the undercabinet light constant report 0 or 7 in the table (they work fine 99% of the time) i was trying to help them have better path to the hub
the one marced "router" are the new one I bought and are rated for 50 devices
-7 under cabinet
i dont know there where recommend here they worked , wife liked, bought more. lol
yes the light group in the hall is STILL having issues even with the new router there (i just move one repeater maybe that will help)
the "routers" above are rated for 50 devices, that is why i put one there to unload the hub - still issues
basically sometimes they work fine all 4 groups come on- sometimes 3 groups come on, sometimes only one group...... its annoying lol i cannot figure out why.
i does take 2-3 seconds for all the groups to come on sometimes also
basically any moting in the hall turn on 13 bulbs via group motion and light groups (4 groups of 3 plus one stand alone)
I have tried many things and cannot get it work right
Even though the undercabinet lights might be showing bad (0 or 7) cost figures as neighbors, they might actually be using an intermediate repeater to reach the hub, in which case their link status as a neighbor is irrelevant (and those repeaters could be moved elsewhere) .
If you feel like experimenting (and at this point, without other tools it's probably your only recourse) you can use one additional data point you have available: the measure of RSSI (signal strength) from a Zigbee device to the hub (if it's a child or neighbor)-- or from the device via the last repeater in the path it is taking to the hub-- by way of the 'Settings-> Zigbee Details -> Zigbee Logging' page.
When you operate one of those undercabinet lights you should see an entry on the Zigbee Logging page, note the number that shows up. It will be a negative number, commonly in the range of -30 to -70. More negative than that indicates a progressively weaker signal. By temporarily removing an intermediate repeater (that you suspect is helping improve the RSSI) you can see how (or if) the RSSI is afffected. If it doesn't change when a repeater is removed, it's likely that the inbound message is using a different repeater, or not using a repeater at all.
You can also try looking at what RSSI numbers show up when you operate your other problematic devices. Again, since this figure only pertains to the last hop, so what you see is only for the final leg to the hub (if its a multi-hop path).
What specifically is this device? Any reports of it being used successfully elsewhere on the forum? There might be some compatibility issue at play here....
The fact that the hub has so many child devices means that those devices don't see any better connections available to the hub. So either there aren't any available (unlikely if you have a surplus of repeaters) or they keep establishing and dropping (due to some incompatibillity). Again, since the hub isn't known for having an especially strong Zigbee signal, its uncommon for it to be preferred as a parent device if there are other adequate repeaters nearby.
my hub doesnt many child devices anymore - that stopped with advice in this thread at most i will see it have 4 but usually 1 or 2
i found this on the forum
uggh this is bad i know a bit about RSSI workign with 4/5g and WIFI
something is wrong
yellow = repeaters/routers
green = Hall light that are not work correctly
the hall router shows a very good signal -54
the Pool route show a good signal -65 (for 2.4ghz wifi that nearly VOIP quality)
others in that range
light in hal show very good signal........... ugh
**also bug? ** activating lights from the GROUP does not report a line to this zigbee log - is that expected?
Ok, good info. Don't know exactly what to tell you other than if you're seeing a correlation in reliability with those devices showing RSSI's in -80's -90's, you need to improve them (possibly by moving a repeater nearby) until their RSSI shows better numbers.
Of course moving a repeater (when you don't know what it is repeating for) may destablilize things for a while. Could be tricky to get all the pieces working without a lot of temporary disruptions; try to keep track of what improves/gets worse with each change you make.
here the HALL lights are the issue and they all have good RSS i just messed with it a bit here is what i figure out
I used my phone to turn off the lights set temps turn off
I DO NOT know if i just have it setup wrong, or this method is not supported , need help
LIGHT groups
HALL pool group (3 lights)
HALL master groups (3 lights)
HALL room front ( 3 lights)
HALL Closet it light (stand alone not a group)
HALL Entryway group ( 3 lights)
HALL ALL LIGHTS GROUP - (all light in the above in one big group, added via individual lights and not groups above though i did try the groups)
IF i turn on and off ONE group, change color etc, then it works fine, I i trying doing the all group then it fails and/or take SECONDS for the all the light groups to come on. (for example 1 or 2 groups will instantly turn then the other 2 groups LAG badly. (also tried having all the small groups in the rule above separately and it also does the same thing)
i'm currently messing with different setups to figure out how to do this, splitting the up the groups might be my only way to get this to work though i would rather not do that. because this i need 5-10 rules for where you enter the hall and what direction you might walk lol
If things work fine individually, effort spent improving RSSI figures by reconfiguring the mesh might eliminate APS-layer retries that you evidently don't notice in those situations. Group messaging multicasts may expose issues you don't see when operating individual devices (or small numbers of them).
