learning the ropes

Shinyoung made a mold and produced a total of 10 static models of Dust to display at the ITP Spring Show.

I also realized that we need a better way to connect the audio output of the Coby MP3 player to the Dust circuit board. For the breadboard, we connected a 1/8″ mini phone plug from the output of the MP3 player into the input of the amplifier section.

I want to remove the headphone jack from the MP3 player for our smaller prototype in order to eliminate the bulky plug.

Time for surgery again.

After poking around for awhile with a multimeter to try to figure out how the jack was wired, I gave up and focused my efforts on freeing the SMD jack from the MP3 player’s board. It was easier to diagnose the wiring of the jack once it was off the boad.

Since I was tired, I soldered the wires on the wrong side of the board at first (wasn’t paying attention to the orientation of the USB pins I soldered on earlier in the day).

In order to reduce the size of the new Dust prototype, I am further hacking apart the Coby MP3 player. I bought an SMD mini USB connector from Digikey to attach to the printed circuit board. My plan is to replace the USB connector on the Coby unit with four male headers which will plug into the main Dust board. It was very confusing to make sense of the USB pinout. The mini connector has 5 pins, but the original large USB connector has four pins.

In the process of coaxing the original USB connector out, I damaged one of the vias and pads on the Coby board. I hope the damage isn’t permanent. You can see the lifted pad right below the “B1″ label (upside down)

I tried to repair the lifted trace and pulled out pad with a short section of wrapping wire.

Notice the wrapping wire soldered onto the end of a male header pin. The next trick is to solder the stripped end of the wrapping wire to the lifted circuit trace. I applied superglue to the bottom of the trace to hold it down and then scraped its surface to expose the copper.

I’ll won’t know if the repair was successful until I get the MP3 player attached to the main Dust circuit board. I’m still a little concerned that the USB wires going between the boards need to be shielded.

I finished another revision of the Dust protoype schematic.

Changes from the last version include:

• Revised amplfier section. The breadboarded amplifier circuit didn’t match the circuit in the physical computing textbook.
• Revised relays. I drew custom parts in the Eagle library for the Omron G6L-1F relays we ordered.
• Renamed parts on schematic so jumpers, transistors, and diodes have matching numbers (eg. Q2, R2, D2, J2). It was difficult to correlate them on the board layout view when their numbers were different.
• Changed motor diode to a SOT-23 package.
• Added landing spots (“Through Hole Pads” in EagleSpeak) for the qprox electrode and the electrode shield.
• Added USB connector

I also finished the first “real” version of the board layout.

It will be a double-sided board with handmade vias in several spots. There was a point yesterday evening when I just couldn’t deal with the complexity of routing things on one side of the board only. I asked Rob Faludi to help me figure out how to add “vias” properly. The whole trick with the vias in Eagle is using the Ratsnest command, which redraws the screen and apparently recomputes connections.

Based on advice from a tutorial I found on Instructables, I drew registration marks on the top and bottom of the board layouts so I can align them perfectly before I iron the toner onto my copper clad board.

Next, I printed out the board to see how things would fit.

Next Steps

• Test amplifier circuit on breadboard
• Battery testing — can we power with multiple watch batteries or do we need a 9-volt?
• Measure the pin spacing on power switch
• Add programming pins for connecting to the FTDI USB port (for upgrading Dust firmware)
• Fix resistor and capacitor package sizes on board design. Resistors are 0805 and there are two types of capacitors: 4×5.5mm and 6.3x8mm
• Etching
• Soldering

This is the schematic for the version of Dust we showed at our final Wearables presentation. I’m still working on updates to shrink the size of the circuitry down.

The following is a record of an electronic surgery performed on 4/18/2007 in the Physical Computing Laboratory at ITP, wherein the patient “Dust” was brought to “life” in the manner of Frankenstein.

We added more details to the state machine which defines Dust’s behavior and started writing code to implement it.

This reminds me of the story of the gingerbread man — you know, he ran as fast as he could… but still ended up in the oven.

We started out with two halves joined together, but then realized that our presentation model should incorporate some of the circuitry. We cut out a door in the back to hold the speaker, LEDS, and vibrating motor.

Polymer clay (in this case SculpeyPremo) bakes for 25 minutes at 275° (or close to it).

After baking, the surface looks more matte.