UV LED Longevity in exposure units

[mark]

I've been doing lots of reading and a few years ago there were quite a few threads about UV LEDs. I have not seen a thread that has looked at the LED usage over time. One thread had the water proof coating turn yellow, but removing it got rid of the problem. On Another forum Sandy King said the strips of LEDs did not last but really nothing since then.

So, people using LEDs:

Are you still using them?
How have they held up?

 

[Fraunhofer]

Hi Mark,
I have an exposure box which is about 1 year old, uses 12 3W UV LEDs. I made in excess of 100 prints, so a total of 10-20 hours of operation. The LEDs are fine as well as the exposure times seem to stay about the same.

I believe, there are two keys to this: One, drive them with a constant current source and make sure that the current falls within specs. Two, properly heat sink them; they get quite warm even with a big aluminum heat sink. If I would builder a bigger unit I would add cooling fans.

Hope that helps a bit.

 

[Mainecoonmaniac]

I love my UV LED exposure box. There's no warmup time and more efficient than fluorescent tubes. However, they are semiconductors and if they get too hot, they'll get cooked.

 

[mark]

If you don't mind me asking, were these the LEDs in strips or did you buy individual LEDs?

 

[Mainecoonmaniac]

If you don't mind me asking, were these the LEDs in strips or did you buy individual LEDs?

They're individual. I soldered them in series and glued them on an aluminum sheet to dissipate the heat. https://www.ebay.com/itm/122516992089

 

[mark]

They're individual. I soldered them in series and glued them on an aluminum sheet to dissipate the heat. https://www.ebay.com/itm/122516992089

Are these better than the strips? And which base did you get. I'm learning as I go.

 

[Mainecoonmaniac]

Are these better than the strips? And which base did you get. I'm learning as I go.

I really can't say for sure that they're better than strips. I never used LED strips before I made my lightbox a few years back and the ones I mentioned emitted the wavelength that is suited for alternative photo processes. The ones I used work with salt prints, cyanotypes and Ziatypes.

 

[REAndy]

Are you connecting them directly to AC voltage? I'm wondering, because if you put enough of them in series, the voltage drop of each LED would consume the 120 VAC household voltage. Then I'm thinking of you made another duplicate string of LEDs in series, and put them opposite polarity as the first set of Led's, as one set turns on, the other's will be off, and then with the AC reverses, the 2nd string would illuminate and the 1st if off. I guess what I'm trying to ask is do you think the 60hz flicker would be canceled using 2 strings, one in 1 polarity, and the other in backwards polarity. Of course, if you are converting to DC what I ask is null and void.

 

[Mainecoonmaniac]

Are you connecting them directly to AC voltage? I'm wondering, because if you put enough of them in series, the voltage drop of each LED would consume the 120 VAC household voltage. Then I'm thinking of you made another duplicate string of LEDs in series, and put them opposite polarity as the first set of Led's, as one set turns on, the other's will be off, and then with the AC reverses, the 2nd string would illuminate and the 1st if off. I guess what I'm trying to ask is do you think the 60hz flicker would be canceled using 2 strings, one in 1 polarity, and the other in backwards polarity. Of course, if you are converting to DC what I ask is null and void.

LEDs only work with DC. My friend helped me design the box which are LEDS soldered in a series and the power supply does connect to AC. It's an old repurposed high wattage laptop power supply.

 

[REAndy]

LEDs only work with DC

Actually an LED (like a regular diode) works like a check valves does in plumbing (lets water flow in one direction only). But in electronics, a diode only lets current flow in one direction. If you connect a diode up to an AC circuit, it will only let current pass when the diode is forwarded biased. A typical use for this is what is called a full wave bridge rectifier, where 4 diodes are used to convert AC to DC, (bumpy DC, but put that thru a capacitor, and you can smooth out that bumpy DC into nice and smooth DC. If you hook up a LED to AC power, the light will turn on and off at the hertz of the AC power. You will need lots of LEDs connected in series because a single LED is NOT going to take 120V. Also a resister is going to be required as a current limit or you will burn out your LEDs. The resister is going to be required if you use DC or AC. This would be for individual LEDs. Typically the strips already have the current limiting resistor in place.

 

[mark]

That went way over my head.

 

[J 3]

There are many kinds of diodes used for different purposes : as a one way switch, as a variable capacitor, to mix radio frequencies, as a voltage regulator, etc... Light emitting diodes (LEDs) of course convert some of the energy they receive as light and convert the rest to heat. They operate in a relatively narrow range of DC voltages, and especially at the higher power ratings are subject to thermal overload (basically when driven incorrectly heat causes them to use more power, which generates more heat, which builds to the point where they fry themselves). Specialized diode drivers have been created to handle these characteristics at higher powers. Some things to note:

1) LEDs are not always more efficient that fluorescents or mercury vapor lamps. It took a very long time to develop efficient LEDs and then even longer to develop bright ones. Also the material and technology changes as the frequency of the LED does. So where as it is easy to find a white light LED that's more efficient than a fluorescent bulb today, the same might not be true at all for a 365nm UV LED. In general the shorter the wavelength, the less efficient the LED will be any given year (they do get better over the years). LEDs tend to produce more directional light though which can be used to more efficiently illuminate the negative. So in total the efficiency is (efficiency of bulb * efficiency of transport to get that light onto the photo chemistry * efficiency of the chemistry at the wavelength put out by the bulb). This equation can (and I suspect often does) come out in favor of the mercury bulb. The price per watt of bulb is certainly higher for LEDs as well. I think LEDs have a place though.

2) LEDs do not always have long life. The ones used for lighting do, but the ones used for curing UV resin and the like only have moderate lives.

