robert moon

So this past weekend was not the most ideal flying conditions but I manage to get some user testing. New York City may not be the most ideal city either. Each park has specific kite-flying restrictions and since it is earlier spring, the Great Lawn in Central Park is not open to the public. Another problem I have run into this weekend is that many parks close at dusk.

On Saturday night (April 19th), Sinan and I arrived to the Great Lawn in Central Park. Even with a warning that Turtle Pond was close, we still made an attempt to get the kite off the ground. Here is Sinan planning to make another run to get the kite off the ground. Unfortunately there was not enough wind to get the kite up.

On Monday evening (April 21st) I decided to try D.U.M.B.O. (Down Under the Manhattan Bridge) right by the river in the hopes to get enough wind for flight. Unfortunately we had to get our flights done before dusk because the park rangers were pretty strict in the closing time of the park. This meant their was too much light in the sky to take decent long exposure photographs.

For my Nature of Code final, I decided to move away from a conversation analysis/visualization from midterm. Midnight Doodler is a drawing tool that uses the night sky as the “canvas” and a kite with controllable LEDs to create lines made visible through long exposure photography. The LEDs can be controlled by either a accelerometer driven light patterns (auto-pilot) or manually controlled LEDs using momentary switches using an xBee radio for wireless communication.

A clear windy evening (~18 m.p.h. winds) is ideal weather conditions. Two users are necessary, but more are welcome. One user is the piloting the kite and the other user controls the LED lights through momentary switches. The LEDs can be dynamically driven by the accelerometer (auto-pilot) or manually turned on/off using momentary switches and xBee wireless radio communication. The camera will need to be set on a tripod and set to ‘manual’ to slow down the shutter speed.(Suggested settings: 100ISO and f22 at 1 minute exposure) Wide angle lens offer more “canvas” area for the light drawing kite.

References

Pika Pika
Uttarayana, Indian Kite-flying Festival
GRL
Charlie Brown

I did some long exposure tests using a LED light pen that I created. I use a DIP female socket so i can change different LEDs for testing. I’m using a 9-volt battery in a 3.3 volt regulator and a momentary switch to control the on/off of the LEDs. I admit this is not optimized. I needed to put something quick together.

I’m using a canon G7, iso100, f11, 15 seconds long exposure shutter speed to shoot down an ITP hallway to test the strength of different LEDs, specifically super LEDs (10,000 mcd).

I spent the greater part of the today working on getting the Xbees talking to each other. I couldn’t have done it without the help of Ben Chao and Tom Jenkins. Thanks a million for getting me jump started. I am also using Tom Igoe’s Making Things Talk and Rob Faludi’s tutorial as references. I was able to wire up an xbee to be configured in Terminal using Screen.

1. In Terminal type this to find USB connection: ls /dev/tty.*
2. Then to start the xbee configuration type: /dev/tty.usbserial-A1000RMU 9600
3. This should hopefully get you to a blank screen, then type AT command: +++
4. If all is good you should get an “OK”
5. So the first thing you want to configure is the network
6. Type “ATID” to see what the Xbee’s current network is
7. To change it, type “+++” then “ATIDxxxx” (xxxx being some number to represent the network)
8. Type “ATID” again to make sure it updates to the new network
9. Type “ATMY” which describes the current Xbee
10. Type “ATMYxxxx” to change the Xbee’s “name”
11. “ATDH” should be set to “0″
12. Type “ATDLxxxx” to assign which Xbee to talk to
13. Finally after making all the changes, type “ATWR” to save the changes

This is pretty straight forward, but one thing that is annoying is that it writes over itself. It’s tough to see if it made any changes at all so I found myself retyping AT commands again just to make sure all the assignments are correct. Once these Xbees get configured, I moved onto setting up an Arduino to pass serial information wirelessly to another Xbee hooked up to another Arduino. More to come in Part 2.

So during Spring Break, I decided to catch up and work on the kite’s components. I decided to perfboard an Atmel8 circuit of a bare bones Arduino with an accelerometer sensor. Here is a picture of the bottom and top of the perfboard (Note the accelerometer is not shown so that the whole board can be seen easily). I mirrored the image on the left so that you can see easily were it corresponds with the top of the perfboard. I have a 3.3 voltage regulator to power the entire board. the bottom horizontal line of female headers is where the accelerometer will go that will cover the Atmel8 chip up until the 16 mhz clock (silver thing). The left hand vertical female headers are 5 analog ins. Just on the right of the Atmel8 chip, are a line of digital ins and out. The bottom 2 are RX and TX (to be hooked up to an xbee radio, also not shown) and the row of 7 female headers are digital ins and outs (to be used for the super bright LED modules on the kite).