Hi Jeff, bit of a late reply. So to calculate the gain I think you just need to integrate the area seen over a day. So ignoring north south adjustments, and just doing east west. And doing the integral for the equinox, and assuming an un-obstructed view of both horizons. (And ignoring edge effects of the atmosphere.. etc) Then the area goes as the cosine. Integrating that I get 2.
formatting link
Compared to the area of a box (In the above picture the box goes from -pi/2 to pi/2) that is pi. (If it was always pointing correctly.) So a pi/2 improvement if always pointing correctly.. about 1.57.. 57% more.
On Mon, 26 Oct 2015 06:56:26 -0700 (PDT), George Herold Gave us:
The panels really do not need to be aimed that accurately. There is no need to maintain 100% perpendicularity with the Sun's position. It only needs to point mostly east in the morning and make a few adjustments throughout the day at maybe ten degree intervals and point mostly west as the day closes. The panels will get full juice potential from about
5 degrees off axis in either direction of full perpendicularity.
So a simple arc plate and follower arm connection on each panel would work, and the telescope computer would be an excellent positioner source. Note that I did not say the telescope drive motor, just the compute engine.
Tilt the base plane up to (90 - Latitude) making it a polar mount.
Adjust the elevation about once per month. If you just want a positioner, you're fine with the relatively fast traverse of a DBS dish positioner. However, if you want to connect it to an optical tracker, you might need to slow down the traverse.
Make yourself one of these for your location: so you can where you end up. Also, your DBS dish pointer can move perhaps 10 kg of antenna without difficulties. The solar panel array under discussion weighs about 250 kg, which might be a bit much for the DBS dish pointer.
I'm curious as to what angular velocity your DBS dish pointer operates. We don't use DBS dish pointers in the USA. Instead, dishes with no rotator and 5 LNB's are common. As a reference, the sun moves at: 360 deg / 24 hrs / 60 min/hr = 0.25 deg/min Your DBS dish pointer will be much faster.
Incidentally, don't point your DBS dish at the sun. I may have experienced a related failure recently. A heat wave coincided with the bi-annual solar outage between Oct 5 through Oct 10. The combination of the quite high ambient temperature, with the 3 meter satellite downlink dishes concentrating some of the light on the C-Band LNBF was enough to burn out both LNBF's. There was no scorching or peeling paint, but the adhesive labels fell off. Oddly, both LNBF's seem to have partly recovered, but are useless as they now are intermittent.
--
Jeff Liebermann jeffl@cruzio.com
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558
On a sunny day (Mon, 26 Oct 2015 09:55:13 -0700) it happened Jeff Liebermann wrote in :
OK.
Wow, that is a lot of work. I think you are radio ham too? I have the 'predict' program installed, both on the PC and some Raspberry (to drive yagi antenna for amsat). Predict also has as menu choice to follow the sun :-)
--== PREDICT v2.2.3 ==-- Released by John A. Magliacane, KD2BD May 2006
--==[ Main Menu ]==--
[P]: Predict Satellite Passes [I]: Program Information [V]: Predict Visible Passes [G]: Edit Ground Station Information [S]: Solar Illumination Predictions [D]: Display Satellite Orbital Data [L]: Lunar Predictions [U]: Update Sat Elements From File [O]: Solar Predictions [E]: Manually Edit Orbital Elements [T]: Single Satellite Tracking Mode [B]: Edit Transponder Database [M]: Multi-Satellite Tracking Mode [Q]: Exit PREDICT
selecting Solar Starting UTC Date and Time for Predictions of the Sun
BTW I wrote some I/O interface for that, there exist several.
Yes, I did not know OP was using such a big thing, I only have a 80 W panel, now fixed on south, is actually for on the boat. Great for topping batteries.
East to west in about ten seconds. The nice thing about these (this is a Motek) is that it has 'goto angle'. This one has now been in use for 15 years, no problems. even though I once opened it and stuck probes in it. To see the Eutelsat bus specs (pdf):
formatting link
At that time I wrote some software for it:
formatting link
GUI:
formatting link
It has a build in geosat calculator, given your location and where the sat is (x degrees east or west), and it will re-calculate where to point the dish. The European Diseqc system uses 13 V / 18 V for polarity and a 22 kHz tone for high and low band.
formatting link
The 22 kHz tones are also used to tell the motor how to turn, and where. The 'goto angle' is a nice feature, works perfectly. Once my small HD USB sat box broke down, and I figured out the circuit, they use a LM 317 from an about 25 V supply to make the switchable 13 or 18 V, and feed the 22 kHz from a micro output pin into the LM317 adjust, I replaced the LM317, it died again, then put it on a heatsink, still working... Anyways if OP wants to send signals he could use a 555 timer to make the 22 kHz and send the commands as low rate digital via one of those (here it is 430 MHz) free frequencies. But 250 kg? No, but you could make something, for 6k$ I think.
