week6

Week 6: Electronics Design!

Group Assignment: Use test equipment to observe the operation a microcontroller circuit board.

Individual Assignment 1: Design and Make a helloworld circuit board

I broke bits, tore up copper, and otherwise wrestled with Clank into this weekend, but once I thought it was in a usable state, and I had a good handle on the Clank workflow, it was time to spin myself up on the next assignment.

Eagle Round 1: ATTiny44 board

I rewatched the class and recitation videos to figure out where to start. It seems like both KiCad and Eagle have their advantages and drawbacks. I decided that reviewing prior years' cohorts and their experiences this week would give me a good sense of which would be better for me. Given my lack of electronics experience, I looked through the arch sections from past years to find my people. I found the documentation from Marianna and Rafa Rafa to be pretty helpful. After canvasing these resources, I downloaded both softwares and poked around in them to get acquainted. I decided to go with Eagle for a couple reasons: 1. I like that it can integrate with Fusion360. 2. For some reason I found the process of adding libraries to Eagle a little bit easier. So, I chose Eagle. Now for the fun part: choosing what to make.

Given how out to sea I felt while trying to learn anything about electronics, let alone a new CAD software, I thought it would be helpful to follow the instructions of those who came before me. The ATTiny44/45 board seemed to be a popular choice in the past years, so I decided to use that as my jumping off point. You can see the board, the chip spec, the schematics and the board layout (without traces) below.

attiny44 — I built the schematic and laid it out pretty much exactly as I saw it, mostly just to get comfortable with using Eagle. I was totally uncomfortable with starting from scratch wtihout a template, and even in following a pretty complete model, I wanted to get a more practiced pair of eyes on my final product before I spent time running traces, exporting and milling something that wouldn't work. So I reached out to anthony so he could check my work. Glad I did. In addition to checking my schematic and board, he explained to me that the ATTiny44 was an older component with less storage space, and it would require that I make some extra processors to make it work. So I scrapped that design and started over with the ATTiny1614 that Anthony ran through in his tutorial.

Eagle Round 2: ATTiny1614

Ok. First things first. The Fab Eagle library does not have the ATTiny1614 footprint. Neither does AdaFruit or Sparkfun. You can find and download it here Everything else I needed seemed to be in the fab library. Now, the data sheet:

My reference board for this project is Neil's ATTiny1614 Helloworld board. See below:

But.... I needed to add a Button and an LED. So we had a little work to do in Eagle to set that up. Luckily, Anthony's tutorial was a good guiding hand.

Here is the list of components I had in my schematic:

1. ATTiny1614-SSNR

2. Conn: FTDI-SMD-Header

3. Conn: 2x2 Pinhead SMD

4. R1 and R2: 1k resistors

5. LEDFAB1206

6. C1: C1206FAB

7. S1: SW262CT-ND

Once I was in Eagle, I pulled the above components from the fab (and other) libraries, and started pushing things around and routing connections. I followed Anthony's tutorial in real time for this schematic and generated a board to start running traces.

To start drawing my traces, I first set my Design Rules to the 16mil tolerance Anthony recommended in his video. I also set my wire width to 16 to beef them up a bit. I then spent a while trying to route traces based on this schematic, but I was unable to make all my connections work. Someone more practiced might have found a successful way to route the traces in that set up, but I decided to make changes to the schematic instead. The resulting setup allowed me to make all my necessary connections. See the schematic and board below:

Time to export. I turned off all layers but the "top," which held the traces. I exported an png image because I planned to use mods to generate my gcode. The exported the traces png at 1000dpi, which looks good, understanding that when I import it to mods, I should double check my dimensions and resolutions, and will likely need to adjust my file to 2000dpi. The png looks like it should though:

I hit a speedbump at this point, though, because I couldn't figure out how to export the interior file for cutting the border that we see in Neil's helloworld boards. I played around a little, but couldn't figure it out, and in the interest of keeping the process moving forward, I created a border in illustrator, making sure to maintain file dimensions. I did have an issue once I imported it into mods. The 2x adjustment to the dpi no longer worked once I ran the files through Illustrator. I may have saved something wrong. This wasn't a huge problem, though. I was able to manually adjust the dpi on both the traces and interiors in mods to ensure that their dimensions corresponded to each other.

Eagle >> Illustrator >> Mods >> Clank

The architecture section has been sharing successes and failures in all things mods and clank, and through some very frustrating moments and a lot of lost bits, I think we've landed on some pretty hard won knowledge about the best mods settings for our clank to run smoothly. The biggest takeaways: slow down the cut rate, increase the number of passes and up the spindle speed. The below settings worked pretty well, but I think we can still do better:

Gcode generated for traces and interiors. Now to wake up Clank again.

It could have gone worse, but unfortunately on my first pass I let my computer fall asleep once even though I told it not to, and that stopped my Clank mid-cut. On the second run, the piece of stock I used was compressed at the edge, so my cuts didn't go all the way through. Oh the joys of Clank. Anyway, I let that one finish up so I could check that my manual adjustments in mods still allowed my components to fit on board footprint. It seems like they do, as you can see above.

Finally, in my third attempt, I was able to mill a board all the way through. I also increased offsets form 4 to 5 to give myself a little more room. I think I could stand to increase that even more now that I have the updated firmware in the z-motor and running a longer cuttime is won't threaten to melt my machine. But I'll save that for next time. I just did a little manual clean up on this one that took about 5 minutes.

With a board cut, it was time to stuff it. You can see my handiwork below. I really want to get better at this soldering business. I need to get more comfortable with the removal process. I think that will help

Anyway, the components are on there. I haven't had time to test it yet, but that is next on my agenda. I am now comfortable with generating a board in Eagle and bring that through to Clank and the post-milling process. What I still need to work on is the actual schematic building phase. I need to become more literate with reading and understanding the data sheets and overall just gain a better grasp of what these boards are doing.

Files: