Building a Variable Contrast LED head for a Durst 138

[Larry Kellogg]

You know, we should figure out how to adapt your design to the Leitz V35, like the Heiland unit. :-)

[Larry Gebhardt]

You know, we should figure out how to adapt your design to the Leitz V35, like the Heiland unit. :-)

I had someone ask me about doing this. I don't think it would be too hard, especially with the stars with multiple LEDs on them. If I remember correctly there isn't a lot of room in the light chamber, but one star on a heat sink should fit. Put a diffuser an inch or so ahead of that and I think it would work (though you might not need it - I forget if there's a diffuser built in). I think 1 green and 1 blue LED should be bright enough. But I've only seen pictures, so I'm not sure of any of this. With one LED each you would want to make sure any driver can handle the low forward voltage.

If you just want to use the color filters in the V35 I think you could get any number of warm white or neutral white LED arrays that could fit. Then just use a simple power supply like the MeanWell LPC-60-1050 plugged right into a timer.

There are a huge number of LED arrays from China on ebay. The prices are very low. No idea of the quality, or if the LEDs are really the name brands they say. But I plan checking a few of them out.

[Larry Gebhardt]

There are certainly many ways to get a LED to illuminate, but after looking at the link what stands out as pure genius is Larry's clever integration of his LEDs into the existing DURST L138 structure. Fantastic work!

Thanks.

[Eric Woodbury]

If I were going to tackle this problem, I'd go with Phillips LumiLEDs. For the blue LED, I'd use something from the UHLV series so that I could get the max contrast during printing. You'll need that 425 nm or shorter wavelength for best results. If you don't care about high contrast so much, you can use cheaper devices. Here's a link to this series of LEDs at Mouser:

http://www.mouser.com/search/refine....9&Keyword=lhuv

The green LEDs are easier to find and cheaper, but nice to keep with the same size and manufacturer.

Before I forget, a warning about eye safety. Although the blues are in the visible band, they can still be blinding and damaging to your eye. Do not operate at high flux without either containing the light or wearing dark sunglasses. Blue light is not seen efficiently by the eye and your iris will not respond properly (it won't stop down enough). Be very careful. Nothing worse than a blind photographer. (no letters please. I understand there are a few blind photogs and I'll bet they are fine people). So, I shouldn't say 'worse'. Still, protect your peepers.

I'm not sure if you need a closed loop setup or not. LED flux output will change some with temperature and age. Temperature you can stabilize and age you can probably live with. High junction temperature is the enemy of all semi-conductors, so keep the junction under 80C and it should last forever. Do your calcs. A little fan could do wonders. I know folks don't like fans, but if you run a small fan at half its rated voltage, it will make your heatsink effectively larger and it won't vibe anything. You can also put the fan in a control loop to get it to control the head temperature.

That said, you could drive the LEDs with a current. I'd probably use a switcher with current feedback from Linear Tech, because I'm so familiar with their products. They make some nice modules so you wouldn't even have to design your own power supply. Some are specifically for LEDs such as crosswalk signs and the rest. I'd figure out a good voltage that is easy to make, then find out how many LEDs you can put in series to comfortably do this. Maybe 3 or 4 in series for a 12 volt system or more for a 24-28V system. Figure out how to space them on your pcb to get even lighting. There is a new acrylic diffusion material, which I have some but have not tried, that is for diffusion. Suppose to be high efficiency. It uses microscopic air bubbles instead of white. Looks white, but they say it's air.

I don't know if possible, but if one could make a simple enough design, then you wouldn't need a pcb. I love pcbs, but they require layout and manufacture and sometimes it's just easier to build a project on vector board.

Except for a small 35mm version of this concept, I'm going to keep using my V54 coldlights. They get the job done and don't distract me from the darkroom. I'm happy to help with design if you want to drop me a line. I do electro-optical analog design all day anyway.

Going back and reading more thoroughly, most of this is what Larry said. Good job. Carry on and good luck. -EW

[Larry Kellogg]

Wow, great stuff, thanks Eric! I think I'm out of my element, LOL. I'm a software guy, and never did that well in my electrical engineering classes.

