Definition request AE5W101V

cansti · **Posted:** Tue Nov 20, 2018 11:59 pm

I will try and log tomorrow to see if the values are populated at least seemingly correct in some ranges. Thank you for helping!

Would you think discrepancy is in LTV code or in definition file? I am surprised by such high numbers in two cells in high RPM ranges which look so out of scale. It is almost the table picks up some random number unrelated to learning values and tries to use it as a numerical value.

dschultz · **Posted:** Wed Nov 21, 2018 5:55 pm

LTV reads what is defined in the definition file.

cansti · **Posted:** Wed Nov 21, 2018 8:42 pm

That makes sense. So definition file needs some more work. I wish there was a way for me to help out a bit more. I can look through code if need be, but I can't get disassemble the ROM - all I can look at is HEX values. If you want me to go through that with your guidance I would be happy to.
Here is my LTV table from today - if that is of any help. all -125.00 values are there, but value for 11,000-11,999 RPM and 1.6-2.29 g/rev load changed.

I wonder if car with AE5W100V ROM sees the same LTV values, or are they OK? Given how close the ROMs are I would not think they would get better results.

dschultz · **Posted:** Wed Nov 21, 2018 10:11 pm

I suspect the FLKC table data is not calculated correctly, usually it's an offset from the IAM RAM location, but here it doesn't seem to line up.
Best bet is to log FLKC, load and RPM so you can figure out which timing/fuel cell needs adjusting if knocking. More ROM decoding is needed and then we can define parameter E173 for the FLKC data location (which I suspect is 0xFF3304).

cansti · **Posted:** Fri Nov 23, 2018 10:08 pm

Ok, so I did a bit of reading and took your idea as to where E173 is to experiment. I modified my logger file to add following bit in the area where E173 is mentioned:

Code:

<ecu id="B02A907007">
           <address length="4">0xFF3304</address>
</ecu>

and when I had a moment today to try LTV I noticed the numbers moved around in the table. See pic.
The above address is still obviously wrong, but it is close - all the fields moved 3 spots "up" if the top left field in the table is "up". Given that I had gibberish in two fields and -125.00 in others, I may have to move everything the other way by two full rows... and see if I got gibberish coming up in the top of the table if I went too far.
Unfortunately I don't know what 0xFF3304 refers to - I understand it is an address but don't know if I can "see" that address in hex editor looking at the image or I just don't know where to look.

dschultz · **Posted:** Sun Nov 25, 2018 9:24 am

cansti wrote:

Unfortunately I don't know what 0xFF3304 refers to - I understand it is an address but don't know if I can "see" that address in hex editor looking at the image or I just don't know where to look.

You won't find this in the ROM as it is a RAM address. More disassembly is require to sort this out.

dschultz · **Posted:** Mon Nov 26, 2018 12:51 am

I confirmed that address 0xFF3304 is the correct address for the FLKC table in RAM.
I'm pretty sure the Editor def is wrong WRT the RPM rows.

Try updating your ROM def file with this:

Code:

    <table type="2D" name="Fine Correction Rows (RPM)" category="Ignition Timing - Knock Control" storagetype="float" endian="little" sizey="3" userlevel="4">
      <scaling units="Engine Speed (RPM)" expression="x" to_byte="x" format="#" fineincrement="50" coarseincrement="100" />
      <table type="Static Y Axis" name="Fine Correction Stored/Applied RPM Ranges" sizey="3">
        <data> Max Range 1 / Min Range 2 </data>
        <data> Max Range 2 / Min Range 3 </data>
        <data> Max Range 3 / Min Range 4 </data>
      </table>

I believe this applies back as far as the 2015 model year.

cansti · **Posted:** Mon Nov 26, 2018 9:32 pm

Thank you - that seemed to have helped. But it is not the only issue. The definition file I used as basis for "mine" seemed to have that section (table definition for ROM ranges) twice, with two different sizey parameter. So I deleted one, and changed the other to match your suggestion. I managed to get the table to show up without any obvious gibberish, but last RPM range is still showing -125.00.
I went back to def file and found that load also has two tables:

Code:

<table type="2D" name="Fine Correction Columns (Load)" category="Ignition Timing - Knock Control" storagetype="float" endian="little" sizey="4" userlevel="4">
      <scaling units="Engine Load (g/rev)" expression="x" to_byte="x" format="0.00" fineincrement=".01" coarseincrement=".1" />
      <table type="Static Y Axis" name="Fine Correction Stored/Applied Load Ranges" sizey="4">
        <data> Max Range 1 / Min Range 2 </data>
        <data> Max Range 2 / Min Range 3 </data>
        <data> Max Range 3 / Min Range 4 </data>
        <data> Max Range 4 / Min Range 5 </data>
      </table>
      <description>These are the engine load ranges that make up the fine learning correction table stored in RAM. These load ranges, as well as the rpm ranges specified by the 'Fine Correction Rows (RPM)' table, determine how each fine correction value is stored as well as applied.</description>
    </table>
    <table type="2D" name="Fine Correction Columns (Load) " category="Ignition Timing - Knock Control" storagetype="float" endian="little" sizey="7" userlevel="4">
      <scaling units="Engine Load (g/rev)" expression="x" to_byte="x" format="0.00" fineincrement=".01" coarseincrement=".1" />
      <table type="Static Y Axis" name="Fine Correction Stored/Applied Load Ranges" sizey="7">
        <data> Max Range 1 / Min Range 2 </data>
        <data> Max Range 2 / Min Range 3 </data>
        <data> Max Range 3 / Min Range 4 </data>
        <data> Max Range 4 / Min Range 5 </data>
        <data> Max Range 5 / Min Range 6 </data>
        <data> Max Range 6 / Min Range 7 </data>
        <data> Max Range 7 / Min Range 8 </data>
      </table>
      <description>These are the engine load ranges that make up the fine learning correction table stored in RAM. These load ranges, as well as the rpm ranges specified by the 'Fine Correction Rows (RPM)' table, determine how each fine correction value is stored as well as applied.</description>

