O2 sensors differences, pros and cons

[Steve Cole]

"When I was at SEMA in November I talked with over a dozen NEW companies now offering wideband tuning applications, this does not include the many already on the market, some of them coming on 10 years. It is also worth mentioning that Wideband sensors are being used more and more in OEM automotive applications, nearly every OE is now using them on core models like the Ford F150."

No doubt about OEM's using widebands in their systems. This has been the case for years. But on which OEM product, the F 150 or any other, is the LSU4.2 sensor used in a primary AFR command role? Correct me if I am misinformed on this, but as far as I know all the automotive OEM's who use this sensor use it in a backup role for emmissions purposes, not a primary AFR command role. What is the reason for this?

This is the big deal that those who are pushing for the cheap Bosch Broad Band sensor not a real Wide Band on a HD keep trying to hide. The Broad Band sensor has been used for years in the OEM automotive world it was developed for them but NOT ONE uses it for primary fuel control! If it was so good every manufacture would dump the switching O2 sensors just for the cost savings alone let along all the electronics and wiring they could drop.

The Broad Band sensor from Bosch is used as a catalytic converter test device and that's it. They are positioned behind the catalytic converter to insure the converter is working properly. The OEM's use as many as 4 switching O2 sensors per vehicle for primary fuel control to meet emissions these days. Any aftermarket people who have to do quality testing and need repeatable results use a good Wide Band system, the cheap ones just do not hold up and are not accurate enough in today's world of testing.

While the cheap broad Band units have there place and we have them as well when it comes time to want to know what's really going on they are not used.

1FSTRK

The corrections on the switching sensors are near zero only because of the limited range of the switching sensor, when you switch to the Bosch Broad band sensor the correction to the output of the sensor over it's usable range are about 28% for fuel mixtures < lambda 1.0 alone and this is just the pressure correction! Then you still need to add in the correction for temperature. So in order to obtain an accurate and repeatable reading you really need to follow the specifications spelled out by the sensor manufacture, Bosch.

The following quotes are straight out of the Bosch specification sheet and as I stated you also need to correct for the temperature.

"Pressure dependency of the sensor signal

A pressure change of the measured gas gives a deviation of the sensor output signal of:
IP(p) = Ip(p0) * p/(k+p) * (k+p0)/p0
The factor k is depending on operation conditions “rich” or “lean” and is for the measuring gas from LSU test bench in section...."

"Sensor characteristic at low or high exhaust gas temperatures

Cold exhaust gas in addition to high gas velocity can lead to a reduced sensor ceramic temperature, when the heater control is not able to keep the constant ceramic temperature. This leads to a deviation of the sensor output signal.
Hot exhaust gas with a temperature above the operation temperature of the ceramic also leads to a deviation of the ceramic temperature and the sensor output signal.

Guide value: a temperature change of the sensor ceramic gives a deviation of the sensor output signal ΔIP/IP of approx. 6%..7% / 100K."


So why is it that none of the aftermarket suppliers of cheap AFR units can bother to do what the sensor manufacture says you have to to make it read properly? It's a real simple deal just add the temperature and pressure measurements and the sensor will operate as specified but until they do the readings you get are NOT going to be accurate and repeatable. The reason they do not do it is because when you add what's needed to get it right they can no longer make a cheap unit. If you use the sensor as specified by Bosch you can expect the results and tolerance they say but if you do not follow the necessary requirements it isn't going to happen!



Jamie

Sorry, but you are DEAD WRONG. The switching sensors are used for primary fuel control. The ECM commands the fuel but adjust/corrects that command from what the SWITCHING SENSORS tells it, that is primary fuel control. There is not one OEM application that uses a Broad Band sensor for this role. If the primary reason to use a broad band is to cover a wider range then why is it that NOT ONE OEM uses it that way?

[ultraswede]

In my car, there is;
broad band sensor, pressure sensor and an exhaust temp sensor..........coincidence..NOT.

[Jamie Long]

Steve, we were all waiting for that response
BTW if you re-read my post you will see that I stated the sensors are not in a primary "command" role, "The sensors are used as a control function"  As we both know the sensors are simply a 0-1V or 0-5V reference which does not command anything, it is how the ECM/ECU interprets/applies this that matters.

[Don D]

 
Here we go again the shootout at the O2 coral. LOL Just funning

[Jamie Long]

This is the big deal that those who are pushing for the cheap Bosch Broad Band sensor not a real Wide Band on a HD keep trying to hide. The Broad Band sensor has been used for years in the OEM automotive world it was developed for them but NOT ONE uses it for primary fuel control! If it was so good every manufacture would dump the switching O2 sensors just for the cost savings alone let along all the electronics and wiring they could drop.

The Broad Band sensor from Bosch is used as a catalytic converter test device and that's it. They are positioned behind the catalytic converter to insure the converter is working properly. The OEM's use as many as 4 switching O2 sensors per vehicle for primary fuel control to meet emissions these days. Any aftermarket people who have to do quality testing and need repeatable results use a good Wide Band system, the cheap ones just do not hold up and are not accurate enough in today's world of testing.

So why is it that none of the aftermarket suppliers of cheap AFR units can bother to do what the sensor manufacture says you have to to make it read properly? It's a real simple deal just add the temperature and pressure measurements and the sensor will operate as specified but until they do the readings you get are NOT going to be accurate and repeatable. The reason they do not do it is because when you add what's needed to get it right they can no longer make a cheap unit. If you use the sensor as specified by Bosch you can expect the results and tolerance they say but if you do not follow the necessary requirements it isn't going to happen!

Could it be that the manufacturers that use "cheap" LSU wideband sensors in their systems may have done the necessary R&D and developed products that incorporate Widebands that not only work properly but very accurately? I guess it is easier for you to say they are trying to hide something or continually state they do not work, all the while there is a wide range of companies/products out there with many happy tuners and end users. Who else is stating these shortcomings outside the Harley community? Are all of these other companies that are using LSU sensors producing products that do not work? We are talking about companies that are industry leaders

[Steve Cole]

Steve, we were all waiting for that response
BTW if you re-read my post you will see that I stated the sensors are not in a primary "command" role, "The sensors are used as a control function"  As we both know the sensors are simply a 0-1V or 0-5V reference which does not command anything, it is how the ECM/ECU interprets/applies this that matters.

So then why are you trying so hard to confuse the issue? Why not just step and state clear as a bell that the Bosch Broad Band sensor is not used in the fuel control role at all in any OEM vehicle. Instead you post

"Both sensors can be effective when used for their intended applications, meaning not just by Bosch or NTK recommendations but by the companies and manufacturers that have tested and incorporated them into their end products."

If you follow the rules as spelled out by the manufacture, I would agree but since they do not............

"As for aftermarket tuning. as a whole you would have a tough time finding a modern air/fuel tuning device that uses a narrowband sensor (again aftermarket, non OEM), while there are many companies that offer wideband based equipment."

Sure there a lot of people offering, still has nothing to do with doing it right, just cheap! Now why not finish the statement and tell people that there is a difference between Broad band and Wide band sensors and the equipment used to properly read them.

"It is also worth mentioning that Wideband sensors are being used more and more in OEM automotive applications, nearly every OE is now using them on core models like the Ford F150."

They have been doing that for over 10 years now, so it's nothing new. Why not just tell the truth and let people know they are used as catalytic converter test devices instead of trying to mislead people into thinking they might be used for fuel control! Bottom line is you are trying to play games with the truth about these sensors.

"Could it be that the manufacturers that use "cheap" LSU wideband sensors in their systems may have done the necessary R&D and developed products that incorporate Widebands that not only work properly but very accurately?"

You cannot do what the sensor is not capable of doing, so you can twist it anyway you like but the outcome is going to remain the same, no matter how hard you try to cover up the truth!

[BVHOG]

I happen to know that a short while back there was a member here who is a tuner(not stroker) that tested the cheap ass broad band sensors against test gas and the resulting readings were within less than one percent. It is entirely up to him if he would like to come forward but I know he personally has no stake in the sales of any tuning devices broad band or narrow band but rather did the test to ensure that his equipment was as accurate as possible. How bout it buddy? post up!!!
Like Jamie stated, many well respected industry leaders rely on the broad band sensors and have been having excellent results with their product.
For another example I had the chance last year to tour the S&S emmisons lab as part of a week of training there and they have a setup that is second to NONE.  When we did the two days on the dyno the company that had the means to use anything they wanted for emissions measurment a hop skip and jump across the street were using dual LM1 units with the cheap ass broad band sensors with the analog system wired into the dyno software. I feel if it's good enough for them then it's more than good enough for me.

