LED spectra

[Andy Davies]

I'm looking for the best LED for microscope illumination and wondered how Olympus manage to get such a broad spectrum. Any ideas as to where they source their LEDs?

[Pau]

Any ideas as to where they source their LEDs?

Of course no, but the small peak around 400nm must correspond to a new generation of violet excited high CRI LEDs like the Yuji VTC https://store.yujiintl.com/collections/frontpage , see http://www.photomacrography.net/forum/v ... 036#203036

I must say that the pictured comparison is not very fair (there are many kinds of commercially available LEDs (and cameras are able to do WB!)
With a cheap Cree XM-L 4000-5000K "neutral white" the result is excellent even with the Canon EOS auto WB, despite the spectrum being much less perfect than the Olympus
The showed sample image could be applicable to a 6000K LED without custom WB

From the Cree XM-L data sheet:


EDIT:
Be aware that violet excited LEDs do emit UV (halogen too) while blue excited LEDs don't at least at the published spectra.

[enricosavazzi]

EDIT:
Be aware that violet excited LEDs do emit UV (halogen too) while blue excited LEDs don't at least at the published spectra.

I don't have specific references on hand, but there are concerns that even the peak radiation wavelengths of blue LEDs (especially "royal blue" ones that emit shorter wavelengths than "normal" blue LEDs) can have a negative health impact, not just their UV emissions.

UV radiation is more energetic than blue, and (especially in the UVC and UVB) is a well-documented health hazard. Blue LEDs, however, emit only minute amounts of UVA at the longer wavelengths near 400 nm, and no UVB or UVC. But still they may not be as safe as incandescent bulbs, which emit mostly longer wavelengths than blue.

Halogen (and even ordinary fluorescent tubes) generate, and may emit through the glass, UV wavelengths. Worn-out fluorescent tubes where the internal phosphor layer has "weakened", in particular, may emit significant amounts of 365 nm radiation (mid-UVA, one of the main Hg emission lines) and possibly shorter wavelengths. I have somewhere UV pictures I took of a traffic tunnel in NY City illuminated with (presumably) worn-out fluorescent tubes, where the tubes light up in UVA with an intensity comparable to the UV in full sunlight illumination.

[Charles Krebs]

I don't know where they (Olympus) source from, but the graph you show looks extremely close to the 5600K Yuji graph below. This likely means the base "die" is a violet one. Read over the comments on the PMN link Pau sent. No one really seems to know definitively about the risks of high peaks of blue radiation and since I might be peering down my microscope tubes for many hours at a sitting I personally prefer LED light sources for this use that do not have strong blue spikes or significant emissions at 400nm and below. As tempting as the specifications are for some 5600k violet based LEDs, I decided to go with the 30H series Brideglux Deco Vero (97 CRI). You might want to consider them as well (although COB LEDs can present other problems). I have included a graph from their spec sheet below.

When retrofitting as microscope with LEDs you may well experience a problem with COB LEDs (These are composed of multiple small LEDs underneath the phosphors). Many microscope lighting arrangements are designed to work with a single small light source such as the filament of a bulb. COB LEDs can show as multiple hot-spots unless diffused. They may be out there, but I don't know of any single die LEDs that have 97-98 CRI "ratings".