Tag: status report

This week, I helped write the Design presentation, created a Solidworks model of the rough idea for how the components would look like, finalized ordering for plastic parts, and discussed more design decisions with the team.

The CAD model really helped cement some of our design decisions such as the configuration the lights would be placed (we decided in a row would look better than zig zag), and the exact spacing of parts (such as deciding the mics will go 1⁄4 and 3⁄4 of the way along the length for equidistance from the center). Because we’ll control all the Neopixels with 1 pin, this means that 1 side of the lights must be on the opposite side as where the pin is, so this will mean our PCB will have traces that travel under the microcontroller to get to the other side. This will probably be fine, since that is a design practice used in industry as well, and it won’t be traveling too far.

To finalize the size of the tubing, we chose ones with the thinnest walls because we’re still trying to minimize the size, but also I did some estimated calculations of choosing an inner diameter tubing that’s slightly smaller than the current width of the PCB. This is because we’re going to try changing the tube shape technique mentioned last week, so the compressed oval should have the dimensions we need. A final consideration that I can think of is to remember to cut some holes on the ends of the PCB so I can tie string to them to help pull the PCB through the tube, since that’ll probably be easier than pushing. Last resort for the tubing would be to cut it open and reseal, but that sounds like we might as well try the plastic sheet idea.

Finally, the idea of swapping batteries came up, so I had some ideas to add velcro to the plastic wrap surrounding it for more convenient access instead of permanently sealing it, but that remains to be seen when we interview our user group on Monday, as discussed in the team report.

Last week, I said I’d try to get to the PCB and Beetle, but unfortunately this week was pretty exhausting for all these deadlines for classes so I didn’t get to that (in addition, the Beetle didn’t get here in time early enough in the week to pick it up and start testing). PCB used to not be on the Gaant chart,and now it is. I would really like to have a PCB prototype done by the end of next week in order to utilize spring break “waiting time for shipping” effectively. The electronic prototyping with the Beetle was scheduled for next week anyways, so that’s still on track technically.

This week, I read some Beetle documentation, wrangled WordPress into displaying weekly reports properly, and made some more design decisions for the layout and materials/manufacturing.

First, I looked into finding polyester cord and cord locks. The reason I chose polyester is because I believe it would be less likely to fray other materials, and is also quite silky so it feels nice. In terms of size, I went with 3mm because it should be compatible with the cord lock size I chose (it has a small size and a simple mechanism), and seems decently thick enough to be strong.

• Amazon.com: INSPIRELLE 3mm Black Satin Cord Rattail Silk Cord Chinese Knot Thread for Jewelry Making, 50 Yards Spool
• Adjustable Cord Lock – Round Ball Style – Single Hole End Toggle for D – SpecialistID.com

Not only will the cord be used for adjustability, but I think it’s a decent way of protecting wires that connect the batteries as well (wrap/knot the thread around the wires, friendship bracelet style). Since this could be done last, it would be easier than putting tubing around the wires because that would need to be done before the soldering (unless you want to cut and reseal the tube, which is painful).

To attach the cord to the bracelet, I came up with using “stoppers” that fit the holes on the ends of the plastic tube, with a hole in the middle for the string to go through. Basically, it looks like a metal disk washer. The stoppers could be glued into place to help seal off the components from outside damage like dust/water, and a knot on the end/some glue can also prevent the cord from slipping through the hole.

Unfortunately, plastic tubing to protect the PCB was a lot more painful to find, which I will get into later.

My initial plan was to CAD model some designs so we can have a more concrete idea of what the bracelet will look like. Unfortunately, that requires some designs first, and drawing on paper is much faster, so see those sketches in the team report. The idea with those sketches was to aesthetically space out the lights such that there wouldn’t be a black hole where the microcontroller is, although it might be a bit hard to tell since they’re not drawn to scale. We’re also not sure what number of lights will be sufficient to not look empty, but also balance power usage. The reason the microcontroller is in the middle instead of the edge is to minimize the max length needed for the traces. Although if there are sufficient benefits to putting the microcontroller on the end, perhaps we will consider doing that instead.

