piclist 1999\03\07\160114a >
www.piclist.com/techref/index.htm?key=solar+panels
BY : Harold Hallikainen email (remove spam text)

On Sun, 7 Mar 1999 12:32:36 -0500 Mike Keitz <mkeitzJUNO.COM> writes:
>
>Exactly how does it work? Do you just adjust the PWM slightly and see
>if  the battery current increases or decreases, or is measurement of the

>solar panel voltage involved too?  Can you just regulate the panel
>voltage to some level and get maximum power out, or does the optimum
>panel voltage vary widely under different light levels?  Measuring the
>battery current seems like the best way since that's really what
>you're after, but it may be hard to do without losing some power in a
>resistor.
>I've heard of these power trackers, but never a description of
>exactly
>what method they use to operate at the maximum power point.
>

You've got it!  The idea is just to adjust the PWM and see if the
battery current goes up or down.  the solar panel voltage does not play a
part in it.  We could measure the panel voltage and current, then adjust
PWM for the maximum of the product, but measuring the output current
seems much simpler.
I see the Thevenin equivalent of a solar panel as a variable voltage
source and a variable source resistance, both of which vary with the
amount of light falling on the panel.  The converter adjusts its load on
the panel to match the source resistance, giving us the most output
whatever, we just adjust for what we want:  maximum output power.  Since
the output power always increases with increasing current into the
batteries, it's pretty simple to measure.  Of course measuring the solar
panel current, we'd find a maximum current when we got zero output power
(driving a short), or maximum voltage when we got zero output power
(driving an open).
Sensing the output current efficiently might be a bit of a problem.  We
don't really need to know the absolute current, just whether it's going
up or down.  So, we could measure the voltage across something where we
are already getting a voltage drop, such as the output diode in the step
up converter.  Another idea might be use of a Hall effect sensor, or
maybe a SenseFET.  So far I've just told the students "Here are some
ideas to play with."  I haven't had to get any of it to work!
They did bring me a circuit someone had developed with LOTS OF PARTS!
It sampled the input voltage and current, multiplied the two (in analog
circuitry) and adjusted a standard switching regulator chip.

Harold
___________________________________________________________________