Page 70 of 72 FirstFirst ... 20606566676869707172 LastLast
Results 1,036 to 1,050 of 1070

Thread: new $EE tuning thing!

  1. #1036
    Fuel Injected!
    Join Date
    Mar 2013
    Posts
    1,469
    The idea is to set cell number and maf value, than press the button once to send the command. The duration of the send command can be until the button is pushed or for a predefined time.

    This will be for setting the ram maf table.

    There will be another global enable switch that will make pcm switch between ram and rom table. It will be also independant of the ram table set option.

  2. #1037
    LT1 specialist steveo's Avatar
    Join Date
    Aug 2013
    Posts
    4,002
    if you just want a clickable send button, you shouldn't use the toggled signal, you should use the clicked signal. toggled is usually for check boxes

  3. #1038
    Fuel Injected!
    Join Date
    Mar 2013
    Posts
    1,469
    I figured that the hard way.

    Here is the latest source code.
    I got some issues to clear with your help and add some new features.
    One of the main things is to send m4 message in two different lenght, based on some of switches in the controller window. Now it got really huge 24 bytes.

    Other issue I got is, 3 different buttons control one bit. When one of the buttons is unclicked the bit is cleared even though the other two buttons are active. I need the bit cleared when all buttons are off.

    The same issue with greying some of the controls because of conflict in m4 message. Some become active when they are not supposed to.

    A nice feature to add is some mode 3 datastream in the controller window. A must for the closed loop controls. So you know the final value that is working the best.
    Attached Files Attached Files

  4. #1039
    Fuel Injected! Terminal_Crazy's Avatar
    Join Date
    Oct 2015
    Location
    Lancashire England
    Posts
    410
    Evening everyone

    I've been out playing with kur4o's 4.90 version.
    I'd just like to say excellent work sir, the individual fuel trims work well.

    However i do have some issues.
    Currently running OLSD as i'm thinking this is the cause of my problems in CL.

    I think my problem occurs because the RHS O2 sensor is sat over 1 pipe of the Mac Mid length headers rather than in the collector, and it sees 1 strong pulse followed by some weaker/mixed ones.
    I haven't checked/can't recall on the LHS where it's sat.

    Watching the graph of the O2mv, the LHS is fairly flat, the RHS sits slightly lower with a higher rate of much leaner spikes.
    RHS BLM's were always higher and i was running the individual fuel trim about 4-5% higher on all rhs cylinders at idle to even out the BLM's..

    car is running idle at around 14.7 which is much lower that it has been.
    Using the fuel trim adjustments tonight altered the overall AFR but didn't really show much change in the spikeyness of the RHS O2's.
    I maxed each out in turn then had all 4 maxed on the rhs, it improved but AFR dropped to low 14's.

    I'll have to double check which cylinder the O2 is sat in.

    I know I've asked before but there doesn't seem any way to correlate which cylinder the data log is referring to so i can reference any changes.

    Over the weekend i tried tracing a data log to work out which cylinders the log represented but again nothing that made much sense.
    As stated previously, Gary Dougs list of the Cylinder ID differs from Kur4o's. The sequence is the same sbc sequence but shifted out.


    Idle is 825rpm / 60 == 13.75 records per second ....... we are logging 10ish per second.
    Target AFR 14.7

    Actual CYL
    ID G.Doug Kur4o
    4 2 1
    8 1 8
    5 8 4
    8 4 3
    6 3 6
    8 6 5
    7 5 7
    8 7 2
    Actual Cylinder
    Gary Kur4o
    log time CylID Lo2 Ro2 AFR Doug
    #500 47.1 67 892 417 14.96 3 6
    #501 47.2 77 888 755 13.84 5 7
    #502 47.3 86 879 825 14.60
    #503 47.4 67 906 892 14.78 3 6
    #504 47.5 86 901 852 14.13
    #505 47.6 57 906 861 14.07 8 4
    #506 47.6 86 901 870 13.78
    #507 47.7 86 901 795 14.55
    #508 47.8 67 892 830 14.07 3 6
    #509 47.9 86 888 715 13.6
    #510 48.0 57 892 835 13.42 8 4
    #511 48.1 86 630 892 14.55 4 3
    #512 48.2 86 897 866 15.08 3
    #513 48.3 67 888 897 15.90 3 6
    #514 48.4 86 910 879 15.25
    #515 48.4 57 928 808 15.08 8 4
    #516 48.5 86 906 786 14.01
    #517 48.6 86 773 808 14.72
    #518 48.7 86 852 839 14.72
    #519 48.8 86 586 870 14.78
    #520 48.9 47 746 826 14.37 2 1
    #521 49.1 86 915 533 14.37 1 8
    #522 49.1 57 924 839 14.31 8 4
    #523 49.2 86 919 746 14.60
    #524 49.3 47 932 733 14.49 2 1
    #525 49.4 86 924 741 14.66
    #526 49.5 77 928 879 13.96 5 7
    #527 49.6 86 928 848 14.84
    #528 49.7 67 928 852 14.66 3 6
    #529 49.8 86 924 417 14.78 5
    #530 49.9 86 924 506 14.37
    #531 49.9 77 924 182 14.60 5 7
    #532 50.0 47 924 160 14.25 2 1
    #533 50.1 86 928 164 14.01
    #534 50.2 77 910 164 14.25 5 7
    #535 50.3 57 804 559 14.19 8 4
    #536 50.4 67 892 835 14.78 3 6
    #537 50.5 86 897 737 14.25
    #538 50.6 57 901 848 14.31 8 4
    #539 50.7 86 626 888 14.90
    #540 50.8 47 502 568 14.84 2 1
    #541 50.9 86 764 790 14.43
    #542 51.0 86 786 861 14.68
    #543 51.0 47 888 786 14.55 2 1

    I was actually trying to work out how many logs behind the wideband was compared to the O2 data but was again inconclusive.


    Thanks
    Mitch
    '95 Z28 M6 -Just the odd mod.
    '80 350 A3 C3 Corvette - recent addition.

  5. #1040
    LT1 specialist steveo's Avatar
    Join Date
    Aug 2013
    Posts
    4,002
    i've used the same headers and ran a different o2 delay time on the further sensor, that helped a bit, but ended up going open loop too

  6. #1041
    Fuel Injected!
    Join Date
    Mar 2013
    Posts
    1,469
    I have just tested the new maf table control feature and it worked great. Ultra fast reatime dial of maf table. Now I need to find a way to extract the table from RAM before it gets reset at ignition off.
    Also tested the closed loop controls and can confirm they all work as they should. It is an updated version with some more stuff added. The first version got some bugs and might not worked as expected.

    Terminal_Crazy
    It is real hard to tune with 02 mounted on one of cylinders only. If there is place to mount the 02 between the collector and y-pipe, go for it.
    The cylinder id also got some delay since it must go from e-side to t-side via spi communication and than the value gets into the aldl datastream.
    You can try custom speed log. First open datalog, main -> uncheck stream button. Open raw window and type 03024301240125, set the delay to 1ms and check loop.
    Fisrt byte on the repons will be cyl id, second and third left and right 02s. You can analyze gathered data and see if there is a missing cyl id in the log.

    Another way to figure the idle trims is to disable cylinders one by one and watch how much the wideband afr and map jumps. You should try to even the jump on all cylinders by the same amount. The goal is to equalize all cylinders. Than the afr readings should be fairly constant with no jumps. Best approach is to tune most of the fuel related stuff in open loop.

    I have an improved version of open loop idle patch. It sets the blm to 128 in blm 18 on low mph and goes to open loop in blm cell 16. It is tested and works great. If you got cl problems on idle only it helps alot.
    Last edited by kur4o; 08-16-2018 at 12:12 AM.

  7. #1042
    LT1 specialist steveo's Avatar
    Join Date
    Aug 2013
    Posts
    4,002
    it's really awesome you got that working. i'm impressed

    if i have some time, i'll try to figure a better stragegy for load/save of that table.

  8. #1043
    Fuel Injected!
    Join Date
    Mar 2013
    Posts
    1,469
    It turned to be one of the easiest patch and super simple also. I spend much more time on the eehack interface.
    I also added in the aldl datastream maf cell number and maf cell offset. The offset show the interpolation factor between maf cells. At 0 zero you are close to current cells value, at 255 you are almost at the next cell value.

    I got 3 more slots for realtime control of eside tables and I hopefully have found free enough ram space to put that
    SO the next thing will be ve tables. Setting tables` cells one by one will be huge pain, but it will work.
    The control interface will have one byte for cell 1-17 and one byte for row 1-9 and 1-17[for extended rpm table], and two control bits for each table, enable table modification and switch between ram and rom table.

    If you are willing to make a easier to load interface for eehack will be great. Some crawling algorithm that sends one by one cell values with 0.5 second delay when changed.

    I will try to make pcm, at ign off to copy ram table to another location that will be easy to extract on next ign on cycle. The i/o window will be great for extracting with mode 2 command.

  9. #1044
    Fuel Injected!
    Join Date
    Mar 2013
    Posts
    1,469
    Hi,
    Steveo

    I am trying to configure a wideband on pin D25. It has 0-25.5 volts sensing input. 0-5volts will be in $00-$34 hex range on that input.
    It is supposed to be much more stable input with no pull up/down resistor. Pure voltage sensing input.

    I configured the wideband for 11.3-16.4 afr with 51 points spread. So one in hex will correspond to 0.1 voltage change.

