[SOLVED] Simon Tech Zwave network very slow responding after full reset via Simplicity Studio

Setup, Configuration and Use Installation
1
  • Platform information:
    • Hardware: RPi4 8GB 240Gb SSD
    • OS: Raspbian 10 (buster) - Kernel: Linux 5.10.103-v7l+
    • Java Runtime Environment: 11.0.6
    • openHAB version: 3.4.0.M3 (persisted items with MySQL and with these bindings: Amazon Echo Control, Astro, ipcamera, MQTT, Spotify, Weather Company, Xiaomi Wifi devices (Mi IO) and Z-Wave)
    • Aeotec Gen 5+ Z-Wave stick (I’m using a USB extension cable since the server is inside a closed rack closet for all the network of the house and other electronic equipments, such as sound system controllers)
    • Aeotec ZW189 Range Extender 7 (there are 4 of these spread around the house)

Ok, here’s again another Zwave slow network problem. So, in this case we are talking about roughly 130 devices from Simon Tech brand, between Lights and Blinds. So no other Zwave devices from other brands are mixed here.

I’ve done already some installations with OH, on my previous one the total quantity of devices was similar but there was a mix of brands and even other norms such as Wifi, RS485 and ZigBee. We are talking about nearly 200 devices and the server have nearly the same characteristics. Everything working perfectly (in this case, around 100 of these devices work in Zwave). In this example, I did had problems with Zwave slow network, but after deleting the “ghost nodes” using Simplicity Studio, everything started working flawlessly.

Going back to my current problem. There is a mix of devices from Simon Tech working with Zwave 5 and others with Zwave 7. This I don’t know if it’s eventually a problem or not.

There were some buttons that were changed, so to clear the network I also used in this case Simplicity Studio to clear the “ghost nodes”. I ended up clicking in “Remove” (which, weirdly, doesn’t even ask for any confirmation at all) and I accidently removed ALL 130 devices from my Zwave stick… so I literally had to add them one by one. After an entire afternoon doing it, I had to first remove the connection from each device to the USB stick (because although they are not in the USB stick anymore, each devices “thinks” it’s still connected, so it had to be removed). And then add one by one.

I noticed some buttons were difficult to add, others were very fast.

Now, my question is, before I complain with Simon Tech so they can bring some specialist from their side to the house, how can I really make sure the problem is really from their devices? RPi usage percentage is perfectly fine (around 30% CPU usage when I am using the server and around 5% in idle, and temperature is also very reasonable, between 38º and 45º).

So what can I do next to try and solve this? Sometimes this is ridiculously slow, like taking more than 1 minute to respond. :frowning:

Thank you so much for your time in advance.

2

From your description, I’m assuming you clicked the controller reset so all the devices were not properly removed, so then had to be factory reset them before readding.

After such a major disruption the radio wave routes may not be optimal for a while, I would try a network heal. It may take some time, but you can watch the progress by the timestamps in the userdata/zwave folder. The other option is to just try to heal the nodes that are slow. You can start a heal immediately by setting the Heal time to the current hour (but nothing will work for a while.)

On the unsolicited advice front; :wink:

  1. use the backup feature of the silabs tool before starting.
  2. With the number of zwave devices, program a spare zstick to a zniffer. You can then see the radio traffic and figure out your problem nodes
1 Like
3

Yes, something like that! :smiley:

And is that “while” supposed to last more than a week? Because this has been a week ago and things still very slow. :frowning:

There are many of them that are slow. The difference is, some are VERY slow, others are “just slow”. Very few work “instantly”.

But very appreciated! :smiley:

You know, I’m a developer for more than 20 years now and I still miss this basic step sometimes… overconfidence I suppose! :frowning:

Have been looking into this and will definitely do it. What is the one you suggest?

And, as always, thank you very much for your time Bob! You are the GOAT.

EDIT: I just noticed something right now… 3 devices are NOT responding. Hours ago, they were ok. Now, they are not. Besides that, I’ve noticed OTHER devices acting like that as well, so these are not isolated cases.

4

A week is long, I assumed this was more recent. Did you have the network heal configured, default is 2am.

If I recall correctly any 500 chip zstick can be reprogrammed as a zniffer. It has been a long time for me and it looks my link to the process in this post is gone, but there has to be tutorials out there.

Not exactly diagnostic, but with Silabs, go to the IMA screen and check out the network health tools. The rediscovery option is the same as an OH heal. It will point out the problem nodes anyway.

1 Like
5

Ok, in this case specifically I think I know now where the problem lies. After talking with a friend of mine, sales person in Simon Tech, it seems between 2022 and 2023 they changed supplier due to the chips shortage and they started using a different Zwave version on the new buttons.

In this specific case, half of the buttons have the old version and the other half, the newer version.

As per what this guy told me, they had experienced exactly this problem that I am reporting in houses with multiple devices and a mix between old and new version.

Solution is: removing all old buttons, send it to factory and have them update it.

Anyway, thank you so much once again Bob!

1 Like