Hi,

I’m looking for engineering-level input on an ongoing stability issue with a Deco XE75 mesh that persists despite extensive configuration changes and testing.

This is not a basic setup and I’ve worked systematically through the usual remedies.

 

 

Environment

• Model: Deco XE75 (EU, V1)

• Firmware: 1.4.3 (auto-updated overnight; previously on 1.4.1)

• Nodes: 4 × XE75

• Backhaul: Wireless only (Cat6 not feasible in this property)

• WAN: FTTP, stable, no ONT issues

• House: UK semi-detached, heavy brick internal walls

• Clients: Mixed (phones, laptops, Sonos, ~10+ cameras/IoT)

 

 

Topology

• Main Deco (Living Room) wired to ONT

• Hallway Deco acts as the hub

• Landing + Bedroom connect via Hallway

• Connection Preference manually set to enforce this topology

Topology is stable when viewed in the app and all satellites typically show Medium or better signal.

 

 

What I have already done

To avoid the usual suggestions, here is what has already been tested and/or implemented:

• Disabled separate IoT SSID → single SSID broadcasting 2.4 / 5 / 6 GHz

• Reduced 5 GHz channel width to 80 MHz

• Confirmed 5 GHz remains on channel 36 (non-DFS)

• Weekly scheduled reboot enabled (03:00)

• Manual reboot tests performed

• Network optimisation largely avoided (except for controlled tests)

• Client roaming / mesh behaviour simplified

• One external extender used only for a single outdoor camera

• Firmware rollback tested previously (1.4.3 → 1.4.1) - stability issues persisted - so moved back to 1.4.3

 

 

Symptoms

• Network becomes unreachable (all clients affected)

• Requires manual reboot (power cycle) of the Deco system to recover

• Occurs intermittently (roughly weekly, sometimes more frequently)

• WAN itself does not appear to be dropping

• Mesh appears to collapse rather than degrade gracefully

 

 

Logs

Observations from the logs:

• No obvious WAN disconnects

• No clean reboot signatures at the time of failure

• Log buffer appears extremely short (often only ~10 minutes retained with no apparant way of changing this - really poor TP Link!)

• Difficult to capture the failure itself before logs roll over

This makes root-cause analysis challenging.

 

 

What I need help with

1. Is this a known stability issue with XE75 wireless backhaul in dense/brick environments?

2. Are there additional logging or debug modes that can be enabled to capture failures properly?

3. Is there guidance on further stabilising wireless backhaul, beyond what I’ve already done?

4. Is 1.4.3 known to address (or introduce) mesh stability issues compared to 1.4.1?

I’m technically comfortable and happy to run targeted tests if given specific instructions.

 

 

Closing

I’m at the point of deciding whether to:

• continue investing time in the Deco platform, or

• switch to a different vendor (e.g. UniFi)

Before doing so, I’d like to give TP-Link one last opportunity to help diagnose and resolve this properly.

Thanks in advance to anyone from TP-Link engineering or advanced support who can engage on this.