Having built one optical ("lightbeam") transceiver
enclosure already, I wanted
to build another
enclosure so that there would be a second unit in existence. (What
is the sound of one hand clapping? What good is a transmitter if
there is no receiver?) Because I wanted it to be quick, easy,
and cheap, I minimized the cost as much as I could - yet, I still
wanted it to be good enough that it was usable and could withstand
being hauled around a bit. While this unit was never optimized,
it has been successfully used in a two-way exchange over a 107 mile
(173km) path!
List of materials:
Comment: Unfortunately, I did not take pictures of the enclosure in various states of construction - sorry.
Assembly - Front half:
Note: There are two identical enclosures - one for receive, and the other for transmit. After initial optical alignment, they are connected to each other to form one unit.
The description below is for just one of the two identical enclosures. Note also that each enclosure consists of a "front" and "rear" half, nested together to provide adjustable focus.
The first step is to characterize the Fresnel lenses. The full-sheet magnifiers that I used have a focal length of about 310mm - a property determined by focusing an outdoor view onto a sheet of paper and then measuring the paper-to-Fresnel distance. This measurement was very important in determining the dimensions of the box.
After removing the glass from the picture frame, I carefully measured it and then cut four sheets of the foam core board. For the top and bottom (the short, or 8" side of the frame) the board was exactly the length of the frame, but for the two sides (the 10" dimension of the frame) the board was cut 3/8" longer so that there was some overlap with the edges of the short side: This allowed the pieces of foam core board to completely enclose the frame, yet provide an interior surface for gluing. As for the length of these pieces, I chose 250mm.
Comment: As pieces are measured, cut, and fitted, it is strongly recommended that they be labeled and marked with to indication the front, top, sides, etc. - and also to note precisely which pieces are fitted against each other!
Using some tape to hold the sides together, I assembled the four sides of the box and then inserted the picture frame, setting it back from the front by about 30mm: This setback performs a slight function as a lens shroud, but mostly it protects the glass lens during normal handling. Once I verified that the pieces of foam core board were of the right length and that the plane of the picture frame was at a precise right angle to the box (that is, that it was set in straight) I used the thermoset to glue everything into place. When doing this, it is best to use a few small dabs in some strategic places to hold everything together - and then re-measuring to make sure everything is lined up before applying the full amount of glue. When applying the glue (which is hot enough to cause a nasty 2nd degree burn) it is best to glue one or two joints and then wait a minute or two for it to set before turning the box around and gluing a few more joints.
Once this portion was completed, the entire assembly was quite rigid - yet light.
As can be seen from the pictures, each enclosure is really two pieces, nested inside each other with the front half of the enclosure sliding inside the rear half allowing one to adjust the focus simply by sliding the two pieces past each other. Because of this, it is important that the glue joints for the front section are inside the enclosure and that the edges are flush so that the two sections fit as tightly as possible.
While it would have made more sense to make the rear half slide inside the front half (as the rear can be smaller, from an optical standpoint) I did not do this. Why not make the front portion smaller? Well, I was sort of lazy. Because the ultimate size determination is that of the picture frame, the front half had to be constructed first. It was easier to build that portion - and then cut the pieces to size to build the rear portion to fit around the outside of that. It was easier to build the second (rear) portion by wrapping around the outside of the front portion than trying to build the rear portion by assembling it inside the front portion.
Assembly - Rear half:
After the front assembly was completed, a 250mm deep box was wrapped around the already-completed front portion. For this, the side pieces were cut to precisely the length of the sides of the front box portion while the top and bottom pieces were cut a bit longer, allowing some overlap for the glue joint. Using the front box as a form, wooden spring-loaded clothespins and weights were used to assure a very snug fit during assembly. It should go without saying that one should be careful to avoid gluing the two halves of the enclosure to each other.
Unlike the front portion, all of the glue joints were done outside the box as to avoid interfering with the tight fit - and this is why the top and bottom pieces are cut slightly oversize.
Once the glue has set I pulled the two boxes partially apart and traced the shape of the back of the enclosure is traced onto another piece of foam core board to be used as the rear panel: Tracing the size allows for a perfect, snug fit. Before gluing the back panel into place, I glued some 2"x2" squares into the corners about 75mm from the rear edge. These function not only as gusset plates to help hold the shape of the box, but also as stops to prevent the two boxes from being pushed together too far - something that could make them very difficult to get apart.
Prior to installing the rear panel, I located the exact center of it (by drawing an "X" on it) and cut a 100x100mm square hole, centered in the panel, for mounting the emitter and detector circuitry. While you are at it, also draw centered horizontal and vertical lines on the back panel to aid in the location of the its center - a helpful guide when installing the electronics. Installing the rear panel, it was glued (with the lines that you drew on it facing outwards) along both the inside and outside, making the rear box quite sturdy.
Installing the lenses:
After all of the gluing has been completed, it is time to install the lenses. Because these Fresnel lenses are extremely thin and flexible (I believe that they are made of vinyl) they need to be attached to the glass in order to be stable enough to be optically useful: Attaching the lenses was done simply with some pieces of clear tape, centering the lens on the piece of glass. When installing the lenses, make sure that they are as flat as possible before applying the tape in order to provide the best optical properties. It is also important that the proper side face outwards: It was empirically determined that for these lenses, the "grooved" side should face away from the emitter/detector.
While these lenses were precisely the same "long" dimension as the glass, they were slightly narrower, leaving a small gap along the "long" edge. For the receiver, I simply masked this "un-lensed" portion with black electrical tape to minimize ingress of stray light, but I didn't bother doing this with the transmitter.
I noticed that these lenses weren't completely flat and tended to bow outwards in the middle. The obvious solution to this would be to sandwich the lens between two pieces of glass. Unfortunately, I only bought three picture frames and had only three pieces of glass. Reasoning that the best-performing lens should be the one on the receiver, I sandwiched only that lens.
The glass is held in the frames with dabs of hot-melt glue: The enclosure was laid "front-down" so that the glass was in the front and the "grooved" side of the lens was also facing front. For the receiver, the second piece of glass was laid on top of the lens, and the entire thing was carefully weighted down (to flatten the lens between the two panes) using a 7 amp-hour 12 volt lead-acid battery. With the lens secured, the edges were dabbed in several places with the thermoset glue. A bead was not run around the entire perimeter, as this would have made disassembly - if needed - much more difficult.
For the transmitter, because I didn't have another pane of glass handy, so I simply cut a strip of foam core and laid it across the lens edge-wise, gluing the ends of the strip to the frame. This served to hold the center of the Fresnel lens against the glass and removed most of the "bulge". (Note: This strip may be seen in the bottom picture of Figure 2 in the left-hand lens as a horizontal bar.) Prior to final assembly, a brief test was done to note the difference in "beam quality" between a Fresnel lens sandwiched between two panes of glass and one that was not. While the center of the beam wasn't affected much, the periphery of the beam was somewhat "cleaner" with the flattened Fresnel. It was also determined at this time that simply holding the center of the Fresnel against the glass fixed "most" of the degradation. Because of this, I feel confident that for the transmitter, the remaining "un-flatness" is will result in only a small amount of degradation. (If I'd had another piece of glass handy, I would have put it in, but now that it is installed, I won't bother adding one later.)
