Using the FRDM-K64F board it is easy to perform current consumption measurements in various processor operating modes. Since the range of values is quite large it makes an interesting comparison too.
The binary for performing the test can be loaded from http://www.utasker.com/kinetis/FRDM-K64F.html
and there is some information about the LLWU wakeup inputs used at µTasker LLWU Support
The results are as follows (at 120MHz whenever running - RTC always operating)
- RUN mode (processor never uses and low power modes) 50mA
- WAIT mode (uses WAIT whenever there is nothing else to do) 27mA
- STOP mode (uses STOP mode whenever there is nothing else to do but doesn't stop the PLL) 3.3mA
- VLPS mode (uses very low power stop whenever there is nothing to do but wakes up on timer or UART events) around 300uA (with peaks during short running phases)
- LLS (moves to low-leakage power mode and waits for a switch to be pressed) 270uA - [SW2 temporarily wakes frrm the mode to print a message and then falls back to LLS / SW3 causes the sytem to move to the WAIT operation again ]
- VLLS0 (moves to very low-leakage 0 mode and waits for a reset to restart operation) - 116uA
- VLLS1 (moves to very low-leakage 0 mode and waits for a reset to restart operation) - 270uA
- VLLS2 (moves to very low-leakage 0 mode and waits for a reset to restart operation) - 270uA
- VLLS3 (moves to very low-leakage 0 mode and waits for a reset to restart operation) - 124uA
My multimeter has problems with measuring in the uA range when there are short high peaks during the regular RUN period in VLPS so that value is not particulary accurate (guestimation).
I haven't analyed to be sure that the VLLSx modes are really correct or as expected but the results may prove of interest as rough comparison/starting point - also for other devices which should be able to achieve much lower values.