# How do you test a solar panel in winter? - Page 3

Posted by mike on February 3, 2008, 4:40 am

Roland Mösl wrote:

I don't have an application? Haven't seen anything that costs less
than the energy it replaces.  But I'm always on the lookout for dirt
cheap solar panels.

I have done some consulting for radio amateurs with repeaters at
mountain sites.

--

Posted by =?iso-8859-1?Q?Roland_M=F6sl?= on February 3, 2008, 8:26 am

And this is exactly the type of application,
where not the peak performance at 1000 W/m² counts.

Here counts the best performance at cloudy weather conditions
to come through some cloudy weeks in December

--
Roland Mösl
http://car.pege.org  cars and traffic
http://live.pege.org  building and live
http://www.pege.org

Posted by spaco on February 1, 2008, 11:03 pm
I am glad you made this post!   One of the questions I have relates to
this issue.   I have heard that some panels do a lot better in cloudy
weather.   With the inexpensive panel I am testing, I get zero output on
a heavily cloudy day and maybe 10% of panel rating on a moderately
cloudy day.   If some panels had DOUBLE my 10% on the moderately cloudy
day, that still wouldn't amount to much.   Can some panels do a LOT
better than others or are they all in the 10% area?
I know there are a million degrees of "cloudy" so please go easy on me.

Pete Stanaitis
-------------------------------------------

Roland Mösl wrote:

Posted by mike on February 1, 2008, 11:19 pm
spaco wrote:

I live near the 45th parallel.  The published insolation numbers for
winter are a small fraction of the summer numbers on a SUNNY day.
On a cloudy winter day, forget it.

A typical solar cell puts out less than a volt maximum.  The power curve
looks like a mountain.  For each light level, there's a maximum power
point (voltage and current).  That sucks for fixed battery voltages.
That's why people use controllers that can sense that peak point and
convert the voltage output to what the battery wants and maximize
the power at the current the solar cell wants to output at its peak.

You've got less than a volt, so you hook a bunch of them in series.
If you don't use an active controller, you want to adjust the number
in series so that the current is maximized at the battery voltage for
the light you have available.  For low light, you want more cells in
series, but that gives you less current per unit area in bright light.
Then for more power, you replicate the whole shebang and put the
series-conected assembly in parallel with others until you run out
of space on a panel that you can manage easily.  I suspect there's
some optimum size that minimizes the total cost of the structure.

You want the system to fail softly.  So, the diodes isolate different
parts so that if one cell fails, it won't drag the others down.
Or if a tree blocks part of the panel, you don't want it to
drag the others down too.  It's like car insurance.  You pay a little,
the diode, to insure against an accident, dead solar cell.

As always, get a second opinion.  By largest solar panel is 6" x 6".
I'm hardly an expert.

mike

--