Fed-Up with Z-Wave battery devices

[sooty]

I've purchased a handful of European battery powered Z-Wave devices over the last year or so mainly for testing purposes both in HS2 and HS3. These consisted of a couple of HSM100's, an Aeon Multi-Sensor and a motion detector and T/H sensor from Everspring.

I've never been able to get any of these devices to work fully and reliably in either HS2 or HS3, so yesterday I decided to update to .88 of the Z-Wave plug-in and have one last attempt before consigning these devices to my HA graveyard.

After about 10 re-scans and lots of red log entries, an HSM100 eventually created the right devices in HS3 but exhibits one problem that it has always had here in that the battery level always shows as 100%. I now also have random false motion alerts and no luminance updates.
I've read Ricks comments on polling these devices but really, I fail to see the usefulness of a device that can only update temp / luminance at 1 hour intervals unless of course you want to be changing the batteries every few days.

The 2 Everspring devices now appear to have battery levels stuck at 60% even with new batteries and again the T/H device seems to eat batteries if the update interval is reduced below 15 minutes.

The Aeon Multi-Sensor is really the only one of the bunch that now appears to work as expected. I had all manner of issues with this device reporting false motion with each wake up in previous Z-Wave plug-ins but since the late .50 versions that now seems to be OK. I still fail to see the usefulness of the device in a HA environment unless it's powered by USB.

So, in view of all the time I have wasted on these devices I have now laid them to rest and perhaps I will take another look at battery powered devices in a year or so.

By contrast, I use a lot of multi-branded battery powered (non 2-way comms) RF devices (250+) in my system via RFXCOM transceivers and plug-in and I don't have anything like the hassle with them that I do with Z-Wave. Battery life on a Visonic motion sensor here is about 2 years in a busy room and 3+ years in a less used room.

I can't believe that all the issues are down to the device manufacturers because the AC powered devices that I've had installed for over a year worked fine in earlier versions of the HS3 Z-Wave plug-in and only started exhibiting problems in later versions.

I expect Rick will argue that it's not an insignificant task trying to support a multitude of Z-Wave devices and I can sympathize with that but given the length of time HS3 has been in development I would have expected things to have moved on somewhat faster. As far as my Z-Wave installation is concerned I'm in a worse state as far as supported devices go than I was in HS2.

It would appear that some other controller / software manufacturers (as HST have done in the past) are applying software workarounds to get over some of the problems with devices that are not fully compliant with the Z-Wave spec however from where I'm standing it would seem that Rick is not keen to do this in HS3 but rather steer away from non-certified or non-compliant devices.

I have 40+ European Z-Wave devices most of which are Fibaro firmware updateable units so I'm hoping Fibaro sort out their issues and make these devices fully compatible with HS3 to save me an expensive hardware or software change.

Rant over..

Paul..

[NeverDie]

By contrast, I use a lot of multi-branded battery powered (non 2-way comms) RF devices (250+) in my system via RFXCOM transceivers and plug-in and I don't have anything like the hassle with them that I do with Z-Wave. Battery life on a Visonic motion sensor here is about 2 years in a busy room and 3+ years in a less used room.

You didn't mention which mix of RF devices comprised the 250+, but if they transmit often, I would hazard a guess that with that many you're also better off not using z-wave and also using a different RF frequency entirely. i.e. to help guard against collisions, use z-wave for infrequent output (like controlling lights) and something on a different frequency for sensors (input).

I have a mix of more than a dozen temp-RH sensors that transmit around 433MHz, and they all perform reliably and last many years between battery changes. It seems that anything with a heartbeat can help you supervise the transmission reliability to whatever level you want. I'm not sure what all the options are for wireless motion sensors (some people here like the wireless X10's because they're inexpensive and easily interfaced with homeseer), but I hope to add some soon. Thanks for your datapoint regarding battery life of the Visonic. Do they work well with the RFXCOM?

