AWR1843AOPEVM: Antenna geometry for AWR1843AoP for People counting demo

Hi Vignesh,

I believe you meant to post this in a different thread. Could you provide your reply there instead of here?

Thanks,

Clinton

The original thread was shown as "locked" . Hence I created a new request

Hi Vignesh,

Is this the thread you are referring to? There was no response to our last communication for months, so it was locked. If it's a different thread, please post a link to it here.

https://e2e.ti.com/support/sensors-group/sensors/f/sensors-forum/1198405/awr1843aopevm-runtime-error-people-counting-demo-on-awr1843aopevm

Thanks,

Clinton

Hello Clinton,

This is the thread I was referring to. Sorry I started working on this only now, due to other commitments.

Regards

Vignesh

I will let Nathan know you have posted here.

Thanks,

Clinton

Hi Vignesh,

What happens when you run the demo with the CLI arguments last suggested: 

antGeometry0 0 1 2 3 0 1 2 3 0 1 2 3
antGeometry1 0 0 0 0 -1 -1 -1 -1 -2 -2 -2 -2
antPhaseRot 1 1 1 1 1 1 1 1 1 1 1 1 

Does the demo run at all? Or does it just show points at incorrect positions?

Best,

Nate

Hello Nathan,

The demo does not run at all if the above configuration is used. However, I could run with the elow configuration:

antGeometry0 0 0 0 0 1 1 1 1 2 2 2 2
antGeometry1 0 1 2 3 0 1 2 3 0 1 2 3
antPhaseRot 1 1 1 1 1 1 1 1 1 1 1 1

But I did not check with the correctness of the angles. I just wanted to share this update with you.

Could you please confirm if the configuration is correct. And additionally, I wanted to keep this thread open, so that I can ask for any problems I have regarding my implementation.

Thanks and Regards

Vignesh

Hi Vignesh,

I don't think this configuration is correct. However, please run the demo with no other changes to confirm/reject my hypothesis. Using the documentation in the detection layer tuning guide:

https://dev.ti.com/tirex/explore/node?a=1AslXXD__1.20.00.11&node=A__AIQPG9x7K34A8l4ZELgznA__radar_toolbox__1AslXXD__1.20.00.11&r=1AslXXD__1.10.00.13

It looks like antGeometry0 should correspond to azimuth and antGeometry1 should correspond to elevation.

Using this image from our last thread, it looks like the virtual TX/RX channels are at the following locations (azimuth, elevation):

TX1, RX1 = (0,0)

TX1, RX2 = (-1, 0)

TX1, RX3 = (-2,0)

TX1, RX4 = (-3, 0)

TX2, RX1 = (0,-1)

TX2, RX2 = (-1, -1)

TX2, RX3 = (-2,-1)

TX2, RX4 = (-3, -1)

TX3, RX1 = (0,-2)

TX3, RX2 = (-1, -2)

TX3, RX3 = (-2,-2)

TX3, RX4 = (-3, -2)

This would lead to the configuration of 

antGeometry0 0 -1 -2 -3 0 -1 -2 -3 0 -1 -2 -3
antGeometry1 0 0 0 0 -1 -1 -1 -1 -2 -2 -2 -2
antPhaseRot 1 1 1 1 1 1 1 1 1 1 1 1

Do you agree? Does this configuration work? If not, what is the error message when you put it in debug mode?

Best,

Nate

Hello Nathan,

The error I got from your configuration was that the num_Tx antennas, num_virtual_antennas calculated from the configuration is 0. The code does not run after that point.

I took the configuration I suggested from the Obstacle detection demo for 1843AOP