> With a system like battery + solar panel the part that most affects
> efficiency is the one with the higher Ri, and that is the panel, so the
> mppt system should concentrate on that point, if you cannot or do not
want
> to cover both.  In practice you could measure the ac ripple on the input
> capacitors of the smpsu and scale it (knowing their esr or something) to
> obtain the required information.

Instead of measuring parameters that effect efficiency, why not just
measure the end result and tweak the solar panel voltage (or current
depending on how you want to look at it) to get the best result.  This way
you don't have to assume anything about the intervening mechanisms, and
you get optimal operation by definition.

For example, if the solar panel is to charge a battery, then the best end
result is the highest battery charge current.  Let's assume that you've
wired up the solar panel such that under useful illumination its voltage
always exceeds the battery voltage.  First you put a big fat capacitor
accross the solar panel, then put a buck converter between the capacitor
and the battery.  The capacitor is big enough so that the individual
pulses of charge going into the buck converter don't significantly alter
the capacitor voltage, and therefore the solar panel operating point.
However, in the aggregate, the duration and frequency of buck pulses
control the current draw, and therefore adjust the voltage on the
capacitor, which is also the solar panel operating voltage.

You really don't give a rat's butt what voltage, power, and current the
solar panel is running at as long as the battery charge current is maximum
for the given illumination.  There is also no need to know the solar panel
impedence, open circuit voltage at high noon in Arizon, it's voltage
versus current curves, or what day of the week it was made on.  All you
need to do is turn the buck converter up or down to get the maximum
desired output, which in this case is the battery charge current.  All the
solar panel characteristics and the buck converter ineficiencies will be
automatically adjusted to by Feedback Magic(tm).

The only tricky issue is that the buck converter to charge current
transfer is not monotonic.  It has a maximum point somewhere within the
range, and falls off on either side of that point.  This kind of
maximum-finding is poorly suited to analog feedback, but can easily be
handled by digital logic.  A small PIC with an A/D and PWM module running
at 32KHz should be able to handle the task just fine.


*****************************************************************
Embed Inc, embedded system specialists in Littleton Massachusetts
(978) 742-9014, http://www.embedinc.com

--
http://www.piclist.com hint: To leave the PICList
mailto:piclist-unsubscribe-request@mitvma.mit.edu