Abstract:
In a multi-sensing, wirelessly communicating learning thermostat that uses power-harvesting to charge an internal power source, methods are disclosed for ensuring that the battery does not become depleted or damaged while at the same time ensuring selected levels of thermostat functionality. Charge status is monitored to determine whether the present rate of power usage needs to be stemmed. If the present rate of power usage needs to be stemmed, then a progression of performance levels and/or functionalities can be scaled back according to a predetermined progressive power conservation algorithm. In one embodiment, a wake-on-proximity function that activates a user interface based on readings from the proximity sensor may be altered while still allowing a HVAC control circuitry to operate as normal.