I assume ... I haven't changed it ...

Log from this evening ... simplified with ECU load target logged.



The error in load seems to be in the BCLO value ... just as I assumed earlier. Any other thoughts?

 

Could you send me your ROM? Do you still have my email address somewhere?

 

I'll see man ... I'm actually out of town right now. I'll see if I can get my ROM and send it to you. Please PM your e-mail ... I definitely do not have it with me.

 

Okay, Mrfred seems to have found the problem ...

I was logging the temp corrected 2-byte (FFFF6B48) load value instead of the basic 2-byte load value (FFFF6B42). The drop in load was caused by the unseasonably warm temps in my area.

Because of this issue I've realized the benefit of the direct boost control method. I've now tuned the car for my desired boost and now temp changes shouldn't affect boost as much ... only load. My airflow and boost values are right where they need to be now, even though the load is still a little low.

Thanks again everyone for all of your help.

 

This is interesting, the original 2 Byte address that jcsbanks used were from the memory addresses used in the boost cut sub routine, so it would seem that the ECU is using different RAM addressees for different routines. So really we should be logging different loads for different applications, for example the temp corrected load might be best used for timing and fuel, where as the uncorrected maybe best for boost tuning.

I'm not convinced that MAP based boost control is what I want, the load based system seen more elegant, with a greater ability to compensate for temp and pressure variations.

By the way, which sub routines do people use to identify the temp compensated load RAM address?

MB

 

TB,

You got the words right, but the RAM variables wrong. :-) FFFF6B42 is the uncompensated load, and FFFF6B48 is the temperature-compensated load (on a USDM Evo IX).

burgers22,

Temperature compensated load is used by the boost control and MIVEC subroutines. Load and fuel subroutines do not. Temperature-compensated load, as defined by the Mitsu engineers, is not so great to use. As the temp goes up, load goes down. Imagine what would happen to timing and boost if this load were used in these subroutines. The car would add more timing, and try to boost higher as air temperature increased. Not so good. This is one reason why I suggest that people move away from temp-compensated load for boost control.

 

Fixed .. thanks.

While revisiting this thread I now see exactly what you are talking about. I had boost control set to follow 6B42 but I was logging 6B48. My solutions were to:

A) Log 6B42
B) Change the load reference for boost to 6B48

It seems to me that setting boost to follow 6B48 would make sense ... since that's the one that fluctuates with temperature, setting it to follow the defined load levels would make sense. It would basically stabilize that value and make it act like 6B42. But, 6B42 would then fluctuate with temperature ... so there really is no difference.

We really need to just decide on one load value and set EVERYTHING to follow it ... fuel, timing, boost, MIVEC ... all of it. that would take a lot of guess work out of this stuff.

But, now I'm on direct boost control and my target is based directly on the MAP reading, so load is technically irrelevant. I'm still logging 6B48 for my load value since that seems to be the value that my timing map follows. All is well for now ...

 

From your comments, I guess that it is the Mitsu implementation that is the problem.
My thinking was that as the air temperature increases, the air becomes less dense, so more boost is required to deliver the same air mass. From the information in this thread, it suggests that the Mistsu temperature compensation changes the load by to great a degree, is this done via a map or a subroutine?

MB

 

That is what it does, but who wants to dial in more boost when the AIT is goes up? It just causes more knock. Better to stick with the same boost level or perhaps even dial down boost a little with very high AITs.

 

I can see what you are saying, and I don't have enough experience tuning to know if high ambients will make the IAT be so high as to start knock, so am merely thinking ( or posting) out my thoughts. It could be that as the stock car is fairly low on boost and no where near the edge tuning wise, the stock compensation table might be set to take this into account.

If the air comp map is used to adjust the load against temp, it should be fairly easy to set this map flat, or change the slope to your desired tilt.


MB

 

its ****house really, temp+baro corrected load gets used for timing as well, so you get more boost and more timing when its hotter...

i think 48 is baro+temp not just temp could be wrong thou

 

If we were to flatten the air temp comp and baro comp maps, would this have the effect of negating the the variations see during ambient temperature fluctuations and pressure changes. Or are there other factors involved.

tephra, is that MUT 48 you are referring to?

MB

 

He means FFFF6B48 which is temp and baro compensated load. The load for timing is a little bit different when the AIT > 25C. Its not as strongly affect by temperature. Check out this thread:

https://www.evolutionm.net/forums/sh...d.php?t=288214

Best example is in post #16.

 

Thanks mrfred, I see what you mean. Sorry to labor this point, but if we flatten the temp and baro maps, will this negate the effects of temperature and pressure on the loads that use these corrections. As you know I'm pretty limited with what I understand in IDA, so I find making sense of it myself rather to much.

MB

 

FFFF6B42 (USDM Evo IX ROM) is neither temp nor baro compensated, so the easiest solution would be to use the equivalent load variable for your ROM in the timing, fuel, and boost subroutines.