I’m running Home Assistant on a Raspberry Pi 5 with a memory card, configured with the ZHA integration. My Zigbee coordinator is a Sonoff Zigbee 3.0 USB Dongle Plus v2, and I have 39 devices connected to it, including 10 routers spread across different rooms. The coordinator is placed away from any interference sources, and the apartment is about 100 square meters with mostly brick walls.
The issue I’m facing is with low LQI values across the network, even though most devices are quite close to each other. Additionally, there are noticeable delays in many of my automations. My automations are organized into scenes triggered by Zigbee buttons and standard Zigbee switches.
During troubleshooting, I observed high channel utilization. Initially, I was on channel 25, which consistently showed usage over 98%, as flagged by log warnings. I monitored other channels over a few days and found that channel 11 had lower utilization, around 3-5%. Thinking this would help, I switched the entire network to channel 11. However, this didn’t resolve the problem; instead, channel 11 quickly spiked to 95% utilization, while channel 25 dropped to 5-10%.
Has anyone encountered a similar issue or have suggestions on potential solutions? Any insights on optimizing the Zigbee network in this setup would be greatly appreciated.
I use default Pi power supply.
In the logs, I can see the specific devices where the delays occur. For example, in a scene that turns off all the lights, some switches experience delays of 15-20 seconds. Interestingly, this often happens to a switch that is just over a meter away from the coordinator’s antenna. It seems more likely that the problem is with channel utilization. Currently, I’m on channel 11, and here’s the current situation:
“energy_scan”: {
“11”: 98.93395819824867,
“12”: 43.057636198227904,
“13”: 19.00785284282869,
“14”: 9.713248103580147,
“15”: 62.257682586134884,
“16”: 46.26944564832987,
“17”: 46.26944564832987,
“18”: 2.2107128772756957,
“19”: 46.26944564832987,
“20”: 10.914542804728702,
“21”: 21.09014924761344,
“22”: 6.011489450827149,
“23”: 10.914542804728702,
“24”: 39.90320178295578,
“25”: 3.6632469452765037,
“26”: 78.25348754651363
Forget about the energy scan for a minute - you said you already moved your channels and that the utilization moved with them.
Check your logs and the automation traces for anything which stands out, and paste one of your problematic automations (and its trace) here.
I know there have been reports of lights not being turned on together when you switch on/off a bunch of them at once instead of using zigbee groups, so it could be that. However, the traces will probably paint a clearer picture
I’ve never had much luck turning on many (>5) lights at once. I tried putting them in zigbee groups (not HA groups) and that helped some. So from my experience, long delays getting 10+ devices to change at the same time is not surprising. For the automations I have that need to change many devices, I’ve ended up breaking them up into separate steps and putting in delays between them to get them to work.
Hi there, any luck with solution? I guess I am in the same situation.
I am superb new with HA.
Bought a HA Green and Sky Connect. Swapped out all my wifi switches to Zigbee and jump from Smartlife into direct ZHA. Most if not all my devices - approximately 48 devices (11 routers - heater switches / smart plugs) direct ZHA.
Setup approximately 50-60 automations mainly motion / presence / door / Reolink Detection as sensors to trigger lights on / off and other automation.
Thus far looking great except some automations especially trigger on / off lights tend to have massive delay. The delay may take up to few secs to minutes. I am pretty sure my 2.4 WIFI is not interfering my Zigbee which is in Channel 25.
Not sure if adding more smart plugs as routers or just add in the SONOFF DONGLE P in as router will help?
Hi there, I haven’t reached a final solution yet, but I can share what I’ve tried so far.
I added 3 additional router devices (smart plugs), which had a short-term positive effect. For a few hours after restarting HA, everything works without noticeable delays. However, after some time, delays return and can range from several seconds to a couple of minutes.
When I run a Trace on any automation right after a restart, it completes in 5-6 seconds. Running the same trace a few hours later shows delays of 1-2 minutes.
I’m currently testing two potential solutions:
I ordered an SLZB-06 Zigbee coordinator, which connects via the network rather than USB. My hope is that this might alleviate some bottlenecks or improve overall stability. Unfortunately, I haven’t received it yet to confirm if it helps.
Since my HA is running on a Raspberry Pi 5 with an MMC card, I decided to replace the card with an NVMe drive. However, this change introduced its own set of issues, which I’ll address in a separate question as it’s a more complex topic.
If you’ve tried similar steps or have additional insights, I’d love to hear them!
