As they say, a picture is worth a thousand words:



I am not 100% with the algorithm, hell I am not 50% with it, but this should help some people visualise what's going on..

edit - yes that's compressor surge...

 

good to see you working on this. so, the plot shows load tracking with the lower of MAP and MAF calcs, but now for the inevitable questions:

- what is this this load used for - table axes and general load-based control of engine function? is it used for the air flow value in the fuel pulse width calc?

- back when i was doing evo 10 disassem, i recall there being uncomp load, baro+airtemp comp load, etc., just like in the 8/9 disassem. the vals were calced from what i call a master load. is the load in your plot one of the compensated loads, the master load, or something earlier than the master load?

 

Great! More grayhair for me!

 

Does it not follow MAF after 5000rpms?

 

Ah, you're right. Lighting was poor when I looked earlier. Will edit my post.

 

As far as I know everything comes from Load_RAW (or your master load).

My log is of the Load_RAW...

I will check the IPW calcs...

 

So some logging i did during spool with an underscaled MAF showed Load jumping quite a bit. i couldn't work out if it was the compressor, or it was switching from MAP to MAF. I suspected the latter... but I'm probably wrong
higher RPM it always tracked MAF which was too low

D - if you deliberately OVERscale the MAF, do you think it will still track the lower of the two up top (which would then be MAP)?

Will try and did up the logs...

 

should do...

The algorithm is a bit more complicated than the "lower" of the two.

I think the main thing is we can log it and then you can see IRL

 

By chance, last night, I was going through the function responsible for main fuel map lookup... and was wondering about the load-selection code right there. It compares two load variable values, then applies The Chosen One for use by the fuel map scale.

The annoying thing, I thought, was that it didn't slap The Chosen One into any useful (ie. loggable) address. Quite unusual for the Mitsu code - their stuff usually looks very "debug" friendly, in that you can dip in and catch interim calc results in most places.

I was actually planning on rejigging things so it put "the" fuel load in a more permanent spot (kind of like that old Boost Error tweak), so I could log exactly what was going into the fuel map lookup.

Or was I over-thinking things? That has been known to happen occasionally... ahem...

Rich

 

Are you talking about the load lookup routine at the top of the fuel map lookup routine?

 

Yep. I don't have the code in front of me, but it grabs what I presume are two load candidates, compares 'em, chooses one, and drops it in the load-scaling address.

Edit: Just correcting myself: that subroutine was actually performing checks of a couple of other variables (not a comparison between the loads).

What irked me was that it didn't put the chosen one somewhere I could log, straight off the bat! How rude...

Rich

 

You're correct in your assessment. That piece of Evo 10 code is essentially a direct port of the Evo 7-9 SH2 code.

 

dave, any chance i can get you to get a log with the maf 25% overscaled and then another with map 25% overscaled? i think that would give some interesting info on the master load calc without putting your motor in jeopardy. (for those wondering how this can be safe, if the ecu somehow decides to use the higher load rather than the lower, this will cause more fuel to be pushed and cause the ecu to pick more conservative values from the timing map.) perhaps do it at a lower boost and maybe only alter the higher airflow cells so it still runs normally off-boost. i'd do it myself on 3dman1's evo, but atr doesn't doesn't have these loads available for logging yet.

 

Is there a reason why the load switches from following the MAP sensor to MAF sensor in middle of a WOT pull?

What is going on when the ecu switches from the MAP sensor to the MAF sensor? Would this cause an inconsistent AFR where the ecu is switching from the MAP sensor to MAF sensor? If this is a problem, is there any way to stop the ecu from switching to the MAF sensor in the middle of a pull aka Speed Density?

I am a newb at all this buy i love learn. Any the only way to learn is to ask questions. Thanks!

 

Nice work Teph, Sub'd