So, I'm willing to help contribute money to get something produced but I probably won't be able to build it myself.

As for cold lights, I'm glad they work well for you, I'm not always happy with the results, probably because some of my negatives do not have enough contrast.

[Larry Gebhardt]

Thanks Eric. Have you tested the 425nm vs the 450nm (royal blue) vs the 465nm (blue) LEDs with VC paper? I didn't do any testing between the 450nm and the 465nm LEDs, but I did look at the Ilford spec sheets. The one for the new Classic shows peak sensitivity in the blue range at about 435nm, so it may make a big difference. However I do get good results with the 450nm royal blue LEDs. My assumption was the 450nm may make the head not as fast, but wouldn't really result in lower contrast. That may not be a safe assumption.

Thanks for warning people about the blue LEDs and the potential safety issues. I find the green also need to be treated with care. They are both too bright to safely look at when running anywhere near full power.

As far as temperature goes I found that there is very little rise in temp with the simple heat sink (a thick aluminum plate) when the LEDs are on for only a few minutes. The area around the LEDs was measured well under 80C, but I don't know a good way to really measure the junction temp. I have a finned heat sink as well, but decided it wasn't needed for how I use the system. I'm not sure what temp the system would stabilize at with prolonged use.

The Linear Tech drivers look good. Thanks for bringing them to our attention. Their site seems to have some nice tools to find the correct driver.

Is the diffusion material you mention called Makrolon? I've thought about trying to find some small samples of this, but there isn't a supplier near me. I didn't find one online.

[Tin Can]

Thanks Eric. Have you tested the 425nm vs the 450nm (royal blue) vs the 465nm (blue) LEDs with VC paper? I didn't do any testing between the 450nm and the 465nm LEDs, but I did look at the Ilford spec sheets. The one for the new Classic shows peak sensitivity in the blue range at about 435nm, so it may make a big difference. However I do get good results with the 450nm royal blue LEDs. My assumption was the 450nm may make the head not as fast, but wouldn't really result in lower contrast. That may not be a safe assumption.

Thanks for warning people about the blue LEDs and the potential safety issues. I find the green also need to be treated with care. They are both too bright to safely look at when running anywhere near full power.

As far as temperature goes I found that there is very little rise in temp with the simple heat sink (a thick aluminum plate) when the LEDs are on for only a few minutes. The area around the LEDs was measured well under 80C, but I don't know a good way to really measure the junction temp. I have a finned heat sink as well, but decided it wasn't needed for how I use the system. I'm not sure what temp the system would stabilize at with prolonged use.

The Linear Tech drivers look good. Thanks for bringing them to our attention. Their site seems to have some nice tools to find the correct driver.

Is the diffusion material you mention called Makrolon? I've thought about trying to find some small samples of this, but there isn't a supplier near me. I didn't find one online.

How much Makrolon do you need?

[Larry Gebhardt]

How much Makrolon do you need?

I'd like to test the diffusion box with it in place of the acrylic, so a piece about 9"x9" would do it. Do you have a supplier for this that's easy to order from for small amounts (single sheets)?

[Tin Can]

I'd like to test the diffusion box with it in place of the acrylic, so a piece about 9"x9" would do it. Do you have a supplier for this that's easy to order from for small amounts (single sheets)?

No I don't, but have a piece I can spare for experiments. I don't find it to be miracle diffusion material, but as you know it's specifically for LED point light diffusion. It's really used for heat molding large LED display signs, like a Burger King sign.

PM with what you want and address. No charge for it, no charge for shipping.

I don't have any more free samples, so anybody else, please do not inquire.

[Drew Wiley]

Any plastics shop can cut you a piece of diffusion acrylic. What is important is not to order plain"white" plexiglas or whatever, because it will be too dense, but what they generally market as "sign white", because it will be translucent, to a degree obviously dependent upon the actual thickness. Thicker plastic can hypothetically be ground down at the edges and corners to match any illumination falloff due to your specific system, lens, etc, though there are a few tricks to doing that. Otherwise, concerning another question, Chinese LED arrays tend to be temperamental due to fragile wiring. They don't seem to last very long.