Michael Ang

ITP students are a rather eager bunch. Not content with just taking classes, the actually teach other classes/workshops called DriveBys as well. Normally they are on Thursday nights, but there is such an excess of keen that they happen on Monday nights now as well.

This Monday, November 12 at 9:15pm at ITP, Nick Sears and myself will teach a drive-by on how to make a printed circuit board for your electronics projects. Moving your circuit to a soldered board will make it much more reliable and rugged than if it stays on a breadboard. Anything you intend to transport or keep running without your continued Tender Loving Care should end up soldered. (Some would say wirewrap is also acceptable, but I’ve never gotten good at that.) Creating a schematic of your circuit will help you create the circuit in the first place and is useful for debugging and communicating your design (for example, to your future self).

We will cover how to create schematics in Eagle PCB, lay out a board and get it fabricated. We will also briefly discuss making your own boards using homebrew techniques, and give an introduction to surface-mount soldering.

Here are some relevant links:

• Presentation slides (ppt, NEW!)
• Nick’s post about software tools for creating PCBs
• CadSoft/Eagle homepage
• Mac OSX requires X11 software as well
• Eagle tour with complete list of steps for making the schematic, laying out the board, and producing the output files
• Basic Eagle tutorial (the straight lines in the schematic are kind of ugly)
• More detailed tutorial (pdf, talks about quirks on Linux)
• Eagle schematic tutorial on Instructables
• Eagle 4.1 Tutorial/Manual (big pdf)
• Short Eagle tutorial including how to build custom parts
• Eagle part libraries (e.g. for AVR)
• Nice tutorial on PCB layout
• Some online PCB fabricators
  • BatchPCB (SparkFun frontend to Gold Phoenix PCB)
  • Gold Phoenix (based in China)
  • Custom PCB (based in Malaysia)
• Free online board checking
• Limor’s PCB resources
• ITP tutorial on soldering perf boards (an alternative to printed boards)
• ITP tutorial on etching your own boards
• Build Your Own Printed Circuit Board (book) by Al Williams (flipped through and looks good)

Advanced links:

• Mixed analog/digital board layout techniques(pdf)
• Circuit board layout techniques from Texas Instruments

There should be a video up sometime after the event (thanks Tikva!)

After creating some crazy wigs out of blue insulation foam through the highly inefficient process of 3D modelling and laser cutting it turned out my wig was too big. So I took out a few panels and cut it down with a Leatherman (to a tolerance of a few thousandths of an inch…. sure). Long live the Foamhawk!

Attaching it to my head was a challenge. I wanted something that would be secure but also have some give for when I (inevitably) collided with something. (It turned out to be the refrigerator door at the deli across the street from school.) The foam didn’t seem like it would take an adhesive very well (either melting or not sticking). In the end I came up with a system where bobby pins were attached to wire held to the foamhawk with screws. Nice and secure, but not too secure, and easily fixable in the field.

Foamhawk pictures (this site)
Foamhawk pictures (Flickr)

Yesterday my Assistive Technology class visited the Adaptive Design Association in midtown Manhattan. Adaptive Design creates custom equipment for children with disabilities so they can more fully participate in daily life including education and play. A lot of the work they do is creating custom seating. If a child can’t sit properly for extended periods of time it makes it very difficult for them to learn.

ADA primarily uses cardboard as their building material since it is easy to work with, easy to obtain, and fairly durable. Since each piece is basically custom they need to be able to build and adapt quickly. They have developed techniques to give the pieces a finished appearance and you’d hardly guess it was cardboard. ADA is a non-profit and you can get involved and help out!

It’s remarkable how much a difference something built with simple materials and technology can make. The largest cost seems to be the labour — about 2-3 full days are required to build a standard chair. Of course I started thinking using a laser-cutter. They already have an established workflow but it might help for designs that are fairly standard. But the bigger gains are probably to be made from documenting their processes and applying more people in more cities to the problem. Also there are some smaller things that might help within their existing hand-tool approach like larger glue spreaders and jigs for cutting the paper strips that line the cardboard.

ADA site visit pictures (this site)
ADA site visit pictures (Flickr)

For Halloween my friend Stefan and I decided to go way geeky and make some wigs out of blue foam using the Eyebeam laser cutter. Stefan and Karl have been working on scripts in Maya to automatically divide a 3D surface into profile curves suitable for rapid fabrication so this is proof of concept for ongoing work. It’s also silly. We took digital photos of Stefan’s head from different angles, then imported them into Maya to use as reference for creating a 3D surface. Then Stefan’s script generated the profile curves and we exported them to CorelDraw using the Maya vector renderer. From there we manually tweaked the curves to get the spikes and shapes we want (in future this will be algorithmically driven) and cut them into 1″ blue insulation foam on the laser cutter.

Unfortunately mine didn’t really turn out. I got the scale wrong by about 25% so it was larger than intended. Must have been the foam fumes.

Many thanks to Eyebeam for giving us time on the laser cutter!!