[1FSTRK]

The Bosch broad band LSU sensors prices on line range from $48.98-$66.94
The Bosch 12028 narrow band sensors prices on line range from $11.92-$18.95

I do not know why the price of the sensors is constantly being brought up unless it is to repeatedly use the word “cheap” in a derogatory way in an attempt to discredit any one using them.

While the more expensive wide band sensors do exist, because no one uses them, they need not even be brought up

Technically correct or not, the people selling tuning devices that use broad band sensors are not the first or only ones referring to them as wide band and I personally do not believe it is done to miss lead as much as to describe the wider range that it will read

[FLTRI]

I happen to know that a short while back there was a member here who is a tuner(not stroker) that tested the cheap ass broad band sensors against test gas and the resulting readings were within less than one percent. It is entirely up to him if he would like to come forward but I know he personally has no stake in the sales of any tuning devices broad band or narrow band but rather did the test to ensure that his equipment was as accurate as possible. How bout it buddy? post up!!!

If you are referring to me, yes I did do a comparison which continues today as in the past, to assure accurate reading from the O2 sensor mounted in the Dynojet vacuum pump.
That said I have had plenty experience with LSU4 sensors as applied to tuning 2,4,6,8, and 10 cylinder race engines and can confidently report that these sensors are great for tuning on the dyno, but NOT reliable when continuing high heat and vibrations are applied constantly. This is why race engine builders never recommend running and relying on broadband sensors while racing.
Race teams would MUCH rather leave these sensors in while racing because they can more closely set AFR on the "ragged edge" of being "too lean" for best fuel mileage because better mileage require fewer stops to refuel....and that alone can, and has, won many races and championships!

It is for the above reason I feel very strongly that using broadband sensors mounted in the pipes for tuning, then removing them and running the engine in open loop is the best option for these HD bikes, for it is the harsh environment they need to located for "good" O2 sampling that will degrade them faster than when used downstream as with automotive applications.

HTH,
Bob

Bob

[Steve Cole]

The Bosch broad band LSU sensors prices on line range from $48.98-$66.94
The Bosch 12028 narrow band sensors prices on line range from $11.92-$18.95

I do not know why the price of the sensors is constantly being brought up unless it is to repeatedly use the word “cheap” in a derogatory way in an attempt to discredit any one using them.

While the more expensive wide band sensors do exist, because no one uses them, they need not even be brought up

Technically correct or not, the people selling tuning devices that use broad band sensors are not the first or only ones referring to them as wide band and I personally do not believe it is done to miss lead as much as to describe the wider range that it will read

I was the one that called them "cheap" and it was to try and break them out from the expensive real Wide Band sensor.

The truth is there are three types of these sensors and people have tried repeatedly to mix two of them together. The problem has come from the names put on them. In some Bosch documents they are called Wide Band and in other Bosch documents they are called Broad Band sensor. Do I think Bosch did it by mistake... no, marketing at it's finest. Everyone knows Switching or OEM sensor but everyone doesn't know there is a world of difference between a Broad Band and a real Wide Band sensor. With the Bosch documents calling the Broad Band sensor a Wide Band sensor it let's those selling them use the name of the much better sensor, so how is the consumer to know there is a real difference. If people would have stuck to Switching, Broad Band and Wide Band there would have been no confusion but that's not the way it is.

As for your statement that no one uses a real wide band that cannot be further from the truth when you come to the testing industry. I have worked in labs all across the county and a few in Europe and they all use real Wide Band sensors for testing purposes when the measurements need to be correct. I can tell you that HD, GM, Ford, Chrysler and most every OEM worldwide uses them, along with just about every professional racing and development firm out there. If you are trying to get repeatable and accurate results you use them.

Only the aftermarket industry has taken the Broad Band sensor to make inexpensive AFR units with them. The Broad Band sensors is much cheaper than a Wide Band sensor and it has to be for what it was designed to do. The Broad Band Sensor does not have the accuracy or range of a Wide Band sensor. The Broad Band sensor is built from the same base Technology as the OEM Switching Sensor.