It should be mentioned again that these flexible "Sheet Magnifier" Fresnel lenses are of rather poor optical quality. In comparison with the rigid, molded Fresnel lenses, they produce a rather "dirty" pattern - most notably a somewhat "fuzzy" image and the tendency to produce a weak "X"-shaped pattern of spurious light. It is mostly for this reason that I didn't go through too much trouble to flatten the the transmit lens.
Comment: On the outside of the front half the enclosure, I drew a line at the location of the plane of the Fresnel lens. This provided a handy reference for measurement of the focal plane of the lens during the "rough focus" setup.
Mounting the electronics:
As mentioned, a 100mm x 100mm square hole was cut into the back panel for mounting the electronics. The detector and emitter assemblies are mounted on a plate made from a piece of foam-core board that is somewhat bigger than the hole - about 150mm square or so. Onto these I drew both an "X" and a cross - on both the front and back of this plate - to locate the center as well as to provide alignment with the center-locating lines that were drawn on the rear panel. The idea here is to align the LED and the detector's photodiode to the exact center of the mounting plate and, with the lines, be able to mount the plate in the center of the enclosure's rear panel.
Figure 3 provides several views of the mounting of the enclosure. The top picture shows the installed panels with the electronics packages. These panels are held in place with "grabber" (or drywall) screws. Note that because this foam-core board is rather fragile, it takes screws with very pointed tips and deep, aggressive threads to hold things into place. Even so, these screws cannot hold into the board material very strongly - but the panels should be very secure as long as reasonable care is taken in handling of the enclosure. In order to improve "hold" several layers of foam board (and/or thin wood) could be glued together inside to provide a stronger grip.
The middle picture shows the detector. This is the original prototype version of the Version 3 optical receiver mentioned on the Optical Receivers page on this site, and it was used simply because it happened to be laying around on the workbench. As can be seen from the picture, some extra pieces of circuit board material were soldered into place to provide mounting from the screws, as well as side-shields to minimize RF and electric field ingress. Note that if this unit is used near a transmitter site, further shielding will probably be required! I was pleased to note that inside my house, it was remarkably unaffected by nearby electromagnetic fields.
The bottom picture shows the emitter assembly - and some explanation is required here. My first experimentations with the 3-watt Luxeons were with this unit - a Luxeon emitter (not a star) epoxied to a small aluminum heat sink. I soon noted that, for continuous operation, that the original, small heatsink was inadequate. When mounting the emitter for this project, I rummaged through my collection of heat sinks and found a larger one (from the same computer monitor, I believe) that had fins that would jam into each other and interlock. I simply smeared some heat-sink compound on the mating surfaces, pushed the two heatsinks together, and then put dabs of high-temperature epoxy on the lot to hold them together. Testing showed that there was excellent heat transfer between the two, thus solving the "small heatsink" problem.
To further facilitate mounting to the foam core board, I cut a small piece of sheet aluminum and used it to spread the distribution of force from the mounting screws (as well as increase the heat sink area) to provide a solid mount. On the rear of the panel, some small washers were also used to distribute the force and prevent tearing through the board. (I may put larger washers on the back panel, though...)
Comments:
Matching the emitter to the lens:
Once the emitter assembly was mounted, I checked the size of the "circle of light" produced by the Lambertian pattern of the Luxeon emitter at a distance of 310mm (the focal length) and compared it to the size of the Fresnel lenses. This test - along with one using the lenses themselves - showed that most of the light was spilling out beyond the edges of the Fresnels, resulting in a significant loss in available radiant energy and meant that a secondary lens was needed.
It is normal practice to use a PCX (Plano ConveX) lens to change the "spot size" but I had on hand some cheap, plastic DCX (Double ConveX) lenses (20mm diameter with a 34.5mm focal length) that I was dying to use. A quick test showed that this tiny DCX lens was nearly as effective as a much larger, heavier, and more expensive glass PCX lens - and it was obvious that it would be much easier to mount the small, plastic lens.
The "mounting tube" was made by rolling some paper around another tube (a 6LN8 - a pentode-triode combination to be precise) and taping it. I was then able to force the plastic DCX lens into the end of the paper tube (which was fractionally smaller in diameter than the lens itself) which held it fairly securely. I then carefully applied some clear, fast-setting epoxy at the boundary between the paper tube and the DCX lens, taking care to keep as much of it off the optical area of the lens as possible. After a few minutes (and exposure to the warm heatsink of the powered-up LED) the epoxy had set, so I carefully trimmed down the paper tube until the "spot size" of the LED at 310mm approximately matched the size of the Fresnel lens - something that occurred when the lens was about 5mm away from the top of the Luxeon's dome. At this point, the paper tube was epoxied (using the high-temperature epoxy) to the small heatsink and centered over the top of the Luxeon emitter. This technique is simple, crude, yet effective and rugged.
Comment: I believe that significantly better efficiency (more light from the LED being directed toward the lens) could have been obtained if a different secondary lens was used. Because this was a "quick 'n dirty" transceiver, its somewhat inferior - but adequate - performance was the acceptable tradeoff in order to have a 2nd unit onhand for testing.
Aligning the optics - emitter:
Alignment and focusing of the emitter was fairly straightforward - especially after having done so on the previous enclosure. During construction, I was careful to make the picture frame parallel with the front of the enclosure as doing so allows the use of a carpenter's square to determine the pointing of the lens when a laser level is clamped to the carpenter's square - see this page for more details. In a nutshell, the use of the square-laser combination allows one to determine precisely where the lens is "pointed" - and one can easily verify that the emitter is located within the center of the focus spot of the lens.
After verifying alignment, the mounting plate is screwed to the back panel and the beam is focused: Focusing is done simply by sliding the two boxes in and out of each other. While rough focus can be done indoors, provided that one has a fairly long (30 feet or 10 meter) distance, this will only give approximate results. For final focusing, I took the entire enclosure outside and used the side of a house on the next street over (about 300 feet or 100 meters away) as a target, focusing for the "sharpest" square pattern. It should be noted that with the secondary lens, the focus is somewhat "closer" to the lens than the focal length of the Fresnel lens alone.
Once the precise focus was found, I marked it with the silver-colored marking pen and then secured the position with the four "grabber" screws - one on each side of the enclosure.
Comment: As the beam is focused, one should be able to resolve the "square" shape of the Luxeon's emitter. If good quality lenses are used, the bond wire and connecting electrodes should be visible in the focused image at relatively short distances. Best focus is obtained when this "square" pattern is at its sharpest - and this should be done over a large a distance as possible. Considering the accuracy of typical Fresnel lenses, a distance of about 300 feet (or 100 meters) is likely to be "close enough" to infinity to achieve a good focus. With these somewhat poor-quality "sheet magnifier" lenses, the square shape of the Luxeon's emitter was somewhat blurred, but still generally recognizable.
