Results 1 to 20 of 20

Thread: Differences Between '11 & '14 GT500 MAF Data

  1. #1
    Advanced Tuner CCS86's Avatar
    Join Date
    Nov 2017
    Location
    Austin
    Posts
    923

    Differences Between '11 & '14 GT500 MAF Data

    Hi guys,

    I was under the impression that 2011 - 2014 GT500s had identical MAF housings. But comparing downloaded tunes (labeled stock) there are fairly significant differences in the MAF curves, and the differences are a bit lumpy.

    Comparison of the upper end:
    2011 vs 2014 GT500 MAF.jpg



    Closer view of the bottom end:
    GT500 MAF 3.jpg





    Tables:
    2011 vs 2014 GT500 MAF tablular.jpg




    Did the stock values really change this much, year over year, for the same MAF housing?
    Last edited by CCS86; 3 Weeks Ago at 05:14 PM.

  2. #2
    Senior Tuner
    Join Date
    Jan 2013
    Location
    Hawaii
    Posts
    1,478
    5.4L with 1.8L supercharger with 6250RPM redline vs 5.8L with 2.3L supercharger with 7000 RPM redline. A lot changed between 12-13.

    I see identical curves, down to the lumpiness around 200 period. Which is more due to the large jump in period values at that point. The 13-14's just have a few points removed, 1.22, 2.17 and 3.64, to extend the curve below 100 period to achieve higher airflows.
    I believe on a flow bench you put in fixed known airflow values, so those don't need to change, and you record the MAF frequency/ period from them. They certainly got slightly different values for period, but not an extreme difference and more so at the low airflow values. Maybe a different flow bench calibrated slightly better, hopefully.

    11 vs 14.PNG

    With those differing points removed so the same X and Y values are only shown.

    11 vs 14 three points removed.PNG
    "We can never be right, we can only be sure that we are wrong"- feynman

  3. #3
    Advanced Tuner
    Join Date
    Feb 2012
    Location
    TEXAS
    Posts
    228
    2013-2014 GT500 MAF Stock.JPG
    2013-2104 gt500 stock maf

  4. #4
    Advanced Tuner CCS86's Avatar
    Join Date
    Nov 2017
    Location
    Austin
    Posts
    923
    Quote Originally Posted by murfie View Post
    5.4L with 1.8L supercharger with 6250RPM redline vs 5.8L with 2.3L supercharger with 7000 RPM redline. A lot changed between 12-13.

    I see identical curves, down to the lumpiness around 200 period. Which is more due to the large jump in period values at that point. The 13-14's just have a few points removed, 1.22, 2.17 and 3.64, to extend the curve below 100 period to achieve higher airflows.
    I believe on a flow bench you put in fixed known airflow values, so those don't need to change, and you record the MAF frequency/ period from them. They certainly got slightly different values for period, but not an extreme difference and more so at the low airflow values. Maybe a different flow bench calibrated slightly better, hopefully.

    11 vs 14.PNG

    With those differing points removed so the same X and Y values are only shown.

    11 vs 14 three points removed.PNG



    The M122 is 2.0L, but I see your point. Sure a larger engine, with a larger blower, and a higher redline needs some more headroom in the MAF function. That would explain the rescaled axis values, but not all the mid-range differences.


    You say that you see identical curves, so I guess you didn't look at my chart or table. My first post shows the two curves overlaid with each other and there are clear differences. In fact I used a forecasting lookup chart I built to compare MAF curves with different axes, and spit out actual % differences. That's in my post too.

    The way you are visualizing those curves is not best. Using a rolling average with such sporadic spacing of points gives very poor results. That 200 us "lump" you see is a perfect example. There is no lump. Using a scatter chart with smooth lines works far better for this:

    GT500 MAF.jpg

  5. #5
    Advanced Tuner CCS86's Avatar
    Join Date
    Nov 2017
    Location
    Austin
    Posts
    923
    Quote Originally Posted by mstang_man View Post
    2013-2014 GT500 MAF Stock.JPG
    2013-2104 gt500 stock maf

    Is this data that has been converted for use on a pre 2011 voltage based MAF?

    I'm using it on a 2012, so I have 40 cells of MAF vs period to fill.

  6. #6
    Senior Tuner
    Join Date
    Jan 2013
    Location
    Hawaii
    Posts
    1,478
    Your first post looked like you manipulated the data, unintentionally or intentionally I'm not sure. If you were not paying attention to the R squared value of your trend line you extrapolated your airflow values from you would only see what you wanted to see.

    Again the period values are the Y values from the flow bench. So for a given X they do come or slightly different, but nothing like you are talking about.

    I was looking at, and comparing the raw data, unmanipulated, it's the only way you should be comparing two sets of data. Again I see they are the same trend. One that any of the excel trend lines wouldn't fit well. Manipulating the data then looking, is not the better way to look at this.
    "We can never be right, we can only be sure that we are wrong"- feynman

  7. #7
    Advanced Tuner CCS86's Avatar
    Join Date
    Nov 2017
    Location
    Austin
    Posts
    923
    Quote Originally Posted by murfie View Post
    Your first post looked like you manipulated the data, unintentionally or intentionally I'm not sure. If you were not paying attention to the R squared value of your trend line you extrapolated your airflow values from you would only see what you wanted to see.

    Again the period values are the Y values from the flow bench. So for a given X they do come or slightly different, but nothing like you are talking about.

    I was looking at, and comparing the raw data, unmanipulated, it's the only way you should be comparing two sets of data. Again I see they are the same trend. One that any of the excel trend lines wouldn't fit well. Manipulating the data then looking, is not the better way to look at this.



    I don't know where you are coming up with this "manipulated" data claim.

    I asked whether the data I posted was truly the stock data. Since I didn't read out these files, I have no way of knowing for sure.

    I have not manipulated the data in any way. The graph I posted is the two tables plotted on the same axes, and it's clear to see the curves are different. Yes, there are matching airflow values like you said, but since they are associated with different period values, they lie on a different curve.

    I am not using trendlines, because unless you want to decompose the curves into a bunch of shorter sections, the fit will be garbage.

    My comparative tool linearly interpolates between points, to lookup values using a mismatched set of periods. This works pretty well. The error from the linear interpolation is minimal compared to the differences in the curves, and I stand behind the values I posted (again, if that is the stock data).

  8. #8
    Advanced Tuner CCS86's Avatar
    Join Date
    Nov 2017
    Location
    Austin
    Posts
    923
    If you still doubt the way I am looking at % differences with my interpolation calculator, here is the graph, but with the addition of my 2011 curve interpolated to match the 2014 period values.

    You can see that it completely obscures the original 2011 curve, so the fit is almost perfect, but with matching period values, so percent difference can be calculated.

    GT500 MAF 2.jpg

  9. #9
    Senior Tuner
    Join Date
    Jan 2013
    Location
    Hawaii
    Posts
    1,478
    You insist on coming up with your own airflow values out of the period values. You seem to want to manipulate the data to see something that is not there.
    All you needed to do was remove the three airflow values that are no longer in the newer transfer. line up the airflow values, they are exactly the same. Look at the percent difference between the period values. You would see they are with in 2%, much closer to 1% or less in most of the transfer, with the exception of the low airflow values. Everyone knows the most inaccurate part of the MAF transfer is the low airflow values.

    This is all the stock data, and it all comes from a flow bench. You need to think like that is how these numbers were established, not like you're graphing fuel trims. You can expect things to be perfect and exactly the same, but they usually never are, even in very well controlled situations. Gas flow is not always perfectly laminar even in the flow bench.
    "We can never be right, we can only be sure that we are wrong"- feynman

  10. #10
    Advanced Tuner CCS86's Avatar
    Join Date
    Nov 2017
    Location
    Austin
    Posts
    923
    Quote Originally Posted by murfie View Post
    You insist on coming up with your own airflow values out of the period values. You seem to want to manipulate the data to see something that is not there.
    All you needed to do was remove the three airflow values that are no longer in the newer transfer. line up the airflow values, they are exactly the same. Look at the percent difference between the period values. You would see they are with in 2%, much closer to 1% or less in most of the transfer, with the exception of the low airflow values. Everyone knows the most inaccurate part of the MAF transfer is the low airflow values.

    This is all the stock data, and it all comes from a flow bench. You need to think like that is how these numbers were established, not like you're graphing fuel trims. You can expect things to be perfect and exactly the same, but they usually never are, even in very well controlled situations. Gas flow is not always perfectly laminar even in the flow bench.




    I don't know what else to tell you murfie.

    You are looking at zoomed out charts separately, not overlayed, using a rolling average and claiming the two curves are "identical".

    This is flat out, objectively wrong. Just plot both MAF curves on the same chart in Excel and zoom in.

    Tell you what, I did it for you:
    Attached Files Attached Files
    Last edited by CCS86; 4 Weeks Ago at 10:02 PM.

  11. #11
    Advanced Tuner
    Join Date
    Feb 2012
    Location
    TEXAS
    Posts
    228
    Quote Originally Posted by CCS86 View Post
    I don't know what else to tell you murphie.

    You are looking at zoomed out charts separately, not overlayed, using a rolling average and claiming the two curves are "identical".

    This is flat out, objectively wrong. Just plot both MAF curves on the same chart in Excel and zoom in.

    Tell you what, I did it for you:

    the data i posted is stock direct out of my SCT software, no values or manipulation..

  12. #12
    Advanced Tuner CCS86's Avatar
    Join Date
    Nov 2017
    Location
    Austin
    Posts
    923
    Quote Originally Posted by mstang_man View Post
    the data i posted is stock direct out of my SCT software, no values or manipulation..

    Ah, you probably went to the Airflow vs Voltage table, instead of Airflow vs Period.

  13. #13
    Advanced Tuner
    Join Date
    Feb 2012
    Location
    TEXAS
    Posts
    228
    2013-2014 MAF Period.JPG
    2013-2014 Period VS None

  14. #14
    Advanced Tuner CCS86's Avatar
    Join Date
    Nov 2017
    Location
    Austin
    Posts
    923
    Quote Originally Posted by mstang_man View Post
    2013-2014 MAF Period.JPG
    2013-2014 Period VS None

    Thanks man!

    That matches my 2014 stock data perfectly.

    If I can just validate the 2011 data I have, I'm right back to my original question

  15. #15
    Advanced Tuner
    Join Date
    Feb 2012
    Location
    TEXAS
    Posts
    228
    2011 Shelby500.JPG

    2011 Shelby GT500

  16. #16
    Advanced Tuner CCS86's Avatar
    Join Date
    Nov 2017
    Location
    Austin
    Posts
    923
    Here's a closer view of the bottoms of the 2011 vs 2014 curves.

    This look "identical" to you murfie?


    GT500 MAF 3.jpg

  17. #17
    Senior Tuner
    Join Date
    Jan 2013
    Location
    Hawaii
    Posts
    1,478
    YES. Both probably came from a flow bench and are with in the error margin of that device, ~+/- 1%. Close enough for Ford to pass their required emissions standards with.

    When you look at these two curves in the image below you probably think "They are soo different". I think they look identical. It's because I understand which axis is the representation of reality and which is the scalar, and how that effects what the curve looks like. As a Hyperbolized example, when you look at a full size car compared to its 1/10 scale model or 1/25 scale model. I would say they all look identical, if it was actually a well made set of models. I wouldn't let the size difference effect my judgement, as I understand the scaled models purpose to just be a representation of reality, not the exact thing.

    scale vs actual.PNG

    If you don't understand why you are looking at error in this transfer totally wrong after this, I don't know how else to explain it to you.
    "We can never be right, we can only be sure that we are wrong"- feynman

  18. #18
    Advanced Tuner veeefour's Avatar
    Join Date
    Oct 2016
    Location
    Poland
    Posts
    753
    Quote Originally Posted by murfie View Post
    YES. Both probably came from a flow bench and are with in the error margin of that device, ~+/- 1%. Close enough for Ford to pass their required emissions standards with.

    When you look at these two curves in the image below you probably think "They are soo different". I think they look identical. It's because I understand which axis is the representation of reality and which is the scalar, and how that effects what the curve looks like. As a Hyperbolized example, when you look at a full size car compared to its 1/10 scale model or 1/25 scale model. I would say they all look identical, if it was actually a well made set of models. I wouldn't let the size difference effect my judgement, as I understand the scaled models purpose to just be a representation of reality, not the exact thing.

    scale vs actual.PNG

    If you don't understand why you are looking at error in this transfer totally wrong after this, I don't know how else to explain it to you.
    +1

  19. #19
    Senior Tuner
    Join Date
    Jan 2013
    Location
    Hawaii
    Posts
    1,478
    Here the confusing scaling values removed for you so you see it easier.

    Book1 update.xlsx
    "We can never be right, we can only be sure that we are wrong"- feynman

  20. #20
    Advanced Tuner CCS86's Avatar
    Join Date
    Nov 2017
    Location
    Austin
    Posts
    923
    Quote Originally Posted by murfie View Post
    YES. Both probably came from a flow bench and are with in the error margin of that device, ~+/- 1%. Close enough for Ford to pass their required emissions standards with.

    When you look at these two curves in the image below you probably think "They are soo different". I think they look identical. It's because I understand which axis is the representation of reality and which is the scalar, and how that effects what the curve looks like. As a Hyperbolized example, when you look at a full size car compared to its 1/10 scale model or 1/25 scale model. I would say they all look identical, if it was actually a well made set of models. I wouldn't let the size difference effect my judgement, as I understand the scaled models purpose to just be a representation of reality, not the exact thing.

    scale vs actual.PNG

    If you don't understand why you are looking at error in this transfer totally wrong after this, I don't know how else to explain it to you.


    You seem to be ignoring the fact that a matched set of [sensor period] and [flow rate] is the fundamental data of a MAF curve. You can't ignore one of those things. You are so focussed on the flow bench methodology of using a specific sequence of flow rate values, filing in period output to generate the curve, that you ignore the importance of period values being different for a given flow rate; which means the calibration of that housing is different.

    I don't know what you are trying to prove with your modified chart, by completely re-writing all the flow rate values, and turning it into a sequence which only shows the spacing of the period values. You seem to think you are "normalizing" the scaling of the curves, but in reality, you are skewing the real scale to fit your preconceived notion that the curves are the same. What you are doing is putting the model car much closer to you relative to the real car, giving it the appearance of being the same size, then claiming they are identical. We are talking about the science of measurement, not in optical illusions, so this is not valid.

    For a given MAF tube size, physical sensor, and mass flow rate of air, you will get a specific frequency/period output from the sensor. That is a matched set of data that associates a physical phenomenon (mass flow rate) with an electrical signal. You need both of them (X and Y) to define that condition. You need a sequence of defined conditions to generate a characteristic calibration curve. You cannot then take that curve and arbitrarily scale it, then claim it is still a valid calibration. Just like you couldn't take a strain vs voltage output curve from a bathroom scale, modify it and claim it is the same.

    In this chart you are so focussed on the blue lines being horizontal, that you ignore the red lines:

    GT500 MAF 4.jpg