Prototype number zero mainboard and no magic smoke 2007-06-09

After about two weeks of almost constant work in the evenings and late into the night, I've managed to produce the first prototype mainboard and a number of sensor module prototypes will be completed today. The electronics required aren't particularly advanced but I've had to greatly improve and develop a number of my design and fabrication skills.

The goal of the prototype 0 mainboard is to be able to capture and record the inputs from a number of different sensor modules and store the data for future processing. The main components for the mainboard are a CPU (Analog Devices EVAL-ADUC7020MK Minikit), power supply regulation (LM2576, LM1117), host interaction (MAX232), SD/MMC card adapter, connection points for digital and analog inputs, and a Dual UART (SC16IS752) for interfacing with serial sensor modules. I also added loads of prototype pads which should allow me correct any mistakes or add missing features.

I used the open source tools gschem and pcb to design and layout the schematic and circuit board.

Image 1: Example of CPU schematic from gschem.
Image 2: Screen shot showing the top and bottom layers of the board from pcb.

I took a number of photos of the steps I did when going from the circuit board layout to the finished board. I used the photo-resist process to produced a double sided board, I didn't plate the holes but I soldered pins on both sides of the board to join the layers. Notes (almost tutorial) for each photo are from left to right, top to bottom.

  1. I used a laser printer to print out the appropriate mirrored layers (solder, component) on to clear film (acetate) sheets.
  2. I lined up the holes and created an envelope which will contain the board.
  3. I used pre-coated photo-resist copper clad boards, I cut the board to side and then clear fully remove the film.
  4. I placed the board inside the envelope, and taped the corners which I had cut at an angle to securely film while being exposed.
  5. I have a UV box and placed and exposed the board on each side for 4 minutes.
  6. It is possible to see the slight change in color caused by the exposure.
  7. Clear full pour the developer onto the board in a tray. Note: These chemicals are hazardous and damaging so wear rubber gloves and eye protection, and preform these steps in a shed or work room.
  8. Gently rock the tray, you should see the board appear, turn the board so that both sides are developed.
  9. When the board is developed remove it from the developer and rinse it thoroughly in water. Look for any imperfections, the photo-resist I used is green in color so the imperfections should clearly standout. If there are any area which need a touch up use a suitable (Red Staedtler Lumocolor 313) permanent pen to make the corrections.
  10. I then placed the board inside another clean tray which was inside a larger tray containing hot war, like a bain-marie or double boiler. I then carefully poured in the etchant (Ferric Chloride, FeCL3).
  11. Gently rock the tray, after about five minutes depending on the side of the board temperature, and concentration of the etchent you should start to see the copper dissolve away.
  12. Once the etching has finish rinse the board thoroughly in water.
  13. The photo-resist lay needs to be remove, by exposing the board in the UV box placing back in the developer or rubbing with a polishing block.
  14. Drill the many holes, you'll get the best results using a drill-press or desk-mounted drill. Polish and clean the board before soldering.
  15. I used a combination of solder paste and wire solder depending on the type of component.
  16. The top side of the board.
Image 3: Mosaic of photos showing the steps involved in producing the board. (Click to see larger version of image)

Here is the finished board.

Image 4: Mainboard top.
Image 5: Mainboard bottom.

The design of the board isn't perfect. ;) Two problems I've noticed so far is that I never added any LED indicators to show if the power is working etc., and I never linked the IRQ pin on the Dual UART to the CPU.

I was very cautious when first powering up the board, and used a very basic current limiter circuit, so no magic smoke was released. And when I plugged the serial cable into my computer and reset the CPU I was greeted with the fabulous words 'Hello World!'.