When using the CC3000 Data Logger GUI, if you just leave it alone to broadcast some logged temp data every 20 seconds, the radio power plot on the GUI will look like this:
If you are careful, you can measure that the bases of those little power blips are 2.8 seconds wide and that they peak at 128mW. With a period of 20 seconds, that yields an average radio power of 9mW! ( .5 * 2.8 * 128 / 20) Pretty great for a WiFi gizmo!
Unfortunately, that seems to be pretty inaccurate. If you remove the "RF PWR" jumper from the EVM board and install an ammeter, you get a plot that looks like this:
Using a very conservative model of 20% of the time at 10mA and 80% of the time at 100mA, we can see an average current of 82mA going to the radio, which at the 3.6V supplied by the Experimenter Board, yields an average power of about 300mW. At the time that the GUI estimated 9mW average power, the reality appears to be more than 30x as much. Something to keep in mind in case your application is similar to this demo application.
It is worthwhile to note that while streaming accelerometer data, the power plot on the GUI is fairly accurate.