I ordered some freshnell lenses from ebay just to play, to make fire in the wild, man those are cheap, easily set your place on fire too:
formatting link
And I noticed you can use this in place of a Manta, with some effort look through it with 2 eyes to get depth while soldering that tiny SMD stuff. But I think solder plumes will darken or damage it over time.
It occured to me that as you know when the thing points up tracking, you can sync your micro timer each day... So it is a clock too :)
Then auto EPS motors are not overkill. I am also sizing the right motor fo r my van, which has a leaking hydraulic rack and pinion. I am trying to re place it with electric, perhaps with BLAC motor. 12V 30A or 24V 15A should work, with a second battery for EPS.
Next door to a previous workplace, there was a radio tower that worked on a similar principle. The dish antenna was laying on the ground and pointing up, and there was a reflector at the top of the tower (roughly
50 or 60 feet up) angled at about 45 degrees. The antenna had a pointy fiberglass radome on it to keep water and crud out. The EE professor I worked with said that at one time, there was an idea that the loss over the extra "air" distance from antenna to the reflector was less than the loss over the extra "wire" distance of stringing co-ax up the tower. Also, the antenna was easier to service on the ground. I don't think it caught on, because I haven't seen any other towers built that way.
It's called a "periscope antenna". Such antennas have been effectively banned by the FCC for many applications[1]. The problem is that the reflector at the top of the tower doesn't capture all the RF sent in its direction. Even at such short distances and narrow beamwidths, quite a bit of the overspray hits the tower, other antennas, or mounting structure, which then bounces the RF in unwanted directions. When microwave links are licensed and coordinated, there is quite a bit of concern about such problems because there could easily be other users nearby on the same frequency who would not appreciate interfering RF sprayed in their direction.
In the above photos, notice that the periscope antenna is often the ONLY antenna on the tower. That's because one cannot put other antennas in the path or they will scatter the RF.
Although they still can be licensed, the FCC has taken a dim view of passive microwave reflectors, which look like giant billboards on mountain tops. Same problem. They can scatter RF in undesired and uncontrolled directions.
It might be true that traversing the same distance through the air has less loss than the equivalent piece of coax. However, if I include the reflection and possible overspray losses in the reflector, pretend that waveguide doesn't exist, and ignore any gain that might be provided by a similar size coax cable driven antenna, then your EE professor didn't bother running the numbers.
[1] (c) The Commission shall require the replacement of any antenna or periscope antenna system of a permanent fixed station operating at 932.5 MHz or higher that does not meet performance Standard A specified in paragraph (c) of this section...etc...etc...etc
--
Jeff Liebermann jeffl@cruzio.com
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558
While the panel illumination level can be easily calculated from cos(error_angle) real solar panels have some form of anti-reflective coating or the panels have grooves to minimize direct back reflection from the silicon, thus causing the photo-electric conversion.
These structures are optimized for reducing back-radiation from perpendicular radiation, but unfortunately at the expense of off axis radiation.
Thus the electric output drops faster than simple geometry would suggest. For this reason at least a few degree error is acceptable, but with pointing errors more than 10-15 degrees, the electric output is more significant.
In addition, following the sun own to horizon is not very efficient due to the Air Mass (AM) losses close to the horizon. This is clearly evident, since you can look directly into the midnight sun (at high latitudes of course).
When I was working there, I remember there being at least one other antenna, but it was on the other side of the tower from the reflector. That was was a conventional dish antenna, with co-ax to the antenna, pointing more or less horizontal.
It's a guyed tower, three vertical tubes in a triangle, with cross- braces between the tubes; the reflector stuck out from one vertex of the triangle, not along the flat side.
Looking at Google Street View from May 2014, the reflector is still up there, but I think it may not be in use; there are some rod antennas now bolted to the same vertex of the triangle that the reflector is on. Those antennas weren't there before. I can't tell if the dish is still on the ground; it's on a property that the Street View car didn't visit.
I don't know what purpose the periscope antenna originally served. It may have been a private communications link for the company that originally owned the whole set of buildings. The buildings had since changed hands, but the new owners maintained the tower.
If I remember right, the dish on the ground was sized somewhere between an old C-band satellite TV dish, and a modern DirecTV dish.
