Z-Wave Network Stability Adversely Impacted, Seemingly After Having Added Aeotec Home Energy Monitor

tzest · January 16, 2020, 5:54pm

Hi, I have (or had!) a very stable Z-wave network mastered off a Qolsys IQ 2 Plus from Surety. The Z-wave network consists of a couple light switches, 2 Schlage BE469ZP door locks, an ADC-T3000 t-stat, 3 ADC-S2000 remote temperature sensors and, most recently, an Aeotec ZW095 (Gen 5) home energy monitor.

Prior to adding the home energy monitor, the rest of the nodes on the network have been up and functioning perfectly for months. I had been proactive in monitoring signal strengths and running network rediscovery upon adding any new nodes. I never had a single instance of interrupted communication to the Z-wave mesh.

Strangely, after adding the Aeotec ZW095 home energy monitor (added months after everything else on the network), I have been getting intermittent instances every 1-3 days where each of the two door locks (at different times) would be reported as having gone offline. With time or forced use, the devices immediately come back online again.

All Z-wave devices (including the locks) have their nearest neighbor and route going directly to the IQ panel. The door locks each have multiple neighbors. Additionally, the locks are each “closer” (distance-wise and, I believe, signal-strength) to the panel than the home energy monitor. If I look at the routing on the panel, they ordinarily have “good” signal strength, but when I look at one of the locks when offline, of course it shows as “unreachable”.

After adding the home energy monitor, I relearned the network a couple different times, the most recent time after forcing a change of state in all possible nodes to forcibly wake them up. I am puzzled how having added a single Z-wave device (that should act as a repeated) could do anything other than strengthen the stability of the network.

Any thoughts…?

jwcsurety · January 16, 2020, 7:22pm

Adding a repeating device shouldn’t have a negative impact, correct. Whenever an issue concerns Z-wave locks, the first thing I like to rule out is battery levels, because that might be causing an issue coincidentally.

When is the last time the lock batteries were swapped?

Borderline low battery levels on locks has resulted in some signaling inconsistency in the past. It would be good to try replacing the batteries first, run a network rediscovery, then monitor the network a short time. Does this improve their function?

tzest · January 16, 2020, 8:48pm

The batteries have never been swapped, but the system itself is pretty new (maybe 7 months old for the oldest component). The batteries for both locks were installed as a fresh set and health is reporting as 96% for both locks… Is there any reason to suspect battery health reporting may not be accurate and try replacing the, anyway?

jwcsurety · January 16, 2020, 9:01pm

Lock battery life-span is dependent on usage, but is usually no more than a year, so 7 months is not out of the realm of possibility for starting to get low.

Even with the percentage report, I would still try swapping the batteries, at least in one lock, as a test. It is something I would always try to rule out if locks are the problem. If batteries are the problem, fresh batteries should result in an obvious improvement.

One other thing that might be an issue with a Z-wave network when adding more devices: network signals can only jump four times, so they will only go through 5 devices to the destination.

Adding devices can alter routes, although in a normal home it would be extremely rare for a device to be far enough away to necessitate four jumps.

xeon · January 22, 2020, 6:02pm

just wanted to add to this, just got my Aeotec Home Energy Monitor installed and running, and also did experience some weird Z-wave issues myself, for example, my “garage” (Linear) went offline and on, and when I activated my scene to “away”, it didn’t lock my doors. Could be a coincidence (I’ve had the Garage go offline before on and off before the Energy monitor was installed, but it corrects itself, somehow), but never had issues with my Z-wave locks.

I’ll report over the next couple of days.

xeon · January 23, 2020, 4:13pm

I can confirm that indeed there has been some weird things going on with my Z-wave devices since the installation of the Aeotec ZW095 HEM. 1 of 3 of my door locks went offline while armed stay for about 6 hours and malfunctioned according to my ADC.com app (it was grey out), however, there was no malfunction notification on my IQ2 panel, infact, when the app was saying it was offline, I just unlocked it from my panel, then it woke up in the app. This lock, has never given me issues. My Garage also went offline momentarily according the app, but my garage works fine.

Wondering if this is a Z-wave network issue in my home, or an adc.com thing?

Has the OP solved his problem? Battery levels for the lock in question was 60%.

I have a total of 8 Z-wave devices, and they are all within 35 feet from the panel.

EDIT: another thing I forgot to mention, when i activate scenes which involve the door locks, it used to be that it would be almost instantly that the door locks, now its about 2-3 minutes for it to lock. I have to check if its the app that is delayed in updating or if actual lock is delayed in locking

tzest · January 23, 2020, 4:40pm

Sorry, I’ve been keeping tabs on this thread, though I have not yet taken any further action at this time:

Jason, I’m confident that the batteries are still good, though I understand your desire to remove a variable nonetheless. I’ll see if I can swap batteries this weekend. Also, I can’t see the 4-hop limitation as being pertinent here - mine is a small network with all devices in pretty close proximity to one another (one signal-strength exception worth noting is an outdoor light switch that’s in a metal box, so I have a plug-in switch acting as a repeater as close as possible to it on the inside wall).

I’m attaching some pics I took from the panel from one of the events when a door lock was offline.

xeon, thank you very much for sharing your experiences! My problem is not resolved, but, as I mentioned, I also have not tried Jason’s suggestion to try replacing batteries (ADC indicating that they’re at 96%).

My experiences seem similar to yours: The device go offline (and registers as “malfunctioning” per ADC) and comes back online either on its own with time and with no action, or simply by manually locking/unlocking, which immediately and consistently wakes the device back up and has it successfully started communicating again.

Based on what I saw at my panel (see pics), this seems like a Z-wave issue, more than an ADC thing. That having been said, I’m not yet convinced that this is an issue unique to my network, especially given our shared experiences.

Perhaps we can both try replacing the batteries in our locks, share the outcomes, and see if Jason has some additional thoughts…? May I ask what type of locks are you using?

xeon · January 23, 2020, 4:55pm

That sounds like a good idea. I’ll change the batteries on the next offline instance (whenever that may be). In the meantime, I will check out my z-wave maps to see how the connections are formed. Can you/someone remind me how to get those screens again?

The lock in question is the Kwikset 914 Touchpad (but its called Weiser here in Canada)

jwcsurety · January 23, 2020, 7:01pm

My experiences seem similar to yours: The device go offline (and registers as “malfunctioning” per ADC) and comes back online either on its own with time and with no action, or simply by manually locking/unlocking, which immediately and consistently wakes the device back up and has it successfully started communicating again

Locks are operating in a power save mode as battery operated devices. It is not uncommon for them to be affected first with supervision errors, but this is definitely odd for a specific device type to cause it, and especially since the lock doesn’t appear to be using the new meter (or any other device) as a repeater.

Your network map seems odd, with further devices having stronger signals than a nearer one, in fact used as a repeater.

If you have not already tried it, before the battery, try power cycling the panel. I’m curious if a panel reboot might result in a different network map.

tzest · January 24, 2020, 3:42am

Jason, for what it’s worth, I had rebooted the panel in between some of these events. But I understand you’re referring specifically to the network routing map here.

The tstat is an ADC-T3000, whereas the 3 temp sensors are ADC-S2000-T-RAs. The RTSes were learned in at their installation locations. I found their specific association with the tstat as a repeater interesting, too, but figured that was just how they preferred to handle their associations.

Here is how things look today under nominal conditions:

…Here is how things look immediately following a reboot of the panel:

…And here is how things look ~1 hour following the panel’s reboot:

jwcsurety · January 24, 2020, 2:14pm

Just to verify, you definitely have a GEN 5 ZW095 HEM? I’d like to check with Qolsys to see if they can replicate or confirm any potential cause of this behavior.

My main guess as to why this might happen is that the ZW095 HEM may be hitting the Qolsys panel with repeated statuses in quick succession, impacting the radio’s ability to process other signals.

tzest · January 24, 2020, 2:56pm

Yes, just confirmed - mine is an Aeon Labs Aeotec ZW095-A Gen5 HEM.

xeon · January 24, 2020, 3:56pm

thanks for the information!

Yesterday when I got home, I reset the panel AND did some Z-wave diagnostics such as network rediscovery, device rediscovery and ran the tests, etc. Too soon to tell, but so far so good.

Hopefully the HEM is not bogging the IQ2 panel down! (supports 119 z-wave devices).

tzest · January 24, 2020, 4:46pm

xeon, how frequently do you see this from your end? I went back and filtered activity on the ADC web site by device “Tamper/Malfunction”, going back to the date of my HEM’s installation. From my end, I counted 17 distinct events across 47 days. Where events involved multiple devices across adjacent time spans, I counted each device event separately.

I found one series of events where both door locks went offline within a minute of each other, which was interesting, especially in light of Jason’s theory.

jwcsurety · January 24, 2020, 4:55pm

I’ve sent some info to Qolsys and asked them to look into it. Will follow up here with any response!

xeon · January 24, 2020, 5:59pm

Thanks Jason for following up with QolSys.

Tarek, using the same criteria as yourself, I counted 5 distinct events over 2 days (my HEM recently only installed about 3 days ago), however, none since the reset/diagnostic procedure I described in a previous post.

Its only 1 device that is troublesome, my back-door lock which also happens to be the furthest away from panel, but still within 40 ft from the panel.

xeon · January 25, 2020, 6:15pm

Just had another instance of an offline lock. (garage door). In this case, the lock is 5 ft from the panel, which is strange. Here’s my zwave network

Just after panel reboot

This after some time from the reboot.

So, why does the node network change so often?

Here’s an ideal node map

jwcsurety · January 27, 2020, 3:32pm

Where is your home energy meter mounted, how close is it to the electrical panel? Is the transmitter mounted to the metal panel housing? It should be a little away from the panel and away from power lines. It doesn’t need to be too far, but if it is too close the metal/power lines can interfere with the signal.

xeon · January 27, 2020, 4:23pm

My HEM is located beside the electrical box (about 6 inches) and under the wires going into the electrical box (another 6 inches). Is this far enough? I can easily move it to do tests. The HEM is mounted on plywood.

At this point, my Z-wave network is very unreliable (at least my locks are), so i’ll try anything to get this resolved.

jwcsurety · January 27, 2020, 4:55pm

Try moving that a little further away from the electrical panel and wires if possible. Any improvement?