See the SRM section 6.4 "Current Measurement". Use a differential
scope probe or multimeter across the resistor mentioned there to take
measurements.
-Andrew
I hate to keep telling people to read the manual. It seems obvious to me, but maybe I am over looking something.
Gerald
I’m sorry I didn’t see that part. Is this the only way to do it? Is it possible to do it maybe using some software tools?
Do you mean SW running on the board? You can do that. But if you do that, you are testing the current of the SW that reads the current, not your algorithm
Gerald
Yes I mean like some linux tool that can measure power consumption. I found this utility called “PowerTOP” that seems to do this but I’m not sure if it’s working under Angstrom.
What I would like to do is for example measure how much power is consumed during a JPEG encoding of an image.
You could use PowerTOP or a similar tool but that may not give as
detailed information as using a scope. If you can toggle a GPIO at the
start of your algorithm's code, use the GPIO to trigger the scope, and
then toggle another GPIO (or the same one to the other state) to show
when the operation completes, you can get very detailed data on your
power consumption.
Or, use another beagle's A-D converters to do the measurement (but
please observe the voltage levels and do some isolation). Then run
your power consumption data collection routines on the other beagle.
A software tool that can sample quickly to give you detail will consume
power, leading to Gerald's comment. Using an external device will give
you the most accurate and precise data. A scope would be best, an
Arduino or another beagle's A-D would be second best, a multimeter 3rd.
-Andrew
The problem with doing it the instruction manual way is that it doesn't
truly provide what he's looking for - power consumption. Measuring amps
across the shunt is handy to know the max amperage the board draws but
is not useful to volts X amps will approximate peak power, depending on
how saggy the power supply is.
I'm going to assume that the OP is interested in power consumption
because he wants to run the board from batteries. Making that
measurement requires a power meter. Such as this one
http://www3.towerhobbies.com/cgi-bin/wti0001p?&I=LXBHDT&P=7
(Not endorsing the vendor, just the first one google returned).
I have an earlier version without the LiPo balancing. It is a very good
product and according to my lab standard watt-hour meter, very accurate.
It samples rapidly so it will catch the momentary high power draw of an
intense compute routine in the software.
While you're ordering I suggest getting several of the mating Jones
Super Plugs to make with the meter. They're the RC standard, rugged as
heck and vibration-proof.
John