I wondered and have been meaning to ask. Is there anyway to force "3" in that list? I wondered if you have, as I do, a Z Wave stick running PC Controller; whether it's possible to send some command to an individual device that's hanging on to a crap route. There seems to be a lot of additional functionality in PC Controller and I wondered whether there are any shortcuts to sort things out.
There is a way to mess with routes but I have been reluctant and not knowledgeable enough in the subject to give it a go..
My understanding about #3 is it is a very expensive operation so should not be done lightly. Also I'm not sure you can get the best device to route through this way..
1 Like
What I've done a few times is simply unplug the repeating device that is in the route you are trying to change. Let the mesh do its thing and re-energize the disabled device.
2 Likes
Just for the benefit of others reading this in the future, I've actually had great results using the Aeotec repeaters.
I had devices that just couldn't or wouldn't work with the hub, despite having repeating devices (plug-in modules, switches) available to act as repeaters, spaced between the devices in question and the hub. I then added two repeaters (each about 1/3 of the way between the hub and the problem device area) and the problems resolved within a day. I've now used this approach in two buildings with great success.
I've no idea why this works for me while others have had no success, but for a future reader of this thread I'd argue that "Your Mileage May Vary" is a phrase to keep in mind for all things related to Z-wave. User experiences and what does and doesn't work varies widely between users, sometimes for reasons no one really fully understands.
3 Likes
My prior observations on this topic include:
Also, the dedicated repeater has more freedom regarding antenna design, as they do not have to position it around relays, TRIACS, or other devices. No guarantees that anyone is doing good RF design or firmware implementation, though.
1 Like
By "Dark Magic" you mean not at all magical....right?
1 Like
Interesting, my experience does not follow the location relationship.
In my initial test of the Aeotec Extender 7 (firmware UNK)
I had some more distant devices pairing with a very old 300 series Leviton plug in dimmer. It was sitting on the floor under a table.
I unplugged the Leviton and replaced it with the Aeotec Extender 7.
Went back and checked the "more distant devices" and they paired with a Thermostat much further away than the Aeotec.
Hmmm.... Then I decided the distant devices might have paired before I plugged in the Aeotec.
OK I removed the Thermostat. The distant devices jumped to a dimmer (again further away than the Aeotec). This dimmer is in a metal wall box.
I tried a few other combinations and finally gave up. At least the dimmer was a Plus device.
I sent the Aeotec back to Amazon.
I would be nice if SiLabs documented route behavior. From what I've seen and read, it seems that route choice may add weight based on how long a route has existed (e.g. how long it's seen a neighbor). That makes some sense, better to have a route via a device which is consistently available than to bounce around between ones which come and go. That may also explain why people sometimes see it take weeks to settle down after a change.
3 Likes
Videos 8 and 9 touch on it a bit. Still not explicit full documentation of the base algorithms, weighting, tuning though.
1 Like
It is one of those "all else being equal" things, while all else is never actually equal. Like you, I have an old Evolve LDM-15 plug-in lamp dimmer that is one of my most prolific routers. It is also huge, so it wouldn't surprise me if it had a proper dipole antenna in the thing.
I am not sure how much application-specific RF design goes on with smart home products. Probably minimal. So, your approach of returning the ones that don't work for you makes perfect sense.
1 Like
“Any sufficiently advanced troublesome technology is indistinguishable from magic.”
A.C. Clarke - edited for relevance and clarity 
9 Likes
I think my most quirky route situation is a device which has two repeaters on the path from the hub to it. However, another device uses this as its only hop to reach the hub. Clearly one of these two situations is nonsense, and perhaps both.
My actual "repeater" device is, ironically, one of the few devices in the house that didn't decide to go direct to the hub. It has two devices on the way to it, and nothing repeats through it.
That is consistent with my observations. I have 6 devices routing through a Leviton dimmer that was one of my earliest devices installed. This is in spite of the fact that I have numerous Aeotec and Ring range extenders scattered around the house that were added after the mesh was established.
Or, perhaps you just have a unique definition of "crappy" that is different than the z-wave spec.
If a route is "working", why change to something else that might fail? 
Alternatively, if a device is experience many route changes searching for a good/stable connection, a new repeater at the right location will sometimes be picked up pretty quickly.
1 Like
There's one case where a repeater is special, it will support smart locks and beaming. Not all z-wave plus devices support beaming.
1 Like
I'm about 80% sure they're v6, and no idea as to firmware. I don't see either listed when I pull the device up on the hub from my iPad, which is what I'm currently using. Sorry.
I have been seeing this with the gen 7 not 6 so I guess that makes sense!
1 Like
The older I get, the more I wish companies would just land on something that works reliably, and then stay there. But I'm a curmudgeon, so...
3 Likes
FYI I've been comparing nightly changes to my mesh (which had its last device installed 2 days ago). Here is what I observed. Note I have not used "repair" on any device.
-
If you have a device that is not used in any rules, nothing changes in the mesh for that device. Add one rule (like turn it off daily) and it becomes a device that the mesh actively changes.
-
The mesh tries hard to eliminate hops. I have one device that had 1 hop and was switched to direct despite a big drop in RSSI. Interestingly, I've seen this device toggle back to using hops like the algorithm is trying to find the best situation.
-
I've seen RSSI change for some devices like +/- 5DB from day to day. I don't know if this is mesh behavior or just random RF noise interference.
