CC1312R7: Problems with interoperability using the BZ Phy, between Silabs and TI's solutions

Hi Thiago,

thank you for reaching out to us.
The issues that we found during our last testing got resolved with the SDK version 8.30.1.1: https://www.ti.com/tool/download/SIMPLELINK-LOWPOWER-F2-SDK/8.30.01.01
Do I understand correctly that you use this version for testing?

Looking at your captures I can see that in the BZ capture both TI nodes complete the security exchange can you help me to understand the issue that you see compared to the NA capture and also explain me what kind of device 1c07 is?

Kind regards,
Theo

Hi Theo,

We are indeed using the mentioned SDK version.

1c07 is a node that should be ignored.

interoperability_logs.zip

We have generated new captures for BZ and NA regions. The devices are the same as the previous setup, and for this test there are 2 captures, a tcpdump capture and a capture generated by Silabs Border Router daemon (wsbrd). In the .zip file attached by Thiago, it was just the capture generated by wsbrd.

There is also a wsbrd_BZ.log file, which is the log of wsbrd for the BZ test. We are aware this log is not from TI, but it could be useful to understand the problem.

Even though the BZ US-IE problem was resolved, there are still some connection issues. For instance, f024 completes the security exchange, but then it starts sending "PAN Advertisement Solicit" again.

Thank you.
Regards,

Mauricio

Hi Theo,

We have contacted Silicon Labs, and they mentioned that the BZ band has changed the PHY specifications (Revision 2V03). Did TI implement these changes?

Thank you.

Regards,

Mauricio.

Hi Mauricio,

thank you for sharing all the additional information.
I'm in contact with our R&D experts to check the logs and the mentioned specification.

Following the behavior of the TI router node f024 I see that it was sending PAS messages but also data messages. Was the node fully joined so that you could communicate to it and did it start then sending PAS messages in certain intervals?

I'm trying to trace down in which state the nodes state machines are when this behavior occurs.

Kind regards,
Theo

Hi Theo,

The f024 node never fully joined to the network.

I'm not sure which states the nodes are, because the ns_trace logs are disabled. I have disabled the traces because the firmware may throw a hardware exception when the traces are enabled.

However, from the blink pattern, I can at least say that the nodes never went to state 4.

Thank you,

Regards,

Mauricio.

Hi Mauricio,

thank you for the clarification. We are in the process of checking the PHY and the shared data.
I will update you as soon as I got feedback.

Kind regards,
Theo

Hi Theo,

We found the problem, it was indeed the PHY specifications (Revision 2V03) changes.

We tested a version of Silicon Labs border router before the BZ PHY specification changes, and the TI nodes connected without any issues.

Thank you.

Regards,

Mauricio.

Hi Mauricio,

thank you for the update and the test!
I'm following up with R&D to implement the changes according to this revision so that the issue will be resolved with future releases.

Kind regards,
Theo