Posted by ohara5.0 on June 27, 2008, 2:49 am
The thread on buying Si wafers got me thinking about some useful
info, or at least useful to somebody:
Si wafers can be polished to a roughness of less than 8 angstroms rms
although the very thin ones are rougher.
You can buy ultra-thin mica, as thin as 12 microns in sheets as large
as 3" X 3" for very cheap and it is transparent and essentially
smoother than 3 angstroms rms. It is ultra-flexible.
Ultra-thin flat glass is available, as thin as 5 microns and very
flexible, roughness less than 5 angstroms rms but wavy.
Large sheets of plastic can be lacquer coated and then coated with
gold or silver or Al to be insanely smooth although wavy.
A friend of mine has tested solar cells to find that they are much
more efficient at high temps. This means that cheapo concentrators
enable them to work at higher efficiencies.
The efficiency of a parabola for concentration of sunlight can be
improved on by using a shape called a compound parabolic concentrator
(Google "CPC, OptIcs" or "Winston Concentrators"". These can be made
from reflective mylar.
Posted by coj on June 27, 2008, 6:57 am
OK, I stand corrected on buying solar Si wafers.
Solar cells aren't usually polished to be very smooth, usually they will
have some texturing to roughen the surface (often in pyramids) so that
light reflected off the surface might hit another pyramid and be absorbed.
Solar cells peform *worse* at higher temperatures, check the gamma
derating factor on spec sheets which is defined as 1/Pmax * dPmax/dT and
is usually between -0.2 t0 -0.45%/K.
The design of the front grid of a cell is an optimum between resistance
in the fingers/tabbing and the shading, concentrating light onto a cell
that hasn't been designed to take the extra current will result in
series resistance loss.
Apart from reflections, most photons not producing current will end up
heating the cell. For example if a cell is 15% efficient, almost 85% of
the light input heats the cell. Look for the definition of NOCT which
shows how modules at 20C ambient and 800W/m2 usually end up around 47C.
If you did a 3x concentrator onto a module not designed for it then the
temperature would get up to around 100C,above what the solders and
plastic coatings could reliably take.
Concentrators only use the direct light (so aren't any good in cloudy
skies), need to be tracked (more expense) and need good thermal sinks to
stop the heat frying the cells. You might get a tracking gain of 35-40%
insolation at a sunny site and this would be seen by a non concentrator
cell on a tracker.
Certainly it can be done (and is done at many bright locations) but it
isn't just a simple case of adding lenses and mirrors to ordinary cells.
You don't need ultra smooth reflectors for concentrators as you are
concentrating the sun's disc onto a finite size. Maximum reflectivity
and low degradation of the material plus ability to be cleaned is more
Posted by Eeyore on June 27, 2008, 2:01 pm
This is the opposite of the currently received wisdom.
Posted by ohara5.0 on June 27, 2008, 11:55 pm
It is possible that they were specially configured for high temps. He
did seem to get very high performance out of them with concentrated
sunlight at high temps.\
I agree, roughness of materials is not important for solar
concentrators but may be for solar sensors.
Posted by LU Wei on June 29, 2008, 3:05 pm
firstname.lastname@example.org wrote on 2008-6-28 7:55:
It's not possible -- high temperature will exponentially decrease the
minority carrier's life. The high performance is resulted from the
concentrated light, not the high temperature -- I think if he could
control the temperature the performance will be even better.