Advice for DIY UV exposure box

[electrosonic]

I am looking to make a simple, small exposure box using UV LEDs. There are many projects on the web about this. The thing that is unclear to me is how to figure out how far to put the led array behind the glass. Should I move it far enough away so that each LED's illumination hits the entire exposure surface or is there some magic distance based on the viewing angle of the leds?

Any insights?

Thanks,
Andew.

[R.G.]

I didn't bother with LEDs. I bought two 24" fluorescent fixtures and unfiltered UV bulbs (note: NOT germicidal bulbs, which are a vision hazard). I made an exposure table out of an old printer stand that let me alighn stuff on top, mount the bulbs under the top surface, and expose the films about 6" down. Worked the first time. I did one test exposure (I have a chemical photography hobby in the deep past too) and picked out the exposure. All subsequent exposures worked fine.

I even did a vacuum plate for the exposures. I put a plexi sheet over the plywood exposure surface, grooved the surface and put some plastic foam in the groove. Laying down the unexposed PCB, film, then plexi and turning on the vacuum pump (actually, the compressor from a discarded window AC unit) held the film and PCB in alignment.

You want to get enough LEDS far enough away and in a regular array so that there is no significant variation on the illuminance. Varies per LED. No magic numbers. Details matter.

[Keppy]

There's quite a bit of tolerance. I expose my boards under a single 60W incandescent bulb for 10 minutes at a distance of about 3-4 inches. It's never failed me for the boards from Small Bear. I had some other boards as well (from Jameco I think) that took a few tries to get right, but I got those to work as well. Not that you should emulate my method by any means, but don't sweat the fine details in construction. Almost anything will work if it's kinda even and you find the right exposure time.

[Perrow]

I've a 1W uv led coming from an eBay vendor. It's got something like 110 or 120 degree "viewing angle" and is as close to optimal wavelength as I could find. I'm hoping it'll work as well or better than the sun lamp I'm currently using

[davent]

Up between the joists above my workbench is a dual 4' fluorescent fixture with "regular" bulbs. I attached a small clear piece of plexiglas to the joists below the fixture to give me a small shelf to hold the exposure packet. Packet sits about an inch below the bulbs. Never had any issues with regular fluorescents and positive resist pcb boards. 


dave

[electrosonic]

I've a 1W uv led coming from an eBay vendor. It's got something like 110 or 120 degree "viewing angle" and is as close to optimal wavelength as I could find. I'm hoping it'll work as well or better than the sun lamp I'm currently using

That's one thing I was considering,  I would think a point source (i.e. single LED)  would give a sharper exposure.

Andrew.

[Mike Burgundy]

That's one thing I was considering,  I would think a point source (i.e. single LED)  would give a sharper exposure.

That is true for shadow puppet style exposure, but introduces problems. An LED is not a perfect point source (nothing really is) - the light intensity varies with viewing angle. Most high-yield LEDs have a fairly narrow viewing angle. The highest intensity is dead-center, (light perpendicular to the board you're exposing), to either side it gets less and less (Edit: often rather quickly - the output graph looks like a bell, often like a bell *stretched so much it looks more like a rounded-off spike*). This gets compounded by the fact that the light hits the board at an angle - even a perfect point-source (radiating the same energy in *all* directions) would illuminate the edges of the board less than the center due to angle of incidence. The net result is the possibility of an over-exposed center WITH under-exposed edges of the board.
There will also be distortions of the design due to the angle of light.
Putting the mask as close to the photo-resist as you can (ie touching) will eliminate the need for sharply defined light rays. This will work just fine unless you want to etch stuff so small normal human beings cant solder to it by hand.
Making the light diffuse (more sources, reflective surfaces around the sources) eliminates differences in exposure, a correct mask takes care of the rest.
My fave idea is butchering an old scanner, fitting it with several UV incandescents (stolen from a facial tanning machine) and a nice reflective (silver/Alu-foil) surface, and using the existing glass plate as your work surface. Place mask and board face-down, close lid, weigh down with a book to make sure the mask really clings to the board, and fire up. Takes a little bit of experimentation to find out the perfect exposure time for your setup, once you have it you're good.

[davent]

How sharp do you need? This was exposed about an inch from a regular flourescent for ~9min, nothing very exact but the resolution/sharpness is pretty decent.



A purpose built exposure unit may be nice but for me it's just something else i'd have to find a place to store and that's where it would be most of the time.


Take care,
dave

[electrosonic]

I have previously exposed a PCB with an ikea stainless steel dome lamp with a 13 W compact fluorescent bulb. It worked but the traces were thinner then I wanted and the text I put on the board disappeared. I was thinking that the diffusion of the light was causing the problem, but maybe I just overexposed the board. Maybe I just need to calibrate my exposure time.

Andrew.

[ORK]

Sounds like your layout wasn`t lying flat on the copper surface. (Stray light underneath the layout film).

[davent]

Hi Andrew, A few things to try.

Make sure the printed side of your transparency is next to the board when you expose it.

Make sure there's excellent contact between the transparency and board. I use a small frameless picture holder (Ikea) to create a sandwich with the glass front/transparency/pre-sensitized board/backer board. The holder has a clip on each side to squeeze it all together, super easy to slip on and off.

Do a test exposure with a transparency and a strip of board. From past experience my exposure time is usually around 9 minutes so if i'm using a new brand of board i'll expose a small strip of board to a range of exposure times to find the correct time.

In very subdued  light (don't need total darkness or a safelight as you would with photo film and paper) make a packet with your narrow strip and transparency put it under your light but cover all but about half an inch of the board with a piece of card stock, turn on your light. After two minutes move the card so another half inch is exposed, another two minute expose another half inch, do the exposing so you have six sections with different cumulations of exposure times, last exposure time is four minutes.  (Doesn't have to be half inch wide sections, use whatever works best for you.) Develop board and you see the results for exposures of 4, 6, 8, 10, 12, & 14 minutes, pick the shortest time that results in the clearing of the resist from the board. You also want to go with the shortest time in the developer to get the clear result. Developing can go pretty quick so you need to hover over this to pull and rinse the board once it's fully developed.Your light and brand of board are going to vary from mine so you may need to have an expanded range of test exposure times to find the time best for your setup.

I expose with the packet about an inch from the light. Used to use the undercabinet flourescent light's in the kitchen, big stack of phone books and fat electronic catalogues to get the packet up to the lights.

dave