Results 1 to 6 of 6

Thread: logging "Commanded" VE?

  1. #1
    Potential Tuner
    Join Date
    Dec 2013
    Posts
    9

    logging "Commanded" VE?

    Is there a way to log the "Commanded" VE used during operation? What I want is to get a timestamped VE that get's used in the fueling calculations in every line (row) of the excel-based (.csv) output of a log. That way I can take the AFR % Error from the WB02 for each reading and calculate a "Corrected" VE.

    The typical procedure of using the HPTuners histogram of AFR % Error and applying that to the VE table is causing my tunes to 'hunt' continuously...one day I'll add 2% to a given cell then another day i'll subtract 2% from a given cell, for example...doing some statistics on my logs is showing a bimodal distribution (2 humps in a histogram, instead of typically and hopefully, 1).

    Here is an example of what I'm trying to do (see attachment which is an excel output of a log at idle)...however in the attachment the "Commanded VE %" is a calculation performed in Excel using the CylAirmass calculation inverted to find VE [VE=(CylAirmass*0.28705*((IAT+459.67)*5/9)*8cylinders)/(6*MAP)]...but there seems to be lots of error in this method so I need realtime data from the scanner...and the "Corrected VE %" is just "Commanded VE% x (1 - AFR Error %). For example in this cell of the VE table (30kPa,800rpm) my VE per the Primary VE table should be 55.92%.

    Example log at idle.JPG

  2. #2
    Senior Tuner LSxpwrdZ's Avatar
    Join Date
    Dec 2008
    Location
    on the Dyno
    Posts
    1,825
    If your lambda error PID is setup correctly you should get the same final result...

    The other software calculates VE the way you describe rather than just applying the error to the existing graph. It wont work on HPTuners as the manifold temperature PID is missing and is needed to accurately calculate VE %.
    James Short - [email protected]
    Located in Central Kentucky
    ShorTuning
    2020 Camaro 2SS | BTR 230 | GPI CNC Heads | MSD Intake | Rotofab | 2" LT's | Flex Fuel | 638rwhp / 540rwtq
    2002 Camaro | LSX 427 | CID LS7's | Twin GT5088's | Haltech Nexus R5 | RPM TH400

  3. #3
    Senior Tuner
    Join Date
    Nov 2002
    Location
    Central Florida
    Posts
    2,503
    Maybe your bimodal distribution is from lumping steady
    state and transitional data together. This is the flaw in
    the paste-histo tuning method, if you don't qualify the
    incoming data then your histogram can contain a lot of
    maggots. Tasty, tasty maggots.

    If you have one MAP and RPM then you can of course
    run your finger down the table and pick off the VE (less
    any traversal-filter time domain effects). I still prefer to
    work that way (since nobody has me on the clock) -
    start with a patch of data you know is worth working,
    get your RPM, MAP -> VE cell, tweak that cell based
    on wideband. Repeat for other positions where you
    really trust the data and ignore where you don't. Then
    at the end, you smooth in the untouched cells (by
    hand, one by one or using the interpolation feature
    between two trusted points) to fit the ones you
    worked.

    Slow, compared to magic box, but way more integrity.

    Now another piece of the puzzle is that you're expecting
    to get closer than 2% stable accuracy day to day, which
    to me seems a bit optimistic. How good is good enough?
    If you don't have "spec limits" then the job is never done.

  4. #4
    Potential Tuner
    Join Date
    Dec 2013
    Posts
    9
    Yeah I've been thinking about 'spec limits'....how good is good enough...some say +/-1%...

    If I track the corrections I've made over the last few weeks they look like this for a particular cell (in this case 55kpa,2000rpm and 50kpa,1600rpm respectively)...these are the 'corrected' VE values after logging that particular run and applying the AFR % Error to the VE table in effect at that time...I'm not sure if the variation is due to temperature or there is some natural frequency of the control system (I know that sounds crazy)...so I guess my solution is to average all of these 'corrected' VE values over the last few weeks and see what that does...so for example I would target 77% VE for the 55kpa,2000rpm and about 75% VE for 50kpa,1600rpm

    55kpa,2000rpm cell VE corrections based on AFR % Error after various logs.JPG50kpa,1600rpm cell VE corrections based on AFR % Error after various logs.JPG
    Last edited by Throttler; 02-06-2015 at 11:52 AM.

  5. #5
    There's no PID for the commanded VE in the LS1 PCMs, the VE table lookup is never saved to a RAM location
    I count sheep in hex...

  6. #6
    Potential Tuner
    Join Date
    Dec 2013
    Posts
    9
    Thanks Chris, is there a way to back calculate what VE was used however using the parameters that are saved to RAM? If this is a simple calculation like the one I posted above I would suppose this could be done easily, but if the code uses all kinds of model inputs, biases, kickers, etc then that's probably not possible...we'd have to 'look under the hood' to understand how the software works then replicate that using a PID or in excel...