Light Box

Another televisor build using club circuits. Geared towards identifying where to get all the materials, tools, how to 3D print stuff, and troubleshoot.

Moderators: Dave Moll, Andrew Davie, Steve Anderson

Light Box

Postby Andrew Davie » Wed Feb 15, 2017 8:28 pm

First thing I'm going to 3D print is the light box enclosure for the LED array. Inside the box I plan to put mirrors so that there will be more light escaping through the front. Also, I think it should be fairly straightforward to design the out "window" for the light to be the EXACT size of the frame (that is, input the hole diameter, first hole radius, last hole radius, number of scanlines, then it should build a box exactly right. So, this is where I'll do that. We'll be using OpenSCAD as our "programming language" to create STL files (which are suitable as pre-processing files for printing on a number of printers). Also, I plan to put a few "shelves" inside the box to allow various diffusing layers to be placed inside. Should be a bit of fun and a good primer to using OpenSCAD and 3D printing. I'll put the item up on Thingiverse but should also be available in this thread too. I'll include "NBTV" in the names of any printable items I put on Thingiverse so they're easy to find.

Right, there are a number of measurements that are going to drive the design of the "box" - and these are "parameters" for the OpenSCAD "program" I'll be writing. Let's list those.

1) radius of scanline hole
2) radial distance from disk center to first (inner) scanline hole
3) radial distance from disk center to last (outer) scanline hole
4) thickness of box wall
5) "shelf" positions (there may be more than one)

The current thinking is that the mirror bits are optional and glued in around the shelves
We'll assume that the inside-back does not need a mirror, so the wires can enter anywhere. The circuit board will mask most light going backwards anyway.

Those above parameters will give us the window size, but we need these things too...

6) LED circuit board dimensions (x,y,z)
7) LED circuit board thickness (so we can make a slot for it to slide or clip into)
7) distance of top edge of circuit board from exit window
8) diameter of the power wires, so we can build a nice exit for them from the box.
9) location of the wire exit points.

I think making the light box the minimal dimension possible, encapsulating the contents, makes sense, so the X/Y/Z of the box will be derived from the other parameters.
It makes the assembly assumptions easier (that is, that no mirror on the back inside is required).

Finally, I think a screw-in backplate makes sense because that gives easy and sturdy access.
I think I've got all that I need, so I'll start making some measurements using my existing LED array. Let me get a picture...

Image

OK, that was interesting. This was actually pretty well constructed. 40 LEDs - no wonder it's so bright. According to their 55000 mcd spec (Yep I know fairly meaningless) but anyway, just saying... that would be 2200000 mcd. Ha! Basically it's a standard "experimental prototype board" with holes, mounted with a closely spaced array of very bright LEDs in a staggered array of 8 strings of 5 LEDs on a board dimensions 70 mm x 40 mm give or take in a box 80 mm x 55 mm x 30 mm and held together with electrician tape. For a diffuser I used some dog-food-can plastic lids and some greaseproof paper. I also have silvered card in there, acting as a mirror on the sides. Interesting. It was a pretty good display (incredibly bright, even in bright sun) so I'll be duplicating that I think.

By the way, ALL units will be metric as far as possible.

So, thinking about the curved auto-calculated framing window - perhaps that's a bad idea :( -- it would require the angle of the light box to be set to the tangent of the midpoint of its position, and we don't want that, really. Unless we design things so that's actually what we DO want. Mhhm. That would be the requirement that the center point of the curved lightbox window was horizontally coincident with the center of the disk, or vertically coincident. That's not really that hard to arrange, I suppose. One could always put in an option for a custom size window with no curvature, so let's go with the speccy one for now.

One thing I considered doing is having a moveable light box mount so that you could swing from horizontal position at the side of the disk to vertical position at the top of the disk, so you could watch stuff in either aspect. Will keep that in mind :)

Right, well that's enough to make a start. I'll head off and do a bit of OpenSCAD programming, put that up on Thingiverse, and post a link. From then on, the up-to-date version will be posted there and well, everyone can print their own :P Actually, Thingiverse now lets you order prints of the stuff on their site, so if you don't have/want a 3D printer, you can hopefully get this printed without too much trouble. Of course I have to design something first.

So, be back soon!
Attachments
box1.jpg
(498.51 KiB) Not downloaded yet
User avatar
Andrew Davie
"Gomez!", "Oh Morticia."
 
