# LED Light Table

Joined May 11, 2008
17
Hey all,
Finally getting going on this, now that i've found out that i've passed all my exams.
Basically, I'm making a Light Table (Or Light Box) using LED's. Traditionally, these are made using fluorescent lights. However, when you're tracing drawings for 5 hours over them, you roast and become very dehydrated. So i figured one out of LED's would be a)less painful on the eyes b)lighter, skinnier, and generally more mobile, and c) less roasting.

Basically, im not too knowlegable about circuits. I know a limited amount, but not enough to make this circuit. I've a few questions, so maybe someone could help me out:

I did some research, and i've decided to use a normal DC adaptor. The wire itself coming out of the plug is split in 2, so i'm guessing thats a + and - wire. (is there some way to tell which one is which?) It's an old Minidisc charger. The other problem with this is that it's only a 3v charger. My questions for this is, 1) How can i tell which wire is + and which is - (One of the wires has a white stripe on it by the way) and 2) will 3v be enough to power the leds?
All the info text on the lable says:
Input: AC230-240V~ 50Hz 5W
Output: DC3V - 500mA

Problem 2: LED to Dimension Ratio??
Ok, so the dimension of the illuminated surface is 600 x 900 mm (or approx 64 x 35 inches). Now, i know how to diffuse the LED's. They're actually going to be sort of dual-diffused. Firstly, the LED's themselves will be diffused, but also, the plexiglass they'll be illuminating will be diffused too. The main problem here, is deciding how many LED's ill need. Fixing them to the surface is no problem at all. They'll be approx 50-75mm (2-3 inches) away from the plexiglass.

Problem 3: LED colour?
This is a tricky one. White would be the obvious choice here. HOWEVER, i also came across an interesting article which explained how blue led's are better for your mood than any other light. To quote: "446 to 477 nm wavelength light is supposed to be the most effective for treating seasonal affective disorder (second to natural sunshine, of course)." which is blue.... SO the new problem is, do i chance Blue LED's? They probably would be bright enough that we could still see the pencil through it, right??

Problem 4: On/off switch?
With this one, i was just wondering how hard it would be to put a simple on/off switch in to the circuit?

AND FINALLY, Problem 5: The circuit
All these things in mind, could someone possibly give me an idea of what way to go about the circuit, like for resistors, series/parallel etc? i just do not know where to start.

Thanks all,
tm

#### Wendy

Joined Mar 24, 2008
21,914
That wall wart isn't going to be very efficent. LED's are current devices, and newer types drop over 2.5VDC each. Generally it is more efficent to have a row in series and regulate the current for all of them. This can be done with precision with a really simple current regulator like a LM317, or a simple resistor. I would recomment something like a 24VDC power supply, but this is a personal preference. The reason for the higher voltage is fewer parts, your power supply would require a resistor for each LED, and at 20ma per LED you would max out around 25 LEDs (actually less).

You'll need to research some datasheets from LED manufactures, although I suspect some of the other guys will be of more help there.

A simple toggle switch is cheap and easy, and can be had from Radio Shack or anyone else, so no sweat there.

Here is a circuit we worked out for another poster, but the resistance will have to be adjusted to match your specific LEDs.

#### Wendy

Joined Mar 24, 2008
21,914
Those power supplies would work fine, that isn't too critical. I can't speak to the color and suitability of the LEDs, but their forward drop is 3.9 V and the recommended current is 20ma. This means you could have 5 LEDs per chain using a 220Ω resistor. The 200ma supply could drive 10 chains, but I wouldn't use more than 8 for safety. The 400ma could drive twice as many, but isn't regulated, so you would need to measure the voltage drop across the resistor and calculate the current, I'm assuming you know ohms law.

Current = Voltage across resistor / Resistor value

One other thing, these LEDs have a fairly tight beam of light coming out of them, so you will probably need a diffusion screen. They are simple sheets of plastic that are pretty cheap to buy from local hardware stores.

Last edited:

Joined May 11, 2008
17
5 chains of 8... hmmm... 40 led's should be plenty, provided they're diffused well, right??

#### Wendy

Joined Mar 24, 2008
21,914
I would thing so, it will very bright, but even with diffusion it will be a pretty localized, almost like a spot light. If you built this it would be interesting to see how it works out. What part of the world are you? The 220VAC gives some clue.

Just a suggestion, there are a lot of smarter guys on this site, give it time for some more feedback to be posted. I'm sure there will be other ideas.

Joined May 11, 2008
17
im in ireland. The lightbox is for university. I'm studying architecture, so there's a lot of tracing.

#### Wendy

Joined Mar 24, 2008
21,914
OK, I just figured out what you were doing, I thought this was a SAD light box.

#### SgtWookie

Joined Jul 17, 2007
22,210
The viewing angle of the LED's you're considering is 4°. You would wind up with 40 moderately bright dots on your plexiglas diffuser. You would not have room enough inside the box to get far away enough from the diffuser to spread the beam.

You need more like a 80° to 110° viewing angle. Since you are familiar with architectural drafting, you must have at least a fleeting acquaintance with mechanical drawing. Figure out how far away LEDs with that kind of viewing angle would have to be from your plexiglas diffuser in order to ensure a reasonably even illumination and diffusion.

Power supply: it's not really necessary that the power supply itself be regulated, if you use current regulation in each string, like Bill was saying. For the most simple current regulation that would give a stable current, an LM317 or LM317L per string would work. However, that leaves you with no practical means of lowering the intensity of the LEDs should they prove too bright. For something like that, you would need a PWM drive in addition to the current regulation.

Also, you may not be happy with all blue LEDs. You may wish to use both white and blue LEDs, but have separate intensity controls so that you can vary the color anywhere between all white and all blue.

Before you invest in dozens of LEDs, I suggest you just obtain one or two of various colors that are of the superbright variety. Be warned that you should not look directly at superbright LEDs for more than a moment, as you can permanently damage your eyesight.

#### Ron H

Joined Apr 14, 2005
7,014
Have you considered edge-lighting the panel? It seems like this could give you more even illumination.

#### Audioguru

Joined Dec 20, 2007
11,251
Those LEDs are in a clear case, not a diffused case.
Their light beam is extremely narrow at only 4 degrees. Get LEDs in a diffused case with a 40 degrees or more beam.

Joined May 11, 2008
17
ok, so here's where i've gotten to so far.

I know the led's are clear cased. There's a very easy way to diffuse them using sandpaper.

Secondly, i've come across micro LED's with a 120 degree viewing angle, and they're not super bright (see here: http://ie.farnell.com/1367505/optoelectronics/product.us0?sku=dialight-corporation-5989290102f )
Could these work??

Also, i'm really confused about the whole regulating thing... I thought it was just a question of basically DC + > on/off switch (/ground) > strings of [resistor + led's] > DC -

#### Wendy

Joined Mar 24, 2008
21,914
I wouldn't worry too much about the regulation part of it, the LEDs are where it's at. I like the suggestion about getting a few to try out, there are LOTs of flavors out there, shop around. Some of the guys might be able to recommend better sources.

Your basic concept of power supply, switch, banks of resistors/LED chains seems pretty on track to me. The diffusion is another issue, and can be a big one, so take your time and do it the way you really want it to be.

Figure the forward voltage drop is additive (5 X 3.9V = 19.5V), the difference in voltage will be how you pick the limiting resistor (24V - 19.5V = 4.5V, 4.5V / .02A = 225Ω, rounding off to 220Ω). If you mix types just use their Vf as part of the math.

Last edited: