Page 6 of 46 FirstFirst ... 234567891016 ... LastLast
Results 101 to 120 of 913

Thread: Turbonator LM codebase

  1. #101

    Re: Turbonator LM codebase

    The dipsh*t that wrote that post has his head up his *ss. All he does is talk...blah blah blah. What a moron.

    I've been looking to integrate some of Rob's code tweaks into my BB-based cal, but before I do that I'm trying to finish my "uncal.pl" script so that I can push my table mods back into the source instead of doing it by hand. Kind of a pain correlating everything...needs the LSt file, etc. The other option is doing a binary diff to create a patch and then reapplying it after reassembling. It will work as long as the table addresses and sizes don't change.

    There he goes again. Blah blah blah... Idiot.

  2. #102
    boostaholic Turbo Mopar Contributor
    Join Date
    Mar 2008
    Location
    Arizona Bay
    Posts
    1,097

    Re: Turbonator LM codebase

    Quote Originally Posted by MiniMopar View Post
    The dipsh*t that wrote that post has his head up his *ss. All he does is talk...blah blah blah. What a moron.

    I've been looking to integrate some of Rob's code tweaks into my BB-based cal, but before I do that I'm trying to finish my "uncal.pl" script so that I can push my table mods back into the source instead of doing it by hand. Kind of a pain correlating everything...needs the LSt file, etc. The other option is doing a binary diff to create a patch and then reapplying it after reassembling. It will work as long as the table addresses and sizes don't change.

    There he goes again. Blah blah blah... Idiot.
    Wow, where did that come from?

  3. #103
    Hoosier Daddy?? Turbo Mopar Staff Clay's Avatar
    Join Date
    Dec 2005
    Location
    Terre Haute, IN
    Posts
    5,173

    Re: Turbonator LM codebase

    Quote Originally Posted by risen View Post
    Wow, where did that come from?
    He is the guy that wrote that post.....
    Clay
    1986 GLH-T - Flat black......... learn it, love it, LIVE IT!
    1988 LeBaron GTC Turbo


    SDML Readers Rides
    C/S Registry
    Stuff For Sale

  4. #104

    Re: Turbonator LM codebase

    I'm a little punchy today.... It's nice outside and I want to be out there wrenching, not sitting here pretending to care about work.

  5. #105
    boostaholic Turbo Mopar Contributor
    Join Date
    Mar 2008
    Location
    Arizona Bay
    Posts
    1,097

    Re: Turbonator LM codebase

    Quote Originally Posted by Clay View Post
    He is the guy that wrote that post.....
    I'm still trying to figure out who is the dipshit, and whether or not I should just keep to myself....

  6. #106
    turbo addict moparzrule's Avatar
    Join Date
    Dec 2005
    Location
    Carlisle, PA
    Posts
    7,081

    Re: Turbonator LM codebase

    He was referring to himself, calling himself a dipshit.....

  7. #107

    Re: Turbonator LM codebase

    LOL...I was referring to the post that Clay linked to....

  8. #108
    boostaholic Turbo Mopar Contributor
    Join Date
    Mar 2008
    Location
    Arizona Bay
    Posts
    1,097

    Re: Turbonator LM codebase

    Quote Originally Posted by MiniMopar View Post
    LOL...I was referring to the post that Clay linked to....
    Oh, sorry for a second there I thought you were refering to Clay or I. Man am I confused.

    As a side note I've been using perl quite a bit at work lately perhaps I can give you a hand with that script. I have a perl script that takes a .tbl and .bin and generates a spark map for boost vs rpm (3d map). It's not completed, but it mostly works, and might work as a base for your script. I can send it to you if you want or post it here.

    Although, I would think that the standard unix diff and patch would probably work pretty well for what you need to do. Have you tried it and found it to be a pain?

