Nichia UV LED Lamps

[Alex H]

I've never been able to get much at all with my 292BP10. Used it on a D300, I think it was. Most likely I hit a sensor limitation, but don't know for sure. However, I should repeat that test with the D600 and K5, just for grins.

 

How about your 280 ???

 

I would expect your results to be the same with D600 and K5. For my tests I have to wait few more days for the good test subject. We just had few inches of snow again yesterday, but it had already melted. It will be a subject for a separate post in the nearest future.

[rfcurry]

Andrea,

I don't replace the Wood's filter with glass, the lamp is bare to the air.

[Andrea B.]

I am curious about whether there is any possible danger from the bulb suddenly breaking?

[Shane]

I have been using the UVP BP-100AP for indoor (and lab use) for probably 15-20 years. We used to affectionately call it the "death ray", eye protection is mandatory. They are an excellent indoor UV lamp. Replacement bulbs will cost you a fortune. Be careful moving them/dropping them when hot as it can kill the bulb. I had recommended this lamp back on Nikongear somewhere. I measured their output spectrum and found with the UV filter in place it does leak some blue and red light, no surprise if you check them in the dark against a shiny metallic subject e.g. spoon or ball bearing.

 

I think both Klaus and I went through the Metz route. I personally didn't like the whole battery setup and it didn't have the versatility of the SB-14 for easily applying flash filters if required. Vivitars 283 & 285 are OK but their output is low so either multiple flashes, get up close or dual flash units. These units have been manufacturer in many different countries over the years resulting in variable reliability, and some have high sync voltage. The big Sunpak was OK (sometimes) unfortunately QC was poor and allowed the use of different window and flash tubes resulting in "you never know how much you need to modify it" so trial and error. Tried the Quantum 150W system but too expensive and not much gain over the Nikon SB-14 power wise. The 400W flash systems are nice and can be purchased fairly reasonably but heat output can be an issue. Then there is the heat ray Bjørn uses.

 

Again about 10 years ago in the lab I was experimenting with UV LEDs and most of them have pathetic output unless you get something specialized like the Nichia UV LED. Output power still hasn't changed that much except on the high end options.

 

But no way to photograph it because it was such a short interval. And too dark.

Before someone corrects me, I do not mean fluoresce. I do mean phosphoresce.

 

Andrea come on - put on your thinking hat ~ multiple exposures and image stacking.

 

The most challenging for me was 30 exposures 15 secs each on slide film for a diamond that "phosphoresced" (actually room temperature thermoluminescence a different process) but had a rapid decay, so longer exposures were no good.

[Andrea B.]

Shane, lots of great additional info about UV flashes. Thank You !!

(I need to add some of these comments to the Stick.)

 

((I have a thinking hat?? Wonder where did I put that darned thing? :D))

[JCDowdy]

((I have a thinking hat?? Wonder where did I put that darned thing? :D))

 

Thinking caps sure have come a long way, I am saving up for a new one!

http://www.kickstart...er-interface-fo

[Andrea B.]

omg !!! That's quite a device.

 

I read a bit about it on the Kickstarter link. And was even slightly tempted to support it. Although after one see's ones brain waves a couple of times the thrill might wear off.

 

Gotta love Kickstarter. I supported a retrospective book for a photographer. And got a lovely book.

And also supported a lens holding device to wear on a belt. Forgetting that I don't actually wear belts.

[igoriginal]

As Iggy said it is pretty easy to wire up a board with a range of LED's to a battery to display a range of wavelengths.

I have seen 365nm LED's with two legs, but shorter wavelengths, I'll have to see what Iggy has or knows ?

I guess I could hope to find a couple of LED's around 320 & 340nm or so, that should cover anything I am likely to need to pass UV for photography with the available Baader or Andrea UV filters.

 

I have them all stashed in my home studio / lab, and haven't opened the packages, yet. I made some crazy (and dare I say overly compulsive and irresponsible) purchases all throughout last year. And now, I am paying the price, by being broke.

 

Now, I have all of these unopened boxes and packages littering the floor of my home studio / lab, and have yet to be sorted out. My wife has been rolling her eyes at me, too. :D (Even though it is MY studio! Haha.) So, I better get my butt in gear, and get to organizing them all. I have just been preoccupied with other work matters.

 

Ummm ... actually, one of the packages of LEDs that I have, are supposedly marked 280 - 400 nm, I kid you not! (Although I paid a near-fortune for just a package of 10 count. $299). Yes, some industries supposedly now use much broader-spectrum LEDs. Apparently, they exist! (You just won't find them used in commercial toys, like flashlights or torches).

