I became aware of remote phosphors recently through a video on laser headlights by mikeselectricstuff where he shows a piece of remote phosphor in passing at about 5 minutes in. I’d never seen this material before and it got me interested in having a play around with it.
For a bit of background, there’s not really any such thing as a white LED. LEDs are essentially monochromatic as a result of the quantum mechanics which underpins their operation, which makes producing white light with an LED alone essentially impossible. One way around this is to use a red, green and blue LED in combination but by far the most common way to produce white light from LEDs is to use a blue LED and a yellow phosphor: the phosphor absorbs some of the blue light and re-emits is as various longer wavelengths, which appear as yellow. The combination of blue and yellow appears as white. The phosphor is typically embedded within the package of the LED.
Remote phosphors don’t substantially change this, except that the phosphor is separated out from the LED. One potential advantage of this is that the remote phosphor acts as a diffuser, giving the light a softer glow and more even light distribution. Another is that they essentially remove the influence of the emitter on the quality of the light produced: if a remote phosphor is used, the phosphor gives the same CRI regardless of what blue LED is used behind it, allowing more flexibility with component substitutions. I bought a few Intematix Chromalit parts from digikey to have a play around with. The effect is really bizarre, as you can see below.
My first target to make use of this technology was some LED downlighters which came with my flat. These are pretty awful, producing flickery light.
I thought these were GU10 or MR16 but actually when I took them apart I found that the bulbs are some strange proprietary type called G40. Despite being apparently purpose-built for LED bulbs, the fitting seems to suffer all of the failings of LED retrofits: the compact form factor places hot LEDs next to a hot driver circuit, reducing the reliability of both, and the lack of space leads to insufficient decoupling and large amounts of flicker. This seems like a stupid design decision when the ceiling space gives ample opportunity to separate out the driver.
I ordered some 5W blue LEDs and a constant-current driver module off aliexpress and fitted these in the original fitting. The phosphor I use here was CL-927-LR-PC: 90CRI at 2700K so it should produce a very good quality warm white!
I’m pretty happy with the outcome, at least for a first attempt. The quality of the light seems good, although what isn’t obvious from the photos (given the camera’s auto-exposure) is that it’s not quite as bright as I’d hoped. I used a 600mA LED driver with a 2x series LED package, for about 3.6W. The light it replaced was 5W. I’ve ordered a larger 3x series LED package and a driver to suit it, so in the future I’m going to have a go at upgrading the LED. I think I might also need to have another look at the thermal design. The transfer from the LED package to the fitting case doesn’t seem great. I’m hoping I can find some way to scrape off the paint and improve contact.
Having done this, I’m inspired to have a go at using this material for some other projects: it’s also available in strips, which I think I might use for some under-counter lights in my kitchen.