  9. #109

    Re: Turbonator LM codebase

    uncal.pl is a mess, but at the moment it loads the LST file into a table and then loads the table data from the BIN. The hard part is slipping the tables back into the code as assembly instructions because the code is inconsistent in how it depicts tables (bytes, pairs of bytes, etc). At the moment it stomps on the code and makes it unbuildable. That's when I went to the diff method. I do something like this so that it is also slightly human readable:

    $ xxd mycal.bin > mycal.hex
    $ xxd origcal.bin > origcal.hex
    $ diff -uN origcal.hex mycal.hex > mycal.hex.diff

    Then build the new cal and apply the patch:

    $ xxd newcal.bin > newcal.hex
    $ patch newcal.hex mycal.hex.patch

    Finally, convert the hex dump into binary again:

    $ xxd -r newcal.hex newcal.bin

  10. #110
    boostaholic Turbo Mopar Contributor
    Join Date
    Mar 2008
    Location
    Arizona Bay
    Posts
    1,097

    Re: Turbonator LM codebase

    Quote Originally Posted by MiniMopar View Post
    uncal.pl is a mess, but at the moment it loads the LST file into a table and then loads the table data from the BIN. The hard part is slipping the tables back into the code as assembly instructions because the code is inconsistent in how it depicts tables (bytes, pairs of bytes, etc). At the moment it stomps on the code and makes it unbuildable. That's when I went to the diff method. I do something like this so that it is also slightly human readable:

    $ xxd mycal.bin > mycal.hex
    $ xxd origcal.bin > origcal.hex
    $ diff -uN origcal.hex mycal.hex > mycal.hex.diff

    Then build the new cal and apply the patch:

    $ xxd newcal.bin > newcal.hex
    $ patch newcal.hex mycal.hex.patch

    Finally, convert the hex dump into binary again:

    $ xxd -r newcal.hex newcal.bin
    Sorry, I should have been more clear. I meant running diff/patch on the source code before it gets compiled. Something like this:

    risen@sleipnir:~/turbonator_cal$ diff 87T2_Master.asm 87T2_Master_new.asm > cal.patch
    risen@sleipnir:~/turbonator_cal$ cat cal.patch
    731c731
    < hex 54 10 FF33
    ---
    > hex 54 10 FFFF

    If the lines get moved around in the new cal, you would have to edit the numbers on the left (I believe) of the change line, so it puts it in the right place. I have to do this with configuration files @ work all the time, and it takes some care, but it's saved me a boatload of trouble many times over.

    Although, I think the best would be to csplit the cal into 3 files (one for def's, one for tables, and one for code). Any new cal just split out the tables and cat yours in the middle. That's probably even less trouble, just a csplit and a cat.

    risen@sleipnir:~/turbonator_cal$ csplit 87T2_Master.asm /RomBegin/ "/Begin code/"
    29112
    130441
    361163


    The xx01 file should be tables, to reconstruct you could do

    cat xx00 <your table file> xx02 > output.asm

    I don't know how you're doing it with binary diffs, I'd have broken something long, long ago.

  11. #111

    Re: Turbonator LM codebase

    As long as the table addresses and sizes don't change, the binary diff works. This has allowed me to rebuild the code a couple of times and reapply the tuning that I spent so much time on. The problem is the calibration tools all work on the binary only. Patching the assembly only works if you already have the tables in assembly form.

  12. #112
    Visit www.boostbutton.com... Turbo Mopar Contributor ShelGame's Avatar
    Join Date
    Dec 2005
    Location
    Whitmore Lake, Michigan, Unite
    Posts
    9,918

    Re: Turbonator LM codebase

    Yeah, I've been thinking we need an editor that can write out the table data (DATA, not just the definitions) into the form that the assembler uses so that we can edit tables before assembling the binary. Wish I was a C++ programmer. I may just have to break out my old C++ text books and re-learn it. VB is so much easier, though...
    https://db.tt/SV7ONZpQ
    Rob Lloyd
    '89 Daytona C/S

    2.5 T1 Auto
    13.24 @ 100.5mph
    NHRA #3728 AF/S

    boostbutton.com
    tuning wiki

  13. #113
    Hoosier Daddy?? Turbo Mopar Staff Clay's Avatar
    Join Date
    Dec 2005
    Location
    Terre Haute, IN
    Posts
    5,173

    Re: Turbonator LM codebase

    oh, stop it, you guys are making my head hurt!!!!!!!!!!!!
    Clay
    1986 GLH-T - Flat black......... learn it, love it, LIVE IT!
    1988 LeBaron GTC Turbo


    SDML Readers Rides
    C/S Registry
    Stuff For Sale

  14. #114
    boostaholic Turbo Mopar Contributor
    Join Date
    Mar 2008
    Location
    Arizona Bay
    Posts
    1,097

    Re: Turbonator LM codebase

    I took my perl script that generates a spark map and hacked it apart to make a script that will dump the hex values for a table to your screen. I tested it against a couple builds of the 87T2_Master code and the values looked ok. It took about an hour to get working, as the other code was already working pretty well. You can just run the script for a usage example.

