Final Project: A Dancing Toy Bee

Introduction

My final project is to create a toy bee that waves his wings to "dance" to music.

In the spirit of spiral development, I plan to develop different iterations to achieve the desired effect eventually.

V1.0: Microphone detects if there is sound. If there is sound, send a signal to servo motor to trigger movement of the wings.
V 2.0: Implement better control of servo motor movements (speed and direction) to achieve better effect of the dance.
V 3.0: Implement better detection of sound by adding low pass filter and more refined coding in translating different kind of sound to different movements.
V 4.0: Integrate the wing movement feature into the overall project, which includes movement of other parts of the toy bee and an embedded speaker in the bee.

This is how far I got to by Tuesday night: The bee waves his wings if you press the button. I also implemented microphone but it has some problem. VIDEO

Getting Started: A Failed Attempt to Integrate with Previous Project

The wing movement feature is intended to be a feature to add on to our prototype of BeeWhiz. So my original wish is to add this development effort to our (one and only) old prototype. However, the prototype stopped working a few weeks ago and our robotic engineer is no longer able to support BeeWhiz project.

This is how it looks like when it worked! VIDEO

So I first set out trying to fix our old prototype by replacing two boards, of which our previous robotic engineer suspect the chips were burned during user test.

       

After the replacement, the bee still does not function. I tried to troubleshot other part of the prototype, but it turned out that it's very hard since we did not have any documentation for the previous work. And our robotic engineer put hot glues around all the connection to increase the strength of the connections. This was a quick and dirty fix in development process, but as a result it is almost impossible to see what's inside every glue clot without heating it up and potentially damaging the connections.

This is how it looks like inside the prototype. I took pictures to try to figure out the connections among different parts (Arduino + hacked furby + stuffed animal outfit).

       

As time is running out, I decide to park the idea of fixing the old prototype and build the new feature into it. Instead, I decided to develop the wing movement feature as a stand-alone project and leave the fixing and integration after this week. @HTM, if you know of anybody who would be interested and available in helping with rebuild / fixing the BeeWhiz prototype, please let me know. Much appreciated!

Restart the Project: Taking a Stand-alone Module Development Approach

I bought this buddy from Amazon as the toy bee that I will implement the wing movement function on.



3D Design

I designed the 3D models of the wings in Rhino and printed two sets of them, one set from GSD 3D printer and another set from MakerBot in our lab.



Here's the design file. And below are two sets of wings that I printed. The blue ones are from Maker Bot and the white ones are from GSD 3D printer. And I used a knife to take off the support structure from the blue wings.

       


Electronics

I redesigned two of Neil's boards in Eagle, the microphone input board from Input Devices week and the servo motor board from Output Devices week. Then I milled the boards and stuffed it.

       

At first, I didn't solder on the microphone since I want to make sure that my chip is working. After I successfully programed my chip, I added on the microphone. For polarization of microphone, the end with marks is negative. I also used below wire clippers as a temporary setup to prepare to test out my code before I do a permanent setup.

       

I successfully programmed both of the boards. Here are my code and make files.

The connection between the two boards are established as follows:

If ACD>700, set PB2 of microphone board as high and a wire goes from PB2 of microphone board to PB2 of servo motor board. On servo motor board, if PB2 is high, trigger the servo motor movement via PWM.

However, both boards experienced some problems when I actually power it to see the effect.

At first, the servo motor didn't move as it's supposed to (VIDEO) because I didn't set the fuses correctly. The board was using the internal clock instead of external 20MHz crystal.

The fix for that is to change the make file code into: avrdude -p t44 -P usb -c avrisp2 -U lfuse:2:0x7E:m. You can also just run this line in command window after generating hex file.

IT WORKED!

The microphone board has a bigger problem. The board doesn't have any serial communication when I ran python terminal. And I also don't see any voltage changes from oscilloscope. The voltage between ground and PB2 didn't (always) move according to the sound. VIDEO

  

After checking and re-soldering the board multiple times with the help from many great people, I still couldn't make the microphone board work. :(

I decided to simply my project and add a button as input device. That was easy! Shake hands with my bee and it will wave his wings for you.

Assembly

And I continued to assemble the wings, bee, the battery and other parts.

       

One More Try for Microphone...

In the spirit of spiral development and ultimate resilience, I decided to try to use a SparkFun microphone board, which has built-in amplifier in it. I made headers for the board to connect it with the ISP header on the servo motor board. I was hoping to use one of the MOSI pin to take audio signal and use that signal to trigger servo motor movement. But unfortunately, I wasn't able to get the c code to work yet.

   

Learned a Lot from Final Project

During final project I've learned many things and improved many skills, including: