This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

AWRL1432:ATE calibration & Factory calibration difference

Part Number: AWRL1432

Hi there,
I use the parameter configuration of gesture_recognition in our custom board,found that noise occurs on different PCs (RD map is output),

Normal graph (The horizontal axis is doppler; the vertical axis is range):

Exception graph:

After referring to the information above e2e, try to change factoryCalibCfg,

"factoryCalibCfg 1 0 40 0 ​​0x1ff000 \n\r" =>Original

"factoryCalibCfg 0 1 40 0 ​​0x1ff000 \n\r" =>After changes, It is found that RDmap has no noise

I checked from the code and found that when factoryCalibCfg's restoreEnable is set to 1, factoryCalCfg.isFactoryCalEnabled = false (as fig.1),

which will prevent ATE calibration from performing the action of retrieving information (as fig. 2), but on e2e I know that 1432 has factory calibration & ATE calibration , the two are not the same,

but changing the factory parameters in the program will simultaneously affect the ATE calibration. What is the relationship between them?

fig.1

fig. 2

Best Wishes,

YAN ZUH, CHEN

  • Hi,

    Can you please let us know which specific demo you are using? Please also provide the version. This will help our team support this question better

    thank you

    Cesar

  • Hi Cesar,

    The software version I use is MMWAVE_L_SDK_05_03_00_02, and the demo uses gesture_recognition.

    YAN ZHU CHEN

  • Thank you

    This demo is supported only for xWRL6432 devices, not AWRL1432. Did you port this demo to AWRL1432?

    Thank you

    Cesar

  • Hi Cesar,

    In the ti\MMWAVE_L_SDK_05_03_00_02\examples\mmw_demo path, there is gesture_recognition in mmwdemo, and the program code is divided into XWRL64XX and else. Do you mean that this package of code cannot be used for AWRL1432?

    Thanks

    YAN ZHU CHEN

  • Hi,

    You are correct, I see that this demo is supported in SDK 5.3.0.2 on 1432

    C:\ti\MMWAVE_L_SDK_05_03_00_02\examples\mmw_demo\gesture_recognition\xwrL14xx-evm\m4fss0-0_freertos\ti-arm-clang

    Let me check which team supports this demo for 1432

    thank you

    Cesar

  • Hi Yan Zhu Chen,

    For production samples(Trimmed) Some Calibrations like RX IFA calibrations are performed at the TI factory and results are trimmed in eFuse.
    But for pre-production samples(Untrimmed) these calibrations are done at the customer side itself using ATE emulation SDK package which calculates and stores the trim values in flash.

    We recommend you to use trimmed samples if you are using un-trimmed ones.

    Thanks,
    Kundan

  • Hi Kundan,
    Thanks for your reply.

    I have the following two questions:
    1. I have previously completed ATE calibration through the tool provided by the sdk and written it into Flash (C:\ti\MMWAVE_L_SDK_05_03_00_02\tools\Ate_Calibration\readComPort)
    So would you suggest that I change the setting to "factoryCalibCfg 0 1 40 0 ​​0x1ff000 \n\r"?

    But as far as I know, ATE calibration is not the same as factory calibration.
    2. Can you provide the specific content of ATE calibration and factory calibration?
    I found x432 Calibration Sequence from e2e, but it didn’t clearly tell me which one is ATE calibration and factory calibration?

  • Hi Yan Zhu Chen,

    Yes we recommend to restore the factory calibration data to avoid re-running the cals everytime.

    The Image which you have shared is for production(trimmed) samples where you should not be concerned about the ATE calibration. Since they are done at TI Factory itself this has been not mentioned here.

    Thanks,
    Kundan

  • Hi Kundan,
    But as far as I know, if I change the first two configurations of factory to factoryCalibCfg 0 1 40 0 ​​​​0x1ff000 \n\r,
    It will read the last information stored in Flash (because saveEnable=0; restoreEnable=1). If I have previously set it according to the default value (saveEnable=1; restoreEnable=0), will it restoreEnable the value read? Is it the correction value obtained by setting saveEnable=1; restoreEnable=0 before?

    In addition, I still want to know why when adjusting the Factory parameters in the program, it is also adjusted to ATE calibration (like my initial question, please refer to fig. 1 & 2)

    Thank you.

  • Hi Yan Zhu Chen,

    If you do save-enable=1 then FECSS performs all the relevant calibrations and saves the calibration data into flash.
    Once you do restore-enable=1 then If valid calibration data is present, it will restore the data from FLASH(latest stored using save enabled) and provide it to FECSS.

    Regarding the second question I will come back to you by Wednesday.

    Thanks,
    Kundan

  • Hi Yan Zhu Chen,

    if you do save enable=1 that essentially means that you wish to perform the factory calibrations. For that you would essentially need ATE calib results to be restored(In untrimmed samples). So that is why by adjusting the parameters it is also getting adjusted in the ATE Calib.

    Following is the recommended flow:
    Perform ATE Cal → restore the ATE Cal data → perform Factory Calibrations and save the overall calibration results → restore the overall calibration data.

    Thanks,
    Kundan