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.

IWR6843AOPEVM: Is it possible to reduce the speed resolution to 0.01 m/s?

Part Number: IWR6843AOPEVM

Demo visualizer: https://dev.ti.com/gallery/view/mmwave/mmWave_Demo_Visualizer/ver/3.5.0/ 

Sensing estimator: https://dev.ti.com/gallery/view/mmwave/mmWaveSensingEstimator/ver/1.4.0/ 

I tried the above two methods.

It was impossible for me to reduce the velocity resolution by 0.01 m/s in two methods, but how much is it possible to reduce the velocity resolution by the method of the respondent or other source?

  • HI, Jeongwoo:

    The velocity resolution can be lowered by

    1) increasing the chirp repetition time, this can be achieved easily by increase the idle time.   Note that this will reduce the maximum velocity time

    2) increase the num-loop in the frame.   This may have a limit due to memory limitation. 

    Note that both of these two steps will increase the frame duration. Hope it is fine with you. 

    You can find the formula of velocity resolution in the section 2.2 of the following document.  

    I understand that the tools you mentioned may not have all the flexibles to design the chirps.  You can start with the one you have and keep increase the two parameters I mentioned until it reaches your desired level.

    Best.

    Zigang

  • Can I know the best velocity resolution value that exists so far with this product?

  • HI, Jeongwoo:

    There is not a really a limitation on how low velocity resolution can be.   As an example, in the in-cabin sensing demo,  we have following configurations:

    profileCfg 0 60.00 205 11 41 0 0 97.0 1 64 2200 0 0 36
    chirpCfg 0 0 0 0 0 0 0 1
    chirpCfg 1 1 0 0 0 0 0 2
    chirpCfg 2 2 0 0 0 0 0 4
    frameCfg 0 2 220 0 200 1 0

    There are 3 chirp ID that covers TX1, TX2 and TX3, each chirp is with duration 205 + 41 = 246us. So the total chirp repetition time is Tc = 246*3 = 738us

    The velocity resolution is labmda / (2 * Tc * 220) = 4.0e-3 / (2*738e-6*220), which is about 0.0123.   Because in this demo, we are trying to identify the very small movement of people inside the car.

    Note that we have num_loop = 220, and frame duration is 200ms.  

    You can always further increase num_loop and idle time to lower your velocity resolution further.  Again, there is not really any limitation.  But you have to consider the total radar cube memory you can use (768K for xwr6843), the maximum velocity you may want to have and the frame duration you would like to have.   

    Best,

    Zigang