ANALOG BOARD FOR ROBOT CONTROL

1. Patrick Le, patle@ee.pdx.edu
   Analog software engineer, knowledge of FPAA tools from Lattice.
2. Justin Collum, jcollum@ee.uswest.net
   Manager, sound engineer. Knowledge of waveform editing and public-domain sound software.
3. Jason Popovits, popo@ece.pdx.edu
   digital software, WWW Master. Knowledge of digital FPGA tools from Lattice.
4. Phil Tran, tranphw@ee.pdx.edu
   PCB designer, design tools expert. Knowledge of Tango, ORCAD and schematic Capture.
5. Hugh Quach, Hugh.V.Quach@intel.com
   Sound Engineer. Works closely with Justin Collum.

PROJECT HISTORY

1. WEEK 1. November 27.
   1. Perkowski: What is the project about.
   2. Perkowski: Tasks assigned.
   3. Group: Discussion of tasks and milestones.
   4. Group: Find out on WWW all information about sound processing software and applications.
   5. Group: Think and brainstorm about using voice recognition and voice synthesis for a small demonstration walking robot.
2. WEEK 2. December 5.
   1. Here are new information for your group.
   2. The task of the project is to be able to store (from a microphone or other audio input) the digital sound (speech, song, effects or music) in a digital form, process it in software and next create a waveform to be played back.
      This sound should be tuned by software and by commands given to the filters in FPAA.
   3. The principle is to demonstrate usefulness of dynamically changed parameters of a field programmable analog device.
      person to contact in Lattice is Jim Krebs, email jim.krebs@latticesemi.com
      You will have to use the following chips: ispPAC10, ispPAC20, ispPAC80.
      ispPAC80 allows to realize fifth-order low pass filter. They have 8000 filters in table.
      You can realize 4 order low-pass 10-100 kHz filter with external filter with these chips.
      There are new versions of pSPICE modesl for PAC10, PAC20 and PAC80.
      "Stored" voice to be send out to speaker is in ``.wav'' format. Learn about this format. There is much information on the Internet.
      There are some laptop sound cards that can be used.
      You will have to strip the header from the .wav file.
      There is a tool for it, called COOLEDIT. It edits the textfile in hex that corresponds to the .wav file.
      (I do not know how the textfile is created yet).
      You will have to keep the sound file and sample it at slow rate, try different rates.
      8 kHz sampling will give you 10-12 seconds of speech.
      The speech can be stored in EPROM or RAM. You will need EPROM burner. Find out what we have at PSU,
      and what else we need. There are 512K EEPROMS (from whom?).
      You will have to adjust digitally, from general purpose controlling software, the amplitude, volume,
      and filter parameters. This will change the voice to a voice of robot, child, woman, etc.
      You will need a 44 pin PLD board. Jim has one.
      You will need to program counter in EPLD (to go through the waveform, like a generator project from the 271 class) and use A/D from the chip (check which PAC chip?)