Measuring battery life

Poor battery life is affecting the take-up of too many devices. One of the technical reasons holding back the uptake of Internet of Things (IoT) products is poor battery life. While power has always been crucial, from a hardware perspective these devices require layers of software to operate. This can mean that choosing the wrong protocol or a failure to take into account the impact of software updates can help to ruin the overall user experience. It offers an integrated method for performing automated measurements and testing, explains Samuelsson.

In low-power embedded systems the choice of SoC or MCU is perhaps the most significant, and semiconductor vendors have taken many approaches to saving power. Sometimes peripherals, coupled with a direct memory access (DMA) controller, move data to and from SRAM without the processor needing to be active. But from the microcontroller’s datasheet alone that is hard to assess, says Samuelsson. By providing control over the supply voltage while measuring the current consumption, various power source strategies can be evaluated. If the impact of sub-circuitry needs to be understood, the differential analogue input on the expansion connector, together with a small resistance in the supply line, allows the power impact to be clearly separated out from the total power consumption.

The second key element, according to Samuelsson is the serial data RX input of the expansion port. Serial communication data can be captured via this interface at rates of 9600 bits per second (bps) up to 4Mbps. Any logging messages output by the application are then time-stamped and synchronised with the other time-domain measurements. A real-time operating system (RTOS)-based application could, together with instrumented trace logging via a UART, be tracked as it switched between tasks and into and out of its idle-task and low-power modes. The final element is the single-ended analogue input that is available when differential-mode power measurements are not being made.