    I can probably use perl2exe to create an executable for all the windows folks (if anyone wants it) or they can get a perl runtime (which can be a PITA). It doesn't currently dump out any FF bytes after the table, but that should be a simple update if wanted. I would like to eventually add this and the config flag to chem2, but I dunno when I'll be able to get to it. Since I can't seem to post attachments here's the code:

    Code:
    #!/usr/bin/perl
    #program to read .tbl and .bin and print table info out in format
    #usable by chemasm for LM's
    #
    #usage: dump_entries($table_offset, $FILE_WITH_CAL_DATA );
    sub dump_entries()
    {
            $off_info = $_[0];
            $CALFILE = $_[1];
            $offset = hex($off_info) - hex(C000);
            #seek to position in bin, and read the number of entries for this table
            seek( $CALFILE, $offset, SEEK_SET );
    
            #read and convert number of entries
            read( $CALFILE, $in, 1);
            $num_tbl_entrs = unpack("C", $in);
    
            #dump number of table entries
            printf("hex %02x\n" , unpack("C", $in));
    
    #       print STDERR "number of entries in " . @$tbl_info[0] . " : " . $num_tbl_entrs . "\n";
    #       print STDERR "offset of table: " . $offset . "\n";
    
            #for the number of entries in the tables
            #read through the next set of bytes and dump them out
            for( $i=0; $i < $num_tbl_entrs; $i++)
            {
                    #read x value from ith table entry
                    read($CALFILE, $in, 1);
                    printf("hex %02x " , unpack("C", $in));
    
                    #read y value from ith table entry
                    read($CALFILE, $in, 1);
                    printf("%02x ", unpack("C", $in));
    
                    #if not at last entry, dump slopes for this entry
                    if ($i < $num_tbl_entrs-1)
                    {
                            #rise
                            read($CALFILE, $in, 1);
                            printf("%02x", unpack("C", $in));
                            #run
                            read($CALFILE, $in, 1);
                            printf("%02x\n", unpack("C", $in));
                    }
                    else
                    {
                            printf("\n");
                    }
            }
    }
    
    my($input_tbl, $input_bin, $maplist);
    my(@raw_tbl_entries);
    
    if ($#ARGV != 1)
    {
            print STDERR "Usage: dump_tblinfo.pl <base filename> <comma seperated list of tables to dump> \n";
            print STDERR "Example: dump_tblinfo.pl 87T2_S1 WOTMAP,HOTMAP,CLDMAP,PEFTBL\n";
            exit(1);
    }
    
    $input_tbl = $ARGV[0].".tbl";
    $input_bin = $ARGV[0].".bin";
    
    #substitute commas for pipes to be used in regex
    $_ = $ARGV[1];
    s/\,/\|/g;
    $maplist = $_;
    
    open(TBL,"<$input_tbl") || die "couldn't open file $input_tbl";
    open(BIN,"<$input_bin") || die "couldn't open file $input_bin";
    binmode(BIN);
    
    # read and parse table
    while( <TBL> )
    {
       @raw_tbl_entries = split(' ', $_);
    
            if(@raw_tbl_entries[12] =~ m/$maplist/)
            {
                    print "@raw_tbl_entries[12] \n";
                    &dump_entries(@raw_tbl_entries[0], BIN);
            }
    }

  15. #115
    Hybrid booster
    Join Date
    Dec 2005
    Location
    Dallas, Texas
    Posts
    741

    Re: Turbonator LM codebase

    Looks like a party going on in here!
    Thanks Cordes for pointing me to this subject from Moparchem.
    I race with an A-413 trans and will see if the v4 code will work or not on my end. I just changed the torque converter only to find it was tighter than my previous 3800 stall and all I can get is 4 psi brake boosting so I have been playing around will timing retard with hardly any luck at all. So, let's play!

