Project  - Voice Activated Eblock

Team Members: Eric Frohnhoefer
CS 179J: Senior Design Project in Architecture / Embedded Systems

Engineering Notebook / Log

Week 1
Date Hours Summary Ideas/Issues/Plans Links/Documents
Mon, Jan 5        
Tue, Jan 6        
Wed, Jan 7 3
  • Briefly discussed plans with Ron
  • Created website with additional links relating to voice recognition IC's
  • E-mailed Susan.
  • What is the average power consumption of an Eblock?
  • How much energy is stored in a battery?
  • What toys are out there that responded to voice commands?
  • What voice recognition projects have others done?
Thur, Jan 8 3
  • Started a comparison sheet between various voice recognition IC's.
  • What is going to be the target cost of Eblock?
  • What are the trade-offs between various voice recognition IC's?
Fri, Jan 9 3
  • Looked at various software voice recognition solutions.
  • What kind of software solutions exist and what are their trade-offs?
  
Sat, Jan 10        
Sun, Jan 11        
Week 2
Date Hours Summary Ideas/Issues/Plans Links/Documents
Mon, Jan 12        
Tue, Jan 13 2
  • Meet with Ron to discuss proposal.
  • Post written project proposal.
 
Wed, Jan 14 1
  • Researched more voice IC's and worked on comparison matrix.
  • Downloaded datasheets to post on website.
  
Thur, Jan 15 1
  • Joined mailing list
  • Send e-mail about project meeting
  • Topics to discuss include join project, plans for clapper, plans for voice recognition
Fri, Jan 16        
Sat, Jan 17 2
  • Met at 10:30 in Science Library for about one hour.
    1. Eric Frohnhoefer
    2. Ron Feliciano
    3. Long Nguyen
    4. Edward Lee
    5. Rima Fata
    6. RJ Jareno
  • Next meeting Wed 9-10am in Science Library.
  • Tasks to be completed before next meeting.
    1. Complete comparison matrix for Voice IC's (Eric).
    2. Joint project proposal (Long).
    3. Preliminary clapper schematic and part list (RJ).
  • Uploaded and updated parts of website website.
  • Tentatively decide to do joint voice recognition project.
  • Choose clapper as fall back project
Sun, Jan 18        
Week 3
Date Hours Summary Ideas/Issues/Plans Links/Documents
Mon, Jan 19 1
  • Researched open source voice recognition projects that may be suitable for embedded systems.
  • We may be able to cut programming time down if we already have sample code to work off of.
  • Unable to find and example of code for use in an embedded system but found and interesting project called Sphinx
Tue, Jan 20 3
  • Worked on power estimations.
  • Posted second proposal.
  • Energy numbers a slightly off according to Duracell's datasheet.
  
Wed, Jan 21 2
  • Group meeting in Science Library to set outline of project goal and distribute work load on group proposal.
    1. Eric Frohnhoefer
    2. Ron Feliciano
    3. Long Nguyen
    4. Edward Lee
    5. Rima Fata
  • Next group meeting at 2:30 on Saturday in Surge 283.
  
Thur, Jan 22 2
  • Put together group proposal
  • Look into how voice recognition is done.
  • Looked for sample FFT and Windowing code.
  
Fri, Jan 23 1    
Sat, Jan 24        
Sun, Jan 25        
Week 4
Date Hours Summary Ideas/Issues/Plans Links/Documents
Mon, Jan 26 3
  • Group meeting in Surge 283 at 6:00. I arrived early to set Voice Extreme software one the computer in the lab.
    1. Eric Frohnhoefer
    2. Edward Lee
    3. Rima Fata
  • Edward, Rima, and I where able to program the Voice Extreme to respond to Yes/No commands.
  • Edward and Rima also worked on their clapper circuit.
  • Ways we can reduce false positives?
Tue, Jan 27 1
  • Posted wrong trade off analysis correct link added.
  
Wed, Jan 28        
Thur, Jan 29        
Fri, Jan 30        
Sat, Jan 31 3
  • Group meeting in Surge 283 at 2:30 to work on first prototype.
    1. Eric Frohnhoefer
    2. Edward Lee
    3. Rima Fata
    4. Ron Feliciano
    5. Long Nguyen
    6. RJ Jareno
    
Sun, Feb 1        
Week 5
Date Hours Summary Ideas/Issues/Plans Links/Documents
Mon, Feb 2        
Tue, Feb 3 2
  • Added status LEDs to first prototype.
  • Worked on my contribution paragraph.
  • Looked in to getting headers for Voice Extreme.
  • For bread board prototyping I plan to purchase some IC headers, ribbon cable and a standard 1mm header to connect the Voice Extreme to the bread board.
  • Tested the PIC to see if it would successfully operate at 3 volts.  Test worked and eblock seems to be functioning normally.
Wed, Feb 4 1
  • Looked into how many MIPS are in a PIC
  
  • EEMBC (Updated their benchmarks)
Thur, Feb 5 2
  • Looking into how to put the PIC and Voice Extreme into sleep mode.
  • It looks like it is feasible to put the PIC to sleep in the clapper eblock and voice eblock. However the savings in the voice eblock will be minimal because the voice extreme consumes most of the power 26 mA where as the PIC consumes 5 uA.
  • There is no good way to sleep the Voice Extreme without affecting realibility even more.
  • Tentatively plan to implement low power mode in the PIC.
  • Is there any additionally power consumption for wake up mode??
 
