oxigen sensor tutorial

[max]

Very in-depth explanation about the function of an oxigen sensor in an engine, long and short term fuel trims and reset purpose.

This video finally made clear to me how an oxigen sensor works and why it's output produces a variable wave instead of a straight line graph.

In short; the sensor operates within a narrow band and measures the exhaust gas, tells the ecm if the value is too rich or too lean and tries to keep the optimum stoichiometric air/fuel ratio at 14:1.

The ecm then adjusts the short term a/f ratio to maintain this value which is important for the functioning of the catalytic converter.

Over time the ecm learns this trend and produces an internal "map" of values for various engine speeds and other inputs, this is called "long term fuel trim" and is stored in memory to provide a base line for a/f control.

If there is a fault in the fuel system ( vacuum leak ) the ecm re-learns the "map" and changes some values in it, once the fault has been repaired these "wrong" values need to be reset by disconnecting the battery for 10 minutes which does a reset and then re learns a new set over time.( not sure about this but a good possibility )

This may be important info for the ones that remove the battery from the bike for storage!:exclamation:

As i am not knowledgeable enough to give any advice, may well be wrong in what i just stated and may not relate to our bikes i want you to not rely on my explanation, however there may be some info here that explains IF there is some unexplainable subtle behavior change with battery replacement.
Not trying to panic anyone, nothing is wrong, just some careful subtle advice with a little back ground.

If there is a better or more correct explanation out there i am sure we would like to hear it.

 

[GoldOxide_imp]

Sure, if the "user" map is not stored in non-volatile memory (I do not know if it is or is not), then when the battery is restored, factory settings are probably used as the basis of the re-learn sequence and subsequent new "user" map.

 

[Toro5xi_imp]

It’s not only the fuel trims which are lost but also settings related to the IAC, Idle Air Control. I have changed battery’s on cars and they will stall or nearly stall until the idle speed settings are relearned. In the dealer ship we would keep the ECM powered via the DLC, data link connector, to avoid losing settings. Another method is to use alligator clips and an external battery connected to the battery cables when changing out the battery.
The long term fuel trim is accessed via the scan tool and a good indicator as to the health of the engine and engine management systems. The short term fuel trim is designed to switch from lean, low voltage, to rich, high voltage as measured by the O2 sensor so not usually used for diagnostics other than to verify the system is switching constantly from slightly rich to slightly lean as you mentioned for the catalytic converters to eliminate hydrocarbons, carbon monoxide and nitric oxide.
As you mentioned the long term fuel trim percent indicates if the system is adding or subtracting from the hard coded fuel maps to compensate for dirty air cleaners, weak fuel pressure, manifold leak etc. Typically if that percent exceeds 10 percent in either direction the check engine light is set and a trouble code is stored.
Finding and correcting diagnostic codes is not always easy as the code will say if the engine is running lean or rich but the technician has to often eliminate many different possibilities before finding the root cause. That being said I enjoyed driveability work.

 

[jimgl3_imp]

so would it just be wishful thinking to think one could do a tsr and full exhaust with o2 sensor and reset the ecm by battery disconnect and it would relearn with the new intake and exhaust, eliminating the need for dynojet or similar module?

 

[Toro5xi_imp]

That is a good question and I don’t have enough knowledge of emissions work on motorcycles to give a comprehensive answer. The O2 sensor only directs the fuel injector pulse width when the car or bike is running in closed loop operation. I have been told on motorcycles that they operate in closed loop for a small percentage of time in comparison to automobiles, which is more my background.
Just to add more background and indicate why custom fuel maps are designed is that maximum power is achieved at richer ratios than stochiometric or 14.7 to 1. When custom fuel maps are uploaded the O2 sensor must be disabled or the computer will try and adjust the fuel ratio back to stochiometric in essence disabling the custom fuel map.

 

[max]

Toro thanks for all your valuable input, i was hoping guys like you would add to the story, i have also heard the ox sensor is only used at low speed.

I am considering mounting a small voltmeter on the bike so i can see the output of the o2 sensor and see if that is true and having a look at providing the ecm with something like 0.7 volts ( middle of the range ) to simulate the sensor and reduce/eliminate the slightly hesitant behavior of the otherwise lovely engine.

Main reason for posting was the insight it gave me in the workings of the fuel trims and i thought some of us may appreciate the video i stumbled across, any thoughts?

 

[Toro5xi_imp]

Hi Max

The voltage meter hooked up to the O2 sensor is a great idea. There is 1 wire that you can back probe which sends the voltage signal to the ECM that you want to tap into. The wiring diagram will tell you which wire you want to back probe.

