FWIW. most engineering systems are designed using "engineering standards". In the auto industry these are the SAE (or equivalent(country) standards) like SAE J1349 and J1995 that address temperature and humidity corrections. These are likely incorporated in the ECU algorithms to maintain the tune for proper emmisions.

 

guys if you can dig up my thread on multiple load values then you will get another piece of the puzzle...

 

I can think if a couple possibilities:

- Winterized gasoline (oxygenated)
- Cold air cannot hold as much moisture. Could be that the reduction in water content allows a faster burn rate.

 

Winterized oxygenated gasoline was brought to my attention at sometime earlier yesterday. I had completely forgotten about the difference in gas from winter to summer, and lately I've been having a very minute knock issue at low rpm's. The person in particular I had mentioned this to told me about the differences in gas and how it might have an effect on my knock during the winter and my tune over all in general.

You make a valid point with this. I just did a datalog of the IAT the other night and thought it looked strange lol. It really should be on the intake manifold somewhere to achieve the most consistent results, but who am I to argue with Mitsu's engineering genius. How would be the best way to go about finding what values this table accesses and modifies? I know each MUT table is usually semi-unique between rom ID's, so would logging a specific area and watching it's actions help in determining what it does?

 

I am considering a very basic test to determine what the temp correction table does ... log my car as it is, then make a drastic change to the temp compensation table and see what happens. Very simple.

Currently, my boost curves are a little erratic because I'm letting the error correction table handle everything, so if there is a change in load values with the temp compensation table it might not be as apparent. However, I think this test needs to be done.

I was told yesterday that the temp compensation table modifies the fuel to compensate for weather changes, but I'm still not sold on that. I think the table has to do more than just modify fuel.

 

Temp compesation modifies timing values. do some logs on a hot a nd cold day, and you will see there will be adiference in timing values applied by the ECU.

Just alter these in the specific map. EcuTek can change these.

 

If this is the case it would sort of account for my timing logged at 3 counts of knock being a decent bit different than what the table actually is set to for that range, for some reason it increased timing above what it was set at. But at the same time, decreasing timing decreases power production i.e lower load cell, and we pretty much know from personal experience that during colder months you tend to hit a higher load. So does this table advance timing when it's colder and decrease in warmer temps? Decreasing timing has a tendency to richen up the mixture a hair, and advancement can lean it out. Wouldn't you want to enrich the mixture during the colder weather due to a denser charge? I'm going over the MUT table to see if I can find any patterns, I don't know whether it's relevant or not but according to the evo7 parent base the IAT is part of the fuel category.

 

As I stated earlier, I saw timing advance with colder temps as well ...

I've logged with changes to the temp compensation table and I'm uploading images now ... findings are, well, interesting ...

 

So, What I've done here is logged my current tune and then modified only the Air Temperature Compensation Table and logged again for comparison.

I have 4 log files total. Logs 1 and 2 are with my current temp comp table and will be the baseline for the test. Logs 3 and 4 are with the over modified temp comp table and were recorded in the exact same location as logs 1 and 2. So, log 1 will be compared to log 3 and log 2 will be compared to log 4 since they match each other in terms of geographic location.

The logs are from an Evo IX w/ v.15 ROM. 3" DP and TP w/ 2.75" cat-back. 3.5" FMIC w/ stock piping. Stock BC with modified pill. Stock air box and paper filter.

Logged air flow values are 2-byte. Load values are 2-byte corrected per JB's findings. Boost logged with JDM MAP. AFR from LC-1 reading ~14.8 @ idle.

Logs performed in 3rd gear.

Here is the data:

Ambient Temperature ~ 47*F (~8.5*C)

Temp Comp Table 1 ...



Log 1, location 1 ...



Log 2, location 2 ...



Temp Comp Table 2 ...



Log 3, location 1 ...



Log 4, location 2 ...



Comparison Graphs ...

The baseline curves have lower opacity.

Log 1 vs Log 3 in location 1



Log 2 vs Log 4 in location 2



--------------------------

So, what do you guys think ... it looks to me that the only thing that really changed was the logged air temp. The timing is a little different between the logs, but it seems to correspond with the slight variations in the load curve.

The only other observation I have is that there seems to be a slight variation in the boost:load relationship as the air temp values change.

Just to note, when I witnessed the increased timing values and knock occurrences in colder weather, I was logging the basic 2-byte load and not the corrected 2-byte.

I honestly think this is a forest > trees situation and the temp compensation table simple modifies the intake temp:voltage translation for the ECU. Performing this same test while simultaneously logging the basic 2-byte and corrected 2-byte would be useful I think ...

 

are you logging all 3 load variables?

when you get to this level of detail the timing routine becomes very complicated, I havn't really got the time (now) to go through and make sense of it. It needs someone with a lot of "engine" knowledge to go through and match up bits of code with "what should be happening".

 

I'm currently only logging the single corrected 2-byte load that JB suggested, which I believe you initially found ...

 

So, a quick summary ...

1) Many people report increased knock occurrences in colder weather
2) Many people report reduced knock with an upgraded FMIC
3) Some have reported increased timing values in colder weather
4) Winter/colder air is denser
5) Winter air may have lower humidity
6) The ECU makes adjustments for air intake temps via the Air Temperature Compensation map
7) The Air Temperature Compensation map seems to modify the ECU's registered intake temp only
8) The ECU cannot correct for increased FMIC efficiency due to the location of the air temperature sensor

Any thoughts are appreciated here ...

 

Are any of the ECU gurus able to determine how this table is used or if there is a different temperature compensation table available. there have been other errors in the IX XML definitition as well.

 

That table definitely changes the voltage:temp relationship. I think it was just an assumption that it changed the fuel calculations.

If the table worked differently on the VIII, maybe someone can perform the same test I've done on a VIII and post results.

 

Are you going to show some logs of colder weather? or warmer weather?