I deleted the bottom one - but I still ended up with same result. So I am not sure what if anything can be done about -125.00 values. To me it seems as if they are in the wrong spot or scaled wrong (but would scale be off only on that row?). Hope this is helpful to others too.

dschultz · **Posted:** Mon Nov 26, 2018 11:00 pm

Those "duplicate" tables your deleted have different table names, but that doesn't matter as they don't apply here.
The code might work differently than the older FLKC system pre-2015. Maybe they no longer include a row or column for the last range value+.
If you log RPM, Load and FLKC, do you get that -2.80 value in the defined Load/PRM cell range?

cansti · **Posted:** Tue Nov 27, 2018 12:11 am

Hmm ... maybe I missed something but those tables (including the code I posted) *seem* to have identical names and they were one above the other - just as I posted them. I literally copied and pasted the definition file and while I admit being new to this I can't see the difference in those two table definitions including the name (other than one space after name in the second "table"):

<table type="2D" name="Fine Correction Columns (Load)" category="Ignition Timing - Knock Control"

<table type="2D" name="Fine Correction Columns (Load) " category="Ignition Timing - Knock Control"

I am not trying to prove my point - being new to this, I genuinely want to know if I am missing something or is it really the same ....

I have seen -2.75 FLKC only once in my logs, at RPMs a little higher than it showed in the table - around 3,400 RPM - but since then it has not gone over -1.00 according to some logs I have taken. I am not sure if it is real knock or just an oddball - in 3rd gear pulls I don't get anything over -1.00 typically. But that was a while ago and I didn't log in last week or so.

One of the reasons why I wanted to get to the accurate enough def files is so that I can create proper log cfg file and log with SD card - having a laptop in the car is not the optimum for everyday logging :-)

.

dschultz · **Posted:** Tue Nov 27, 2018 10:51 am

cansti wrote:

I can't see the difference ... (other than one space after name in the second "table"):

<table type="2D" name="Fine Correction Columns (Load)"

<table type="2D" name="Fine Correction Columns (Load) "

Bingo!, you are now a definition master :wink:

cansti wrote:

I have seen -2.75 FLKC only once in my logs, at RPMs a little higher than it showed in the table - around 3,400 RPM - but since then it has not gone over -1.00 according to some logs I have taken. I am not sure if it is real knock or just an oddball - in 3rd gear pulls I don't get anything over -1.00 typically. But that was a while ago and I didn't log in last week or so..

Just to confirm that the learned knock is correct, Read LTV to see if a cell is non-zero, if so, then log Load/RPM/FLKC (no pulls just regular driving in that Load/RPM range) to see if that FLKC value reported is truly in the Load/RPM cell reported.
We still need to do a code review to see if there truly is a 5400+ RPM row (or even a 2.30+ Load column).

SD logger config is not dependent on any of the LTV information.

cansti · **Posted:** Tue Nov 27, 2018 1:55 pm

if spellcheck was the only thing required to master that ... :wink:

but at least now I know the file has potentially more duplicates like it had for lod and rpm.

I will try some logging this week to see how it turns out. Understand what you are trying to get now.
I was just uncertain how to create SD logcfg file, so first wanted to at lease confirm I grasp the def file. Can’t use online tool - newest logger def file has not been incorporated yet - so I have to try and piecemeal it together manually. And since I am new to this I wanted to take it in steps.

Will run some logs and let u know how it turns out.

dschultz · **Posted:** Tue Nov 27, 2018 9:43 pm

This is up to date.
http://romraider.com/logcfg/

cansti · **Posted:** Wed Nov 28, 2018 10:29 pm

I could have sworn that only a couple of hours prior to your post online config tool was still using logger v330... thank you!

I did a log of daily drive today; and took LTV shot before and after. In all of them the FLKC was 0.00 at all times. See screenshots. The last row is still -125.00 and that seems out of place. I don't think that last row exists in newer ROMs, or at least in 2017 ones. If I can log anything or help in any other way, let me know.

I have also created logcfg.txt from the online tool, based on my rr log profile and tried logging directly to SD card - with no success. I tried logging to SD card in the morning on my way to work, but no logs got created (trigger should have been a defroster); and on the way back I logged to laptop and managed to get a nice long log. So not sure what the problem with SD card might be. It fits really lose in the Tactrix, but I can't find any recent requirements on the card itself. I got 64 GB one, reputable mfg, so don't know where to look. Google will have to be my friend....

cansti · **Posted:** Thu Nov 29, 2018 10:23 pm

Short follow up in case others get stumped with SD logging. My problem was that I was too careful with SD micro card - you really have to push it into the slot fully. It actually it ends up fully inside the slot. Once I have done that, logging worked.

One interesting point: I had it triggered by rear defogger switch, but it self-created 5 separate log files rather than one large one -even though I didn't stop logging in between. So it seems there is a size limit as to how big a file it can generate at 352 kB. But it is certainly convenient way to log.

RomRaider

Forum rules

Definition request AE5W101V

Who is online