6. Electronics design¶
I know very very little about electronics. Which is weird, considering the world I live in. This week therefore was straight uphill for me. I got an overview of the field by Neil’s talk and the excellent lecture Erwin Kooi gave us - overview also meaning a clearer picture of actually how much I do not know.
General concepts like ‘ground’ and ‘tx’ and ‘rx’ or ‘I’: I have to look them all up.
So I bought this yesterday, and started reading - in the hope I will grasp this week’s assignment on a level that is not merely ‘going through the motions’.
After a very long day I know a bit more about transistors, microprocessors, power, electrons and what (Watt) not. And also some more on microprocessors and pins and… Actively using this knowledge is still… but at least I am not scared of it anymore - which I think I was. I clearly remember my despair in highschool when it came to math, physics and the lot.
Most of my time went into making Kicad obey my orders. I think I designed an echo hello-world board, based on the specifications Henk gave us, and following, amongst others, last year’s tutorial of Aalto Fablab and the blog of Nicole Bakker.
Some keywords, I will elaborate on those and illustrate them later on:
-the tutorial was in an earlier version of Kicad so I spent a long time finding the right buttons;
-part of those buttons I could not see anyway because of the Zoom configuration;
I made a schematic combining the instructions from Aalto and Nicole, ending up with this one: I like the esthetics and lettering a lot.
-I lost my Fab library (and found it again thank you Jonathan);
-A lot of oldschool puzzleplay went into laying my traces - I painted myself in the corner quite a few times. I’m not sure if that is because I do not know the right tricks, or if it is a matter of planning and insight. (I did not find any subreddits on the challenge of tracing);
_ I had to face KiCads Design Rule Check telling me there were a few parts unconnected, even though I cannot see how (because they are it seems). I include the picture of the error, to which I have no solution yet. But it is where I stop for today.
Today I thought I managed to get a design where everything is connected the way it should be. In the morning I found out that I wasn’t there yet - before milling the thing I asked Henk to check, and several incongruities in the design showed up. One connection to the VCC I missed (the dead pin could not be a dead pin), and this mixed up the whole routing. I have no pictures of that I’m afraid. With the settings I used in Kicad there seemed to be no solution - so I was happy Henk found a way by changing the copper ‘fill’ margins - a more delicate design was the result, finally functional it seems.
When by then the machine became available, I could mill my board. Many things went wrong there. Board that was left proved to be just a little bit too small; new board was fine, but having stopped the millingprocess because of the above, both the Roland and mods had to be restarted. In the right order. Which took several attempts. Then, operating mods proved to be more complicated then I remembered - twice I milled the negative of what I needed. In the meantime the copperplate wasn’t tight enough, the drill wore down, and I had a strange framing line around my board that was not in my final design. Also, I felt pressured because there were more people wanting to use the machine, so when I finally managed to mill a seemingly correct board but the outline wasn’t correct, I just took the thing off the bed and decided to solve it somehow afterwards.
At that point I spent basically my whole day on this, no more time for soldering let alone testing. But Henk was reading us poetry in the meantime so I did not end up totally frustrated. Thank you Ben for patiently answering all my questions, either through your documentation or live.
To be honest, I circled around all the groups taking measurements, but I cannot say I really got what each and everyone was doing. I took a few pictures, but more with an eye for visual esthetics then for actual data. Boris Karloff is in one of them.
I found parts of this ‘hard science’ difficult to follow. Therefore I caught up a bit only later on, trying to debug my boards.