Week 05

Electronics Production

Assembling Clank

This semester, due to COVID restrictions on shop access, CBA has equipped us with an at-home-mini-fablab including a kit for a desktop CNC to mill PCBs. The machine is affectionately known as Clank. Last weekend I followed Jake's instructions to assemble Clank. Fun to put together!

Responsive image
Begining to assemble Clank.

Responsive image
Responsive image
Many faces of Clank.

Responsive image
Assembled Clank.

Once all the hardware was assembled I got the controller up and running and did some jogging.

Clank lives!

Calibrating and Characterizing Clank

Next it was time to start running some gcode on Clank. When the controller is loaded, the default gcode that loads along with it is the toolpath to flatten the bed of the machine. This was done with a 1/8" end mill. One thing I learned the hard way during this process is that it's important to make sure the computer doesn't fall asleep while you are milling or the connection to the machine will be lost and the position will freeze.

Flattening the bed.

With the bed flattened it was time to characterize the cutting quality of the machine. I did this using a file that tests the internal and external cutting of the machine. The gcode was generated in mods. The first toolpath engraves the surface using a 1/64" end mill and the second toolpath cuts the part out of the stock using a 1/32" end mill.

Responsive image
Design used to test machine.

The first time I ran this file, the stock was not flat on the bed because it didn't quite fit in the pocket that was made when flattening the bed. As a result, you can see that on one half of the job the bit didn't break through the copper whereas on the other side it did.

Responsive image
The result of unlevel stock.

On the next try I had more success, but as you can see the cutting is quite fuzzy and the traces on the smaller end of the spectrum did not hold up very well.

Responsive image
Fuzzy milling.

In one of the tutorial videos, Jake mentions that he was having success using a v-groove engraving bit as an alternative to the 1/64" end mills. I gave this a shot and the results were quite clean.

Responsive image
Clean milling.

Power Supply Cover

This week I also designed a cover for the power supply where the 110v is exposed. An .stl file for this is linked at the bottom of the page. It's a bit bulky but it clips onto the existing prints and power supply without any hardware or the need to disassemble anything.

Responsive image
Power supply cover.

Milling and Stuffing PCBs

With the machine calibrated,it was time to take a crack at milling PCBs. The assignment this week was to make an in-circuit programmer. There are two categories of boards we could make: one for programming AVR microcontrollers and one for programming ARM microcontrollers. To be honest, I'm still wrapping my head around the differences between these options but this week I was mostly focused on fabrication. Hopefully the higher level understanding of the electronics will come in the next few weeks. In the end, I made one of each with the ARM board ultimately serving as a testing ground for figuring out soldering.

With the significant caveat that we are not yet designing our own boards, here is the basic workflow for making the programmers:

  1. Start with existing PNG of design.
  2. Use mods to generate g-code for engraving and cutout toolpaths.
  3. Run g-code on Clank.
  4. Stuff board with components.

Responsive image
Design for ARM board.

Responsive image
Milled ARM board.

Responsive image
Stuffed ARM board. Oof.

Soldering was definitely a challenge. I've soldered in the past but mostly for making sculptures. The surface mounted components are quite small and my usual techniques were not going to cut it. I abused the first board I tried to solder pretty badly mostly by overusing the heat gun but figured out a process that works for me along the way.

What I find works well is to start by putting a dab of solder on all of the pads for the component I want to attach. Then I take the heat gun in one hand and melt the solder as I place the component with the other. To attach the microprocessor or microcontroller I first get some solder on all of the traces and then wick away the excess leaving really just a film of solder on the traces. Then again I take the heat gun in one hand and melt the solder as I place the component with the other. From what I understand the process described above is essentially reflow soldering just without the solder paste.

Once I had this process worked out I stuffed the AVR programming board with FTDI chip and had much better results.

Responsive image
Design for AVR programming board.

Responsive image
Milled AVR programming board.

Responsive image
Stuffed AVR programming board.

File: 201009_power_supply_cover.stl.3dm