Jump to content
UltravioletPhotography

Current State of UV LED Technology, KOPP Glass Article, 21Mar16


Recommended Posts

Thanks John. Very interesting.

Although they show the Nichia LEDs listed in the first chart of the downloadable pdf file, I can't find the Nichia LEDs compared to the others in the 'performance data' chart in the same pdf,

which would be quite handy for comparison.

Link to comment

John, thank you so much for the link to this UV-LED report. Very interesting to read about the substrates.

 

"Commercially available" means something slightly different to us UV-photographers who want to buy a UV-LED torch on Ebay. :D

Link to comment
Andy Perrin

I actually have built a UV light box before (dimmable too) and one thing I found out is that LEDs use different voltages depending on what color they are. (Specifically, the bandgap in eV corresponds to the voltage in volts.) They all use about 20mA at full brightness. Here is my box:

post-94-0-10177100-1458956787.jpg

 

The LEDs are just 365nm. I probably should have modified a UV flashlight or something, but I really needed it to be dimmable and it seemed just as easy to buy the LEDs I needed and make it myself. In retrospect, it was not as easy as it seemed, but at least it works.

 

Anyway, to modify a ring like that, you would need to make sure enough current is being supplied to bring them up to 20mA, which probably means replacing whatever current limiting resistor is in there at the moment.

Link to comment

I would like UV-LEDs at 310, 320, 330, 340, 350 and 360 nanometers all together in one light in such a way that the beams blend. Maybe I'd add 370nm if it was narrowband enough. I wonder if this is feasible?

 

I'd also like to be able to buy UV-LED torches at each of those wavelengths separately. This is probably in the realm of possibility now. But likely would be very expensive at anything but the usual 365nm which has found commercial popularity.

 

UV-flash - or sunlight if one is lucky enough to get a sunny day - is still the only option for broadband UV illumination in the UVA region.

 

Andy, good for you on building your own light! What LED chips did you use for that?

Link to comment
Andy Perrin

No chips, Andrea, just old-fashioned LEDs in the clear housing you can see. They were bought back in 2010 here. (That is a wayback machine link, they've been discontinued in favor of newer versions.)

 

If you want to make your own LED lights, that shop, LEDDNA, is pretty great. They have a wide range of stuff to choose from.

Link to comment

Yeah, I was being a little too specific identifiying the entire finished LED light with its chip.

 

Sometime try building a light box with mixed wavelength LEDs? Then you could "paint with light" and maybe get a photo with a mix of the wavelengths??? I don't know why I want a mixed wavelength LED torch so badly when UV-flash works perfectly well. I suppose it is because a torch is smaller and easier to handle and easier to use?

 

Can each LED be wired separately in your light box to provide the different voltages? I would assume so, but I am electrically challenged.

:D I didn't grow up having a dad who regularly did wiring, etc. I can do a mean spark plug cleaning & gapping though. Why he thought I needed to learn that I don't know. That's what boyfriends were for. (......j/k......)

 

Hey, Andy, there might be a cottage industry for you in retrofitting a macro light ring with UV-LEDs or building mixed-LED torches!!

 

Srsly though, I don't like being restricted to just 365nm in UV-LED fluorescent work. And I totally dislike the 365nm results in plain old UV photos. I spent one entire year ('09 maybe??) using only 365nm lighting for the botanical work. That whole years worth of UV photos is pretty much unusable. Recently I've tried B&W conversions. Those seem a bit better.

Link to comment
Andy Perrin
Can each LED be wired separately in your light box to provide the different voltages?

 

Andrea, yes. If you just want a box of LEDs that turns on and off and has a bunch of different wavelengths, that's pretty easy. You wire each color in parallel with a different resistor on that branch to adjust the voltage appropriately. (It was making the box dimmable that drove me nuts, because LEDs just don't want to dim in a proportional way. They like to dim exponentially.)

Link to comment

UV-flash - or sunlight if one is lucky enough to get a sunny day - is still the only option for broadband UV illumination in the UVA region.

