ZBT-2 lower LQI than ZBT-1

Hello everyone. My zbt-1 was actually reaching end of life, with frequent disconnections (4-8 times per day) from the zigbee network. So I decided to bite the bullet and purchase the new ZBT-2. Migration was smooth, since the device was automagically recognized by HA (running HAOS on a qemu VM). Once i selected to configure the integration, it migrated the ZHA network without a glitch.

And here comes the punchline: Network connectivity and LQI are much worse than used to be with zbt-1. With zbt-1 the lowest LQI was > 150 and now the low LQI on some 5 routers is < 18.

  1. Τhe new ZBT-2 was connected on a USB-2 port with it’s cable and placed on top of the rack. The cable run outside the rack, in exactly the same path as the previous USB cable with ZBT-1.
  2. I experimented by changing the USB ports (used both the old one that ZBT-1 was connected, and another usb2 that was unused previously).
  3. My network includes more than 25 routers (mostly Nous Plugs, aqara plugs and some lights from IKEA) and some 30 end devices (motion and door sensors)

I have searched the net, spent some hours on AI (gemini, claude and the likes) but to no avail.

Has anyone else had the same experience?

As long as you don’t have performance / connectivity issues do not worry too much about LQI, it’s a very unreliable metric. It can be reported vastly different from coordinator to coordinator. You are using an extension cable to it, not connecting it straight in the USB port?

Nope I’m using only the cable that came with the packaging. Seems robust enough.

That’s what I meant - not directly into the port. :slight_smile:

I replaced my old ZBT-1 with the new ZBT-2 just for performance reasons and it works definitly better (although i’m using it as Thread Border Router). Did you check the sticks configuration like baudrate and stuff?

Everything checked: Zigbee channel is 20 as in ZBT-1 (it actually migrated the setup from the old one). Port settings as per instructions. Speed / response on the network is good, (i would say better than the previous but I will do some more testing).

1 Like

I’ve tried various Zigbee controllers over the last several years and they all report different LQIs with devices and controller in the exact same physical locations.

The only apples to apples comparison you can ever make is if two controllers are using the exact same Zigbee chip. Because then it comes down to PCB and antennae design.

Unfortunately LQI calculations will vary across different Zigbee chips. As others have pointed out, if you are not having network or reliability issues, then the LQI readings should be used for relative issues within the the same network, ie, if you move a device to a different location then does the LQI go down or up.

2 Likes

No, it’s fully supported and there is nothing to wear out…
If you were having problems with your mesh before, that likely won’t change with a different coordinator…

The connections come from the mesh for the most part, not the coordinator stick.

1 Like

I guess he meant the stick’s life itself was reaching it’s end with frequent disconnections and not the end of software support.
Whatever, you are right in regards to an insufficient mesh without enough router devices might be the most common cause for disconnections.