Light source power and distance ...

[Karmelo Martin]

Hi all,

I'm thinking about making a custom light source using LEDs, and don't know which is the most appropriate power the light source should have and also which is the distance the light source must be set from the paper...

The light source will be used to expose contact sheets, if it makes any difference. I'm considering to build it with LEDs to have no warmup time and no loss of color over time. Also, the light should be a bit more directional this way, I imagine.

Can anybody help, please?

Thanks in advance
Carmelo

[Robert Hughes]

Hi all,
Can anybody help, please?

OK, but you may not like the answer. LED lamp assemblies require voltage/current regulated DC power supplies that are not difficult to design and construct - if you already know enough electronics to do that sort of thing. But I would guess from the tone of your post that you are not an electronics whiz.

So - like anything involved with electricity - know what you're doing or go to somebody who does. Otherwise it's not worth the trouble.

(I've blown up more stuff over the years than I care to remember - including white LED's.) :eek:

PS - white LED's do change color over time, so if you're concerned about constant color, LED's may not be your best choice. Have you considered fluorescent lights? They're much simpler to wire up and maintain than LED arrays. But they do have a warmup time - perhaps you could put a sliding shutter between the lights and your contact printer so you can have the lamps running prior to printing.

How much light do you need? I use the bathroom 60 watt incandescent light for ~5 seconds when I contact print on photo paper, but go outside when doing alt prints.

[Karmelo Martin]

Haha...

Yes, you're quite right, I've also blown up many things in the past, I was thinking about buying a simple current source for it, and yes, I guess I'll have to take a lot of care not fry the whole thing...

But my original question was more about luminic power, I don't want to build up a system that yields too long, or (even worse) too short exposure times. So basically I need an idea of the lumens the source must send.

Kar

[rdenney]

12-volt LEDs are available and can be wired in series with a resistor--the resistor is on the downstream side and pulls current through the LED. Remember that an LED is a diode, so the DC current will only go one direction through it. Use Ohm's Law to determine the size of the resistor needed to achieve about 80% of the rated maximum forward current. With a well-regulated 12-volt power supply (which will probably be more like 11.5 or 13.8 volts, depending on which one you buy, and that actual voltage should be what you use in the Ohm's Law calculation). Adequate and well-regulated power supplies are easy to find, even at places like Radio Shack.

Use high-quality resistors so that they won't change their value when heated by current.

All non-lased light sources follow the inverse square law, meaning that if you double the distance, the intensity of the illuminiation drops by the square of 2, or 4. You can use the standard f-stop scale to set distances, actually. If you want a source that is 4 inches from the contact frame to be a stop brighter, move it to 2.8 inches from the frame. If you want it to be a stop less bright, move it to 5.6 inches from the glass.

You'll need a fairly tight array to achieve adequate and even illumination--LEDs do not have smooth patterns of falloff and you'll need to integrate a lot of LEDs to mash all that waviness together. Typical commercial bright LED's have a maximum forward current of about 20 milliamps, so at 13.8 volts you'll need something like a 1KOhm resistor in series to load it to some useful (not full) percentage of its rating. Each resistor will consume a quarter watt of power, so an array of LEDs in large numbers will produce some heat. You can wire a bunch of LEDs in parallel going into a single resistor if it has sufficient power rating (at least twice what it needs to be by calculation). Many LED circuits are more sophisticated than this and pulse the LED to reduce consumption and heat dissipation, but this will work.

Disclaimer: This is a bare outline of a fairly easy technology for electrical types. If it confuses you at all, you might want to study up or get some hands-on help.

Oh, and I think you'll find that even so-called white LEDs...aren't.

Rick "thinking this might be easier with a flourescent light table as a starting point" Denney

[Robert Hughes]

Haha...
I need an idea of the lumens the source must send.

Easy answer! It, uh, depends... Are you using standard photo paper (I'd guess approximately ISO 1-3 compared to film)? How much loss will your light diffuser introduce? How tightly do you intend to pack those LED's next to each other? Do you want to use white LED's or mix and match red/green/blue (each with separate controls for adjusting color balance)?

If you are set on LED technology (sounds like you are somewhat familiar with it), keep in mind that each LED needs a separate current limiting device (even a resistor will do), as some LED's "fire" sooner than others and would hog the current if you tried to drive them in parallel groups, leaving you with hot spots at least, and fried electronics soon after. As Rick says above, series circuits may alleviate the hot spot issue. Also, heat buildup from several devices in close proximity may become an issue.

As for how much light you need: try it out. Mock up a prototype, perhaps using an incandescent light and diffuser, run a few contact prints until you get a workable exposure, measure the intensity with a light meter, then cobble up your LED array to meet the illumination requirement.

[Karmelo Martin]

Thanks rdenney,

Do not waorry, it doesn't confuse me at all. Actually I'm quite used to this terminology as I have a degree in physics.

My problem is more related to the light quantity I need to get a properly exposed paper within a reasonable time. Robert gave me a hint saying a 60 watt bulb in his ceiling makes a 5 sec exposure (Potency and distance), but I cannot extrapolate it very well to the LED infrastructure as LEDs are a lot more efficient than a tungsten bulbs (They dissipate a lot less heat, so more percentage of the power is used to emit visible light).

