Test for high current draw during commissioning

[bigjosh]

There are several failure modes that result in the current draw being too high and this will make the batteries go dead prematurely. We should test for this defect before units go to customers.

The easiest way to test this would be to use the built in current measurements on the EZ-FET programming board. Alas, the tools to access this are closed source and the documentation is not usable.

We could make our own current sensor, but then there would need to be a way to integrate that into the programming hardware and too much work.

So here is an idea: during the "First Start" programming step, we wait for the power supply voltage to fall below a given voltage and then we check how long it takes to fall to another lower voltage. During this time between the two voltages, the rate of the decay is dependant on the total power being drawn from the 1uF decoupling capacitor. If we find the current draw was too high, then we can fail the unit on the next step.

I think the best way to implement this is to continuously check the voltage during the "First Start" page. While we are connected to the programmer, the voltage should be 3.3V. Once we disconnect from the programmer it will start falling. When we see it hit, say, 3V then we put up a new message like "A Check" and set an RTC for, say, 1 second and go to sleep. When we wake 1 second later we check the voltage again and if it is above some threshold (2.5V?) then we set a "power draw OK" flag for the next step. If we get to the next step (battery insertion) and that flag is not set, then we know either the voltage was too low, or that the draw was so high that we ran out of juice before we ever woke up.