Week 9

Output Devices

This week I wanted to experiment with flexible circuits and LED arrays. This is in consideration for my final project.

I used the vinyl cutter to make the circuits and went through some material explorations in order to get a circuit that was still flexible, but that did not melt during soldering. My original intent was to make the circuit transparent, but I soon learned that this is very difficult unless you have a heat resistant material (even if turning down the temp on the soldering iron, and being meticulous about it). Since I did not have an abundance of time to order clear heat resistant material, I may attempt this again in the future.

After numerous attempts (using the one layer layout on clear plastic sheet, layered circuit on clear plastic sheet, layered circuit on insulated tape), I finally settled for a thin layer of clear mylar, heat resistant tape, copper, and another layer of heat resistant tape for the via layer.

With the vinyl circuits, there is not a whole lot of forgiveness - if you solder something incorrectly it is difficult to fix since the pads just have a tendency to unstick. It is important to get it right the first time!

Now that I have a better understanding of the workflow/commands to program the board, the programming process went a whole lot smoother, though I did have a weak battery that lead to some confusion!

I got the LED running and it is pretty cool that it can take different shapes. This will be something that I continue to play with over the next weeks.

Week 8

Input Devices

I first started by milling the boards. I milled the temp sensor board as a baseline this week, as this will be helpful toward a project that I am working on for urban farming. I also finished stuffing the Fabduino board and was hoping to get to hook up multiple sensors. This did not work out for this week, as I was still trying to debug my temperature sensor. I will continue to work on the Fabduino this week.

This week I had quite a bit of difficulty actually programming the sensor. This was largely due to the fact that I did not understand the workflow to program a sensor. Big thanks to Jennifer and Matt who helped me through this. For those who may have the same issue (and I think there were a few), I will run through the steps:

1. Program board through the FabISP and make sure all of your wires are connected properly to the pins. Also check the continuity of your board to make sure there are no poor solders.

2. Run the bootloader in Arduino to fuse the bits if needed (make sure you have the proper board, serial port and programmer selected).

3. Compilation: Run the makefile from terminal. This translates the source code (from C code or assembly code) into machine code (.hex file) and allows the board to understand the C code when uploading. For Mac OS X you will need to have Crosspack installed to compile C programs and Gavrasm installed to compile assembly code.

4. Upload: Run the C code from terminal or Arduino IDE. **It is necessary to run the makefile first otherwise the board will not understand the C code.** This step writes the machine code into the program/flash memory of the microcontroller.

5. Python code is used for visualization of the information received from the sensor only! Run the python code from terminal. Make sure that you input your serial port key (you can find this number in terminal by typing: ls /dev/tty.*). This should give you back something like dev/tty.usbserial-XXXXXXXX. Paste this into your python code (port = "/dev/tty.usbserial-A80016SY", for example).

This is a very useful CBA site for programming microcontrollers which I did not stumble across until later on.

I also wanted to practice programming for the arduino, so I continued with some work with the sensors and the arduino/breadboard. There is a video of some successful readings from a temperature sensor and photo sensor for a project I am working on.

Week 7

3D Scanning and Printing

This week I began developing/testing some ideas toward my final prototype. I am testing the possibility of harnessing small amounts of energy through peltier devices in direct contact with skin and the environment (Neil - I have to try this out to see if it will work!). The first images show the model that I created in rhino, a series of chained links that would each have a surface mount LED and tiny peltier device attached to the top and bottom of each link respectively. The model is misleading, since it seems like the parts are large, though each chain is 1.5 cm in diameter. The idea is to create a flexible substrate on which to mount the necessary components. These components could then either be cast in some sort of pliable material or could be embedded in fabric. Perhaps they work better alone, or perhaps they are not necessary at all - that is what this week's prototyping is testing.

The 3D printed model is super delicate! I realized after I tried removing a few pieces on my own (and then was told that was to be handled by John and Tom - my mistake) with an exacto-knife as per the online tutorial - that the links would continue to break using this method. In the end there were a couple of issues that we will try to resolve when printing again. First, when printing on the machine that does ABS plastic, the layers are not as accurate for doing small parts - it is better to use the machine with the acrylic. Second, something went a little strange with the mix of the materials for the acrylic machine on my batch - the substrate and final material mixed together and bonded. We are going to try to run the print again to see if this fixes it. UPDATE: Ran the print again at a larger scale, the model is more robust. Additionally once I took the parts home (both for the larger and smaller printed chains) and put in warm oil bath to remove the extra substrate, the links were MUCH cleaner. Now they are moving and flexible. I still will need to reconsider the design for the final project, but this is a good first pass.

