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.

IWR6843ISK-ODS: Questions about the "Range Bias and Rx Channel Gain/Phase Measurement and Compensation"

Part Number: IWR6843ISK-ODS

Hi all,

In below documentation, we can do some calibration for the antenna.

{mmWave_sdk_DIR}/packages/ti/demo/xwr68xx/mmw/docs/doxygen/html/index.html#Calibration_section

However, I am not quite understand how to do this.

1. What is the meaning of "Set a strong target like corner reflector"? Does it means that place the sensor in front of the wall?

Or we need to place the sensor like below?3 side need to have strong target

┌────────────────────────
│                           ▲
│                           │
│                           │
│                           │ x m
│              x                    
│ ◄─────── Sensor
│                         
│                          │
│                          │ Distance: x m
│                          │
│                          │
│                         ▼
└─────────────────────────

2. Is there any requirement for the height of the sensor when do the calibration?

3. What is the meaning of "search window"? And how to define this parameter when do the test?

 If we set for example 0.25, do we need to move the sensor while testing?

4.  The document state "The window size is recommended to be at least the distance equivalent of a few range bins", what is the meaning of "bins"?

Please help if you know somethings.

Thanks.

Tom

  • Hello

    Have you had a chance to look at how FMCW works and the concepts of Rand, Angle and Velocity bins?

    Training.ti.com  has some of this covered as well.

    For the Calibration:

    Would it be possible for you to try the values and see for yourself the correction  with and without  it?

    Thank you,

    Vaibhav

  • Hi Vaibhav,

    Thanks for your advise. I will check those concepts of the radar system.

    Let me explain why we want to have calibration.

    We have a custom pcb which is the same design of the IWR6843ODS.

    However, we found that the range of radar sensing area can't reach 6m x 6m , but just  ~3m x ~3m.

    We test it with the overhead_demo_visualizer.

    We can see from the gui.

    When the target running out from the sensor ~1.5m, the target stop on this area even the target is going out further.

    So, we try to do the calibration to see if we can get some improvement. 

    Thanks,

    Tom

  • Hi, Tom:

    For calibration, you want to use a point target (such as corner reflector) instead of a wall.  And the you want to place this point target at the bore sight of the sensor.  

    For overhead mounting, depend on the mounting height, the distance may be limited by the FOV (field of view).  You can try to mount a little bit higher to see any improvement.

    Best,

    Zigang

  • Hi Zigang,

    If we don't have corner reflector, what is the alternative targets that we can use? Can we use just a piece of metal plate?

    There are many size of corner reflector. Which size of corner  reflectors we can use?

    What is the meaning of "bore sight"? Should not have any objects between the target and the sensor?

    We tested and not much improvement of the range if we mount a little bit higher.

    There are no issue on the IWR6843ISK-ODS evaluation board with the same cfg file and set up.

    Do you know why our custom pcb will have such narrow FOV(almost by half compare with the EVM )? Antenna issue?

    Is there anything we can do to increase the FOV of our custom pcb.

    Thanks.

    Best regards,

    Tom

  • HI, Tom:

    Yes, a metal plate should work too.  The size of corner reflectors depends on the distance, you want a small corner reflector like 4cm if you placed at 0.5m or 1m range.  Bore sight means zero degree in azimuth and zero degree in elevation relative to the sensor.   And you do want to have clean environment (no clutter) around the target range.  

    If the FOV is worse then our EVM and you want to redesign your antenna, please send a separate e2e thread. 

    Best,

    Zigang

  • Hi Zigang,

    Thanks for your explanations. It helps a lot.

    Few more questions about the calibration procedure.

    In profile_calibration.cfg, there are parameter "measureRangeBiasAndRxChanPhase 1 X D" which is set the calibration.

    We know "X" is the distance between the corner reflector and sensor.

    What is the meaning of "D" search window ?How do we set it correctly?

    How it will affect the calibration result? e.g. if we set to 0, will there have bad calibration result?

    Thanks.

    Tom 

  • Hi, Tom:

    It is explained in the SDK Doxygen Documentation.  file:///C:/ti/mmwave_sdk_03_05_00_04/packages/ti/demo/xwr68xx/mmw/docs/doxygen/html/index.html#Calibration_section

    Usually I would set to about 0.1 or 0.2m.   It depends on how confident you are about the distance.  If you are not very confident about the target range, you can set to 0.2m, but be sure there is no larger clutter within 0.2m of that target. 

    Best,

    Zigang

  • Hi Zigang,

    I have read this documentation many times before.

    I don't understand the concept of "search windows".

    So,  the "search window" will affect the sensor how it send the calibration signals?

    For example, if I set 0.2m, then the sensor will try to find this target within the range? Like below diagram?

                              +------------------+
           0.2m            |                      |       0.2m
    <-------------------- | Target           |  ---------------->
    ^                           ------------------+                    ^
    |                                                                        |
    |                                                                        |
    |                      Sensor calibration                     | 
    |                        signal range                            |
    |                                                                        |
    |                                                                       |
    |                                                                       |
    |                        +------------------+                     |                       
     |-------------------|      sensor      | ------------------|
                             +------------------+  

    Please advise if I understand correctly.

    Thanks.

    Tom

  • Hi, Tom:

    After range and Doppler FFT, the algorithm need to find the peak.   For the calibration, instead of doing a CFAR for peak search, it will only do max peak search on zero Doppler bins around range window (X-D,  X+D), where X is the rough target distance and D is the search window size.  Your picture shows the x and y, but the algorithm only observe the signal power over range (sqrt(x^2 + y^2)).   

    Best,

    Zigang