Rough Idle
- Thread starter captngaryr
- Start date
Yes. Primary are in the manifold in front of the cats and secondary are on the outlet of the cats. The primary sensors do the fuel mixture control / correction. The only function of the secondary sensors is to monitor the operation of the cats. They have no control function.
Tip - you need to do this when the engine has come up to operating temperature. When the engine is cold it runs in open loop fuel control and doesn't go into closed loop fuel control using the O2 sensor until it has warmed up enough to come off of warm-up fuel enrichment.
Tip - you need to do this when the engine has come up to operating temperature. When the engine is cold it runs in open loop fuel control and doesn't go into closed loop fuel control using the O2 sensor until it has warmed up enough to come off of warm-up fuel enrichment.
Reran the test with the Primary O2 sensors disconnected and the engine up to temperature. The engine still seemed rough to me. It's not rough enough to shake all the bolts loose, but doesn't seem right. How bad would the engine run if the timing belt were off one tooth?
Gary
Gary
Can't comment on how rough a one tooth error would cause the engine to run or whether it would have any effect at all; however, there was a recent thread on Prime where an owner was reporting a mixture out of range error code which was eventually attributed to a one tooth out error in replacing the timing belt. Did you just have the timing belt replaced?
Your car is still on the young side; but, if you get a chance put it up on the hoist and have a look at the front engine mount. If that has deteriorated / failed that can result in a lot of noise and engine vibration.
Your car is still on the young side; but, if you get a chance put it up on the hoist and have a look at the front engine mount. If that has deteriorated / failed that can result in a lot of noise and engine vibration.
Old Guy
I've only had the car for about a month and part of that time it was in a shop for unrelated work. I've probably driven the car for a total of 100 miles. The previous owner said he never noticed such a problem. The timing belt was replaced about 2 years ago, but again if you've driven an NSX that was running smoothly, you'd notice something's not quite right with this one. The guy who replaced the t-belt is well know and a respected NSX mechanic in the area where I bought the car, so it's pretty unlikely that he could have gotten the timing wrong, but stranger things have happened. Compression checks out fine with values within a couple of psi. I didn't notice the problem when I traveled to check out the vehicle, but with so many question spinning in my head, I could have missed it. I'm not the brightest bulb in the box, sometimes. I was under the car this morning and it looks like a brand new car underneath, but I didn't specifically look at the engine mounts. I'll do that next time I'm under there, Other rubber parts such a exhaust mounts looked brand new. I don't know how the CEL system detects a misfire, but I'm not throwing any codes so I haven't a lot to go on. I have a pretty basic OBDII reader, but that shouldn't matter if there's no CEL light on the dash. I tried listening to all the fuel injectors with a stethoscope, and they all sound the same (for what that's worth). I don't have a fuel pressure gauge, but there may be one in my future. If you have any other things I can check, fire away.
Thank you for all your help,
Gary
I've only had the car for about a month and part of that time it was in a shop for unrelated work. I've probably driven the car for a total of 100 miles. The previous owner said he never noticed such a problem. The timing belt was replaced about 2 years ago, but again if you've driven an NSX that was running smoothly, you'd notice something's not quite right with this one. The guy who replaced the t-belt is well know and a respected NSX mechanic in the area where I bought the car, so it's pretty unlikely that he could have gotten the timing wrong, but stranger things have happened. Compression checks out fine with values within a couple of psi. I didn't notice the problem when I traveled to check out the vehicle, but with so many question spinning in my head, I could have missed it. I'm not the brightest bulb in the box, sometimes. I was under the car this morning and it looks like a brand new car underneath, but I didn't specifically look at the engine mounts. I'll do that next time I'm under there, Other rubber parts such a exhaust mounts looked brand new. I don't know how the CEL system detects a misfire, but I'm not throwing any codes so I haven't a lot to go on. I have a pretty basic OBDII reader, but that shouldn't matter if there's no CEL light on the dash. I tried listening to all the fuel injectors with a stethoscope, and they all sound the same (for what that's worth). I don't have a fuel pressure gauge, but there may be one in my future. If you have any other things I can check, fire away.