The 3D scanning on the other hand was very tiring! I scanned multiple things multiple times with different parameters and with different colors/reflectivity. The scans that I was getting were poor at best! I understand the use of this process for scanning one of a kind pieces (like ancient artifacts), but to spend this much time scanning an object that I could model more quickly and more precisely in a modeling program made this whole process seem tedious and more of a headache than it was worth.

Week 6

Embedded Programming

This week I worked on programming and making multiple boards as I wanted to get a general feeling for the different types I could work with. I also wanted to advance on a portion of a project for plant sensors that I am researching and included this towards this weeks' exploration.

The tutorial for Eagle CAD this week with Adam was super helpful, as this was the first time I have used Eagle with a clearer understanding - and found it much more useful than my past attempts! I created the .png for the ATtiny 44 with and LED and resistor with Eagle CAD.

Also thanks to David and Ed for great tutorials for the Arduino programming and ModKit. In total, I milled the hello.ftdi.44 board, the hello.arduino.168 board and also the Fabkit board. I have mostly soldered the components on the boards, though I have not programmed them all yet, but will continue working to do so. I do have my hello.ftdi.44 board blinking the LED. I had to troubleshoot the FabISP because it was not being recognized in my system information/usb panel. Once this was working and I connected my FabISP and hello.ftdi.44 board with the arduino tutorial from David - it was a fairly smooth process (minus a few hitches with the boot loader).

I also put together a kit called Botanicalls that I plan to hack for a project that I am working on. It is a smart little kit that monitors how much water your plants need and tweets the results to you. I have been working with some of the programming for ethernet and am continuing to work on that. So far I have the probes inserted into the soil of my plant and have it up and running to tweet. I would like to figure out a way to sense other aspects of plant needs such as sunlight, nutrients and also water for a hydroponic system. I also looked at David Robert's hello.plant example in a past HTM project.

Week 5

Molding and Casting

This week I created my own version of the Yoshimoto Cube, which is a magic puzzle that has a myriad of folding sequences as you can see in the video. It was very difficult and time consuming to make because it is very precise! Ultimately I am happy with the results - though I do think that I would use a different material for making this project (I tried casting plastic, but also was not super happy with the results).

I used the Shopbot to create the first part of the mold from the milling wax - this was a fairly easy process, though it took much longer to mill than expected (or predicted by the program). In the end I had to stop the milling before the final cut, I only have the rough cut, because there were others waiting and it was taking too much time. In the end, I actually liked the patterning of the rough cut, and feel it adds a nice texture to the design. I then casted the blue silicon for the second part of the two part mold. This took longer than anticipated to dry, but worked beautifully in the end for the mold.

I casted many parts, since I knew that there would be some errors and wanted to make sure I had spares. I tried casting in both the HydraStone and in plastic which I dyed with a pink highlighter. Believe it or not the highlighter dye worked really well! Unfortunately the plastic by nature produces lots of bubbles! I may revisit this material at a later time.

Because I had to work with rough cut pieces that were not the exact size anticipated, I needed to do a lot of manual labor that should not have been necessary in terms of sanding. I would have tried to mill the piece again, but the shop was completely out of wax! Precision with this project is key!

One of the most difficult aspects was deciphering how the joints were to be made and the patterning which is very complex for this project. After testing with masking tape as a flexible hinge, I was finally able to create the hinge from a plastic coated paper.

Week 4

Make Something Big

This week I used the Shopbot to create a rolling Conversation Cart. I used the cart during member week at the Media Lab to discuss ideas for my GrowPOD project. The idea was to engage people in a short conversation by offering them a tagged apple, then to discuss my "elevator" pitch with the boards attached to either side of the cart. If the person was interested in the project, I had my laptop available to discuss the project in more detail. I also created custom stools that would pull out of the cabinet of the cart for longer conversations. By moving around the Media Lab, I was able to talk to many different kinds of people - and I also gained many interesting perspectives and feedback on the project.

Thank you to those who donated their OSB to me for this project. I used three sheets of 4'x8' OSB and two 2'x8' insulated foam sheets to make this project. Also thank you to John and Tom for being extra patient and helpful!

The rolling cart was assembled using small 1/2"" dovetail joints. I used brackets on the interior of the cabinet to hold in place, but the cabinet itself did not require wood glue to stay together (I was avoiding using wood glue if possible). I used large ball bearing casters on the bottom to roll the cart around, and I also installed metal handles to push the cart and metal hinges to open the cabinet. I was experimenting with laminated foam and OSB for the seating. At first I was trying to make completely out of layers of OSB, but the seats started to become very heavy. Instead I decided to mix in layers of foam to lighten the stools - I also liked the play of materials. The cabinet was designed to hold the interlocked stools, as well as an extra shelf to hold more apples and business cards etc. Altogether, I feel it was a really successful project and I am proud of the results! I am planning to use the cart for my demos so that I can make them moveable as needed.

