Schematic/Board:
The schematic was designed based on recommended layouts, and previous experiences with PIC microcontrollers. Using the Autorouting function of Eagle, and many hours of routing and re-routing, the first revision of the EMG Data Transciever has been built. For getting the board manufactured I decided to go with BatchPcb as the service seemed like a really good deal for my own printed circuits. The main reason that it took me so long to design the board, besides the fact that I was getting accustomed to using Eagle, was that I had to create my own package for one of the components.
A crystal whose package was not included in the Eagle library needed to have one created for it. In the process, I learned how to design my own packages and libraries. The other main problem was that in creating the pads(the parts that connect the SMD component to the board) I added 3% roundness to them. This was so little that it could not be noticed visually, however the BatchPCB's DRC Bot kept giving me errors. It took me many days to resolve that problem. As this is the first PCB I have ever designed, I believe that it came out pretty well.
Prototype:
Upon arriving back to campus after the winter break, I found that the components were waiting for me to be put together. As I recently got the AOYUE 968 Hot-Air Rework/ Soldering station, I believed that soldering the components would be easy, but I was very wrong. Rather than making it easy, It made it possible, but still far from simple. This board forced me to re-evaluate the techniques and tools that I had. I have had Copper Braid laying in my toolbox for years now and I have never even touched it until now. For soldering the QFN parts that I have never had the opportunity to use, It proved to be essential. 2-pin SMD components are exponentially more troublesome than their DIP counterparts. As the 0603's are so small that they stick to the soldering iron, I found that the best technique was to apply solder to one of the pads, then hold down the component with tweezers as you solder the one side into the pad. At this point, soldering the other side is easy. With using the Hot-Air gun, I found that it is useless for 2 or 4 pin components. The air just blows away the component, and it can be lost forever. The air gun greatly helps though, with large components with many pins, as it can apply the heat much more evenly than a soldering iron.
In the end I got the board built only to realize that the PIC32 needs a 10uf capacitor on one of the pins to function. Luckily, I was able to scavenge two 5uf capacitors from scrap electronics, and with the loss of one of the 8 I/O ports, I was able to get the board working and recognized by the programmer. Time to celebrate, with sleep.
PS: I am currently working on Rev2 of the board which will be smaller, and run on li-ion batteries
