LED Blacklights vs Flourescent Blacklights?

[Draco]

The sub 4 dolor 365nm 3 watt leds came into day and they are better then i had expected, i don't have a way to measure the output in mw but here is a shot from my living room, exposure time is 2 seconds with the aperture of about F2.0, the led is 8 feet from the wall.

distance about 3 feet but this rock normally florescence under short wave UV

the led does have some visible light but is better then what i was using, i am running one led at 700ma, 4 will be over kill for my original idea

[absolom7691]

Nice! Is this just with one LED? If so, that is better than what I expected as well. I am wondering if you could use a UV dichro or a woods filter to kill the visible light. I am curious if that would work on the higher output LEDs. I have no doubt the higher output LEDs generate more visible light as well.

[Draco]

yeah it's just one. I dont have a dichro that would do that and the filter glass i own is for short and medium wave, i could try my medium wave one, i would need to pop it out of the light though, it's in a portable UVC/UVB light that i take with me when i am out hunting at rock shops, i just got it working again, it had a bad lamp and corroded battery terminals from storage but it does now fully work, i could try that, the glass is just hot melt glued on, on my portable, my larger UVC only the filter glass is silicon-ed in, i took a normal fixture that you find at Halloween stores and put a UVC lamp in and made an aluminum cover and glued in hoya UVC pass filter glass, at lest i think it's hoya, i just looked at http://www.uvsystems.com/index.php for LW filter glass and it's much cheaper then the SW stuff so if my budget permits i will get some filter glass from them and test it
They had the filter i could use but a damned 35 dolor minimum order limit
Now the dichro glass i do have for UV reflects it so i will try an experiment when i can find the gass in my hoard

When i hold my hand over the led i can feel the heat from the light it's self, that surprised me a bit too, i have it mounted on a heat sink and am driving it with a jaycar kit built http://www.jaycar.com.au/Kits%2C-Sci...logue/p/KC5389 driver, it's large by todays standards but drives this thing right at 700ma and i think i can series two. I have a higher frequency Chinese switch mode driver but i have blown leds with it lol, little touchy on the voltage

[Draco]

one thing now to test as well is the lumen maintenance. I did find a kit for a uv light sensor so that will be the next project to see if i can measure the output, originally i wanted it to measure the uv that my lizards must have. https://www.adafruit.com/products/1918 hopefully that will work

[dkumpula]

Interesting stuff! I've been looking at these UV LEDs for a while as well. My goal is to provide enough UV light to fluoresce walls in a 19 ft x 10 ft space. So far, the DMX fixtures using the 365nm diodes are still quite expensive IMO. (US$2,800 for 60W!) I purchased one of these 10W 365nm chips to experiment with.

I really wish I had the ability to measure luminosity from these as Chinese vendors frequently overstate the specs on their products. That said, they are so cheap, even if they doubled their actual output in their specs, it still looks cheap compared with purchasing ready-made fixtures.

-David

[Draco]

I have found that my laser power meter a sper scientific 840011 will respond to these leds but i don't know what the correction factor should be as it's outside of it's listed response curve, it's lowest is 400nm with a correction of x10.4 but given that i am measuring about 150mw with the sensor right on top of the emitter, there is some visible light at that distance and at one inch i get about 41 mw, the uv light sensor i built i am still playing with, i think the formula from adafruit may be wrong, still working on it though.
I did manage to kill one, i was using it on a Russian simple drive and the interlock is different, it came out of interlock with the current way to high for a split second and blew the diode. I have found that these types of leds are more intolerant to spikes then other more conventional diodes are. in the past i used a Chinese constant current driver to run 4 higher wave length uv leds and blew the entire array when i was trying to adjust it's set point, i did get it to correctly run another 4 after i learned from my error, these 365nm diodes i have been using a kit built driver for luxeon star leds and runs 2 of these in series very well. the kit is older but jaycar still has them, called KC5389 i think

[Draco]

I have been running these as often as i can to try to get some idea of the lumen maintenance these may have, although i have not been running them at elevated temps yet

[Draco]

I have run into some problems with the uv leds, the two i am running with the jaycar driver have both had the outer lens melt, one totally failed, since they are two in series i added resistors in parallel to make sure each led is getting a balanced voltage, i am using 330 ohm .5 watt resistors to make sure each led is running at 3.7 volts with a total current of 660ma and still had a lens fail after that arrangement but that led is still operating, there is some type of clear rubbery coating over the led element, i have a led mag light brand flash light do the same thing, i have contacted the vendor for more information. I also checked the driver it's self for ripple and find it's ripple is at 80mv with a period of 70 micro seconds so i doubt that has caused any damage, the total failure of one led could have been due to an uneven voltage across the leds and added the resistors to balance that, i have never liked operating leds in series for that reason and these may not be as robust as a visible wavelength led. i am going to try and get the vendor to replace the leds but i suspect i will still have a lens failure as it seems to just be a polycarbante lens held on my silicone glue. mouser is also out of stock on the better quality led with a listed lead time of 14 weeks
I have these mounted to a heat sink that is over kill and keeping them running fairly cool. I hope this info is of use to any one interested in these and testing goes on

[Draco]

done a little bit of more resurch and found that the better quality 365nm uv leds have a glass outer optic and not the crappy plastic one, my remaining array is still running and the vendor did refund my money for the two that had the outer lens fail and since i included the resistors into the array they seem to be operating in a normal fashion for now, mouser does have other leds from the same maker listed in stock but at less then half of the one that is out of stock and sadly the want the same amount of money, i did come across the Nichia NCSU033B that has a glass optic so that is a possibility as well and has a nominal output of 540mw with a min of 495mw

[Draco]

here is the part from digikey http://www.digikey.com/product-detai...032-ND/4976722

and they have them in stock