Developed with Plugwise Emma in mind, but universally usable with any Home Assistant climate entity that exposes
current_temperatureandtemperatureattributes.
This release is intentionally built from the stable v1.2.1 baseline and only includes the rounding improvements from PR #8 (rounding_direction).
Unwanted functionality introduced in newer lines is intentionally excluded from this recovery patch.
ClimateSync is a HACS-ready Home Assistant custom integration that implements delta-based thermostat synchronisation. It reads the heating demand (delta between target and current temperature) from multiple source climate entities (rooms) and drives a single destination thermostat by continuously adjusting its target temperature.
Instead of simply syncing the highest source setpoint to the destination, ClimateSync uses the maximum delta (largest gap between target and current temperature across all rooms) and adds it to the destination's own current temperature:
destination_target = destination_current_temperature + delta_max
This approach was designed with the Plugwise Emma in mind. The Emma is connected to the central-heating boiler (CV) over OpenTherm and internally calculates a desired CV water temperature based on its own current delta. By feeding the Emma the home-wide maximum delta on top of its own measured temperature, ClimateSync lets the Emma function as a full OpenTherm controller — it sees the "hardest working" room's demand and adjusts the boiler modulation accordingly.
Even without an Emma, this delta-based method is more accurate than copying setpoints because it accounts for how far the destination already is from equilibrium.
- Event-driven updates — reacts immediately to temperature changes.
- Periodic resync (configurable, default 60 s) to recover from missed events.
- Anti-flap: only sends
climate.set_temperaturewhen the change exceeds a configurable threshold (default 0.2 °C). - Rate limiting: maximum one service call per 10 seconds (configurable).
- Rich diagnostic sensors including a
sensor.climatesync_statusthat makes desyncs visible. - No controllable entities — all control is internal via
climate.set_temperature.
- Open HACS → Integrations → ⋮ → Custom repositories.
- Add
https://github.com/Patrick1610/ClimateSyncas an Integration. - Search for ClimateSync and install.
- Restart Home Assistant.
- Copy
custom_components/climatesync/to your<config>/custom_components/directory. - Restart Home Assistant.
Navigate to Settings → Devices & Services → Add Integration → ClimateSync.
Select one or more climate entities that represent the rooms whose heating demand should be tracked. Each selected entity must expose current_temperature and temperature attributes.
| Field | Default | Description |
|---|---|---|
| Destination climate entity | — | The thermostat that ClimateSync will control. |
| Idle temperature | 5.0 °C | Target temperature sent to the destination when no room has a positive delta (all rooms are at or above their target). |
| Maximum setpoint | 35.0 °C | Hard ceiling for the destination setpoint. |
| Rounding mode | 1 decimal | Step size used for the computed setpoint. |
| Rounding direction | nearest | Whether the computed setpoint is rounded down, normally, or up within the selected rounding mode. |
After setup, open the integration → Configure (⚙ gear icon) to get the same 2-step wizard again. You can change:
- Step 1: add or remove source rooms
- Step 2: change the destination thermostat, idle temperature, maximum setpoint, rounding mode, rounding direction, and advanced options:
| Option | Default | Description |
|---|---|---|
| Destination thermostat | — | Change which thermostat is controlled. |
| Idle temperature | 5.0 °C | Temperature sent when no room needs heating. |
| Maximum setpoint | 35.0 °C | Hard ceiling for the destination setpoint. |
| Rounding mode | 1 decimal | Step size used for setpoints. |
| Rounding direction | nearest | floor, nearest, or ceiling rounding within the selected mode. |
| Resync interval | 60 s | How often ClimateSync checks even without state changes. |
| Minimum change threshold | 0.2 °C | Only send a new setpoint if the change exceeds this. |
| Minimum send interval | 10 s | At most one service call per this many seconds. |
For each source climate entity (room):
current = current_temperature attribute
target = temperature attribute
delta = max(target - current, 0)
(if either attribute is missing/unavailable → delta = 0)
delta_max = max(all room deltas)
If delta_max <= 0:
setpoint_raw = idle_temperature # No room needs heating
Else:
setpoint_raw = destination_current_temperature + delta_max
# The destination target is set to its own current temperature
# plus the largest demand across all source rooms. This means the
# destination "feels" the same heating gap as the hardest-working
# room, which is critical for OpenTherm controllers like the
# Plugwise Emma that modulate boiler output based on their own
# observed delta.
rounded_setpoint = round_setpoint(setpoint_raw, rounding_mode, rounding_direction)
setpoint_final = min(rounded_setpoint, maximum_setpoint)
If abs(destination_current_target - setpoint_final) > min_change_threshold:
If time_since_last_call >= min_send_interval:
climate.set_temperature(destination, setpoint_final)
The rounding mode determines the step size. The rounding direction determines where the raw setpoint lands within that step size. ClimateSync applies epsilon-safe floor/nearest/ceiling rounding to avoid floating-point edge cases around step boundaries.
nearest is the default and preserves the historical ClimateSync behaviour for existing installations. If an older config entry does not contain rounding_direction, ClimateSync treats it as nearest.
| Mode | Example input | Result |
|---|---|---|
0.5 steps |
19.3 | 19.5 |
1 decimal |
19.33 | 19.3 |
2 decimals |
19.333 | 19.33 |
| Direction | Behaviour |
|---|---|
floor |
Always round down within the selected mode. |
nearest |
Round to the nearest step, matching the old behaviour. |
ceiling |
Always round up within the selected mode. |
Plugwise Emma examples:
| Mode + direction | Raw setpoint | Sent setpoint |
|---|---|---|
0.5 steps + ceiling |
19.1 | 19.5 |
0.5 steps + ceiling |
19.5 | 19.5 |
0.5 steps + ceiling |
19.6 | 20.0 |
0.5 steps + floor |
19.1 | 19.0 |
0.5 steps + floor |
19.9 | 19.5 |
1 decimal + ceiling |
19.11 | 19.2 |
1 decimal + floor |
19.19 | 19.1 |
For Plugwise Emma, 0.5 steps can be useful because Emma commonly accepts half-degree setpoints. Combining 0.5 steps with ceiling can help when small RoomMind deltas would otherwise disappear through normal rounding, for example a raw setpoint of 19.1 becoming 19.5 instead of 19.0.
