We have a MAKE:Princeton meeting tomorrow! This one will be inside and away from flame as we discuss sustainability, energy, and ways to make your home more energy efficient. Discussion will include an analysis of (1) how energy is used in a home, (2) how energy is wasted in a home, (3) how wasted energy can be reduced or reclaimed, and (4) different ways to think about energy efficiency.

And, since this is make, feel free to bring your projects to show and tell and bring a friend.

Where: Room 039, East Pyne Hall, Princeton University Campus
When: Tomorrow, 15 April 2009, 7:00pm
What: MAKE:sustainability
Who: You and your friends.

See you there! (and sorry again for the last minuteness).

I have parts I’d like to cast for my Stirling Engine project. Additionally, Ken suggested we make some sort of small memento, such as MAKE:Princeton plaques or keychains, so everyone can bring something home.

Because of the nature of this activity, we cannot do it inside. Also, for safety, we want to do it during daylight hours. Thus, we will be conducting this meeting at a new location (Nick’s back yard) and at an earlier time (5pm). We realize that some of you will be coming from work, so don’t feel bad about showing up late; we’d rather you get there late then miss the whole thing!

Where: Nick’s Back Yard, 222 King St, Princeton
When: 1 April 2009; 5pm–dark (feel free to show up late)
What: Casting aluminum in sand and plaster

Bring: friends, ideas, enthusiasm, and scrap aluminum (don’t bring cans, as their high surface area causes them to oxidize instead of melt).

A lot of people came out; the tag line “melt metal” really piques your interest. We were in a race before the sun went down, and despite repeated mistakes, we were able to build a sand mold, melt some aluminum, and pour to create a duplicate.

Ken and I first dug a hole in the ground and built a cylinder of bricks within it. We positioned a steel pipe (for air flow) and then covered the walls in dirt. We used a broken vacuum cleaner as our air source, and charcoal as our fuel:

Next we rammed up a mold. In this technique, we build a sand mold in two halves, separable along the horizontal. We must have done this four times as the top half kept collapsing. Our problems: we didn’t have enough clay in the sand, we didn’t tamp the sand enough, and we didn’t use a parting sand between layers. Once we realized this, it got a lot easier.

Once we had that done, we got to start up the fire. The fire started slowly; the tape on the vacuum came loose and we didn’t get enough air at first. Soon enough, a good fire started, and our crucible turned bright red:

Just as it was starting to get dark, we had enough liquid aluminum for the pour:

Here’s a video of the pour:

We let it cool for a while, and when it came out it looked like this. Not perfect, but not too bad either:

It was a grand old time. Can’t wait until next time!

As you may guess, this is not something that can be done inside. As such, I’m going to host the meeting in my yard. Since we need daylight, I’m going to start the meeting early. Don’t worry if you can’t show up at the start time, this will take a while to complete, and I’ll be going at it for quite some time.

Where: 222 King St, Princeton, NJ 08540 (Butler Apartments)
When: Wednesday, 18 Mar 2009; 5pm–finish
What: backyard metal casting of aluminum, with sand casting.

Streetside parking is available. Also, the Princeton University “East Line” bus will take you to King St from the Woodrow Wilson School or from the Engineering Quad at Princeton University. It’s an easy walk or bike ride from campus.

“We’re very happy to announce that the Arduino Duemilanove has been upgraded to a more powerful microcontroller: the ATmega328. It’s fully compatible with the previous ATmega168, but with twice the memory. That includes flash memory for storing sketches (32 KB instead of 16 KB), RAM for holding variables (2KB instead of 1KB) and EEPROM for saving data when turning off the board (1 KB instead of 512 bytes). We’ve also raised the speed (in the bootloader) for uploading new sketches from 19200 baud to 57600 baud.

