AWR1843: How to perform calibration on custom board

Hi, 

Which specific calibrations are you trying to perform? We have range bias and phase compensation parameters we can collect without a DCA1000.

Below you can find information about how to perform the calibration.

<SDK_INTSALL_DIR>/packages/ti/demo/xwr18xx/mmw/docs/doxygen/html/index.html#Calibration_section

Regards,
Tim

Dear Tim,

I went through the process of phase compensation and got the 24 (12real + 12 imaginary) parameters to understand the phase compensation. with the visualizer i was able to verify that the calibration is actually working and the shift of FOV has be removed. 

But when i try to implement that compensation in the medium range radar from Automotive toolbox following the location mentioned in https://e2e.ti.com/support/sensors-group/sensors/f/sensors-forum/1120155/awr1843-where-are-the-calibration-parameters-in-automotive-lab-demo-mrr this thread. I updated the 12 values for the point cloud by converting the values as per the Q15 format using the tool https://chummersone.github.io/qformat.html.

But i m still getting the same kind of response from the radar. I have updated the POINT_CLOUD_PROCESSING values. but how to get the calibration parameters for MAX_VEL_ENH_PROCESSING.

These are the set of 12 values:

How to get the MAX_VEL_ENH_PROCESSING parameters for the Phase compensation?

Regards
Kunal

Hi,

For your application do you use both POINT_CLOUD_PROCESSING and MAX_VEL_ENH_PROCESSING?

Thank you

Cesar

Hi Cesar,

For my application, I am using only the MAX_VEL_ENH_PROCESSING.

Regards

Kunal

Hi Kunal,

Author of this application is currently unavailable. Can you please wait till first week of January for more details?

Regards,

Samhitha

Dear Samhitha,

I am not in a condition to wait this long. Already the turnaround time is high via E2E forum. I have to wait 24 hours to get a reply. The product we are working on is going into production from Q1 2024-25. We have to complete engineering and testing by 1st week of january2024.

This is going to be a big business impact. Hence we are looking for priority support. Kindly arrange something to resolve this on priority.

Regards
Kunal

Hi Kunal,

Hi Samhitha,

Thank you for the response, We are using MRR120, which uses all the 3 TX channels. I got to understand the process but i am not able to get the CFG file to work with the live visualizer. file generated using https://dev.ti.com/gallery/view/mmwave/mmWaveSensingEstimator/ver/2.3.0/

This is the config file i was able to get as per my understanding. But when i run this config file i get "Exception: ./mss/mmw_cli.c, line 274." in the Console.

If you can get me the cfg file for this application it will also be helpful.

Regards

Kunal

Hi,

The SDK mmwave demo does not support enabling 3 TX at the same time.

Can you please try the following:

For the 12 calibration values you have collected, can you please average the values for Tx1, TX2, TX3?

In the data collected, the first 4 values should be for TX1, the next 4 values for TX2 and the next 4 values for TX3

thank you

Cesar

Thanks Cesar for the response,

I have implemented these values as you have suggested. The performance also improved. But there is still some misalignment that i am observing. I hope there is a better way to perform calibration on the MRR and get a precise calibration parameters.

Regards

Kunal

Dear Cesar,

Any update on this calibration process. I have tried doing some calibration and got to understand that i am not able to get the MRR to align correctly at the center.

I hope this screenshot helps you understand my issue.

Here 2 trackers are there at 10 meters. Both of these are of Same corner reflector placed in front of the Unit. The point at center is the USRR and the Point which is slightly right side is MRR.

I have followed the same process of averaging the calibration values i have received as you have suggested earlier. but worked for the first time, the accuracy was acceptable but its not the case for all. The shift in the above screenshot is not acceptable. and this is also after the 3rd time i have performed the calibration on the same device.

Kindly provide the resolution at the earliest. we are already behind the timeline and there is still no concrete update on getting the process of getting the MRR calibration parameters.

Regards
Kunal

Dear TI team,

Any update on this?

Hi Kunal,

Thanks for your patience.
We are trying to find an alternative solution. We will get back to you as soon as possible.
Did you perform calibration for both MRR120 and USRR configuration? Did you update the parameters for both the processing paths (if you are using both of them) in rxGainPhaseParam_Init function?

Regards,

Samhitha

Hi Kunal,

If you have a DCA1000 board to capture raw data, then try referring to the following E2E.
AWR1843: Where are the calibration parameters in automotive lab demo mrr? - Sensors forum - Sensors - TI E2E support forums

Regards,

Samhitha

Hi Samhitha,
Thank you for the response. But i have already gone through this thread. My issue is the radar I want to calibrate is on a custom board. It doesn't have any provision to connect DCA 1000 over LVDS port. Thats the reason i am not able to capture raw data.

I was able to get the parameters for USRR but averaging them and then getting the parameters for MRR and not giving me accurate results.

Looking forward to your response.

Regards
Kunal

Hi Kunal,

Can you try making changes to the code to capture a frame of ADC data on L3 and then take a memory dump in .bin format from the CCS?
After capturing the ADC data, you can perform the steps listed in the E2E thread.

AWR1843: Where are the calibration parameters in automotive lab demo mrr? - Sensors forum - Sensors - TI E2E support forums

Regards,

Samhitha

Hi Samhitha,

It will be helpful if you can point me to the right function which is processing this value in the MRR. Since the code is quite huge and i am recently assigned to this task and i don't know where in the code this L3 ADC data is being processed.

Thanks and regards

Kunal