I just got the new ZWA-2 wave controller from Nabu Casa. It replaces an Aeotec Gen 5 controller. The Aeotec controller was at firmware 1.02. I was able to successfully backup the old Aeotec controller and restore the NVM to the new ZWA-2 controller and all of my devices connect to the new controller. However, I now experience frequent and sometimes extended periods of time where the new ZWA-2 controller reports as “Jammed”. In ZwaveJSUI it reports “The controller is unable to transmit”.
I have a read a few topics on this and it appears this issue affected Aeotec and SiliconLabs controllers a year or more ago with a few specific versions of the Zwave SDK. I also read that nodes which are particularly chatty like switches which also report power readings can cause a jammed state. I changed all 5 of my switches with meters to hopefully reduce these jammed states. I checked my background noise (RSSI) and it averages about -110dBm which I believe indicates virtually no or extremely low background noise and interference.
This is the first time I have ever experienced this. My old Aeotec Gen 5 controller never had a jammed state. I run ZwaveJSUI in a docker container on my Synology NAS. If I switch back to my Aeotec controller, everything is fine. What’s going on and how do I fix this?
The controller has 14 dropped RX messages and 36 dropped RX total commands. No dropped TX commands or messages. Only two nodes have 5 or less dropped RX commands. This is all after 2 hours up time. The nodes have considerably less dropped RX commands than with the Aeotec controller. They would have close to 50 by this time.
So as I understand, ZWave controllers at and above 700 series are notorious for having soft locks at times. I encountered many jams that you describe when I upgraded to a Zooz 800 series. It became very problematic at a certain point, and I believe it was partly due to a newer sdk released by silabs.
Subsequent firmware versions introduced some workarounds to avoid getting stuck in jammed states. Zwave JS UI, for example, has a soft restart function that essentially resets the controller to get it unstuck from a jammed state. Another firmware revision reduced the number of jams I’ve encountered, but it still sometimes happens. I dont think there is a fix for this until SiLabs really works hard to get it under control, so unfortunately there may not be any solution for your issue, aside from the workarounds, like soft resetting the controller when it locks up. Sorry mate. I know its not the news you wanted to hear.
Thanks for the info. Yes I had read about that on both GitHub and either a Reddit post or a HA community thread. I guess I misunderstood the issue. I thought it was an issue that affected a couple certain controllers and not all. I guess though the issue is in the SDK and all controllers must use/implement the SDK, then it makes sense that all controllers above a certain firmware would be affected. I guess I could just revert back to my Aeotec controller in the meantime as there wasn’t really anything wrong with it, just some oddities I thought were due to an older stick.
I would still suggest you try the usual troubleshooting stuff when diagnosing radio issues. Although the zwave controllers can be susceptible to jams, I dont think it should be that bad, especially on the newer sdk revisions. Make sure your usb connection is not usb 3, the antenna placement is ideal, the device is on the latest firmware, etc. Try and see if there is also a soft recovery function in your setup.
There is a soft reset option. It’s always been disabled for me. However, if I read the note correctly it seems disabling doesn’t apply to newer controllers like the ZWA-2. It doesn’t imply these controllers just automatically soft reset, but it seems to do so anyway. Anyway, turning it on doesn’t seem to do anything different in terms of issues I experience.
Unfortunately, My NAS only has USB 3.0 ports on it. I do have a 6 foot USB extension cable between my controller and the NAS. I could maybe find a USB 2.0 hub and put that in between my NAS and controller if the extension cable isn’t enough. I know firmware is up to date on both my Aeotec and the new ZWA-2.
So I plugged my Aeotec controller back in and unplugged the ZWA-2. I put it in the same exact spot as the other one. The average RSSI went from -110dBm to -83dBm with the Aeotec controller. Now I’m not sure which is “better”. I’ve read conflicting info on signal strength and the noise floor, but the Aeotec seems to be stable and generally work better.
I had the same issues with the 700 series controllers. SiLabs has broken zwave with this nonsense. Once I transitioned all my devices (except my locks) to Zigbee, things got a lot better. How many Zwave devices do you have?
I was hoping the ZWA-2 meant that Nabucasa had squashed all the bugs with silabs, but seems like its the same old story. The older 500 versions had issues too, but they didn’t have these problems.
Full Disclaimer, I am not a radio engineer nor that knowledgeable of the subject, so my thoughts are my own and may be partially incorrect, so dont take my word as full truth.
That said, I do believe the reason you are seeing higher interference with the new radio is because the antenna is much bigger and therefore more susceptible to environmental noise. From my experience with troubleshooting my own zwave difficulties, the interference from noise did not appear to influence the controller jamming or locking up. It would happen at random, and inconsistently. What did consistently cause the controller to jam was when the network would become very busy, such as reinterviewing a node or healing the network.
Keep in mind, this was not normal behavior nor did it happen to me UNTIL the controller firmware was updated to a version with Z-Wave SDK 7.19. This sdk version was arguably the worst with the network difficulties, and it almost became impossible to use z-wave, even though I only have about 25 nodes.
Subsequent firmware revisions did help, and while I cant remember the exact firmware revision that fully solved the issues, I can say that sdk version 7.22.1 was when I saw a decrease in the number of issues I was encountering.
The only change in the network was the controller firmware. Position, placement, server setup, number of nodes, etc all remained the same.
My controller is now on a firmware that runs sdk 7.23.2 and I have not encountered a single jam or lockup.
Its POSSIBLE that it has occurred silently, but I do not experience any side effects, as I believe the workarounds and fixes in place have made jams rare, and solve any jams that may occur by soft resets.
I should also mention that part of any troubleshooting step is to do a full wipe of whatever you are troubleshooting. Unfortunately, this can be time consuming so I understand the hesitation in doing so, but I would suggest factory resets of the controller and nodes. Especially when considering migrating networks between major controller revisions, theres bound to be potential issues.
Unfortunately I cannot help any further, aside from sharing my experiences, the usual troubleshooting steps that are commonly mentioned, and ideas that might help.
Radios are weird and can be very finicky.
FWIW, I’m also seeing the controller is jammed log message. I did not migrate my Z-wave network, instead moved all my devices over individually. I was hoping this would put me in a better state.
I think you will find the Jamming is a long running issue with the SiLabs SDK, and has been around quite a while. 700 and 800 series controllers seem particularly susceptible to it, hence why you didn’t get the issue with the old 500 series controller.
There are methods built into ZWaveJS to try and compensate for it, but not much can be done until SiLabs fix the underlying issues with the SDK.
This is not unique to the ZWA-2 Virtually all 700 and 800 series controllers suffer from the issue.
The issue is just as present in the 7.23.2 and higher SDK’s as it is in 7.23.1, so don’t expect updating the ZWA-2 to 7.23.2 to fix the issue.
Quite honestly, unless you have a specific reason to use the 800 series controller, a ZWA-2 is not going to be that much benefit over the old Aeotec 500 series controller. A lot of people got the ZWA-2 to support Nabu Casa, but you really won’t find it a much of an improvement over other controllers, unless you have a lot of long range devices that communicated directly back to the stick and don’t use the mesh. Upgrading just for the sake of upgrading is not always a good thing, and sometimes you will find more bugs in the new version than in older ones.
I upgraded my aeotec 500+ old stick to ZWA-2 hoping to get a better zwave network and to be more future-proof.
Worst decision I ever made.
My ZWA-2 is costantly jammed ( Controller is unable to transmit)
Zwavejsui soft resets are able to mitigate the issue, but yesterday I had to unplug the controller to have my network back. And I lost a difficult to physically reach node.
My network was OK with the old stick
I don’t know if I can restore a 700 backup on a 500 controller. Otherwise I will have to reset and reconfigure 10+ nodes I added after the controller migration.
To be honest, I turns off noisy features like power monitoring or temp/lux/humidity monitoring on devices that didnt need them enabled and my network jams a lot less. This has been a problem for way too long though, and im starting to think silicon labs cant solve it.