If you use the Sensor as it was designed to be used it will do what the manufacture of the sensor says it will do. The specifications have to be followed to get the expected results. So if one was to control the pressure and control the temperature you would not have to correct the output.

[FLTRI]

Just a bit of technical info that gets in the way of opinion:
http://www.ecm-co.com/product.asp?5220#options
HTH,
Bob

[Jamie Long]

So does anyone have any experience and predictions for how long the broadband sensor functions properly in the HD application? Or is it just a crapshoot?

Or more directly to the point, if I buy the newest and bestest tuning system from another thread, should I consider doing a mileage-based sensor replacement or wait until it fails?

I would recommend contacting the manufacturer of the product you are intending on using. You will recieve firsthand information from the company that designed and tested the particular product as opposed to one of their competitors

[BVHOG]

I happen to know that a short while back there was a member here who is a tuner(not stroker) that tested the cheap ass broad band sensors against test gas and the resulting readings were within less than one percent. It is entirely up to him if he would like to come forward but I know he personally has no stake in the sales of any tuning devices broad band or narrow band but rather did the test to ensure that his equipment was as accurate as possible. How bout it buddy? post up!!!

If you are referring to me, yes I did do a comparison which continues today as in the past, to assure accurate reading from the O2 sensor mounted in the Dynojet vacuum pump.
That said I have had plenty experience with LSU4 sensors as applied to tuning 2,4,6,8, and 10 cylinder race engines and can confidently report that these sensors are great for tuning on the dyno, but NOT reliable when continuing high heat and vibrations are applied constantly. This is why race engine builders never recommend running and relying on broadband sensors while racing.
Race teams would MUCH rather leave these sensors in while racing because they can more closely set AFR on the "ragged edge" of being "too lean" for best fuel mileage because better mileage require fewer stops to refuel....and that alone can, and has, won many races and championships!

It is for the above reason I feel very strongly that using broadband sensors mounted in the pipes for tuning, then removing them and running the engine in open loop is the best option for these HD bikes, for it is the harsh environmet
nt they need to located for "good" O2 sampling that will degrade them faster than when used downstream as with automotive applications.

HTH,
Bob

Bob

Nope, not you. Someone else here.
From the above quote am I to read that you are taking the new closed loop bikes out of closed loop?

[TXP]

I too use the gas to monitor the accurracy of the LSU4.2 on my dyno regularly. When it drifts .4 to .5 off its base test number with the gas it is replaced on my dyno. Depending on usage and humidity levels I generally use 3 to 6 sensors per year. How many do you average Bob? Taking this into account, I wonder how long this sensor will actually be effective in an "autotune" application. The other thing to me is the 4.2 is purported to be accurrate to within 1 point. The NB switching sensor measures within 1 point. So by definition isn't the NB switching sensor going to provide more accurate feedback to the ECM? And isn't this the main reason OEM's use the switching sensor for fuel control information instead of the broadband which as stated, is typically used in a backup/check role? I have seen no real evidence that would lead me to believe otherwise.

[FLTRI]

From the above quote am I to read that you are taking the new closed loop bikes out of closed loop?

I was referring to aftermarket broadband sensored EFI systems NOT stock switching sensored EFI systems. Sorry for the misunderstanding.
Why did the MOCO move the O2 sensors downstream on baggers Could it be the sensor housing heat issues created by radiant heating from the pipe right outside the port? Could it be a test for the future O2 sensors and locations for all other models...as we know the MOCO does when trying new systems and parts?
Bob

[FLTRI]

From the above quote am I to read that you are taking the new closed loop bikes out of closed loop?

...but NOT reliable when continuing high heat and vibrations are applied constantly. This is why race engine builders never recommend running and relying on broadband sensors while racing.
Race teams would MUCH rather leave these sensors in while racing because they can more closely set AFR on the "ragged edge" of being "too lean" for best fuel mileage because better mileage require fewer stops to refuel....and that alone can, and has, won many races and championships!
EDITED for addition, spelling, and proper English:
It is for the above reason I feel very strongly that using broadband sensors mounted in the pipes for AUTO-tuning systems work fine, then remove them and run the bike in open loop. IMO, this is the best option for these bikes, for it is the harsh environment they need to be located in for "good" O2 sampling that will degrade them faster than when used downstream as with automotive applications.

