Electronics Production – Milling the FabISP

Step 1 - Milling the Board

While learning to use the milling machine in the shop I was thankful for the clear directions provided by Rob and the patience of my peers. The process began by ensuring the mill machine and the computer were speaking to each other. Once the connection was established I cleaned the surface of the mill machine with a little alcohol and vacuumed up any dust from the last board. Next I placed down some double-sided tape on the surface - making sure not to overlap them I firmly placed the copper sheet in place, ensuring no corners were up and that the board was secure. Using the fabmodules I uploaded the image of the traces, selected the Roland Mill as the output, and 1/64 as the process. Before I oriented the machine I gently placed the drill head in the shaft. I then used the x and y coordinates to get the drill to the right location on the board I found one of the most important steps in milling is to correcly zero the z of the drill so that it is firmly in contact with the board. Initally, my traces didn't come out so well because of this mistake. Lastly, once the traces are done and dust cleared, it was time for the outline. It was important to change out the drill head as well as change the process to 1/32 on the fabmodule.

Step 2 - Deburring + Prepping

Once the board was milled for the traces and interior it was time to get it cleaned up and ready for the next phase. To deburr the board I took some alcohol to it was a paper towel. I saw others use sandpaper, toothbrushes, soap and water in their deburring process. One of the most important steps in the whole process for me turned out to be how I prepped myself for the soldering. Initially, I jumped right in. I practiced on board a bit and instantly got frustrated with how sloppy I seemed with the whole thing. I also avoided the microscopes at first, but soon found my eyes are too weak to complete the task without assistance. Another key learning was preparing the components. Rob graciously printed out a bunch of copies of the board. I placed a piece of double stick tape on this sheet. On top of the tape I laid out the components and labelled them on the side. This helped me stay organized as I was soldering. Ultimately, the best positioning for me when soldering was with a stool so I could sit higher up, a microscope I could adjust, a lamp to light the area, and the soldering iron on my right handside.

Step 3 - Stuffing the Board

There are few things in recent memory that have challenged my grit more than my first day of soldering. I was easily frustrated by how unsteady my heads felt in figuring out the right balance of heat, solder, and positioning on the pad. Everyone has their tips, but I quickly learned a few key tricks for my process: 1) get comfortable - this includes singing 90's hits to yourself to calm your nerves 2) cut a 3 inch piece of solder instead of keeping it attached to the spool 3)make sure the iron tip is shiny 4)heat the pad liberally (more on this later) 5)the microscope is your friend. I also found it helpful to start with a big piece like the mini-USB and then work from the inside to the outside of the board. The copper braid came in handy once I realized how applying some solder on the braid once it was hot and in the area helped pull the solder away from the desired location.

Step 4 - Programming the Board

Soldering was so challenging for me that I was thankful I was done and on to the programming phase. For some reason I thought this step would be really easy...it was not. First, the smoke test. Within seconds my little PCB started to smoke. I looked at the online tutorial for tips and saw the tip to inspect the tiny pins of the tiny USB. Sure enough mine where attached with solder. I used the braid to seperate the solder and on the next try it passed the smoke test. Next programming...this took a significantly long time to troubleshoot. Initially, the computer was not recognizing the board. We inspected every component and found two big issues: 1) One of the diodes seemed to have blown (perhaps during the smoke test) 2) Many of the traces I thought I had beautifully soldered were not actually soldered properly. Nils and Rob helped me see how some of the feet of my components were not interacting with the copper. I had to go back through and resolder a tone of the pads. What I realized I had done before was not heat up the pad enough. I think I was heating up the component and touching the solder to the component, and not getting enough connection on the pad. When I went back I let the iron heat up the pad for several seconds before I added more solder. We were able to program the board, but then we were having trouble with the computer recognizing my board as a tiny USB. We started to try and isolate the problem by checking values on various parts of the board. Ultimately we found that one of the resistors near the tiny USB was the issue. This was very interesting, because to the naked eye and underneath a microscope none of us could tell the component was not soldered correctly. Once I resoldered to make sure all values were working the way they were suppose to, the computer was able to recognize the tiny USB.

Step 5 - Desoldering the Board

I used braid method as my main method of desoldering. Again, I found it to be very useful to apply some solder to the top of the braid once it was heated. I did not tried to use the flux once, but did not find it to be worth the mess it added to the process. When removing SJ2 on the board I used the heat gun to help loosen the solder. After about 30 seconds of sustain heat I was able to have the board fall away from the component. I desoldered SJ1 and SJ2 on the PCB.

Step 6 - Making IDC ISP Cable

When making the IDC ISP cable we hit a bit of a bump. The rainbow colored cable we had originally tried was not fitting into the clamps as well. Upon further inspection we (Rob, Angie, Brendon, and I) noticed there was better cable available. This process simply entailed selected 6 connected cables and then cutting a piece of the six-packed cable that is about 10 inches in length. Rob suggested securing the clamps using the vice. Once the cable was put together I tested it on the computer in the shop.