 

Supposedly, they are using such LEDs in botany research, as well as micro-organism response, and even epidemiology research.

 

The thing about a mixed UV-LED panel is that the LED beams don't spread much so the light doesn't blend into a broader UV spectrum.

I've had trouble blending my various UV-LED torches. You get one central spot that is blended, but away from that things get blotchy.

I guess you all can tell I'm not much enamoured of UV-LED so far. "-)

 

Oh, and as for all of this talk about LEDs providing "blotchy" or "uneven" light coverage. That's simply not always true. It depends on something called "beam spread", as well as other aspects of the LED design and manufacture. Not all LEDs are created equal.

 

Of course, you're going to get strange, uneven, and undesirable results, if you use commercially-available "torches" (flashlights). They are supposed to be "flashlights", for Pete's sake. So, they cast a very narrow, "spotlight"-like beam, by default. That's what they are designed to do ... and not illuminate evenly in all directions (lest someone wants to blind themselves, while using their own flashlight). B)

 

Alas, though, you cannot judge all LEDs, by just a limited experience to torch-designated designs. Not unless you've had access to a much broader array of LEDs, for many types of applications outside of flashlights. I am almost willing to make a wager, that there are LEDs out there, that WILL rival any other multi-directional UV-based output. You just have to pick the right types.

 

(True, that for the time being, their UV "brightness" will never surpass a hot-blooded UV lamp based on compressed gas / vapor discharge or filaments ... but, then again ... just wait another 5 to 10 years. LED technology has dramatically improved, in a relatively short period. Including an ever-increasing efficiency in light output-per-surface area / wattage ratio).

 

But, no point in yammering over this ... while all of my packages still remain littered all over my floor, for the time being. I suppose we shall see, eventually. :D

[colinbm]

Hi Iggy

Thanks for your help with the Metz flashes, much appreciated, I have a couple on the way :D

 

Those 280-400nm LEDs sound like the real McCoy, no good in the box though, It would be nice to have a camera that could use that range.

 

I just was reading about this very topic of LED spread.

http://www.clearstonetech.com/UVTech.html

This site has lots of interesting links embedded in the text too :D

Cheers

Col

[colinbm]

While I wait for some UV LEDs to arrive to test, I have set up my trusty Sigma Foveon DP2, full spectrum camera with the IRC removed.

I have revisited the giant 400 watt Mercury UV party lamp. I set up the camera with an Amici Prism Spectroscope mounted centrally on the lens.

This prism spectroscope gives a non-linear spread of the colours of the light spectrum. The UV end is scale expanded & the IR end is compressed.

I got this nice shot of the emission lines from 365nm to 850nm. Afterwards I applied a BG39 filter & it completely eliminated the IR end.

Cheers

Col

 

[igoriginal]

Ok. So I went ahead and opened up one of the packages.

 

To my astonishment, the data included with this particular batch actually lists the UV LEDs of having a 230-400 range! These have also been already pre-wired individually with 330-ohm ceramic resisters, in order to make them ready to connect to a standard circuit powered by a 9 volt battery. How convenient! (They also include chrome mounting collars, for each LED!)

 

Attached to this reply is a screen shot of the associated specs (see photo).

 

 

However, note that the "beam spread" of these particular models is listed at 20-degrees ... hence, I'd have to mount them on a gently-tapering spherical-shaped panel, in order to facilitate an even spread. Yes, it can be done. It's just that no one has ever done it, yet (to my knowledge), because the UV photography field still remains quite a niche, after all. I, therefore, endeavor to be one of the pioneers to try my hand at manufacturing UV photography-dedicated (and custom "tuned") lighting panels.

 

I do believe, though, that in a separate package laying somewhere on my studio / lab floor ... I have UV LEDs listed as 280-400nm range, but with a wider beam spread (70-degrees). So, I will look around, and report back.

 

 

Those 280-400nm LEDs sound like the real McCoy, no good in the box though, It would be nice to have a camera that could use that range.

 

Getting such a camera would be quite easy, actually. Alas, it would be finding a lens and an associated filter that could allow for such a deeper transmission (and not break the bank), that would prove to be quite difficult for most people.

