Haven't seen a concise record of the costs and benefits of slapping a 700 series device in the middle of a plus and non-plus mesh with a C7. I have seen some say that the whole system is bottlenecked by the slowest device, which would imply that to reap the benefits of 700 one would have to replace every single device with a 700 series device.
I am presuming that's not quite the case. Can a gen7 end device benefit from the better battery life if it has to route through a gen5 repeater? I know smart start and s2 are unaffected.
Will a gen7 end device choose the fastest route to the hub and choose to skip slower repeaters if possible? This would be the best scenario I think, but what determines it and can it be predicted? This is all good to know if one wants to reap any measurable benefit from the 700 series. Determining how to build the most efficient zwave mesh is a whole nother ballgame when there are 3 different versions to keep track of.
It is good to note that smart products still have some decent resale value and one can upgrade for a relatively low cost in exchange for a little legwork.
True - but only in the routing path. If you have an old non-plus device and something is routing through it, yes all of the messages can only go the 40kb max the non-plus device supports (thus potentially slowing down any 500/700 devices routing through it).
If you have 500/700 series devices, and they don't happen to route through a non-plus deivce then they can still go max speed of 100kb, even though there are other non-plus devices on the mesh.
Having a non-plus device on the mesh DOES NOT automatically slow down your entire mesh.
That said, I still highly recommend getting rid of 100% of non-plus devices. They are old, usually quirky (no auto status reporting, etc), and simply are not worth the headache.
Also keep in mind most non-plus devices do not support Network Wide Inclusion, so another way non-plus devices can give you a headache (during pairing of other devices on your mesh).
And most non-plus devices do not support beam forming, so can cause issues if FLIRS devices like locks and some thermostats try to route through them.
In terms of 500 vs 700 - I would not bother proactively replacing 500 series devices with 700 unless:
I needed the extra range 700 series devices can provide for a specific application
I had battery powered devices that I wanted longer battery life on. 700 series battery powered devices will almost always have much longer battery life than a similar 500 series device.
All that said, a mesh full of 700 series devices MAY work better than a mesh full of 500 series devices, only because the longer range of the 700 series would likely mean less routing/fewer routing hops between devices and the hub. Less routing = less mesh bandwidth/radio airtime used/less chance of a packet error/less change of a device in the middle doing something "odd".
Hey thanks. Is there a way to see which devices are non-plus without tearing them out of the wall? I know I have some non-plus switches and outlets, just not sure which they are.
So if I'm understanding correctly, the best thing would be to upgrade routers first - then end devices, unless I can be sure the end device will use a 700 router or go straight to the hub.
And 700 end devices can't take advantage of the better battery life without routing thru other 700 devices (or direct to hub)?
Howw does that work exactly?
You can always tell by looking at the inClusters on the device detail page (or if you have a C-7 hub, on the settings -> zwave details page).
If the 1st cluster is 0x5E it is plus. If it has no 0x5E it is non-plus.
700 battery powered devices will/should have longer battery life no matter what they route through. It is really a matter that the new silicon/radio used in the devices uses less power than 500 did.
Got it. I have more non-plus devices than I thought. All my First Alert smoke detectors are non-plus. Older end-devices shouldn't slow down a mesh though right? Only routers?
That said, the non-plus end devices will slow down what they are repeating through as those routers have to drop down to 9.6/40kb for any messages they are routing from that device.
Not a huge overhead doing that though, and it happens with any 500/700 series device that picks 9.6/40kb as its message rate too (out of my 72 500/700 series devices 13 are running 40kb anyway, and 1 is even 9.6 - a device on the very, very edge of my mesh outside).
But how much does a 9.6 vs a 100 kb really pan out. We talking about very tiny messages, at least that show up in logs. Or are there large numbers of hidden background comms with larger payloads?
All depends on the device and use. If it is power reporting it could be a ton of messages. If a non-plus router has a lot of devices routing through it, it could be a ton of messages in the end. If there is interference causing high retry rates, it could be important. If there are S0 devices involved it could be important. Etc.
That is why it is tough. To your point, how impactful it is "depends". Could be ok for one person, and not for another.
There is no one "right" answer for all meshes. Just factors to consider.