Thank you for all your help,
Gary
Last edited:
Is the engine coming up to full operating temperature (needle a nudge under the mid point)? If the engine is not up to normal operating temperature it may not be coming out of warm up enrichment and going into closed loop fuel control off the O2 sensors. My 2000 definitely has more vibration when cold.
Does your OBDII scanner allow you to check the status of something called the emission monitors which are the software routines that the ECU goes through to check the status of 'stuff'? I am a little fuzzy on this particular detail of OBDII. There is definitely a misfire monitor and a fuel monitor (related to O2 sensor operation) and a bunch of other monitors. I don't know whether these two particular monitors need to go through a drive cycle to set them. If the monitors are not operational then you might have a misfire or an O2 sensor / fuel mix problem; but, the ECU is not generating an error code because the monitor is not active. If your scanner tells you that your monitors are set / OK then the absence of error codes really is because of an absence of detected problems and you need to look elsewhere.
I just suggested the engine mount as a last resort. Given the age of you car and relatively low miles I would not expect a problem; but, it is a low cost item to inspect.
Does your OBDII scanner allow you to check the status of something called the emission monitors which are the software routines that the ECU goes through to check the status of 'stuff'? I am a little fuzzy on this particular detail of OBDII. There is definitely a misfire monitor and a fuel monitor (related to O2 sensor operation) and a bunch of other monitors. I don't know whether these two particular monitors need to go through a drive cycle to set them. If the monitors are not operational then you might have a misfire or an O2 sensor / fuel mix problem; but, the ECU is not generating an error code because the monitor is not active. If your scanner tells you that your monitors are set / OK then the absence of error codes really is because of an absence of detected problems and you need to look elsewhere.
I just suggested the engine mount as a last resort. Given the age of you car and relatively low miles I would not expect a problem; but, it is a low cost item to inspect.
So, I've checked the plugs, coils, O2 sensors and for stored codes. The fuel injectors sound the same with a stethoscope. The engine mounts look fine. Has anyone got any other ideas on what I can check?
Thanks,
Gary
Thanks,
Gary
Injectors 'sounding the same' just means that the solenoid operated valves in the injectors are operating. It does not mean that the injectors are flowing equal amounts or that the offsets are equal. The only way to definitively check that is removal and ship them to a cleaning service like RC Fuel Injection or a local injection shop if you have one you trust. If one of the injectors is slightly clogged, the O2 sensor could correct the fuel mix on the other two cylinders on that bank to give the correct AFR on the aggregate of the three cylinders. This would result in two cylinders running rich and one cylinder running lean which would make for uneven engine operation; but, as far as the OBDII is concerned everything is OK. This kind of problem would normally show up as high fuel trim - hence my suggestion about a scanner that can report fuel trims. If the fuel trims are small, then you pretty much know that your problem is not fuel related.
I am aware that the stethoscope is a very incomplete test but it was all I had. I agree with what you are saying, and am looking to get a better OBDII App. so I can check things like fuel trim. Can you recommend a good one for an iPad? My old one isn't very complete.
Thanks,
Gary
Thanks,
Gary
I have only plugged a scanner into my NSX once, and that was out of curiosity to see what came up which was nothing. I don't have a scanner and was using my son's so I have not useful advice. I know Larry B. had a PC based scanner that he was a fan of. Perhaps he can advise on a suitable tool.
Interesting dilemma. Do you just pull the injectors to get them checked and cleaned which would probably be less than the cost of a good scanner with the risk that it might not be the injectors. Or, do you purchase a good scanner and then find out that there is no problem with fuel trim. Or you purchase a scanner and then find out you have a trim problem and then have to send the injectors out for cleaning. Time to get out the textbook on risk analysis!
One observation. Even if it is not 'the' problem, having the injectors on a 17 year old car cleaned and flow tested would never be a total waste of money.
Interesting dilemma. Do you just pull the injectors to get them checked and cleaned which would probably be less than the cost of a good scanner with the risk that it might not be the injectors. Or, do you purchase a good scanner and then find out that there is no problem with fuel trim. Or you purchase a scanner and then find out you have a trim problem and then have to send the injectors out for cleaning. Time to get out the textbook on risk analysis!
