- thassaj commented on DrewPaulDesigns's instructable TRIPLE YOUR SOLAR ARRAY'S OUTPUT!4 months ago
- thassaj commented on DrewPaulDesigns's instructable TRIPLE YOUR SOLAR ARRAY'S OUTPUT!4 months ago
OK - I did an approximate calculation, taking into account that a solar cell will pick up less light per unit area when it is at an angle to the sun, taking into account the reflectivity of silicon solar cells with no anti-reflection coatings as a function of angle (Optics Communications 172 (1999) 139-151 - A. Parretta, et, al), and taking account of the fact that for a smaller V angle you will pack in more solar cells in the same area. What I find is that for a V angle of 15 degrees (roughly what you used?) you should get roughly 40% more power per unit area than using a single solar cell taking up the same area. I assume the array is pointing directly at the sun. I'm unsure of your power figures, but I think there should be a measurable improvement, given all the assumptions above. If the cell has an antireflection coating then the above calculation won't be valid.Posted at 15:24 UK time (Instructables seems to be adding 10 hours).

I think this is apretty good idea: the main plus point is that as a solar cell doesn’t absorb100% of the photons with an energy above the semiconductor band gap energy, itmust be reflecting some of these photons back out and they are lost. It is thislost energy that can be partially recaptured in this configuration. By usingthe V shape, the photons of interest get a second chance to be absorbed by thepn junction and converted to electricity. And then of those still reflected, theyget a third and fourth chance, etc.One thing that would be interesting to know is thereflectivity (percentage of photons reflected) at a given wavelengthcorresponding to usable energy (i.e., energy higher than the band gapenergy). If we knew this, and also the figure for power generation, both as a function of angle of incidence on the solar cell, it would be possible (using the solar powerspectrum) to calculate the optimal angle for the V as well as the maximumtheoretical performance enhancement, where by performance I mean power outputratio compared with cells that face the sun directly. This would represent a theoretical upperbound on the benefit of the V configuration.I bet this data is sitting around on the internet … I wonder if creating “inverted pyramids” by folding in twodimensions would be any better?

I think this is a pretty good idea: the main plus point is that as a solar cell doesn’t absorb 100% of the photons with an energy above the semiconductor band gap energy, it must be reflecting some of these photons back out and they are lost. It is this lost energy that can be partially recaptured in this configuration. By using the V shape, the photons of interest get a second chance to be absorbed by the pn junction and converted to electricity. And then of those still reflected, they get a third and fourth chance, etc.One thing that would be interesting to know is the reflectivity (percentage of photons reflected) at a given wavelength corresponding to usable energy (i.e., energy higher than the band gap energy). If we knew this, and also the power generation function, as a function of angle of incidence on the solar cell, it would be possible (using the solar power spectrum) to calculate the optimal angle for the “V” as well as the maximum theoretical performance enhancement, where by performance I mean power output ratio compared with cells that face the sun directly. This would represent a theoretical upper bound on the benefit of the V configuration.I bet this data is sitting around on the internet … let’s see …I wonder if creating “inverted pyramids” by folding in two dimensions would be any better?

View Instructable » - thassaj commented on briandegger's instructable Silhouette cut Public Lab Mobile spectrometer4 months ago
This thing works amazingly well. Just pointing it at a bright area in the sky (taped to my Samsung Galaxy Note II phone) I captured what amounts to a solar spectrum. The resolution is so high you can even see the Fraunhofer lines (black lines)!

View Instructable » - thassaj commented on 344185151's instructable 利用电子垃圾——废旧光盘制作光谱仪4 months ago
Very nice one!

View Instructable »

OK - here is an embarrassingly simple calculation for the maximum amount of improvement that should be possible. A silicon solar cell with sunlight hitting it at normal incidence has a reflectivity of about 23% (assuming 50:50 of s-polarized and p-polarized light). The figure comes from the Optics Communications paper in my previous post. This reflected light represents the lost energy that could be captured. So if 100% of light hits the solar cell, 23% gets reflected off and 77% is not reflected, then the largest increase we can have is going from 77% to 100%, which would be (100-77)/77 = about 30%. So assuming that our V configuration has a narrow enough angle that the vast majority of light is absorbed, that is the best we can do. How narrow should the V be? If the V is 20 degrees, then it can be shown that there will be 9 reflections before the light bounces back out of the V. The number of reflections is I think 180/angle. But with reflectivity of 23%, after 3 reflections, the remaining reflected light is 0.23^3 = 1.22% (nearly nothing). So the tight angle with many cells is overkill.The simplest V we can make is with two panels in a V taking up the place of one panel. Then the opening angle is I think then 60 degrees, leading to three reflections.So we should see a maximum of a little under 30% increase in this configuration, and little if any improvement for more "Vs" (i.e., a tighter angle in the V and more solar cells per area).I would appreciate it if someone could check this calculation!

View Instructable »