Aligning the optics - detector:
Focusing of the optics - especially the detector - is, perhaps, the most awkward part of the construction of the entire project. Unlike the emitter, there is no spot being cast to show the precise alignment. Now, it should be possible to simply substitute an emitter -located in the precise position of the photodiode- to accomplish this, but I have yet to do so. Using the indoor test range, I was able to get approximate focus - a setting that should be "fairly close" to optimal. Through previous experience, I noted that if one focus in the indoor test range, the proper focus for an infinite distance is one that is slightly closer to the lens.
As for the other enclosure, I checked receiver focus and sensitivity using a dim, diffuse, red LED attached to a wall about 33 feet (10 meters) away. This LED was modulated by a signal generator and by varying the drive, the intensity of the LED could be adjusted from a "dim, but easily visible" setting down to a "I can't see it unless I'm right next to it." As with the other enclosure, an easily-audible (and speech-capable) signal was obtained from an LED that was not readily visible at the full 33 foot distance.
Later, the enclosure was dragged outside. As expected, the waxing moon was more than enough to completely saturate the receiver, and fairly distant streetlights (and flying aircraft) were clearly audible.
Comment: After completing this enclosure, the focusing of the transmitter and receiver was re-checked using some simple equipment on a test range and both the receiver and transmitter were found to be within a fraction of a dB of their optimal settings in terms of focus and parallax.
Marrying the two enclosure:
The experiences of the Australian group (Chris, VK3AML and Mike, VK7MJ) show that it is most convenient to gang two lenses together and precisely align them, thus eliminating the need to aim more than one box. In this case, two separate boxes were built - mainly because it was easiest to do so - but there was still the matter of putting the two together and making sure that they were in alignment.
First, the two enclosures are laid next to each other and just the two rear screws holding in the large, square plate on the top and bottom are inserted: This will loosely couple the two sections together.
As can be seen in Figure 4 some scrap pieces of foam core board were used as "plates" to tie the two together. As with the rear plates and focus adjustments, "grabber" screws were used to secure the plate. On both the top and bottom, two plates were used: A larger, square plate near the back (the one with "TOP" written on it, both normally and upside-down) and a smaller plate at the front of the enclosure: Two similar plates were used on the bottom side as well. In addition to plates on the top and bottom, there are two small plates (or straps) mounted on the rear panel to further improve rigidity.
Because the entire purpose of these plates is to gang the boxes together and ensure stability, it is very important that the two boxes actually be pointed in the same direction and are properly aligned. Note that due to parallax, this must be done with a special target that takes into account the distance between the two lenses. As the distance approaches infinity, the beamwidth of the two lenses will cause overlap, effectively negating the parallax.
The task of alignment was done in a way similar to that done with the other enclosure: A paper target was attached to the far wall (just above the LED used for testing the receiver) and marks were made on the paper: One mark, representing the center of the transmitter beam, was spaced the same distance from the receiver-test LED as the centers of the two lenses, while two other marks placed at the distance from the the center of the transmit lens and the boresight of the laser level to locate both azimuth and elevation of the enclosure.
First, the receiver was carefully aligned on the test LED. Further verification of the receiver's aiming was done by pointing a laser pointer at the test LED: A "hiss" from the laser pointer was noted (along with some receiver desense) and by moving the laser pointer side-to-side and up and down with respect to the test LED, the proper aiming of the receiver could be verified.
Once the receiver was aimed, it was necessary to be very careful to avoid disturbing the position of the receive portion of the enclosure. At this point, the emitter is activated and the transmit enclosure is shimmed and shifted as necessary to get the center of the transmitter's beam to line up with the marks on the paper. As proper alignment is achieved, the other plates may be added to "lock" things into place.
After installing the plates, re-check the alignment once again by peaking the receiver on the LED with the test signal and then verifying that the center of the transmit beam was in the center of its respective target. If minor adjustments are to be done, the screws holding mounting plate or the emitter assembly may be removed, the position of the emitter adjusted slightly, and the screws reinstalled through new holes. Note that if this sort of adjustment is required, it is easiest to do so with the transmitter - after verifying peaking of the receiver - because the transmit beam can be easily seen...
Adding a septa:
As mentioned on the page "Optical Communications for the Amateur" (by Chris Long, VK3AML) a device called a "Septa" may be used to reduce stray light such as that encountered in an urban environment. The structure of a septa is simply that of a series of parallel tubes that only allow light from the direction of the source to enter the detector and it would, ideally, be constructed of thin, opaque material painted flat black to minimize reflection: A series of thin metal plates is often described as it would cause minimal blocking of the desired light. Practically speaking, there is a limit on how narrow the viewing angle of the septa will allow and this angle is directly related to how long the tubes are and how small each of these tubes might be.
Because the optical quality of these sheet protectors is rather poor, they are quite susceptible to light from angles far removed from the desired direction. Because of practical reasons, the length of the septa (and the number of dividers) was limited. Also, because I was using scrap foam core board left over from the construction of the enclosure, I had to take into account the limited amount of material onhand and the fact that the thickness of the sheets would block some of the light.
The installation of the septa is rather simple: Because the picture frame is set back by several centimeters (to protect the glass, mostly) I constructed the septa to simply slide in, being held in place by friction. The structure of the septa is simply a series of rectangular tubes created by dividers cut from foam core board, all held into place with "hot melt" (thermoset) glue.
Although it is difficult to tell from the pictures, you might notice that from the front view, there appear to be fewer vertical dividers than are visible from the rear view, but this is because some of the dividers are about half the length of the others: Simple geometry will demonstrate that the divider placed midway between the two adjacent ones need not be as long as those dividers in order to provide complete blockage.
This septa greatly narrows the view of the lens while blocking, perhaps, 10-15% of the light - less than 1dB, and weighs about 12 ounces (0.3 kg.)
Was the septa worth the trouble?
Another accessory (not yet shown) is an optical attenuator plate . Designed to be used in an urban environment over fairly short distances, this "optical attenuator" is simply a piece of cardboard with holes in it. This attenuator plate is placed against the Fresnel lens, blocking 90-95% of the incoming light. Because of the holes are distributed over the same area as the Fresnel lens, the capture area of the lens remains about the same and most of the reduction of scintillation provided by the larger area is maintained.
Because of the relatively high light pollution encountered in an urban environment, the limiting factor in the sensitivity of the receiver is not the sensitivity of the detector itself, but the amount of "dilution" caused by other light sources. When this attenuator is used with a septa, the beamwidth of the septa itself is narrowed somewhat by the holes: If two matching attenuator plates are used with a septa (one at the lens and another in front of the septa) then the beamwidth may be narrowed even more.
What this means is that if the background light is well above the noise floor of the receiver, dynamic range of the receiver itself can be compromised by this extra light. Also, at relatively short distances, the full sensitivity of the receiver may not be required and the attenuation of "background" noise sources might be welcome.
Like the septa, the optical attenuator was only tried once as an experiment and is not among our normal compliment of equipment.
