Making an UV enlarger

[ic-racer]

What's interesting to me is that in the early days of this thread replies consisted of reasons why this just wasn't possible and this negativity lasted until 2021. Now we have a working example so what couldn't possibly be done is done

Is it just possible that there are lessons here for those whose ideas of what works and what doesn't or won't tend to be rather fixed

pentaxuser

What was the key item to make this possible? LED with less heat than arc lamp, allowing long exposures without burning up the lamphouse?

 

[koraks]

What was the key item to make this possible? LED with less heat than arc lamp, allowing long exposures without burning up the lamphouse?

Indeed. Back in 2012 when this thread started, UV LEDs with this kind of power were simply not available commonly - or indeed, at all. See this publication from 2013 (published 2014): https://iopscience.iop.org/article/10.1088/0268-1242/29/8/084004/pdf
LEDs that were cutting edge technology, only available from select manufacturers at exorbitant prices, are now cheaply and commonly available.
Any other available technology than LED would indeed bring serious heat dissipation problems that are very unpractical to solve in a home setting.

So yeah, LED technology has been the defining factor here.

 

[ic-racer]

Are any high power LED UV lamps MR-16 base?
Are diffusion enlargers out of the question? Seems like UV would damage the mixing box plastic.
I don't know anything about reciprocity failure of UV sensitive materials, but if the enlarger is not that efficient, would multi hour or day exposures work?

Thanks.

 

[htmlguru4242]

Are any high power LED UV lamps MR-16 base?
Are diffusion enlargers out of the question? Seems like UV would damage the mixing box plastic.
I don't know anything about reciprocity failure of UV sensitive materials, but if the enlarger is not that efficient, would multi hour or day exposures work?

Thanks.

I doubt that a UV lamp of this power in MR-16 would be possible with present tech, due to heat dissipation. High power LEDs can be damaged or outright burnt out almost instantly if they aren't properly cooled.

I'd imagine that condensers / lenses with a point source are being used here from an efficiency perspective. From an optical standpoint, I can't speak to whether a diffusion enlarger would work. Regarding plastic damage, at the wavelengths, powers, and exposure times used, UV-induced damage seems unlikely.

 

[koraks]

From an optical standpoint, I can't speak to whether a diffusion enlarger would work.

No doubt it would, and the issue brought up by @ic-racer would not bug me too much. First of all, a big array of multiple LEDs is already somewhat close to a diffuse light source in practice, and an additional diffusor can be fashioned out of frosted/ground glass. It won't be damaged by UV, but it will of course eat up quite a bit of the useful energy.

I doubt that a UV lamp of this power in MR-16 would be possible with present tech, due to heat dissipation.

Agreed. That's currently out of the question. A true point source (i.e. similar emitting geometry to a filament bulb) LED with high power (100W and above) is currently not very feasible - maybe in a lab setting with special provisions for liquid cooling. Even if brought to the marketplace, it would never become a drop-in replacement for a halogen bulb given te cooling apparatus required.

 

[Petrochemist]

No doubt it would, and the issue brought up by @ic-racer would not bug me too much. First of all, a big array of multiple LEDs is already somewhat close to a diffuse light source in practice, and an additional diffusor can be fashioned out of frosted/ground glass. It won't be damaged by UV, but it will of course eat up quite a bit of the useful energy.



Agreed. That's currently out of the question. A true point source (i.e. similar emitting geometry to a filament bulb) LED with high power (100W and above) is currently not very feasible - maybe in a lab setting with special provisions for liquid cooling. Even if brought to the marketplace, it would never become a drop-in replacement for a halogen bulb given te cooling apparatus required.

In lab settings a Deuterium lamp is the historical source for high energy UV, this has now typically been replaced by unfiltered Xenon tubes. I've seen a Xenon lamp that claims 150W.
AFAIK LEDs are unlikely to replace either of these for spectroscopy as wide bandwidth sources are usually desired.
Where very high power monochromatic sources are wanted, Hollow cathode lamps & Electrodeless discharge lamps are used but these give highly precise wavelengths

Unfortunately each of these needs a highly specific power supply, and I haven't found them practical to sneak home even when the instrument using the source is being scrapped.

 

[htmlguru4242]

In lab settings a Deuterium lamp is the historical source for high energy UV, this has now typically been replaced by unfiltered Xenon tubes. I've seen a Xenon lamp that claims 150W.
AFAIK LEDs are unlikely to replace either of these for spectroscopy as wide bandwidth sources are usually desired.
Where very high power monochromatic sources are wanted, Hollow cathode lamps & Electrodeless discharge lamps are used but these give highly precise wavelengths

Unfortunately each of these needs a highly specific power supply, and I haven't found them practical to sneak home even when the instrument using the source is being scrapped.

Given the expense, PSU requirements and finite lifetime of the aforementioned light sources, I think that the scientific community would welcome broadband high-power UV LEDs, but I feel like these are some years away, though I've seen some stuff for IR/VIS. Decent light sources can be assembled from multiple UV leds of different wavelengths, as their spectrum is wide-ish, but they become exponentially more expensive and lower power at wavelengths below ... 365nm or so. Obviously this kind of a configuration also has the additional disadvantage of requiring optics to combine multiple light sources in a uniform way - not that complexity or expense has ever been an issue in high-end spectroscopy and such, but I suppose that since D2 and Xe lamps work just fine, why fix what ain't broke?

For what it's worth, power supplies for both show up on ebay from time to time. I've hesitated to buy them because they're often "untested", and we all know what that means, but it might be worth a try. A 100W+ Xe short arc lamp could be a lot of fun, at the very least.