 

In fact, I have two full-spectrum-converted cameras that can record UV down to 180 nm! (That's no typo or exaggeration!) That's really that not unusual, actually. What makes this possible, is that the individual whom I paid to do my camera conversions used a different glass, rather than the usual and quite common "Schott WG280." Instead, he implemented Spectrosil 2000, and this permits my camera to record UV radiation down to 180 nm!

 

But ... like I said, the limiting factor, then, is not my camera ... but finding a lens / filter combo that could transmit that low. Even the popular Baader-U would utterly fail at such an attempt.

 

(I have considered trying my luck with "pin hole" body cap attachments on my Spectrosil 2000-converted camera. Granted, the image would never be sharp enough to be used as "documentary / archive quality" ... but, at least I will get a sense of what kind of UV coloration is possible, down to 180 nanometers, since a body-cap-drilled pin hole attachment is nothing but air, and no glass standing between the UV light source and the sensor).

 

I would also have to wear tons of sunscreen (above SPF 100), and wear protective goggles, in order to use some sort of germicidal lamp that is capable of radiating down to the UV-C range for such deeper UV photography (in conjunction with an indoor shooting environment) ... since there is practically almost no usable UV-C in sunlight, at the terrestrial level (below the tropospheric upper boundary of the atmosphere).

 

Anyway, heading to bed, now. Will stay in touch, all. Thanks!

 

Oh, and Col .... I like your shot of the emission lines. I particularly love your "rich full spectrum" photograph / panorama, in the same photo set on flickr. Thanks for sharing!

[colinbm]

Thanks for looking Iggy.

I have some pin hole shots their too.

The hole needs to be small, like 0.3mm & round & burr free :D

Col

[Andrea B.]

In the absence of any actual measurements, pardon me while I retain some skepticism about photographing at 180nm with commercially available DSLRs however be they modified.

 

We do know that the UV-Nikkor/Rayfact or Coastal 60 or other dedicated UV lenses can go to around 200nm and such filters do exist. That should be close enough to 180 - assuming some carcinogenic, cell-destroying source of illumination around 200nm can be found - and you make the photograph from the next room using robot arms. "-)

 

BUT, I need to see some measurements of sensor sensitivity to UV for the mentioned cameras and understand whether the Bayer dyes can pass all UV or not. My closest experiment to 200nm was with the UV-Nikkor, a 290nm narrowband, and my filterless D300. I can tell you that it is very painful trying to get a photograph in strong sunlight at 290nm. I got one, but it required a lot of massaging in the editor. I tried again this week with the Pentax K5 and got almost nothing. But the K5 does have some kind of clear glass filter from its Kolari conversion.

 

 

Edit: Coastal 60 does not go into the 200s. My error. "-)

[Alex H]

In fact, I have two full-spectrum-converted cameras that can record UV down to 180 nm! (That's no typo or exaggeration!) That's really that not unusual, actually. What makes this possible, is that the individual whom I paid to do my camera conversions used a different glass, rather than the usual and quite common "Schott WG280." Instead, he implemented Spectrosil 2000, and this permits my camera to record UV radiation down to 180 nm!

 

I would like to see a hard evidence on that. Do your cameras have the sensor cover glass replaced? Do your cameras have Bayer array and microlenses removed?

[Alex H]

We do know that the UV-Nikkor/Rayfact or Coastal 60 or other dedicated UV lenses can go to 200nm and such filters do exist.

 

One small correction, Andrea, if you please. Contrary to UV-Nikkor/Rayfact, Coastal UV-VIS-IR 60mm 1:4 Apo Macro has a transmission of approximately 40% at 300 nm and approaching its limits close to 275 nm.

Specification sheet here: http://www.jenoptik-inc.com/literature/doc_download/29-1-uv-vis-ir-60mm-slr.html

I wish it would go deeper though :D

[colinbm]

Hi Alex

Do you know the material that the micro-lenses are made from & the transmittance please ?

Cheers

Col

[Alex H]

Hi Alex

Do you know the material that the micro-lenses are made from & the transmittance please ?

Col

 

I wish I was able to provide exact data, but I am afraid it is proprietary information of sensor manufacturers.

[colinbm]

That is probably why some astro-photographers know to scrap off the CFA & micro-lenses............

Col

[Alex H]

That is probably why some astro-photographers know to scrap off the CFA & micro-lenses............

Col

 

I thought it was more to do with the increase of resolution when Bayer array is removed and the image is processed without interpolation of RGB data. I know of very few attempts to do astrophotography in UV spectrum from the surface of our planet. Contrary to astrophotography in infrared spectrum.

