_TZE204_hlx9tnzb report flood and route flap after reboot

What happened?

Hello. I have some _TZE204_hlx9tnzb ZS-SR-EUD-1 dimmers. When the dimmers are already paired and experience a simultaneous power cycle (e.g., after a power outage), they start an endless message storm and topology change.

This has been tested and reproduces identically on both TI CC2652 (zstack) and EFR32MG24 (ember) coordinators

Short logs:
master_bedroom_damir_dimmer— linkquality: 104 → 96 (22.07.2026, 22:03:04)
master_bedroom_damir_dimmer— linkquality: 92 → 104 (22.07.2026, 22:03:04)
master_bedroom_diana_dimmer— linkquality: 68 → 76 (22.07.2026, 22:03:04)
master_bedroom_diana_dimmer— linkquality: 72 → 68 (22.07.2026, 22:03:03)
master_bedroom_damir_dimmer— linkquality: 96 → 92 (22.07.2026, 22:03:02)
master_bedroom_diana_dimmer— linkquality: 84 → 72 (22.07.2026, 22:03:02)
master_bedroom_damir_dimmer— linkquality: 92 → 96 (22.07.2026, 22:03:01)
master_bedroom_damir_dimmer— linkquality: 88 → 92 (22.07.2026, 22:03:00)
master_bedroom_diana_dimmer— linkquality: 64 → 84 (22.07.2026, 22:03:00)
stairs_dimmer — linkquality: 180 → 176 (22.07.2026, 22:02:59)
master_bedroom_diana_dimmer— linkquality: 76 → 88 (22.07.2026, 22:02:58)
stairs_dimmer — linkquality: 172 → 180 (22.07.2026, 22:02:57)
master_bedroom_damir_dimmer— linkquality: 88 → 92 (22.07.2026, 22:02:56)
master_bedroom_diana_dimmer— linkquality: 80 → 76 (22.07.2026, 22:02:55)
stairs_dimmer — linkquality: 176 → 172 (22.07.2026, 22:02:54)
master_bedroom_damir_dimmer— linkquality: 92 → 88 (22.07.2026, 22:02:53)
master_bedroom_diana_dimmer— linkquality: 88 → 80 (22.07.2026, 22:02:52)
master_bedroom_damir_dimmer— linkquality: 80 → 84 (22.07.2026, 22:02:50)
stairs_dimmer — linkquality: 172 → 176 (22.07.2026, 22:02:49)

I made a test instance of zstack+z2mtest (ember+z2m is prod) with only one dimmer and issue still persist. New join - ok, reboot - not ok.

I wrote to MOES:
I am very sorry, but our technical team has not yet conducted in-depth research on the Home Assistant platform,
so we may be unable to resolve this issue for the time being.

They provided me some useless doc.

Also i tried with no success
[12, null, tuya.valueConverter.raw],
[15, null, tuya.valueConverter.raw],
[17, null, tuya.valueConverter.raw],
[19, null, tuya.valueConverter.raw],
[20, null, tuya.valueConverter.raw],

Can you help me please?

What did you expect to happen?

No response

How to reproduce it (minimal and precise)

Join, power cycle

Zigbee2MQTT version

2.13.0 (unknown)

Adapter firmware version

{"maintrel":1,"majorrel":2,"minorrel":7,"product":1,"revision":20230507,"transportrev":2}

Adapter

ZStack3x0

Setup

os: #1 SMP PREEMPT_DYNAMIC PMX 7.0.12-1 (2026-06-09T21:07Z) - 7.0.12-1-pve - x64
node: v24.19.0
ha: false

Device database.db entry

No response

Debug log

ok_after_rejoin.txt
spam_after_device_reboot.txt

protocol_hlx9tnzb_20260811.pdf

Notes

Last health check

time: Tue Aug 11 2026 13:53:42 GMT+0300
os.load_average: 1.8, 28.85, 52.62
os.memory_percent: 19.9125
process.memory_percent: 16.423
process.uptime_sec: 611

Start by comparing ok_after_rejoin.txt with spam_after_device_reboot.txt and reviewing protocol_hlx9tnzb_20260811.pdf. Reproduce the issue with the ZStack3x0 setup by joining the dimmer and power-cycling it; done means the reboot no longer causes an endless message storm or topology changes.