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.

AWR1843AOPEVM: calibration of the AWR1843AOPEVM for MRR

Part Number: AWR1843AOPEVM
Other Parts Discussed in Thread: AWR1843AOP

Tool/software:

Hello Ti team,

I am using the MRR application on the AWR1843AOPEVM and logging radar data over a 500 kbps baud rate in Classic CAN_frame.

For testing, I moved from 6 meters towards the radar, staying within the -1m < x < 1m range. The 0xD1 message consistently appears in the packets, but my position data is not always present. Sometimes, my position appears in one packet but is missing in the next, and  some times it occurred the 4th packets of radar after getting first data .

please suggest  any solutions?

Thanks

Mitesh

  • Hi,

    I recommend using first the MRR Matlab GUI first and make sure that it works as desired. 

    Usually MRR should be tested at larger distances

    thank you

    Cesar

  • Hello Mitesh,

    If this thread is finished lets resolve it.

    Best,

    Gabriel

  • Hello Cesar/Gabriel,
    The application is running in MRR_USRR mode, which means it is capable of detecting nearby objects as well. So, if I move toward the radar within a distance of 6 meters, my position data should be included in the regular packets generated by the USRR configuration.

    if i am wrong please correct me .
    Thanks
    Mitesh

  • Hi,

    In order to try to get better performance at close range you can utilize the USRR 20m config. In order to use it you need to access mrr_config_consts.h in the common folder. From there you can change the USRR config used at this location.

    I provided the USRR20 config as well if not already present (place in common folder as well). After this is done import the projectspecs to CCS and rebuild.

    mrr_config_chirp_design_USRR20.h

    Best,

    Gabriel

  • Hello Gabriel,

    After including mrr_config_chirp_design_USSR20.h and building the project, I tested it using the MRR GUI Visualizer. The static objects are not being displayed. Does this application detect static object ?

    For moving objects,  points do appear, but I noticed some points showing up at farther distances, like around 7 meters—even though there is no actual object at that range but i am moving under 4 meter .

    Thanks 
    Mitesh

     

  • Hello,

    After conducting further research, it seems that the MRR demo was not intended for use on the 1832AOP it was intended for use on the 1843BOOST due to temperature constraints. The reason there are issues with the USRR subframes is because they are using MIMO and it will be negatively affected in detection as you are seeing. This E2E could be helpful since they calculated the appropriate USRR on the AOP board. But, to clarify the MRR demo was not meant to be run on the AOP board, and it explains the issues you are seeing.

    (+) AWR1843AOPEVM: AWR1843AOP - Sensors forum - Sensors - TI E2E support forums

    Best.

    Gabriel

  • Hello,

    Thank you for the clarification regarding the MRR demo and its limitations on the AWR1843AOP due to temperature constraints and MIMO-related issues in the USRR subframes.

    If the MRR demo is not intended for use with the AWR1843AOP, could you please suggest which demo or application is best suited for this device to detect both static and moving objects effectively?

    I am specifically looking for a reliable configuration or lab demo that is optimized for the AWR1843AOP hardware.

    Looking forward to your recommendation.

    Thanks
    Mitesh Singh

  • You have two options that I know of for demos for the AWR1843AOP. One is present in the mmwave_sdk_03_06_02_00-LTS at the following path. The binaries needed are present here

    C:\ti\mmwave_sdk_03_06_02_00-LTS\packages\ti\demo\xwr18xx\mmw\

    Another option is present in an older radar toolbox version. Here is the link to the user guide to get started.

    Obstacle Detection AOP

    Best,

    Gabriel

  •  

    Hello Gabriel,

    Thank you for your response.

    I have a few follow-up questions regarding the demo binaries and firmware you suggested, specifically from the following directories:

    • C:\ti\mmwave_sdk_03_06_02_00-LTS\packages\ti\demo\xwr18xx\mmw\

    • Obstacle Detection AOP (for AWR1843AOP)

    Could you please clarify the following:

    1. Detection Range & Field of View (FOV):
      What are the typical detection range and field of view supported by these demo binaries? Are there any configuration files where these parameters can be tuned?

    2. CAN Interface Support:
      Do these demos transmit object data (e.g., TLV output) over the CAN interface? If not, driver source code available that shows how to implement CAN data transmission for these binaries?

    3. Object Detection Capabilities:
      Do these demos support detection of both static and dynamic objects? If so, is there a way to distinguish between them in the output data?

    4. Performance and Use Case:
      Are these binaries optimized for any specific use case (e.g., parking assist, short-range radar, obstacle detection), or are they intended as general-purpose demos?

    5. Data Output Format:
      Is there documentation available that describes the TLV output structure used in these demos?

    I’m particularly evaluating these binaries for use with the AWR1843AOP, so any additional guidance or documentation specific to this device would be greatly appreciated.

    I have an additional question regarding the MRR application.

    If I enable only the #define SUBFRAME_CONF_MRR  and comment the #define SUBFRAME_CONF_USRR,( Means only MRR mode)

    will the application work on the AWR1843AOP? 

    My primary objective is to detect both static and dynamic objects up to a distance of 100 meters. I chose the AWR1843AOP because TI mentions that it supports detection ranges up to 150 meters.

    Given this requirement, I’d like to know whether the suggested demo binary fulfills these needs, or if there is an alternative firmware or source code that would be better suited for achieving reliable object detection at that range. Could you please advise?

    Thank you once again for your support!

    Best regards,
    Mitesh

  • Hello,

    I will give you background on how to access this information for the SDK demo.

    1) You can access this information by looking at the profile present in the SDK 

    C:\ti\mmwave_sdk_03_06_02_00-LTS\packages\ti\demo\xwr18xx\mmw\profiles\profile_3d_aop.cfg

    It has descriptions of the profile in the same file.

    2) There is a CAN driver example also present in the SDK that can be accessed here

    C:\ti\mmwave_sdk_03_06_02_00-LTS\packages\ti\drivers\can

    3) Yes, demos have support to detect both static and moving objects. In order to distinguish them you are going to have to do research into the TLV's and see how to do that because I do not believe any of the 1843 visualizers can do that.

    4) Again, refer to the profiles in order to observe performance and what they are optimized for. There is even a tool which I will provide a link for where you can input parameters for a profile in order to see its performance. It has compatibility with the 1843 so you can use this to try to create new configs with desired characteristics. You can even input known profile parameters such as "profile_3d_aop.cfg" to get more details on its expected performance.

    mmWaveSensingEstimator

    5) Regarding data output format there is more information available in the mmw demo for awr18xx documentation present here.

    C:\ti\mmwave_sdk_03_06_02_00-LTS\packages\ti\demo\xwr18xx\mmw\docs

    Navigate to this file path and open the index.html file. I will provide a link as well if you cannot find it. This provides specifics of the SDK demo and some limitations. Navigate to the section "Output information sent to host" to get more info on the TLVs.

    Millimeter Wave (mmw) Demo for XWR18XX

    For the MRR demo, yes it will work as intended with the MRR subframes. You are going to need to change the profile in the MRR demo to achieve desired parameters in a respective .h file such as mrr_config_chirp_design_MRR80.h, the estimator tool can help.

    Please as this thread has deviated from the original question quite a bit, if you have further questions please create a new thread so we can help the best we can.

    Best,

    Gabriel