Tuners! Help me understand 02 sensor voltages

[hrdtail78]

I disagree with the ranges you are stating for NB and fuel requirements our Harley's need.  I have heard the cooling myth before, but nobody can explain the latent heat in fuel and how more fuel is actually added when someone targets 13.8 instead of 14.3.  That is .5   Is the latent heat of that much more fuel doing anything?

Sensors have grown up over the years. They are pretty fool proof now if you follow the manufacturers installation instructions and test them in air before putting them in, they have an air test procedure. O2 sensors nowadays are pretty problem free and work as designed.

And what aftermarket system uses "wide band" sensor that follow the manufactures installation instructions?  The sensor manufacture?  Not the tuner's.  I don't want to get too far into that but it needs to be mentioned.

What are the sampling rates of the "WB" sensors you are using?  What are the NB sensor sampling rates?  At what crank position does "WB" sensors poll?  What crank position does the NB poll data?

This is stuff I would look into before claiming something is junk and has no place while tuning an engine.  I use both narrow bands and broad bands while tuning.  They both have their places in the HD tuning world.



Congrates on the new dyno, Joe.

[lonewolf]

And what aftermarket system uses "wide band" sensor that follow the manufactures installation instructions?  The sensor manufacture?  Not the tuner's.  I don't want to get too far into that but it needs to be mentioned.

What are the sampling rates of the "WB" sensors you are using?  What are the NB sensor sampling rates?  At what crank position does "WB" sensors poll?  What crank position does the NB poll data?

This is stuff I would look into before claiming something is junk and has no place while tuning an engine.  I use both narrow bands and broad bands while tuning.  They both have their places in the HD tuning world.

I agree.

[07heri]

Is 0 to 1v the operating range?  I'm seeing voltages way way above, like 3.xx in a log.  If they only read up to 1 volt why would a log show higher.  I figure it's showing rich but they must be able to read higher than 1v if the log is showing me numbers way above 1v.  In theory 0 to 1 sounds right, but it looks like they can read higher than 1v.  No?

Would be nice to see a log of a good tune and see how the voltages fluctuate.  Then compare to a crappy tune.  Anyone have some logs for learning purposes?

[joe_lyons]

Bias voltage put through the ground side

[07heri]

Bias voltage put through the ground side

I'm a little slow.  Can you show some math to explain this?

[joe_lyons]

The signal side of the 02 sensor has two wires. One is for (ground) the other is (signal) the ecm will put what's called a bias voltage through the (ground)side to change the range.

[RideFree]

I disagree with the ranges you are stating for NB and fuel requirements our Harley's need.  I have heard the cooling myth before, but nobody can explain the latent heat in fuel and how more fuel is actually added when someone targets 13.8 instead of 14.3.  That is .5   Is the latent heat of that much more fuel doing anything?

Sensors have grown up over the years. They are pretty fool proof now if you follow the manufacturers installation instructions and test them in air before putting them in, they have an air test procedure. O2 sensors nowadays are pretty problem free and work as designed.

And what aftermarket system uses "wide band" sensor that follow the manufactures installation instructions?  The sensor manufacture?  Not the tuner's.  I don't want to get too far into that but it needs to be mentioned.

What are the sampling rates of the "WB" sensors you are using?  What are the NB sensor sampling rates?  At what crank position does "WB" sensors poll?  What crank position does the NB poll data?

This is stuff I would look into before claiming something is junk and has no place while tuning an engine.  I use both narrow bands and broad bands while tuning.  They both have their places in the HD tuning world.

Using O2 sensors properly is a trifecta of the O2 sensor manufacturers installation and operation instructions. First the exhaust manufacturer installs the bungs at the appropriate distance in the pipe from the exhaust valve and at the proper angle. Then the tuner manufacturer creates a product with circuits that properly use the signals generated by the sensors. Then the user or installer follows their part of the O2 sensor instructions which are probably now rendered in the tuners instructions to air test the sensors and then not get any antiseize on the sensor and tighten appropriately. If all that is done properly per manufacturers specs then you have properly working O2 sensors for a long time until ridden into a lake of course.

Two after market tuners that use WB sensors as stated previously are TMax, Power Commander, and PV with AT Pro. You use their products with an exhaust system that has 18mm bungs installed properly and things will work a whole lot better than NB tuning. I've done both and I'm not talking about set it and forget it dynos. I'm talking about tune it yourself then add a different cam, a different air cleaner and tune it yourself again and real-time monitor both cylinder AFR's running at ~13.8 AFR which is below the NB accuracy range. Use WB!

Relative to WB tuning, NB tuning is junk. We only do it because we are forced to.

You can look up that information and let us know here what you find.

[hrdtail78]

"Using O2 sensors properly is a trifecta of the O2 sensor manufacturers installation and operation instructions. First the exhaust manufacturer installs the bungs at the appropriate distance in the pipe from the exhaust valve and at the proper angle. Then the tuner manufacturer creates a product with circuits that properly use the signals generated by the sensors."

Do you get a free unicorn also?

Both sensors have their problem, and there place.  I don't only want to target 13.8 or richer.  Lower kpa areas I target AFR ratios inside the NB's range.  The frequency of the sample at a known crank position puts ECM read NBs, VE mapping the best way it can be done.  They are not junk.

[rageglide]

Is 0 to 1v the operating range?  I'm seeing voltages way way above, like 3.xx in a log.  If they only read up to 1 volt why would a log show higher.  I figure it's showing rich but they must be able to read higher than 1v if the log is showing me numbers way above 1v.  In theory 0 to 1 sounds right, but it looks like they can read higher than 1v.  No?

Would be nice to see a log of a good tune and see how the voltages fluctuate.  Then compare to a crappy tune.  Anyone have some logs for learning purposes?

The values you see above 1V are the voltage required to keep the O2 in spec.  Drop exhaust temp and the heat circuit needs to kick it up to insure the O2 is running at the most accurate temp.  Raise temps (lean running) the heater circuit voltage will drop.

[joe_lyons]

The ecm uses the bias voltage as a test for how warm the sensor is.  If you notice that when you first start up that your sensor voltage is 5 volts, that is the ecm using its thermometer to see how hot the sensor is and as it warms up either through a heater circuit and/or without(exhaust temp alone) the voltage drops .  Then to keep an eye on the temp it uses the bias voltage (which varies between cals).  Bias voltage and heater circuit are two separate circuits and only work side by side. 
Here is a cool video that does a little bit of explaining
https://www.youtube.com/watch?v=v8XrM-7BuOg

[Karl H.]

Great Video! 

[rageglide]

  

[scott45acp]

What does the O2 sensor actually measure?  If it was oxygen then it could not read afr below 14.6 because the 02 would be used up.  If it was hydrocarbon then it could not read above 14.6 because there would be no unburned fuel.

[rageglide]

What does the O2 sensor actually measure?  If it was oxygen then it could not read afr below 14.6 because the 02 would be used up.  If it was hydrocarbon then it could not read above 14.6 because there would be no unburned fuel.

Wikipedia has lots of good info about how an O2 works.  They don't measure AFR.  This is a good explanation from Wikipedia, better than I could say explain it.

"The sensor does not actually measure oxygen concentration, but rather the difference between the amount of oxygen in the exhaust gas and the amount of oxygen in air. Rich mixture causes an oxygen demand. This demand causes a voltage to build up, due to transportation of oxygen ions through the sensor layer. Lean mixture causes low voltage, since there is an oxygen excess."

http://en.wikipedia.org/wiki/Oxygen_sensor

[07heri]

Is 0 to 1v the operating range?  I'm seeing voltages way way above, like 3.xx in a log.  If they only read up to 1 volt why would a log show higher.  I figure it's showing rich but they must be able to read higher than 1v if the log is showing me numbers way above 1v.  In theory 0 to 1 sounds right, but it looks like they can read higher than 1v.  No?

Would be nice to see a log of a good tune and see how the voltages fluctuate.  Then compare to a crappy tune.  Anyone have some logs for learning purposes?

The values you see above 1V are the voltage required to keep the O2 in spec.  Drop exhaust temp and the heat circuit needs to kick it up to insure the O2 is running at the most accurate temp.  Raise temps (lean running) the heater circuit voltage will drop.

I dont have a heater circuit on an 07.  Just 2 wires.

[joe_lyons]

But it still uses the bias voltage as a check on the temp of the sensor.  Not that it can do anything about it.

[07heri]

The ecm uses the bias voltage as a test for how warm the sensor is.  If you notice that when you first start up that your sensor voltage is 5 volts, that is the ecm using its thermometer to see how hot the sensor is and as it warms up either through a heater circuit and/or without(exhaust temp alone) the voltage drops .  Then to keep an eye on the temp it uses the bias voltage (which varies between cals).  Bias voltage and heater circuit are two separate circuits and only work side by side. 
Here is a cool video that does a little bit of explaining
https://www.youtube.com/watch?v=v8XrM-7BuOg

Good video Joe.  I have a way high voltage reading on the rear.  I also have a high VE number, 108, at idle.  Even though I dont have a heater circuit if I'm way rich it wont allow the 02 to heat up and thus give me a high voltage reading...similar to a bad heater circuit causing a high voltage?

[joe_lyons]

Bingo sir.  Thats exactly what the guy said in the video.  You could have a junk sensor.  Does it become active at any point in time? 

[Tsani]

But it still uses the bias voltage as a check on the temp of the sensor.  Not that it can do anything about it.

So, it is normal to have delays greater than 40 to 60 seconds for the sensor to reach the switching threshold in a sensor with out heaters?

[joe_lyons]

That seems a little slow.  If you rev it up or go for a ride do they come on quicker?

[Tsani]

That seems a little slow.  If you rev it up or go for a ride do they come on quicker?

