RVAsec Badge Build 2014
This Project page is still a work in progress! Timeline for the 2015 Badge Project
Since the start of the RVAsec conference here in Richmond, HackRVA has been building the attendee hardware badges. The project has been a great opportunity for members to work together on a long-term computer hardware fabrication. The goal of the badge is to provide each attendee with an open and completely "hackable" piece of hardware, loaded with entertaining games and puzzles that encourage attendee-attendee interaction.
The current firmware, including links to datasheets, can be found on our GitHub Page
The past two years (2014 and 2013), HackRVA and its members have built the badges from bare copper boards. Members etch, pick-and-place, solder, and QC every badge. The following PCB Creation and Component Stuffing sections detail this process. Relevant QC steps are described within these sections.
There are many PCB manufacturers that offer relatively cheap manufacturing of single-layer boards. Still, drawbacks include the increased cost and the "Christmas surprise": a large number of costly boards appearing on our doorstep with systemic issues. Additionally, the hacker look and the learning experience of creating our boards is generally considered very valuable.
We start with bare copper boards (1/2 ounce), apply a resist compound in the pattern of the desired board trace, and finally we remove the exposed copper (not covered by resist) with ferric chloride.
The following subsections describe this process, including some tips and general do's and don'ts.
We experimented with both the toner transfer and photo-resist film. In the end, the majority of the production boards were fabricated using the toner transfer method due to issues in the photo-resist process. Note that a slightly botched resist application can often be saved with a steady hand and a permanent marker. The marker will resist the etchant fine, so simple draw-in any missing pieces of the resist.
The toner transfer is just that: we print the schematic onto a material that readily transfers the toner to the copper board when heat is applied. We do this by coating high-quality photo paper with dextrin (a starch compound) and then directly printing the design using a *need printer model here*. We've found that applying the dextrin with a high-quality paint brush (fine bristles that, no streaks) is fast and effective. We also used our t-shirt press *at what temperature and timing* to release the toner followed by a soaking in tap water that removes the paper. Generally, high pressure and heat should do the job. When etching, be careful to not let the toner fall-away; it's relatively sensitive and may not last for long in a hot ferric bath.
We bought photo-resist film and paste from several sources, with varying success. The photo-resist intended for screen printing applied well, but didn't stand up to the ferric bath well enough, since this compound is water soluble. A similar paste, but specifically intended for PCB etching, was also tested (link??). It was very resilient to scrubbing, water, and the ferric, though applying it evenly consistently proved very difficult. However, a higher quality brush or roller applicator may work well.
After your resist is in place on the copper board in the pattern of your desired copper trace, you can begin to etch your boards.
Though we did some experimentation with *other strong chemical that burns*, we found Ferric Chloride to work best. It is available at most Radio Shacks and in bulk online. It's best to use it heated up to around 115 degrees Fahrenheit. A higher temperature will etch the copper away faster, but the heat may also erode away your resist faster. So it's a bit of a balancing act and takes some experimentation. Some slight agitation by hand or with a strong bubbler can really help speed the process up. In fact, without it, you may have very little luck getting an even etch. Also note that glass is probably your best option for a container, though plastics that can stand the heat could also work. Of course, do not use metal.
Cutting and Drilling
You should now have at least one board with your desired trace on it. If you need to make multiples, you'll have a larger caopper board with many smaller duplicate traces on it. We used a tile saw to cut these boards out and we didn't have any problems. We also drilled any through-hole points by hand using a drimel and stand. The process of making these holes could be improved using a CNC, though we had issues getting correct and consistent registration. Your mileage may vary.