back -> Nathan Hall-Snyder - Final Project

Project Plan and Review: Strobe mp3

Document Written 5/10/2010. Updated 6/4

Component List:

• Processor: Arduino Teensy

  

• Screen: Sparkfun 16x2 Black on Green LCD Display:  http://www.sparkfun.com/commerce/product_info.php?products_id=255

• Two Sparfun Rotary Encoders ( 1 volume, one interface): http://www.sparkfun.com/commerce/product_info.php?products_id=9117

• One Large Button: http://shop.ebay.com/?_from=R40&_trksid=m570&_nkw=ZB2-BE101C

• 10k pot 

• Audio Out Jack

• MP3 Decoder: http://www.sparkfun.com/commerce/product_info.php?products_id=8954

• ~100 ultra-bright white LEDs

• Lasercut LED Enclosure. Drawing: Stuff for Lasercutting.DXF (WARNING: TAB SIZES ARE INCORRECT)

Unsolved (and solved!) Technological Challenges:

• Interface/Graphic LCD

• Reading SD card contents

• Solved with an array of strings as the character list.

• Unfortunately, this limited the number of songs the device can deal with due to limitations on program memory. 

• Reading ID3 Data

• Never happened due to time constraints

• Scrolling

• Implementing reasonable scrolling on a 2 line display took several iterations of interface design, but I eventually arrived at something that worked.

• Description:
• Say you're scrolling down, the screen displays:
• > Song1
•    Song2
• And next
•    Song1
• > Song2
• and next
•   Song2
• >Song3 and so on
  

   

• Choosing/playing songs

• Implemented with a global variable for the current song playing. Turned out to be simple.
  

   

• Display/Visualizer

• Analyzing Analog Audio data to generate info for visualizer

• Developing SPI database for LED controller

• Doing all this fast enough to keep up with music

• What I learned:
• Sample parts take too long to arrive.
• Simple is better
• Beat Detection: Is actually really difficult.
• I tried for two days trying to get beat detection on the arduino to work.
• I tried hardware filtering with software filtering, and couldn't get anything reliable.
• The closest I got was using a lowpass filter in conjunction with a high-pass filter with a comparator...e.g. when the power of one was higher than the other, it pulled high. This worked about half of the time, but wasn't reliable enough to run a strobe.
• The sampling frequency of the arduino is 500 hz, not enough to analyze audio.
• It turns out that I would have needed a dsp chip, or two arduinos, to do peak detection and play mp3s. Arduinos don't have enough memory or speed to do what's required.              
• Therefore, I implemented a simple strobe with adjustable frequency.

• Power Supply (5v regulator with filter caps, 12v supply)

• Presented one of the more difficult challenges. I experimented with different linear regulators, trying to get the Arduino to boot properly.
• I ended up using regulated 5v, from a regulator with filter capacitors, which I soldered directly to the screen (which needed 5v) and to the USB input to the ardunio (allowing it to regulate down to 3.3v, for the logic circuits)

Plan B:

Same as plan A, only fewer LEDs, using LED drivers that don’t use SPI to communicate.

Plan C:

Same as plan A, using the internal PWM of the Arduino to control the LEDs instead of a seperate driver

Project Timeline:

5/14: Enhanced mp3 player functionality (mute, volume control)

5/16: Accessory parts received

5/20: LED driver functionality

5/23: Packaging design, circuit board made

6/2: Finished, party

Project Timeline Actual

5/14: Enhanced mp3 player functionality (mute, volume control)...3 days late

5/16: Accessory parts received...never arrived

5/20: LED driver functionality....never arrived, plan C adopted

5/23: Packaging design, circuit board made...packaging design/made completed ahead of schedule. No circuit board etched, perf-board used

6/2: Finished, party....completed. Rock on.