3) There are several places that sell electronic parts in small lots and these pages have the spec-sheets for the stuff they sell. It can be a great place to back of the envelop figure out what you're going to get with any given LED (Life time, efficiency, illumination angle, power requirements, heat sink requirements, etc).

 

[Kino]

They're individual. I soldered them in series and glued them on an aluminum sheet to dissipate the heat. https://www.ebay.com/itm/122516992089

What wavelength did you choose and how many? How big is your exposure box?

 

[Mainecoonmaniac]

What wavelength did you choose and how many? How big is your exposure box?

My box is 13"X 16". I use a total of 12 400nm-410nm LEDs.

 

[Kino]

My box is 13"X 16". I use a total of 12 400nm-410nm LEDs.

Thank you. I had no idea it took that many to cover that area. Did you have to experiment to avoid hot spots?

 

[Mainecoonmaniac]

Here's a pic.

 

[Kino]

Nice. Looks like a big extruded heat sink there...

 

[removed account4]

that looks beautiful !
my tanning lamp is disgraced

 

[Mainecoonmaniac]

Nice. Looks like a big extruded heat sink there...

In hindsight, it's overkill. I'd use a thick sheet of aluminum instead.

 

[Mainecoonmaniac]

that looks beautiful !
my tanning lamp is disgraced

You're so kind. It's cobbled together with scrap lumber and parts from eBay and a discarded laptop power supply. It works.

 

[J 3]

That is a really cool looking UV box Mainecoonmaniac. One question though. What kind of chemistry are you exposing with it? At 400nm I could see salt print working, but Some Sandy King articles had lead me to believe you couldn't get any significant action on Platinum or Carbon Transfer chemistry with 400 which is why I was mentioning 365nm. It would be cool if I were wrong because there are a lot better LED options at 400nm.

 

[Mainecoonmaniac]

That is a really cool looking UV box Mainecoonmaniac. One question though. What kind of chemistry are you exposing with it? At 400nm I could see salt print working, but Some Sandy King articles had lead me to believe you couldn't get any significant action on Platinum or Carbon Transfer chemistry with 400 which is why I was mentioning 365nm. It would be cool if I were wrong because there are a lot better LED options at 400nm.

Thanks. I've exposed cyanotypes, Ziatype, Kalitype and salt prints. When I first made the box, I tried one LED on cyanotype and it worked. Knowing that, I built the box. What ever process you plan to use your box on, try a small strip of LED or a single one first. The average standard processing time for the process is about 7 minutes. My first concerns about LED light box are selecting correct wavelength, the largest size I plan to print and keeping them cool enough so they don't cook. So far, it's worked well. I use it on a darkroom timer and it turns off when the exposure is done. I also built a bigger florescent box that uses actinic tubes I got from Petco. I have to leave the box on for the duration of the print session because they require warm up time. Good luck in building your box. If it's well built, it will give a lifetime of good use.

 

[Minas Stratigos]

Thanks. I've exposed cyanotypes, Ziatype, Kalitype and salt prints. When I first made the box, I tried one LED on cyanotype and it worked. Knowing that, I built the box. What ever process you plan to use your box on, try a small strip of LED or a single one first. The average standard processing time for the process is about 7 minutes. My first concerns about LED light box are selecting correct wavelength, the largest size I plan to print and keeping them cool enough so they don't cook. So far, it's worked well. I use it on a darkroom timer and it turns off when the exposure is done. I also built a bigger florescent box that uses actinic tubes I got from Petco. I have to leave the box on for the duration of the print session because they require warm up time. Good luck in building your box. If it's well built, it will give a lifetime of good use.

Hi. I am using a LED box for Pt/Pd, about 8mins exposure for a Dmax of 1.35, Na2 mix. I use 2 x 5m LED strips, cut and connected in parallel on an 8AMP power supply. It works for 5 months and i am not facing any trouble from the box itself.

 

[DonF]

Actually an LED (like a regular diode) works like a check valves does in plumbing (lets water flow in one direction only). But in electronics, a diode only lets current flow in one direction. If you connect a diode up to an AC circuit, it will only let current pass when the diode is forwarded biased. A typical use for this is what is called a full wave bridge rectifier, where 4 diodes are used to convert AC to DC, (bumpy DC, but put that thru a capacitor, and you can smooth out that bumpy DC into nice and smooth DC. If you hook up a LED to AC power, the light will turn on and off at the hertz of the AC power. You will need lots of LEDs connected in series because a single LED is NOT going to take 120V. Also a resister is going to be required as a current limit or you will burn out your LEDs. The resister is going to be required if you use DC or AC. This would be for individual LEDs. Typically the strips already have the current limiting resistor in place.

LEDs are very intolerant of reverse voltage and will break down far more easily than proper rectifier diodes. Connecting an LED to AC will cause it to fail in short order, even with a proper current limiting resistor. When reverse biased, that resistor drops zero voltage, meaning the entire reverse voltage appears across the LED, subjecting it to breakdown even more quickly.

Best,

Don

 

[Dibbd]

LEDs are very intolerant of reverse voltage and will break down far more easily than proper rectifier diodes. Connecting an LED to AC will cause it to fail in short order, even with a proper current limiting resistor. When reverse biased, that resistor drops zero voltage, meaning the entire reverse voltage appears across the LED, subjecting it to breakdown even more quickly.

Best,

Don

Most LEDs are rated for 5V reverse bias and some say they are not designed for reverse bias. However, I've seen a cheap chinese plug extension with a single indicator LED in series with a resistor right across the AC mains. It clearly wouldn't pass any safety standards but it did seem to work. It leads me to believe that the 5V reverse bias limit advertised is just an arbitrary figure and in practice they are much more robust. Having said that, I don't condone connecting strings of LED's directly across the mains at all.