Certainly it would help to pick out the high fliers and improve the mesh to eliminate them. Your screenshot shows a half dozen weaker than -85 which is about as they get; apparently the one at -90 is working for you even though its signal is almost completely buried in noise. That's more a testament to Zigbee's DSSS modulation than the quality of your mesh, lol.
I wouldn't expect that.. maybe someone using group messaging could comment (I don't use Rule Machine and don't use groups).
One other point: Since 'kitchen undercab peninsula' isn't a child device, that means the -90 RSSI figure it reports is the signal strength of the repeater it is using to reach the hub... so every device using that repeater will also be hampered by that weak signal path (same goes for the -88's and -86's etc.... they could in fact all be using the very same repeater, so improving that weak repeater may improve them all).
is there a way to figure out what is using what? that way i can track it down ?
i notice the one kitchen one FINALLY jumped on one of my new route but still reports bad RSSI ( i don't know what heck is in that wall if that is the case)
but other device like the kitchen motion report -70 and they are in 5 feet apart
is it possible the kitchen under cabinet are interfering with each other very bad?
There are tools (hardware and software required) that will scan the child and neighbor tables of all the Zigbee routers in your network and from that info generate a network map that shows the mesh topology.
The XBee device (you need to use the Zigbee version) and XCTU are the hardware and software components that are discussed here: Everything Xbee
I use a Digi Xstick2 (not sold anymore, it's a precursor to the XBee) and XCTU (which is free). The resulting map can take a long time (multiple scans a taking a few minutes each) before your devices all show up, and though you can customize the layout to some extent using it is very fiddly as the graph changes with each device discovered. Also you need to refer to the shortID's (rather than symbolic names) to make sense of it all. Still, it's probably the best way to get the information you're looking for.
One way to find out is to turn them off... wait a while and see what happens. Its always easier to do the opposite (gradually build out a network while observing how it performs--albeit with the limited diagnostic info you have available-- then add new stuff.... and repeat) but at this point experimenting might be your only option if you want to get to the bottom of this.
a "like" on one of your posts on this topic is not enough to give the appreciation due for the time you spend helping folk get to the bottom of this mesh stuff with your understanding of it.
Aw shucks... thanks. Half the time I post this stuff because when/if my mesh eventually breaks I can find out if what I've been recommending actually works, lol.
I’m jumping in a bit late here but I see you have some Sylvania lights. These make horrible repeaters for non-bulb devices. They may work fine for a year or may get confused frequently. I don’t know of any rhyme not reason to the patterns. I’ve seen both good and bad behavior. It did start to get worse as I added more motion sensors.
For me the solution was to create two meshes. I have a C-5 running my lights and some repeaters. Then I use a C-7 to run all the remaining Zigbee devices. Things have been very stable since splitting out the lights. I get Sengled lights for any new purchase but I still have a bunch of the legacy Sylvanias.
I would suggest taking a good look at the Sylvania lights.
unfortunately i haven't had any time over past couple weeks to dig into, the Hall light are still not working correctly, can walk down the hall, randomly it will work perfect then sometimes half to none work.
can literally walk back 2 minutes later and get exact opposite results. for example i can leave my master activate all motion as i walk dow nthe hall/ nothing - grab a snack and on the way back they all turn on perfectly lol
today i have a bit of time, I tried splitting the rules up from one into 4 rules and same results. i have noticed the light on the motion sensor stays on longer if this i sanything like Z-wave it should just be flash
im might pull all my reapters today and do the 20 minute thing and see what happens
UPDATE:
I bought the new H8 - still had the issue ( i want it for matter plus the long range)
I then turned off my kitchen under counter lights and i think that was at least most of the issue they stiill do not work sometimes but its much better
Child Data
No information for Child 0
No information for Child 1
No information for Child 2
child:[null, 9932, type:UNKNOWN]
child:[null, DB79, type:EMBER_RF4CE_CONTROLLER]
child:[null, F440, type:UNKNOWN]
i would rather not do that. because this i need 5-10 rules for where you enter the hall and what direction you might walk lol
