Skip to content

Electronics production

For this week we had to produce a PCB, program it and test it. I plan to mill a FTDI-UPDI programmer

  • Linked to the group assignement
  • Documented how you made (mill, stuff, solder) the board
  • Explained any problems and how you fixed them
  • Included a ‘heroshot’ of your board

To milled my PCB I use our newest low budget milling machine from Sainsmart.

I began by assembling the machine … The machine is pretty straithforward, it is composed of 3 axis without any endstops.

It took me almost 1 hour to assemble everything. I quickly checked if everything work correctly by powering the main baord. You can control the machine using a tiny controller (at your own risk) or with a computer.

After reading the manual I move the 3 axis and observed a strange behavior… The X axis was inverted…

This may happend when a machine firmware isn’t well configured, but this machine was shipped with the firmware already uploaded in the board. As every board from this vendor should receive tghe same firmware, the problem comes from elsewhere…

Another way to fix that is to switch the stepper phase. After swapping the black wire and the red one as well as the red wire and the black one. I plugged bacj the connector and the X axis move in the correct direction !

I then use my computer to control the machine using Candle, a GRBL controller. Candle is very convenient. I tried to “air” mill several demonstrating programm suscessfully.

I decided to use V bit to begin carving since they are pretty cheap… I not very confrotable breaking the straigth ones…

I use the two wrench to tighten everything correctly and the bit seems to be fixed in place. I started the spindle and except the heavy vibration everything was OK. We used doudle sided duct tape to set the plate in place.

This machine as no endswitch to automatycally set the zero position. It has to be set mannualy. To set the zero coordinate, you can use Candle to move to the desired position and press XY zero the set the zero. To set the Z zero, the process is a bit more complex. When I was in intership at the waag Amsterdam, Henk explained the zeroing method on the Rolland milling machine. He was moving the bit just above the copper plate manually and then untighten the bit to let it fall gently on the copper. This technique was very useful. But on this machine this is impossible because the bit move upward when you tighten it…

I tried instead using a piece of paper and manulally approached the bit to the sheet of paper until I cannot move the paper anymore. This was a good technique but I figure that the paper was 0.1mm thick so I was 0.1mm above the plate. Moving downward by 0.1mm solve the major issue :)

I tried to mill the USB-UPDI program from another student and got this result :

Obviously the plate wasn’t horizontal and need to be leveled. We leveled the plate manualy using a tiny piece of paper to raise by 0.4mm on the other side of the plate.

I decided to mill my own file after this failure to be sure understanding every step. I began by setting up the correct parameters on Fab modules using the online program Mill 2D PCB PNG I used the default feed and speed value since we were using those before. Unfortunately fabmodules refuse to read my PNG file… The vectorization isn’t working every time. I find a way to fix it by running mods locally… But I also tried to vectorize the file by myself using inkskape. Then I use the Mill 2D PCB SVG to prepare my programm.

I first used the linetest file to charaterize the machine. I got this result using a basic V mill.

Engraving the traces is OK with V mill but cutting trough the plate is difficult. The mill witdh grow as you go deeper into the material and I forgot to set an offset in mods. In other words my linetest is cropped…

I tried to mill the USB - UPDI programmer, I used a V mill to cut the traces and then a V mill to cut trough the material (unsing this time the right mill diameter) I measure the V mill diameter at 1mm from the tip to set the cutting diameter on Mods I got that result

The traces in the center are really thin and the V mill was sheared a little. As a result The thin traces were removed. I then try to use a 1/64 flat mill and got that result :

Unfortunatly we haven’t any FT232 chip left… So I mill this circuit for … Fun !

So I tried to mill a FTDI - > UPDI programmer instead.

I use this hello.USB-UPDI.FT230X file :

I repeated the same process and got this board :

I used a soldering station to tin every smd pad and then solder the resistor and the pin header.

Here is the final result !


Last update: September 23, 2021