Charging session stops itself after ~10–150 seconds, "Command rejected for DP 140: None" in the log

[bjoern477]

Environment

• Integration: Tuya EV Charger Local (custom_components/tuya_ev_charger)
• Device: dé Mobile Wallbox 11 kW, 3-phase
• Connection: local, protocol 3.5
• sensor.connection_health: 99.9–100% success rate throughout (no connectivity issue)

Summary

A charging session started via switch.charging_session (or by the charger itself on cable plug-in) reliably runs for anywhere between ~10 seconds and ~150 seconds, then self-terminates: sensor.status drops from charging to paused, sensor.total_power drops to 0.0, and the log shows Command rejected for DP 140: None. This happens whether the session was started by the integration's own switch.turn_on, by the charger's own auto-start, or right after a full power cycle of the charger.

Reproduction (same day, same charger, three independent resets in between)

Time	Event
06:48:13	switch.charging_session → on, status → charging, total_power → 0.371 kW
06:48:24	(11s later) status → paused, total_power → 0.0, switch → off (not commanded by any automation)
06:52:11	switch.charging_session → on again, status → charging, total_power → 1.459 kW
06:52:21	(10s later) same self-termination: status → paused, total_power → 0.0

Both stops happened with no automation, no user action, and no surplus/threshold logic involved — this was a manual on/off test with all our own automations disabled.

What we already ruled out

• Not caused by our own Home Assistant automations. We disabled every automation touching this entity before the test above; the stops are not requested by anything in HA.
• Not a connectivity issue. sensor.connection_health stayed at 99.9–100% the whole time; polling never lost the device.
• Not button.ready_to_charge-fixable. Pressing ready_to_charge before retrying switch.turn_on does not prevent the next rejection.
• Not button.reboot_charger-fixable. Same result after rebooting the charger via that button.
• Not fixed by a full physical power cycle of the charger (unplugged and reconnected).
• The only thing that reliably "unstuck" it (temporarily) was toggling the physical switch/button on the charger itself — which then allowed one more ~10s burst before the same failure recurred.

Log excerpt

ERROR (MainThread) [custom_components.tuya_ev_charger.tuya_ev_charger] Command rejected for DP 140: None
homeassistant.exceptions.HomeAssistantError: Unable to start charging session.
  File "/config/custom_components/tuya_ev_charger/switch.py", line 65, in async_turn_on
    await self._async_set_charging(True)
  File "/config/custom_components/tuya_ev_charger/switch.py", line 77, in _async_set_charging
    raise HomeAssistantError(...)

This error recurred identically 14+ times over a ~3 hour window the previous afternoon, and again this morning after every reset attempt listed above.

Additional context

• We also saw a genuinely successful, uninterrupted multi-hour charge on the same charger the previous afternoon (16:00–18:00, battery rose smoothly from 73% to 80% with no gaps), so the charger and installation are capable of sustained charging — this failure mode seems intermittent/state-dependent rather than a permanent hardware fault.
• sensor.status occasionally also gets stuck on unknown for several poll cycles (30–60s) after a reset, even though the vehicle is physically plugged in the whole time.
• Happy to run tuya_ev_charger.profile_assistant or share a DP dump while charging if that helps — let us know what's most useful to attach.

Question

Is DP140 (x_do_charge) known to be rejected while some other condition/DP isn't in the expected state (e.g. right after a charging attempt, before some handshake/ack completes)? Is there a recommended sequence (e.g. a specific DP to poll/reset) before retrying switch.turn_on after a self-stop like this?