OmniBoard KIT7: TD-SP01 Text-to-speech Module (056)

Get this page as PDF

last update: February 4, 2009.

  

OmniBoard Projects Home

RS232 text input is converted to robotic speech.

  • Specifications:
    • CMU American English dictionary for pronunciation lookup (with lexical stress) of 125000 words.
    • Phonemes, dictionary and user sample files stored on SD/MMC card.
    • DIY user can make his/her own phonemes if needed.
    • Interfaces with PC or other hardware RS232 serial port.
    • DAC board with audio amplifier, small speaker and RS232 level converter included.

     

  • User control via terminal (Level converter/RS232-2 serial com port @115200 bpS:) see below for the command set.
    • speak single word or whole sentences.
    • spell single letters.
    • speak phonemes.
    • play up to 50 user sample files.
    • easy-to-use commands.

     

  • What does it sound like? Here are some recorded mp3 examples:
    • "Module ready." mp3 (when booted.)
    • "Are you ready for some text ?" mp3
    • "I am a machine." mp3

     

  • Assembled and programmed pcb.
  • Battery supply +3V5 ... +6V, integrated voltage regulator.
  • Hardware is compatible with other omniboard projects and can be upgraded.
 

KIT7 (left) + DAC board (right) + flat mini speaker (top)

 

 
  Free schematics, diagrams and downloads available.

 

  
  Downloads:  
   

 

  • Phonemes, dictionary and user sample files stored on SD/MMC card: zipped files. This contains 3 directories:
    • "Dictionary" CMU American English dictionary for pronunciation lookup of 125000 words (with lexical stress)  and index file.
    • "Phonemes pcm8" 41 phonemes, 8-bit unsigned PCM 44100Hz mono WAV format.
    • "User Samples" 50 user samples, filenames "050.wav" up to "099.wav", 8-bit unsigned PCM 44100Hz mono WAV format.

     

  • Bootloader: dsPICprogrammer, dsPICbootloader.hex, howto

Shop:

How does it work?

The fast dsPIC searches each word in the dictionary index, then in the dictionary itself.

When the word is found, the corresponding phonemes are spoken through a DAC and an amplifier.

Example: how to say "pronunciation" with separate phonemes, after dictionary search:

PRONUNCIATION P R OW0 N AH2 N S IY0 EY1 SH AH0 N

  
 

 

Get this page as PDF

 

 

       

  

RS232 controls and  Command set

  • User control via terminal (Level converter/RS232-2 serial com port @115200 bpS with "type communication":) command examples in green below. These commands may be mixed in any way. Unknown words (those that cannot be found in the dictionary will be replaced by the nearest word found.)

     

  • speak single word or whole sentences: type text and return. The module speaks when a return, a point or exclamation/question mark is received. Examples:
    • Are you in for some fun?
    • Yes I am, thanks.
    • And now for something completely different!

     

  • spell single letters: same as above. Examples:
    • A B C.
    • X Y Z.

     

  • speak phonemes: type phoneme  and #, example: to play the phoneme: "AE": type AE#

 

  • play up to 50 user sample files: type the filename (050...099) with a backslash: example to play filename 098.wav: type 098\

     

 

 
 

terminal screen output
  Connections:  
   
How to make it work?
  1. Use a supply or battery between +3V5 ... +6V
  2. Serial terminal through RS232-2 or via the level converter on the DAC board.
  3. MMC or SD(HC) cards must be formatted as FAT32.
 
In case you need to reprogram the dsPIC from scratch, then use the ICSP+ connector X3, connected to a hardware programmer, like the dsPIC USB programmer: WinPic800.

Omniboard ICSP+ (X3 connector) connection pins:

  1. !MCLR/VPP
  2. PGD (data)
  3. GND
  4. PGC (clock)
  5. +3v3
  6. leave open
  7. leave open
  8. leave open (closest to pcb corner)

DAC Board diagram

 

DAC board connections:

 

KIT7 ICSP+ (X3 connector)

DAC board, JP1

AD5601 connections:
1. !MCLR/VPP
2. PC13/PGD 1. SDIN
3. GND 2. GND
4. PC14/PGC 3. !SYNC
5. +3v3 4. +3V3
6. leave open
7. PG2/SCL1 5. SCLK
8. leave open (closest to pcb corner)

 

 

KIT7 RS232-2 (X1)

 DAC Board, JP3
 

internal RS232 connections
1.      
2. PIC-TX 2. from PIC-TX
3.      
4. PIC-RX 1. to PIC-RX

 

 

DAC Board, JP2

speaker connections
1. spk-
2. spk+

 

 

DAC Board, JP4 PC serial connector
   
1. to PC-RX 2. receive data
2. from PC-TX 3. transmit data
3. GND 5. signal ground

 

 

PC Serial connector

  

 

 

 

Images: (click images to view full resolution)

    

pcb top view pcb bottom view
Diagrams: (click images to view full resolution)
   
  pcb top diagram pcb bottom diagram  
Component Datasheets:

 

Check out our development tools page.

 

Piggy-back, side view
 
 

 

 

<< back to the OmniBoard HomePage

TechDesign Electronics © 2005-2009