Fine-tuning the elevation:
For our first optical contact, I used the wooden enclosure while Ron and Gordon took this enclosure to a point across the valley. In setting up, they noticed two problems that complicated the precise aiming of the enclosure and keeping the aim steady:
Spot Quality comparisons:
As was expected, the higher-quality optical acrylic Fresnel lens used in the wooden ("first") enclosure produced better-quality "spots" than the vinyl "full-page magnifier" lenses. During testing, I decided to do a direct comparison between the two.
Ideally, one would have done everything possible to make the two tests equal, but there were some unavoidable differences that may affect the accuracy of the direct comparison of intensity:
Of more interest was the "quality" of the spots that the two boxes produced even when optimally focused. As can be seen from Figure 6 the "main spot" (the brighter "square" portion) is almost identical in size, but the lower spot (from the page-magnifier Fresnel lenses) is not only dimmer, but more light is spread out beyond the main beam perimeter. It should be pointed out that this effect is apparent not only from the pictures, but is arguably more visible when view with one's own eyes. Although not visible in the picture, there is a sort of faint "X" pattern weakly emitted from the page-magnifier lenses that seemed to be totally absent from the higher-quality acrylic lenses.
Again, it should be noted that in the case of the "fuzziness" of the spot of the vinyl page-magnifier lens, this peripheral energy could not be removed by adjusting focus. In the case of the images in Figure 6, the spot was projected onto the surface of a 33 foot (10 meter) diameter white satellite dish that was about 200 feet (60 meters) distant. At this distance, the beams had "mostly" collimated, but were very slightly out of focus as compared to the normal test distance that I'd been using of about 525 feet (160 meters.) The satellite dish was chosen because it was the only relatively large, flat, white surface that was available at a reasonable distance. Note: One can see the subreflector assembly in the bottom of these pictures, along with some lines from the sections of the main reflector as well.
Comments:
Keywords: Lightbeam communications, light beam, lightbeam, laser beam, modulated light, optical communications, through-the-air optical communications, FSO communications, Free-Space Optical communications, LED communications, laser communications, LED, laser, light-emitting diode, lens, fresnel, fresnel lens, photodiode, photomultiplier, PMT, phototransistor, laser tube, laser diode, high power LED, luxeon, cree, phlatlight, lumileds, modulator, detector
 |
List of materials:
- Two Fresnel Lenses.
- The cheapest, large-sized Fresnel lenses that I could find were of the "full sheet magnifier" types. While these are readily available at many stationary and office supply stores, I found them for $2.00 each at American Science and Surplus, part number 8665 (LINK). These Fresnels are very thin (only 1-2mm at most) and very flexible, about 7.25" by 10.25" (184x260mm) and have a focal length of approximately 12.25" (about 310mm.) They are of mediocre optical quality - but "good enough."
- More recently, inexpensive (ranging from $4 to $10, depending on brand and the retailer) "rigid" Fresnel lensed, molded from optical acrylic, have been appearing at various book stores and office supply houses. A number of brand names have been noted, such as "Basic" and "Bausch and Lomb". These are roughly 8-1/4" x 10-3/4" (21 x 27cm) and are variously listed as having "2x" or "3x" magnification and with a variety of focal lengths. The samples that we have obtained have a focal length of about 32cm - about the same as the "flexible" lenses described here. If you use these lenses, picture frames are not required as they are, themselves, fairly rigid - but their "ridged" fronts should still be protected from scratches during transit and from dust when being stored! These lenses appear to have optical properties superior to those of the aforementioned "flexible" vinyl type.
- A word of warning, however: These inexpensive "rigid" page magnifiers seem to come in a variety of focal lengths, ranging from about 9 inches (230mm) to over 24 inches (600mm). The longer focal length lenses are very difficult to efficiently illuminate and will require a very large box! For this reason, I would strongly suggest that one chooses a lens with a focal length of 13 inches (330mm) or less. To determine the focal length, one can simply focus a distant object (say, a building across the street) onto a piece of paper and measure the distance between the lens and the paper: Most of these lenses are packaged in a way that allows such a test to be conducted without removing them from their box. Note: For this measurement to be at all accurate, use an object at least 100 feet (30 meters) distant!
- DO NOT rely on the "2x" or "3x" magnification
rating to tell you something about the focal length of the lens:
I have seen seemingly-identical lenses - with the same focal length -
that are variously labeled as "2x" or "3x"! The only way to be
sure about the lens you receive is to check the focal length as
described above!
- If you are using the "flexible" Fresnel page magnifiers mentioned above, you'll need some 8x10 picture frames. I picked several of the cheapest wooden 8x10 picture frames that I could find - $3.00 each at WalMart. It is important to make sure that they have clear glass as many of them have slightly frosted glass in order to minimize glare. Actually, I picked up three frames (in case I broke some glass) but ended up using just the glass from one of them in order to flatten one of the Fresnel lenses - more on this later - but I should have really picked up four frames!
- Note: If you are using the rigid Fresnel page
magnifiers, these are rigid enough to support themselves and a picture
frame may not be required. Note, however, that doing so leaves
the fronts of the Fresnel lenses exposed, so care must be taken to
avoid damaging the "grooves" in the lenses and to keep
them free of dust! If you do not install protective sheet of
plastic (or glass) in front of these lenses, it is strongly
recommend that one makes a cover - possibly out of posterboard - to
protect the lens from dust and damage, not to mention the accidental
exposure to sunlight!
- Some "Foam Core" poster board. In the stationary aisle
(amongst the school supplies) I found some 20" by 30" (508x762mm) by
3/16" (5mm) thick flat black boards for $2.88 each at WalMart (prices
noted in
April, 2007 and February 2009.)
This material comes in several colors (black and white being the most
common) and consists of a layer of Styrofoam (of the same color as the
paper) sandwiched by construction paper on both sides. It is very
light and surprisingly strong, and being that it is coated with
paper, it is easily bondable with nearly anything. To assemble
this pair of enclosures (one TX and one RX) I used five sheets of this
board - although with more careful planning, I believe that I could
have done it with just four sheets.
- "Grabber" (tm) screws. Also known as "drywall screws" these are aggressively-threaded Phillips-headed wood screws. These are used to lock the focus into place as well as to attach panels, etc.
- It turns out that I also needed to have a secondary lens for the emitter. When I had ordered the Fresnel lenses, I also ordered a bag of 20 double-convex plastic lenses for $3.98. These are 20mm (a bit more than 3/4") in diameter and have a focal length of 34.5mm (about 1-3/8".) These are American Science and Surplus part number 68093 (LINK). Even though these are really cheap, they seem to be of reasonably good optical quality.
- Some glue. I used "Hot Melt" (thermoset) glue with a glue gun to assemble these boxes, although with the paper covers, white glue and pieces of heavy-bond paper could have also been used, provided one allowed plenty of time for each set of glue bonds to dry. Silicone sealant could have been used to provide good bonds - again, assuming that one was willing to do the "glue - then wait until it cures before gluing some more" trick - a process that would require several days of intermittent construction. The thermoset glue is nice in that it's ready after just a few minutes, but the bad part of using thermoset glue is that if this enclosure is left in a car on a hot, summer day, it's likely to disassemble itself. (I may reinforce the glue joints with silicone sealant as a "backup" adhesive...)
