AFR/Lambda table size, what is required?

[Fred]

When tuning FreeEMS, one will first configure the AFR table to their ideal target contour, and then tune the VE (an actual VE!!!!) till observed AFR matches desired AFR. Thus AFR will be integral to all equations unlike MS variants.

So, I realise that AFR requires much less finess than actually tuning the engine to meet those targets, but how much less?

Originally I had said that it would be like this :

8x8 <= AFR Table <= VE Table

However a few of you suggested that 3 points in a triangle were enough. I tend to disagree about that though. I'll put up some options to vote for, but by all means, suggest your own if it's not present. Bare in mind the assymetry is because of the idle/cranking region being included too, but you might argue that the edge values are enough for that. Certainly a single extra column is all that is required anyway. Above all, please justify your opinions. An unjustified vote is one that might be skewing the results without good reason, and we don't want that do we! :-)

Fred.

[johnd]

the bigger the tables the more defination you can,
as with all variables
--maybe not equally spaced points but user definable to suit engine characteristics
once you are on high engine loads/rpms then you will not need as many points--same as with ignition,as lomg as you can interperlate in all directions then you will end up with a better map ,or enable you to dial out a black hole in the power curve

[Fred]

All tables will have movable bins.

I'm not convinced that we need the extra space for Lambda target, but because it won't be in ram it is OK for it to be big.

Fred.