>At roughly 3% duty cycle he will really have to put the power
>to the segments, is this really safe?

Provided the average duty cycle is not exceeded, and the maximum current
rating not exceeded then yes it is safe.

>I see how that can be used to overcome the drive power limitations
>of a PIC I/O pin.  What I don't see is how this really gains us
>anything.  To get a certain amount of brightness out of the display
>it's going to take a certain amount of "average" input power to the
>display regardless of whether we turn on one segment at a time, or
>all the lit segments in each display digit (or for that matter all
>the segments of all displays at once) .  Is the idea here to not exceed
>some total output limitation of the PIC? (i.e. even though each PIN is
>capable of 25mA, we can't drive seven pins at 25mA simultaneously
>obtaining 175mA output).  What am I not understanding here?

I think the idea is to be able to multiplex the display without having any
driver transistors for the common cathode or common anode connection. This
should be easily do-able with the modern hi-efficiency leds available, at
currents that the PIC pins can handle, even with a 3% duty cycle.

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics