Note: Don’t look directly into the UV LEDs. It harms your eyes.
This project is featured at electronics-lab.com. There is also a part two, with a timer, check it out here.
I have been using the toner transfer method for about 9 years with great results. Occasionally I would need a board with finer traces and I would use UV method. My exposure setup allowed only a single sided board to be made and it was composed of a 30cm UV tube, some holders, a table, board and film. Exposure time was rather long as the tube was far away from the board.
A few months ago I was looking for some LEDs at one of my suppliers and found that they also had UV LEDs. My mind went directly to upgrading my not so great setup, so I began looking on the internet for info. It turns out some other people have done it, and this instructable confirmed it was really possible. A LED UV box has a clear advantage over a tube one: it can be made in any size, depending on the needs.
I bought 25 LEDs and began to experiment. This way I was able to find a compromise between LED density (spacing), distance between LEDs and board and cost. I wanted something compact, I don’t make large PCBs and those 30cm tubes were not right for this job.
The LEDs are OSSV53E1A from OptoSupply and they have an 140° angle. This means that the first plane with uniform density can be obtained at a shorter distance from the LED, for a given spacing between LEDs which means smaller height for the board. Their peak wavelength is 405nm, which is higher than recommended by the boards’ manufacturer. Results showed they emit sufficient UV light in the responsive spectrum of the photo resist to allow for a perfect exposure.
I wanted to use a common voltage so 3 LEDs in series with a 91Ω resistor are connected to 12V to ensure a 20mA current.
The final setup will contain a LM317 for stabilizing as I don’t plan on using my variable regulated supply all the time. There is another reason for this, due to the non linear nature of the LEDs, varying 12V with +/-5% creates much larger current variations which influences the required time. Putting a regulator on the box simply allows more flexibility in choosing the power supply.
On the bottom board I have added red LEDs. These LEDs help me align the top and bottom films. I have used low brightness, cheap, matte LEDs as I have to look into them when aligning the films. Connection is similar, 3 LEDs in series with a 220Ω resistor, except for the last row which contains 2 LEDs and a 330Ω resistor. These are not really necessary, aligning the films can be done in a dim ambient light.
For safety reasons and extra functionality I added a switch to turn off the UV light when the box opens, turn on the red light and stop the timer. This makes the box more high tech and gives it a professional behavior.
I used nuts glued on the interior walls of the box to hold the bottom 4 mm thick plexiglas panel. The choice was simply of what was easily available: nuts. Don’t use normal glass as it blocks ultraviolet light. There is no top panel mounted, because it is part of the aligning procedure: I put the first film on the bottom, then the board and the second film is taped to a smaller piece of plexiglas. Aligning means aligning the smaller piece of the plexiglas so that the two layers overlap perfectly. Having the film attached to it makes it easier. Extra markings outside the board help this (ghost vias). Part two should contain a more detailed tutorial on making double sided boards.
The final assembly allows an area of about 10x15cm of board size, with the interior dimensions of the box being 18cm long, 11cm wide and 5.5cm deep for each of the two halves. Exterior dimensions will depend on the thickness of the material used. Each board contains 7×12 UV LEDs, with 3 and respectively 4 holes between them( the actual spacing between the center of the LEDs is the same). The distance between the top of the LEDs and the board is 3.5cm, including the 4mm plexiglas panel. Check out the detail below to see the LED arrangement on a normal 0.1″ or 1.27 spaced board:
I tested with a small board from 2 minutes up to 8 minutes. Starting with the minimum of 2 minutes proved not to be the best idea, but my long exposure time old setup made me choose this value.
What you can see in the picture above, from 2 to 7 minutes exposure if perfect. At 7 minutes some over exposure starts to show up, only in large copper areas, due to the fact that the film was printed with a laser printer. This result is very different from what I have obtained when using a fluorescent tube: below 12 minutes it was clearly underexposed and above 14 it was clearly overexposed, leaving only a small interval where it was right.
So far I have made 4 boards with 2 minute exposure and they came out perfect. When free time will be available I will investigate what the results are for less than 2 minute exposure. As any time between 2 and 7 minutes seems to work, there is no immediate need for a timer. Still, part two should contain a basic circuit for this.
Check out the gallery below for extra details:
The finished panels:
Bottom plexiglas panel installed and a nut:
The safety switch:
UV on(switch is not used):
Building your own
In order to build your own you first must decide what the required size is for you. With my setup, boards up to 10 x 15 cm may be exposed. For this, you will need the following materials:
- 168 UV LEDs, OSSV53E1A;
- 56 91 ohm 1% 0.25W resistors(5% is also ok);
- 66 red diffused LEDs;
- 18 220 ohm resistors for red LEDs;
- 6 330 ohm resistors for red LEDs;
- 2 prototype boards, 13 x 25 was the smallest size I found that fitted;
- 1 switch for the lid capable of 1.5A minimum;
- 2 hinges;
- 1 box;
- misc: screws, nuts, glue etc.
The box may be made of your choice of material. I’ve built it out of two identical halves, interior dimensions 18 x 11 x 5.5cm each. The material is 8mm thick OSB, for which you will require the following parts:
- 4 11 x 5.5cm pieces;
- 4 19.6 x 5.5cm pieces;
- 2 19.6×12.6cm pieces.