UV LED Flashlght

On a lark, I bought one, advertized to have a spectrum at 395 nM.  There is a lot of visible light with this unit.

Things that respond well, any thing with daylight optical brighteners; paper, many colored plastics, laundry detergent.  Skin, fingernails, dead plant material.

It excites a red fluorescence in my aquarium plants, much less with terresterial plants.  It also lights up the aquarium water green, I don't know what that's about.  I don't see any effecct with tap water.

I have some of the fluorite spheres that are sold on E-Bay that fluoresce green and continue to glow for hours afterwarde, they respond well to the LED light. 

Many of the rocks I find in the Mojave desert have a thin layer of calcite, Mn activated, that glows orange with the normal longwave UV (351 nM, I believe).  With the LED light the same areas fluoresce light pink.

 

Oh, and cat urine.  I have cats.  Yuck.

Re: UV LED Flashlght

Using a UV filter (B+W 403) over the LED flashlight blocks most of the visible light, and the calcite fluorescence is again the normal orange.

The pink that I see with the unfiltered LED still looks to me like a fluorescence of some type, not just a reflection of the light from the LED.  I haven't figured out how to tell for sure.

Re: UV LED Flashlght

Added a picture with the rock illuminated with a 405 nm laser pointer

 

I tried 361 nm UV LEDs in a flashlight, much less visible light but not enough UV to do any good.

ed.a's picture

Re: UV LED Flashlght

Hi Charles - your blue laser fluorescence image is very interesting and revealing for a couple of reasons. Because the laser is almost purely monocromatic, we know that anything visible in the image is either an artifact of the camera responding to the excitation source, or fluorescence. Your image shows both very nicely.

Note the camera's red channel response in addition to the blue. That's different for each brand and model, but lots of cameras are designed to have the red channel respond to deep violet light. The eye does something similar, and the makers probably want a photo of a prism or spectrum to show some violet. That side effect really messes up fluorite photos, where intense blue areas start turning violet as the red channel saturates. Even stranger things happen in the camera related to color saturation, depending on the camera's extra processing for "vivid" images. Shooting in "raw mode" really helps there.

There's one spectral range where your camera has little response to the excitation light, and that's the green channel of your photo, where you can clearly see the fluorescence:

I muted the red and blue channels so you can see the green alone. The excitation leakage into the green that remains is probably an artifact of the camera saturation algorithm. Funny how interefence fringes from the laser appear along with shadows of  microscopic particles on the dispersing lens. What's really clear though is the small crystal and  little blebs.

Try this yourself. I used PC Irfanview (free) to edit the channel gains. It's a very cool tool. Alternatively, ImageJ has lots more image analysis features and plugins, and runs under Java on any platform, but is trickier to use.

You'll definitely want a blocking filter on the camera. That, plus a smoother lens+diffuser for the laser, should give you some nice images showing red/orange  fluorescence, and maybe into the blue, too. If you're looking to capture the pinkish emission you'll need to record fluorescence in all three colors, without the excitation light.

Here's a trick - If it gets down to just seeing the bright visible specular reflections of the light source leaking through, try adding crossed polarizers. One goes on the camera as normal, and another one on the light. Test to see that they both block the deep blue. (Even expensive polarizers tend to leak blue light - so this won't likely work with deeper UV sources. Circular polarizers will work, facing each other.)

I also meant to follow up on your last image that used a 47B over the 395nm LED. I found some filter specs for the Wratten 47B. It looks like less than 10% of the 395nm gets through but it passes the more visible blue leakage quite well. The camera picks that up, probably even with a blocking filter.

I mentioned that I'd found a "laser cleanup" filter on eBay. Unfortunately, it too has a very sharp cutoff - blocking over 75% of even 405nm light. Also, since it's a dichroic, any off-axis path (off the reflector in the LED flashlight) raises the cutoff frequency with the angle (like the anglular shift in a difraction grating). I tried fixing that with a shiny aluminum tube over the UV LED, but it still was no good. Besides being too small, in my flashlight it's worse than just using regular UV filter glass.

Looking forward to your next images,
Ed.A

p.s. Image attachments are now working again, but your gallery with descriptions is a great upload location.

Re: UV LED Flashlght

"The pink that I see with the unfiltered LED still looks to me like a fluorescence of some type, not just a reflection of the light from the LED.  I haven't figured out how to tell for sure."

Insert a UV blocking filter between your LED and the specimen (perhaps you could use a common camera's filter).  That will block all light below 400nm - then you're sure that no UV is getting to your specimen.  You just have to be sure whatever filter material you use blocks LW (glass won't do it).  But - note that a couple of minerals (at least) will still fluoresce at 405nm (visible).  Some Greenland sodalites glows orange even under a blue light, and rubies fluoresce red under blue light.

Mark

Re: UV LED Flashlght

That makes very much sense, until I start messing with different filters.  I set up a gallery with shots taken with either the LED filtered or the camera lens filtered.  It's not clear to me what I have learned.  Photos are small, no color correction was done, just changed the size in Photoshop Elements.

