None of my aqara devices have ever routed through them.
they don’t change parents very readily if at all, I’d still recommend joining through the nearest routing device.
I was having trouble following the Specific Configuration for the Tube EFR32 Gateways, and don’t have this part in in config. Would that cause my issue?
no, as of last HA there was a change to use the fw settings not hard coded settings in the bellows library.
the only thing to consider is if you want source routing on or off.
How do I do enable that?
zha:
zigpy_config:
source_routing: true
I will try that now.
Also I am getting this warning: " * Zigbee channel 20 utilization is 87.33%!"
you can look in the diagnostics data to see what the energy scan looks like for all the channels.
2023.6 will include a button in ZHA to move the channel:
From current beta:
I’ve run into an issue with my Ethernet base coordinator when recovering for power loss. We have had a few power outages recently (Electrical work on the house, and weather events) and every time once power has been restored the coordinator needs to be power cycled.
Power will come back, and no zigbee devices will be working, so i go into HA, and see Z2M isnt running. I restart Z2M, but it refuses to start citing the coordinator isnt present. I restart the coordinator, then start up Z2M and all is fine.
Is this a normal issue? It seems like if the coordinator starts up before the network is restores (takes a couple minutes to boot up the unifi gear) then it doesnt start up properly?
Hi @Eatoff ,
I would also lose connection with the coordinator every 1 month or so. I did not think of possible power outage (I don’t recall having any and no storms our side) but the symptoms are the same yours:
Zigbee devices are not reporting and only restarting the coordinator via its IP address helps.
Good to know it’s not only mine and I wonder what could be causing it?
Have you tried a full re-flash of the esp32?
I’ve heard of this a few times but nothing widespread and have never been able to replicate it. I wonder if there is a timing issue with dhcp, and if using a build of esphome with static ip would help.
Let me know if you want to try that. The yaml is online with commented out static ip config. Happy to work with you or anyone on this.
1 Like
Happy to give that a go. I have the coordinator added via the ESPHome integration, but no config or anything for it like I do for the AirGradient in the screenshot:
Shoot me a private message with the ip and the gateway and I’ll compile a binary for you to try.
1 Like
Thanks, sent you a PM
So upon power restoration today, had the same issue (that was with updated firmware, not the custom version with static IP you made for me).
How the issue presents is Z2M being unable to start. When attempting to access the coordinator via browser it times out. Power cycling the coordinator brings it back online again.
I’ll give your custom firmware a go tomorrow. Just updating here for anyone following along
Hi all,
I need some (a lot of) assistance.
I have an old Tube gateway (square model, with ethernet and USB. The PCB says efr32 pro v2.1)
It has started to play up where it would seem to drop off the network (the ESP side of things pin low/reset etc) would become unavailable and then the zigbee devices would drop off the network. To me that points to a network/ESP/power type issue.
I have decided to blow the dust off it and update it’s firmware/recheck all the connections. I have successfully updated the ESP firmware, but I’m struggling with what files I need and how to update the other firmware(s)?
There is NCP and bootloader, and within those there are different files?
I have seen the youtube video about setting the bootloader pin, rebooting the gateway and sending the firmware file, but I don’t get a response in my serial window.
Please help…
Take a look at the hardware - is the usb port on solidly? This can sometime become loose, but this is more common in situations when the device falls with a cable in the port.
Updating the esphome fw is a good start but I’d go a step further and suggest doing it over serial. On the model you have you can do this with the addition of a 2 DuPont jumper wires. (
).
Scroll down here for the hard ware configuration:
The reason to do this over serial is to clear out any “rusty old bits” in the esp32 flash storage, as it does a full erase first vs the OTA where it just copies over the existing fw flash.
For the efr32 flash - I definitely need to update docs. But now recommend this tool:
It can put the module into bootloader and update the fw. The file you would need is the .gbl file ncp folders.
make sure zha is disabled.
the command to flash would be like this, replacing the ip address and path to the fw file.
universal-silabs-flasher --device socket://192.168.1.136:6638 --bootloader-baudrate 115200 --baudrate 115200 flash --firmware ./build/release/ncp-uart-hw.gbl
Changing the ip address and file path/name for the gbl file.
Note I’d try any address the esp32 serial flash before even looking at efr32 firmware updates as I think the esp32 connection is most likely the issue.
I’m away from home this week so don’t have access to hardware etc to do any physical troubleshooting but happy to help the best I can here, but feel free to direct message or email me at [email protected]
Many thanks for the quick and very helpful reply, it’s gratefully appreciated 
The gateway is mounted up on a wall, so for the moment I have just done an ESP OTA update and it seems to be working better 
I’ll see how it goes over the next day or two and look to taking it down to do a serial update and the other firmware.
Many thanks,
J.
Hi Tubes,
I need help please to change the IEEE address of my POE co-ordinator. It seems to work fine, but it always defaults back to the original “b4” IEEE address. Any suggestions?
How did you know it’s reverting? What does the info command from zigpy-cli show?