Maybe it was an EE student instead; this was ~10 years ago. I know that the prof I am thinking of had done some RF stuff in the past; one of his jobs included shooting radar at prototype aircraft parts. If the radar didn't see anything, that was a good result.
Although two fixed flat mirrors would be better. The simplest design with a gain of two is half a hexagon with the PV at the base.
\_/
light entering the wider aperture normal to the PV array hits it after one relfection. It is the simplest of the non focussing flux concentrators and also doubles output during diffuse illumination.
Much more than 2x gain and you tend to cook the PV cells ruining their efficiency unless you have a combined water cooled panel.
The more complex designs with realistic gains of upto 10x take the simplest form of parabolas arranged so that the focus of one is at the base of the other. This trick means any photon entering the front aperture will after a finite number of reflections hit the PV plate.
These designs are borrowed from early HEP neutrino detectors to leverage the detection sensitivity of expensive PM tubes. They also appear in solar power books from the late 1970's oil crisis onwards.
Another cute one is spiral based but hell to make.
On Tue, 27 Oct 2015 08:42:32 +0200, snipped-for-privacy@downunder.com Gave us:
All of the roadside help phone units do not even move at all. The same is true of the FAA "outer marker emitters" that get their battery fed in this manner. And one would think that they are placing one of the more efficient variety panels. So "those structures" must be optimized for noon-time peak output, and not even care about the beginning and late day reductions.
Such programs use Keplerian elements to track the various satellites. For completeness, thee's data available for tracking various geosynchronous satellites (which don't change position), the moon, sun, planets, some large space junk, military satellites, and possibly some comets. Here's some for tracking the sun:
Most of the interfaces are designed for driving ham radio azimuth and elevation type antenna rotators.
The OP did not specify what they are using. Somewhere in the discussion, topic drifted to sizing the drive motor, which involved estimating the gearing required, for which I offered a "typical" gearing arrangement and specifications for a rather large WattSun solar panel tracking array.
How many degrees is east to west? Assuming the full 180 degrees, that's: 180 deg / 10 sec = 18 deg/sec = 1080 deg/min Your dish pointer is 4,320 time faster than the 0.25 deg/min movement of the sun. The problem with moving too fast was detailed in a previous rant on optical tracking, where the tracker would erratically try to follow reflections, clouds, automobile headlights, aircraft, etc. The advantage is that your motor doesn't need to run for very long each time it changes position to follow the sun. It also helps to have a high gear ratio to reduce the size of the drive motor.
--
Jeff Liebermann jeffl@cruzio.com
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558
You can't ignore (or discount) the effects of atmospheric attenuation: There's quite a bit less light available at low angles than at high noon. At really low angles, the losses are huge. See table below:
SEA is "solar elevation angle" AM1 is "Air Mass with sun directly overhead".
Solar Insolation SEA W/m2 % Drop cf AM1
Offhand, I would guess(tm) from the table that trackers that work at elevations below about 20 or 30 degrees are not worth the cost or effort.
I don't think I could produce a complete model for atmospheric attenuation, mostly because it varies by altitude, weather, season, temperature, air turbulence, air pollution, etc. Worse, the above article mentions that the curvature of the earth needs to be considered. I guess losses through the glass solar panel covers should also be considered (Brewster's angle). I'll see if I can find something where someone else has done the work.
--
Jeff Liebermann jeffl@cruzio.com
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558
On a sunny day (Tue, 27 Oct 2015 10:46:28 -0700) it happened Jeff Liebermann wrote in :
Yup, plenty of data, I added ISS parameters by hand (from NASA).
Exactly, given the 'predict' numbers it would be sufficient to just change position maybe every hour or so.
I am not sure but from what I have experienced with this panel so far sun is not always visible through the clouds, and there is most of the time a lot of scattered light. Not sure what the real gain would be to always keep at the right angle to the sun.
As to 180 degrees, you can program 2 end stops in the thing, there are also 2 micro switches for the real limits. I never go the full 180 as there is 1) no sat of interest there, and 2) it hits the wall. So the end-stops are important for if you want to wave it around by hand, I never do that, I select a program (set a timer) and it will go exactly to the right angle for that sat. So in setting it up (there are many ways) I tune to one station's transponder, move the motor to the calculated position, and then by hand move the dish until I get best signal (beeper). and fix it on the motor shaft (you also need to use the right elevation, there is a scale on the mount). the motor shaft is bend in such a way that if you move east or west from from due south then the thing will tilt the dish, following the earth curvature, so polarization changes so it is still correct for the sat you point to, From that moment on all other positions in degrees are OK. It is quite accurate (to a degree or so).