One observation. Even if it is not 'the' problem, having the injectors on a 17 year old car cleaned and flow tested would never be a total waste of money.
I got an OBD Fusion scanner app and it does show fuel trim percentage. The percentages (in the garage) run from about -1% to about +15%. I don't know how to tell if that is normal or not. One bank seem about 2 or 3 percent higher than the other. When the car is fully warmed up the roughness seems less, but peaks at about 2000 RPM. I'm wondering if that's just a resonance as a result of the 90 degree V6 and I'm making a mountain out of a mole hill. I haven't run the OBD scanner on the road yet, as the weather has been crappy around here lately.
More to come,
Gary
More to come,
Gary
Last edited:
Three thoughts - first, I doubt that you are making a mountain out of a mole hill because you have too much experience; there probably is some problem. Second, bad gas is something to look into. Last is to go back over any work which was done ( PPI as an example ) between the time when the car ran properly and now looking for things that aren't right. In my experience when a car starts having a problem after being worked on recently the problem can often be traced to a ground that isn't tight, some aged plastic part that developed a crack by being disturbed, etc. This type of thing happens to the best mechanics and is just part of working on old cars.
Last edited:
The NSX ECU has two sets of fuel trims, a long term trim and a short term trim. The long term trim is like a long term running average of the short term trim. If the short term trim is small and going positive and negative then the long term trim should be close to zero. If the short term trim values are consistently positive or negative the ECU will start to accumulate long term trim values. The ECU adds short and long term trim to calculate the actual fuel correction and the objective in accumulating long term trim seems to try and keep the short term trim close to zero. I also expect that large long term trims are what the ECU uses to trigger an error code. As you have discovered, the NSX should have separate trims for the front and back cylinder banks so potentially you could have 4 trim values; front short, front long, back short and back long.
If the trim values are bouncing around a lot I expect that you are looking at the short term trim values which are prone to fluctuate. However, I am surprised that they are bouncing up to 15%. That seems like a rather high fluctuation. I have a 1970s vintage cast iron four banger that I have retrofitted with digital EFI. This engine has a pre NOx limits camshaft with a fair amount of overlap and lift which makes for somewhat erratic idle in particular because I am forcing the idle AFRs up to around 15. The short term trim typically bounces +/- 3% at idle. I would expect (but have never checked) that the NSX with its milder base cam profile should have short term trims that do not fluctuate much and should be close to zero or small single digits at idle. For the short term trim values to have any meaning you need to be looking at them when the engine is at normal operating temperature with the ECU running in closed loop at steady load. Do not look at short term fuel trim during engine transient conditions (accelerating and de accelerating). The trims may be all over the place because the ECU shuts fuel off during de acceleration and adds extra fuel during acceleration.
If you can access the long term trim values those will have more meaning. If these are large numbers and particularly if they are uneven from front to back, that is an indication that something is likely amiss (which could include a failing O2 sensor).
The short term fuel trim values are the instantaneous fuel correction that the ECU is applying to hit the target AFR ratio, which at low loads on an emission controlled car is normally 14.7. If your short term trim is bouncing up to +15% I expect that means that it is increasing the length of the fuel PW in the base fuel map by 15%. This would normally occur because the O2 sensor has measured an AFR that is higher than 14.7 (running lean). This could occur because you had a misfire or an incomplete combustion event. A misfire will cause unburnt fuel to enter the exhaust manifold. It will also cause an equivalent amount of unburnt O2 to enter the exhaust manifold. O2 sensors don't measure AFR, they measure O2 content in the exhaust relative to ambient air. When an O2 sensor registers an increase in O2 in the manifold it interprets this as running lean and the ECU tries to correct this by increasing the fuel trim. So, if you are seeing repeated large positive spikes in short term trim you could be having a mild misfire event or some kind of incomplete combustion that are not severe enough to trigger the misfire detection monitors.
As a double check, some scanners can directly access the O2 sensor output voltage. When the engine is operating at 14.7 AFR, the sensor voltage will be somewhere around the 0.2 - 0.6 volt range. If you can access the scanner output and you see the O2 sensor voltage dropping below 0.2 volts that is an indication that you could be having a misfire / incomplete combustion event. If this fluctuation is periodic it is also possible that you could have one lazy injector that is causing one cylinder to run lean causing a periodic lean spike to show up. Don't bother with the secondary O2 sensor voltages since they should pretty much be rock steady.
