I have a detached garage where I have 3 Ecolink Zwave Plus sensors as well as an Inovelli switch.
The garage is only 8 feet away from the nearest Inovelli switch in the house which is right next to a sliding glass door.
The switch in the garage is terribly unreliable. Sometimes it works great, other times it will be a delayed response. Other times it will not respond remotely, but when physically toggling, the on/off will show correctly in the Smartthings app.
The battery powered Ecolinks work perfectly and they are further in the garage than the switch is. They work regardless of whether the switch is there or not.
I have gone through 3 switches to eliminate the possibility of them being defective. All switches will only pair if I am in the house. They will not pair in the garage. The Ecolinks paired immediately and without the switch installed and no other repeaters. I have done Zwave repairs each and everyone I tried a new switch in the garage. After I pair in the house, I move out to the garage and do a Zwave repair. Usually, that switch fails to update mesh/route. It will still “kind of” work though.
On the other hand, a GE Zigbee switch controlling pool Lights in a metal box even further away outside with no repeaters near it works flawlessly.
Anyone have any ideas on what is going on here? These Inovelli switches are supposed to have powerful repeaters built in and I have over 30 of them in the house - all work great. The garage is only 8 feet away and battery operated Zwave and Zigbee devices are performing flawlessly.
If you have battery operated zwave devices in the garage I would definitely keep a hardwired zwave device (that works) to act as a repeater. At the very least it should help make the batteries last longer and supposedly make the communication more reliable.
Are you putting the zwave switch in a metal box? If so, is the front of the switch facing away from the house? Often the antenna is right behind the paddle.
If you are pairing in the house it is very possible that the route it settles on will not be ideal once you relocate the switch. I would definitely do a network repair (even though ST keeps saying it causes more issues that good) to ensure the path to the hub is optimal. Of course this will only work if you have a good RF path to a zwave repeater.
Can you add a zwave hard wired device anywhere else that is also close to the detached garage? Maybe the RF path to the switch next to your glass door is actually not optimal and another one might work better. To have more placement optioons you could also consider a Zwave outlet. For testing you could use a Zwave plug in switch.
Signals from 900MHz (zwave) and 2.4GHz (zigbee) devices behave quite differently when it comes to their ability to get from point A to B through materials. Generally, the higher the frequency the worse they are at dealing with obstacles.
You might have significant interference close to the zweave frequency… Baby monitor maybe? Maybe a bit unlikely but a neighbor might be the source of the interference (if close enough) so don’t just focus on your devices.
Can you try to pair the switch in the garage outside, or far away from obstacle before putting it in the electrical box? The idea is for it to find the closest node even though it is weak instead of it trying to use whatever node was chosen when pairing it inside your house.
I’ve had similar issues where control was poor but status readings worked… that typically went way with an exclude / include. I was never certain why the communication worked better one way than the other.
You’ve gotten very good advice so far. It’s not likely to have anything to do with battery versus hardwire, but sometimes it’s the specific location, down to within three or 4 inches, that’s a problem.
So far you’ve tried swapping out one wired zwave switch for another, I understand that. As a field tech, my next step would be to take the switch out of the wall all together and move one of the battery operated devices which is working in the garage to the location where the switch is. See if it fails in that location. If it does, that tells you there’s a micro location issue, something right in that wall or on that wall or as was suggested maybe a frequency interference from something like a baby monitor or old cordless phone that just happens to be hitting that spot.
There’s no harm in trying a zigbee switch or WiFi Switch there instead of a Z wave switch if you think the battery operated devices aren’t using that switch as a repeater anyway.
Sometimes it just takes a lot of trial and error to find out what the issue is in a particular spot in a particular building. If you have network tools, it’s way easier, but smartthings doesn’t give us any.
As it happens, if you’re one of those people who will just go crazy until you figure out exactly what’s happening, there is an installer tool set that’s on sale right now for $189, but that’s a lot of money to spend just to satisfy your curiosity.
Without that kind of tool, you just have to keep trying different combinations of devices and locations until you find something that works for you.
If I were in your situation and I felt that I really didn’t need a Z wave repeater in that location I would probably shift to a Wi-Fi switch just because they have great range and they are likely to work the first time this particular kind of set up. But obviously that’s up to you.
