I bought a dedicated Raytech 6 + 6 W UV lamp about 5 years ago for about 400 dollars as it was the best price/quality solution for photographing mineral fluorescence back then. That was then and now is now. The pictured UV torch sells for 20 dollars and is way more powerful in terms of UV flux than my Raytector.
The torch is based on Nichia 365nm UV LED and uses one rechargable 18650 Li-ion battery.
I did a little comparison and tried to evaluate the max. relative luminous flux per unit area from the two sources for an area that is small enough to be inside the cone of light from the flash. I had about the same brightness with the following values of exposure. This is for about one square centimeter and both Raytector and torch as close as possible to the fluorescent subject, For wider area the results would not have been so bad for Raytector
Torch ISO100, 0.5s, 5.6
Raytector ISO1000, 20s, 5.6
In terms of max luminous flux per unit area the torch is about 400 times more powerful than my Raytector. Part of this is explained by the fact that the big and clumsy Raytector is difficult to position really close to the subject whereas with the torch this is easy. This means that the torch makes low budget microphotography of fluorescence possible. I am using a 20x 0.42 Mitutoyo microscope objective with long working distance attached straight on a 100mm tube lense. So this is a single 10x image straight out of camera and shot with ISO100, 0.5s, f5.6 (fluorescent Hazel leaf)
Using the same setup with same exposure parameters, just replacing the torch with Raytector gives this image
Really nothing to see here. Should have used ISO1000 and 20s, but then the noice would have been really bad and even little spill visible light would have pretty much washed out the fluorescence.
Here's the torch
http://www.gearbest.com/led-flashlights/pp_277704.html
Here's discussion about the torch
http://www.photomacrography.net/forum/viewtopic.php?t=33123&postdays=0&postorder=asc&start=0
Here's a test for the torch
http://savazzi.net/photography/cheap365nm.html
HenriOne thing to be said about the inexpensive LED lights now available is the lenses have no specified filtering capacity and they can range from about 365 to 395 nm.
For a recent family outing at a location where long wave light would be quite sufficient I bought 4 Ray-O-Vac LED lights at US $9.97 each. We were in the woods on a sunny day but they gave off enough light to locate good specimens for further evaluation. Great family fun at a bargain price. Everyone found something and everyone had their own light. Thing is, the unfiltered LED light behaves much differently than filtered. I read the specs on the $20.00 light you cited and it said nothing about the filtering capacity or quality of the lens. A friend of mine has a generic LED light he paid $40.00 for. It does have a good filter on it. All other factors equal, his light gives a much truer color when compared to a high quality light. The lights I bought tend to look washed out. Sliding a piece of good long wave filter over my lamp vastly improved the results. The results then compared favorably with the filtered LED and my small Raytech. So the bargain price LED lights were sufficient for the needs of a day trip. For good photography, I would not use them.
Henri, nowhere near the information you presented, just empirical experience and an awareness of good filters for UV lights along with a respect for the cost of appropriate filter materials. Maybe it will give you ideas on what to look for regarding filter lenses and price comparisons. Hello Alfred,
these new 365nm Nichias are way beyond any other cheap "UV" torches previously available. There is very little need for a filter cutting the long end as there is really very very little visible light emitted. Look at the spectral power distribution graphs available in the last link in my post. Henri, let us know when you find an economical lamp for the short wave ultra violet range, or one that has all three long, mid and shortwave range selection.
Steven Hello Henri.
Interesting topic and expected results.
I have been experimenting with LEDs lately and plan trying also UV LEDs this same way:
Here 3 x 5W Yuji COB LEDs are brought within 2cm of the target. Plenty of flux.
This same setup could be used with UV LEDS. I thought about using 254nm ones maybe with band pass filter.
Does anyone have experience of these shorter wave UV LEDs with minerals.
Jyrki Hello Jyrki,
this LED stuff is evolving fast. 365nm leds comparable to the 365nm Nichias cost several hundred dollars just 2 - 3 years ago. Nichias only sell LW LEDs from 365 up, there could perhaps be some consumer regulation issues involved with SW sources.. I bought two 365nm Nichias for 40 dollars including postage, which is an amazing bargain. The UV flux they generate is more than enough for 20X microscopy. Now all that is missing are the amazing images :-)
In case you are not aware some guys at photomicrography.net are playing with similar constructs as yours. Thanks Henri, I will look at Nichias and wait to see those images.
