Emissions Conundrum
Discussion
Following on from my recent exhaust noise issues, today I had a mobile chap come out to test the emissions. I was expecting just to have to tweak a couple of points on the ECU map, but it's all turned out a bit more complicated.
Background: 3.0 v6, separate exhaust per bank with cats, Emerald ECU, mapped on Emerald's rollers back in October. Emissions then were fine, well inside the 0.02 CO / 200 HC limits I need.
I've done nothing to the engine, other than take the exhausts off, and drive it 70 miles. It's been run a number of times on the drive to check various things.
So today, we did the warm up (2-5000 rpm until the fan cut in at 97 degrees - about 10 minutes).
We did both slow and fast (2500-3000rpm) tests, a number of times each, each time leaving the probe in for several minutes.
C0 was very high, HC about right. I reduced the fueling at the map positions concerned (can't go wrong here as the software shows you which cell it's reading from), plus the ones either side as it extrapolates.
At fast idle, on bank 2 (Offside), it came down but only to about 0.5. On bank 1 it was about 4.0
Slow idle, bank 1 was in, just, at 0.03; bank 2 was still up at 3.5-4!
Unfortunately his lambda sensor wasn't working; however the narrowband on the car suggests that it's rich, which would match.
Soooooo...I'm at a loss to think of what can have changed, and also to account for the difference between banks. Is there anything I should check for mechanically speaking which could cause it to run so rich? Injection is sequential but there's no sound of any misfire or anything obviously wrong. It drove fine and revs fine.
It's on a single plenum and single throttle.
I'm also not sure of what I can do about it, short of dumping it somewhere with a rolling road and a full gas analyser and falling on their mercy...so any and all help appreciated.
Background: 3.0 v6, separate exhaust per bank with cats, Emerald ECU, mapped on Emerald's rollers back in October. Emissions then were fine, well inside the 0.02 CO / 200 HC limits I need.
I've done nothing to the engine, other than take the exhausts off, and drive it 70 miles. It's been run a number of times on the drive to check various things.
So today, we did the warm up (2-5000 rpm until the fan cut in at 97 degrees - about 10 minutes).
We did both slow and fast (2500-3000rpm) tests, a number of times each, each time leaving the probe in for several minutes.
C0 was very high, HC about right. I reduced the fueling at the map positions concerned (can't go wrong here as the software shows you which cell it's reading from), plus the ones either side as it extrapolates.
At fast idle, on bank 2 (Offside), it came down but only to about 0.5. On bank 1 it was about 4.0
Slow idle, bank 1 was in, just, at 0.03; bank 2 was still up at 3.5-4!
Unfortunately his lambda sensor wasn't working; however the narrowband on the car suggests that it's rich, which would match.
Soooooo...I'm at a loss to think of what can have changed, and also to account for the difference between banks. Is there anything I should check for mechanically speaking which could cause it to run so rich? Injection is sequential but there's no sound of any misfire or anything obviously wrong. It drove fine and revs fine.
It's on a single plenum and single throttle.
I'm also not sure of what I can do about it, short of dumping it somewhere with a rolling road and a full gas analyser and falling on their mercy...so any and all help appreciated.
Exact readings of both CO, HCC and Lambda for both banks would be helpful.
Is the car single point, multipoint injection ? batch fire, sequential ?
Compression all good on both banks ?
Are all injectors known to be in good order ? No air leaks in the intake etc ? Exhaust ?
Was the testers gas analyser definitely working correctly ? You mention he was also using a lambda sensor ?
Is the car single point, multipoint injection ? batch fire, sequential ?
Compression all good on both banks ?
Are all injectors known to be in good order ? No air leaks in the intake etc ? Exhaust ?
Was the testers gas analyser definitely working correctly ? You mention he was also using a lambda sensor ?
I think he meant the lambda reading on his gas analyser, if the O2 sensor is duff you won't get a lambda reading.
He's getting opposite readings on each bank under differing conditions, one bank is rich at idle, the other is rich at fast idle, assuming that all the sensors aside from the lambdas (MAP/MAF, coolant temp, TPS etc) are all applicable to both banks then we need to know if it's in closed loop, what the correction factors are and what the lambda sensor readings are on each bank to make any progress.
He's getting opposite readings on each bank under differing conditions, one bank is rich at idle, the other is rich at fast idle, assuming that all the sensors aside from the lambdas (MAP/MAF, coolant temp, TPS etc) are all applicable to both banks then we need to know if it's in closed loop, what the correction factors are and what the lambda sensor readings are on each bank to make any progress.
Steve H said:
I think he meant the lambda reading on his gas analyser, if the O2 sensor is duff you won't get a lambda reading.
He's getting opposite readings on each bank under differing conditions, one bank is rich at idle, the other is rich at fast idle, assuming that all the sensors aside from the lambdas (MAP/MAF, coolant temp, TPS etc) are all applicable to both banks then we need to know if it's in closed loop, what the correction factors are and what the lambda sensor readings are on each bank to make any progress.
A gas analyser does not use a lambda sensor to calculate lambda.He's getting opposite readings on each bank under differing conditions, one bank is rich at idle, the other is rich at fast idle, assuming that all the sensors aside from the lambdas (MAP/MAF, coolant temp, TPS etc) are all applicable to both banks then we need to know if it's in closed loop, what the correction factors are and what the lambda sensor readings are on each bank to make any progress.
Hence the need for all information from the gas analyser, not partial info.
That's correct Steve but but most 4-gas setups use a laser bench to measure CO, CO2 and HCs and a chemical sensor to measure O2; if the O2 sensor in the analyser is faulty you will get all the other readings but not O2 or lambda. My guess is that's what the OP meant when he said the guy's lambda sensor wasn't working.
This would also be why not all the info from the analyser was available.
This would also be why not all the info from the analyser was available.
Gassing Station | Engines & Drivetrain | Top of Page | What's New | My Stuff