Posts: 1590
Joined: Wed Jan 24, 2007 4:42 pm
Location: Queensland, Australia

Re: Light Box

Postby Andrew Davie » Thu Feb 16, 2017 12:14 am

OK, I've done the OpenSCAD programming for the parametric lightbox. "Parametric" means that lots of the dimensions are defined as constants, and the program will correctly resize everything else when you change one of those. It's pretty cool. Currently setup for the original LED board I did, it will be really easy to change. I did a small video of the design - I'll have a go at printing it tomorrow. What you see if you watch the clip is a screen rendering of the design, being rotated. There are two parts - the box and the lid. They are side by side so they can be printed flat. The box has a curved aperture window which is exactly the right arc shape for whatever Nipkow disk dimensions you enter as the parameters. It has a bit of a tolerance too, so you can get it to display a bit more than a frame in size. At the end of the video (after a bit of a delay) you can see a pink thing inside - that's my mockup of my LED board, to give some idea of how it sits in the box. Finally, inside the box are a number of shelves; these are to hold the plastic diffuser shelves (if you want them). Each "layer" on each shelf would be glued onto the shelf so it holds firm. Finally, although the backplate is snap-fit (should be really tight), I've put a screwdriver notch in the base so that you can leverage it off. The shelf inside prevents light entering/escaping through that slot. For wire entry/exit, just drill holes where you want 'em.

For the video see Screenshot of NBTV Lightbox v0.0
The source code is attached, if you want to play with it, or print it.
Attachments
LightBox.scad
(4.94 KiB) Downloaded 526 times
User avatar
Andrew Davie
"Gomez!", "Oh Morticia."
 
Posts: 1590
Joined: Wed Jan 24, 2007 4:42 pm
Location: Queensland, Australia

Re: Light Box

Postby Klaas Robers » Fri Feb 17, 2017 3:50 am

Andrew, do you know the princples behind the optical sphere of Ulbricht? It is a (large) sphere, painted white at the inner side, a kind of white that "reflects" incident light by almost 100%. As far as I know it is Magnesium Oxide, a very white powder. Each time light hits that inner surface, all directivity of the light is removed, it scatters it back in all directions.

Spheres of Ulbricht are used to measure the light output of lamps. If you place a lamp inside the Ulbricht sphere, all light is running inside the sphere from wall to wall and cannot escape. If you look into the sphere through a small window, such that you cannot see the filament of the lamp directly, it is just white inside and you don't see any structure. Everything is equally white (or red if you placed a red lamp).

One of the most difficult things of a light box is that you can see the the individual LEDs. You may try to diffuse that by placing a diffusing frosted glass in front, but you still will see the individual beams of the LEDs. This is especially if you use red, green and blue LEDs.

But if you use the principle of Ulbricht, so you make a white box, in which you cannot look directly into the LEDs, you can get a perfect mixing of the colours. Then the light is trapped in the box and can only escape via the output window. This implies also that you should paint your Nipkow disc white at the back side. Then 99% of the light is scatterd back into the light box.

I haven't tried it, because I didn't know yet how to construct the light box. But 3D printing is a good idea..... I know Open SCAD and constructed a few things with it.
User avatar
Klaas Robers
"Gomez!", "Oh Morticia."
 
Posts: 1656
Joined: Wed Jan 24, 2007 8:42 pm
Location: Valkenswaard, the Netherlands

Re: Light Box

Postby Andrew Davie » Fri Feb 17, 2017 8:45 pm

I did a 3D print today and it turned out pretty good. I had to cut off one of the shelves manually to fit the circuit board in, and I've rebuilt the model (already updated on Thingiverse) to change how the backplate is attached. But yeah... not bad. But I got to thinking - I have several layers of semi-transparent plastic inside the box design to diffuse the LED lights. I got to thinking - what if I filled the gap between layers with tiny frosted glass spheres. That would bounce the light all over the place surely, and act as a great diffuser. Something like this eBay item. Anyone have any thoughts on this idea? Basically, the LEDs on the board, then in front of that a plastic sheet, then a packed area of spheres, then (possibly) another plastic sheet, then an air gap (or more spheres) and then the final plastic sheet over the arc/display window. All of the spheres would therefore be sealed into little pockets between sheets inside the display box. I'm going to do an order just to see.
User avatar
Andrew Davie
"Gomez!", "Oh Morticia."
 