- Note: When using the thermoset glue with the foam-core
board, note that it takes several minutes for the glue
to cool and set up! This is because the foam-core board is a good
insulator and does not conduct the heat away! To speed up
cooling, you can use a fan or even a hair dryer on it's "non-heat"
setting.
- A silver-colored marking pen to mark surfaces in an obvious way. (We'll be writing on black surfaces, you know...)
- A pen or pencil: These are for making cutting/measuring
marks during construction that don't need to be readily visible after
you have completed the box.
- Some bits of paper for taking notes.
- A straight edge: A metal yardstick or 24" metal ruler works nicely.
- A sharp utility knife. A utility knife has been found to be more convenient than an a hobby knife set such as an "X-acto" (tm) knife.
- A surface on which to cut the foam core pieces that can be
scratched, but will not ruin the knife's edge. A large scrap
piece of cardboard or Masonite is good for this, and an old rug or
doormat can also work as you don't want to ruin your kitchen table!
Comment: Unfortunately, I did not take pictures of the enclosure in various states of construction - sorry.
Assembly - Front half:
Note: There are two identical enclosures - one for receive, and the other for transmit. After initial optical alignment, they are connected to each other to form one unit.
The description below is for just one of the two identical enclosures. Note also that each enclosure consists of a "front" and "rear" half, nested together to provide adjustable focus.
 |
 |
 |
The first step is to characterize the Fresnel lenses. The full-sheet magnifiers that I used have a focal length of about 310mm - a property determined by focusing an outdoor view onto a sheet of paper and then measuring the paper-to-Fresnel distance. This measurement was very important in determining the dimensions of the box.
After removing the glass from the picture frame, I carefully measured it and then cut four sheets of the foam core board. For the top and bottom (the short, or 8" side of the frame) the board was exactly the length of the frame, but for the two sides (the 10" dimension of the frame) the board was cut 3/8" longer so that there was some overlap with the edges of the short side: This allowed the pieces of foam core board to completely enclose the frame, yet provide an interior surface for gluing. As for the length of these pieces, I chose 250mm.
Comment: As pieces are measured, cut, and fitted, it is strongly recommended that they be labeled and marked with to indication the front, top, sides, etc. - and also to note precisely which pieces are fitted against each other!
Using some tape to hold the sides together, I assembled the four sides of the box and then inserted the picture frame, setting it back from the front by about 30mm: This setback performs a slight function as a lens shroud, but mostly it protects the glass lens during normal handling. Once I verified that the pieces of foam core board were of the right length and that the plane of the picture frame was at a precise right angle to the box (that is, that it was set in straight) I used the thermoset to glue everything into place. When doing this, it is best to use a few small dabs in some strategic places to hold everything together - and then re-measuring to make sure everything is lined up before applying the full amount of glue. When applying the glue (which is hot enough to cause a nasty 2nd degree burn) it is best to glue one or two joints and then wait a minute or two for it to set before turning the box around and gluing a few more joints.
Once this portion was completed, the entire assembly was quite rigid - yet light.
As can be seen from the pictures, each enclosure is really two pieces, nested inside each other with the front half of the enclosure sliding inside the rear half allowing one to adjust the focus simply by sliding the two pieces past each other. Because of this, it is important that the glue joints for the front section are inside the enclosure and that the edges are flush so that the two sections fit as tightly as possible.
While it would have made more sense to make the rear half slide inside the front half (as the rear can be smaller, from an optical standpoint) I did not do this. Why not make the front portion smaller? Well, I was sort of lazy. Because the ultimate size determination is that of the picture frame, the front half had to be constructed first. It was easier to build that portion - and then cut the pieces to size to build the rear portion to fit around the outside of that. It was easier to build the second (rear) portion by wrapping around the outside of the front portion than trying to build the rear portion by assembling it inside the front portion.
Assembly - Rear half:
After the front assembly was completed, a 250mm deep box was wrapped around the already-completed front portion. For this, the side pieces were cut to precisely the length of the sides of the front box portion while the top and bottom pieces were cut a bit longer, allowing some overlap for the glue joint. Using the front box as a form, wooden spring-loaded clothespins and weights were used to assure a very snug fit during assembly. It should go without saying that one should be careful to avoid gluing the two halves of the enclosure to each other.
Unlike the front portion, all of the glue joints were done outside the box as to avoid interfering with the tight fit - and this is why the top and bottom pieces are cut slightly oversize.
Once the glue has set I pulled the two boxes partially apart and traced the shape of the back of the enclosure is traced onto another piece of foam core board to be used as the rear panel: Tracing the size allows for a perfect, snug fit. Before gluing the back panel into place, I glued some 2"x2" squares into the corners about 75mm from the rear edge. These function not only as gusset plates to help hold the shape of the box, but also as stops to prevent the two boxes from being pushed together too far - something that could make them very difficult to get apart.
Prior to installing the rear panel, I located the exact center of it (by drawing an "X" on it) and cut a 100x100mm square hole, centered in the panel, for mounting the emitter and detector circuitry. While you are at it, also draw centered horizontal and vertical lines on the back panel to aid in the location of the its center - a helpful guide when installing the electronics. Installing the rear panel, it was glued (with the lines that you drew on it facing outwards) along both the inside and outside, making the rear box quite sturdy.
Installing the lenses:
After all of the gluing has been completed, it is time to install the lenses. Because these Fresnel lenses are extremely thin and flexible (I believe that they are made of vinyl) they need to be attached to the glass in order to be stable enough to be optically useful: Attaching the lenses was done simply with some pieces of clear tape, centering the lens on the piece of glass. When installing the lenses, make sure that they are as flat as possible before applying the tape in order to provide the best optical properties. It is also important that the proper side face outwards: It was empirically determined that for these lenses, the "grooved" side should face away from the emitter/detector.
While these lenses were precisely the same "long" dimension as the glass, they were slightly narrower, leaving a small gap along the "long" edge. For the receiver, I simply masked this "un-lensed" portion with black electrical tape to minimize ingress of stray light, but I didn't bother doing this with the transmitter.
I noticed that these lenses weren't completely flat and tended to bow outwards in the middle. The obvious solution to this would be to sandwich the lens between two pieces of glass. Unfortunately, I only bought three picture frames and had only three pieces of glass. Reasoning that the best-performing lens should be the one on the receiver, I sandwiched only that lens.
The glass is held in the frames with dabs of hot-melt glue: The enclosure was laid "front-down" so that the glass was in the front and the "grooved" side of the lens was also facing front. For the receiver, the second piece of glass was laid on top of the lens, and the entire thing was carefully weighted down (to flatten the lens between the two panes) using a 7 amp-hour 12 volt lead-acid battery. With the lens secured, the edges were dabbed in several places with the thermoset glue. A bead was not run around the entire perimeter, as this would have made disassembly - if needed - much more difficult.
