S2 authenticated or unauthenticated for Range Extender

Hi - trying to work through an issue with two switches that are located outside the house, so range is an issue. They worked OK on ST, however making the move to HE, and they are problematic, so I purchased an Aeotec Range Extender 7. I have a couple of questions, as they are still not working correctly.

  1. All my Zooz devices are S2 Unauthenticated. When the Aeotec 7 paired, it asked me for the DSK, so I entered it, and it shows up as S2 Authenticated. Should the S2 Unauthenticated devices be able to route through it still? (not sure if that's the correct term)

  2. The devices are not routing through this (I have actually added it in S2 Authenticated & S2 Unauthenticated modes, even though I do a Z-Wave repair after adding it. They go through other switches/devices (occasionally - mostly they just won't pair or respond), but not this repeater, even though it's now outside & <10 feet from them. (have tried it just inside the door, too, which is still only about 15' away...) Am I doing something wrong?

Thank you for any help - pulling what's left of my hair out over this... =:)

This thread and this post on particular seem relevant to your question:

But to summarize, secure devices can route thorough non-secure nodes and vice versa, so that shouldn't be your problem. Perhaps you have not given it enough time, or maybe your devices judge that other repeaters are better for some reason.

Thanks, Bert!

So does the “overnight zwave repair” do something different than initiating it manually?

I think the nightly process is a "cleanup," not a repair. The cleanup removes "ghost nodes," nodes that the Z-Wave radio thinks are there but Hubitat no longer has a device for, something that can happen if a Z-Wave device is force-removed. I'm not sure what else it may do but have not heard of it being an actual Z-Wave repair, the general advice for which is to not run one if you don't need it and to consider the new per-node repair option on the C-7 if you do.

1 Like

Correct, nightly cleanup is not a Z-Wave repair from what @bcopeland has said about it, just a clean-up as you noted.

You could make your life simpler by unchecking all security options when the S2 dialog (below) comes up when adding new devices. Other than for locks/entry devices many of us feel that security is an unnecessary complication on devices like light switches, extenders, etc.

3 Likes

If you do not de-select it on discovery, is there a way to turn it off after the fact?

no. But if it paired correctly the 1st time - leave it alone. Once paired and working there is zero reason to re-pair it non-secure.

If it isn't working, well that is a different story. Then remove/re-pair it.

2 Likes

Ha! Thanks for adding that additional commentary to your no :slight_smile: I was just going to ask.

Good point - I was not suggesting to remove/re-pair working devices to remove security. [eek!] :wink: Thanks @JasonJoel for clarifying that.

Agreed. I have a Zooz ZEN16 MultiRelay that paired securely but then was slow as molasses in responding to commands, to the point of being unworkable. It controlled an outdoor Fireplace LED controller that had to be turned “on, off, on” within about 3 seconds in order to put the LED controller into a color-cycling mode, and with each on/off command taking several seconds, the LED controller couldn’t be controlled. Re-paired with no security, operation returned to normal.

First, thank you all for your replies - community support is what makes platforms so great!

I think I understand it better now - sounds like I could do either mode, but some of the other posts I read indicate that S0 might be faster/ more reliable for some?.

S2 and non-secure are same speed. S0 is always slower.

Always. Because it takes 3 messages for what non-secure or S2 only take 1 message for.

oops - yes, thanks - unauthenticated is what I had meant, not S0... :slight_smile:

I realize this thread is 4 months old, but I have a question that feels related.

Background
I migrated from ST to a Hubitat C7 hub with firmware version 2.2.4.158 a few weeks ago and I was having a lot of problems with delayed and lost events. All involved devices were Z-Wave.

While carefully experimenting with 1 set of changes at a time, I noticed that I had 2 Honeywell-Jasco switches that I had included with "Security=None". They were located together in a 2-gang box. A third Honywell switch, located about 5-10 feet away was included with "S2 Unauthenticated". After excluding & resetting the 2 "Security=None" switches and then re-including them with "Security=S2 Unauthenticated", everything stabilized. It was a dramatic improvement!

Prior to the change it would take a long time after a hub reboot for the routing tables exposed by @tony.fleisher's excellant Z-Wave mesh tool to populate, and many of the routes seemed very circuitous. After the change, those routes were showing up in just a few minutes and they seemed much more reasonable.

My questions

  • Is there a well-understood reason why the change from security="None" to "S2 Unauthenticated" should have such a dramatic improvement in my network stability?
  • I have 1 Zooz ZEN30 that I have included as "Security = None" based on these posted instructions from the Zooz folks. While it is working reliably, it has the worst "LWR RSSI" measure (15db) of all my devices, shows the hub as it's only neighbor despite another device that is closer showing the Zen30 as a neighbor and can never be successfully "repaired". Could this be related to the use of "Security=none" vs. "Security=S2 Unauthenticated"?

Thanks in advance for any insights you can provide!

Marc

FWIW: My Jasco/GE/Honeywell Z-Wave Plus switches are mostly joined w/out security, though I do have two joined w/S2 Unauth, and all of the switches function w/out issue or problem, extremely consistent, never have to worry about them. I have never seen a relationship between S2 security and their performance, mixing S2 and no security w/these switches has not been an issue for me. (Also have some earlier generation GE Z-Wave switches that don't support S2 and they also play nicely.)

In your case it may have been that simply excluding and re-including the switches resulted in them making a better connection/routing to your mesh.

I don't have any Zooz swiches so can't speak to that specifically. However:

I don't know why you would want to repair a switch that is working reliably. :slight_smile: Sometimes looking at the Z-Wave details page can lead you down rabbit holes that you really don't need to go down. If things are working, they probably don't need fixing. :smiley:

2 Likes

Spot on with this. (Do you mind if I quote this in the documentation for the Z-Wave Mesh tool?)

2 Likes

15db wow that is high!1 I don't know much about HA BUT A RADIO IS A RADIO

+15 DB IS REALLY HIGH

Really high signal can interfere with each other.

Yes. The act of excluding and re-including causes routes to be redone.

1 Like

My thanks ot everyone for all the great info!

Valid point. I'm in that phase where I am reading a bunch of stuff, experimenting and learning just enough to get myself into trouble :slight_smile: Good point and I am going to let it go.

The "LWR-RSSI" measure is a measure of the difference between "Last Working Route" RSSI and the background RSSI & should be >=16, per this Silicon Labs doc, page 18.

Still learning, so please bear with me. Is my ZEN30 still operating as an S2 device when it is joined in "Security = None" mode, or is now operating as a less-efficient S0 device? Based on reading JasonJoel's quote, I believe it is the former (more efficient), but want to make sure I am understanding correctly.

Thanks again everyone for your patience with the "new guy".

Marc

1 Like