GENERAL DESCRIPTION
The  ML2250  family is  a  2-channel  mixing  speech  synthesis  device  with  an  on-chip voice  data  (i.e.,  phrases) storing mask ROM and a flash memory. Besides playing the built-in voice data, this device can output voice data that  is  input  from  outside  the  device.  This ML2250  family  allows  selecting  the  playback  method from  the  8-bit PCM, non-linear  8-bit  PCM, 16-bit PCM, 2-bit ADPCM2, and 4-bit ADPCM2 algorithms. Andthe sound volume is adjustable as well.

FEATURES
Non-linear 8-bit PCM, 8-bit PCM, 16-bit PCM, 2-bit ADPCM2, and 4-bit ADPCM2 algorithms
Serial input/parallel input selectable
Phrase controltablefunctioni.e., user definable phrase control table function
2 channels mixingfunction
Master clock frequency:  4.096 MHz
Sampling frequency:  4.0 kHz, 5.3 kHz, 6.0kHz, 6.4 kHz, 8.0 kHz, 10.7 kHz, 12.0kHz, 12.8 kHz, 16.0 kHz, 21.3 kHz, 24.0kHz, 25.6 kHz, 32.0 kHz, 42.7 kHz, 48 kHz
Maximum number of phrases:  256 phrases
Sound volume adjustment function built in(2 soundsindependently adjustable in 29 steps)
External voice data can be input
14-bit D/A converter built in
Built-in low-pass filter:  Digital filter
Package:  44-pin plastic QFP  (QFP44-P-910-0.80-2K)
                                       (ML2251-XXXGA/ ML2252-XXXGA/ ML2253-XXXGA/
                                        ML2254-XXXGA /ML2256-XXXGA / ML22Q54GA/ML22Q58GA)
             33-pin W-CSP        (P-VFLGA33-5.03X5.78-0.80-W)
                                       (ML2253-XXXHB/ ML2254-XXXHB)
                                       (ML2256-XXXHB)