In djozz’s test, he said that the Lite-On maxed out at 2a. I’d be kind of scared to run 3 on a fet. Wouldn’t that kill them?
i’m talking about Vf, if it is high for particular led, 1 cell may not have enough voltage for that led to pull anything beyond 100 or so mA. let alone 3.
i may have missed something, but looks to me he used a diy test stand, and used more than 1 cell for tests. like i said not all of those leds are the same, 3535 is a popular format, those leds are made by different manufacturers, Vf is not same for all. we know nichia works great with 1 cell, but those that i bought would not.
these are the ones i bought, from this seller, great leds, but not for 1 cell.
If I look at the VF, the Liteon should work ok but close to max as a parallel triple on a FET driver with a not too high drain 18650. But perhaps start with not bypassing the springs and measure current that way first.
But mind that you can’t use a standard plastic (PMMA) optic with 365nm leds, it absorbs the UV, so only as a mule or in a light with triple reflector.
I’m kind of a voltage dummy, thanks for the explanation.
I don’t know that about the optic…I’ve got a lot to learn.
Thinking about 4 x NiMH and some 7135s…
I’m using generic 365nm “bead” type UVLEDs on 20mm stars in a C8. Not as bright, lumenwise, but a nice tight intense beam.
standart plasitc lenses are polycabonate, they are no good for 365nm, pmma is ok for 365, i used ledil 20mm triple in my build, carclos are polycab. and would not work, ledil triple is pmma, also a bit higher, it also has a small rim making it a bit bigger than carclo lens in diameter, basically no good for most drop ins.
I decided to go with just one of the Lite-On leds in a S2+ host with a 1.5a buck driver.
I did not know that Carclo optics were polycarbonate and Ledil optics PMMA, thanks for correcting me. I searched for transmission spectra of these plastics and came up with this graph. I did not do an extensive background search on the origin of the graph (on first look it looks like a trustworthy source), but at least this source shows that what they call 'standard PMMA' is no good for 365nm leds either. There is a 'UV-T PMMA' that does transmit deep into the UV. I will try to dig a bit deeper because I like this kind of stuff :-)
when i was building my light i powered up the leds with no lenses, shined the light on $100 jamacan bill, there are 2 different uv dies there, one lights up even with 400nm, but small part in the center only lights up with lower band uv like 365, than i put carclo lens, both dies stopped fluorescing, one completly, other one almost, it was barely glowing, than i replaced them with ledil pmma ( from here, top one), both dies were fluorescing same way as it did with no lenses. there prbly are loses with pmma lens, but i could not see them with a naked eye. but since things were lighting up with that lens, and also 3\16 window over it from pmma as well, it is good enough for me.
ledil makes pmma lenes for their lz1 leds, that are rated 365nm, http://www.ledengin.com/products/lenses#LZ1
Perhaps Ledil uses a non-standard PMMA then, which would be good news. I have a Ledil triple optic, so I can test it one of these days 
Bumping this thread to see if anyone has tested the SST-10 emitters yet. I purchased one of each of the lite-on and SST-10 from mouser. According to the datasheets the SST-10 should put out more radiant power at the same 500mA current (492-665mW vs 720-810mW), but was curious if anyone had done a test yet.
I also got the SST-10 in the royal blue, deep red, and far red to compare to XP-E/XP-E2 color emitters. Waiting for the green and regular red to become available.
I have a 365nm SST-10 waiting for a test for a few months now. First I had to update my test method, but it works now and now I have to check two already tested 365nm leds (276A and the LiteOn) to form a new reference. Then the SST10. I’m not looking forward to the work atm (sorry, led testing is always a bit boring), but it is in the planning soon.
Yeah I can understand it being boring. Working on making my own pvc “lumen tube” integrating sphere, but I want to figure out a way to automate testing before I push it into service.
Idea would be to set a few limits, and just hit a button and have it automatically ramp and hold current and take light measurements until it sees a decrease in luminance, then save all the data to an SD card to put into excel.
Yeah would be interesting how this performs
the thermal pad seems promising
just saw this one on mouser 30% cheaper
https://www.mouser.de/ProductDetail/Luminus-Devices/SST-10-UV-A130-E365-00/?qs=sGAEpiMZZMu4Prknbu83y0xUekQfhx0SNwykuKeIiMDCbrPKy%2FDTAw%3D%3D
But with more current efficient UV LEDs lately I have seen for Nichia one 233B that the forward voltage bumped by 0.5V up to 4.2V,
so you need a Buck/Boost driver
Sounds good, Jensen! I usually envy people who have automated their test equipment, HKJ seems a master in automation.
I myself mostly managed to avoid computers and automation all my life, and so all my testing is done with stand-alone equipment and pen and paper. This has the disadvantage of it being somewhat labor-intensive, every time again, on the other hand it gives very direct control over the methods, I’m continuously doing checks and almost immediately notice it if anything does not work as intended.
And in my case computers eat up lots of time, usually spent on things unrelated with what I want from them.
By the way, with led testing for BLF, there’s only so much that can be automated: I still have to do the reflows and mounting with care, make relevant pictures and think up and write a readable and informative post. All that takes way more time than the measurements, calculations and typing in the numbers and making a graph in Excel take.
(I have a ready to post led test for the red 219B that I never posted because I did not find the motivation. Btw, it performs a tiny bit under the red XP-E2)
I can confirm the SST-10 is at least as good output as the lite-on with maybe a bit less waste light. Don’t have a test to prove if it’s better or not quantitatively, but both running 1.4A (4x7135) they look very comparable, with the SST-10 maybe having a slight advantage.
And I can confirm that again with a direct test against each other. Still have to crush the numbers and make a nice post about it.