The O2 Sensor Voltage Targets are what controls the Air-Fuel Ratio (AFR).

Regardless of what the map says, the ECU will adjust the
fueling until it reaches the O2 sensor targets. 

This goes against the popular conception, but it's a better characterization of the engine controls.  Any arguments?

sign216 wrote on 05/22/12 at 22:06:00:

The O2 Sensor Voltage Targets are what controls the Air-Fuel Ratio (AFR). Regardless of what the map says, the ECU will adjust the fueling until it reaches the O2 sensor targets.

Not entirely regardless: It won't adjust more than +/- 25% (or is it 20%? I can't remember)


sign216 wrote on 05/22/12 at 22:06:00:

This goes against the popular conception, but it's a better characterization of the engine controls.  Any arguments?

No objections... but what do you mean would be the popular conception?

Raz, by "popular conception" I mean that "it's all about the map."  That the map (inj duration) is the primary control.

It seems that the O2 target voltages are the real control.  That is what determines the engine's AFR.

I bring this up because I want to adjust 4% richer to account for E-10 ethanol fuel.  To change the map for this fuel, I'm guessing I need to slightly change the O2 sensor targets.  Otherwise the O2 sensor will keep pushing me toward the stoichiometery of pure gas, and not the correct AFR for E-10.

Your probe signal is Lambda driven, not ARF. It's measuring the amount of surplus oxygen in the exhaust gas. No need for you to change anything.

Didn't we have this just last week or so? I remember it because I also was wrong with that.

and it's all about maps. The ARF thingy comes afterwards, as a correction only.

(lambda) x (stoich of fuel) = A/F

For gasoline: 0.9 x 14.7 = 13.2

With some ethanol in it, we may end up with 0.9 x 14.1 = 12.7

See? As soon as you replaced the fuel, you already corrected the equation.

I've scoured the internet for papers on the introduction of E10, and the consensus is that O2 sensors will function adequately with E10.  It is only with higher levels of alcohol that the sensors fail to cope.  For engines to run higher alcohol levels, they must have alcohol sensors to detect when the fuel has changed.  Then the ECU can adjust the engine parameters.

Different fuels have different emissions, to include oxygen levels.  Yes, E10's emissions are similar enough so that standard engines will run fine.  Not as well as pure gasoline, but still fine.

So yes, the standard map and O2 sensor will give me adequate results.  Adequate.  Not optimum, not superior.  It just gets by.  The engine will run fine, but that's all.

If I wanted mediocre results, I could stay with the factory settings.  MyECU is about improving the performance.  To get exceptional results.
That's my goal.

There are two separate things that happen when you use E10 instead of pure gas:

1. The theoretical stoichometric ratio goes from 14.70 to 14.13. This is 4%. This is what you have been appearing to talk about all the time. And this one IS corrected by the sensor without you changing the target. So you'd change the map by 4% but not the targets.

2. Even after the compensation in #1, there may be a little more tweaking needed. At this point it has NOTHING to do with stoichometry. I wanted to postpone that discussion until you fully understand #1. You did not appear to do that so far, but I may have been reading you wrong. Your last post may refer to #2. Your earlier posts really seems to discuss #1.

I have seen figures (for some certain unknown engine) that said that to tune it for E85, you first corrected the stoich by adding some 50% more fuel (#1), and then you had to add another couple of percents (#2) for getting max power. For E10 I believe this last correction would be truly neglible but as long as you understand these are separate things I endorse your experiments.

Raz, Although Lambda is defined as "1" the oxygen sensor doesn't read Lambda.  It reads oxygen content, putting out a voltage based on that.  The ECU reads sensor voltage, to adjust fueling richer or leaner, to achieve the target sensor voltage, not the target Lambda. 

This works fine if the ECU is programmed for pure gas and running pure gas.  If it's running something else, it will be at less than optimum. 

Yes, as you said, for E10 the difference is small.  And the O2 sensor will be "close enough."  But "close enough" just isn't close enough for me.  And I think it wouldn't be good enough for your bike, either.

Edit:  I was looking to change the O2 sensor targets because I thought max power for E10 was at a different % of Lambda.  Research shows that although the max power AFR is different, the Lambda % is still the same as gas (0.86), even though for E85 it is markedly different (0.71)
Although E85 really changes things, E10 is still surprisingly close to pure gas (at least with the limited data on hand).  I don't see why, but if that's the data, that's the data.

 

sign216 wrote on 05/28/12 at 10:26:21:

Raz, Although Lambda is defined as "1" the oxygen sensor doesn't read Lambda.  It reads oxygen content, putting out a voltage based on that.

I know nothing about your sensor but my Bosch LSU 4.2 does output a voltage reflecting lambda (between 0.65 and free air), and nothing else. This is regardless if you run gasoline, diesel, ethanol, methanol, LPG, CNG or E85. Without knowing what fuel it is in there, you do know the lambda from reading the sensor voltage. Only if you also know the stoich of the actual fuel used, you can calculate AFR and other things. But what you get is Lambda.

raz wrote on 05/28/12 at 17:09:17:

sign216 wrote on 05/28/12 at 10:26:21: Raz, Although Lambda is defined as "1" the oxygen sensor doesn't read Lambda.  It reads oxygen content, putting out a voltage based on that. I know nothing about your sensor but my Bosch LSU 4.2 does output a voltage reflecting lambda (between 0.65 and free air), and nothing else. This is regardless if you run gasoline, diesel, ethanol, methanol, LPG, CNG or E85. Without knowing what fuel it is in there, you do know the lambda from reading the sensor voltage. Only if you also know the stoich of the actual fuel used, you can calculate AFR and other things. But what you get is Lambda.

Raz, the sensor gives a voltage, and nothing else.  The ECU's program interprets that as a Lambda.  Depending how you've set the ECU, the same voltage can mean different Lambdas.  This is good, as it allows for tailoring the ECU to different fuels.

sign216 wrote on 05/28/12 at 22:15:29:

raz wrote on 05/28/12 at 17:09:17: sign216 wrote on 05/28/12 at 10:26:21: Raz, Although Lambda is defined as "1" the oxygen sensor doesn't read Lambda.  It reads oxygen content, putting out a voltage based on that. I know nothing about your sensor but my Bosch LSU 4.2 does output a voltage reflecting lambda (between 0.65 and free air), and nothing else. This is regardless if you run gasoline, diesel, ethanol, methanol, LPG, CNG or E85. Without knowing what fuel it is in there, you do know the lambda from reading the sensor voltage. Only if you also know the stoich of the actual fuel used, you can calculate AFR and other things. But what you get is Lambda. Raz, the sensor gives a voltage, and nothing else.  The ECU's program interprets that as a Lambda.  Depending how you've set the ECU, the same voltage can mean different Lambdas.  This is good, as it allows for tailoring the ECU to different fuels.

We are really not getting anywhere. I give up here.

It is all too easy to overthink these things. You have a good base to start with. Much easier and more fun to make small tweaks to things like the voltage targets and see what the results are on the ride.