All entities are attached to a ClimateSync device. Sensors (setpoint, deltas, destination target) are regular entities; the status sensor is classified as diagnostic.
| Attribute | Description |
|---|---|
destination_entity_id |
The controlled thermostat |
destination_current_temperature |
Current measured temperature at destination |
destination_current_target |
Current target temperature at destination |
delta_max |
Max delta used for this computation |
rounding_mode |
Active rounding mode |
rounding_direction |
Active rounding direction |
raw_setpoint |
Unrounded destination_current_temperature + delta_max, or idle temperature when no room needs heating |
rounded_setpoint |
Raw setpoint after rounding mode and rounding direction |
final_setpoint |
Final setpoint after rounding and maximum-setpoint clamp |
idle_temperature |
Configured idle temperature |
State: the final setpoint that ClimateSync wants to apply (destination_current_temperature + delta_max, rounded and capped by maximum setpoint).
| Attribute | Description |
|---|---|
room_deltas |
Map of {entity_id: delta} for all rooms |
leading_room |
Entity id of the room with the highest delta |
State: the maximum delta across all rooms.
One sensor per source climate entity.
| Attribute | Description |
|---|---|
source_entity_id |
The climate entity this sensor tracks |
current_temperature |
Last known current temperature |
target_temperature |
Last known target temperature |
raw_delta |
target - current (may be negative) |
State: max(raw_delta, 0) — the effective heating demand for this room.
Shows the destination thermostat's actual current target temperature in real time, making it easy to compare against the computed setpoint without switching to the destination device.
| Attribute | Description |
|---|---|
destination_entity_id |
The controlled thermostat |
destination_current_temperature |
Current measured temperature at destination |
State: the destination's current temperature attribute (its active target).
States:
| State | Meaning |
|---|---|
ok |
Everything is in sync, no issues. |
rate_limited |
A setpoint update was suppressed because the last call was too recent. |
destination_unavailable |
The destination climate entity is unavailable or unknown. |
missing_source_data |
One or more source entities have missing/unavailable temperature attributes. The integration continues with delta = 0 for those rooms. |
apply_failed |
The climate.set_temperature service call threw an exception. Check last_error. |
mismatch |
The destination's actual target deviates from the desired setpoint beyond the threshold. ClimateSync will attempt to correct this on the next cycle. |
Attributes:
| Attribute | Description |
|---|---|
last_update_time |
ISO timestamp of the last evaluation |
last_service_call_time |
ISO timestamp of the last climate.set_temperature call |
last_desired_setpoint |
What ClimateSync computed as the ideal setpoint |
last_applied_setpoint |
What was last actually sent to the destination |
current_destination_target |
The destination's actual temperature attribute right now |
rounding_mode |
Active rounding mode |
rounding_direction |
Active rounding direction |
raw_setpoint |
Last unrounded setpoint |
rounded_setpoint |
Last rounded setpoint before maximum-setpoint clamp |
final_setpoint |
Last final setpoint after rounding and clamp |
mismatch_seconds |
How long (seconds) the desired and actual setpoint have been diverging |
mismatch_since |
ISO timestamp of when the mismatch started (null when in sync) |
resync_count |
Number of periodic resyncs since startup |
apply_attempts |
Total service call attempts since startup |
apply_failures |
Total service call failures since startup |
evaluation_count |
Total evaluation cycles since startup |
skipped_anti_flap |
Times a setpoint update was skipped because the change was within the threshold |
skipped_rate_limit |
Times a setpoint update was skipped due to rate limiting |
last_error |
Last exception message, if any |
Some thermostats (e.g. Plugwise Emma) only accept specific temperature steps. Use the 0.5 steps rounding mode in that case. If small RoomMind deltas are rounded away, use Rounding direction = ceiling so a raw setpoint such as 19.1 is sent as 19.5 instead of 19.0.
An external automation or the user may be overriding the destination's target. ClimateSync will keep trying to reapply on every evaluation cycle and resync. Check if another integration or automation is fighting over the thermostat.
The source rooms are changing temperature very rapidly. Increase min_send_interval in the Options Flow to reduce chatter.
One or more source climate entities are offline or do not expose current_temperature / temperature attributes. ClimateSync treats those rooms as delta = 0 and continues.
The destination thermostat is offline. No service calls are made. ClimateSync will recover automatically once the entity becomes available again.
Check last_error in the sensor.climatesync_status attributes. Most likely the climate entity does not support the climate.set_temperature service or the entity id is wrong.
ClimateSync uses only the standard climate.set_temperature service and reads standard climate entity attributes (current_temperature, temperature). It works with any climate integration that follows the standard HA climate platform contract, including but not limited to:
- Plugwise Emma / Smile
- Generic Thermostat
- ESPHome climate components
- Z-Wave thermostats
- Zigbee thermostats (ZHA / Zigbee2MQTT)
- Google Nest (via the Nest integration)
MIT — see LICENSE.