Nathan Dennler: Failure, Recalibration, and Iteration

So last week said I’d try to document more about the actual process of making the clothes, but a few pictures in, I realized that it actually just looks like a big ball o’ fabric up until the very end. It’s really one of those things you have to see in person and interact with to really understand what is going on. To that extent, I’ll save you a lot of scrolling and abridge everything.

So I first started out with the pants this week. Basically the schema for making any garment is similar: you first buy a pattern, you get the parts for the piece that you are trying to make, you cut the fabric into the pieces, and you just follow the directions on the instruction sheet.

I got the pattern for these pants back in week one, because Joanne’s always puts their patterns on sale at the end of a season (!!!). They were around 2 dollars each, whereas they can usually go for around 15-20 dollars during the season if you try to buy them online.

So once you have your pattern, you can look at the pieces in the handy dandy instruction sheet to figure out which pieces you need, which are all conveniently numbered in a sequential order (unlike some of the electronics I have used — more on that later).

Then you lay out your pieces on your fabric, which becomes an interesting packing problem. The instructions suggest a layout, but you can sometimes find better ones if you are so inclined (though the suggested ones are pretty good).

Then you cut the pieces out. My mom always used to tell me that this is the most important step, since you have to match all the edges up, and if you cut things out crooked, your final garment will also be crooked.

Then once you have the pieces all cut out you have to follow the steps of the garment. I think when you first start, this is the most difficult thing to understand since there is a lot of sewing jargon, but then each garment has uniquely named pieces. So you have to learn the terms that relate to general practices (e.g. dart, press open, notches, facing, lining, contrast) versus which parts correspond to the specific clothing item you’re working on (e.g. yoke, side front, collar). You then have the issue of having to follow a 2D projection of a 3D mutable object, which is sometimes difficult. As you practice more, though, I do think it gets a bit easier. Here are some examples of instructions from the pants:

After you follow all the steps (and do a lot of head scratching and spatial processing), you get the final thing!

So following these abstracted steps again, I made the jacket. Unfortunately, the material for this jacket wrinkles like no-one’s business, so every time I store it and take it back out, I have to re-iron it. Hopefully I can find a different place to store it that isn’t the Foisie locker and just keep it on a hanger.

After I finished the main parts of the clothes, I started working on interfacing with the microcontroller. From my experience with robotics, I know that there is often a lot that goes into being able to just interface with the controller you are using. This time around it actually wasn’t too bad–I just had to install a couple of boards, have something fail in my environment, reinstall Arduino IDE and then reinstall the boards. The whole process only took like an hour! I experienced some cognitive dissonance when the successful upload message is printed in a bright red, making it look like something bad happened–but it did work.

Here is the successful upload message, for reference all the header inclusion errors are written in the same color

Then I went to work on the actual controlling of the LEDs. This is when I realize that the Trinket doesn’t have a datasheet? I didn’t think it would be too much of an issue since there are only like 8 pins, but that would be too easy. I was pretty easily able to just do a regular digital write on the pins to blink the LED, but then I wanted to get a more gradual fading, which would require analog writing.

I looked at the product page and found out that there were indeed 2 analog output pins on the board, but there was no information on which two they were. I did some searching and found out that the label for the pins are “A1” and “A2”. I tried using A2, and my foolish self thought that would correspond to pin number 2. I uploaded my code, and nothing happened!! One more thing, since the board is so small there is no serial communications bus to communicate with your computer, so there cannot be any debugging!!!!! WooHoo!

As it turns out, my controller was indeed not broken and A2 corresponds to the pin labeled zero. I can only presume this is because when you spell out “two” the last letter is an ‘o’ which looks like a zero. After that was solved, the code did end up producing the organic ‘heartbeat’ effect I was hoping to achieve!

Leave a Reply