I thought the fueling on my 2014 CB was very good so never really dug into it. You sound like you are interested in this stuff. It might be worth looking at the calibration of the TPS, throttle position sensor. I am not specifically aware of it on the CB but I had a Bandit where it was possible to change the position of the TPS so the ECM thought the throttle was open more than it actually was so more fuel was added.

This stuff makes logical sense if you just understand the ECM is taking input from the various sensors, then references a vlook up table to see how long the injector needs to be on to provide the proper quantity of fuel to obtain the 14.7 to 1 ratio we have talked about. We used to put holes in the vacuum hose running to the MAP sensor, manifold absolute pressure, so the ECM thought the engine was under a greater load than it actually was and add more fuel to richen the mixture.

Where it gets tough is that some cars use a speed density system, some rely on the mass air flow sensor, and they all use different sensors. Without access to factory flowcharts or other documentation it is difficult to know where to start. I am sure it has gotten more complicated over the years but it is very logical from a cause end effect stand point. For instance at start up the ECM needs to know the temperature of the air to compute or infer the air density. Denser air needs more fuel to obtain the proper ratio we have been talking about. If the ECT, engine coolant sensor, has failed or is no longer calibrated properly than the ECM will use the wrong air mass quantity and thus add the wrong quantity of fuel.

Sorry for the long winded post I guess I am just trying to encourage anyone who is interested in this stuff not to be intimidated. Once you understand the basics it really makes a lot of sense.

 

[dave]

Thank you Max.

 

[max]

Cheers Dave,

Hi Max

The voltage meter hooked up to the O2 sensor is a great idea. There is 1 wire that you can back probe which sends the voltage signal to the ECM that you want to tap into. The wiring diagram will tell you which wire you want to back probe.

I thought the fueling on my 2014 CB was very good so never really dug into it. You sound like you are interested in this stuff. It might be worth looking at the calibration of the TPS, throttle position sensor. I am not specifically aware of it on the CB but I had a Bandit where it was possible to change the position of the TPS so the ECM thought the throttle was open more than it actually was so more fuel was added.

This stuff makes logical sense if you just understand the ECM is taking input from the various sensors, then references a vlook up table to see how long the injector needs to be on to provide the proper quantity of fuel to obtain the 14.7 to 1 ratio we have talked about. We used to put holes in the vacuum hose running to the MAP sensor, manifold absolute pressure, so the ECM thought the engine was under a greater load than it actually was and add more fuel to richen the mixture.

Where it gets tough is that some cars use a speed density system, some rely on the mass air flow sensor, and they all use different sensors. Without access to factory flowcharts or other documentation it is difficult to know where to start. I am sure it has gotten more complicated over the years but it is very logical from a cause end effect stand point. For instance at start up the ECM needs to know the temperature of the air to compute or infer the air density. Denser air needs more fuel to obtain the proper ratio we have been talking about. If the ECT, engine coolant sensor, has failed or is no longer calibrated properly than the ECM will use the wrong air mass quantity and thus add the wrong quantity of fuel.

Sorry for the long winded post I guess I am just trying to encourage anyone who is interested in this stuff not to be intimidated. Once you understand the basics it really makes a lot of sense.

I thought the fueling on my 2014 CB was very good so never really dug into it. You sound like you are interested in this stuff. It might be worth looking at the calibration of the TPS, throttle position sensor. I am not specifically aware of it on the CB but I had a Bandit where it was possible to change the position of the TPS so the ECM thought the throttle was open more than it actually was so more fuel was added.
Have seen video's of someone doing a similar thing to a ktm, increase the tps "idle voltage" by 0.2 volt and get more grunt, not sure if that bike had an o2 sensor or not that would fight that situation, actually not looking for more HP on this bike, i actually "tootle" around more than anything but i can feel a need to improve the slight hunting it has between 2000 and 3000 rpm which would indicate the o2 sensor playing a part in that range and many have had success with their sensor eliminators.
Although i can put one on and have temporarily done so with the desired effect i am cautious to what might happen with the cat converter and the o2 sensor itself, so i like to find another way to stabilise the ecm fuel control whilst maintaining all inputs in line.

To do this i need a better understanding of what the consequences are, hence the video that i thought might be handy for others as well

Hi Max

The voltage meter hooked up to the O2 sensor is a great idea. There is 1 wire that you can back probe which sends the voltage signal to the ECM that you want to tap into. The wiring diagram will tell you which wire you want to back probe.

I thought the fueling on my 2014 CB was very good so never really dug into it. You sound like you are interested in this stuff. It might be worth looking at the calibration of the TPS, throttle position sensor. I am not specifically aware of it on the CB but I had a Bandit where it was possible to change the position of the TPS so the ECM thought the throttle was open more than it actually was so more fuel was added.

