Yea I actually had a pretty stable ZHA setup for over 2 years on the sonoff dongle-P, I cant pinpoint what happened actually. I might have expanded too much or some update broke it. I decided then to get an Ethernet coordinator when I read some glowing reviews about them.
With the UZG-01 in hand I migrated slowly to Z2M. I have the Z2M and MQTT in its own docker container, it is not running with Home Assistant. However experience hasnt been great with the UZG-01.
I ordered an SMLIGHT SLZB-06 and 2 of those arrived as duds, so I had to get support to send me a 3rd one which finally works. In the meantime I got another UZG-01 in the hopes the first was a dud.
At first the new UZG-01 was quite stable for 4-5 days, however on the new firmware which keeps updating every other day for one bug fix or another (there should be release cycles), I cant seem to find a stable release. With the latest firmware, it lasts 4 days compared to the older UZG-01 which lasts 2 days before I have to hardware restart it.
I gave up on these for now, paper weights really now. Not blaming you @mercenaruss, seems like you are a one man team so I can understand not easy to support but based on open issues in git and on this forum, there seems to be a serious issue either with the zigbee chip or something else that is causing hangups.
I am using the good SLZB-06 now, so far 2 days no reboots and signal is stronger compared to the UZG-01 so plus there.
Cant recommend these at the moment, but I hope to revisit them in a couple of months, I am hoping it is a firmware and not a hardware issue. Shame to leave them as paper weights and hoping I can recommend these in the future once the firmware is stable.
Currently we have a single developer and is Koenkk,he is well aware about problems with firmware for Zigbee chip.
He is trying to solve it: HERE ,it’s the same for all people who are using Tuya devices,is a problem on SDK from Texas Instruments,we simple people cant do nothing about this. P7 is working only on last SDK, P2 is still working with old SDK and is kinda stable. I was thinking many time to go back to P2 chip and forget about all this,but i see things are moving forward.
I understand perfectly the problem,you can see my upper post.
From other hand we have firmware on this devices:
ESP32 firmware,what is answering for LAN and Bridge connection.
We moved to unified firmware XZG because we was able to add there OTA upgrade for Zigbee.
Long story short: SMLight forked our opensource firmware and of course worked on it,but closed the source files when they added OTA.So thi why we have now so active development.
CC2652P7 firmware - what is not stable now,because of issue mentioned everywhere.
I still hope issue with P7 will be solved soon by Koenkk or Texas Instruments.
Currently i recommend to all - upgrade to XZG - there you can try 1-3 zigbee firmwares, for many people 1 of 3 zigbee firmwares solved the problem they had,but all resumes to Tuya devices.
Hello @all , today was discovered some strange things on some boards with XZG firmware.
People who dont have network coming up please follow next procedure:
yes I am already on the firmware 20240316 but still having the issues with hangups, not sure if related to the P7 chip or ESP32, weird thing is I have 2 of these UZG-01 but they behave differently. One hangs after 1.5 days the other after 4 days so behavior is inconsistent.
Anyhow I am going to remove my tuya stuff slowly and restest in a month once I am back from my holidays. I am hoping we have some more stable firmware then. Hoping have more positive results.
One thing I also notice that the LQI for the SMLIGHT SLZB-06 is much higher than UZG-01. Is it because the P7 chip still doesnt support +20dbi ?
Hello guys!
I am observing this thread since i got mine UZG-01, about 2 months ago.
since I connected it to my HA and replaced sonoff dongle (btw that migration went without any problems and with no need to reconnect anything!) i have 0 problems with the coordinator.
So i am very confused by the problems discussed in here and the explanation that it is a problem with P7 chip itself.
Granted… I am on a pretty old firmware. I like to jump on the latest, and greatest software immediately when it arrives, but reading all those issues i am afraid of doing so. Am i missing something? Am i just lucky or that old firmware is saving my butt, and I should definetally not upgrade any time soon?
Yeah,you are right more than 90% of my customers never had any issue at all,
but who have currently have one of two issues:
Problem with Zigbee firmware what is affected by SDK and Tuya devices.
Random dying LAN8720 chip,what make LAN network disappear( Planning to move on other chip what’s more stable) . This users should contact me over email for a replacement.
Agree XZG is in havy development - daily development and releases. But we need feedback on Discord and Telegram XZG groups/channels to make it universal.
Is working with know devices: SLZB-06,TubesZB,ZigStar old and new devices.
Firmware for ESP32 is stable with small issues what dont affect your zigbee network,but affects only Zigbee OTA.
So XZG is must go anyways soon.UZG-01 firmware will remain at current version 0.2.6.
Is there any timeline on getting this addressed? I just took delivery of my UZG to improve my zigbee network stability from a USB dongle, and was disappointed to learn here that the 20db talked about on the product page only applies to the older devices, so the UZG may well not actually improve things for me at all at only half power.
Very interesting,is the same chip,is the same config.
So you say that some devices was dropping constantly from UZG-01 and on SLZB06P7 they are stable?
Correct, at least for the 4-5 days since I’ve done the change. The problems started after I upgraded to the 20230507 firmware. I’ve tried all the versions after that as well. I’ve tried to get a hold of the older firmware without luck.
Devices would stay online for about 24 hours after I re-paired them, then start dropping off again. Random devices, one day a in-wall dimmer from Namron would drop off, the next day a Hue bulb. The most notorious ones were a Aqara vibration sensor, a Namron K4 switch (battery powered) and a IKEA repeater (I have two, both took turns dropping off). There is no pattern as to what devices would drop off, and some of them came back online after I switched the breakers off and on again.
I also had some crashes from time to time.
No such issues yet with the SLZB. So at the moment the UZG-01 is a paper weight here in my house.
Edit 24.06: This also appears with the SLZB-06p7 so the UZG-01 is not at fault.
After each firmware change is recommended to Erase NVRAM, so Z2M will restore proper backup to new firmware/device.
I answered to your ticket,will have a small remote session to identify the problem.
Like i said many times, i have a feeling that my decision to move to P7 chip was a mistake. Everything was more stable with P2.
I still have hope that this will be solved ASAP by Koenkk.
If no any movement to this direction till end of the month i will go back to P2 chip and forget all this nightmare with Tuya devices and route table.
Thought I would add my experiences:
I have a UZG-01 (p7 chip) and Z2M on my HA installation.
The UZG-01 was pretty unstable (ESP 0.26 and Zigbee 20230507). I run quite a lot of TUYA devices (24 of them and mainly smartplugs). Typical uptime was around 4 days before all devices started going offline and I had to restart the UZG-01 to bring them back.
Since then I have moved the UZG-01 to ethernet (was WiFi) and the ESP firmware to 20240602. Zigbee is still 20230507.
I have also changed the reporting/polling intervals on the TUYA smartplugs to be much less frequently as I was aware this could produce a lot of Zigbee traffic.
For devices that have a Polling option (BSD29_1, TS011F_plug_3) I set the ‘Measurement poll interval’ interval to 360 seconds.
For the devices that have Reporting options (TS011F_plug_1, TS011F_plug_3) I set the ‘Min rep interval’ to 360 seconds for all 3 Reporting options.
The UZG-01 has currently been up 10 days without errors. I do not pretend to know if the reliability problem is really fixed, or which of the above changes had a beneficial impact.