Week 3

Printed Circuit Board

This week I created a FabISP (in-system programmer) by milling a custom printed circuit board on the Modela machine. During this process, I learned that it was important to have a sharp end mill (blade) for the milling, otherwise you need to slow down the cutting so that you get smooth edges. The first board I milled came out with rough, uneven edges.

Next, I soldered the surface mount components for the kit. I have a good deal of experience soldering, but this is the first time I have ever tried to solder surface mount components - and boy are they tiny! I followed Neil's recommendations to start soldering components from the center of the circuit out (this is very helpful so that you do not have to maneuver around components). I also began by adding a little solder to one of the pads for each component and then laying the component on top, melting the solder and using this connection as a base to hold the component in place. This is very helpful!!! Another important point is to follow instructions for the solder jumpers - they must be jumped to program the board and then broken before hooking up to your computer.

I then connected the fabISP to the computer and downloaded the firmware.zip and CrossPack AVR - changing the makefile as noted here and then running the avrdude. One thing that I did not realize - silly me - was that the FabISP must run from another micro-controller to install, it can not be programmed without one (in other words avrdude is an actual physical piece of hardward, not just programming code). Jennifer gave me this helpful hint! After running the program I got a green light! Then I broke the connection between SJ1 and SJ2. I hooked up to my computer and the USB port under system information did not recognize the device. I checked all the solder joints for continuity, checked the resistance of the resistors, and re-soldered areas that I thought may need better connection.

Week 2

Press-fit Construction Kit

I was interested in making a beautiful and simple press fit construction kit for jewelry. I became fascinated with developing a flexible joint for the system, that became an integral part of the design. I also wanted to use the vinyl cutter exclusively since I have previous experience with the laser cutter, and because I wanted to cut thin sheets of metal which is not possible on the laser cutter.

To begin with, I was experimenting with floral patterns, and then realized after testing and making some simple models, that this was not the route I wanted to take. I began sketching a system to that was less tedious, and decided to implement. The final design incorporates components that are modular and scalable - the chaining together of the pieces gives the desired length for the user, and the different size components allows people to create their own pattern.

In order to get the desired rigidity for the joint to move smoothly - I experimented with different laminations of the thin copper on different backings. I finally determined that a thin layer of acetate sandwiched between two layers of copper was best.

Another material that I wanted to explore was using colored sheets of thin plastic, that when overlapped would produce a color-wheel effect. The joints and color transparencies would make the connections visually active.

Week 1

Final Project Proposal

As a designer by training, I believe in conceptualizing and creating functional, useful and beautiful works. I am very interested in pursuing a final project that I have been excited to build for some time. The project is also complementary to my research within my lab group. I plan to build an installation using the energy from a (or multiple) custom made Stirling engines or thermogenerators, that would involve the different modes of fabrication learned in the class. A Stirling engine runs through temperature differentials between surfaces, otherwise known as the Seebeck Effect.

Two strategies that I am considering for the design and application of this custom system:

1. Capturing waste energy of city systems such as excess heat from the subway that could be transformed into cool air.

2. Application for micro scale stirling engine/thermogenerator - a fabric that could be worn on skin, creating energy through temperature differentials from the body and the environment.

Project Update

What is the goal?

The goal is to capture small amounts of underutilized energy and convert into useful energy that will accumulate over time.

What is the prior art?

There are a number of models built of sterling engines (and I am also researching thermogenerators/peltier effect), but the sterling engine assemblies are often used as diagrammatic displays the engine rather than being used toward a purposeful application. Thermogenerators are often used to cool off CPU's which is the largest application for the technology.

What are the systems and components?

The system will require input, processing and output. The input will be the initial heat (or cold) entered into the generator. I would like to design this for use in clothing applied to warm skin, so the input would need to be embedded in some kind of fabric that would not dissipate, rather accentuate, the thermal characteristics needed for the system to work. The processing portion would be the generator - and its design. I would like to break down the sterling engine/thermogenerator to develop a series of small scale generators. If possible, I will try to hack into existing devices to work as efficiently as possible. The output will be to cool (or inversely to heat) an area contrary to the initial input. I would also like this to emanate from the fabric which will be used as a conduit to transfer the cool air to parts of the body needed to be cooled.

What questions will you need to answer?

I will need to figure out whether this will work at a small scale - whether the effect will be great enough to make a difference. I will need to figure out a way to make this in a manner that is cost effective, and will need to hack into parts that are available. I have found some parts, but need to test how effective they will be.