Another annoyance is that the Beetle doesn’t have all the desired dimensions (the 25mm length doesn’t include all of the USB port, and no height listed), so we will just have to guesstimate in the CAD and PCB. At first, it seemed like the height of the PCB + components would only be about 5mm, which is great because many plastic tubes can be that size. Unfortunately, since tubes are circular, we also need to consider our width dimension, which is going to be about 30mm when accounting for some buffer room because of the length of the Beetle. We can’t rotate the Beetle to make the width “smaller”, because then the USB port would be hard to access for the user (perpendicular to the PCB board length is easier to reach, instead of being along the PCB length because of the angle the wire would have to bend to avoid other components). Finding a 25mm tube was a lot harder, and it’s also a bit of a waste of space in the height dimension. Bendable, oval tubes are also quite hard to find. So we have some options that we are going to experiment with and hopefully at least one of these prototypes will work:

1. Buy a smaller diameter plastic tube, and use heat and weights to smush into an oval shape. McMaster-Carr
2. Buy those plastic sheets that are used to cover tables, and cut out the correct shape, use hot glue to wrap into a tube shape. Amazon.com: NECAUX Custom Multisize 1.5mm Thick Clear PVC Table Cover Protector – 14 x 14 Inch Waterproof Crystal Soft Plastic Square Tabletop Protective Pad for End Table/Night Stand/Side Table, 2pcs : Home & Kitchen

Since we initially thought we didn’t need a PCB board and now we do, we will be designing that concurrently while waiting for the orders of the other parts to come in, so I don’t think we have encountered any “blocking” actions and can be relatively on schedule. If the other parts come by next week we can unit test/sanity check some things before we put it onto a PCB. I’m not sure if it would be ambitious to have a PCB ordered by spring break? That way we can use spring break to wait for the shipment.

For next week, I plan to:

• Design some CAD models of bracelet layout (they are going to be simplified boxes with guesstimated dimensions if I can’t get pre-made CAD of the objects)
• Try PCB prototypes if CAD goes well
• Read more Beetle documentation
  • If the beetle comes in I can start poking it with hello world and other sanity checks
• Think of more ideas for materials/prototyping/edge cases for how it’s built?

I spent some time searching for small microcontrollers. There seemed to be many promising results at first, but I’ll go into more detail below:

• Gemma is 1” diameter, so nice and small. In addition, it comes with a tutorial for how to control a bunch of Neopixels in different colors with just 1 pin, so this is a good sign that we don’t need to individually pin each LED (which would save a lot of wiring and pins). Unfortunately, Gemma was eliminated because a) there wouldn’t be enough pins left to hook up 3 microphones and Bluetooth, and b) it doesn’t support USB charging of batteries, which would be very convenient instead of asking the user to take out the battery and recharge: Adafruit GEMMA M0 – Miniature wearable electronic platform : ID 3501 : $9.95 : Adafruit Industries, Unique & fun DIY electronics and kits
  • Assembly | Gemma Color Touch Pendant Necklace | Adafruit Learning System
• Circuit Playground Express has lights directly on the board, so that would also conveniently save the need to buy separate lights. It’s also a bit bigger than Gemma, at 2” diameter, but it also comes with more pins. This was eliminated because a) our initial idea was for lights to be around the wrist for better viewing, and b) no Bluetooth built in: we estimate that controlling Bluetooth would be more difficult that controlling lights, so we’d rather have the Bluetooth interface be easier. Circuit Playground Express : ID 3333 : $24.95 : Adafruit Industries, Unique & fun DIY electronics and kits
• FLORA is 1.75” diameter, so in between the above 2, but it also has many pins, which is nice. Since it’s marketed as a wearable electronic, that’s a good sign for being safe to use close to the body. Unfortunately, this one also doesn’t have a built-in charger or Bluetooth. FLORA – Wearable electronic platform: Arduino-compatible [v3] : ID 659 : $14.95 : Adafruit Industries, Unique & fun DIY electronics and kits
• The only rectangular one of the list at 1” length, Trinket M0 also decently small with many pins. Unfortunately, like the FLORA, it was eliminated for not having a built-in charger or Bluetooth. Adafruit Trinket M0 – for use with CircuitPython & Arduino IDE : ID 3500 : $8.95 : Adafruit Industries, Unique & fun DIY electronics and kits

Conclusion: although these candidates ended up being eliminated, these possibilities were still important in helping us narrow down our use cases and think more deeply about implementation. The first big takeaway is that we’ve proved that we only need 1 LED pin to individually control a whole array of lights (with the help of the libraries listed in the Gemma pendant link tutorial), which also helped us discover that NEOPixels were a nicer alternative to LEDs. Second, we had a good discussion about ranking what we value most in our parts, such as whether Bluetooth or LEDs were more important, and also how to charge the battery. Initially, we thought it would be fine to take the battery out to charge, but since the battery wires are likely going to be soldered to the switch, that’s probably not possible. This isn’t something we would’ve thought of if it weren’t for comparing all the different features of the microcontrollers we had available and ranking what we valued in our use cases.

I think the project schedule is on-time with the new Gantt chart changes, now that we have our main parts decided upon. However, we shouldn’t slow down the pace given that it did take changing the Gantt chart.

Next week, I plan to start finding 3D models of all the components we will be using, and trying out different arrangements to see how they’ll fit together, so we have an estimate and can be prepared before parts arrive. We can also analyze responses from our user survey to see if there are other design implementations we can narrow down, and go from there.