UV Haze and Skylight filters are glass, regular photograpfy filters.  Wratten are Kodak gel filters.  B+W 403 is glass.

Camera was left in automatic, so the exposure changed with different conditions, I can provide that data if anyone cares.

 

The rocks are resting on a Kodak gray card if anyone wants to play with the colors.

Re: UV LED Flashlght

The way I understand your picture postings - the first group of pics are taken with the filters between the LED and the rock.  It seems that the only real change is seen when you use the b&W visible blocking filter.  The visible blue/purple light is filtered out and you're left with the fluorescence of the rock.  All the other filters you placed between the LED and rock would have filtered out the UV, transmitting only visible light.

The 2nd set filters the light reaching the camera - correct?    It seems that the only thing that made a significant difference is the 2E filter (which - according to my feeble memory - is yellow).  The 2A and 2B are also yellowish I think, but not as much as the 2E.  Yellow filters blue, so those filters would cut down the amount of blue reaching your camera.

Conclusion - 395nm LEDs are near useless for photography (IMHO).  They simply put out too much visible light.  You must filter them with a UV bandpass filter.  Filtering at the camera is probably too extreme and results in off-color images.  Nicha 365nm LEDs on the other hand are great for photography as they emit very little visible and lots of UV.  Does your final image (below, w/ the 2E filter) look like the real rock?

(BTW - it looks like you can now upload images to a "temp" directory for inclusion in forum posts.  Sure would be nice it the settings were changed to allow images from the galleries to be included in these forum posts).

Regards
Mark
www.minershop.com

Re: UV LED Flashlght

Your take on my picture posting is correct.  I pretty much agree with your conclusions.  None of the pictures look like what I see with my eye, the 3 Volt flashlight is close, but too warm.  The rock as I see it when illuminated through the 403 filter looks visually the same as when I illuminate it with a real UV lamp.  The photo doesn't.

 

The thing I am after now is the red or pink that I see with most filters and that gets stopped by the 2E.  I'm thinking that there is a fluorescence that is being excited by energy between 420 and 440 nm.  I have ordered another filter (47B) that should block  green through red light to see it the pink is really from the rock.  However there is such a long attenuation tail of the 2E that it could be blocking red that leaks past the 2A.

To my eye about half power attenuation seems to be the discernable limit, I decided this by just blocking a portion of the aperature in front of the UV LED light

ed.a's picture

Re: UV LED Flashlght

Hi Charles,

Try using both the 403 and the 2E filters, and you should get some really nice results.

I have the B+W 403 filter and a 3 watt 392nm LED flashlight, and tried a comparison. The 403 passes less UV than Hoya U-325C filter glass, but it is a very good excitation filter because it passes very little visible light. The Hoya will be leaky by comparison. The 2E yellow as a barrier filter over the lens should be a good match for the 403.

Your photo with the 403 but no lens filter looks blue because the camera is reponding to the near UV. The 2E looks pale because of the visible leakage from the LED. You can test the combination of the filters for good fluorescence contrast by shining the flashlight through both filters. You should get black.

The 403 seems to pass only 5% of the 302nm, but for photos it just means a longser exposure. For field use, it blocks too much light. You might try some woods glass as recommended by redhat. I've used pieces of broken BLB blacklight bulbs over early UV LEDs with some useful results. By wearing yellow safety glasses, you can get pretty good contrast.

BTW, If there's no cyan flourescence to be found in the specimen, a full yellow filter works fine. I've done this to block the 436nm mercury line from blacklight bulbs. Funny how that purple color is what most people think of as ultraviolet. It's probably also the inspiration for Terry Pratchett's fictitious "octarine" as a magical color.

Note that the camera can also get tricked into boosting the blue. If it's on auto-white-balance, it will think the lack of blue light in the scene means it's lit by incandescent wires, and will boost the blue (sometimes a lot) to compensate. I suggest picking a white balance for consistency. If your camera is capable of RAW mode, use that for extra dynamic range and the ability to choose any white balance in post processing.

Be aware also that cameras don't have perfect RGB filters. The red channel often responds to blue light. That's great for shooting concerts, but makes for horrible rendering of fluorite. Some cameras also respond strangely to deep red, getting either orange or purple/pink depending on whether the green or blue channel, or both, are responding to the red. To test your camera, try shooting the spectral display of a halogen lamp reflecting off a CD (or use a diffraction grating if you have one.)  A free image analysis tool from NIH called ImageJ can plot the camera color response across the spectrum.

UV Laser pointers are getting cheap, and they put out nearly monochromatic 405nm light. No excitation filter is needed, and the light can be diffused for fluorescence imaging. Even a 5mw pointer is powerful enough for prospecting at a distance. I took my UV laser shopping, and noticed that some of the clear safety goggles also block the 405nm (through the packaging). They're not spec'd that way, it's just a side effect of the acrylic they're using. The clear goggles are way better than yellow when stumbling around in the dark. Try bringing binoculars, too. For really long distance prospecting, the laser beam can be diverged and collimated backto a tight spot with the binoculars. Then you might want a telescope for viewing...  But I digress.