 

Totally agree, I was thinking of something more like a modeling light to focus with which switches off just as the UV xenon strobe sitting on the hotshoe flashes.

Link to comment

I would like UV-LEDs at 310, 320, 330, 340, 350 and 360 nanometers all together in one light in such a way that the beams blend. Maybe I'd add 370nm if it was narrowband enough. I wonder if this is feasible?

 

 

With a prism it might be. A grating could work too but the losses would be higher.

Link to comment
  • 4 months later...

In the report I couldn't find the Ledengin UVA LEDs I am using (LZ1-10UV00-0000, mounted on star board), and they are much better than anything that I have earlier used. I am using those emitting at 365 nm (nominal) as they are extremely efficient resulting in that one can have an LED emitting 1200 mW of UVA radiation and dissipating not more than 2700 mW of heat, and so usable with small star-shaped heat sinks. To give you an idea of how much 1.2 W of UVA is, we can compare it total UVA in sunlight (AM 1.5): about 30 W per square meter. They cost about 30 € when buying less than 10 pieces from Mouser. Their emission spectrum extends slightly less into the visible than in other 365 nm LEDs I have earlier used. This small difference in the spectrum makes them also preferable to older types.

 

 

28975244730_c23a5677dd_z.jpg

 

 

The rulings on the background board are 10 mm apart.

We mount them onto the heat sink with self-stick heat transfer pads that are readily available pre-cut to size. (I ordered from Mouser Electronics and Digikey two international electronic components sellers.)

 

These emit a lot of UV, so be careful, and use eye protection and other protection as needed!

 

(Sorry about the image links, I do not seem to understand how I can upload images through the comments editor.)

Link to comment
Upload image with the "More Reply Options" and (after that) with "Choose File"/"Attach This File." By the way your links above are marked "private" when I try to open them, unfortunately.
Link to comment

To link from Flickr, click the Share Photo arrow to the lower right of your Flickr photo.

Next click BBCode on the popup window.

(Also select an image size if the default Medium size is not large enough.)

Copy that BBCode and paste it into the message box here.

 

The Choose File/Attach This File works only for images which reside on your PC.

Link to comment

John,

 

No, I have not compared them to current Nichia LEDs. According to data sheets the Ledengin ones seem to have about 10% higher output, but there is quite a bit of variation between "bins" so probably this is not too important. Initially I chose to buy Ledengin because they were easier to obtain ready mounted on a star board. The heat sink in the photograph does not require any active cooling, in fact it does not even get hot. I will give part numbers and prices in case you want to assemble some:

 

LEDs:

LZ1-10UV00-0000 (365 nm) Mouser part # 897-LZ110UV000000 € 32.85 Digi-key part # 1537-1032-ND € 31.94

LZ1-10UB00-00U4 (390 nm) Mouser part # 897-LZ1-10UB00-00U4 € 15.59 Digi-key part # 1537-1137-ND € 15.36

Heat sinks:

Ohmite SA-LED-113E (0.5" thick) Mouser part # 588-SA-LED-113E € 2.27, Digi-Key part # SA-LED-113E-ND € 2.26. With a thermal resistance of 5.33 C/W is fine for 2.7 W dissipation. (outer diameter is 2.75")

Thermal pads with adhesive on both sides, any of these should work:

t-Global Technology LP0001/01-L37-3F-0.45-2A Digi-Key part # 1168-1854-ND € 0.33

t-Global Technology LP0001/01-L37-3F-0.3-2A Digi-Key part # 1168-1853-ND € 0.33

 

It is easier to solder the wires before sticking the Star board to the heat sink.

 

I drive them at between 800 mA and 1000 mA (with a voltage drop of about 3.7 V). I use an electronics workbench power supply. As the voltage drop is small, they can be easily connected in series if needed.

Link to comment
Pedro, thank you for listing all the parts so carefully. This will be of help to those who would like to try putting together their own UV-LED.
Link to comment

Please sign in to comment

You will be able to leave a comment after signing in



Sign In Now
×
×
  • Create New...