  16. #116
    Visit www.boostbutton.com... Turbo Mopar Contributor ShelGame's Avatar
    Join Date
    Dec 2005
    Location
    Whitmore Lake, Michigan, Unite
    Posts
    9,918

    Re: Turbonator LM codebase

    Has anyone else tried it? Any more feedback? I found a couple of small things I changed, but nothing that should cause the breaking up...
    https://db.tt/SV7ONZpQ
    Rob Lloyd
    '89 Daytona C/S

    2.5 T1 Auto
    13.24 @ 100.5mph
    NHRA #3728 AF/S

    boostbutton.com
    tuning wiki

  17. #117
    boostaholic Turbo Mopar Contributor
    Join Date
    Mar 2008
    Location
    Arizona Bay
    Posts
    1,097

    Re: Turbonator LM codebase

    Quote Originally Posted by ShelGame View Post
    Has anyone else tried it? Any more feedback? I found a couple of small things I changed, but nothing that should cause the breaking up...
    Haven't had a chance to, been busy moving into the new garage ATM. I may get a chance this weekend. If not this weekend then I'll try to do something more in-depth with some datalogs next weekend.

  18. #118
    Hybrid booster
    Join Date
    Dec 2005
    Location
    Dallas, Texas
    Posts
    741

    Re: Turbonator LM codebase

    Same for me. I will be able to once this rain stops!

  19. #119
    boostaholic Turbo Mopar Contributor
    Join Date
    Mar 2008
    Location
    Arizona Bay
    Posts
    1,097

    Re: Turbonator LM codebase

    Quote Originally Posted by ShelGame View Post
    Has anyone else tried it? Any more feedback? I found a couple of small things I changed, but nothing that should cause the breaking up...
    I tried v4 today a couple times. For some reason the idle fueling is way off after scaling in chem2. Tried to bring it back in line based upon the fueling guide in chem2 but no go, still really lean. Funny thing was, the idle fuel seemed OK with the doubled pumpeff table (v1), but once the pumpeff table was scaled properly to be 50-150, it seems like idle is really lean once the start to run transition is complete. It's possible that it's unrelated, but I figured I'd mention it.

    Anyways, I tested with the anti-lag both on and off, and it acted the same way. The car revs to the limiter around 3k rpm, then dies off, hits the limit again, and then goes back to idle (and stumbles a little when it gets there). I didn't get a chance to try it with both anti-lag and staging limiter off to see if the car ran alright, maybe next weekend.

  20. #120
    Visit www.boostbutton.com... Turbo Mopar Contributor ShelGame's Avatar
    Join Date
    Dec 2005
    Location
    Whitmore Lake, Michigan, Unite
    Posts
    9,918

    Re: Turbonator LM codebase

    Quote Originally Posted by risen View Post
    I tried v4 today a couple times. For some reason the idle fueling is way off after scaling in chem2. Tried to bring it back in line based upon the fueling guide in chem2 but no go, still really lean. Funny thing was, the idle fuel seemed OK with the doubled pumpeff table (v1), but once the pumpeff table was scaled properly to be 50-150, it seems like idle is really lean once the start to run transition is complete. It's possible that it's unrelated, but I figured I'd mention it.

    Anyways, I tested with the anti-lag both on and off, and it acted the same way. The car revs to the limiter around 3k rpm, then dies off, hits the limit again, and then goes back to idle (and stumbles a little when it gets there). I didn't get a chance to try it with both anti-lag and staging limiter off to see if the car ran alright, maybe next weekend.
    OK Thanks. I'll dive into it again this week.

    EDIT: Already dove into it. I found a problem with how I was calculating the PumpEff re-scale. That was likely the cause of the lean running and the break-up, since it would cause an overflow and an effective PumpEff of ~20-30% instead of 70-80%. That's not enough fuel and it would cut out, same as the rev limiter.

    So, v5 is now posted on the original post. I think (hope?) this fixes it.

    Everyone - please let me know how it works out if you get a chance to try it.
    Last edited by ShelGame; 03-23-2008 at 02:31 PM.
    https://db.tt/SV7ONZpQ
    Rob Lloyd
    '89 Daytona C/S

    2.5 T1 Auto
    13.24 @ 100.5mph
    NHRA #3728 AF/S

    boostbutton.com
    tuning wiki

Page 6 of 46 FirstFirst ... 234567891016 ... LastLast

Tags for this Thread

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •