Lit Plexiglas Experiments
My objective was to find the magic combination of parameters that would provide the most consistent light through Plexiglas. After about 20 different experiments, over 40 photographs in controlled lighting environment, I think I've got it. Some of the crazy, and not so crazy ideas are discussed here. If you want to skip to the punch line, read the last couple of paragraph. Things I tried include:
LED Interface with the Plexiglas
Tight hole vs. loose hole
Blind hole vs. through hole
Epoxy vs. no Epoxy
None of these showed any significant difference. The through hole was a bit better, but not much.
Angles
90, 75, 45 degrees toward the viewing surface
45 degrees away from the viewing surface (bouncing off back side)
Of these, the 45 degree toward the viewing surface showed the most promise.
Polished vs. Smooth
This was very revealing. The light will bounce off (stay inside) if the angle of incidence was less than around 50 degrees. It caused lots of hotspots and was hard to control. It would be very bright in one spot, but would be almost completely dark ¼” away depending on the geometry. The convex shapes made this less apparent, but it was still too harsh. On the other hand, the roughed up surfaces caused the light to distribute more evenly but was not as bright. For both the polished and rough surfaces, putting a piece of tin foil on the surface made things slightly (maybe 10%) brighter.
I also tried roughing up the surface of the LED's. This made the light much more distributed, reducing the hotspot effect, but brought down the brightness.
Custom Shapes:
Lots of crazy stuff was tried. Concave top surface, convex top surface, rounded corners, 45 degree corners, mushroom shape with light aimed toward the back, etc, etc. etc. In all, I made 16 different test pieces. Of all the parameters played with, the shapes had the greatest effect at changing the light characteristics.
Conditions:
I rigged up a couple of pieces of black construction paper; one vertical, taped to a box, and another horizontal, taped to the table. I made small corner stops for the Plexiglas and camera. This made each shot very consistent regarding the distance and image size. It was handy for merging the images together for comparison. Each piece of Plexiglas was 3/8” thick by 2” by 4”. In all cases, the Plexiglas protective paper remained on the sides. I used the same LED for most all of the experiments.
I kept notes in a notebook. Each shot was numbered, and documented in a fairly crude manor. In general it included the picture id, polish/none, test piece #, and/or other attribute. On a red stick-um I wrote the picture id to know which picture was which.
Conclusion:
I was not completely satisfied with any of the results. A couple of the custom shapes came close to getting an evenly distributed light. But they were finely tuned geometrically. I was concerned about how to scale it. For example, I wanted to be able to make a 2 inch light, or maybe a 2 foot light. The custom geometry approach was not going to scale well. I was thinking about putting LED’s every couple of inches, but that would lead to hotspots. I could put a bunch of sanded down LED’s about 4 inches away from the back, and it would be consistent, but would not be practical to mount.
On a whim, I tried putting the LED’s ¾” apart. Perfect, well almost. After all this time in the shop with compasses, sanders, Forstner bits, etc, the best result came from bruit force. A simple line of evenly spaced LED’s ¾” apart. Looking straight on you could barely make out the different light sources (much less than any other test). This went away by lightly sanding each of the LED’s. The light defused a bit more causing a lot of overlap between the LED’s light. Sure, I lost a little intensity by sanding them down, but with a whole lot more LED’s, it is VERY bright.
Best Results:
For a 2” thick slab with a rough sanded face, rough sanded LED’s epoxied in ¾” apart. Holes drilled about ¼” deep using a #9 bit. If using a thinner slab, you might want to mount the LED’s a little closer together to avoid hotspots.
I took a lot of photos of the process, but will not bore you with the details. A few choice ones are presented below.
1) Polished vs. rough 45 degree backside piece. The brighter one was from the polished piece.
2) The set-up
3) Various custom shapes attempted
4,5) The best resulting experiment. The right side is a little dim, but it is because I only had 7 LED’s wired up, but needed an 8th in the last hole.
With these experiments complete, I can prep some Plexiglas pieces and get them mounted in the CP and Bezel. Now back to your regularly scheduled cab build.