Thank you all for the suggestions. The switch in the garage is facing the opposite direction of the house. The nearest switch in the house is facing the opposite direction of the garage. Basically, the 2 switches are back to back and practically lined up with 8 feet of separation (and 2 exterior walls).
I was hoping to have a switch that has scene capability which is why I am using Inovelli everywhere else.
I think I’ve told the story before, but we had a real life case where we kept losing communication with the controller late at night but only once in a while. It was driving us crazy. It would last about 20 minutes and then resolve itself. We did all kinds of frequency sweeps and checking cabling and rebuilds and nothing mattered. Still apparently randomly every once in a while we would have these outages.
So one of our guys decided he was just going to sit in that room every night for three weeks to see what was going on. ( cameras weren’t an option at that time) and he found it.
Turned out one of the programmers on the client staff had this all night Big Data Analytical program that had to be run every so often. And he liked to babysit it, because if it ran into any problems, he’d have to tweak things and restart it in order to get it done in time. So he would volunteer to work the graveyard shift in the server room on those nights.
Everybody else who worked there was right-handed. This guy was left-handed. When he stood up from the desk he would push the rolling chair back in the opposite direction that everybody else did. And it tended to roll to a stop right in front of an open server cabinet that had the hub that kept going out.
He’d take a break, go get some coffee, get back in about 15 minutes, get back in the chair and re-position in front of the desk – – and the hub would reconnect to the network! Crazy but true. And if we hadn’t had a field tech who wanted to just sit there and watch what happened, we probably never would’ve found it.
What is the ideal direction to have a switch facing? If I were to install a Zwave receptical, which direction should the front of the switch be facing inside the house if I’m trying to send the signal out to the garage?
Should the front face towards the garage or the back face toward the garage? Is it the same for switches?
It’s a 360° signal once you get about a foot out from the radio, so it’s less about the orientation then it is the obstacles that the device may encounter in those first 12 inches.
It is up to each individual device manufacturer how they lay out the internals of each model, there’s no one way to do it. Some will have the antenna pointing out through the front of the switch, some towards the bottom of the switch, most don’t have it pointing Towards the back of the switch since that’s a guaranteed obstacle.
I’ve seen signal interference caused by a mirror on the bedroom dresser which just happened to be very close to the wall switch.
I figured I could play with the placement of the repeater because I can’t change the location of the switch. Since the 3 Ecolinks are working perfectly in the garage, I figure maybe a repeater in a different spot than the switch could help the switch bounce the signal.
Without a layout diagram it is hard to give suggestions but one thing I would try is to get an extension cord, connect a plug in zwave repeater, place it between the garage and the house, closer to the opposite side of the wall where the garage switch is. The idea is to test whether the issue is that something is making it hard for the garage switch to receive the signals from the hub or for its signals to make it out to the hub. A repeater placed outside the garage, close enough to the garage switch might be able to overcome the obstacles and relay the messages to the house mesh network. Just something I’d check…
Mirrors are plated with silver and maybe other metals. That is what makes them shiny so I am guessing they are not too RF friendly especially if all the walls are covered. Also, is the garage structure traditional drywall, wood frame and possibly brick? The more material or the more metal there is the worse it gets.
My observation on the antenna was related to the fact that a few zwave switches i took apart had the antenna laying against the aluminum yoke of the switch. While the antenna is omnidirectional, its performance must be better in the direction of less obstacles. The yoke is metal and thus must have some effect, plus as @JDRoberts pointed out, the back side of the switch is bound to have the most ‘immediate’ obstacles… wires, box, wall, etc. In your case, you have two nodes pointing their best lobe away so that likely has a negative impact.
@JDRoberts can likely correct me if I am wrong, but what I am thinking is, I believe, represented by the diagram below. Best performance is when the 2 green lobes are facing each other. Obviously communication can still work in any of the directions but the distance/quality will likely be diminished accordingly. The diagram I found is most likely representing a directional antenna but it served to show what I was thinking.
If the “Ecolinks” you reference are hardwired zwave devices, and they work in the same exact spot, then @JDRoberts might be onto something when pointing out the antenna location/orientation varies across brands/models.
I have the toolbox that @JDRoberts referenced. It is a fun tool to have if you want to see the inner workings of the mesh network. I had a device outside in an electrical panel driving me crazy, much like in your case, so I went for it. A few shots (not all inclusive) of what you can see with it: