How to generate Red light from a Blue-Dominated Spectrum

Question

I have a white LED light source, the spectrum of which looks as follows ---

want to filter out red light from this white LED source. But as you can see the spectrum dwindles in the red zone.

And therefore, when I put red filter (absorptive filter) in front of my white LED light source, the brightness suddenly drops down to abysmal values.

Is there a way in which I could use this spectrum and get red lights without compromising on the brightness? I thought about interference and then I came across dichroic filters. But I'm not sure if that'll do it.

I also came across these filters called LEE LED filters which are filters specifically meant for LEDs. But I do not understand the scientific principle they work on. I would be grateful if you can follow the link and explain to me the scientific principle. From the spectrum shown in the link, these filters can give you more than 85% red. But the overall transmission is just 10.1% of the original intensity. So it's kinda bleh. Or maybe I don't understand properly.

A red LED light source would be just perfect for me, but a high wattage (50W-100W) red LED light source is hard to come across in the market.

Let me know if there are other options you guys can think of, for producing cool red LED light. Right now I use a tungsten lamp with RED absorptive filter, but there is so much heat (infrared) in it that my houseflies get fried up. [ I work in an insect flight lab and we use red light for high speed videography, since houseflies dont see well in the red zone]. So I want to switch to LEDs.

Answer 1

As poster states,

A very bright RED LED lamp (~50W) with a lamp tripod or something would be perfect for me.

Given the fact that one can easily buy off-the-shelf, 50 to 150w, LED flood lamp, with heat sink and drive electronics, but they typically comes with white LED.

One may consider simply replacing the central coin sized LED module and make corresponding small change in drive V and A.

Either do it yourself or if, in (small) batch quantity, manufacturer can do it for you.

For example, I am using this 10W LED, which is 9 LED die, 3 series in one set and 3 set in parallel. It is 350mA, 10V. The constant current driver board is adjustable up to max. 5A.

The RED LED module shown (which I do not have on hands, but, same principle as the 10W I have) is 10 LED by 10 LED, each die is one 1W LED, module is about 32V and 3.5A.

The example spec. is white LED (which is actually blue then changed to white) and voltage is slightly different from red but is same range 3.x to 4.x V

Commercial 100W driver with CE, 85 to 250V AC in, out DC 20 to 38V, 3A

Answer 2

Your "white" LED is actually a very bright blue LED combined with a phosphor that absorbs some of the blue light and re-emits it as the broad yellow peak you see in your spectrum.

While it would be theoretically possible to substitute a different phosphor that emits the red light you seek, I don't know of any commercial products that do this.

However, very bright red LEDs are readily available. They're used as the strobe lights on emergency vehicles, such as ambulances, fire engines, police cars (depending on your jurisdiction), etc.

Answer 3

You can't add light with a filter that isn't there in your spectrum to start with.

Operating theatre lights often use pure red LEDs to augment the spectrum of 'white' LEDs. I would imagine that you need good colour rendering at the blood end of the spectrum in that environment.

Answer 4

I'm afraid you understand it perfectly, although you don't much like it.

Once your light is generated, there really isn't anything you can do with the unwanted frequencies except filter them out. There just aren't any good ways to convert one frequency to another. (DPSS lasers are one exception, but not applicable here). Dichroic filters are used because they have relatively sharp cutoff characteristics and so can produce narrow passbands which provided "purer" colors, but they don't actually convert light from one frequency to another.

As a general rule, incandescents produce about 10% of their power as visible light, and virtually all the rest as IR, which is why your red-filtered light is very hot - the red filter is not eliminating the IR. LEDs provide considerably better efficiency, but as you've noticed, not much of it is red. With not much there to begin with, any red-filtered version will be, in your words, abysmal.

What you need, I think, are IR blocking filters, or cold filters, for your incandescents. There are two types. The most common is used in cameras and other sensors, and has the problem that they won't handle a lot of heat. The other variety is used in things like LCD projectors, and they are designed to take heat, although cooling design is also important. One source is http://www.mecanusa.com/polarizer/ColdFilter.htm , but you'll notice that these puppies can be expensive.

As to LEE LED filters, a general, non-technical outline is here http://www.leefilters.com/lighting/led-02.html#led-explain but it's actually pretty simple. Regular filters work on incandescent light, with a standard black-body intensity curve. As you may have noticed, your LED intensity has this big fat spike in the blue. The LEE LED filters just have what it takes to compensate for this spike.

Answer 5

An longpass optical filter with a cutoff at 480nm would do the job.

I mean a "real" optical filter, like interferometric filters with very sharp transition.

They might cost 10th of $

For the brightness, this is the best solution. Because the filter is quasi ideal, almost all the energy in the red part would pass through it. If this is not bright enough, that means there is not enough power into the emitted light at the first place. No way to boost this. Except adding LEDs or focusing the output light to the ROI.