Too many Z Wave repeaters?

Lounge
#1

I know there is a similar thread on this topic, but regarding Zigbee. I wanted to poll the groups collective thoughts on whether in your experiences there can be such a thing as having too many Z Wave repeaters.

I have grown a reasonably sized mesh network. My Hubitat is centralized on one side of the house and device installations progressed outwards. During all the device installations, I did have troubles with one device in the garage (lock installed in a steel door, with radio on opposite side of door), so I installed a switch and a Aeotec repeater in the garage. That solved it. As I started to expose dashboards throughout the house, I noticed that certain tiles were not always accurate for the state the device was in. Sliders were also not always responding appropriately. One day, I decided to install another Aeotec repeater in the middle of the house. After the Z Wave repair, I noticed a significant difference in dashboard tiles being accurate, sliders responding, devices taking action immediately, etc. The much better experience remained consistently. So, my thinking was hmm, why not install some more repeaters.

I then later installed two more Aeotec repeaters on opposite axis's of the house, about midway between the Hubitat and installed devices. Then executed another Z Wave repair. I have started to notice a much less reliable system again.

I wonder if too many repeaters can be a thing? Is the hop limitation of Z Wave and the number of repeaters affecting stability? Just not sure. I am going to try and move the two extra repeaters further out and/or remove them again and see if things improve.

What has been your experiences? Does anyone know of any tools I can use to better diagnose the mesh network performance?

Thanks in advance.

#2

I have a love/hate relationship with the Aeotec Range Extender 6 repeaters. When they are included the significantly improve the entire mesh and are an excellent addition to the zwave mesh.

However I have experienced at least 3 times now (in about a 5 month span), I notice that the repeaters green light starts no longer randomly blinking as they do when first installed. Shortly after I notice this no surprise I start seeing devices not functioning properly. At which I have to remove each (of my 5) one from my mesh, doing a zwave repair after removing each one to properly re-route the other devices more efficiently. Then re-include the Aeotecs one by one also doing a zwave repair after each to build the mesh back.

Unfortunately there is nothing in the generic driver to help with determining the status of these and when a problem occurs.

#3

I have been using this driver written by @srwhite for Iris z-wave repeaters with the Aeotec Range Extenders. Worked fine for the last ~2 months. The driver periodically checks to make sure they are online. None of them have gone offline after I switched to this driver. My z-wave toolbox indicates they are still functioning appropriately as repeaters.

1 Like
#4

I was using that driver this last time they stopped working. The driver kept logging that they were working fine Even though when I range tested them by holding in the button until it turns purple a couple of them blinked red stating no communication, but the driver kept saying they were online. But the minute I excluded the repeaters from my mesh the devices started working again sporadically.

But I'm about to add them back because there is no consistency without them. So it's a love hate about once a month repeat this process.

#5

I use srwhite's driver too and never had an extender drop off. What firmware version are those extenders on? srwhites driver shows that to you

All of mine have version 1.5

#6

I'll have to let ya know when I include them back. Right now they are excluded.

#7

How are your locks holding up? :wink:

#8

Right now 2 of my 3 are (somewhat) staying online, my outside shed keeps dropping off.

#9

Back to the topic.

Yes, Zwave does have a limit when it comes to number of hops. Any Zwave message cannot use more than 4 hops between the hub and the device. Zigbee does not have this limitation.

However, this does not mean that you can only have 4 repeaters. The idea is still the more the merrier. The Zwave protocol will try to avoid using more than 4 hops and this is all done during a Zwave Mesh repair. All the neighborhood tables are build then and the devices will know which path to go. The 4 hop max comes more into play if you try to cover a great distance. Zwave is limited when it comes to how far you can stretch the network

#10

I was not aware of the Z Wave Toolbox. Interesting looking tool to have. It's now ordered. I will also switch the driver for my repeaters to see how that might help. I do think the Z Wave Toolbox may be exactly what I need to optimize the mesh though.

#11

This is great! Thanks for pointing out this driver!

S.

1 Like
#12

I bought the toolbox not long after I obtained my Hubitat. For me, it's never quite reached truly useful. That may be my own lack though, it's been on my network since about day 7 with Hubitat, rock Solid and does something...lol

#13

So I looked into this and 3 of my 5 are version 1.4 and 2 of them are version 1.5

I contacted Aeotec about getting the latest firmware so I could update them and they referred me to this page. Range Extender 6 (ZW117) firmware update. (1/15/2018) : Aeotec Group

However the instructions say to "unzip" the zip file.....except the link provided for the USA download isn't a zip file it ends in .rar at which Zensy tools doesn't appear to want to accept it.

Waiting to hear back from them.

1 Like
#14

rar is a different zipping algorithm. I think a tool like "7-Zip" can extract it.

3 Likes
#15

Point of comparison. All of mine are device version 1.5. And they haven't fallen off.

1 Like
#16

I have the same results with 1.5.

1 Like
#17

So far so good. I updated one and on the "test communication" of that Iris driver last night I was getting 16 frames returned, and after the update I'm getting all 30 now.

I also have another plan of attack that I'm now trying for my locks. My hub #3 was just a spare backup in case one of the two productions went down. Now it houses a second zwave mesh with ONLY these Aeotec Range Extenders and the locks connected by HubConnect. It was a pain getting it all actually connected and it gave fits for the first few hours but it seemed to have settled down and now the reliable locks hasn't missed a 5 minute poll for the past 6 hours.

But a few days from now will tell.

1 Like
#18

All have updated firmware now. 4 of the 5 now instantly send back all 30 test frames received SUCCESS. 1 of them (which is my most furthest away in a metal building) still is sending 0 test frames received and FAILED, even though it joined back into my HE mesh.

Will give this a day or so to see if it improves, however the locks are still consistently polling every 5 minutes without missing a beat even when I excluded one extender at a time and after I'm done.

Thanks @dan.t for suggesting looking at the firmware.

Edit: Didn't take long 2 of the 3 locks have fallen off again, and the range extenders are not sending full amount of test packets. Another test crashed and burned. and I'm still on 2.1.5.124 so it's not in reference to other posts on the update messing up their zwave.

2 Likes
#19

@waynespringer79 I am having the same issues that you described with the Aeotec extender. I have four of them and am still trying to figure out best placement. I replaced the generic driver on all with the @srwhite version that was suggested. One in particular is 25-30 feet from the hub, but was still failing on all 30 test communication attemptseven when I moved it slightly closer to the hub. In a fit of desperation I swapped that extender with another one (also roughly 30 feet from the hub but in a different direction) and suddenly they both have success on 30 of 30 attempts consistently, over the past two hours. That makes no sense to me at all. For now I'm going to start a z-wave repair and see what it looks like in the morning.

Download the Hubitat app