Further Xiaomi Pros and Cons

It will be hard to tell whether zigbee 3.0 or Z-Wave 700 will be the technology I need first. I'm speaking to a manufacturer next week about Z-Wave 700 delivery timelines for their product line. There will definitely be some products by some manufacturers commercially available by third quarter. So maybe I have a year before it starts to get painful.

Zigbee 3.0 is already out, but I haven't seen any devices yet that wouldn't work with backwards compatibility. So I'm not sure how much of a hurry that really is.

Well why would you want the latest version of Android or IOS? :joy::rofl:
The device would certainly still work with the previous version :stuck_out_tongue:

Having taught English in Vietnam, I would guess that the use of the word "shall" (or other terms) in the ZigBee specifications may have been incorrectly interpreted/translated by the engineers who worked on the firmware of Lumi's ZigBee-based devices.

Perhaps this is a matter of someone explaining it to the right person and then a firmware update propagated by the user's Xiaomi Gateway? I don't see that happening, sadly.

ZigBee 3.0 by design is backwards compatible. So a ZB 3.0 end device will just work with an non ZB 3.0 coordinator like the Hubitat hub.

If that was the case would be more fun.
But you are right. Shall in most languages is seen as a recommendation of what should be pursued (see what I did there with should) and not something that really needs to happen.

Fingers crossed that this is the case with the Xiaomi version.

Certainly possible, that is why they define their terms at the beginning of the document:

7 Likes

Interestingly, in Zigbee 3.,0 they changed the wording/requirement a little... Summary:

Processing Mgmt-Leave and NWK-Leave requests are now mandatory but handled differently depending on the Zigbee device type. Routers accept these commands from any node while end devices only accept the command from its direct parent. Coordinators ignore all leave requests. The Mgmt_LQI_req command is mandatory for all devices (including end devices) and provides a standard way of performing network topology discovery.

So if they did it based on language is just silly.

I would say that was more of a requirement to avoid the type of questions we are having here now.
Happy they did.

1 Like

Due diligence by the writer does not necessarily lead to due diligence by the reader.

If Lumi has made official claims that their ZigBee-based devices are ZigBee spec compliant, then there's a case for informing them that their devices are not.

But that whole process of finding out who to talk to in a huge Chinese corporation, and then waiting for a response, change, and implementation of the change? It would be easier to work out a Hubitat-native LAN to Xiaomi Gateway LAN solution, I believe.

3 Likes

Yep, agreed.

I agree that every situation and personal budget ability plays into this. As you said earlier it is great to have options as choices.

But I honestly think Xiaomi is missing out on sales at least to some degree because of this.

Ageee, but it also has to be said that Xiaomi can be budget unfriendly too for someone that already has built a decent Zigbee mesh.

$20 Xiaomi Motion Sensor + $15 Temp/Humidity Sensor vs $20 ST Multi Sensor + $25 ST Motion Sensor. Or heck, the $32 Zooz 4-1.

The ST setup is $10 more, but you don’t have to replace (time and money) plugs, switches, etc, etc. The Zooz 4-1 is few dollars cheaper than the Xiaomi setup.

IMHO it seems that at this point in time going all in with Xiaomi up front can be cost effective, but I doubt new HA users would know or think about this. By the time most realize this, they are probably invested too much to care.

Also any device that you have to resell is most likely at a loss (time and money) that has to be taken into consideration too.
A few plugin outlets might not be bad, but how many people going to rip out switches, resell plugs, etc, etc (on top of buying new Xiaomi compatible replacements) just because they cause issues with a temp/humidity sensor they just bought?

You left out needing to new buy repeating/routing devices. While it may not be much, there is still the point I made about removing and reselling devices you had already.

Using the cheapest Amazon prices for the setup you quoted + 4 IKEA outlets came out to $386 not including any extra shipping. I get that you could get them cheaper probably too.

I think you could buy a similar setup using other brands for the same price. I mean that Zooz 4-1 is cheap, Peanut plugs can be had for $15, the new ST devices are pretty cheap too.

Your pricing is way off for the xiaomi stuff, like double assuming you are quoting USD

I quoted my setup and at the time the cost of the devices were 222$ I order them from Aliexpresa.
The IKEA plugs are roughly $10 each.

I sorted by price for each item, I said that is Amazon prices. Sorry, I didn’t go searching on dark web Chinese sites.

Prices can vary for sure, you can get the Zooz 4-1 for $25 right now from other places but you’ll pay $8 shipping. Could also wait for them to be delivered via Carrier Pigeon or get it in two days. Time and $ vs Time or $ can always be contested.

Somel said about $250, add the IKEA outlets, and it’s about at what my price point was. So both he and I must be doubling the price?

Precisely...I don't personally know anyone (myself included) that can afford to spend even $500 in one fell swoop on home automation stuff. I started out with switches and gradually expanded to dimmers, contact sensors, motion sensors, etc. Xiaomi is definitely cool and interesting, but I have four Iris plugs that I am not going to replace just to make Xiaomi work.

So I read through the Xiaomi hub LAN communication guide today (google translated from chinese).

Looks pretty straight forward at first glance... It is all multicast (hub discovery, periodic device status reporting) and unicast (device discovery, commands, and responses).

@chuck.schwer Not having written a LAN app for HE before, are there any decent starting points or documentation for this? Not sure if I can do it all in HE, or need a pi/man in the middle, but would need to:

  1. Listen for/receive multicast from end device
  2. Send unicast to, receive unicast answer from end device

I DO think Xiaomi would be great if you started that route and stuck with it.

I don’t know if other devices have issues with using the IKEA or Xiaomi outlets, switches for routing or not. I assume it’s just the other way around.

This whole thread makes me happy that I am 50/50 right now with Zwave and Zigbee devices though. :blush:

Middle of last year I had about 5 devices when I moved to HE. My order of z-wave ZigBee and nest products towards end of last year came in at about £950 :open_mouth:. I had been planning it for a long time though :joy:. It was hard to press the order button!

But although I looked at Xiaomi, the rush for me was too much.

I spent about double that on replacement devices when I was troubleshooting Zigbee issues back in December & January. :unamused: The upshot was that once everything was fixed, I had a bunch of spare devices which I've used on a variety of pet HA projects. :sunglasses:

2 Likes