The methodology presented in this paper aims to characterize impairments shown by real devices which are usually neglected on standardized tests but that become very important in massive IoT scenarios. For instance, we have measured that real BLE scanners are not able to scan continuously even though they are configured to do so. Besides, we have also found and demonstrated that some manufacturers seem not to apply any backoff mechanism although it is mandatory. These two unexpected behaviors have a significant impact on the performance of massive wireless sensor networks based on BLE. So, it becomes necessary to characterize these and other impairments. The proposed tests are based on device current consumption measurements and their association with the information obtained from upper layers. We describe a new low-cost generic measurement setup and provide all the necessary data (configuration parameters, scripts, etc.) for applying the proposed methodology. As an example, we use it to profile the behavior of Bluetooth Low Energy devices. Furthermore, the proposed setup can also inspire researchers to characterize other wireless technology devices, like Wi-Fi, Zigbee, LoRa, etc.