For the transmitter, because I didn't have another pane of glass handy, so I simply cut a strip of foam core and laid it across the lens edge-wise, gluing the ends of the strip to the frame. This served to hold the center of the Fresnel lens against the glass and removed most of the "bulge". (Note: This strip may be seen in the bottom picture of Figure 2 in the left-hand lens as a horizontal bar.) Prior to final assembly, a brief test was done to note the difference in "beam quality" between a Fresnel lens sandwiched between two panes of glass and one that was not. While the center of the beam wasn't affected much, the periphery of the beam was somewhat "cleaner" with the flattened Fresnel. It was also determined at this time that simply holding the center of the Fresnel against the glass fixed "most" of the degradation. Because of this, I feel confident that for the transmitter, the remaining "un-flatness" is will result in only a small amount of degradation. (If I'd had another piece of glass handy, I would have put it in, but now that it is installed, I won't bother adding one later.)
It should be mentioned again that these flexible "Sheet Magnifier" Fresnel lenses are of rather poor optical quality. In comparison with the rigid, molded Fresnel lenses, they produce a rather "dirty" pattern - most notably a somewhat "fuzzy" image and the tendency to produce a weak "X"-shaped pattern of spurious light. It is mostly for this reason that I didn't go through too much trouble to flatten the the transmit lens.
Comment: On the outside of the front half the enclosure, I drew a line at the location of the plane of the Fresnel lens. This provided a handy reference for measurement of the focal plane of the lens during the "rough focus" setup.
Mounting the electronics:
 |
 |
 |
 |
As mentioned, a 100mm x 100mm square hole was cut into the back panel for mounting the electronics. The detector and emitter assemblies are mounted on a plate made from a piece of foam-core board that is somewhat bigger than the hole - about 150mm square or so. Onto these I drew both an "X" and a cross - on both the front and back of this plate - to locate the center as well as to provide alignment with the center-locating lines that were drawn on the rear panel. The idea here is to align the LED and the detector's photodiode to the exact center of the mounting plate and, with the lines, be able to mount the plate in the center of the enclosure's rear panel.
Figure 3 provides several views of the mounting of the enclosure. The top picture shows the installed panels with the electronics packages. These panels are held in place with "grabber" (or drywall) screws. Note that because this foam-core board is rather fragile, it takes screws with very pointed tips and deep, aggressive threads to hold things into place. Even so, these screws cannot hold into the board material very strongly - but the panels should be very secure as long as reasonable care is taken in handling of the enclosure. In order to improve "hold" several layers of foam board (and/or thin wood) could be glued together inside to provide a stronger grip.
The middle picture shows the detector. This is the original prototype version of the Version 3 optical receiver mentioned on the Optical Receivers page on this site, and it was used simply because it happened to be laying around on the workbench. As can be seen from the picture, some extra pieces of circuit board material were soldered into place to provide mounting from the screws, as well as side-shields to minimize RF and electric field ingress. Note that if this unit is used near a transmitter site, further shielding will probably be required! I was pleased to note that inside my house, it was remarkably unaffected by nearby electromagnetic fields.
The bottom picture shows the emitter assembly - and some explanation is required here. My first experimentations with the 3-watt Luxeons were with this unit - a Luxeon emitter (not a star) epoxied to a small aluminum heat sink. I soon noted that, for continuous operation, that the original, small heatsink was inadequate. When mounting the emitter for this project, I rummaged through my collection of heat sinks and found a larger one (from the same computer monitor, I believe) that had fins that would jam into each other and interlock. I simply smeared some heat-sink compound on the mating surfaces, pushed the two heatsinks together, and then put dabs of high-temperature epoxy on the lot to hold them together. Testing showed that there was excellent heat transfer between the two, thus solving the "small heatsink" problem.
To further facilitate mounting to the foam core board, I cut a small piece of sheet aluminum and used it to spread the distribution of force from the mounting screws (as well as increase the heat sink area) to provide a solid mount. On the rear of the panel, some small washers were also used to distribute the force and prevent tearing through the board. (I may put larger washers on the back panel, though...)
Comments:
- In series with the LED (a Luxeon 3-watt emitter) is an LM317-based current limiter set to allow only up to 2.25-2.5 amps of current to flow. This prevents immediate damage to (or destruction of) the emitter should the modulator malfunction and dump excess current into the LED.
- Philips is apparently phasing out the Luxeon I, III, and V lines in favor of the lower-power Luxeon Rebel devices. Since I have not used those other devices, the techniques described here may not directly apply. For the time being, however, the Luxeon III devices are still available from various sources.
Matching the emitter to the lens:
Once the emitter assembly was mounted, I checked the size of the "circle of light" produced by the Lambertian pattern of the Luxeon emitter at a distance of 310mm (the focal length) and compared it to the size of the Fresnel lenses. This test - along with one using the lenses themselves - showed that most of the light was spilling out beyond the edges of the Fresnels, resulting in a significant loss in available radiant energy and meant that a secondary lens was needed.
It is normal practice to use a PCX (Plano ConveX) lens to change the "spot size" but I had on hand some cheap, plastic DCX (Double ConveX) lenses (20mm diameter with a 34.5mm focal length) that I was dying to use. A quick test showed that this tiny DCX lens was nearly as effective as a much larger, heavier, and more expensive glass PCX lens - and it was obvious that it would be much easier to mount the small, plastic lens.
The "mounting tube" was made by rolling some paper around another tube (a 6LN8 - a pentode-triode combination to be precise) and taping it. I was then able to force the plastic DCX lens into the end of the paper tube (which was fractionally smaller in diameter than the lens itself) which held it fairly securely. I then carefully applied some clear, fast-setting epoxy at the boundary between the paper tube and the DCX lens, taking care to keep as much of it off the optical area of the lens as possible. After a few minutes (and exposure to the warm heatsink of the powered-up LED) the epoxy had set, so I carefully trimmed down the paper tube until the "spot size" of the LED at 310mm approximately matched the size of the Fresnel lens - something that occurred when the lens was about 5mm away from the top of the Luxeon's dome. At this point, the paper tube was epoxied (using the high-temperature epoxy) to the small heatsink and centered over the top of the Luxeon emitter. This technique is simple, crude, yet effective and rugged.
Comment: I believe that significantly better efficiency (more light from the LED being directed toward the lens) could have been obtained if a different secondary lens was used. Because this was a "quick 'n dirty" transceiver, its somewhat inferior - but adequate - performance was the acceptable tradeoff in order to have a 2nd unit onhand for testing.
Aligning the optics - emitter:
Alignment and focusing of the emitter was fairly straightforward - especially after having done so on the previous enclosure. During construction, I was careful to make the picture frame parallel with the front of the enclosure as doing so allows the use of a carpenter's square to determine the pointing of the lens when a laser level is clamped to the carpenter's square - see this page for more details. In a nutshell, the use of the square-laser combination allows one to determine precisely where the lens is "pointed" - and one can easily verify that the emitter is located within the center of the focus spot of the lens.
