Page 1 of 11 12345 ... LastLast
Results 1 to 20 of 203

Thread: Quick guide to simultaneous VVE/MAF calibration for gen4+

  1. #1
    Advanced Tuner
    Join Date
    Aug 2010
    Location
    Detroit, MI
    Posts
    926

    Quick guide to simultaneous VVE/MAF calibration for gen4+

    I still see a lot of people using the old method of disabling the MAF to calibrate their VVE on 4th generation LS stuff, that generally being E38 controllers and up. This isn't necessary anymore because you can log those GMVE values in real time with HP Tuners. All you need to do is use the proper math parameter(s) and filters to transform the data into the classic '% error' you're used to, which I will describe here. One thing to remember, as you always should when calibrating anything airflow related...it must be at steady state. So you should be using a method of your choosing to filter out transients, which I won't get into here. This method also assumes using short term fuel trims (LTFT disabled), so any open loop guys or calibration into the power-enriched range will have to incorporate fueling error from a wideband.

    The math:
    GMVE STFT % error = ((MAF_af)*(1+(STFT_avg)/100)-(GMVE_af))/(GMVE_af)*100

    Because steady state will always be looking at the MAF for airmass calculations, you must base your GMVE airflow error on the MAF airflow error. First, I normalize the STFT percentage so 0% becomes 1. A STFT of 5% would yield 1.05. Then multiply this by current MAF airflow to get an error corrected MAF airflow. From this value you subtract GMVE airflow to get a raw GMVE airflow error. Then, divide that error by the GMVE airflow to get your normalized VVE airflow error. Multiplying by 100 turns this back into % error to make pasting into the editor simpler.

    To break the GMVE airflow calculation down a bit...

    GMVE_af = (MAP*GMVE/IAT)*RPM/60*4

    Take MAP in MEGApascals (important) and multiple by the raw GMVE value from CAN and divide by the IAT in KELVIN (also important). This gives a single cylinder airmass. To get flow rate, multiply by RPM, divide by 60 to get revolutions per second, and multiply by 4 for the number of induction cycles in one revolution. If you have a 4 or 6 cylinder, multiply by 2 or 3, respectively.

    The full math parameter: GMVE STFT =
    (([16.71]*(1+(.01*[6.156]+.01*[8.156])/2)-[12.56]/60*4*[11.92]*[2312]/[2127.240])/([12.56]/60*4*[11.92]*[2312]/[2127.240]))*100

    Take this error data and plot it in your histograms and you would normally and save a bunch of time and disabling stuff in the calibration. As always, the more data points you get the better you filter out data noise.

    Screen capture of the parameter working...

    Capture.PNG

    Notice at the cursor point, the average STFT in the third graph is showing -4.4%, so a MAF airflow overestimation. However the GMVE airflow, being slightly lower is showing only 0.3% error, so it's right on for that operating point.

    GMVE STFT % error from MAF.MathParameter.xml


    Update (06Nov2020): See the post below for information on improving this process using MAT versus IAT (if parameter is available)
    https://forum.hptuners.com/showthrea...l=1#post631001
    Last edited by smokeshow; 11-06-2020 at 05:19 PM.

  2. #2
    Advanced Tuner ttz06vette's Avatar
    Join Date
    Sep 2008
    Location
    Charlotte, NC.
    Posts
    242
    Good stuff...Couple questions.
    So does it matter what RPM you are enabling MAF in your tune?
    In your example at that cursor point, what change would you make, multiply your MAF table by .978 at that rpm/frequency?(50% of the -4.4%)
    What would Your histogram table axis’s be if you set this formula up?

    Thanks for the thread, can’t wait to give it a try.

  3. #3
    Advanced Tuner
    Join Date
    Aug 2010
    Location
    Detroit, MI
    Posts
    926
    Quote Originally Posted by ttz06vette View Post
    Good stuff...Couple questions.
    So does it matter what RPM you are enabling MAF in your tune?
    In your example at that cursor point, what change would you make, multiply your MAF table by .978 at that rpm/frequency?(50% of the -4.4%)
    What would Your histogram table axis’s be if you set this formula up?

