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.

AWR1843: FOV problems

Part Number: AWR1843

Hi,

we are facing a issues with AWR1843. I will describe the situation as follows: during the module (AWR1843) testing process, we identify false detection when there isn't an object in the way of the sensor's data as seen by the MRR/SRR visualizer. When a sensor is in a static condition and mounted at a height of one meter, fault detection does not occur. When the sensor is in a moving condition, fault detection occurs at random. Kindly helps us to resolve this problem.

  • Thanks Sriram for your question.  

    A few follow up questions:

    1. What is your configured Vmax from your chirp configuration?  If your Vmax is too low and your sensor is moving very fast, then you could get false detections.  

    (+) AWR1843: How to correct erroneous detection caused by radar shaking - Sensors forum - Sensors - TI E2E support forums

    2. You can also try to adjust the CFAR threshold.  A higher threshold will make you less vulnerable to false detection.  

    the above thread that I link to has some good suggestions to resolve erroneous detections. 

    Regards,

    James

  • Hi,

    which CFAR threshold should be changed to avoid false alarm.


    obj->SNRThresholds[0].rangelim = (uint16_t) (10.0f * (float)(1 << obj->xyzOutputQFormat));
    obj->SNRThresholds[0].threshold = convertSNRdBtoThreshold(1, 18.0f,CFARTHRESHOLD_N_BIT_FRAC);

    obj->SNRThresholds[1].rangelim = (uint16_t) (35.0f * (float)(1 << obj->xyzOutputQFormat));
    obj->SNRThresholds[1].threshold = convertSNRdBtoThreshold(1, 16.0f, CFARTHRESHOLD_N_BIT_FRAC);

    obj->SNRThresholds[2].rangelim = 65535;
    obj->SNRThresholds[2].threshold = convertSNRdBtoThreshold(1, SUBFRAME_MRR_MIN_SNR_dB, CFARTHRESHOLD_N_BIT_FRAC);

    /*! @brief peakVal thresholds are varied as a function of range */
    obj->peakValThresholds[0].rangelim = (uint16_t) (7.0f * (float)(1 << obj->xyzOutputQFormat));;
    obj->peakValThresholds[0].threshold = 4250;

    obj->peakValThresholds[1].rangelim = 65535;
    obj->peakValThresholds[1].threshold = 0;

    obj->peakValThresholds[2].rangelim = 65535;
    obj->peakValThresholds[2].threshold = 0;

  • Hi Sriram,

    apologies for the delay, I am getting the answer.

    Regards,

    James

  • HI,

    we identify false detection when there isn't an object in the way of the sensor's data as seen by the MRR/SRR visualizer. When a sensor is in a static condition and mounted at a height of one meter, fault detection does not occur. When the sensor is in a moving condition, fault detection occurs at random.

    Hi,

    A radar sensor performs usually better when there is mouvement.

    Usually fault detections can be caused by the Hardware. Are you using a custom board? Could you try same experiment with TI board to see if there is a difference?

    Thank you

    Cesar

  • Please see the description for cfar threshold given on page 28 of mmwave_sdk_03_06_00_00-LTS/docs/mmwave_sdk_user_guide.pdf

    You can alter its value in the profile you are using