Poly(gon)wig pictures

(Technical notes after the jump.)

(more…)

On day 3 I played some more with surfaces in SolidWorks. The Dimonte Group “curvy stuff” tutorials are great. SolidWorks “solids” are actually macros that create surfaces! I still haven’t gotten my head around SolidWorks enough to create a snap-together part out of a lofted surface.

On day 4 I wrote some code in Processing to create scaled cubes according to a simple algorithm, captured it using OGLE and printed it out. The goal is to take away the barrier from thought (as expressed as code) and physical objects. Working with SolidWorks feels something like sculpting. It feels good, but the work is definitely being done with your hands. Writing an algorithm and having it create physical objects feels completely different. Like the objects come directly from ideas.

Pictures from Eyebeam day 3/4

I decided to learn SolidWorks since it’s designed for solids modelling and seems a perfect match for the 3D printer. Brian (another 3D intern) and Michael (the R&D Tech Director) tend towards Maya. Maya seems like the tool to use for dealing with characters and geometry captured from graphics-oriented programs (using the incomparable OpenLab creation OGLE of course), and SolidWorks is stronger on the mechanical engineering side and designing parts from scratch.

SolidWorks seems awesome so far. You build parts by choosing an operation (e.g. extrude), drawing a sketch, then choosing the parameters for the operation. The complete list of operations and sketches is stored as a graph, and you can go back and change the parameters at any time. So for example if you design a box with filleted corners and an inset lip, you can go back and change the xy dimensions of the box and hey presto the box is rebuilt This was something that really frustrated me about SketchUp — once you extruded a circle, for example, you couldn’t go back and change the radius of the circle at all.

To test the accuracy of the printer and the properties of the ABS plastic it prints in I created a 1″x1″x”0.5″ box with two mating halves that holds a colour-changing LED. On one half there’s an extruded circle that holds the coin battery (with a slot cut out to pass one of the leads from the LED). The two halves fit together very snugly — I used “best vertical quality” in the Catalyst settings and this gives ridges along the vertical faces that seem to help. I put in a small tab to try to hold the battery in, but it was too small and broke off when I removed the support material. The battery stays in well by friction fit anyway.

The walls of the box are 0.06in thick, and the LED light shows up nicely through them. The thin walls where the halves mate are only 0.03in thick and there are some gaps in the printed plastic where the thin walls meet the thicker walls below (Catalyst says the minimum wall thickness is 0.04in, so this may be past the ability of the printer).

The Blinky Box feels nice in the hand and there’s something about the translucent-white-plastic-yness of it that’s kind of compelling. The 3D printer is dirt simple to operate and there’s something empowering about being able to create the kind of compound curves in plastic that once seemed the exclusive domain of designers at huge companies. The 3D printer at the OpenLab seems well-suited to making portable/wearable devices since it can make relatively sturdy/lightweight 3D forms (square boxes in the hand or against the body are lame). Definitely need to do experiment along those lines!

Pictures of LED Blinky Box design and reality (time from finished design to reality = 30 minutes!)

The kids at Eyebeam are the coolest, the 3D printer is more fun than should be legal, and SketchUp is no longer my girlfriend.

I made a compound curve in SketchUp using extrude along a 3D path, exported using the su2stl plugin and loaded the STL file into Catalyst (the software that takes an STL geometry file and generates tool paths for the printer).

But Catalyst complained about open curves. So the geometry coming from SketchUp was not a proper solid. I tried exporting to .obj and generating the STL from Maya, but same problem. I was sad. I tried to fix up the geometry in SketchUp, but once you extrude in SketchUp you basically just have a mess of polygons, so it was no joy.

Meanwhile Brian had printed out a nice little model of a hand from Maya. I tried drawing another (simpler) curve in SketchUp, but each time I tried to run the su2stl Ruby script it erased the curve! Bad mojo.

So no prints on Day 1 and I resolved to learn a better tool.

Pictures of test geometry in SketchUp and errors in Catalyst

I’m pleased to announce that I’m interning for the 3D Printing/Digital Fabrication position at the Eyebeam OpenLab.

Here’s the job description:

3D Printing/Digital Fabrication

The OpenLab houses a Dimension BST 3D Printer, which has been used internally for a number of projects, including OGLE and Robot Clothes. We are seeking an intern to coordinate and encourage use of this tool within Eyebeam, and to help bring in external collaborators who intend to use the 3D printer in ways consistent with the OpenLab’s goals. You will also probably get to do some 3D printing of your own.

Skills Required:

* Familarity with fundamentals of 3D surface and solid modelling

It will be a fairly short internship (one month) and I will see how far I can push the printer in the service of art/coolness. Current plans are to do some basic studies to figure out the tech, then some algorithmically generated forms and possibly some 3D parts for Blue Flower!

The current flagship use of the 3D printer at Eyebeam is for their OpenGL Extractor (OGLE) which lets you pull 3D models from most programs that support OpenGL. Very cool stuff. Pic of a model extracted from World of Warcraft and printed into plastic below.