A few months ago, I replaced my C8 hub with a C8-pro. All of the routines and devices were transitioned from the old hub. After doing so, I'd have an occasional issue where one of two lights in a room wouldn't turn on after entering the room.
Once the issue started, it wouldn't start working again until I opened up the automation in webcore, simply hit "save" without making changes. This weird fix would work for about a week, then the problem would return.
I could confirm the light worked by turning it on and off before entering the room. It also would work fine if I turned in on after other light turned on.
Since this seemed to only happen when the webcore automation powered up the light, I thought it might be because too many commands were sent by zigbee at the same time. As a test, I added 500ms delay before sending power-on. So far, this seems to have resolved the issue.
Has anyone else seen this happen? Could the speed of the C8-pro be faster to the point that it's overrunning the zigbee network with commands causing network collisions?
Thanks
The difference is a slightly faster CPU and more RAM, but that shouldn't change what you are seeing.
That seems more like a ZigBee issue, or if you turned the light on/off when the hub was rebooting, then it could lose state (think it's on, but it was off). I've had that happen based on my rule configuration.
Because this could be an issue I have 'command optimisation' off in all my pistons.
I think the zigbee network can be overrun if a command is issued to turn off multiple lights simultaneously.
I have a piston that turns all lights off if a button is pushed in our bedroom when we went to bed.
Sometimes not all lights would turn off. I put a 100ms delay between each each turn off command (they were all in one turn off command) and they now all turn off when the button is pressed.
I'm not saying any of the above is your issue but........ might be worth giving it a try
1 Like
If possible, this would be a situation to use zigbee group messaging.
1 Like
It's good to know that I'm not the only one that adding delay solved a problem.
I'll have to give this a try, but right now, it's working with the 500ms delay.