[Andrea B.]

There is some false colour solar imagery here: Extreme Ultraviolet Images of the Sun

Absolutely beautiful and fascinating.

 

An angstrom is .1 nanometer.

So 170-300 angstroms is 17-30 nanometers, which is indeed "extreme ultraviolet". :D

That website calls this range "soft X-rays".

 

I include Solar Astronomer on my list of What I Want to Be When I Grow Up.

[igoriginal]

I would like to see a hard evidence on that.

 

I would, too. Which is why I said, earlier, that I want to "get a sense of what kind of UV coloration is possible."

 

Never did I make any absolute claims. If I already knew, without any doubt, what was and wasn't possible ... then I wouldn't need to experiment, would I? :D

 

So, yes. I want to obtain some hard evidence, too (of what the real world capabilities are, rather than claims made by the seller / converter of the camera). I am right there with you.

 

Do your cameras have Bayer array and microlenses removed?

 

So, am I to assume that the Bayer Array, on every single camera, is incapable of deriving any information from UV wavelengths below the typical 300nm line, that the majority of UV photography is usually limited to? Why is the 300nm line so special? And where exactly does a Bayer Array's ability to sense radiation stop, then, if not 300nm?

 

280? 250? 230?

 

Has anyone actually tested a bare (unprotected) Bayer Array in various cameras, for these purposes? If so, is there an on-line link or reference that I can read? I would greatly be interested, if such information is available .... because it would save me a lot of time and conjecture. Thanks!

 

(Otherwise, I would have to do my own experimentation, as planned ... to see what is and isn't possible, even with a Spectrosil 2000 glass directly mounted in front of my sensor, which has replaced the removal of the hot mirror and AA filter).

[Andrea B.]

BTW, the developing "consensus of opinion" is that our broadband DSLRs probably cannot photograph below 300nm.

This is anecdotal and experience-based evidence at this point.

 

If I knew where to go, then I would take my broadbands and pay to have them measured just so we could finally have some facts.

[Shane]

In fact, I have two full-spectrum-converted cameras that can record UV down to 180 nm! (That's no typo or exaggeration!) That's really that not unusual, actually. What makes this possible, is that the individual whom I paid to do my camera conversions used a different glass, rather than the usual and quite common "Schott WG280." Instead, he implemented Spectrosil 2000, and this permits my camera to record UV radiation down to 180 nm!

 

Just because the ICF was replaced with a 180nm transmitting glass doesn't mean it will perform in that region.

 

In a front illuminated sensor, the cover glass, Bayer array, microlenses, overlying passivation, polysilicon etc will all attenuate deep UV penetration either due to absorption or reflection. In addition, the junction depth on consumer DSLRs & P&S is optimized for green wavelengths around 540nm making the junction way too deep to ever receive 180nm wavelength light. Even a back illuminated sensor would require special construction to be optimized for UV.

 

From an earlier post I made

"For optimized 350nm you would require a junction depth around 0.01um i.e. 10nm (10nm of p+Silicon is about 70 atomic layers!!). These depths also do not account for overlying layers such as passivation which unfortunately have an additional attenuation effect on UV. Also against the odds, is the fact that UV light is more highly reflected from these structures than blue light."

 

The photon absorption length (penetration depth) in silicon for 180nm wavelength is around 6nm.

 

With a typical junction depth around 2um, around 90% of 540nm wavelengths are absorbed in the space-charge region compared to NONE for UV wavelengths. UV/Blue wavelengths are typically collected through lateral diffusion of the carriers that are generated on or close to the vicinity of a photodiode peripheral. Dark current at the surface causes recombination centers and defects which absorb surface generated electron hole pairs. This results in poor UV/Blue sensitivity.

 

Reflectivity of silicon in the UV/blue region exceeds 50% reaching 73% at 270nm. Any anti-reflective coatings would require optimization for UV wavelengths. Polysilicon structures would require being replaced by indium tin oxide for enhanced UV response. Silicon nitride, silicon oxy-nitride passivation etc would need to be replaced with silicon dioxide. Cover glass, Bayer array and microlenses would need to be removed. The junction depth would need to be reduced to totally impractical depths. The list goes on. Not impossible goals but certainly not reality for a consumer camera.

[igoriginal]

... assuming some carcinogenic, cell-destroying source of illumination around 200nm can be found - and you make the photograph from the next room using robot arms. "-)

 

Hardly something so dramatic, Andrea. Hehe. B) You're getting a bit carried away with the "robotic arms." Although it is hilarious! :D

 

(Keep in mind, though, I am not planning on photographing with X-ray or Gamma-ray, here. :D)

 

My mentality is to put on a crap load of the strongest sunscreen, wear protective clothing on top of that, and put on a welder's mask or goggles (At least that would be sufficient for UV-B range, but probably not even enough for UV-C). Of course, that would be quite clunky, unwieldy, and cumbersome ... and likely not realistic ... since you need finer articulation of equipment and button pressing through the entire procedure.

 

Thus, realistically, the best thing to do (even if I manage to actually find the right kind of equipment for the test), is to set everything on a timer, then leave the room for a moment, while the exposure is taken. This is something I imagine would be the least cumbersome, if such a test is possible at all.

 

(The only other thing to consider, is to perhaps erect a small, opaque tent around the entire testing area, to protect the paint on my walls, as well as the artwork in my room, from damage).

 

I imagine that my biggest hurdle for such a test ... would be to either find a light source that can actually transmit UV-B range only (without leaking any other spectral range), or finding a narrow-band filter that can transmit a specified UV-B bandwidth.

 

BUT, I need to see some measurements of sensor sensitivity to UV for the mentioned cameras and understand whether the Bayer dyes can pass all UV or not.

 

And this is the principle question that has captured my imagination: Just how far down the radiation spectrum can a bare (unprotected) Bayer Array pick up on? Why are we to assume that the dyes inside the Bayer Array cannot interact with any bandwidth of radiation? What's to stop them from doing this? Why would there be sudden resistance to any specific bandwidth? (Regardless of whether or not the incoming information is interpreted "correctly", or assigned "proper color.").

 

My assumption is that the dyes in the Bayer Array would respond to any radioactive range, short of being destroyed (once the energy wavelength would become damaging to the Array, itself).

[igoriginal]

Just because the ICF was replaced with a 180nm transmitting glass doesn't mean it will perform in that region.

 

In a front illuminated sensor, the cover glass, Bayer array, microlenses, overlying passivation, polysilicon etc will all attenuate deep UV penetration either due to absorption or reflection. In addition, the junction depth on consumer DSLRs & P&S is optimized for green wavelengths around 540nm making the junction way too deep to ever receive 180nm wavelength light. Even a back illuminated sensor would require special construction to be optimized for UV.

 

From an earlier post I made

"For optimized 350nm you would require a junction depth around 0.01um i.e. 10nm (10nm of p+Silicon is about 70 atomic layers!!). These depths also do not account for overlying layers such as passivation which unfortunately have an additional attenuation effect on UV. Also against the odds, is the fact that UV light is more highly reflected from these structures than blue light."

 

The photon absorption length (penetration depth) in silicon for 180nm wavelength is around 6nm.

 

With a typical junction depth around 2um, around 90% of 540nm wavelengths are absorbed in the space-charge region compared to NONE for UV wavelengths. UV/Blue wavelengths are typically collected through lateral diffusion of the carriers that are generated on or close to the vicinity of a photodiode peripheral. Dark current at the surface causes recombination centers and defects which absorb surface generated electron hole pairs. This results in poor UV/Blue sensitivity.

 

Reflectivity of silicon in the UV/blue region exceeds 50% reaching 73% at 270nm. Any anti-reflective coatings would require optimization for UV wavelengths. Polysilicon structures would require being replaced by indium tin oxide for enhanced UV response. Silicon nitride, silicon oxy-nitride passivation etc would need to be replaced with silicon dioxide. Cover glass, Bayer array and microlenses would need to be removed. The junction depth would need to be reduced to totally impractical depths. The list goes on. Not impossible goals but certainly not reality for a consumer camera.

 

Thank you so much for that very detailed information. That is most helpful!

 

So, essentially, unless I have access to very specialized testing equipment that is very specifically designed for such testing ... then I should forget about even trying, huh? :D

 

(Am I to assume, then, that the Spectrosil 2000 converted "Astrophotography-optimized" camera in my possession is mostly a gimmick? Or do you think it is possible to at least photograph down to the upper 200nm range, under the right conditions? Wouldn't a Newtonian or Dobsonian reflector telescope, using mostly mirrors - instead of refractive glass - be capable of photographing emission lines below 300nm? The reason I ask, is because I own a few telescopes, including a Newtonian, and I have mated my cameras for astrophotography before. Just haven't tried playing around with my Spectrosil 2000-converted camera, yet.)