Much has been made on these forums about how to finesse zigbee-wifi coexistence. And much has been made also about the importance of a strong mesh (both zwave and zigbee). I'm not looking to re-tread either of these discussions.
Rather, my question for those folks who have multi-hub environments is whether there is any potential problem associated with inter-hub zigbee coexistence? Specifically, if two hubs are in adjacent areas of a home, is it important that they be on different zigbee channels, or do the hub radios sort out for themselves who owns what traffic?
And, perhaps more importantly, if two hubs with zigbee radio traffic get too close to each other, whether they are on separate channels or not, can it cause a problem more serious than the modest waste of having two hubs serving an area that a single hub might be able to cover with proper repeaters?
I have three zigbee hubs in a small house (two Hubitats and one Almond router). The Hubitats are on channel 15 & 20. Not sure about the Almond. But there are no issues I am aware of. Nothing has dropped off, and sensor trigger times are rapid.
You'd see it as slow responses on all your Zigbee gear that are on the same channel.
All of the protocols we Home Automation enthusiasts use every day are resilient. They expect collisions. They have to... you have a motion sensor that suddenly detects motion... it's going to try and send that to the Hub.. at the same time, the hub wants to tell some other device to turn off because the timer hit zero. They will collide.. they solve that by RANDOMLY backing off before resending and then Listening to the airwaves for emptiness. In the middle of backing off, some other device jumps on. It's all random so that's baked into the protocol.. to overcome adversity and be resilient.
All that backing off and retrying would be what you and I would call: DELAY.
I am not an RF engineer, and most people here aren't either, so take my advice with a grain of salt. However, my understanding of the Zigbee (and actually Wi-Fi, too) spec is that it performs clear channel assessment before sending a message, meaning that interference should be minimal. (As stated above, the re-tries this involve may make things slower. I do not know the practical, typical outcomes of this.) You also definitely won't get messages on a different hub that were intended for another--the message sent also includes the controller's (or device's) ID. So, that's the theory. From a practical perspective, many people have multiple Zigbee (and/or 2.4 GHz Wi-Fi, Bluetooth, and whatnot) networks in the same home--or if not they are in range with one or more of the same from neighbors--without problem. This, too, suggests that you are unlikely to run into problems.
That being said, I wouldn't make things harder than they need to be. If you have two separate Zigbee networks, I'd put them on two separate channels. Keep your hub at least a few feet away from your Wi-Fi AP (I think this is Hubitat's recommendation, if not likely a good idea regardless). If you're able to, choose Zigbee and Wi-Fi channels with minimal overlap. You often see this chart floated around with Wi-Fi vs. Zigbee channels and their frequency ranges, and because of it, people often suggest 15, 20, or 25 (some devices don't like the higher channels, though) as good Zigbee options, but I can't help but feel that everyone is misintepreting that chart by assuming surrounding Wi-Fi networks are on Wi-Fi channels 1, 6, or 11 (which are indeed popular because they overlap minimally with each other, but that's by no means a guarantee that your neighbors--or you--did this). But back to Zigbee, if that is all you're concerned about, Zigbee channels do not overlap each other, so even adjacent ones there should be fine (or even the same channel, but again, not something I'd do if you can avoid it).
Devices send so few and such small packets I can't imagine two or even three "Z" networks sharing a channel hindering each other. It's a different model than Ethernet, no streaming going on, no downloads of gig-sized files.
I'm an Engineer, but not the right kind to be asserting one way or another. This is just my opinion.
