Also Im not know smart guy on Excel so clipse hit me up on Pm how do that Im jackass when come to that part. lol

 

Tried this out today, seems to be working great for me. I noticed my WOT AFR's have gone a bit rich since it's cooled down, and this fix undid that for certain - back to 11.6-12.0 AFR where I want to be w/ methanol. Not much of a change, but my fuel trims are very close to 0.

Somebody asked before, and no, this change will absolutely not affect your ECU's ability to compensate for weather changes.

 

Thanks tephra for figuring this out! You're a genius!

I have some mixed feelings about disabling LTFT from affecting WOT. On the one hand, it's great to not have this extra factor affecting things. On the other hand, it's possible that the weather compensations don't make up for the cold weather enough, and that without LTFT, AFR's can get too lean as Hiboost experienced when the weather is cold.

The concept of LTFT affecting WOT makes sense - if in closed loop, the car can't achieve the right stoich afr, it will adjust fueling based on a learned percentage to keep AFR's correct despite any long-term issues such as seasonal changes, clogged air filter, fuel differences etc. However, as we have seen, LTFT can vary WAY too much in a single day. It needs to be more stable in order for it to work properly when used to adjust WOT.

One specific concern is that in most areas, they add more ethanol to standard fuel over the winter. I believe (or maybe just blindly hope) that LTFT would adjust for this and keep AFR's safe. My main concern is that without the LTFT affecting WOT, I will need to more closely monitor AFR's. Note that I have a wideband but it is not permanently installed. Anyone else have any experiences with disabling LTFT in WOT? I guess it's hard to say because who knows what will happen to AFR's (without LTFT) in the summer when it heats up again.

 

I did some more testing on the way back from work, just richening up my tables .2-.3 AFR ended up putting me back to ~11.7 range on average for 1st- 4th. The most sensitive gear to AFR changes is 5th just as full boost kicks in at 4k on the highway with high load, all seemed fine with 12.0 AFR's. I have noticed that right after a reset that doing a 5th gear romp at the boost threshold right before everything has learned is likely going to cause knock at higher boost levels until the car has learned airflow a bit, one of those odd quirks about the Evo X.

I also noticed that I may be right on the edge of running too lean during spoolup as well, I may try and increase the threshold of where the LTFT kicks in at partial boost but will always be absent at WOT. Is there a table we can make to give us a rough guess of l what hertz airflow = Load assuming 5000 RPM's or so?

300 Hz = 100% Load
?? hz = 120% Load
?? hz = 140% Load
?? hz = 160% Load
?? hz = 180% Load
?? hz = 200% Load
?? hz = 220% Load
?? hz = 240% Load
?? hz = 260% Load
?? hz = 280% Load

I probably would want to test if crossing over at 180%-200% load would work well or not. If I had to guess that would be 800 - 1000 Hz maybe?

 

Chris - Log the memory address of MUT1A - that's MAFHz, formula is x*6.25

 

Will do!

 

That sounds like a good idea to have this work at like 180-200 load. Rather than work at part throttle as well. Since the AFR doesn't play such a large issue at part throttle especially if your injectors are tuned/scaled correctly and pretty close to 0.

 

I get about 1100Hz @ 280 Load.

 

It depends on the RPM you're talking about...

At 4000 rpm, I have 275 load and 1087 Hz.
5000 - 257 and 1263 Hz
6000 - 227 and 1313 Hz
7000 - 194 and 1325 Hz

 

Thanks for those data points guys!

I decided to test 600/593 Hz crossover combo and so far that seems to be in the 8-12 psi range at 140-160 load or so on my car depending on RPM. That should let the car fiddle with it's LTFT adjustments based on gas quality at light boost and then when the car is well into open loop it will switch over to no LTFT adjustments and just go by my AFR map. I'll have to report back after some time behind the wheel since I need to duplicate that light boost situation where it felt like it was going too lean before. I blame this winter blend gasoline we have here now.

 

That's a good place to start about 140 load.

 

Mine is identical to this in my 55590006 ROM from my 2010 Ralliart.

The compensations from LTFT explain my f'ed up results at the track last time I was there! I was running way rich, so did a fast re-tune... after the re-tune (re-set fuel trims) I was running way lean... put the original tune back in, and was at the correct A/F ratio's.... I have high hopes of this improving consistancy... just wish I knew about this before my dyno runs (way rich).

Keith

 

I would check that your injector scaling and latencies keep you within +/- 3% for consistency.

 

So, #1 I need to look up the address in my own MUT table, and #2 put 23 in front of what ever number I find... I think I got that part now... but I have no idea what #3 means...

In my case the addresses are:

MUT0F = STFT. . . . . . . . . . . . . . . 808A21
MUT0C = LTFT_Idle . . . . . . . . . . 80861B
MUT0D = LTFT_Cruise . . . . . . . . 80861A
MUT0E = LTFT_High . . . . . . . . . . 8089FD

So I would use 23808A21 for STFT and so on for my numbers??? Still not sure what #3 means in this...


Keith

 

In evoscan, you have to specify returnbytes equal to 1 or 2 in each formula. I "think" but could be wrong, that you just look at the last digit in the address to determine if it's odd or even.

808a21 -> 1 which is odd
80861B -> B which is 11 which is odd

These are hex addresses, so A = 10, B = 11, C= 12, D = 13, E = 14, F = 15.

The part I'm not sure about is, if it'd odd, do you request 1 or 2 bytes? I would check other formulas in the evoscan data.xml for examples to determine whether you should specify 1 or 2 bytes for odd versus even addresses.