Channel sounding issues

[Patrick49]

Hello, 

I'm having some issues with the channel sounding on the nRF54L15.
I'm using two nRF54L15-dk and two custom bords with the nRF54L15.I'm using SDK v3.0.2 for my software and work on Windows 11.

The first one, is that I'm getting some broken ifft values equal to 0, sometimes, as shown below. This issue is happening both with the dk and my custom boards.

Since the ifft should be the best measuring method, it's problematic to have some null values.

The second one is with a custom board. When I run my program for about 30-35 minutes, a warning about ranging data overwritten shows up. Therefore, it doesn't stock new values after 4096 values ? and we need to reboot if several errors appear to resume normally the measurements. 

This warning doesn't show up with the dk when I run the same code and for a longer period of time as well. 

My code can be found in the archive in the attached documents.

CS_code.zip

Thanks and best regards,

Patrick 

[Simonr]

Hi

1. Can you share some more details on your testing environment here, and if you've made any specific changes to the channel sounding sample projects?

2. Regarding the "Ranging data overwritten". You can try increasing CONFIG_BT_RAS_RRSP_RD_BUFFERS_PER_CONN to allow more rangings in one connection.

Best regards,

Simon

[Patrick49]

Hello, 

1. I've tested both the default channel sounding sample as well as my modified version that only weight the average of the three methods. The test was indoor, both devices were placed on a table in a short range.

2. I tried this option on the reflector since it's this device that has de RSSP defined. But it doesn't change anything with the default channel sounding sample, whether I put the value at 1 or 10. This picture is taken after 35 minutes of testing, the logs can be found in the attached document as well. 

Logs_CS_40min.txt

Anyway, I'll try to modify my program so it doesn't only loop on those measurements so that I won't get this error again. Also I'm seeing that some values are doubled.

Best regards, 

Patrick