It is important to understand the effect of reading out and buffering the sensor data in ultra-low leakage RAM, which enables transmission of one much longer data packet with less protocol overhead. This is possible with both Standby and Sensor Controller solutions, as two dedicated RAM areas are available: one for the Arm Cortex-M4F and a smaller one for the SCE peripheral.

The duty-cycle approach cannot buffer the data, as it gets powered off and all RAM content is lost. Hence, the sensor data must be sent out wirelessly before the power-off occurs leading to the transmission of additional wireless overhead, which is at the beginning of each packet sent over-the-air. In this report the Sub-1 GHz protocol overhead consists of 4 bytes preamble (0101.. sequence) followed by 4 bytes synchronization pattern (also called SYNC word) and 2 bytes packet sequence number (or just one byte to code the packet length).

Assuming the total packet length is 250 bytes (including 10 bytes overhead), see Section 3.2, both the StandBy and Sensor Controller solutions can send the data of 60 humidity and temperature values collected in 1 hour, when the sensor readout happens once per minute. In contrast, the Duty-cycling approach will transmit redundantly 59 times the identical wireless protocol overhead of 9 or 10 bytes, with nearly the highest power consumption, as these bytes are sent over-the-air.

Figure 4-5 Power Efficiency due to Data Buffering and no Unnecessary Wireless Overhead

Obviously, the data buffering with longer data packets delivers additional savings, which amount to 30%, see Figure 4-5. This is a significant factor and must considered, when using the decision tree, see Figure 4-2. Depending on the amount of extra wireless overhead in a given application, the Duty-cycling solution may even require a much longer inactivity period than initially estimated to deliver the lowest power consumption.

Using the Sensor Controller Engine in the CC1352P MCU (and the CC13xx and CC26xx family devices) for sensor readout achieves equal or lower power consumption than using the main Arm Cortex-M4F MCU. If the application has long inactivity periods of many minutes or more, then Duty-cycling of the full system can deliver even lower power consumption with the addition of a nano-timer and load switch devices.