I would need to go over the logs but I believe so. Stock softail headpipe with CS slip ons. Takes the front about 30 seconds for the front O2 to start dropping off of 5100 and the rear follows about 10 seconds later. About a minute or so later you see them fluctuating "normally".  Heating the engine up faster does change the time some but not by much. I think its a combination fuel cooling issue as well as placement. The time does change with restarts like after a regassing for instance but the engine temps can drop fairly well too so I assume everything is sinking heat. These were new sensors but I know that doesn't mean they can't be faulty to some degree and I will be trying another set. Once everything is warmed up, bike runs fine and as it warms up seems to do well. And at that point it doesn't matter because it is in open loop in that area. But sometimes I wonder if it is not due to ECM coding or is what we see in the data recording actually data from the sensor at the moment it is occurring? I know that may sound like a strange question but is it possible that the ECM doesn't poll the sensors or output O2 data to the data plug until a certain amount of time has passed? I could see programing doing that and was wondering.

Not looking to start a mess about WB v.s. NB or what device is better. Just trying to understand a few things about the sensors and also the ECM because who knows what they do with the coding. But I do know from working PLC's and such that that is a possible scenario.

[Jamie Long]

That seems a little slow.  If you rev it up or go for a ride do they come on quicker?

I would need to go over the logs but I believe so. Stock softail headpipe with CS slip ons. Takes the front about 30 seconds for the front O2 to start dropping off of 5100 and the rear follows about 10 seconds later. About a minute or so later you see them fluctuating "normally".  Heating the engine up faster does change the time some but not by much. I think its a combination fuel cooling issue as well as placement. The time does change with restarts like after a regassing for instance but the engine temps can drop fairly well too so I assume everything is sinking heat. These were new sensors but I know that doesn't mean they can't be faulty to some degree and I will be trying another set. Once everything is warmed up, bike runs fine and as it warms up seems to do well. And at that point it doesn't matter because it is in open loop in that area. But sometimes I wonder if it is not due to ECM coding or is what we see in the data recording actually data from the sensor at the moment it is occurring? I know that may sound like a strange question but is it possible that the ECM doesn't poll the sensors or output O2 data to the data plug until a certain amount of time has passed? I could see programing doing that and was wondering.

Not looking to start a mess about WB v.s. NB or what device is better. Just trying to understand a few things about the sensors and also the ECM because who knows what they do with the coding. But I do know from working PLC's and such that that is a possible scenario.

The operation you are noting is normal for models with non-heated O2 sensors. When the sensor relies on EGT to get up to temp the amount of time it takes closed loop to come online is relative. As far as closed loop itself there are several weighted factors in the code that that affect when the integrator starts rolling, in this regard most importantly run time vs engine temp.   

[Jamie Long]

Here is a chart which references O2 voltage vs. AFR (assuming 14.7 stoich). Note the non linear output of the sensor, specifically what happens to voltage below 14.2 and above 15.0

[Tsani]

That seems a little slow.  If you rev it up or go for a ride do they come on quicker?

I would need to go over the logs but I believe so. Stock softail headpipe with CS slip ons. Takes the front about 30 seconds for the front O2 to start dropping off of 5100 and the rear follows about 10 seconds later. About a minute or so later you see them fluctuating "normally".  Heating the engine up faster does change the time some but not by much. I think its a combination fuel cooling issue as well as placement. The time does change with restarts like after a regassing for instance but the engine temps can drop fairly well too so I assume everything is sinking heat. These were new sensors but I know that doesn't mean they can't be faulty to some degree and I will be trying another set. Once everything is warmed up, bike runs fine and as it warms up seems to do well. And at that point it doesn't matter because it is in open loop in that area. But sometimes I wonder if it is not due to ECM coding or is what we see in the data recording actually data from the sensor at the moment it is occurring? I know that may sound like a strange question but is it possible that the ECM doesn't poll the sensors or output O2 data to the data plug until a certain amount of time has passed? I could see programing doing that and was wondering.

Not looking to start a mess about WB v.s. NB or what device is better. Just trying to understand a few things about the sensors and also the ECM because who knows what they do with the coding. But I do know from working PLC's and such that that is a possible scenario.

The operation you are noting is normal for models with non-heated O2 sensors. When the sensor relies on EGT to get up to temp the amount of time it takes closed loop to come online is relative. As far as closed loop itself there are several weighted factors in the code that that affect when the integrator starts rolling, in this regard most importantly run time vs engine temp.

Thanks Jamie. Makes sense.

[07heri]

Bingo sir.  Thats exactly what the guy said in the video.  You could have a junk sensor.  Does it become active at any point in time?

The sensor only has 3k or so on it, but could still be junk.  I'm betting more on an exhaust leak (messed up my VE's on last couple auto tune runs) and waiting for the gaskets from the dealer.  Once I get it sealed back up I was going to manually drop the high VE at idle down a bunch and see if the sensor responds accordingly.  The last auto tune run ended up at 108 VE at idle on rear and it stinks rich.  If it's a rich condition I should be able to see it start lowering voltages as I drop the VE's.  Assuming the sensor is good and the leaks are tight.  Do you agree?