Yup, it’s confirmed that the core clock divisor works as described. Running a while (1); at 24 MHz drops the current to about 60% of the current at 48 MHz. As always, running at 24 MHz means that it will take twice as long to perform the same set of instructions than running at 48 MHz, so while the peak current consumption will be lower at 24 MHz, the energy consumed to perform an identical set of instructions is actually higher at 24 MHz than 48 MHz.
Hey robin_hodgson, i tried this clock division with a freertos program and it works. So nice, again thank you for the information!
-Turo
Thank you adam and El_Duderino for the current measurements. I also want to voice agreement with Adam that a low power microcontroller would be a great buy. So much effort goes into making them small but then large batteries are needed because things like sleep current are high. A true low power artemis would be incredibly useful for so many projects
Also, a Hackaday project was recently able to get an idle LISP editor to around 2.64mW: https://hackaday.io/project/184340/log/215909-snake