After verifying alignment, the mounting plate is screwed to the back panel and the beam is focused: Focusing is done simply by sliding the two boxes in and out of each other. While rough focus can be done indoors, provided that one has a fairly long (30 feet or 10 meter) distance, this will only give approximate results. For final focusing, I took the entire enclosure outside and used the side of a house on the next street over (about 300 feet or 100 meters away) as a target, focusing for the "sharpest" square pattern. It should be noted that with the secondary lens, the focus is somewhat "closer" to the lens than the focal length of the Fresnel lens alone.
Once the precise focus was found, I marked it with the silver-colored marking pen and then secured the position with the four "grabber" screws - one on each side of the enclosure.
Comment: As the beam is focused, one should be able to resolve the "square" shape of the Luxeon's emitter. If good quality lenses are used, the bond wire and connecting electrodes should be visible in the focused image at relatively short distances. Best focus is obtained when this "square" pattern is at its sharpest - and this should be done over a large a distance as possible. Considering the accuracy of typical Fresnel lenses, a distance of about 300 feet (or 100 meters) is likely to be "close enough" to infinity to achieve a good focus. With these somewhat poor-quality "sheet magnifier" lenses, the square shape of the Luxeon's emitter was somewhat blurred, but still generally recognizable.
Aligning the optics - detector:
Focusing of the optics - especially the detector - is, perhaps, the most awkward part of the construction of the entire project. Unlike the emitter, there is no spot being cast to show the precise alignment. Now, it should be possible to simply substitute an emitter -located in the precise position of the photodiode- to accomplish this, but I have yet to do so. Using the indoor test range, I was able to get approximate focus - a setting that should be "fairly close" to optimal. Through previous experience, I noted that if one focus in the indoor test range, the proper focus for an infinite distance is one that is slightly closer to the lens.
As for the other enclosure, I checked receiver focus and sensitivity using a dim, diffuse, red LED attached to a wall about 33 feet (10 meters) away. This LED was modulated by a signal generator and by varying the drive, the intensity of the LED could be adjusted from a "dim, but easily visible" setting down to a "I can't see it unless I'm right next to it." As with the other enclosure, an easily-audible (and speech-capable) signal was obtained from an LED that was not readily visible at the full 33 foot distance.
Later, the enclosure was dragged outside. As expected, the waxing moon was more than enough to completely saturate the receiver, and fairly distant streetlights (and flying aircraft) were clearly audible.
Comment: After completing this enclosure, the focusing of the transmitter and receiver was re-checked using some simple equipment on a test range and both the receiver and transmitter were found to be within a fraction of a dB of their optimal settings in terms of focus and parallax.
Marrying the two enclosure:
 |
 |
The experiences of the Australian group (Chris, VK3AML and Mike, VK7MJ) show that it is most convenient to gang two lenses together and precisely align them, thus eliminating the need to aim more than one box. In this case, two separate boxes were built - mainly because it was easiest to do so - but there was still the matter of putting the two together and making sure that they were in alignment.
First, the two enclosures are laid next to each other and just the two rear screws holding in the large, square plate on the top and bottom are inserted: This will loosely couple the two sections together.
As can be seen in Figure 4 some scrap pieces of foam core board were used as "plates" to tie the two together. As with the rear plates and focus adjustments, "grabber" screws were used to secure the plate. On both the top and bottom, two plates were used: A larger, square plate near the back (the one with "TOP" written on it, both normally and upside-down) and a smaller plate at the front of the enclosure: Two similar plates were used on the bottom side as well. In addition to plates on the top and bottom, there are two small plates (or straps) mounted on the rear panel to further improve rigidity.
Because the entire purpose of these plates is to gang the boxes together and ensure stability, it is very important that the two boxes actually be pointed in the same direction and are properly aligned. Note that due to parallax, this must be done with a special target that takes into account the distance between the two lenses. As the distance approaches infinity, the beamwidth of the two lenses will cause overlap, effectively negating the parallax.
The task of alignment was done in a way similar to that done with the other enclosure: A paper target was attached to the far wall (just above the LED used for testing the receiver) and marks were made on the paper: One mark, representing the center of the transmitter beam, was spaced the same distance from the receiver-test LED as the centers of the two lenses, while two other marks placed at the distance from the the center of the transmit lens and the boresight of the laser level to locate both azimuth and elevation of the enclosure.
First, the receiver was carefully aligned on the test LED. Further verification of the receiver's aiming was done by pointing a laser pointer at the test LED: A "hiss" from the laser pointer was noted (along with some receiver desense) and by moving the laser pointer side-to-side and up and down with respect to the test LED, the proper aiming of the receiver could be verified.
Once the receiver was aimed, it was necessary to be very careful to avoid disturbing the position of the receive portion of the enclosure. At this point, the emitter is activated and the transmit enclosure is shimmed and shifted as necessary to get the center of the transmitter's beam to line up with the marks on the paper. As proper alignment is achieved, the other plates may be added to "lock" things into place.
After installing the plates, re-check the alignment once again by peaking the receiver on the LED with the test signal and then verifying that the center of the transmit beam was in the center of its respective target. If minor adjustments are to be done, the screws holding mounting plate or the emitter assembly may be removed, the position of the emitter adjusted slightly, and the screws reinstalled through new holes. Note that if this sort of adjustment is required, it is easiest to do so with the transmitter - after verifying peaking of the receiver - because the transmit beam can be easily seen...
Adding a septa:
As mentioned on the page "Optical Communications for the Amateur" (by Chris Long, VK3AML) a device called a "Septa" may be used to reduce stray light such as that encountered in an urban environment. The structure of a septa is simply that of a series of parallel tubes that only allow light from the direction of the source to enter the detector and it would, ideally, be constructed of thin, opaque material painted flat black to minimize reflection: A series of thin metal plates is often described as it would cause minimal blocking of the desired light. Practically speaking, there is a limit on how narrow the viewing angle of the septa will allow and this angle is directly related to how long the tubes are and how small each of these tubes might be.
 |
 |
 |
Because the optical quality of these sheet protectors is rather poor, they are quite susceptible to light from angles far removed from the desired direction. Because of practical reasons, the length of the septa (and the number of dividers) was limited. Also, because I was using scrap foam core board left over from the construction of the enclosure, I had to take into account the limited amount of material onhand and the fact that the thickness of the sheets would block some of the light.
The installation of the septa is rather simple: Because the picture frame is set back by several centimeters (to protect the glass, mostly) I constructed the septa to simply slide in, being held in place by friction. The structure of the septa is simply a series of rectangular tubes created by dividers cut from foam core board, all held into place with "hot melt" (thermoset) glue.
Although it is difficult to tell from the pictures, you might notice that from the front view, there appear to be fewer vertical dividers than are visible from the rear view, but this is because some of the dividers are about half the length of the others: Simple geometry will demonstrate that the divider placed midway between the two adjacent ones need not be as long as those dividers in order to provide complete blockage.
This septa greatly narrows the view of the lens while blocking, perhaps, 10-15% of the light - less than 1dB, and weighs about 12 ounces (0.3 kg.)
Was the septa worth the trouble?
- Despite the poorer quality of these "page magnifier" Fresnel
lenses, the septa has not been found to be necessary - even in
an urban environment, in the presence of city lights, although its
addition would certainly reduce the hum from them. While it was
tried once, the minor improvement was not judged to outweigh its
awkwardness, so it has never been used since!