“Thanks to some tough negotiating by Gianluca Martino of Arduino and Smart Projects, this won’t raise the price of the board. Look for distributors to roll out the upgraded boards soon. Current Duemilanove or other Arduino boards can be upgraded by replacing the ATmega168 with a bootloaded ATmega328 (you’ll need to buy a pre-bootloaded one or use a hardware programmer). Be sure to select “Arduino w/ ATmega328″ from the Tools > Boards menu.”

Ken

Our next meeting will take place on Wednesday, 4 March 2009. Normal time and place: 7–9pm in East Pyne Hall, room 039.
There have been a few interesting suggestions for a topic. Ken has one great suggestion: everyone bring in some electronic item—one that is already broken, or that is okay to break. We’ll tear them apart, attach an oscilloscope, try to reverse engineer, and look for opportunities for re-use. On that note, I am particularly interested if anyone has things with solar cells on them, such as pocket calculators.

Also, Rob asked if anyone knows something about electrolyzed water, and is seeking someone with a chemistry background.

See you there! Bring a friend!

As unprepared as I was, everyone else brought in cool toys. Ken made an Arduino-based animated LED heart valentine for his wife. He uses the “charlieplexing” technique, in which you try to drive as many LEDs with as few wires as possible. Only one LED is lit at a time, but you cycle through them so quickly that the human eye can’t tell. Looks great already, but we talked a little about resistor choice to brighten the display. By the end of the meeting, he had resoldered it to double the brightness. We also suggested that he add some sort of a light sensor, so it can conserve battery.

(in case the embedded video doesn’t work, check here http://flickr.com/photos/npj/3291388819/)

Walt (of Hacktory fame ;) brought in his impressive Arduino-based display. The arduino is pulling image data from the flash chip (center, bottom row), and writing row at a time to the LED drivers (three of them, center row), who in turn drive the LEDs. I think he said he has four levels of grey for each LED, and has a refresh rate of about 50 frames per second.

In this video, Walt showed how this thing really rocks. Image data is from some flickr group about pictures of words.

(in case the embedded video doesn’t work, http://flickr.com/photos/npj/3291388823/)

Walt wants to build these as modules, 16×8 each, that can be plugged together to build a huge display. He wants to offload most of the signal generation from the arduino onto a programmable oscillator. Excellent project, and I want to help where I can.

Also, Walt mention an excellent supplier for those 8×8 modules seen in his project. Anyone remember the link?

We had been talking about dismantling a scrap TV for a while.  With the impending change to digital (has it happened? what’s up?), we figured there would be a lot of these in dumpsters soon, so it would be nice to know how to SAFELY dismantle and extract useful parts.  I took the power transistors, rob took the potentiometers.  And I put the tube on my officemate’s desk for the surpise on his face in the morning.

PLEASE PLEASE PLEASE be cautious, and don’t open a TV unless you know how to discharge it’s power supply.  If you don’t know what that means, don’t open a TV, even if it’s been unplugged for years!

Good stuff.  I’m glad to hear Ken’s success with the Arduino, and I look forward to Walt’s manufactured wall-o-LEDs.

Next meeting in two weeks!

Recounting last meeting’s coverage of the Kreg pocket hole jig, Jon showed us a Kreg jig in a carrying case that he purchased recently and a shelving project that he did for his wife. He talked about the challenges and successes of putting together melamine shelves.

Also recounting last meeting’s coverage of mold making and casting, Ken showed an acrylic project. Using a small plastic film canister and easily available casting acrylic (fiberglass casting resin, readily available at home centers is tinted grey and will not work for this, so Ken recommended Clearlite casting acrylic available from TAP plastics http://www.tapplastics.com), Ken has preserved wedding cake for his two son’s weddings as a memento. (Pictures to be posted

shortly.)

Ken also spoke about the use of polymer clays (Fimo or Sculpey) as a modeling tool for making parts. Polymer clay works well for non- functional, decorative items, such as moldings or frills, but for practical parts such as knobs or handles it is too fragile. Instead, sculpt the item you wish to create in polymer clay, fire it at 250 degrees for 20-25 minutes per ¼”, and use the resulting object to cast a negative mold. Then recast the object in acrylic (clear or dyed, or paint it afterwards) and you can create multiple, sturdy items.