Fri, Feb 6        
Sat, Feb 7 4
  • Group meeting in Surge 283 at 11:00 to improve on first prototype.
    1. Eric Frohnhoefer
    2. Edward Lee
    3. Long Nguyen
    4. RJ Jareno

  • We went to Electronics Warehouse to purchase parts to help us prototype the Voice Extreme cost was $.73.

  • Also added a switch. Now the user is able differentiate between block by first saying a number the "yes" or "no"

  • Spent some time updating group proposal.
  • Still need to find ways to improve accuracy of recognition. After adding a second word false positive have be almost completely reduce however it doesn't recognize words when it should.
  • Accidentally left eblock on over night current it lasts only about 3 days :(
  
Sun, Feb 8        
Week 6
Date Hours Summary Ideas/Issues/Plans Links/Documents
Mon, Feb 9        
Tue, Feb 10 1
  • Saw Susan's picture in the Highlander.
  • Put together header for the Voice Extreme
 Plan to draw out the bread board circuit before plugging in the Voice Extreme.

Consider changing from 9V battery to 2 AA batteries.  Approximately the same size but more energy.

Need to look into power saving ideas.

  
Wed, Feb 11        
Thur, Feb 12        
Fri, Feb 13        
Sat, Feb 14 4 Meet with Ron to work on voice extreme and low-power mode for the PIC. Looked into how a low-power mode can be implemented in the clapper and future e-blocks.

WDT is able to wake the PIC up from sleep mode.  With a postscalar setting of 1:128 it times intervals of about 2.3 sec.

Built circuit from Application Note 0552 and modified it for use in an eblock.  Whenever the state of the button changes an interrupt is triggered and a packet can be sent.

  
Sun, Feb 15 2   Simulation shows the sleep code is correct however I have been unable to program the PIC.  
Week 7
Date Hours Summary Ideas/Issues/Plans Links/Documents
Mon, Feb 16 2 Completed Circuit as shown in Application Note 0552. Also completed C code for the project since only ASM code was provided. Everything seem to be working as planned.
  • Need to modify circuit and code so that change on single PIC causes interrupt.
  • Need to add in communication protocol.
  • Seems to be an issue with denounce in the button.
  
Tue, Feb 17 2 Working to modify code and circuit.
  • UART not configured properly.  LED eblock's error LED is solid yellow. Might be making a logic mistake when setting outputs.
  
Wed, Feb 18        
Thur, Feb 19        
Fri, Feb 20        
Sat, Feb 21        
Sun, Feb 22 1 Website update.    
Week 8
Date Hours Summary Ideas/Issues/Plans Links/Documents
Mon, Feb 23 2  
  • Sleep function doesn't seem to work properly. It appears that the PIC is in a state of contentious interrupt because out state LED seems to be flashing. Not sure what could be causing this.
  
Tue, Feb 24 6
  • Ron and I measured the amount of current drawn by the PIC.
  • Completed the Voice Extreme Circuit with Ron.
  • Connect VE circuit to PIC circuit for 2nd prototype..
  • Website update.
  • Everything seems to be running fine.  The only thing left to complete is to properly configure the PIC's WDT so that a packet is sent at 3 sec intervals.

 
Wed, Feb 25        
Thur, Feb 26        
Fri, Feb 27        
Sat, Feb 28        
Sun, Feb 29 3 Meet with Ron to make auto cad drawing of circuit. Also went shopping for parts.    
Week 9
Date Hours Summary Ideas/Issues/Plans Links/Documents
Mon, Mar 1        
Tue, Mar 2 6
  • Worked with Edward and Ron to complete our final prototype.
  • Researched the 7th Habit for presentation.
    
Wed, Mar 3        
Thur, Mar 4        
Fri, Mar 5        
Sat, Mar 6        
Sun, Mar 7        
Week 10
Date Hours Summary Ideas/Issues/Plans Links/Documents
Mon, Mar 8 4 Completed almost all the required documentation.    
Tue, Mar 9 5
  • Completed all documentation.
  • Attempted to modify our final design but it ended in disaster.
 Killed the VE!!!! Had it plugged in for awhile and it started to smoke.  
Wed, Mar 10      
Thur, Mar 11        
Fri, Mar 12        
Sat, Mar 13        
Sun, Mar 14        

Last Updated: 01/06/04