Another accessory (not yet shown) is an optical attenuator plate . Designed to be used in an urban environment over fairly short distances, this "optical attenuator" is simply a piece of cardboard with holes in it. This attenuator plate is placed against the Fresnel lens, blocking 90-95% of the incoming light. Because of the holes are distributed over the same area as the Fresnel lens, the capture area of the lens remains about the same and most of the reduction of scintillation provided by the larger area is maintained.
Because of the relatively high light pollution encountered in an urban environment, the limiting factor in the sensitivity of the receiver is not the sensitivity of the detector itself, but the amount of "dilution" caused by other light sources. When this attenuator is used with a septa, the beamwidth of the septa itself is narrowed somewhat by the holes: If two matching attenuator plates are used with a septa (one at the lens and another in front of the septa) then the beamwidth may be narrowed even more.
What this means is that if the background light is well above the noise floor of the receiver, dynamic range of the receiver itself can be compromised by this extra light. Also, at relatively short distances, the full sensitivity of the receiver may not be required and the attenuation of "background" noise sources might be welcome.
Like the septa, the optical attenuator was only tried once as an experiment and is not among our normal compliment of equipment.
Fine-tuning the elevation:
For our first optical contact, I used the wooden enclosure while Ron and Gordon took this enclosure to a point across the valley. In setting up, they noticed two problems that complicated the precise aiming of the enclosure and keeping the aim steady:
- The bottom of the enclosure was not flat: When set on a
flat surface, it tended to rock back and forth. Fortunately, Ron
brought some cedar shims and was able to wedge the enclosure so that it
could be stabilized. The fact that it was being set up on a
rough, stone wall made the setup and aiming even trickier.
- While azimuth adjustment was accomplished simply by turning the box left and right, adjusting the elevation was very difficult, but accomplished with the help of the cedar shims. Even when aimed, no-one dared touch the unit out of fear of disturbing the aim as it was very easily moved.
Spot Quality comparisons:
As was expected, the higher-quality optical acrylic Fresnel lens used in the wooden ("first") enclosure produced better-quality "spots" than the vinyl "full-page magnifier" lenses. During testing, I decided to do a direct comparison between the two.
 |
 |
Ideally, one would have done everything possible to make the two tests equal, but there were some unavoidable differences that may affect the accuracy of the direct comparison of intensity:
- It would have been ideal to have used exactly the same LED for both tests, but this would have been difficult. Instead, the same LED current was used (1.1 amps) for each test. Because the two LEDs are believed to from the same production run, it is likely that their optical outputs are very similar, given the same operating current.
- Different secondary optics were used. Ideally, the same set of optics would have been used for both, but this wasn't very practical. In the case of the wooden box, a large (48.5mm diameter with a 51mm focal length) glass PCX (Plano-ConveX) lens was used to optimally focus the LED's output onto the Fresnel lens, while for the second "cheap" enclosure (the one using the Vinyl page magnifiers) a 20mm diameter acrylic DCX (Double-ConveX) lens with a 34.5mm focal length. It is likely that in this second case, there is some inefficiency in "photon transfer" because of this non-ideal lens.
- Precise focusing of the secondary lens:
- For the first (wooden) enclosure, two different focus points were manipulated: The spacing of the secondary lens to the LED, and (independently) the spacing of the assembly consisting of the LED and secondary lens from the Fresnel lens. For the first part, the spacing between the LED and the secondary lens affected how much light was reaching the Fresnel lens: If the secondary lens is too close to the LED, the back of the Fresnel is "overilluminated" with much of the light spilling out past the sides, while if the distance between the LED and secondary lens is too great, the Fresnel is "underilluminated" with only a smaller portion of the center of the Fresnel getting light - something that will also reduce the effective aperture of the Fresnel and increase the severity of scintillation. The latter case has another downside: As the spacing between the LED and secondary lens is increased, the virtual spot size of the LED also increases, which also increases beam divergence. For best results (e.g. maximum luminous flux at a distance) a compromise must be reached between proper illumination of the Fresnel lens and virtual spot size. Essentially, one adjusts both parameters until maximum flux at the distant target is achieved.
- For the second enclosure (the one made of posterboard - the very one described above, on this page) the adjustments were rougher, going mostly by "eyeball." I have little doubt that a couple of dB of improvement in distant flux density could have been achieved - but since it is considered to be only a "low performance" device that was thrown together quickly, I have no plans to try further tweaking to make it work better.
- It should be noted that both boxes were carefully
focused (e.g. the distance between the LED/secondary lens and the
Fresnel) to achieve the best optical output.
Of more interest was the "quality" of the spots that the two boxes produced even when optimally focused. As can be seen from Figure 6 the "main spot" (the brighter "square" portion) is almost identical in size, but the lower spot (from the page-magnifier Fresnel lenses) is not only dimmer, but more light is spread out beyond the main beam perimeter. It should be pointed out that this effect is apparent not only from the pictures, but is arguably more visible when view with one's own eyes. Although not visible in the picture, there is a sort of faint "X" pattern weakly emitted from the page-magnifier lenses that seemed to be totally absent from the higher-quality acrylic lenses.
Again, it should be noted that in the case of the "fuzziness" of the spot of the vinyl page-magnifier lens, this peripheral energy could not be removed by adjusting focus. In the case of the images in Figure 6, the spot was projected onto the surface of a 33 foot (10 meter) diameter white satellite dish that was about 200 feet (60 meters) distant. At this distance, the beams had "mostly" collimated, but were very slightly out of focus as compared to the normal test distance that I'd been using of about 525 feet (160 meters.) The satellite dish was chosen because it was the only relatively large, flat, white surface that was available at a reasonable distance. Note: One can see the subreflector assembly in the bottom of these pictures, along with some lines from the sections of the main reflector as well.
Comments:
Optical quality:
- As mentioned before, the optical quality of these "full sheet magnifiers" is much poorer than that of the thicker, molded Fresnel lenses. It would appear that while the main beam is of similar intensity, they suffer from a larger amount of weak, spurious side responses. Fortunately, these contain relatively little of the total energy, but on receive, they could result in a response to off-axis light sources that could prove to be a source of interference or desensitization.
Weight of the enclosure:
- With the receiver and emitter installed, the entire enclosure weighs in at about 6-7 pounds (approx. 3kg).
Safety concerns when in direct sunlight:
- Because the electronics are permanently installed and the material used for construction is so lightweight, instant damage could be done if the lens is aimed anywhere near the sun - so don't!
Keywords: Lightbeam communications, light beam, lightbeam, laser beam, modulated light, optical communications, through-the-air optical communications, FSO communications, Free-Space Optical communications, LED communications, laser communications, LED, laser, light-emitting diode, lens, fresnel, fresnel lens, photodiode, photomultiplier, PMT, phototransistor, laser tube, laser diode, high power LED, luxeon, cree, phlatlight, lumileds, modulator, detector
Copyright, KA7OEI 2007-2009,
Last
update: 20091230