Alternatively we also talked about using polymer clay as a one-shot negative mold for casting acrylic or plaster items. Once the medium has set up the polymer clay can be ablated away, leaving the cast object.

Nick showed us the results of his tube-casting, which utilizes the tube mitering template software that he designed.

http://www.cheaphack.net/2009/01/easily-miter-tubes.html.  After creating a complex design of mitered tubes, printing out the templates, taping them together using card stock, and then bracing the tube-template-mold in sand, he filled the tubes with plaster.  There were some voids, and as it turns out the mold was somewhat tilted, but he ended up with a post-modern sculpture which could be a desk ornament or a hat rack. One thing he has learned from this casting experiment is that due to the fine nature of plaster, all of the seams, tape and even some minutely raised lettering on the cardstock were imprinted in the plaster.

New member Walter brought us news of a DIY group in Philadelphia called The Hacktory http://thehacktory.org/. They have a shared space for working on projects, meet once a month, offer a variety of classes, and … have just received a 3D printer from a fellow inventor.

Walter told us tales of putting this thing together and making it work, and then filled us in on his own Arduino/LED matrix projects.

His goal, to make an Arduino driven 90×90 LED matrix that displays monochromatic grey scale full motion video. As part of his project he is exploring the design and manufacture of sub-components that can be daisy-chained together.

Many of us have expressed interest in a “field trip” to Philly to visit The Hacktory.

Rob has suggested that he will bring in a discarded (but working) B&W television next meeting for us to experiment with, and if Ken can find it, he will dig up his Timex/Sinclair to produce RF output for the TV.

The next meeting of MAKE:Princeton takes place TOMORROW NIGHT… As always, it will be in Room 039 of East Pyne Hall, from 7–9pm. Bring your inventions and your friends to enjoy some diy and hackery. Directions and parking info is available in earlier posts.

Before this meeting, there is a colloquium by Andrew Ng of Stanford about Intelligent Robots; details are in the previous post.

At this meeting, we’ll be discussing techniques and inventions, as well as have a get together. The list of topics is not complete yet, but it will at least include Nick’s pipe mitering software and mold making.

See you there!

STAIR: The STanford Artificial Intelligence Robot project

Andrew Y. Ng
Stanford University

DATE/TIME : Wednesday, February 4, 2009, 4:15 P.M.
LOCATION  : CS Room 105

This talk will describe the STAIR home assistant robot project, and the
satellite projects that led to key STAIR components such as (i) robotic
grasping of previously unknown objects, (ii) depth perception from a
single still image, and (iii) multi-modal robotic perception.

Since its birth in 1956, the AI dream has been to build systems that
exhibit broad-spectrum competence and intelligence.  STAIR revisits this
dream, and seeks to integrate onto a single robot platform tools drawn
from all areas of AI including learning, vision, navigation, manipulation,
planning, and speech/NLP.  This is in distinct contrast to, and also
represents an attempt to reverse, the 30 year old trend of working on
fragmented AI sub-fields.  STAIR’s goal is a useful home assistant robot,
and over the long term, we envision a single robot that can perform tasks
such as tidying up a room, using a dishwasher, fetching and delivering
items, and preparing meals.

In this talk, I’ll describe our progress on having the STAIR robot fetch
items from around the office, and on having STAIR take inventory of office
items.  Specifically, I’ll describe: (i) learning to grasp previously
unseen objects (including unloading items from a dishwasher); (ii)
probabilistic multi-resolution maps, which enable the robot to open/use
doors; (iii) a robotic foveal+peripheral vision system for object
recognition and tracking. I’ll also outline some of the main technical
ideas—such as learning 3-d reconstructions from a single still image,
and reinforcement learning algorithms for robotic control—that played
key roles in enabling these STAIR components.