daymean

iTrader: (2)

How different would those addresses mentioned above be for this 96260009 rom, I am also using the 96260009 rom.

logic

iTrader: (2)

Try here: https://www.evolutionm.net/forums/ec...l-dummies.html

Despite the name, there's information there for a few Evo VIII ROMs as well.

fostytou

iTrader: (15)

This is excellent. I really had no idea how easy this was and just stayed away from it because people said it was inconsistant. I had often wondered if there was an easy way to make it lean out for just a bit, but didn't have the time to research.

I plan to enable this very soon and see what I can get, especially since my fuel table AFR #s are basically dead on for my actual AFR.

03whitegsr

iTrader: (8)

I think the reason people say its inconsistent is because it gives you different AFRs when you tune, particularly comparing 3rd gear to 4th gear pulls. But, after the work others have done, it has become apparent that this is exactly the effect this feature has. It leans out (or richens up if you prefer) based on how long you have been above the activation load.

To demonstrate where I think this idea comes from, let’s create a hypothetical tune. Let’s say you were to tune the car in 3rd gear from 2500-7500 RPM and dial in the main fuel map to give you a solid 11:1 AFR across that RPM range with the feature turned on. In the main fuel table, you would find that if you set the main fuel map to 11:1, the car will run very lean right when you first get on the throttle at 2500 RPM. As the revs rise, you see the AFR slowly taper to a richer condition. You may think that "oh, the values in the main fuel table just don't correspond to actual AFRs and I need to richen up the low end to get the numbers I want."

So you tune the fuel map to get desired AFRs and the low end is very rich and then upper revs are fairly close to the desired AFR numbers. Now you test it out in 4th gear and you find that the car starts off around 11:1 as targeted at 2500 RPM but then tapers to a richer AFR faster (RPM based) then it did before. Now it seems like the car just isn't consistent because you had tuned 3500 RPM for 11:1 in 3rd gear but now it's giving you 10.5:1.

What really happened though is that 3500 RPM happened let's say 3 second after getting on the throttle at 2500 in third gear, but it takes 8 seconds to get there in 4th gear. The "lean spool" counter is still high at 3 seconds of time on throttle in 3rd gear so it is leaning out the car considerably. However in 4th gear, because you have been above the activation load for 8 seconds the "lean spool" counter is much lower and thus it is leaning out the AFR considerably less.

This is why I suggest tuning with "lean spool" disabled and then activating and tuning the lean spool tables after. You will find that the main fuel table is actually VERY CLOSE to real AFRs once you disable the lean spool feature, provided you have the injectors and MAF scaled correctly. With stock injectors and stock MAF scaling, I have found that my main fuel table is within probably 0.2 AFR point across the entire table. That's pretty accurate and is very convenient because I can pretty much type in a desired AFR and I know I'll be VERY close to that number on the first pull.

Mitsubishi did actually use logic behind these computers and it is my opinion that if you come across values that don't seem to make sense, it's probably just that our understanding of that table or scaling is wrong.

Appauldd

iTrader: (22)

You are the man ! !

ace33joe

iTrader: (2)

Great info! I am also very interested in utilizing the lean spool function. I guess playing with load threshold and AFR mapping would work great, but I am also interested in understanding those trailing time (or decay rate) tables.

It seems 94170008 has slight different address offset as I wrote here:

https://www.evolutionm.net/forums/ec...ling-time.html

Anyhow, I hope many ECU gurus can chime in so that we can fully understand lean spool function.

Edit: mrfred got a good post about this decay rate, but do we have any update on this?

https://www.evolutionm.net/forums/ec...uestion-4.html

nitz

iTrader: (3)

it seems as though the LS Load Threshold and LS AFR Mapping addresses have not been identified for 96260009. anyone with the know how care to take a crack at that?

nitz

iTrader: (3)

i've got lean spool disabled and my rc1000 injectors scaled to within 3%, but i find that with increased rpm i actually have to increase the values in the fuel map to keep the car from going very rich. afr's are close to map values early on but don't hold. i've not scaled my maf. would scaling maf make a difference when u're injectors are scaled and trims dialed in?

tapers to 11.0 by 5500 and is flat after that.

LOAD % 300
rpm fuelmap
4000 11.1
4500 11.2
5000 11.3
5500 11.4
6000 11.6
6500 11.8

fostytou

iTrader: (15)

By MAF scaling I believe they are referring to the overall "Airflow/Hz Raw Scaling" table. Using this table you can get your high RPM #s to match without screwing up everything else.

EVO6RS11

Yo fostytou, I am intrigued by the idea whole MAF scaling. I would love to know how it is done.

nonschlont

iTrader: (3)

the info you need is in this thread.
https://www.evolutionm.net/forums/ec...hout-pics.html

EVO6RS11

Yow respect

03whitegsr

iTrader: (8)

From ~800HZ-1600Hz the factory scaling gives me AFRs that match the wideband very well. However, I have noticed that once the airflow maxes out the 1-byte channel at 1600 Hz, the AFR will slowly taper richer, despite a constant AFR being set in the main fuel table. I would bet this is what you are seeing as well.

I would assume that Mitsubishi didn't really intend for the car to run much above 1600Hz and thus didn't scale above that in the MAF tables. This is why I will be rescaling my MAF tables above 1600Hz. I would bet that it would only take 1-2 points above 1600Hz until it reaches the point of maxing out the MAF because it seems to be very linear on how it tapers on the AFR.

Appauldd

iTrader: (22)

So you think it is possible to rescale the MAF to a higher max Hz? I would think by going any higher that you would lose resolution (accuracy).

fostytou

iTrader: (15)

You can just log 2 byte airflow, check the difference between AFRMAP and Actual AFR and then apply an average correction at those frequencies that are pretty far off. The DSM MAF was only very reliable to around ~27,000hz if I remember correctly, but the evo MAF is much more capable.

I think you can't get more resolution by scaling it down because the inaccuracy is a physical property of the design of the MAF rather than a limitation of the digital signal.