I do have a line on a mercury lamp from an old print / graphics arts shop (which I believe is basically a mercury vapor streetlight arc tube without the protective glass bulb), but I'm not sure how suitable that would be for an enlarger application, because the UV and heat output is ... considerable to say the least. I'm also much more comfortable working with LED voltages as opposed to the 200+V at high current for arc lamps.

I have a couple of brand-new deuterium lamps hanging around (they were like $15 each at the MIT Swapfest), and I'd love to have a power supply for them. I've got a really nice optical bench from an Ocean Optics USB2000 (which I also got for a song and a dance) that I'd love to make part of a UV/Vis spectrophotometer.

I've always been blown away by the build quality, precision and attention to detail of high-end scientific & optical equipment. But I suppose that's what you expect in a market segment where "inexpensive" products start at $3K +

 

[koraks]

I just came across a few posts from someone who's also working on an UV enlarger: https://groups.io/g/carbon/topic/uv_enlarger/107920492
He uses two acrylate fresnel lenses to make a virtual point source. Seems to work surprisingly well!

 

[alanbradford]

Koraks, Thanks for the details, Alan

 

[koraks]

Very welcome; to be clear, I'm not the author of the posts I linked to - just an admirer!

 

[sigurd]

I'm the one doing the UV enlarger mentioned over at the carbon transfer printing group at groups.io. Saw this last post just now. I learned a lot from this thread since Douwe Krooshof did his enlarger. If you read this Douwe, thank you, and the same to all contributors of this thread.

My alt process of preference is carbon transfer printing. And although I would consider this enlarger project to still be in the test stage, I can print 11x14 inch prints from either 4x5 inch or 6x7cm negs within reasonable printing times. This is quite suprising to me, considering the amount of UV needed for carbon transfer.

I have some issues with unevenness of light that I hope to get rid of. Also, there are tweaks to be done to make it more light efficient that will further reduce printing times.

Plan to do a new horizontal enlarger setup in a couple of weeks, as per a tip from Gene Townsend over at Groups. Because of limited ceiling height, a horizontal enlarger will give me the opportunity to test a more varied selection of lenses. Not only enlarging lenses, but also taking and repro lenses. If some lenses with bigger entrance pupil will work, new possibilities open up with bigger and higher wattage LED arrays.

Has been a very interesting project so far.

 

[koraks]

Hey, welcome aboard @sigurd! Super thrilled to reading your comment above; please keep us update on the horizontal enlarger project as well as the light source upgrades. On this forum, @AndrewBurns is also working on imaging systems for alt. process prints, and he's been eyeballing carbon specifically as well. The twist is that he's trying to do this with a direct digital imaging method, with the main contender being 3D-printing oriented LCD's as the imaging device. There's some interesting reading on that here: https://www.photrio.com/forum/threa...negative-in-alt-process-contact-prints.207526

 

[sigurd]

Hey, and thanks! Will do.

Wow, the project of @AndrewBurns is interesting.

 

[AndrewBurns]

Hey, and thanks! Will do.

Wow, the project of @AndrewBurns is interesting.

Cheers! I consider the contact printing project to be basically done, it works quite well and very reliably, as long as you only want to make a print as large as your LCD screen. I couldn't manage to get the prints to reliably tile together and after a few weeks of trying decided that I would enjoy spending my free time and energy elsewhere. I'll definitely still keep my contact printing setup around as it's probably going to be the fastest and easiest way for me to make small alt-process prints.

What I'm currently embarking on is a digital UV projector, basically making a horizontal UV enlarger of almost exactly the same design as yours, but using an LCD screen in place of the negative. I'm using a 400W UV light source because the LCD screen eats up so much of the light, and an EL-Nikkor 210mm enlarging lens (which my research indicates should work well at 380nm). The LCD screen is fairly similar in size and aspect ratio to a 5x7 negative.

I plan to start a new thread to cover my design and experimentation when I have more to show.

 

[sigurd]

Very interesting stuff! Looking forward to the thread.

What are the physical dimensions on that 400w LED? Problem I encountered was with the LED chip not being a true point light source, and that it needs to not be bigger than the entrance pupil of the enlarger lens to harness all the light. There might be viable ways around it, but always on the expense of light efficiency. The eternal tradeoff in building a UV enlarger...

 

[koraks]

What are the physical dimensions on that 400w LED?

Always a lot bigger than the entrance pupil on a typical lens. Your concern is valid; it's presently a limitation for UV enlargers.

 

[AndrewBurns]

Yep the LED array is something like 30x75mm active area, larger than the lens entrance pupil (which is fairly large given the lens is 210mm f5.6). The LED is an array of discrete surface-mount modules, each of which has its own fused quartz lens with a 60 degree projection angle. Each module is rated to something like 15W maximum power dissipation, and they measure a few mm square, so you can roughly work out how much power you can really cram into the entrance area of a lens.

The reason I'm still going larger is that I'm banking on some of the light from the edge LED chips still making it through the lens anyway, we'll see how well that goes. It's not going to be a true point source, but it never will be unless you can get away with significantly less power. I'm sure that with some additional optics you could compress the light emitted by an array of LEDs down to a smaller effective size, but every optical element introduces its own losses and I don't have the skills or abilities to design such a system. So for the time being I'm going to have a crack with what I've got and see what I get.

 

[sigurd]

Yeah I see. Wonder if something like an Apo-Ronar 600/9 or similar could work. Not sure about number of lens elements in that or UV transmittance, but it has a crazy big opening of something like 65mm. Even though it is f9, the light loss might be compensated with your large LED array.