4. Electronics Production

Group Assignment

Characterize PCB production rules

The goal of this group assignment is for us to know the potential gaps between our traces, or how thin the traces could be. Since the machines were always occupied and the booking system at Fab Lab Barcelona was not really effective, our class couldn't have a chance to do many group tests. In the end, we were able to finish only 1 group test with the participation of the whole class.

The machine

We used the Roland MonoFab SRM-20 machine to mill this test file provided by the instructors. Detailed specs of the machine:

  • Work area: 203 x 152 x 60mm
  • Loadable workpiece weight: 2kg
  • Operating speed: 6mm/min - 1,800mm/min
  • Spindle speed: 3,000 – 7,000rpm
  • Input Format : RML-1
  • Material: Modeling Wax, Chemical Wood, Foam, Acrylic, PCB
Roland SRM-20

Preparing the file

We prepared the .rml files using Fab Modules. The workflow of Fab Modules is pretty straightforward, and here are some recaps of steps followed:

  • Select .png input format and load the .png file
  • Select .rml output format
  • Select the proper process: PCB traces (1/64) or PCB outline (1/32). This will automatically define the proper cut depth (0.1mm for milling traces and 0.6mm for cutting).
  • Select SRM-20 in the output machine. Modify settings to origin 0,0,0 (x,y,z); zjog = 12 (to make sure the milling bit will be lifted up while moving across the workpiece and avoid damaging both the traces and the fragile bit itself); and home 0,0,12 (x,y,z).
  • Select the proper direction. If there are thin traces on the board, we need to select the conventional direction in order to avoid broken traces.
  • Click the calculate button to calculate the toolpath and click the save button to save the .rml file.
group test file

Milling

We used Roland VPanel controller to adjust the milling start point, the feed rate, and the spindle speed. Detailed summary of how to use the milling machine and VPanel:

  • Choose the correct milling bit for the job. We were provided with 2 bits: the 1/64" used for milling the traces on the board and the 1/32" used for drilling holes and cutting.
  • Insert the proper milling bit into the machine and manually adjust the Z origin.
  • Set origin X/Y and Z in Vpanel. Click the Spindle button to test if the milling bit can cut through the copper layer.
  • Click the Cut button, then add the .rml file and start milling by clicking the Output button!
group test file

Result

We had an issue with the outlines cutting. One side was not cut completely. We managed to generate the file multiple times and did the process again, but we still had the same problem. However, we could tell from the failed test that using the default settings of Fab Modules, the machine is able to mill up to 0.010mm thin trace.

group test result

Personal assignment

Producing an In-Circuit Programmer

So for my personal assignment I decided to produce the FTDI programmer that was available for the Series 0,1 chips.

Preparing the file

I actually used the Mods page for producing the files so the process went a bit different than the group assignment. Here's the steps I took to produce the .rml files for milling.

  • Go into the Mods page ( http://mods.cba.mit.edu/ )
  • Right-click and select Programs > Open Server Program > Roland > mill > SRM-20 > PCB png
  • change the last module in the file ( WebSocket device ) to use the save file module
  • Load the .png file into the read pcb module
  • select either mill outline or mill traces in the set PCB defaults module
  • click the calculate button in the mill raster 2D module.
mods - PCB millingmods - PCB milling
mods - PCB millingmods - PCB milling

Milling

With this process I got 2 .rml files which I moved into the milling machine and started milling.

To start milling I had to use 2 different drillbits and change them between files, I used the 1/64 bit for the traces, and the 1/32 bit for the outline.

With the milling I ran into the issue of choosing the wrong pcb file so I ended up milling a wrong circuit, so I had to run the milling again with the correct circuit, but once I did that the pcb got loose during milling ( probably I didn't press enough the PCB into the double sided tape ) and it got destroyed, so I had to run the mill yet one more time to get it done correctly.

PCB millingPCB millingPCB milling

Soldering

Once I had the finished PCB I had to make sure to take out some extra traces that were left behind, and with that I was ready to start soldering.

Soldering ComponentsSoldering ComponentsSoldering Components

Testing

I used a USB extender to plug the FTDI programmer to my PC and I used both dmesg and lsusb to check that it was recognized correctly.

testing