This is for geo sat of course. If not stationary you need elevation control too.
It is really very very simple, but I cannot find that simple picture I have in mind online, just imagine the amount of crap people have to digest,,,
Well, no, it doesn't do anything useful in diffuse light.
The product E = (area * projected solid angle) of any optical system can't exceed that of the source (or equivalently of the detector). If it could, you could make a perpetual motion machine. (For more discussion, search for "sufficiently small heat engine" on sci.optics about a dozen years ago.)
The projected solid angle of a hemisphere is pi, so for a flat detector, E = pi * A.
The Sun subtends a full angle of about 0.5 degrees, so its solid angle is pi/4 * (0.5*pi/180)**2 = 0.00006 sr. The maximum theoretical concentration is thus pi/0.00006 = 52000. People have made practical concentrators up to 2500 suns that I know about. That's a few percent of the power density of the solar photosphere, so cooling gets to be a challenge!
The solid angle subtended by the yearly path of the Sun is of course much larger, +- 23 degrees in elevation and more than +-90 in azimuth. If we say we want +-23 by +-60, that's about 1.3 steradians, so any stationary concentrator is limited to a factor of pi/1.3 ~ 2.4.
In diffuse light, the source subtends pi steradians itself, so no concentration gain is possible.
Cheers
Phil Hobbs
--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC
Optics, Electro-optics, Photonics, Analog Electronics
160 North State Road #203
Briarcliff Manor NY 10510
hobbs at electrooptical dot net
http://electrooptical.net
Sure.. I mean we can look at the sunset's with no eye damage. I was just trying to give you the "physicist's spherical cow" model of how to think about it. (I took me half a beer sitting by my creek to realize that the base of my rectangle was pi radians long. (I started with one... crazy numbers.)
Sure.. I can make another "spherical cow" model for that too. (Treat atmosphere as a layer X miles thick and do the geometry.) But to me exact numbers are not that important. A 30-40% gain seems reasonable.
I guess losses through the glass solar panel covers
Do they put AR coatings on solar panels? I think you get something like a ~30% reflection from the index of refraction mismatch of air to Silicon.
Yes you are right about this simplest one. My mistake. It doesn't do anything helpful for diffuse illumination.
It is never enough to be a perpetual motion machine but it is enough to be interesting. I have made a 10x gain one out of aluminium foil once.
You can prove that all photons crossing the entrance aperture will eventually hit the PV array (assuming perfect reflectivity).
No argument there.
Classical flux concentrators are imaging devices. These are not and they aren't breaking any of the rules of thermodynamics. They just put the PV array in a position where the sun apparently subtends a much larger angle. Ray tracing designs allow you to optimise them.
formatting link
It isn't very well written :(
OP might find this link interesting:
formatting link
Beware several other links I found searching had hostile malware lurking on their websites despite containing benign technical content.
For an imaging system that is true, but for a non-imaging system it isn't although there are obviously hard constraints on the available solid angle around the target.
The problem with efficiencies and loss calculations is that they are the product of many such efficiencies and losses. Make a few too many assumptions about accuracy, reasonable numbers, and expected outcomes, and final overall efficiency product will be spectacularly wrong. In this case, the problem is that trackers generally cannot be pointed at angles lower than about 20 degrees above the horizon, making the recovery of low angle light a rather futile exercise. The dusk and dawn alignment errors of the typical single axis tracker adds yet another source of loss.
Yep. Ordinary uncoated glass panel covers will reflect perhaps 70% of the light that hits it. Coated glass will reflect only about 3-5%. You can easily tell by the lack of glare and reflections off the panels (which makes the neighbors happy). The bad news is the silicon coatings can pass quite a bit of far IR, which will heat up the solar cells, reducing their efficiency. However, things are looking better:
That's for pure silicon (n=3.44) which shows a reflectance of about
30%. SiO2 (n=1.45) shows a reflectance of 3%.
--
Jeff Liebermann jeffl@cruzio.com
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558
I have Welford's book on nonimaging concentrators, and read it with interest just last year, so I'm well up on them. The radiance conservation rule applies regardless of the optical system--imaging, nonimaging, fibre, anything you like.
Nope. Radiance conservation applies regardless. Otherwise you could use a concentrator to make something hotter than the Sun, i.e. heat would spontaneously flow from colder to hotter. Ain't happening.
Cheers
Phil Hobbs
--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC
Optics, Electro-optics, Photonics, Analog Electronics
160 North State Road #203
Briarcliff Manor NY 10510
hobbs at electrooptical dot net
http://electrooptical.net
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.