-Reiner. Radiating power in 254nm is only couple of mW but they are small in size and if they can be brought close enough to the object it could work.
Area to illuminate could be in range of 10-20 mm2 and distance 10-20 mm. 
Hi Jyrki,
If you're looking to get high-quality UV LED sources for macro / micro work for example Thorlabs and Roithners could be one place. But then you would probably end up spending a few hundred I think on LED+mounting-cooling and controller/ beam expander.
https://www.thorlabs.com/newgrouppage9.cfm?objectgroup_id=2692
Laser needs even lower power levels for equivalent work, but then properly stabilised, narrow bandwidh UV laser light systems are even more expensive due to the need for pricey controller / power units. This is eventually what I'm dreaming of...
I would definitely use a shortpass filter myself if the wavelength of the cheap UV LED is not guaranteed and stabilised, or even if it was claimed to be fairly narrow. If the power level / wavelength fluctuates then you'll be getting various fluorescence reactions from different activators. A good shortpass filter from, say, Edmund Optics or Thorlabs might be worth the cost.
https://www.thorlabs.com/newgrouppage9.cfm?objectgroup_id=1001
On the other hand, if you want to filter out the cheap UV light's violet / blue light then indeed a longpass filter positioned in front of the LWD objective would be recommended.
https://www.thorlabs.com/search/thorsearch.cfm?search=longpass+filter
Personally, I don't trust unstabilised cheap UV LEDS / Lasers, as how can their operating tempature, voltage, current and wavelength possibly be gauranteed? When you're operating cheap lights for 5 minutes, 10 minutes or an hour, there's bound to be variation. That's why quality stuff (so far) costs more - but hopefully one day in the near future things might change..?
You see, even though my own new 3000€ (+ VAT) laser is a stabilised SLM system, its wavelength has to be confirmed and entered into the system every single day after it's stabilised in a few minutes....One day it can be 531.5xxxxx and another it can be 532.00xxxx. Gas lasers are another matter.
This is just some thoughts after carrying out my own modest photo / spectroscopic / Raman work... with both cheap and expensive light sources.
Cheers, 
Alfred, the flashlights with the Nichia LED do not come with a filter. You have to order that separately. It is called a ZWB-2 filter. Not sure I can give a link here to a source, but just google it. They are about $2-$4 depending upon the source. The flashlight is very much better than the other previous LED longwave torchs. Minerals not known to fluoresce previously are glowing now. Elmwood calcite even glows a little at the edges. The Convoy light (which is the brand sold by Gearbest and a number of other resellers from China) and the Nichia 365nm LED are game changers in the fluorescent mineral world. They are fantastic little beasts. The only problem is what Paul mentioned - minerals not known to fluoresce are glowing now. This means that we now need display lights using this LED or similar in order to show off our new finds. It also means that if you use this light while collecting (in the field or at a show), you might find something that looks great, but wont light up at all using your regular tools. I was at Don Newsome's house (maker of the Superbright), and we went through his extensive collection with this little light. It caused many existing spectacular short wave only specimens to fluoresce in new colors long wave, implying there are more minerals there than were previously known.
As for the filter, if you are using the light for photography, the ZWB1 filter (20mm, by the way) will cut out even more visible light (including some visible UV at the upper end of the range). It does so at a cost of a lot less 365 transmitted light, but for photography, you can just take a longer exposure. The ZWB2 filters are very cheap and usually available.
Everyone who has seen one of these lights in action has decided to buy one. It isn't as cheap as $25 because you probably want to get a couple of batteries and a charger, and that can cost you $40 if you get quality materials, but it is still pretty cheap, easy to stick in a pocket, and can easily cause minerals to fluoresce visibly even in a lighted room. (AND it is the best pet stain detector I have ever used ;-)
(BTW, I think there are many problems with 254 LEDs. Way too expensive, too dim, short lifespan and they self fluoresce. I wouldn't hold my breath waiting for one.) Thanks William, I just bought one ;)
By the way, after reading this thread I decided to get my hands on one of those Convoy light.
