Thread devices can't connect my Thread border router.

[Assaf Tal]

Hey all,

In order to research Thread protocol and its behaviour, I bought a Nordic dongle nRF52840 and used it to make a Raspberry pi a Thread border router.

Currently, I succeeded to make my Raspberry pi a thread border router, I can control the Thread network details and can list them (parameters like External PAN ID, Network name, Channel, etc). My Raspberry pi is connected directly to a router with Ethernet cable and I have internet connectivity from my Thread border router (can ping Google DNS with my IPv6 from the Raspberry).

My next goal is to connect a Thread device to the Thread network. I have a Nanoleaf smart light bulb that supports Thread that I tried to connect to my Thread network, unfortunately with no success. When my Nanoleaf smart bulb is looking for Thread available networks, it can't find one, however when I look under settings under Thread networks section I can see my Thread border router with only the following parameters: External PAN ID and network name (don't see other necessary information like Channel number, pass key, etc).

My question is: can I connect a Thread device to this network? I know that there is a DK from Nordic that I can buy and configure it as a Thread device, but I want a "real" Thread device to connect to this network. Maybe there is an option to publish the other Thread network parameters to make the Nanoleaf smart bulb find the network while it looks for Thread available networks?

Thanks,

Assaf

[Izvor]

It should be possible to join the Thread network with a Matter over Thread device. I haven't tested with a production device yet, but I have( as you have done) created an OpenThread Border Router using a nRF52840 Dongle. Then I have created a Matter over Thread device using a Silicon Labs EFR32xG24 Dev Kit. I then joined this device to the Thread network and it's working fine. You should not need to be concerned about the channel number etc. You do of course need to use the correct Thread dataset in the Matter pairing command. How did you try to pair your device to the Thread network? Did you use the chip-tool command?

[Assaf Tal]

Thanks for your reply!
I know that I can buy a DK from Nordic (www.nordicsemi.com/.../nrf52840-dk) and configure it as a Thread device that will be able to connect to my Thread border router.

I want to be able to connect a "real" Thread device which I can control him by it's appropriate app (this time a Nanoleaf smart light bulb). 

The pairing process is by scanning the QR code on the device and the device is looking for available Thread networks. I have no control on the pairing process here, but I thought the device should make it automatically.    

[Izvor]

What app are you using for provisioning / pairing the device? For better control, you could consider using the chip-tool directly to test.

[Assaf Tal]

Nanoleaf app for mobile: https://nanoleaf.me/en-US/integration/mobile-app/.

This is how I should connect this specific device. Moreover, I can find with the application other Thread border routers (for example Google home Nest router that it finds with all the details it needs to connect to the network).

Nanoleaf forum notes that the following border routers are supported: 

1. Nanoleaf
   1. Nanoleaf Lines
   2. Nanoleaf Shapes
   3. Nanoleaf Elements
2. Apple
   1. Apple HomePod mini
   2. Apple TV 4K (2nd Generation)
3. eero (how-to)
   1. eero 6
   2. eero 6 Pro
   3. eero Pro
   4. eero Beacon
4. Google
   1. Nest Hub (2nd Generation)
   2. Nest Hub Max
   3. Nest Wifi
5. Amazon
   1. Echo (4th Generation)

https://helpdesk.nanoleaf.me/en-US/essentials-thread-and-thread-border-routers-15563

Maybe it can't connect my thread border router because it is a raspberry pi that acts as a Thread border router and not one from the list above?

[Izvor]

I have no experience with the Nanoleaf app, but it looks like the reason might be that it only supports the border routers listed. I think your best option may be to use the chip-tool directly. It can be built on a Raspberry Pi with sufficient memory. For me it failed on 4GB and worked on 8GB. Or you can build it on a PC with Ubuntu.