I would need an idea of how many lumens/cm2 are needed to get a correct exposure at... let's say 15 seconds.

Thanks,
Kar

[BetterSense]

12-volt LEDs are available and can be wired in series with a resistor--the resistor is on the downstream side and pulls current through the LED

It doesn't matter where the resistor goes as long as it's in series.



Typical commercial bright LED's have a maximum forward current of about 20 milliamps
Maybe 15 years ago. Modern cree-type LEDs regularly operate at 500ma-1500ma.

The ideal way to drive LEDs is with PWM for dimming.


My problem is more related to the light quantity I need to get a properly exposed paper within a reasonable time. Robert gave me a hint saying a 60 watt bulb in his ceiling makes a 5 sec exposure (Potency and distance), but I cannot extrapolate it very well to the LED infrastructure as LEDs are a lot more efficient than a tungsten bulbs

Well, you could use a meter. But what I would do, is look up the lumens of a 60 watt bulb (or much better, look up the paper datasheet), and considering that LEDs are lambertian emitters, and themselves have datasheets, it should be relatively straightforward, although laborious, to get a good estimate.

[Karmelo Martin]

As for how much light you need: try it out. Mock up a prototype, perhaps using an incandescent light and diffuser, run a few contact prints until you get a workable exposure, measure the intensity with a light meter, then cobble up your LED array to meet the illumination requirement.

I guess you're right, Robert. I was wondering if someone had the same idea before, but probably the best is a try-and-error approach till I get the desired effect.

thanks everybody,
kar

[Robert Hughes]

Of course the PWM (Pulse Wave Modulated) power supply would let you dial in whatever light intensity you want, so as long as you've got "enough" LED's to do the trick, you can work.

[Karmelo Martin]

Hi BetterSense,

Thanks for the accurate detail... You're right... I didn't know the LEDs obey the cosine law... But if it's true, it would be a quite simple problem. A bulb should send the light intensity homogeneously in a spheric wave front and on the other side, the LEDs should deliver a more inhomogeous front, loosing power related to the cosine of the angle of incidence (Assuming the LED source is set completely perpendicular to te paper).

Therefore, I could calculate the total power of a point in the paper summing all the contributions of each LED in the array...

Thanks, this is a good starting point :-)
Kar

[CG]

Why not just use a prewired screw in LED bulb assembly? No wiring. No transforming voltages, no soldering, no resistors... Just hook up your timer to a screw in socket and go.

I happen to use these: www.superbrightleds.com/cgi-bin/store/index.cgi?action=DispPage&Page2Disp=/specs/E27-W24.htm (the green and UV bulbs) but there must be a bunch of companies making similar products by now. I bounce a green and a blue/UV LED bulb off the ceiling to even out the light and bring the exposure time up a bit. They are pretty bright as seen by the paper even though to the eye they don't seem all that bright. I painted a small patch on the ceiling dark to help lengthen exposures. Each bulb is plugged into a separate timer. It works pretty well. They make white bulbs too.

[aduncanson]

Conventional 60 Watt bulbs put out about 520 lumens. They are nearly omnidirectional. If the bulb is 170cm distant from the paper, that would make .0015 lumens/cm^2. However you need to estimate what fraction of that 520 lumens is actinic, or effective at exposing photographic paper.

I use a 50 Watt Halogen reflector flood (with a higher color temp than a conventional bulb) at about 170cm. Based on representative specifications (and a guess about its radiation pattern), I estimate that it makes about .006 total lumens/cm^2 at the paper. I get exposure times on the order of 15 seconds on Lodima, which is much less sensitive and more specifically blue/UV sensitive, than most enlarging papers. I find my set up much too fast for contact printing on enlarging paper, but I would not abide an exposure time shorter than 4 or 5 seconds.

[Karmelo Martin]

Why not just use a prewired screw in LED bulb assembly? No wiring. No transforming voltages, no soldering, no resistors... Just hook up your timer to a screw in socket and go.

I happen to use these: www.superbrightleds.com/cgi-bin/store/index.cgi?action=DispPage&Page2Disp=/specs/E27-W24.htm (the green and UV bulbs) but there must be a bunch of companies making similar products by now. I bounce a green and a blue/UV LED bulb off the ceiling to even out the light and bring the exposure time up a bit. They are pretty bright as seen by the paper even though to the eye they don't seem all that bright. I painted a small patch on the ceiling dark to help lengthen exposures. Each bulb is plugged into a separate timer. It works pretty well. They make white bulbs too.

Mmmhh... Very interesting. So you are getting exposure times of 15 secs (On Lodima) with just two of these bulbs hang on the ceiling?

OK. I think that's all I need for the moment...

thanks a lot for all of your replies guys,
Carmelo

[mhulsman]

Carmelo,

Look at this article http://www.apug.org/forums/forum43/51852-converted-8x10-dichroic-leds.html

30 x Luxeon III leds in a 10 x 10 light head will give about 2750 Watt of light.
That will all fit in a Durst L184/301 head.