Mixture Map and Closed loop lamda

[ellisd1984]

Would it be possible to have an option for the the mixture map function to take into account correction made from closed loop lamda?  I like the idea of having CLL running as an added safety margin when road tuning and it would be good if the mixture map could look at what corrections are are being made to make permanent changes to the fuel map.

[Laimis]

+1

[mapper]

+1

[iecku.tavea racing]

+1

[TechDave]

So you want us to assume for example that the a 10% richer lambda is a result of the 10% more fuel for a given load site?

[mapper]

17 hours ago, TechDave said:

So you want us to assume for example that the a 10% richer lambda is a result of the 10% more fuel for a given load site?

Lamda Error CL ( effective correction needed for Fuel map) = (Target Lambda (1- CL Trim%/100) - actual Lambda) *100%

For example a rich dip after a lean spike: 

target Lambda: 1.0

Actual Lambda= 0.85

Actual Trim = +10% (before response of CL for rich dip) 

Fuel table correction = (1.0* (1-10%/100) -0.85)*100%= 5%

 

[dx4picco]

I am not sure the fuel correction you wrote here is right.

Assuming you are 10% too rich, which means 10% too much fuel in the mixture, this equals being running at lambda 0.909 and not 0.9 like your formula is laid.

Same, 20% too rich is the same as bringing afr back from 0.833 to 1 and not 0.8,

30% too rich is bringing afr back from 0,769 and not 0,7; etc...

 

[dx4picco]

for me your example should look more like:

For example a rich dip after a lean spike: 

target Lambda: 1.0

Actual Lambda= 0.85

Actual Trim = +10% (before response of CL for rich dip) 

Fuel table correction (%)= (1-(Target / (1+trim%/100)) / Actual))*100 = -6,95%

[Vaughan]

Is this still relevant now that Closed Loop Lambda Long Term Trim tables have been added to G4X? (Long term trims should make the CLL tend towards zero overtime)

If it is still relevant, how would you want it to take care of Dual bank Lambda, Average them?

[dx4picco]

Even though I'm not owning a g4x, I see it as two different things.

Or is the LTTt also now taken in account in the mixture map? is the LTTt same size as the fuel table?

In case where LTTt is same size (or close) to the fuel table and it is taken into account in the mixture map then, CLL integration in mixture map wouldn't be needed.

[ellisd1984]

I think this LTT will do what I was wanting the original request to do so consider this one ticked off.

 

[Vaughan]

20 hours ago, dx4picco said:

Even though I'm not owning a g4x, I see it as two different things.

Or is the LTTt also now taken in account in the mixture map? is the LTTt same size as the fuel table?

In case where LTTt is same size (or close) to the fuel table and it is taken into account in the mixture map then, CLL integration in mixture map wouldn't be needed.

You would use the LTT instead of the mixture map in this case, the LTT acts in a similar way to the mixture map in that it is an average of conditions in that cell over a long period of time but it is being continuously updated and stored.

As a side note one of the possible problems I saw when looking at how to include CLL and CLL LTT values in the mixture map calculations was that as soon as you have two banks you run into issues around which CLL you use  and what you would do if the banks have significantly different values to each other.

 

[Confused]

Is LTT stored per load cell, at the same resolution as the fuel table, or is it just an average applied to the whole table?

What happens when you modify the fuel table? Does LTT get reset for this modified cell? Does the saved LTT still get applied to the modified cell, meaning it's still incorrect?

What about us G4+ people? If STFT is being logged, we still want to be able to use that data in the mixture map.

[Vaughan]

The LTTs have one table per bank and these tables have configurable axes so they can be made to match the Fuel table (with the restriction that GP tables have slightly fewer rows and columns than the fuel table) or they can be made to capture other information (I plan to use them for warmup enrich at some point). If the CLL trim is above a particular threshold the current operating cell of the table is incremented by a small amount and if it the current CLL correction is below a particular threshold the current operating cell of the table is decremented by a small amount.

The LTT table values are applied as a trim the whole time and changes remain between power cycles. Currently to update the Fuel table with the LTT values you need to do it manually cell by cell. (The complication here in doing it automatically being the possibility of more than one bank and the possibility of the LTT looking at different runtimes to the Fuel table)

[LeNainCapable]

Hi Vaughan, 

I've been looking into this to help affinate my VE over time, it's close but not perfect.

So basically over time, CLL will populate the LTT tables by storing and averaging the CLL ?

And, a few hours driving later, I can go back, and if say, the 100kpa, 3000rpm cell in the LTT is displaying +2%, and my V.E. for the same cell is at 80, then I have to add 2% so 81,6 ?

Then the CLL shouldn't correct much and the LTT will be closer to 0 ?

Thank you very much

[Vaughan]

it will effectively average the CLL over time by incrementing when CLL says more fuel and decrementing when CLL says less fuel for each cell in the LTT table/s. Yes the LTT value is added to the fuel value as a percentage so a LTT value of 2 and VE (Fuel Table) value of 80 would act the same as a VE table value of 81.6 and a LTT value of 0.

Over time the CLL values should become closer and closer to 0 assuming you have all engine paramters covered, if you aren't correcting for something (say IAT) then you will see different CLL and so different LTT values at different times. If you do see your LTT values increasing some days but decreasing other days in the same cells then you need to look into what could be affecting the engine that hasn't been accounted for. 

[koracing]

I still think it would be useful to be able to know that the mixture map is being applied with the ST corr. and the LT corr. accounted for.  I'd happily just use the LTT map if there was a way to apply it's learned values directly to the ve table of my choice. Like a copy and paste special mulitply by percent as well as if it was possible to use one or two LTT tables to get the axis values to match the VE table in question directly.