The net effect is that we're making 12V from the gpio, which means all the LED
current and switcher inefficiency is coming as current through that gpio. The
-measured LED current is around 7mA, so this is "safe" for the SOC, but leaves
-less engineering margin than we usually like.
+measured LED current is around 7mA, so this is "safe" for the SOC, but really
+not a great plan...
-Changing the circuit to be powered from v_lipo with the gpio hooked to shdn*
-and controlling a FET switch on the output might be smarter someday?
+Need to redesign this in the future to use a transistor/FET switch.