Olin Lathrop wrote:
>
> > Testing the open circuit panel volts only takes a few
> > mS, you can then do the PWM for a few seconds so power
> > lost in the test is almost nothing. Remember it charges
> > the large input cap during the open circuit voltage
> > test so much of that energy is not lost anyway if you
> > keep the test time short enough, ie once that cap voltage
> > rises and stabilises the test is done.
>
> I think you *don't* want the cap there.  It will slow down the rise to
> open circuit voltage and hold the panel at inefficient voltages for too
> long.  Yes, the energy produced by the panel during the test is captured,
> but overall energy production is diminished.
>
> Instead, I would completely open circuit the panel (no cap) for very short
> periods occasionally.  Even if it taks a whole millisecond every second,
> only .1% of the available power is lost.  Actually it should be able to
> stabalize in much less time than that, so there would be no power loss in
> a practical sense.


Sure, that's a good point, but there are two issues;
* simplicity would be to just turn off the main PFET
  buck device and leave the cap across the panel,
  this saves having another PFET before the cap.
* I don't think solar panels really like going from
  full current to open circuit in such a constant and
  brutal fasion...

Since the input filter cap only needs to be effective
at 20kHz or higher, I think a value could be chosen that
allows the panel to rise to full OC voltage within 1mS
or so. Generally the MPPT point is at a fairly high
voltage and the OC voltage is only 20% higher or so.
So the time to rise from MPPT volts to OC volts is not
as bad as you might first imagine.

The micro can cut the buck PWM, then measure the cap
voltage as it quickly rises, then when it stabilises
within a % region the OC voltage test is done and the
buck PWM can continue. So for reliability's sake the
panel load goes through a more gentle transition,
while at the same time the test is done AQAP. :o)
-Roman

--
http://www.piclist.com#nomail Going offline? Don't AutoReply us!
email listserv@mitvma.mit.edu with SET PICList DIGEST in the body