AVM1504D

Rev 1.2

2002/11/6

3

GENERAL DESCRIPTION：

The

AVM1504D is single-chip 3 channels timbre generator that can synthesize (2 ch. melody and/or 1 ch. voice). It is

designed by instruction-based synthesizing method. Total has 504 instruction steps. Total voice is up to 1.5 seconds using

APLUS qualified coding algorithm(LOGPCM). Customer speech data will be edited and programmed into ROM by changing

one mask during the device fabrication.

FEATURES：

1. Single power supply can operate from 2.4V through 6V.

2. With following audio output combination.

A> voice only: all D/A are occupied by voice(VOICE MODE).

B> All 3 channel could play out with arbitrary combination; each melody channel occupy 1/4 of D/A, voice channel occupy 1/2

of D/A (SYMPHONY MODE)

3. The total voice duration is about 1.5 seconds could be partitioned up to 4 voice sections.

4. Voice+mute length could up to 2.7 seconds (6k sampling rate) for each voice section.

5. One 504 instruction steps table, could be partitioned up to 4 sub_tables in SEQUENTIAL mode.

With programmable instructions, Tempo, pitch, playback speed, rhythm, voice section, even STS ON/OFF can be

programmable changed anywhere during synthesizing.

Melody pitch: --3 to ++5 & rest node, include half node. Total has 53 tones.

Voice melody playback pitch: -6# to +4, include half node. Total has 20 tones.

Voice playback speed: 5.35 to 16k Hz. Total has 20 playback speeds.

Voice playback speed (Hz): 5.35 5.66 6.00 6.36 6.73 7.14 7.56 8.01 8.49 9.08 9.53 10.10 10.70 11.33 12.00 12.72 13.47 14.28 15.27 16.0

Voice playback pitch: -6# -7

1

1#

2

2#

3

4

4#

5

5#

6

6#

7 +1

+1#

+2

+2# +3

+4

Melody Tempo: 82 88 95 103 112 123 137 154 176 205 246 beats/minute

Total has 11 different tempo.

Rhythm difference: 1/8 1/4 3/8 1/2 5/8 3/4 1 1-1/4 1-1/2 2 2-1/2 3 3-1/2 4 0 beat

Total has 15 different rhythm

6. Only one trigger TG1 is used as SEQUENTIAL mode with resistive type Schmitt input (270K--1M) for CDS interface.

SEQUENTIAL: if the input was triggered sequentially, the device will response by one sub_table of voice step sequentially

and cyclically, from sub_table 1 to user defined sub_table n (n=1 to 4).

Debounce time: depend on current playback speed (5.3Hz to 16Hz), or trigger during standby.

the dependent speed is 7.6k.

long --- 4.2 ms to 12.5 ms

short --- 150 us to 430 us

In this mode, TG1 could be assigned as different trigger type with following options:

EDGEL/LEVEL, HOLD/UNHOLD, RETRIGGER/IRRETRIGGER.

Because only 1 trigger, it has NO option RESET/UNRESET. So playing sequence will always be sub_table 1 to sub_table n

and then feedback to sub_table 1 cyclically.

7. Only one status output (STS1) has following mask options:

□

STS1:

a>stop high_pulse

b>stop low_pulse

c>busy high_active

d>busy low_active

e>led 6Hz

f>led 3Hz

g>voice dynamic 2/4

h>melody ch1 on at first 1/4 or 1/8 beat, when pure voice, voice dynamic 2/4

i>melody ch2 on at first 1/4 or 1/8 beat, when pure voice, voice dynamic 2/4

j>both melody ch1 & ch2 on at first 1/4 or 1/8 beat, when pure voice, voice dynamic 2/4