MMWCAS-RF-EVM: failed to configure advanced frame for large number of chirps

mostafa Alizadeh

Part Number: MMWCAS-RF-EVM

Hi,

I am trying to configure the radar for 192 different chirps for each chip and repeat them 32 times in a burst in a subframe of the advanced frame. Also, these chirps have unique phase shifts by adjusting the Tx phase shifter for each Tx. From "Programming chirp parameters in TI radar devices.pdf section 6" and from the "radar interface control" of "mmwave_dfp_01_02_05_01" table 5.26, I understood that one can define up to 512 different chirps that they can be iterated up to 512 times in a burst. Thus, we can have up to 262144 chirps in frame. However, when I investigated the received chirps I realized that only the first 10*192=1920 chirps have the specific phase shift values that I set for each chirp. Why? What does limit the number of different chirps with different phases?

Best,

Mostafa

over 5 years ago

0 Ishita Kochar over 5 years ago

TI__Mastermind 31680 points

Hi Mostafa,

Are you getting the captured data for the programmed number of chirps i.e 192*32 chirps in the bin file?

Regards,

Ishita

0 mostafa Alizadeh over 5 years ago in reply to Ishita Kochar

Intellectual 735 points

Hi Ishita,

Yes, I am receiving the data in binary as they are stored on TDA.

0 Ishita Kochar over 5 years ago in reply to mostafa Alizadeh

TI__Mastermind 31680 points

Mostafa,

Just to clarify, the maximum number of unique chirps one can have is 512, and the maximum number of loops can be 255. This is still a large number as compared to your use case. I'll try this at my end and let you know soon. Can you please send me your configuration?

Regards,

Ishita

0 mostafa Alizadeh over 5 years ago in reply to Ishita Kochar

Intellectual 735 points

Ishita,

This is the configuration I gave you for the phase shifter calibration in which 3*64 chirps with specific phase shifts are defined and iterated 16 times. Here is the config file: TxBF_Def_Chirps_Calib_dev4.txt

0 Ishita Kochar over 5 years ago in reply to mostafa Alizadeh

TI__Mastermind 31680 points

Hi Mostafa,

I recommend you to post process these kind of configurations using your own script. Inbuilt post processing of the mmwave studio is for evaluation purposes only. You can use the example post processing MATLAB scripts present in C:\ti\mmwave_studio_02_01_00_00\mmWaveStudio\MatlabExamples for your reference.

Regards,

Ishita

0 mostafa Alizadeh over 5 years ago in reply to Ishita Kochar

Intellectual 735 points

Hi Ishita,

I am wondering if you got my question. I am seeing something strange after the post-processing of the received chirps. I expect to get 64*3 with unique phase shifts with respect to the first phase, which are repeated 16 times. For a typical test, I repeated chirps 32 times in a frame instead of 16, and I got the following after post-processing:

Clearly, the phases are not jumped as I set for all chirps in a frame. It seems that the first 10 repeated set of 64*3 chirps are followed by the phase shifts, and the rest of the chirps have the same phase shifts. I think it comes with a limitation on the number of chirps that can have a phase shift within a frame, which I assumed to be 512*512 chirps. Please correct me if I am wrong since as far as I continued processing with a large number of chirps in a frame, I am encountering more strange things. I guess that the number of different chirps with different phase shifts is not 512*512. Or, in general, I think we can have only 512 different chirps in a frame totally.

Best,

Mostafa

0 Ishita Kochar over 5 years ago in reply to mostafa Alizadeh

TI__Mastermind 31680 points

Hi Mostafa,

I thought you're facing issues because of studio processing. I get your problem now.

By looking at your script, I see that you're only configuring unique 64*3 phase shifts for Slave 3 (Device Id: 8). All the other devices have the same phase shifts for all the chirps. Is this what you're trying to do?

Regards,

Ishita

0 mostafa Alizadeh over 5 years ago in reply to Ishita Kochar

Intellectual 735 points

Hi Ishita,

Yes, there are 64*3 chirps configured for only Slave3 (or the most left chip on the board when the receivers are up), and other chips have the same number of chirps but without any phase shift.

0 Ishita Kochar over 5 years ago in reply to mostafa Alizadeh

TI__Mastermind 31680 points

Hi Mostafa,

Are you seeing this discrepancy in the other plots as well or is it just the phase shifter plot? Are you able to see correct 1D and 2D FFT plots?

Regards,

Ishita

0 mostafa Alizadeh over 5 years ago in reply to Ishita Kochar

Intellectual 735 points

Hi Ishita,

Yes, I am seeing this problem in the range profiles as well:

The range profiles with different magnitudes are those without phase shifts meaning that the phase shifter is not applied on them. I realized that I could define only 1920 chirps per frame totally!

Could you tell me what is wrong?

Best,

Mostafa

0 Ishita Kochar over 5 years ago in reply to mostafa Alizadeh

TI__Mastermind 31680 points

Hi Mostafa,

Having the flexibility to program just 1920 chirps per frame is highly unlikely. Can you run the default script that comes with mmwave studio package and try to post process it? Just ensure the total number of chirps are greater than 1920.

This experiment would probably give us some clarity if there's a genuine issue or something wrong with your configuration.

Regards,

Ishita

0 Ishita Kochar over 5 years ago in reply to Ishita Kochar

TI__Mastermind 31680 points

Hi Mostafa,

Even I'm seeing this issue on my part. Let me consult the design team and get back to you. Thankyou for patience.

Regards,

Ishita

Sensors

Sensors forum

MMWCAS-RF-EVM: failed to configure advanced frame for large number of chirps