    Thanks for the thread, can’t wait to give it a try.
    The MAF is technically always enabled, which is why it is important you filter out transients so you only capture steady state data. So no, it doesn't matter what RPM you disable dynamic airflow at.

    At that cursor point, assuming the error percentages represented a bunch of cell hits, I would multiply the MAF table at that frequency by 0.956 and the VVE at that RPM/MAP by 1.003. You can use percent-half if you prefer.

    The histogram axis depends on what your calibration is set to. I'd use the same axis that is in the editor.

  4. #4
    Advanced Tuner jsllc's Avatar
    Join Date
    Sep 2015
    Location
    Bozeman, MT
    Posts
    701
    Quote Originally Posted by smokeshow View Post
    I still see a lot of people using the old method of disabling the MAF to calibrate their VVE on 4th generation LS stuff, that generally being E38 controllers and up. This isn't necessary anymore because you can log those GMVE values in real time with HP Tuners. All you need to do is use the proper math parameter(s) and filters to transform the data into the classic '% error' you're used to, which I will describe here. One thing to remember, as you always should when calibrating anything airflow related...it must be at steady state. So you should be using a method of your choosing to filter out transients, which I won't get into here. This method also assumes using short term fuel trims (LTFT disabled), so any open loop guys or calibration into the power-enriched range will have to incorporate fueling error from a wideband.

    The math:
    GMVE STFT % error = ((MAF_af)*(1+(STFT_avg)/100)-(GMVE_af))/(GMVE_af)*100

    Because steady state will always be looking at the MAF for airmass calculations, you must base your GMVE airflow error on the MAF airflow error. First, I normalize the STFT percentage so 0% becomes 1. A STFT of 5% would yield 1.05. Then multiply this by current MAF airflow to get an error corrected MAF airflow. From this value you subtract GMVE airflow to get a raw GMVE airflow error. Then, divide that error by the GMVE airflow to get your normalized VVE airflow error. Multiplying by 100 turns this back into % error to make pasting into the editor simpler.

    To break the GMVE airflow calculation down a bit...

    GMVE_af = (MAP*GMVE/IAT)*RPM/60*4

    Take MAP in MEGApascals (important) and multiple by the raw GMVE value from CAN and divide by the IAT in KELVIN (also important). This gives a single cylinder airmass. To get flow rate, multiply by RPM, divide by 60 to get revolutions per second, and multiply by 4 for the number of induction cycles in one revolution. If you have a 4 or 6 cylinder, multiply by 2 or 3, respectively.

    The full math parameter: GMVE STFT =
    (([16.71]*(1+(.01*[6.156]+.01*[8.156])/2)-[12.56]/60*4*[11.92]*[2312]/[2127.240])/([12.56]/60*4*[11.92]*[2312]/[2127.240]))*100

    Take this error data and plot it in your histograms and you would normally and save a bunch of time and disabling stuff in the calibration. As always, the more data points you get the better you filter out data noise.

    Screen capture of the parameter working...

    Capture.PNG

    Notice at the cursor point, the average STFT in the third graph is showing -4.4%, so a MAF airflow overestimation. However the GMVE airflow, being slightly lower is showing only 0.3% error, so it's right on for that operating point.

    GMVE STFT % error from MAF.MathParameter.xml

    So why do you choose to use IAT and not MAT as returned by the ecm which takes into account pressure/boost and other heat transfers?
    2012 ZL1 - Maggie Heartbeat, Port & Polish Heads, Custom Cam, Custom rotating assembly, steel sleeved LS9, No NOS and No water meth. 16psi
    810rwhp and 820rwtq 91 Octane 6400 rpm
    948rwhp and 951rwtq 105 Octane 6400 rpm
    999rwhp and 997rwtq on 60% Ethanol 6400 rpm

  5. #5
    Advanced Tuner
    Join Date
    Aug 2010
    Location
    Detroit, MI
    Posts
    926
    Quote Originally Posted by jsllc View Post
    So why do you choose to use IAT and not MAT as returned by the ecm which takes into account pressure/boost and other heat transfers?
    That's what I had available on the vehicle I developed this on. If you have a temperature sensor on your vehicle that is closer to the cylinder, use that.

  6. #6
    Advanced Tuner jsllc's Avatar
    Join Date
    Sep 2015
    Location
    Bozeman, MT
    Posts
    701
    Quote Originally Posted by smokeshow View Post
    That's what I had available on the vehicle I developed this on. If you have a temperature sensor on your vehicle that is closer to the cylinder, use that.
    Okay, Thank you. Next question as MAF on HPT ends @ 655.35 g/sec it was useless on high HP cars so I used:

    MVE STFT % error = ((DYN_af)*(1+(STFT_avg)/100)-(GMVE_af))/(GMVE_af)*100

    GMVE_af = (MAP*GMVE/MAT)*RPM/60*4

    Which used the dynamic AF value returned in the PIDs. If you remove predicted modifiers DYN and MAF are extremely close. Is there an issue?
    2012 ZL1 - Maggie Heartbeat, Port & Polish Heads, Custom Cam, Custom rotating assembly, steel sleeved LS9, No NOS and No water meth. 16psi
    810rwhp and 820rwtq 91 Octane 6400 rpm
    948rwhp and 951rwtq 105 Octane 6400 rpm
    999rwhp and 997rwtq on 60% Ethanol 6400 rpm

  7. #7
    Advanced Tuner
    Join Date
    Aug 2010
    Location
    Detroit, MI
    Posts
    926
    Quote Originally Posted by jsllc View Post
    Okay, Thank you. Next question as MAF on HPT ends @ 655.35 g/sec it was useless on high HP cars so I used:

    MVE STFT % error = ((DYN_af)*(1+(STFT_avg)/100)-(GMVE_af))/(GMVE_af)*100

    GMVE_af = (MAP*GMVE/MAT)*RPM/60*4

    Which used the dynamic AF value returned in the PIDs. If you remove predicted modifiers DYN and MAF are extremely close. Is there an issue?
    I would use whichever method you typically use when calibrating MAF-only and reach that limit. This won't work for GMVE calibration unless MAF can be used as a baseline. You can strip dynamic airflow down until it is bare-bones MAF only if you want, and if you have the knobs to turn.

  8. #8
    Advanced Tuner
    Join Date
    Jun 2010
    Posts
    293
    sorry to ask but how do you setup it up with WB instead of stft

  9. #9
    Advanced Tuner jsllc's Avatar
    Join Date
    Sep 2015
    Location
    Bozeman, MT
    Posts
    701
    Quote Originally Posted by castorjames View Post
    sorry to ask but how do you setup it up with WB instead of stft
    MVE WB % error = ((MAF_af)*(1+(WB_avg)/100)-(GMVE_af))/(GMVE_af)*100

    GMVE_af = (MAP*GMVE/MAT)*RPM/60*4

    So on a E67 with 2 WBs using lambda:

    WB_avg = ((((WB1+WB2)/2)/EQ Commanded)-1)*100

    using odrer of precedents this would become

    WB_avg = (((WB1+WB)/2/EQ Commanded)-1)*100

    This should work. If you only have one WB the math to average the two would be replaced.

    WB_avg = ((WB1/EQ Commanded)-1)*100

    This is inaccurate as only one bank is used and all of the gen4 engines I have tuned show one bank richer than the other.
    Last edited by jsllc; 07-07-2020 at 12:25 AM.
    2012 ZL1 - Maggie Heartbeat, Port & Polish Heads, Custom Cam, Custom rotating assembly, steel sleeved LS9, No NOS and No water meth. 16psi
    810rwhp and 820rwtq 91 Octane 6400 rpm
    948rwhp and 951rwtq 105 Octane 6400 rpm
    999rwhp and 997rwtq on 60% Ethanol 6400 rpm

  10. #10
    Advanced Tuner
    Join Date
    Apr 2016
    Posts
    328
    Does anyone see an issue tuning the VVE and MAF at the same time with the following formulas, regardless of MAF settings? Assumption is open loop and using a wideband.


    VVE error = DynAir(g/sec) * EQ Sensor / EQ Commanded / VE Airflow(g/sec) * 100 -100

    and

    MAF error = DynAir(g/sec) * EQ Sensor / EQ Commanded / Mass Air(g/sec) * 100 - 100

    the * and - 100 at the end is just another means to give you a negative number when rich/positive when lean, to easily multiply by % in your calibration chart. Otherwise it is just showing you the percent difference between what the Dynamic Air in g/sec SHOULD be (factoring in the EQ error from the wideband), and the 1)VVE g/sec or the 2)MAF g/sec.

    I guess I am just asking if we need to get into the GMVE calculations at all, or can we just log VE Airflow and compare it to a corrected (via WB or STFT error) Dynamic Air, which is what the engine is running on anyways? Repeat the same for Mass Airflow. Right or wrong, it seems to work fine.

  11. #11
    Advanced Tuner
    Join Date
    Apr 2016
    Posts
    328
    I yanked a single frame from a recent log while cruising, and I have the following:

    Mass Airflow 52.2 g/s
    Volumetric Efficiency Airflow 50.9 g/s
    Dynamic Airflow 51.88 g/s

    EQ sensor 0.965
    EQ commanded 0.999

    Plugging these into the equations above, VE is -1.54, and MAF is -3.99, denoted in percent rich or lean (rich in this case).

  12. #12
    Advanced Tuner
    Join Date
    Aug 2010
    Location
    Detroit, MI
    Posts
    926
    Quote Originally Posted by NotSure View Post
    I guess I am just asking if we need to get into the GMVE calculations at all, or can we just log VE Airflow and compare it to a corrected (via WB or STFT error) Dynamic Air, which is what the engine is running on anyways? Repeat the same for Mass Airflow. Right or wrong, it seems to work fine.
    You don't have to use the GMVE calculation in the math parameter, but you'll be limited to the HPT output of 512g/s. The raw calculation avoids that limitation.

  13. #13
    Advanced Tuner
    Join Date
    Jun 2010
    Posts
    293
    JSLLC. Thank you sir. Ill try to understand it. Will do some tests.

  14. #14
    Advanced Tuner jsllc's Avatar
    Join Date
    Sep 2015
    Location
    Bozeman, MT
    Posts
    701
    Quote Originally Posted by NotSure View Post
    I yanked a single frame from a recent log while cruising, and I have the following:

    Mass Airflow 52.2 g/s
    Volumetric Efficiency Airflow 50.9 g/s
    Dynamic Airflow 51.88 g/s

    EQ sensor 0.965
    EQ commanded 0.999

    Plugging these into the equations above, VE is -1.54, and MAF is -3.99, denoted in percent rich or lean (rich in this case).

    Dynamic has other modifiers. It is not a RAW value like MAF. If these are active then you skew the results. A compromise would be to use MAF below 655 and DYN above as most of the modifiers will not come into play (just a thought).

    Do not know why people say MAF ends @ 512. At least on an E67 it is 655.35 g/sec
    Attached Images Attached Images
    Last edited by jsllc; 07-08-2020 at 10:14 AM.
    2012 ZL1 - Maggie Heartbeat, Port & Polish Heads, Custom Cam, Custom rotating assembly, steel sleeved LS9, No NOS and No water meth. 16psi
    810rwhp and 820rwtq 91 Octane 6400 rpm
    948rwhp and 951rwtq 105 Octane 6400 rpm
    999rwhp and 997rwtq on 60% Ethanol 6400 rpm

  15. #15
    Advanced Tuner jsllc's Avatar
    Join Date
    Sep 2015
    Location
    Bozeman, MT
    Posts
    701
    Quote Originally Posted by NotSure View Post
    Does anyone see an issue tuning the VVE and MAF at the same time with the following formulas, regardless of MAF settings? Assumption is open loop and using a wideband.


    VVE error = DynAir(g/sec) * EQ Sensor / EQ Commanded / VE Airflow(g/sec) * 100 -100

    and

    MAF error = DynAir(g/sec) * EQ Sensor / EQ Commanded / Mass Air(g/sec) * 100 - 100

    the * and - 100 at the end is just another means to give you a negative number when rich/positive when lean, to easily multiply by % in your calibration chart. Otherwise it is just showing you the percent difference between what the Dynamic Air in g/sec SHOULD be (factoring in the EQ error from the wideband), and the 1)VVE g/sec or the 2)MAF g/sec.

    I guess I am just asking if we need to get into the GMVE calculations at all, or can we just log VE Airflow and compare it to a corrected (via WB or STFT error) Dynamic Air, which is what the engine is running on anyways? Repeat the same for Mass Airflow. Right or wrong, it seems to work fine.
    Using a HPT "Sensor" does not allow for control of which input you use. Just be aware.
    2012 ZL1 - Maggie Heartbeat, Port & Polish Heads, Custom Cam, Custom rotating assembly, steel sleeved LS9, No NOS and No water meth. 16psi
    810rwhp and 820rwtq 91 Octane 6400 rpm
    948rwhp and 951rwtq 105 Octane 6400 rpm
    999rwhp and 997rwtq on 60% Ethanol 6400 rpm

  16. #16
    Advanced Tuner jsllc's Avatar
    Join Date
    Sep 2015
    Location
    Bozeman, MT
    Posts
    701
    Quote Originally Posted by smokeshow View Post
    I still see a lot of people using the old method of disabling the MAF to calibrate their VVE on 4th generation LS stuff, that generally being E38 controllers and up. This isn't necessary anymore because you can log those GMVE values in real time with HP Tuners. All you need to do is use the proper math parameter(s) and filters to transform the data into the classic '% error' you're used to, which I will describe here. One thing to remember, as you always should when calibrating anything airflow related...it must be at steady state. So you should be using a method of your choosing to filter out transients, which I won't get into here. This method also assumes using short term fuel trims (LTFT disabled), so any open loop guys or calibration into the power-enriched range will have to incorporate fueling error from a wideband.

    The math:
    GMVE STFT % error = ((MAF_af)*(1+(STFT_avg)/100)-(GMVE_af))/(GMVE_af)*100

    Because steady state will always be looking at the MAF for airmass calculations, you must base your GMVE airflow error on the MAF airflow error. First, I normalize the STFT percentage so 0% becomes 1. A STFT of 5% would yield 1.05. Then multiply this by current MAF airflow to get an error corrected MAF airflow. From this value you subtract GMVE airflow to get a raw GMVE airflow error. Then, divide that error by the GMVE airflow to get your normalized VVE airflow error. Multiplying by 100 turns this back into % error to make pasting into the editor simpler.

    To break the GMVE airflow calculation down a bit...

    GMVE_af = (MAP*GMVE/IAT)*RPM/60*4

    Take MAP in MEGApascals (important) and multiple by the raw GMVE value from CAN and divide by the IAT in KELVIN (also important). This gives a single cylinder airmass. To get flow rate, multiply by RPM, divide by 60 to get revolutions per second, and multiply by 4 for the number of induction cycles in one revolution. If you have a 4 or 6 cylinder, multiply by 2 or 3, respectively.

    The full math parameter: GMVE STFT =
    (([16.71]*(1+(.01*[6.156]+.01*[8.156])/2)-[12.56]/60*4*[11.92]*[2312]/[2127.240])/([12.56]/60*4*[11.92]*[2312]/[2127.240]))*100

    Take this error data and plot it in your histograms and you would normally and save a bunch of time and disabling stuff in the calibration. As always, the more data points you get the better you filter out data noise.

    Screen capture of the parameter working...

    Capture.PNG

    Notice at the cursor point, the average STFT in the third graph is showing -4.4%, so a MAF airflow overestimation. However the GMVE airflow, being slightly lower is showing only 0.3% error, so it's right on for that operating point.

    GMVE STFT % error from MAF.MathParameter.xml

    Should check my math but because division is a more expensive cpu operation than multiplication:
    GMVE_af = ((MAP*GMVE/IAT)*RPM/60*4)
    STFT_avg = (stft1+stft2)*.005

    GMVE STFT % error = (MAF_af*(1+STFT_avg)-GMVE_af)/GMVE_af*100

    (MAF_af*(1+((stft1 +stft2 ) *.005)-(MAP *GMVE /IAT *RPM /60*4))/(MAP *GMVE /IAT *RPM /60*4)*100
    ([16.71]*(1+([6.156]+[8.156])*.005)-([11.92]*[2312]/[2127.240]*[12.56]/60*4))/([11.92]*[2312]/[2127.240]*[12.56]/60*4)*100

    MAT
    (MAF_af*(1+((stft1 +stft2 )*.005)-(MAP *GMVE /MAT *RPM /60*4))/(MAP *GMVE /MAT *RPM /60*4)*100
    ([16.71]*(1+([6.156]+[8.156])*.005)-([11.92]*[2312]/[2126.240]*[12.56]/60*4))/([11.92]*[2312]/[2126.240]*[12.56]/60*4)*100

    Should work with AEM can bus WB and MAT
    ([16.71]*(1+([36.238]/[68.238]-1))-([11.92]*[2312]/[2126.240]*[2135.56]/60*4))/([11.92]*[2312]/[2126.240]*[2135.56]/60*4)*100
    Last edited by jsllc; 07-09-2020 at 01:26 AM.
    2012 ZL1 - Maggie Heartbeat, Port & Polish Heads, Custom Cam, Custom rotating assembly, steel sleeved LS9, No NOS and No water meth. 16psi
    810rwhp and 820rwtq 91 Octane 6400 rpm
    948rwhp and 951rwtq 105 Octane 6400 rpm
    999rwhp and 997rwtq on 60% Ethanol 6400 rpm

  17. #17
    Advanced Tuner
    Join Date
    Apr 2016
    Posts
    328
    Quote Originally Posted by jsllc View Post
    Should check my math but because division is a more expensive cpu operation than multiplication:
    ....

    This brings up a question that I've had in the back of my mind for awhile but never found an answer or asked. If you have a bunch of complex maths going on in your Graphs, opened while logging, can it affect the speed or accuracy of data recording? If so, would it be best to just dump all of your graphs from the Scanner while logging, then open them back up afterwards to review? This question would be best put in the Scanner forum, but while we are here...

  18. #18
    Advanced Tuner
    Join Date
    Aug 2010
    Location
    Detroit, MI
    Posts
    926
    Quote Originally Posted by jsllc View Post
    Should check my math but because division is a more expensive cpu operation than multiplication:
    GMVE_af = ((MAP*GMVE/IAT)*RPM/60*4)
    STFT_avg = (stft1+stft2)*.005

    GMVE STFT % error = (MAF_af*(1+STFT_avg)-GMVE_af)/GMVE_af*100

    (MAF_af*(1+((stft1 +stft2 ) *.005)-(MAP *GMVE /IAT *RPM /60*4))/(MAP *GMVE /IAT *RPM /60*4)*100
    ([16.71]*(1+([6.156]+[8.156])*.005)-([11.92]*[2312]/[2127.240]*[12.56]/60*4))/([11.92]*[2312]/[2127.240]*[12.56]/60*4)*100

    MAT
    (MAF_af*(1+((stft1 +stft2 )*.005)-(MAP *GMVE /MAT *RPM /60*4))/(MAP *GMVE /MAT *RPM /60*4)*100
    ([16.71]*(1+([6.156]+[8.156])*.005)-([11.92]*[2312]/[2126.240]*[12.56]/60*4))/([11.92]*[2312]/[2126.240]*[12.56]/60*4)*100

    Should work with AEM can bus WB and MAT
    ([16.71]*(1+([36.238]/[68.238]-1))-([11.92]*[2312]/[2126.240]*[2135.56]/60*4))/([11.92]*[2312]/[2126.240]*[2135.56]/60*4)*100
    Quote Originally Posted by NotSure View Post
    This brings up a question that I've had in the back of my mind for awhile but never found an answer or asked. If you have a bunch of complex maths going on in your Graphs, opened while logging, can it affect the speed or accuracy of data recording? If so, would it be best to just dump all of your graphs from the Scanner while logging, then open them back up afterwards to review? This question would be best put in the Scanner forum, but while we are here...
    These are good points, glad you brought them up. It is correct that my parameter is not optimized to minimize execution time. However, there is one thing to consider that makes this a non-issue: this computation is not executing in a real time environment. In an engine controller, nearly every calculation that takes place is expected to complete in time to service the update rate of the actuator drivers, not to mention the fault monitoring, rationality diagnostics and other things that happen in the background. This is ensured by validating the real-time code in a software in-the-loop environment and keeping the processor utilization under the corporate 80% cap, which is part of what I did at GM. But your PC doesn't need to do this with the math parameter; it can just record the data and post-process it later.

    The trick is knowing when to cut back on the parameters in the scanner. Every modern mid-range PC these days will be plenty powerful so you will never have an issue. But if you're like me and use a light-duty PC for some extra battery life, there's a chance you'll run into a speed issue. I used to see it a lot on my Asus eee PC. It would look like you paused the data recording but time kept incrementing and picked up recording again some time later, and then drew lines in between the last points the scanner had data for. The HPT charts interpolate between variable updates, so it becomes pretty obvious when all variables are paused at once and interpolated. Data collection is the only thing that actually needs to be done in real time with the scanner, so if the scanner hangs when running this and/or other math parameters, you'd be better off pulling it out and running the parameter later when you already have all of your data. Or if you know enough about how HPT executes their own code to manipulate this data, you could make recommendations directly to them

    I've seen this problem only once in industry. Years back, I pulled in too much information in from a GPEC2A in Inca and caused the program to crash. There's no remedial action like interpolating between data points with the industrial calibration software because a lot of post-processing is automated these days. Unless someone looked directly at a temporary hang in a data file, it could get missed and the post processing tool would try to make sense of bad data and produce misleading results. I also don't think ETAS really anticipates this being an issue because it requires thousands of variables to do it and there's really no reason to collect that much data unless you're just too lazy to remove them or switch experiments lol.

    I could talk for a while on any of these topics, so ask if you have any more questions.

  19. #19
    Senior Tuner DSteck's Avatar
    Join Date
    Mar 2008
    Location
    St. Louis, MO
    Posts
    4,856
    Or... you use wideband feedback to back calculate GMVE along with a host of other parameters which you'll find is more accurate and works in either open or closed loop. It does still require steady state (this is where a loaded dyno comes in handy).

    DSX Tuning - Authorized HP Tuners Dealer
    http://www.dsxtuning.com
    http://www.facebook.com/dsx.tuning
    Just say no to bull s***.
    IF YOU WANT HELP, POST A FILE!

  20. #20
    Advanced Tuner
    Join Date
    Aug 2010
    Location
    Detroit, MI
    Posts
    926
    Quote Originally Posted by DSteck View Post
    Or... you use wideband feedback to back calculate GMVE along with a host of other parameters which you'll find is more accurate and works in either open or closed loop. It does still require steady state (this is where a loaded dyno comes in handy).
    There are far too many other variables at play that GM compensates for and the wideband cannot in terms of O2 feedback. I will trust the onboard fuel trims 10 times out of 10 before even the best external wideband.