Posts: 1590
Joined: Wed Jan 24, 2007 4:42 pm
Location: Queensland, Australia

Re: Light Box

Postby Andrew Davie » Fri Feb 17, 2017 9:14 pm

Here's a few views of the 1st 3D prototype light box. Great that the measurements of the circuit board entered into the constants in the SCAD program translate into pretty much exact fits after printing. Here you can see the original LED array I did 10 years ago sitting inside the prototype on its little shelf with a lovely tight fit doesn't need to be glued or attached in any way. I'll be making a brand new LED array, and doing some mods on this box design because, basically, I can and it's easy. What I'm thinking at the moment is putting a "hub" around the back of the Nipkow disk, and a swing-arm going to the light box, so that the light box can be pivoted 90 degrees for viewing at top or right edge of the Nipkow disk. The swing-arm being hollow, the LED wires can go through that and be hidden from sight. I can also build some sort of IR emitter/detector enclosure attached here, too, so it always stays in the same place relative to the position of the LED "window". Keep watching this thread if you want to see that :)

IMG_20170217_192546.jpg
IMG_20170217_192546.jpg (315.6 KiB) Viewed 14016 times


IMG_20170217_192441.jpg
IMG_20170217_192441.jpg (316.94 KiB) Viewed 14016 times


IMG_20170217_192613.jpg
IMG_20170217_192613.jpg (312.55 KiB) Viewed 14016 times


IMG_20170217_192450.jpg
IMG_20170217_192450.jpg (381.44 KiB) Viewed 14016 times
User avatar
Andrew Davie
"Gomez!", "Oh Morticia."
 
Posts: 1590
Joined: Wed Jan 24, 2007 4:42 pm
Location: Queensland, Australia

Re: Light Box

Postby Andrew Davie » Fri Feb 17, 2017 9:18 pm

Klaas Robers wrote:Andrew, do you know the princples behind the optical sphere of Ulbricht?


Hi Klaas - hadn't heard of it, so thanks for pointing it out. I suspect overkill for NBTV - with my last LED array (which was incredibly bright) just two layers of frosted plastic + 1 layer of grease paper *mostly* removed the individual LED visbility. I suspect my new version will be a big improvement on that, with a couple of ideas I have. One of those is to simply spray-paint a silver/metallic paint inside the box to make it much more reflective - and of course the frosted glass spheres that I mentioned a short while ago.
User avatar
Andrew Davie
"Gomez!", "Oh Morticia."
 
Posts: 1590
Joined: Wed Jan 24, 2007 4:42 pm
Location: Queensland, Australia

Re: Light Box

Postby Andrew Davie » Mon Feb 20, 2017 1:25 am

First pass of making the interior of the prototype light box reflective. I had some chrome paint so I thought that would be good enough. I'll give it a couple of coats and the surface should smooth out nicely. Certainly much easier than trying to find and cut mirrors or silvered paper to size. I'll give a couple of coats and see how that goes. Those round 'pads' on the corners are just for keeping the 3D print from warping when I print; they cut off really easily... I just haven't done that yet. This is prototype 2 with the new lid design - press-fit and removes the need for the shelf previously required for the backplate. Now there is no restriction for an exact-size circuit board being inserted.

IMG_20170219_234748.jpg
IMG_20170219_234748.jpg (331.43 KiB) Viewed 14006 times


IMG_20170219_234752.jpg
IMG_20170219_234752.jpg (333.69 KiB) Viewed 14006 times
User avatar
Andrew Davie
"Gomez!", "Oh Morticia."
 
Posts: 1590
Joined: Wed Jan 24, 2007 4:42 pm
Location: Queensland, Australia

Re: Light Box

Postby Harry Dalek » Mon Feb 20, 2017 10:42 pm

Hi Andrew
if you got some 3 colour leds you could make it colour as well ,you might like to try in the future if your interested .
I like pure B/w as its simple but i wouldn't mind trying colour one day .
The electromagnetic spectrum has no theoretical limit at either end. If all the mass/energy in the Universe is considered a 'limit', then that would be the only real theoretical limit to the maximum frequency attainable.
User avatar
Harry Dalek
"Fester! Don't do that to 'Thing'"
 
Posts: 5364
Joined: Fri Sep 26, 2008 4:58 pm
Location: Australia

Re: Light Box

Postby Andrew Davie » Mon Feb 20, 2017 10:55 pm

Harry Dalek wrote:Hi Andrew
if you got some 3 colour leds you could make it colour as well ,you might like to try in the future if your interested .
I like pure B/w as its simple but i wouldn't mind trying colour one day .



In fact I have spent the past day or so thinking about colour. I even did some reading about how it's done, but still a very basic understanding.
Would love to collate colour information into a separate sub-forum.
User avatar
Andrew Davie
"Gomez!", "Oh Morticia."
 
Posts: 1590
Joined: Wed Jan 24, 2007 4:42 pm
Location: Queensland, Australia

Re: Light Box

Postby Harry Dalek » Tue Feb 21, 2017 8:08 pm

Andrew Davie wrote:
In fact I have spent the past day or so thinking about colour. I even did some reading about how it's done, but still a very basic understanding.
Would love to collate colour information into a separate sub-forum.


Yes that would be good for any one looking for colour ,easiest would be garys 2 colour ...one audio channel driving one colour the second channel the other colour ..

Gary has a 3 colour 32 line in video2NBTV v3 so far you would need to make a monitor to use it i think , i can only think the 3 colours per frame 4 hz per colour ?be very flashy oh my head hurts thinking about it !

I think the trick of using 2 colour red blue and making 3 colour ? adjusting the percentage of light from those leds and mixing green leds as the difference from both the red and blue ..bit like pal tv only transmits 2 colours the difference is Green ..
I tried another colour mix comes out better viewing here on line ..i am new to all this too
http://gifmaker.cc/PlayFrameAnimation.p ... ZVClxBDPRR
Attachments
robe-additive-blue.jpg
robe-additive-blue.jpg (9.19 KiB) Viewed 13987 times
robe-additive-red-x.jpg
robe-additive-red-x.jpg (8.72 KiB) Viewed 13987 times
robe-additive-green.jpg
robe-additive-green.jpg (10.38 KiB) Viewed 13987 times
The electromagnetic spectrum has no theoretical limit at either end. If all the mass/energy in the Universe is considered a 'limit', then that would be the only real theoretical limit to the maximum frequency attainable.
User avatar
Harry Dalek
"Fester! Don't do that to 'Thing'"
 
Posts: 5364
Joined: Fri Sep 26, 2008 4:58 pm
Location: Australia

Re: Light Box

Postby Klaas Robers » Thu Feb 23, 2017 1:17 am

Harry, you always NEED 3 signals:
- R, G and B, or
- Y, R-Y, B-Y as in PAL NTSC and SECAM, or
- Y, R, B, or
- any other sensible combination of 3 signals.

In PAL, NTSC and SECAM the Y signal is the black and white representation of the colours R, G and B, where: Y = 0.6G + 0.3R + 0.1B. That is how our eyes see the black and white equivalent. This is needed to come to a compatibility with the existing B&W sets.
For the compatible NBTV-system we choosed for 0.95Y = 0.5G + 0.25R + 0.125B. Digitally this is much easier qua implementation. In all systems you can recalculate R, G, B. In analogue colour TV this is done with resistor values, which add currents in certain ratios and then the absolute multiplication numbers are not so important. Digtally the factors of two make life very easy, just shift over one bit position.

In the compatible system of Vic and me, we encoded the Y signal in the 9 most significant bits of the 16 bits wide words of the L-channel, and R, sample alternated with B, in the 7 least significant bits. For (NBTV) video there is no need for 16 bits depth (resolution) words, for B&W it points out that 7 bits is enough. For sound we NEED 16 bits depth. That we use 9 bits for Y is because of the compatibility with black and white. In the original B&W NBTV CDs the video is encoded as 1 volt positive and the sync as 0.4 volt negative. In this way only 1/4 of the audio range is used. That spoils 2 bits, so in practice the Y signal is encoded also in 7 bits.
User avatar
Klaas Robers
"Gomez!", "Oh Morticia."
 
Posts: 1656
Joined: Wed Jan 24, 2007 8:42 pm
Location: Valkenswaard, the Netherlands

Re: Light Box

Postby Andrew Davie » Fri Feb 24, 2017 12:14 am

Klaas Robers wrote:For the compatible NBTV-system we choosed for 0.95Y = 0.5G + 0.25R + 0.125B.


I don't understand this! 0.5+0.25+0.125 are 2* each, OK - but the sum is not 0.95. It is 0.875
Are you saying you chose arbitrary amounts, in the right proportions, and then calculated what that was in terms of Y?
User avatar
Andrew Davie
"Gomez!", "Oh Morticia."
 
Posts: 1590
Joined: Wed Jan 24, 2007 4:42 pm
Location: Queensland, Australia

Re: Light Box

Postby Klaas Robers » Fri Feb 24, 2017 8:12 am

No, you are right, Y-max is 0.875. It will say that the luminance, the black and white compatible video signal in the 9 most significant bits never reaches to 1.0, but goes until 0.875. Nevertheless this is a slightly higher voltage than on the previous black and white NBTV CDs. So the contrast of the B&W monitor should be set slightly lower. This move was done to make the first decoder in CMOS logic practical realizable.
User avatar
Klaas Robers
"Gomez!", "Oh Morticia."
 
Posts: 1656
Joined: Wed Jan 24, 2007 8:42 pm
Location: Valkenswaard, the Netherlands


Return to Andrew Davie's Televisor #2 (2017)

Who is online

Users browsing this forum: No registered users and 1 guest