<![CDATA[Energy efficiency is a critical requirement for battery-powered Internet of Things (IoT) systems, particularly those operating under duty-cycled conditions. Among widely adopted microcontroller platforms, ESP32 modules are extensively used due to their integrated connectivity and low-power features. In practice, ESP32 WROOM is commonly deployed because of its affordability and availability, while alternative modules such as ESP32 DFRobot are claimed to offer superior low-power performance. However, quantitative experimental comparisons at the module level remain limited. This study presents a controlled experimental evaluation of power consumption characteristics of ESP32 WROOM and ESP32 DFRobot modules operating in normal mode and deep sleep mode under realistic agricultural IoT workloads. Both modules were integrated with multiple environmental and soil sensors and LoRa communication, using identical hardware configurations, firmware logic, and measurement procedures. Power consumption was measured using a dual digital multimeter setup, with each operating condition evaluated over 50 repeated trials. The results show that both modules exhibit comparable power consumption during normal mode operation. In contrast, significant differences emerge during deep sleep mode. ESP32 WROOM consumes 36.907 mW in deep sleep, while ESP32 DFRobot consumes only 0.317 mW. Quantitative analysis indicates that ESP32 DFRobot achieves a deep sleep power efficiency improvement of approximately 99.14% relative to ESP32 WROOM. These findings demonstrate that module-level hardware design plays a decisive role in ultra-low-power performance and provide empirical guidance for selecting ESP32 modules in duty-cycled IoT deployments with significant implications for battery lifetime.]]>
