Bruce
01-16-2004, 08:06 AM
O2 sensors.
We got two types, Narrow Band, and Wide Bands. Either sense the O2 content of the exhaust gases, and from that figure out what the Air Fuel Ratio is.
The Narrow Band, just has a narrow band of where it determines rich from lean. The Wide band has a rather large region of where it can sense rich from lean.
The NB has a high slew rate, ie it runs from ~100mv to ~700 to cover from about 16.5: to 11:1. The exact millivolts and AFRs vary from sensor to sensor, since they're only used to determine if a mixture is richer or leaner then Stoich. A WB on the other hand, uses a range from 0 to 5v to read an AFR of 20:1 to ~9:1. And while the WB isn't perfectly linear in volts to AFR, it is repeatible.
(Stoich, is about 14.7:1, and is when the chemistry of the combustion reaction most perfectly balances out).
The either sensor to know what the O2 content in the exhaust is they need a reference to guage to, ir ambient air. So without exception, they all allow the sensor to breathe to atmosphere thur the wiring. It's a tiny amount but if you were to sholder all the leads, and heat shrink them, then the sensor would quit working.
The NB also need to reach about 600-800dF to begin operating, and in some applications need to heated since they won't maintain that heat from just the exhaust flow. Since the NBs are designed to do anything more then switch from reading rich to lean, nothing it very critical in that reguard, if the output wanders from EGT changes, or back pressure, then no big deal.
Now on the other hand the WB sensors are designed to be accurate. So they have to operate at temps higher then any normal EGT so that EGT temps don't cause their readings to skew due to EGT changes. In the cas of the Honda WB the sensor operates at ~850dC. Yes, Centigrade. If you look into the end of one when it's *hot*, you can see the innerds glowing bright orange. Since the guts are a small deramic based devise, and it operates at that high of temp., the whole trick to them is heating that sensor up without cracking it. ALOT of research goes into that lil circuit.
And since we're now still having to reference to atmosphere, it too needs to be mounted with yet the consideration of it not seeing TOO much backpressure, so for the TRs you need to mount it downstream of the turbo, and angled slightly sensor tip down. The reason for that is so that if any condensaton forms, it doesn't accumulate on the ceramic piece, if it does, that will crack the sensor in it's next warm up cycle.
A WB reading calibration gases can be button on AFR wise. But, that's not to say they are perfect, all the time, under all conditions. Timing, the actual brew of the gasoline, and poisoning from remote sources can effect them. How much is impossible to say, generally not enough to really worry about, but enough that, IMO, I would totally rely on them all the time. You still need to read the plugs to see what's going on inside an engine.
Now, if they are EBP (Exhaust Back Pressure) sensitive, then why did GM mount the stock one pre turbo?. For cold start emissions. The turbo is a huge heatsink. It would delay the time to light-off of the sensor, and since it's only used for switching, it doesn't matter. But, if you're relying on the stock NB as a mixture indicator, don't be surprised when you blow a set of headgaskets. Just as a note, I've read 12.5:1 and 11.8:1 AFRs with a NB reading of 810mv..
AND, the CARB, isn't totally convinced that NBs even do a very good job of reading O2 content. There is/were further investigation by them as to what a NB really is sensing. It might be a combination of gases, primarily O2, but others do slightly effect it.
We got two types, Narrow Band, and Wide Bands. Either sense the O2 content of the exhaust gases, and from that figure out what the Air Fuel Ratio is.
The Narrow Band, just has a narrow band of where it determines rich from lean. The Wide band has a rather large region of where it can sense rich from lean.
The NB has a high slew rate, ie it runs from ~100mv to ~700 to cover from about 16.5: to 11:1. The exact millivolts and AFRs vary from sensor to sensor, since they're only used to determine if a mixture is richer or leaner then Stoich. A WB on the other hand, uses a range from 0 to 5v to read an AFR of 20:1 to ~9:1. And while the WB isn't perfectly linear in volts to AFR, it is repeatible.
(Stoich, is about 14.7:1, and is when the chemistry of the combustion reaction most perfectly balances out).
The either sensor to know what the O2 content in the exhaust is they need a reference to guage to, ir ambient air. So without exception, they all allow the sensor to breathe to atmosphere thur the wiring. It's a tiny amount but if you were to sholder all the leads, and heat shrink them, then the sensor would quit working.
The NB also need to reach about 600-800dF to begin operating, and in some applications need to heated since they won't maintain that heat from just the exhaust flow. Since the NBs are designed to do anything more then switch from reading rich to lean, nothing it very critical in that reguard, if the output wanders from EGT changes, or back pressure, then no big deal.
Now on the other hand the WB sensors are designed to be accurate. So they have to operate at temps higher then any normal EGT so that EGT temps don't cause their readings to skew due to EGT changes. In the cas of the Honda WB the sensor operates at ~850dC. Yes, Centigrade. If you look into the end of one when it's *hot*, you can see the innerds glowing bright orange. Since the guts are a small deramic based devise, and it operates at that high of temp., the whole trick to them is heating that sensor up without cracking it. ALOT of research goes into that lil circuit.
And since we're now still having to reference to atmosphere, it too needs to be mounted with yet the consideration of it not seeing TOO much backpressure, so for the TRs you need to mount it downstream of the turbo, and angled slightly sensor tip down. The reason for that is so that if any condensaton forms, it doesn't accumulate on the ceramic piece, if it does, that will crack the sensor in it's next warm up cycle.
A WB reading calibration gases can be button on AFR wise. But, that's not to say they are perfect, all the time, under all conditions. Timing, the actual brew of the gasoline, and poisoning from remote sources can effect them. How much is impossible to say, generally not enough to really worry about, but enough that, IMO, I would totally rely on them all the time. You still need to read the plugs to see what's going on inside an engine.
Now, if they are EBP (Exhaust Back Pressure) sensitive, then why did GM mount the stock one pre turbo?. For cold start emissions. The turbo is a huge heatsink. It would delay the time to light-off of the sensor, and since it's only used for switching, it doesn't matter. But, if you're relying on the stock NB as a mixture indicator, don't be surprised when you blow a set of headgaskets. Just as a note, I've read 12.5:1 and 11.8:1 AFRs with a NB reading of 810mv..
AND, the CARB, isn't totally convinced that NBs even do a very good job of reading O2 content. There is/were further investigation by them as to what a NB really is sensing. It might be a combination of gases, primarily O2, but others do slightly effect it.