Best Practices for Wireless Hub Placement

[Dean Roddey]

Z-Wave split off from Zigbee way back when it was just a specification being worked on/argued about by the various participants. They figured, well, we'll just go out now with something that works and get a jump on everyone. So they broke off and created something that worked, but that was very primitive. But it was cheap and it was better than X-10 which was sort of the existing standard at the time.
 
But that means that they have since then gone through a long and painful process of globbing more and more stuff onto a standard that was never designed to deal with it. At this point it's honestly pretty woeful technically.
 
Zigbee went forward through a much longer standardization process. I'm sure it was still far from perfect since no one ever gets anything that complex right the first time. But clearly they started at a higher technical baseline. And it's done very well. Outside of the home automation world it massively outsells Z-Wave. The reason it has it because it allows for proprietary implementations because of its ability to support multiple messaging schemes (profiles.) But that has cost it in the home automation world and I think maybe they didn't take that market as seriously perhaps (and reasonably so when it was still very small and mostly very high end.)
 
But the new V3 scheme makes it pretty clear that they have learned that lesson. They can still support proprietary schemes, but they now have a full featured standard profile that everyone can use who wants to make non-proprietary, inter-operable bits. The question is whether there will ever be any business impetus to do that on a wider scale.

 

[upstatemike]

If Hue jumps on board V3 and that opens the door for them to support dimmer switches and plug-in modules that they aren't interested in making themselves it could turn into a full ecosystem to rival Z-Wave.
 
I am kind of surprised that Z-Wave offerings for color bulbs have been so weak and that they pretty much let Hue own that space.

 

[Dean Roddey]

"Z-Wave" isn't a company or organization, so I guess it's sort of wrong to assign motives or failures or whatever to something called Z-Wave. Zen-Sys controls the spec but they don't make any end products themselves. So it just ends up up to individual companies that make Z-Wave based products as to what they want to sell or not sell.
 
I guess that's partly an advantage and partly a disadvantage for Z-Wave. It contributes to lack of 'system-ness', whereas something like Hue is driven by a single company that needs to make things happen, not to wait for someone else to do it. And which is concerned about the coherence (hopefully) of the product and surrounding ecosystem, whereas a bunch of individual companies making Z-Wave devices doesn't really have that sort of focus or coherence.

 

[ano]

But the new V3 scheme makes it pretty clear that they have learned that lesson. They can still support proprietary schemes, but they now have a full featured standard profile that everyone can use who wants to make non-proprietary, inter-operable bits. The question is whether there will ever be any business impetus to do that on a wider scale.

 
I don't think its really fair to say Zigbee "learned that lesson."  Creating the standard that Zigbee has become could not and would not have ever happened had it not been for the many baby steps that the standard evolved through.  Zigbee meets the needs on MANY MANY wildly different industries. Forget home automation for a moment. Zigbee allows your electric meter, water meter and gas meter to be read remotely, allows the customer to monitor their usage, and allows the utility to remotely control devices in your home, all with ULTRA security.  The Zigbee allows large industrial customers to monitor their factory floors. And Zigbee allows Vegas casinos to monitor the mini bars in every room in addition to heating, door security, etc.
 
These are three Zigbee applications in three industries that had vastly different needs and with the different profiles, this is possible.  The Zigbee folks always knew they would need to combine these, but you just can't get from point A to D in one step.  I would say the Zigbee folks had incredible insight on the design of Zigbee.  If you follow Bluetooth or WiFi, you know it wasn't a smooth evolution and they DID just slap features on after-the-fact.  Zigbee REALLY was designed better from the start. Yes it took 3 versions to bring everything together. What are we on now, Wi-Fi version 6? 
 
So now lets look at Home Automation. Short-range wireless for Home Automation is like this little fly sitting on an elephant of other applications. Z-Way, on the other-hand is really only in HA applications, and does it pretty well. Granted.  Dean, you make a great point when you say "Z-Wave isn't a company" and that;s even more true with Zigbee.  When companies introduce new home automation products, they see what technologies are out there, and they pick the one that meets their needs the best. Philips when with Zigbee Light Link because it was out there, and because, at the time, it would have taken much more work to get Z-Wave o do what they wanted the way they wanted it.
 
When Zigbee V3 finally really appears, it will offer many advantages to the Home Automation folks that are pretty compelling. Like access to the millions of industrial Zigbee sensors out there. Also, imagine with a simple Zigbee hub, you could monitor your electrical usage, even communicate with your utility.  YES the utility has to allow this, but if they do, you can do it also. 
 
With or without Home Automation, Zigbee isn't going anywhere. It DEFINITELY has been a long slow process, but it looks like it might be worth it.

 

[Dean Roddey]

Oh, I agree. I wasn't implying that being able to support proprietary applications was a bad thing. It's been hugely successful for them. Just that they took quite a while to get around to really addressing the other side of it. The lesson was more, avoid fragmentation in the general purpose scenario. If you read any conversations amongst the semi- to non-technical in automation related fora, as soon as Zigbee comes up, the discussion turns to the fact that it's 'non-standard'. That's not true and I try to point that out, but the fragmentation of even the 'non-proprietary' side of their story has gotten them a lot of negativity out there.
 
I definitely hope that changes. One huge problem that comes with the 'not a company' thing is that there is no single, widely supported 'gateway' for automation systems to use. It's an enormous problem with Z-Wave. What should be a difficult but not crazy undertaking becomes a soul destroying death march because there's no standard simplified access system for folks like us to use.
 
There are third party things, but they all come with a lot of strings. Zigbee suffers the same thing. Making everyone who wants to access their respective networks deal with all of the complexities of the raw network is just insanity and I cannot believe that they don't see that and haven't worked hard to rectify that problem.

 

[pete_c]

Personally here never have not yet put any sure bet dependences on wireless automation and or use of the cloud for any automation I care about.
 
That said here base light switch automation is UPB powerline. Light switches work 100% of the time.
 
I do tinker with Zigbee and ZWave and WiFi and have done so since introduction.
 
I've mentioned before here that the OmniPro 2 panel base does the heartbeat of the home and it is talking X10, UPB, ZWave and Zigbee today and wired sensors
 
In addition and tinkering today with the OmniPro 2 talking MQTT and to the Samsung Smartthings and to the Amazon Echo devices. 
 
Note tinkering here with no dependences should the two Internet ISP connects debend.
 
I do like the firmware modifications of new WiFi devices (cloud connected) such that I do not have to connect these devices to the cloud to use them.
 
I do now have 5 (6) software automation servers running full time and mostly use these to experiment with.  
 
I like that I have turned my Amazon Echo's in to two way media devices which can be used on demand but not necessarily dependant automation devices. 
 
Relating to the OP of wireless hub placement here don't have any automation dependences so don't care and there will never be a time to create a dedicated infrastructure change in this house or any house to accomodate wireless anything.
 
That said do have a ZWave, Zigbee, Wifi transceivers in the attic (unseen), main floor (unseen) and basement (unseen) of the home. 

BTW here utilize tabletop touchscreen first sold in 1999 and meant for use with the cloud. Built in to these tablet touchscreens were Zigbee and DECT for automation and telephony. Around the same time and in a search for touchscreens you could purchase in wall or tabletop touchscreens from China and all of these had Zigbee chips for automation.

 

[upstatemike]

Another argument for a centralized radio hub closet just occurred to me and that is hop count limits. Z-Wave operates in the 900 MHz band so is likely to penetrate farther and experience less interference than Zigbee running at 2.4 GHz but Z-Wave is limited to only 4 hops while Zigbee can support a dozen or more. I think Insteon is also limited to 3 or 4 hops including powerline repeats. 
 
I wonder how often hop count ends up impacting the stability of a mesh network in large home installations?
 
Lutron and Elk need to talk directly to a primary or secondary hub so hop count is not an issue for them but still believe they need to be central to reduce the number of secondaries required.

 

[LarrylLix]

Another argument for a centralized radio hub closet just occurred to me and that is hop count limits. Z-Wave operates in the 900 MHz band so is likely to penetrate farther and experience less interference than Zigbee running at 2.4 GHz but Z-Wave is limited to only 4 hops while Zigbee can support a dozen or more. I think Insteon is also limited to 3 or 4 hops including powerline repeats. 
 
I wonder how often hop count ends up impacting the stability of a mesh network in large home installations?
 
Lutron and Elk need to talk directly to a primary or secondary hub so hop count is not an issue for them but still believe they need to be central to reduce the number of secondaries required.

If you investigate the Insteon Max Hop specified inside the devices, you soon discover increasing the number of hops increases the time taken for signals  exponentially and becomes ridiculous for HA.
 
Powerline specs for Insteon are reported to by up to one mile in range. Not very likely anybody would have that clean test environment, but I have seen outbuildings over 500 feet away for the last Insteon device  reported successful  a few times now. One used an unloaded 120vac cable with just an unloaded Insteon plug-in device for a repeater at the far end. This is not RF though.