are they high flow or OEM?

 

[NSX_NA1]

Thanks everyone...

The cats are new but high flow aftermarket cats as part of a exhaust system.

There are no additional bungs for a Wideband which thinking about it is a shame as i have a spare wideband sensor and gauge.

The car is a 1995 UK car, it's OBD1.

I might try driving it later at night, it's frustrating as i showed the MOT guy all the articles and explained i had recently reset the ecu.

The car had been sitting for a while.

I have recently done the valve clearances and changed the timing belt, followed the manual, because it was only my time i believe i got everything spot on.

The engine sounds really clean, smooth idle.

I did order some new O2 sensors, they are on order. Doesn't harm to have spares.

It's annoying because like you say, without a gauge i can't see if or when it settles down and runs more within range.

Would i need to follow the reset procedure again, or will it just self calibrate as i go?

I pulled the error codes by jumping the service connector, the only code was TCS 4 long, 4 quick. cleaned the rear right wheel speed sensor, didn't help so ordered a replacement as well.

Thanks

 

[Honcho]

Thanks everyone...

The cats are new but high flow aftermarket cats as part of a exhaust system.

There are no additional bungs for a Wideband which thinking about it is a shame as i have a spare wideband sensor and gauge.

The car is a 1995 UK car, it's OBD1.

I might try driving it later at night, it's frustrating as i showed the MOT guy all the articles and explained i had recently reset the ecu.

The car had been sitting for a while.

I have recently done the valve clearances and changed the timing belt, followed the manual, because it was only my time i believe i got everything spot on.

The engine sounds really clean, smooth idle.

I did order some new O2 sensors, they are on order. Doesn't harm to have spares.

It's annoying because like you say, without a gauge i can't see if or when it settles down and runs more within range.

Would i need to follow the reset procedure again, or will it just self calibrate as i go?

I pulled the error codes by jumping the service connector, the only code was TCS 4 long, 4 quick. cleaned the rear right wheel speed sensor, didn't help so ordered a replacement as well.

Thanks

You don't need to reset it again. Just keep driving and it will continue to learn and trim. I tend to think it might be your O2 sensors being bad, but the odds of both being bad at the same time are pretty slim. Thus, I'd also check the wiring and connectors from the sensors- if it is damaged or too close to the manifolds, it could affect the signal. Something is causing the ECU to add fuel when it shouldn't be. Another thought is your engine coolant temp sensor- if it is not working, the ECU could think the engine is cold and add fuel.

 

[Old Guy]

You don't need to reset it again. Just keep driving and it will continue to learn and trim. I tend to think it might be your O2 sensors being bad, but the odds of both being bad at the same time are pretty slim. Thus, I'd also check the wiring and connectors from the sensors- if it is damaged or too close to the manifolds, it could affect the signal. Something is causing the ECU to add fuel when it shouldn't be. Another thought is your engine coolant temp sensor- if it is not working, the ECU could think the engine is cold and add fuel.

I pretty much agree with Honcho. Definitely do not reset the ECU. With reference to the coolant sensor, that is why I asked what the "Intake coolant bypass" was and whether it was something that might affect ECT readings. Be aware that on your 1995, the gauge temperature sensor and the engine coolant temperature sensor are different. If the gauge is reading fine that does not mean that the engine coolant sensor for the ECU is correct. Check its resistance as per the service manual and make sure the wiring is not damaged. In addition to adding enrichment for start up, an out of range ECT sensor will prevent the ECU from going into closed loop operation. If that happens. you can drive around to the cows come home and you will never ever accumulate the long term fuel trims required to correct the fuel mixture.

Did your "Intake coolant bypass" modification alter the engine operating temperature? If the thermostat is not operating correctly or the the engine is not coming up to operating temperature the engine is likely going to be stuck in a running rich mode.

You said you replaced the intake manifold air temperature sensor. Check the resistance of the sensor as per the service manual to make sure that it is in spec. and check the wiring to make sure that it is not compromised.

You said the O2 sensors were fairly new. Does 'fairly' mean 1 month or 1 year? It would be unlikely for new O2 sensors to be bad (unless they are cheap aftermarket); but, if the car has been driven around for a while and it is running significantly rich then the O2 sensors can be soot contaminated. If the car is also burning oil that will do a number on O2 sensors.

Pull the sparkplugs and have a look at the porcelain insulators around the center electrode. Are they brown to sooty in color? If so, then the fuel mix is definitely on the rich side. If the the insulators are a nice even white to a very, very slight greyish-tan, the fuel mix is probably close to being correct. If one or two spark plugs have an abnormal amount of color you could have a leaky injector(s) which could skew the AFRs. If the car sat unused for a long time with fuel in the system it is always a good plan to remove the injectors and have them sent out for cleaning and flow testing.

If all the sparkplugs look 'correct' then your fuel mix may be mostly correct. The ECU has error detection for abnormal sensor voltages (dead / dying O2 sensor) and for fuel mixture out of range (long term fuel trim is excessive). The error detection is not perfect; but, if you don't have any error codes and your spark plugs look good, the ECU and engine may be just fine and your problem may be your after market catalytic converters. When you had your MOT test, how far out were your HC measurements compared to the standard? If they were just slightly out, that could be a catalytic converter performance problem.

One final thing to check. Since you replaced the complete exhaust system, check for leaks in the exhaust system that might be admitting free O2 in and skewing the O2 sensor reading causing it to increase the injector pulse widths. If there was an air leak on the front manifold this would show up as all of the front sparkplugs being browner than normal ( same for rear). The air leak has to be up stream of the O2 sensor are fairly close downstream.

 

[drew]

Not hard to ruin the o2 sensors with penetrating oil on removal of the old exhaust.

 

[docjohn]

In the US the 95/96 are OBD2....

 

[Old Guy]

OBDII was not a regulatory requirement in the EU in 1996. EOBD did not come into effect until 2000 / 2001. I don't know whether > 1995 EU market ECUs had drive by wire; but, no OBDII error reporting or whether they got the OBDII ECU with no external port and firmware disabled the extended error reporting or whether they got the port even though it was not a regulatory requirement (and good luck finding a code reader). By 1996 the production numbers on the NSX were low enough that it is hard to believe that Honda would create a special hardware variant ECU with drive by wire; but, no OBDII just for the Euro market. The 'tell' would be whether the EU 1996 - 2000 NSXs had pre and post cat O2 sensors. If they just have a pre cat sensor then it is a special ECU with DBW; but, none of the extended OBDII error reporting stuff.

Of course, GB sort of being not all in with the ECU (even back then) could have had different rules.

 

[docjohn]

I know our US 94 was OBD1 with throttle cable but had some features of the upcoming 95/96..

 

[drew]

I believe the USDM year model is N+1 over the rest of the world.