    I looked at the source and couldn`t find where eehack gets the voltage input and how it converts it for wideband usage.

    I will need your help to figure it out.


    EDIT

    Now the edited Maf ram table can be extracted on the next ignition cycle. It will be stored at ignition off at 1a00 address.
    Last edited by kur4o; 08-17-2018 at 02:52 PM.

  10. #1045
    LT1 specialist steveo's Avatar
    Join Date
    Aug 2013
    Posts
    4,002
    the widebands, and every other input, is runtime-configured through the definition csv file (ee_definition.csv). adding new or different inputs is pretty easy.

    if you use the special field specifier WIDEBAND for your wideband input, it'll skip the definition file's math and use the configuration in settings->wideband. this just generates a linear equation anyway.

    related functions are settings_editor::set_wideband_parameters() and datalog_definition::config_wideband(float adder, float multiplier, bool enable)

    you could just as easily NOT specify it as a special WIDEBAND field and do your own math in the definition file.

    if you set their SHORT_NAME to WIDEBAND_AC or WIDEBAND_D27 then it will be able to use them in the dashboard and analyzer by default, otherwise, just use graphing and/or your own log analysis

    if you wanted to add another wideband selection to the settings menu, like WIDEBAND_D25 or something, you could do it. the selected wideband is enumerated in settings.h and stored as a number, and then eehack_settings::get_wb_str() will give you the string that represents that in the definition as a short_name. you would need a new button in the settings ui, and copy the behavior of the other ones (including load/save of that value on open/close of eehack) add a new short_name string to eehack_settings::get_wb_str(), and a new enumerator to the class in settings.h

  11. #1046
    Fuel Injected!
    Join Date
    Mar 2013
    Posts
    1,469
    Thanks alot, I found what I need.

    if(x->special == DEF_SPECIAL_WIDEBAND) {
    x->eq_mult = multiplier * MULT_LINEAR_5V;
    x->eq_add = adder;
    and
    #define MULT_LINEAR_5V 0.0196078431372549

    I need to change MULT_LINEAR_5V to 0.1 to make it work.

    Can I put a check box wideband_d25 and when checked to change MULT_LINEAR_5V to MULT_LINEAR_25V.

    It doesn`t look that easy to link the check box on the settings window to the datalog_definition.cpp math.

    Defining custom math in the definition file looks easy enough to try.

  12. #1047
    Fuel Injected!
    Join Date
    Mar 2013
    Posts
    1,469
    The Ve tables control is done. Now you can run ve tables from ram and change the values in each of the cells.

    The modified VE tables and MAf table are stored in PCM for one ignition cycle.

    Now I need some time to add the patch in eehack.

  13. #1048
    Fuel Injected!
    Join Date
    Mar 2013
    Posts
    1,469
    Finally got some time to finish the latest patch.

    Realtime control of ve tables and maf table added and some other stuff too.

    The provided exe will flash the new patch and enable the extra controls.
    Updated definition file include calculated ve value, current maf cell number and maf cell offset value[the interpolation point of 2 maf cells; 0-255 value; 0=closest to current cell, 255= closest to next cell].

    Flash tested so it wont brick your pcm. Feedback needed.
    Attached Files Attached Files

  14. #1049
    Fuel Injected! Terminal_Crazy's Avatar
    Join Date
    Oct 2015
    Location
    Lancashire England
    Posts
    410
    Evening.
    Just a little bit of feedback.
    I’ve not had a proper chance to test it yet and won’t for a couple of weeks.
    I opened the control window earlier and it’s taller than my laptop screen (1200x1024).
    The bottom row of adjusters was not showing. (First five rows were visible).

    Perhaps it could be split into several smaller windows.

    I’ll give it some hammer as soon as I can.
    Cheers
    Mitch
    '95 Z28 M6 -Just the odd mod.
    '80 350 A3 C3 Corvette - recent addition.

  15. #1050
    Fuel Injected!
    Join Date
    Mar 2013
    Posts
    1,469
    I am aware of that problem. I test it on 1366x768 resolution and it is on the edge.
    The window itself is 530x750 pixels size. I am sure it will work on anything that has at least 800 resolution.

    Can you confirm you are runnning at that resolution and there is no windows enlargements by the OS.

    I can`tthink of easy solution except to make the window wider, than shorter.

Similar Threads

  1. 1badcell and thats not the only thing
    By 1badcell in forum Introductions
    Replies: 2
    Last Post: 12-31-2013, 02:25 AM
  2. Replies: 6
    Last Post: 11-27-2012, 09:03 PM
  3. Replies: 2
    Last Post: 11-07-2012, 05:26 PM
  4. Minor thing.
    By historystamp in forum GearHead EFI Forum Support
    Replies: 7
    Last Post: 01-22-2012, 12:00 AM

Bookmarks

Posting Permissions

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