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.

  • Resolved

IWR6843: txPhaseShifter parameter issue

Expert 1070 points

Replies: 7

Views: 104

Part Number: IWR6843

Dear TI 

              This is from your ti\mmwave_industrial_toolbox_4_2_0__win\mmwave_industrial_toolbox_4_2_0\labs\long_range_people_detection\68xx_long_range_people_det\docs

             Beamforming_in_LRPD Page 4

            

All 3 TX is enabled at the same time. These four subframes are aim to focus on the angle of [-15, 0, 15, 30].
profileCfg 0 61.0 10 6 135 0 8174592 1.8 1 256 2000 0 0 48
profileCfg 1 61.0 10 6 135 0 0 1.8 1 256 2000 0 0 48
profileCfg 2 61.0 10 6 135 0 8668160 1.8 1 256 2000 0 0 48
profileCfg 3 61.0 10 6 135 0 16809984 1.8 1 256 2000 0 0 48

          

           And  Page 2

brief Concatenated phase shift for TX0/1/2, b1:0 reserved (set to 0b00); b7:2 TX0 phase shift value 1 LSB = 360/2^6 = 5.625 degrees b9:8 reserved (set to 0b00); b15:10 TX1 phase shift value 1 LSB = 360/2^6 = 5.625 degrees b17:16 reserved (set to 0b00); b23:18 TX2 phase shift value 1 LSB = 360/2^6 = 5.625 degrees b31:24 Reserved This field defines the additional phase shift to be introduced on each transmitter output. In IWR6843 ES1.0, only 0 degree phase is supported.

    

               angle -15 = 8174592  ???

               angle +15 = 8668160  ???

               angle +39 = 16809984  ???

               angle +12 =  ?       

        

           1.  Our questiion is your description can't match between page 2 & 4 . 

           2.  How to setting the "txPhaseShifter" parameter ?    Can description detail with diagram and  provide tool for txPhaseShifter parameter ?

           3.  Why not provide gui tool for radar configuration ? Parameter scale could confusing user .

           4. If have TX beamforming tuning guide could share for users  ?  

                      

Best Regards

Jasper

  • Hi Jasper,

    1. After writing the document, we adjusted the configuration to cover the following angles:

    -33.75

    -11.25

    11.25

    33.75

    This is why the chirp config does not match the document example.

    2. The document explains how to set the parameter. Section 2 details the phase shift calcuation, and section 3 details how to convert this to the value in the profile Config.

    3. You can use the following matlab script to calculate the phase shift value:

    --------------------------------------------------------------------------------------------------

    theta_desired = -33.75;

    m_ind = [0 1 2];   % antenna distance in unit of lambda  

    phaseInc = 5.625; % step size for the phase shifter

        

    phaseTX_rad = 2*pi*(m_ind*sind(theta_desired));

    phaseTX_deg = (phaseTX_rad)*180/pi;

    phaseTX_deg_wrap = wrapTo360(phaseTX_deg);

    phaseShifter = round(phaseTX_deg_wrap/phaseInc) * [ 0, 2^8, 2^16].' * 4

    ----------------------------------------------------------------------------------------------------

    4. What would you want to see in a Tx Beamforming tuning guide?

    Regards,
    Justin

  • In reply to Justin Curewitz:

    Dear Justin

                      The below value are OK ?  your description is very detail but we are use TI mmwave chip for product design not for university.study.

                      (  Parameters are very much and no direct value look up table for tuning our design become very slow.)

                        we need direct look up table for application , can release look up table or guide for users ?

    1. Look up table  vs  formula :  which one is better for system performance tuning   ?  

    2. We lack of matlab tool for this .

                          

    degree  vs configuration

    -33.75   =   8174592

    -11.25   =   0

    11.25    =   8668160

    33.75    =   16809984

    In the following example, we create 4 subframes.  All 3 TX is enabled at the same time.  These four subframes are aim to focus on the angle of [-15, 0, 15, 30].  
    profileCfg 0 61.0 10 6 135 0 8174592 1.8 1 256 2000 0 0 48
    profileCfg 1 61.0 10 6 135 0 0 1.8 1 256 2000 0 0 48
    profileCfg 2 61.0 10 6 135 0 8668160 1.8 1 256 2000 0 0 48
    profileCfg 3 61.0 10 6 135 0 16809984 1.8 1 256 2000 0 0 48

    Best Regards

    Jasper

  • In reply to Agate:

    Hi Jasper,

    In the document, Beamforming_in_LRPD.pdf, we use the angles [-15, 0, 15, 30].  However, in the chirp configuration, we used [-33.75, -11.25, 11.25, 33.75]

    You can try this python script which is functionally equivalent to the matlab script:

    ------------------------------------------------------------------------------------------------------------

    import math as m

    thetas = [-33.75, -11.25, 11.25, 33.75]

    for theta_desired in thetas:

    m_ind = [0, 1, 2]   # antenna distance in unit of lambda  

    phaseInc = 5.625 # step size for the phase shifter

    phaseTx_rad = [0, 0, 0]

    phaseTX_deg = [0, 0, 0]

    phaseTX_deg_wrap = [0, 0, 0]

    for i in range(3):

    phaseTx_rad[i] = 2*m.pi*(m_ind[i]*m.sin(m.radians(theta_desired)))

     phaseTX_deg[i] = m.degrees(phaseTx_rad[i])

    phaseTX_deg_wrap[i] = round((phaseTX_deg[i]%360)/phaseInc)

    #get sdk value

    phaseShifter = 0

    offset = [ m.pow(2,0), m.pow(2,8), m.pow(2,16)]

    for i in range(3):

    phaseShifter += phaseTX_deg_wrap[i]*offset[i]

    print(phaseShifter*4)

    -----------------------------------------------------------------------------------------------------------

    It will return the values in the chirp configuration in the industrial toolbox, which are: 

    • 14970880.0
    • 10276864.0
    • 6565888.0
    • 1871872.0

    This will allow you to quickly try a different set of angles. Just replace the angles in the list "thetas" with the angles you want to beamsteer towards.

    Regards,

    Justin

  • In reply to Justin Curewitz:

    Dear Justin 

    Thank your fully support .

    Best Regards

    Jasper

  • In reply to Agate:

    Dear Jutin

    try your code as below , result some different .

    0.0
    36864.0
    1871872.0

    Process finished with exit code 0

  • In reply to Agate:

    Hi Jasper,

    Please check that you are using the same array thetas as I am.  Please ensure that indentation is correct in your script.

    Regards,

    JustinphaseShift.txt

  • In reply to Justin Curewitz:

    Dear Justin

                           Check OK (Step by Step) . Thank a lot .

    Best Regards

    Jasper

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.