[Steve Cole]

For another example I had the chance last year to tour the S&S emmisons lab as part of a week of training there and they have a setup that is second to NONE.  When we did the two days on the dyno the company that had the means to use anything they wanted for emissions measurment a hop skip and jump across the street were using dual LM1 units with the cheap ass broad band sensors with the analog system wired into the dyno software. I feel if it's good enough for them then it's more than good enough for me.

So in the S&S Emissions lab where results have to be right there were NO LSU being used but where the results were not important they used Broad Bands. Seems to me thats a company using there heads and spending money  where it needs to be spent. This is the same reason we have them here to use. We know they are not accurate enough but they also cost 1/10th of the Wide Bands. Use the Broad band to get close then switch to the good ones for finish up.

[FLTRI]

A quote from ECM Company as mentioned in post #35 to help those who are interested:
"...with use, sensors can age and when this occurs, the calibration will no longer be accurate. To restore accuracy, the Lambda 5220 can be used to recalibrate the sensors and this new calibration will be stored in the same memory chip. Pressure compensation (P-COMPâ,,¢) improves accuracy at non-stoichiometric (i.e. λ ≠ 1) and non-atmospheric (i.e. P ≠ 101kPa) conditions. For example, a pressure increase of only 34 kPa can cause an error of 0.58 λ at λ = 3. The Lambda 5220 includes a pressure sensor that measures the exhaust gas pressure to avoid this error."
HTH,
Bob

[1FSTRK]

The Bosch broad band LSU sensors prices on line range from $48.98-$66.94
The Bosch 12028 narrow band sensors prices on line range from $11.92-$18.95

I do not know why the price of the sensors is constantly being brought up unless it is to repeatedly use the word “cheap” in a derogatory way in an attempt to discredit any one using them.

While the more expensive wide band sensors do exist, because no one uses them, they need not even be brought up
Technically correct or not, the people selling tuning devices that use broad band sensors are not the first or only ones referring to them as wide band and I personally do not believe it is done to miss lead as much as to describe the wider range that it will read

I am sorry that I was not clear in the original post above.

Does anyone know of any tuning system that uses wide band sensors permanently mounted on a Harley.
If not then they need not be brought into the conversation. I believe they are constantly brought in to cause further confusion.

[1FSTRK]

The Bosch broad band LSU sensors prices on line range from $48.98-$66.94
The Bosch 12028 narrow band sensors prices on line range from $11.92-$18.95

I do not know why the price of the sensors is constantly being brought up unless it is to repeatedly use the word “cheap” in a derogatory way in an attempt to discredit any one using them.

While the more expensive wide band sensors do exist, because no one uses them, they need not even be brought up

Technically correct or not, the people selling tuning devices that use broad band sensors are not the first or only ones referring to them as wide band and I personally do not believe it is done to miss lead as much as to describe the wider range that it will read

I was the one that called them "cheap" and it was to try and break them out from the expensive real Wide Band sensor.

The truth is there are three types of these sensors and people have tried repeatedly to mix two of them together. The problem has come from the names put on them. In some Bosch documents they are called Wide Band and in other Bosch documents they are called Broad Band sensor. Do I think Bosch did it by mistake... no, marketing at it's finest. Everyone knows Switching or OEM sensor but everyone doesn't know there is a world of difference between a Broad Band and a real Wide Band sensor. With the Bosch documents calling the Broad Band sensor a Wide Band sensor it let's those selling them use the name of the much better sensor, so how is the consumer to know there is a real difference. If people would have stuck to Switching, Broad Band and Wide Band there would have been no confusion but that's not the way it is.

As for your statement that no one uses a real wide band that cannot be further from the truth when you come to the testing industry. I have worked in labs all across the county and a few in Europe and they all use real Wide Band sensors for testing purposes when the measurements need to be correct. I can tell you that HD, GM, Ford, Chrysler and most every OEM worldwide uses them, along with just about every professional racing and development firm out there. If you are trying to get repeatable and accurate results you use them.