Amazing ! So I bought a 365 nm flashlight--not the Convoy, but a different brand, purportedly using the same Nichia LED. It works, certainly much better than the 395 nm light I previously had, but maybe my expectations are unrealistic, or my samples are weak, or the light I got is no good? Based on the above reports, for some reason I was expecting knock-your-socks-off, eyeball-blindingly bright fluorescence--but on a number of my specimens, the fluorescence requires putting the light an inch or so away to really make it light up. For reference, the light does a good job of fluorescing things like passports and currency, so maybe I just need better specimens!
That said, petroleum quartz lights up very nicely. The inclusions are easily visible with the light shining from several inches away. Much less purple than the 395 torch.
One of my optical calcite rhombohedra weakly fluoresces pink at 395 nm, and it is stronger at 365 nm but the beam is too narrow to fluoresce the entire piece.
A small fluorite specimen fluoresces pink at 365 and not at all at 395.
I have discovered that a number of miscellaneous specimens I have in my possession have spots where they weakly fluoresce yellow--I am guessing these are bits of chalcedony or calcite that happen to react to this wavelength.
The biggest surprise with the 365 nm light is that an otherwise mediocre sample of spessartine and smoky quartz, most likely from the Wushan mine in Fujian, has what I believe to be chalcedony that moderately fluoresces green, and other microcrystalline material that fluoresces yellow. With a long exposure (10s, f/5.6, ISO 320), and careful mixing of visible white light and 365 nm, I was able to capture what the piece looks like, as well as when it's solely illuminated by UV.
Hi Hiro,
The mineral that fluoresces green on your Wushan spessartine specimen is most likely Opal-AN, not chalcedony.
Russ I agree with Russ, the green is Hyalite Opal. common Botryiodal irregular surface and clear. It's even more awesome to know that the specimen has opal (albeit not of the precious variety)! Also, there is faint red/pink fluorescence of the white crust. Not sure what that is--calcite? But to have so many minerals and at least three different colors of fluorescence makes this piece a lot more interesting. Hiro,
The faint red/pink fluorescent mineral that you observed would be feldspar.
Russ LOL, I'm zero for two...clearly I have much to learn about identifying minerals! Note that these flashlight UV lamps are based on LEDs, and thus are not as powerful as the "real" UV lamps. And, they are long-wave. Nice to have for some things (I myself have one, but bought it to trace radiator leaks in my ol' jalopy; didn't even help much there), but probably not the best for minerals. Mark,
Thanks for pointing that out. Looks like I forgot to take into account that the specimen was illuminated with a long-wave lamp. Calcite would be consistent with that. Calcite is known to occur at that location.
Hiro,
RE: ... I have much to learn about identifying minerals. In one way or another we are all still "learning" here.
Russ The Convoy 365 nm torch is quite interesting, especially since it's relatively cheap.
At work we are using Tank007 UV-L03 torches. With 5 W They are even stronger, but much more expensive, about 150 € each. They don't come with a filter though, but the slight violet is hardly visible, so it's no problem I think. So I discovered that the 365 nm torch I bought is probably the same as the Convoy, just branded differently. I had been using the included battery without charging it fully. Once I charged it, I noticed a substantial increase in light output. If I can get my hands on a Convoy brand torch, I would like to make a side-by-side comparison. Of note, the brand I bought ("Jaxman") was more expensive but it came with a battery, and a ZWB filter (I think "ZWB2") was already installed.
Something else I'd like to try is to see what the ZWB1 filter does, but for now I'm pretty happy with what I've got. There is no comparison - price or performance - between a Convoy S2+ torch and other single LED flashlights. For serious fluorescent enthusiasts a filter is absolutely required. The colors are washed out by the visible light the LED generates, and UV photography will be severely hampered; the visible light will cause pictures to have a blue cast.
The Tank light is an overpriced, underpowered light. There are knockoffs pretending to be like the Convoy but most of them are junk from China, or are designed poorly. They will overheat and burn up.
The Zwb1 filter passes slightly less UVA but is even more effective at blocking 380nm to 400nm light (UV light that is visible to many). It is a good choice for photography, but not needed for everyday use (a brighter flashlight is better IMHO).
All of this is detailed in this review:
