Someone please explain the O2 closed loop bias voltage to me

[rolls]

So I understand how the O2 circuit works, the ECU injects a high impedance voltage and the O2 sensor then sinks it low (0.1v) when lean and pulls it high (0.9v) when rich. Stoich is around 0.45v. I'm unsure what the actual bias voltage is however I'm guessing it is irrelevant here.



This table is what confuses me, is this table indicating the ECU is shifting the bias voltage by -0.003V at 0.2 load at 700rpm? Eg it is shifting the bias voltage slightly negative to fool the ECU into thinking the car slightly leaner than it really is?

Would this mean if I put 0.003 instead of -0.003 it would make the car richer? I read the help comment but it isn't thorough enough.

There is also this second table, which map is actually used?

[murfie]

Bias voltage is applied to electronic devices like microphones that require power to work but that power is not the signal that is being read. It is the electronic operating point also called offset.

This is the same for narrow band oxygen sensors just in a slightly different way. the voltages you brought up are not manipulated by the O2 sensor but actually generated by the sensor when at sensors operating temperature. The oxygen sensor cant produce a signal until it reaches operating temperature. The draw back to narrow band sensors as everyone knows is they go back and forth from rich to lean constantly. This is due to the difference of oxygen levels across the element needed to produce a current flow. Wideband sensors solve this with an oxygen pump.

The pump consist of a heated cathode and anode to pull some oxygen from the exhaust into a gap between the two components. Wired together with the narrow band element a certain amount of current is required to maintain a balanced level of oxygen in the gap. That amount of current required to maintain this balance is directly proportional to the oxygen level in the exhaust.

stoich(.000) would be the certain amount of current to maintain a balanced level of oxygen. The modification to the bias is needed most likely due to the velocity of the exhaust gases and the efficiency of the pump to pull in oxygen.

[rolls]

Ok so the O2 sensor does not pull the signal up or down, the bias voltage is simply the ground supplied to the O2 sensor. By biasing it up or down you shift the voltage the ECU reads? By doing this you effectively shift the stoich point lean or rich slightly.

Eg you shift the ground from 0v to -0.003v, this means when at stoich the O2 sensor is generating 450mv the ECU will then read a delta V of 453mV (450- -0.003), eg slighly rich causing it to pull fuel early. This means negative values make the car run leaner, positive make it run richer.

If this is the point of the maps why not just make the AD converter just look for 453mv in the first place and not bother shifting the ground signal? The fact they haven't done this makes me think this is not actually why they are changing the bias voltage.

This map I showed is for a narrowband only vehicle, it does not have a wideband. Do you know why it has two maps for effectively the same thing, how would it choose which map to use?

[murfie]

the numbers are very small and really don't change the signal that much. Zero the table out if you so choose to. Its probably something they added in to help with emissions/ fuel economy. My car has widebands and its table has a slightly higher resolution that shows a bigger pattern. lean in low rpm/ load and rich in high rpm/ load. again lean and rich when being extremely precise.

o2bias.PNG

[rolls]

the numbers are very small and really don't change the signal that much. Zero the table out if you so choose to. Its probably something they added in to help with emissions/ fuel economy. My car has widebands and its table has a slightly higher resolution that shows a bigger pattern. lean in low rpm/ load and rich in high rpm/ load. again lean and rich when being extremely precise.

o2bias.PNG

So why would they use two tables? What is your guess at which one is actually used?

Also why manipulate the bias voltage when they could simply vary the A/D count they expect to see for "Stoich", you remove an analog output that way and have the same amount of ECU code. Or they could just add a variable injector/air mass offset in those cells instead. Varying the bias voltage just seems odd which makes me think I'm missing something.

I just want to understand why they did it just because it seems odd.

[murfie]

I believe one table would be for the front O2 and one for the back o2.

This patent describes what we are talking about and what you think they should have done.
http://www.google.com/patents/US5255512

Manufactures don't always get to do things the best/ most sensible way, they get to do things the way they have the patent for.


read the description of the fore aft switch. Your car may not have it in hptuners yet.

fore-aft.PNG

[rolls]

Ah back and front makes sense. I'm guessing they use some sort of hardware comparison then and that is why they do it that way.

[Bugasu]

What you have there is the feedforward portion of the bias correction to the front O2.

The Fore-Aft Oxygen Sensor Control uses readings from the rear O2 to help correct the front O2, all in the goal of better A/F control for better emissions.
The system has a feed-forward component you see there, as well as a feedback portion (proportional and integral) based on the expected voltage of the downstream O2 sensor. It's not actually biasing any voltages, the correction is entirely within software.

There's quite a few parts to the bias calculation, not all of them exposed in every strategy for HPTuners. Some for inferred catalyst temperature, inferred catalyst age, and so on.

Regardless, without a second calibrated O2 sensor, you have no reference to really know what the offset should be. The bias is only applied during Stoich Fuel Source, all other times rely on the default UEGO ASIC calibration procedure.

[rolls]

The bias is only applied during Stoich Fuel Source, all other times rely on the default UEGO ASIC calibration procedure.

So if the second O2 sensor has been removed and disabled in the software, would adjusting that bias map simply affect stoich at idle/cruise? Or would it ignore the table entirely if you've disabled it. I guess the only way to know is to test.

What is UEGO ASIC and when is it used? Are you implying there are two different modes of closed loop O2 loops?

Appreciate the info.

[Bugasu]

So if the second O2 sensor has been removed and disabled in the software, would adjusting that bias map simply affect stoich at idle/cruise? Or would it ignore the table entirely if you've disabled it. I guess the only way to know is to test.

What is UEGO ASIC and when is it used? Are you implying there are two different modes of closed loop O2 loops?

Appreciate the info.

Yes there are two different loops, an inner loop based on the upstream sensor (which is a wideband if you're tuning anything recent), and an outer loop based on the downstream sensor.
If you disable FAOSC or set the rear O2 to None, the outer loop is no longer used. This includes the bias map.

The primary O2 loop still has its re-calibration procedure, but what I don't know is just how accurate it'll be without the rear O2. That may be a question for Eric.

[pontisteve]

There is also a diagnostic bias voltage. GM uses 450 mv. When the O2 sensor is cold, it has full resistance on the signal wire. As it warms up, it becomes somewhat of a variable resistor. GM shoots 450 mv into the O2 signal wire (from the PCM). It's a weak voltage, and even hooking up a multimeter to it to read it can cause the voltage to drop some. Once the O2 warms up enough to start building voltage, the diagnostic bias voltage goes away.

If you look at a graph or scantool on a totally cold engine of O2 voltage, you will see it. The O2 voltage will be exactly 450. This is not coming from the O2 sensor, but rather the computer. In probably 15 seconds, the sensor gets hot enough to produce a little voltage of it's own and the diagnostic bias disappears. If you were to unplug the O2 sensor and measure voltage on the O2 signal wire, you would see it.

The bias in the table you were asking about skews how the O2 sensor reads. If you look at the short term fuel trims in closed loop and see the saw-tooth pattern, you will notice the uneven rich and lean shapes. It's my belief that the uneven shapes are the result of the O2 bias. It has the net effect of skewing the air/fuel rich or lean a little, in an attempt to average stoichiometry.

[rolls]

Interesting info, thanks.