Skip to content

Utility-Meter Calibration

Utility meters decoded by SCM, ERT, and SCMplus protocols report a raw consumption counter, but the RF signal does not carry the counter's unit or scale. Different meters report in different granularities, so the integration cannot derive Energy-dashboard-ready values automatically.

Out of the box, consumption is a plain unitless total_increasing counter. To make it eligible for Home Assistant's Energy dashboard, calibrate the device.

Calibration Flow

Open Settings → Devices & Services → rtl_433 → Configure → Device settings, pick the meter, and set its consumption calibration.

Field Meaning
Commodity none, energy, gas, or water. This sets the sensor's device class. Choosing none clears calibration.
Base unit The unit the calibrated counter is expressed in, constrained to Home Assistant units for that commodity.
Scale Multiplier applied to the raw counter so the stored value is in the chosen base unit.

When the meter reports a MeterType or ert_type hint, the commodity is pre-filled from it. You can override it.

The calibration step for an energy meter, with a base-unit selector and a scale multiplier

Once calibrated, the consumption sensor gets a real device class, native unit, and state_class: total_increasing. You do not need to pick the display unit in the calibration flow: Home Assistant can convert convertible units in the entity settings.

Statistics Caveat

Changing commodity, base unit, or scale changes the sensor's native unit or device class. Home Assistant treats that as a non-convertible change for long term statistics.

The entity keeps its ID, but previous long-term statistics are orphaned. Calibrate intentionally, ideally once. Saving a calibration reloads the hub so the sensor is rebuilt with the new unit and class.

Model-Scoped Mappings

For models whose unit and scale are authoritatively known, a contributor can ship a model-scoped mapping in the device library so those meters work without per-device calibration.

The shipped library does not include speculative real-meter consumption mappings. A wrong scale would silently corrupt Energy data.