It would be very useful if there were a table showing the battery life on all (or at least most) of the different wireless sensors, as it would help avoid a lot of after-the-fact disappointment. Maybe there is one somewhere.... Short of that, I suppose you could measure the small amount of currents they consume and then estimate the likely battery life and return to vendor if it's likely to disappoint. It would give a whole new meaning to DIY!

[sooty]

The bulk of my RF devices are Visonic and Oregon with a few other devices from Home Easy etc. Not all devices are transmitters some are controlled from the RFXCOM transmitters so are not constantly generating RF traffic.

I use 2 of the older RFXLAN units with 433 RX, 433 TX and 868 RX modules installed and these have been 100% reliable since 2007. My Visonic devices are the 868 European frequency and yes they have worked perfectly with RFXCOM for many years. The motion sensors are not ideal for HA use because they don't send a "No Motion" message and have an approximate reset time of 3 minutes.

Visonic devices (at least the ones I have) use a system where they send the RF data 3 times with a random wait time in between to reduce the chance of collisions.

I also have a stand alone Hometronic system for heating control which again is RF (868 band). This has been installed for about 6 years now and again has been 100% reliable.

I don't appear to have RF collision issues. The latest RF gear to be added was Z-Wave just over a year ago and initially I only had about 10 AC powered devices. This has since been increased to about 40 devices and these all worked well in HS2 and until the last few Z-Wave plug-in updates they worked well in HS3.

The only real issues I've experienced have been with battery powered Z-Wave devices and I can quite easily live without them.

Remaining Z-Wave issues I have now with AC powered devices are some devices randomly displaying as "Unknown" status and the fact that HS3 doesn't support group 3 associations on my Fibaro devices but this is apparently a Fibaro issue that will hopefully get resolved in the not too distant future.

Paul..

[NeverDie]

Although the next generation of z-wave sounds promising, I'm not as eager to drink the koolaid like I did the first time without first seeing solid evidence of efficacy. I'm convinced the older generation of z-wave devices that I have (mostly circa 2009) often don't seem to handle ordinary transmission failures very well. Even if they theoretically should, in reality their flavor of mesh networking (officially a black box) often disappoints, and the general lack of transparency as to what might be going wrong doesn't help. They're not upgradeable (other than by complete replacement with a newer model), so given all that I figure it's best to just altogether avoid loading them with any additional impairments. That's the long explanation as to why I mostly favor sensor input using non-zwave and on different frequencies. There are plenty of non-zwave wireless sensors out there that seem to "just work" perfectly fine, and so why not leverage that, especially if you have a lot of wireless devices. The z-wave sensors, IMHO, are a better fit for very simple and small setups, and only if tight economics don't allow supplementing with a different wireless sensor technology. When you get to large installations, like what it sounds like you have, it's actually probably more economic to supplement z-wave with a different wireless technology on a different frequency for sensor input, and doing that is likely also more reliable as well because you can leverage mature, wireless technologies that have solid trackrecords. So, at least for now, z-wave is maybe good for light switches--as limiting as that is--as there still aren't completely good alternatives out there for that (at least that I'm aware of). I bought my z-wave light switches and dimmers thinking I would probably have nearly everything be z-wave, but after experiencing z-wave for what it is I now think that only my light switches and dimmers will be z-wave, and probably little beyond that.

[sparkman]

Hi all,

I have a number of battery powered z-wave devices that work well and have had reasonable battery life (6 months-1year+):

- a dozen or so Fortrezz water sensors (although I've had a couple fail right after their 1 year warranty was up so have some Evespring water sensors on order to see how well they do)
- two Everspring luminance sensors
- three Kwikset deadbolts
- two CT30 thermostats

I'm still using HS2 for z-wave and likely won't migrate it to HS3 until the end of the year. I did buy a couple of z-wave motion sensors during the sale and based on what I'm reading, I'm a little worried abut those . I use x-10 motion sensors now and am getting great performance and battery life from them so only wanted to try the z-wave ones to see how well direct associations work with them. The z-wave battery devices do tend to de-stablize the overall network though, so I agree that it's likely best to try and keep as many battery powered devices on different frequencies/protocols.

Cheers
Al