With your inspiration, I've just ordered a small $10 "UV Laser Clean-up" filter off eBay. It passes 80% at 395nm and has a sharp cutoff around 405 to nil at 415nm. It should fit into the flashlight (11mm) and provide really nice contrast for photos and field prospecting.

Please post more photo results, and any more info about the mysterious pink fluorescence. I've found that the longer wavelengths do make for some interesting fluorescence, like yellow-orange from terlingua-type calcites.

Thanks for your posts,
Ed.A

Re: UV LED Flashlght

I've added shots with the LED filtered with the 403 and the camera filtered with the 2E and a number 8 (yellow).  Also shots with the specimen illuminated with a standard Way Too Cool UV setup, long and short wave, no filtering on the camera.   My rock appears to have some uranium in it.  I wonder what the little blue spot in the center is.

I haven't tried adjusting white balance, that can come later.  I'm still trying to figure out what I'm trying to learn.  When I played with white balance before I felt I needed to have the specimen illuminated and quite near the monitor so I could judge the results.  That gives ambient lighting problems.

I am thinking of playing with the 361 LEDs mentioned, I picked up some LED flshlights to see about modifying them.  those have gotten cheap lately.

As expected the 403 and 2E together produce dark.  Where does one get Wood's glass?  Any more information about the 405 nM laser? Maybe pm to avoid advertising in the forum.

Re: UV LED Flashlght

I've added shots with the LED filtered with the 403 and the camera filtered with the 2E and a number 8 (yellow).  Also shots with the specimen illuminated with a standard Way Too Cool UV setup, long and short wave, no filtering on the camera.   My rock appears to have some uranium in it.  I wonder what the little blue spot in the center is.

I haven't tried adjusting white balance, that can come later.  I'm still trying to figure out what I'm trying to learn.  When I played with white balance before I felt I needed to have the specimen illuminated and quite near the monitor so I could judge the results.  That gives ambient lighting problems.

I am thinking of playing with the 361 LEDs mentioned, I picked up some LED flshlights to see about modifying them.  those have gotten cheap lately.

As expected the 403 and 2E together produce dark.  Where does one get Wood's glass?  Any more information about the 405 nM laser? Maybe pm to avoid advertising in the forum.

Re: UV LED Flashlght

Info on 405nm flashlight here: http://www.minershop.com/html/uv_laser.html.  I built a couple two years ago, prices have come down considerably since then.

Wood's Glass - the best source I've found is LW transilluminators on eBay.  Keep in mind that the transmission rate for Wood's Glass is much lower than Hoya 325c.

Regards
Mark
www.minershop.com

Re: UV LED Flashlght

I received the 47B filter which has a maximum transmission at 440 nm, per the label on the package.  I've added a photo showing the pink fluorescence which is usually hidden by the orange excited by deeper UV.  Blue somewhat attenuated by the 2E on the lens.

The mysterious little blue spot doesn't respond to this illumination.

Re: UV LED Flashlght

Hey,

 

I've been working on some UV LED lamps for blacklight art and have found that a Woods Glass filter over the "UV" LEDs helps get rid of wavelengths longer than 400nm.

I have noticed, however, that my Watermelon Calcite cyrstals don't fluoresce nearly as well with the 395 nm LEDs as with some 362 nm Wildfire Effects Master tubes.  I've ordered some 351 nm LEDs from LEDSupply.com to see if that can be improved.

 

RH

Re: UV LED Flashlght

I've done a lot of work with UV LEDs.  The cheaper LEDs all seem to be in the 390nm to 405nm range and emit lots of visible light.  There are other LEDs in the lower ranges (375nm, 365nm most commonly) which emit less visible light.  Filtering with wood's glass is effective in eliminating the visible light, but Hoya glass is better (but of course more expensive).  Both also cut down the UV output quite a bit.  Nichia makes 375nm LEDs in 5mm packages - not a lot of power, but perhaps less visible light.

I have started using Nichia 365nm 3W LEDs in my homemade lamps (high power packages - serious heatsinking required).  These are very powerful and do not emit a significant amount of visibile light (in proportion to the UV output).  I even use them for my photography setup - picture below is a piece of tenebrescent sodalite taken under SW (rear image) and LW (using Nichia LEDs).

The Nichia LEDs are ideal.  But - as with all things ideal there's a "but": they're very expensive.  Around $80 each for just the LED and you have to figure out how to drive them.  If anyone is interested in the details just let me know and I can offer you some help.

BTW - I also have a SW LED!  250nm - Pitifully underpowered (50uw) and the visible light overwhelms the UV output.  Putting a piece of Hoya glass in front of it filters out the visible, but also weakens the UV to almost useless....

Mark

Re: UV LED Flashlght

Hi all,

 

thanks for the interesting discussion. Mark, whats the item no. for the Nichia 365nm 3W LED and were did you bought them? I only get rubbish using Google on that.

 

Thanks,

Stefan

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.
Potential Exploration Links