This stuff makes logical sense if you just understand the ECM is taking input from the various sensors, then references a vlook up table to see how long the injector needs to be on to provide the proper quantity of fuel to obtain the 14.7 to 1 ratio we have talked about. We used to put holes in the vacuum hose running to the MAP sensor, manifold absolute pressure, so the ECM thought the engine was under a greater load than it actually was and add more fuel to richen the mixture.

Where it gets tough is that some cars use a speed density system, some rely on the mass air flow sensor, and they all use different sensors. Without access to factory flowcharts or other documentation it is difficult to know where to start. I am sure it has gotten more complicated over the years but it is very logical from a cause end effect stand point. For instance at start up the ECM needs to know the temperature of the air to compute or infer the air density. Denser air needs more fuel to obtain the proper ratio we have been talking about. If the ECT, engine coolant sensor, has failed or is no longer calibrated properly than the ECM will use the wrong air mass quantity and thus add the wrong quantity of fuel.

Sorry for the long winded post I guess I am just trying to encourage anyone who is interested in this stuff not to be intimidated. Once you understand the basics it really makes a lot of sense.

Don't think it was long winded, i am very happy to hear other's opinion especially coming from a knowledgeable source like the settings for the tps etc, all this helps me to complete the puzzle, i have done some programming and know some basics so i can follow all this theory, but i am not "fluent" in programming like csory is, fortunately i do have a wide background in electronics and some good friends are a great resource which enables me to evaluate some pitfalls ahead of time

One day i wondered how a micro processor worked and with the help of a friend made one that you could step the clock step by step with a push button, you could see the bus and adress lines change with every push, fascinating for me to watch it add 2 and 3 and display 5 in binary.
Very basic for some but i found it interesting and it helped me get an insight in some of it, like playing with electronic lego.
very interesting to hear your input

 

[Houtman_imp]

On quick and only question : can you eliminate the O2 sensor when installing a 4 into 4 Pipe Master ? and will the engine run "perfect "

 

[max]

Well, judging from the long timespan other members have had theirs installed it would appear safe to do so.

From what i can understand the 330 ohm resistor is there in place of the heating element to "fool" the ecm in thinking the heater is ok, the low voltage resulting from the lack of input of the zirconia element may produce a slightly richer mixture but that is JUST SPECULATION from me at this time as i have not had the opportunity to test this out but you could do a measurement with a voltmeter on the disconnected o2 sensor to verify it's value.

A valid test to me would be to disconnect the sense wire from the ecm only, measure the value of it as you present a variable voltage to the ecm and see if the voltmeter with the zirconia element registers a change in the richness of the fuel mixture, the sensor is thus in-circuit but only for readback of the mixture, hope that makes sense.

It may also be a good idea to ask the pipemasters guy on advice, he may have another angle on it and has done quite a few.

 

[Houtman_imp]

Thanks Max, I will send them an email.

 

[max]

Houtman i'm still on the case, my little voltmeter arrived yesterday and installed this morning, it sits on the brake reservoir on the handlebar where i can see it whilst driving ( carefully ).

First impressions; at startup it reads 1.00 v ( ? ), when the bike warms up to where the idle revs are reduced to 1300 the voltage drops from 1 volt to 0 volt in a steady pace ( the element has warmed up ).

Whilst driving you can see the voltage move up to .7 volts initially and after the ecm has noticed a steady throttle input it tries to get the voltage down to 0 volts and oscillates between 0 and 0.7 volts constantly and takes from 1 to 3 seconds to cycle up and down, this corresponds with the engine surging slightly but noticeably.

idling and low load produces a 0 volts reading and when the revs reach 4000 rpm the meter reads steady 0.8 volts.
At this point it stays at 0.8 volts and sometimes goes up to 0.85 volts under more throttle.

So looking at this i believe the o2 sensor is operational below 4000 rpm but the ecm ignores it above this value.

It appears that 0 volts represents a lean( ish ) condition and 0.7 volts a richer condition which the ecm tries to correct down regularly but not at a set rate, it appears a bit hap-hazard.

Have not finished my findings but wanted to let you know early since you are careful with the eliminators on your bike.
Remembering that the sensor has a "narrow band" in which it makes decisions so no big difference is to be expected in the fuel mixture.

We'll keep working on this because i want to use some form of eliminator as well.
and also read the excellent thread from Ferret some time ago, lots of good info in there;
http://cb1100forum.com/forum/showthread.php?tid=6988&highlight=eliminator

There are a lot of eliminators out there functioning just fine, thanks Ferret!

 

[Houtman_imp]

Thanks Max , I have put one of the O2 sensor eliminators on my 2013 as I probably put a 4 into 4 on this bike to make it symmetrical looking.
I also ordered 3 more eliminators from the UK to put on my other bikes.
If you check the manuals and check for a malfunction O2 sensor all three years say : bike will function normally . I REALLY appreciate you doing all this as electronics is not my strong side !

 

[GoldOxide_imp]

Max: You are correct in your comments. Sensing exhaust gases is pretty basic on our vehicle and sampling rate is not very high (in seconds). It is just a sample average at a spot where the sensor is in the exhaust path, so the engineers design around this arrangement or restriction. I suppose if the sensors were moved, the reading behaviour would change a bit. The dynamic range (i.e. you said "narrow band") of the sensor is quite narrow, so it is a bit more care is likely used when consuming and vetting the data in the ECM.

I suspect that if a sensor reading goes out-of-range, the ECM will disregard and try to use some sort of default mode and increase carbon footprint to ensure a continued pleasurable ride. I am not sure how it manages if one or two out of four sensors randomly fail.

 

[Houtman_imp]

Some more questions on the O2 sensor eliminator. I had order 5 of them , put one on my 2013 and thought that the bike run "crispier " Today I took my 2017 out for a 100 mile trip up the mountain from 2000 ft to 4000 ft. Had installed two O2 sensor eliminators. On the ride up everything seemed fine but on the way back after about a mile after leaving the check engine light came on and stayed on , I also felt that the bike did not run as well. When I ordered the parts from Redline Superbikes in the UK the website did mention :not tested on the 2017 CB 1100. Could there be different values with the different years with regards resistance ? I am not an electronics expert AT ALL ! I will pull the plugs to see how they look. Hopefully in the next few days I can take my 2014 with the eliminators installed and see how that one will run with them in.
I have to do all my home work before ordering a 4 into 4 , I always say : you can not outsmart Honda and that is what I am doing now ?

 

[max]

Houtman, simple answer is; replace both and see if the cel stays off.

What i found out in the mean time may make sense to you, the sensors do not get their 12 volt supply for the heater directly from the common rail supplied by the red engine cut out switch and it's relay, it is supplied from the ecm as is the 0volt return, it is "modulated" from the ecm to not overheat the heater element which draws 1.5 amps cold and reduces to 0.8 amps when it is a bit warmer but not yet at operating temperature, this is not according to what the electrical diagram shows you.

I started with a 1500 ohm resistor instead of the heater ( which is what is the eliminator is doing ) and that ( 5 times bigger ) value makes the ecm " happy " and the cel stays off, i then installed a 470 ohm resistor and have been riding with it for abt 200 km with good results, the "surging" has disappeared predictably.
The voltage on the resistor measures 1,5 volt and the voltage ACROSS the resistor is 0,14 volt which is a very low and safe voltage.
The ecm "knows" something has happened but is happy to supply a much reduced voltage to the heater supply.

At the same time the voltage on the sense wire to the ecm is a measured 0.45 volt, a nice "middle of the road" value so i have to do nothing about that, the ecm takes care of that side and should produce a good value for determining the fuel air mixture.

If you or someone has a voltmeter you may first check if the resistor is of the same value as the new ones, has broken/ bad solder joint or something else wrong compared to a new one ( the two pins opposite the locking clip).
Second maybe check the voltage from chassis to the resistor ( 1.5 volt in my case ) and across the resistor ( 0.14 volt in my case ), you may have to dig into the insulation to get access to the resistor wires when it is connected to the bike.

On my bike it seems the value of the resistor is not at all critical but it may be different on your model, you could have several different value resistors and experiment ( or have a trusted tech check these values for you ) to see if they have become critical on the newer model, it will be different from mine after all these years and twice the o2 sensors.

Cheers.

 

[peterbaron]

For those, who would like to know.... tempering with O2 sensor will prevent ECU from going into closed loop, causing to increase voltage, which runs the engine richer.
Also, it will restrict the converter from reducing HC and CO as much as it could, increasing pollution levels. Longevity of the convertor will shorten.
Also, you may notice some side effects....

I found this very interesting:

Dr. Dyno explains 02 sensors and closed-loop systems
+
Dr. Dyno explains 02 sensors and closed-loop systems

 

[Houtman_imp]

I checked the plugs and they look fine , I will remove the O2 sensor eliminators on the 2017 and see if it will run like before I put them on.
Also I might take them of my 2014 (have not run that bike yet with the O2 eliminators ) as I do not want the same problems as I did have this afternoon with the 2017. On the 2013 everything was great with the eliminator. (up till when ?) It makes me a little Leary to mess around too much with the 2014 and 2017 and I would only put the 4 into 4 on the 2013.