I did turn off Device Health this afternoon, before all the last logs I’ve sent. The DTH IS ver 0.2.170515, dated 5/15.
This is in reference to an edge case which only affects some people, but by any chance, do you have any zigbee light bulbs connected directly to the SmartThings hub, and if so which brand?
Yes. GE Links (3), Sengled, Osram GardenSpots (2).
One of the Links has been flakey for the 1+ years I’ve had it; the other two have been solid.
I’ve heard others speak of the GardenSpots causing interference, but I’ve not seen it (until now?). Would gladly rip them all out if that would help the fans.
(P.S. and, not bulbs, but some Iris 3210 Zigbee ZHA 1.2 plugs)
Something else to try and watch the logs when you do this… is get the remote and send a command. I am wondering if it has the similar issue at the time the device doesn’t respond to ST. Just trying to collect as many data points to see if we can’t isolate the issue.
The plugs aren’t a problem, and oddly enough, this is one of the few areas where the GE links are better than other zigbee bulbs. They are not likely to be the problem.
but some zigbee bulbs, when joined directly to the SmartThings hub, are identified as repeaters but occasionally drop messages, making other devices look like they are having problems. I don’t know about the sengled one way the other.
Osram/Sylvania definitely has this problem – – there’s a note about it in the official support article.
Note: SYLVANIA bulbs have a known firmware issue that can occasionally cause them to fail as ZigBee repeaters. This can potentially cause other ZigBee devices to fail to update and/or respond.
It’s unusual for Garden-spots to be the problem, though, because since they are typically outside the house, they aren’t usually used as repeaters by other devices.
You can try killing all power (not just turning off) the zigbee bulbs and see if it makes any difference over a couple of days. You don’t have to remove them from your network or any automations. Just take them off power.
But my guess would be this doesn’t have anything to do with what you’re seeing with the fans. If you had a bunch of Osram/Sylvania bulbs scattered around the house they might be involved, but that just doesn’t sound like your issue.
I’m tired tonight, so I’m not going to going to go into details, but I assume someone has already walked you through potential Wi-Fi interference issues? Wi-Fi interference is one of those things that commonly peaks and ebbs because of changes in usage as household members come and go. For example, I’ve told the story of the zigbee device at my house that kept dropping off every afternoon around 330. I’m pretty sure one of my neighbors has boosted Wi-Fi and a kid that got home from school around then.
Here’s the FAQ:
Obviously, what we’re doing here is just differential diagnosis. Maybe it’s A. Nope, doesn’t look like that. What about B? And so on. Sometimes that’s the only way to troubleshoot these intermittent problems.
Thanks for your attention to this. I have to confess to not knowing WiFi might be conflicting. I have a router within 1’ of the ST Hub that is broadcasting on Channel 1 (also 5gHz Ch 153). Also a Wink Hub in the same cluster.
Now for the worst part. One of the GardenSpot receivers is also 1’ from the ST Hub.
So, what I’ve done: pulled power from all the GE Links (leaves the Sengled removal of which requires an expenditure of political capital that needs some work on my part), pulled power from both the GardenSpots. Reset both Fan controllers. Rebuilt the Zigbee mesh.
Now, one fan (slightly closest) is paired; one is not. I’ll let it all settle overnight and report in the morning. This community is fantastic and I really appreciate everyone’s generous contribution of time.
I think there are a few other brands of lights, and possibly other devices (though I forget the specifics – I think it related to a specific chipset), that can have odd routing problems. GE Link and OSRAM are a couple that come to mind immediately. I’ll checkup on it and report back.
I have my fan controller installed and so far so good.
Just an FYI on “lock” issues;
I am pointing this out because it seems my ST platform has been extremely slow for updating some devices? Like this zigbee device my H801 RGBW devices seem to act like the states do not update but eventually will over 5 minutes. The state is being shown on virtual devices but it never use to update this slow.
Also I have four of these devices installed and all running the 4-speed thermostat without problems right now. It was reported earlier that changing setpoints for them caused issues but I haven’t seen that. I will continue to monitor this one to see if it’s related to the issue of lockup.
Thank you @dckirker. Presently, I have no Links or Osram devices powered (and have rebuilt the Zigbee mesh after removing them). After the Hub reset last night, it still has not found one of my fan controllers.
@JDRoberts after reading your Zigbee - WiFi channel info, I’m not sure what to do next. The extremely close router is Channel 1. Does Ch 1 sit on a possible Zigbee channel? Is it possible, with just the various devices I have, to determine what Zigbee frequencies are being used?
EDIT: Found the Zigbee channel my hub uses. It’s 19.
It looks like Wifi channel 1 and ZigBee channel 19 do not interfere. The ZigBee radio usually hunts for a channel with no interference on network creation.
Digging a little deeper, there are two other 2.4gHz routers in the neighborhood. One is 100’ away and fairly low power on Ch 6. The other is mine on Ch 11. This is in the house, but 50’ from the ST hub and 20’ from the problematic fan controllers.
From that metageek link, it seems like my chances improve if I move that WiFi to Ch 6.
EDIT: Moved mine to Ch 6 and a scan now shows nothing in range on Ch 11.
Wi-Fi 6 is sitting right on top of zigbee 19. So you want them as for away from each other as you can get. Unfortunately, Wi-Fi 11 is also sitting on top of zigbee 19. The only safe Wi-Fi channel would be one.
I know the meta geek chart can be a little confusing, but you want to place two dots as far away from each other as possible.
The trick is that the Wi-Fi channels overlap each other as well! So in the picture, there are what are called “sidelobes” on the side of each of the three colors, and the sidelobes overlap even though the square pieces at the top don’t. This isn’t a problem for Wi-Fi itself, but the side lobes are strong enough to drown out the zigbee signals in that area.
The meta-geek article (which is also linked to from the community FAQ I posted above) has a lot of great information and detail on all of this, including more about the Sidelobes. So I would read it again, this time with the understanding that the Wi-Fi channel sidelobes are overlapping each other.
On this picture, I filled in the sidelobe for Wi-Fi channel 6 with the green cross hatch.
Now let’s look at where zigbee channel 19 Falls.
You can see it’s overlapping the sidelobe areas for both WiFi six and 11. And again, sidelobe activity can be strong enough to prevent zigbee Messages from getting through.
Sideband lobes might not carry WiFi data, but they are fully capable of downing out ZigBee transmissions.
The sideband areas lower in signal strength, so they will only affect devices that are closer to them as opposed to the main area for the channel which is the rectangular part.
You can also see why if you have a choice zigbee 11 and Wi-Fi 11 are the two that are farthest apart and easy to remember.
Your house is going to be a problem because with three different Wi-Fi networks on three different zigbee channels, there will be no escaping Wi-Fi interference no matter where you are in the house.
And remember that the activity on the Wi-Fi also matters, not just the existence. So as people come and go and as activity on each Wi-Fi channel goes up-and-down, you’ll get variable amounts of interference.
To be honest, in your house I’d recommend using zigbee as little as possible and sticking with Z wave. It’s just going to be challenging otherwise.
All of that said, “all home automation is local.” You may find that some zigbee devices in some locations work just great.
Instead of changing your Wi-Fi to channel 6, put it back on channel 1. That would be the best for your zigbee. But again, in an environment like you have you’re just going to have to try and see what works best, and I would run each test for several days to account for different activity patterns. Definitely make sure you have some weekend days and some non-weekend days.
Also definitely separate any Wi-Fi routers or access points from your smart things hub, preferably by about 3 m. But definitely at least 1 m. The signal for both devices is strongest closest to the hub/router, so that’s the area of maximum interference. You just want to give them a little breathing room.
Thanks again, @JDRoberts. I’ll move the secondary router to Ch 1. At this point, I’d rather have WiFi conflicts than WiFi-ZigBee conflicts. Plus, most of my WiFi is on 802.11ac, so conflicts in the 2.4gHz band are tolerable.
One practical hurdle I have is how best to reset the Fan Controller ZigBee module. It’s a problem because I don’t have ready access to be able to power up/down the failing device. Both fans are on a switched circuit, so I can do both easily, but I’m not sure how ST reconnects when it sees two devices gong into reset mode. When I reset last night, I took the easy way and did both - the result was the failing one reconnected and the working one did not. When one disconnects, or locks up, is there a way to prod ST too look for it again without doing the ZigBee reset?
Is it REQUIRED that I purchase some sort of remote for this?
I did not. And if the ZigBee connection was reliable, that would be enough. In my current state of flakey ZigBee, I frequently have no way to control on/off/speeds. So there is a definite benefit, but it is not required.
by any chance does the power for both fans wire up at your old wall switch location? That could be the way to isolate the power going to each fan separately for ease of resetting. You could use a dual switch to power off each one independently.
Well, technically, yes. But both are on one wire. So, at the switch, I only have one load wire going to the first fan then continuing to the second fan.
My next move is to disconnect the working controller so I can use the switch to reset and re-pair the nonworking.
Also on my list of “next moves” is to temporarily run an Ethernet line to the fans and relocate the ST Hub to immediately under the fans.
EDIT: I have done both of the above. Moved the Hub to be below (not yet addressing @JDRoberts point about the spinning metal blades in the way) the fans 7’ equidistant from each controller. From there, did another ZigBee reset of the Hub and reset ZigBee on the previously non-responsive fan controller. Everything came back alive.
So, I’ll leave it this way until it stops. Live Logging in the background.
EDIT 2: 6+ hours and no failures. After the initial hub placement, I moved it slightly away, but still isolated and 25’ from any WiFi router and the Wink hub. Of course the Osram GardenSpots are completely off. Despite an Ethernet cable laying in the floor through half my house, I can leave it this way long enough to see if it stays solid.
Also, I’d be curious what the Wifi channel width is set to on your router? Usually it is defaulted at 20 MHz, but one can select 40MHz for faster speed, at the expense of more RF channel occupation. You’re likely at 20MHz though.
The settings on the near router don’t explicitly give me a channel width option. I specify a “Enable 20/40 mHz coexistence” and a speed of 600mbits/s. I expect they implicitly determine the channel width. Nothing mission-critical connects to the 2.4gHz wireless, so if it helps reduce ZigBee conflicts, it wouldn’t be much sacrifice to narrow the bandwidth; although down at channel 1, I didn’t know if that makes much difference. ?
The second WiFi router, farther away and also channel 1, gives the option to choose modes of 54/145/300. Its only purpose is to provide wired switching and 802.11bg support of a Honeywell thermostat and SkyBell.
P.S. Overnight and still humming along. Pending continued success, I don’t plan on touching anything for a couple of days.