Only the aftermarket industry has taken the Broad Band sensor to make inexpensive AFR units with them. The Broad Band sensors is much cheaper than a Wide Band sensor and it has to be for what it was designed to do. The Broad Band Sensor does not have the accuracy or range of a Wide Band sensor. The Broad Band sensor is built from the same base Technology as the OEM Switching Sensor.

If you use the Sensor as it was designed to be used it will do what the manufacture of the sensor says it will do. The specifications have to be followed to get the expected results. So if one was to control the pressure and control the temperature you would not have to correct the output.

99.9% Of the sensors sold go to OEM or as direct replacement where the make,model and year of the vehicle is given. Are you really trying to tell me that Bosch was tring to fool the oem engineers?

[harleyguynv]

I have run an autotune tuner (with the wide/broad band sensors) on my 07 110 SE Ultra for over 25,000 miles without any problems. I also know of several others that have the same unit, one has over 40,000 miles, and none of them have ever had to replace a sensor.  I've read on the forums about dyno tuners replacing these sensors on their dynos. I would think the bikes would be harder on them. I have yet to run into anyone who has had to replace one on their bike. 

[wolf_59]

If you have a properly tuned engine with the fuel that you would normally be using that is within +/-.1 of the ECM/ECU commanded afr is there or would there be a need to run the broad band sensors over the narrow band sensors? 

[FLTRI]

I have run an autotune tuner (with the wide/broad band sensors) on my 07 110 SE Ultra for over 25,000 miles without any problems. I also know of several others that have the same unit, one has over 40,000 miles, and none of them have ever had to replace a sensor.  I've read on the forums about dyno tuners replacing these sensors on their dynos. I would think the bikes would be harder on them. I have yet to run into anyone who has had to replace one on their bike. 

As mentioned many times on this forum about broadband sensors, they do not just quit working, meaning the sensor output changes over time...as noted in ECM-co.com site I referred to and quoted from.
A close monitoring of fuel mileage can show a broadband degrading. Until total failure the rider prolly thinks his/her bike is still tuned to expected AFR results when it may be as much as 1-1.5 AFR off.
Bob

[FLTRI]

If you have a properly tuned engine with the fuel that you would normally be using that is within +/-.1 of the ECM/ECU commanded afr is there or would there be a need to run the broad band sensors over the narrow band sensors? 

The biggest misconception is the assumption the fuel going into the bike is what is expected. What that means is there is a huge variation in the fuels from stations. Age alone degrades fuel quality. The mix of additives vary as well and change with time, temperature, etc.
The biggest advantage for lambda (closed loop) operation is to compensate for the wide variety of fuel quality which includes octane.
Bob

[1FSTRK]

Here is a good link for as long as it last, you can however download and save a copy of it.

http://www.bosch.com.au/content/language1/downloads/Oxgen_Sensor__Cat_WEB.pdf

On pages D4 and D5 it shows how Bosch recommends to mount the O2 sensor and how to make the bung you mount it in. D4 also shows a drawing of proper placement in the pipe on the lower left side. You can also read about exceptions to there suggestions on D5. Then on page D8 - D11 is the section on the LSU 4.2 sensor.

All I can tell you is that we see a much higher failure rate reported back from the field on LSU sensors then we do on the stock OEM sensor. There does not seem to be a pattern on a specific combination that most the failures occur on and seems to be an even mix across different builds. We have seen these same type of failure rates in the automotive industry for a long time when trying to use this sensor (LSU) as a tuning sensor. In the automotive world these were being sold as the newest thing back 10 years ago. Since then most everyone has learned better. While it will get you close at a wider range than an OEM switching sensor it just doesn't hold up longer term. Many race teams have learned over the years to use them for tuning then remove them due to these failures or step up and buy a true Wide Band sensor system. The typical failure mode on a LSU sensor is to begin to drift away from a close reading to one that can be off as much as +/- 1.0 AFR then shortly after they have drifted that far they will completely die. So the big issue comes from the drift of the sensor feeding wrong information back to the ECM to correct from. Since the ECM does not know any better it begins making corrections to bad data.

Things that kill any O2 sensor range from too much heat, oil in the exhaust, improper placement and anything that allows lead into the fuel.