If the trim values are bouncing around a lot I expect that you are looking at the short term trim values which are prone to fluctuate. However, I am surprised that they are bouncing up to 15%. That seems like a rather high fluctuation. I have a 1970s vintage cast iron four banger that I have retrofitted with digital EFI. This engine has a pre NOx limits camshaft with a fair amount of overlap and lift which makes for somewhat erratic idle in particular because I am forcing the idle AFRs up to around 15. The short term trim typically bounces +/- 3% at idle. I would expect (but have never checked) that the NSX with its milder base cam profile should have short term trims that do not fluctuate much and should be close to zero or small single digits at idle. For the short term trim values to have any meaning you need to be looking at them when the engine is at normal operating temperature with the ECU running in closed loop at steady load. Do not look at short term fuel trim during engine transient conditions (accelerating and de accelerating). The trims may be all over the place because the ECU shuts fuel off during de acceleration and adds extra fuel during acceleration.
If you can access the long term trim values those will have more meaning. If these are large numbers and particularly if they are uneven from front to back, that is an indication that something is likely amiss (which could include a failing O2 sensor).
The short term fuel trim values are the instantaneous fuel correction that the ECU is applying to hit the target AFR ratio, which at low loads on an emission controlled car is normally 14.7. If your short term trim is bouncing up to +15% I expect that means that it is increasing the length of the fuel PW in the base fuel map by 15%. This would normally occur because the O2 sensor has measured an AFR that is higher than 14.7 (running lean). This could occur because you had a misfire or an incomplete combustion event. A misfire will cause unburnt fuel to enter the exhaust manifold. It will also cause an equivalent amount of unburnt O2 to enter the exhaust manifold. O2 sensors don't measure AFR, they measure O2 content in the exhaust relative to ambient air. When an O2 sensor registers an increase in O2 in the manifold it interprets this as running lean and the ECU tries to correct this by increasing the fuel trim. So, if you are seeing repeated large positive spikes in short term trim you could be having a mild misfire event or some kind of incomplete combustion that are not severe enough to trigger the misfire detection monitors.
As a double check, some scanners can directly access the O2 sensor output voltage. When the engine is operating at 14.7 AFR, the sensor voltage will be somewhere around the 0.2 - 0.6 volt range. If you can access the scanner output and you see the O2 sensor voltage dropping below 0.2 volts that is an indication that you could be having a misfire / incomplete combustion event. If this fluctuation is periodic it is also possible that you could have one lazy injector that is causing one cylinder to run lean causing a periodic lean spike to show up. Don't bother with the secondary O2 sensor voltages since they should pretty much be rock steady.
Last edited:
The previous owner (I think you know him) drove the car after the PPI and didn't notice the problem (I believe him). The car has not been worked on since the PPI. I noticed the problem the day the car arrived. So, either the problem is so subtle as to not be noticed by PO, or the car broke in the truck on the way across the country or I have a vivid imagination. Arggh! This is driving me nuts. The new OBDII scanner shows nothing, and it appears to be a very complete scanner. I'm not sure how much fuel trim is normal, however.
All the best,
Gary
All the best,
Gary
Wow! You are the best! That is exactly what I wanted to know. I have access to short term and long term % fuel trim for both banks. It seams that the short and long term trim are going mostly positive and seldom negative. That surprises me. I also have access to the O2 sensor voltages. I plan to take the car out tomorrow (the weather is finally supposed to break around here) and run the scanner on the road. I will collect some data and get back to you.
Thanks,
Gary
Thanks,
Gary
In the end, I took her to the local Certified NSX Mechanic and he assured me that, that although the vibration is perhaps slightly more than my previous NSX, the car is quite normal. There is a intrinsic secondary imbalance in a 90 degree V6 and they believe that is what I am feeling. We checked the plugs, coils, vacuum leaks, flywheel, OBDII and engine mounts and all check out fine. For all who help me in this fools errand, I thank you.
Gary
Gary
Similar threads
