FEDL2250DIGEST-12
Issue Date: Apr.6, 2009
ML2251/52/53/54/56-XXX,
ML22Q54/Q58
2-Channel Mixing Oki ADPCM Algorithm-Based Speech Synthesis LSI
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. And the sound volume
is adjustable as well.
The ML2250 family incorporates a 14-bit D/A converter and low-pass filter.
It is easy to configure a speech synthesizer by externally connecting a power amplifier and a CPU to the ML2250
family.
The ML2250 family line-up includes 2 types of products: with on-chip mask ROM, and with on-chip flash
memory.
• ML2251/52/53/54/56-XXX
This is a CMOS single chip speech synthesis device with an on-chip mask ROM. Products with 5 types of mask
ROMs are available in the ML2250 family depending upon the total playback time length.
• ML22Q54/Q58
The ML22Q54/Q58 is a speech synthesis device with a 4-Mbit flash memory built in. The voice data can be
easily written to the flash memory using a special tool. The on-chip flash memory product is suitable for the
diversified low volume production or short delivery time applications that the on-chip mask ROM product
cannot support. The ML22Q54/Q58 is most suitable for evaluation because the circuit configuration is the same
as the on-chip mask ROM product. As it is easy to write to build in-flash memory, it is able to combine fixed
message and variable message.
1/38
