Page 6

Single-Chip Voice Recording & Playback Device for Multiple 40- to 60-Second Messages

Microprocessor Controlled Message Management

to help simplify microprocessor controlled message manage-

ment. When controlling messages the microprocessor essen-

tially toggles pins as described in the message management

sections describe previously. The /Busy, /Strobe, and

/M7_END pins are included to simplify handshaking between

the microprocessor and the APR9600

The /Busy pin when low indicates to the host processor that

the device is busy and that no commands can be currently

accepted. When this pin is high the device is ready to accept

and execute commands from the host.

The /Strobe pin pulses low each time a memory segments is

used. Counting pulses on this pin enables the host processor

to accurately determine how much recording time has been

has a total of eighty memory segments.

The /M7_END pin is used as an indicator that the device has

stopped its current record or playback operation. During

recording a low going pulse indicates that all memory has

been used. During playback a low pulse indicates that the

last message has played.

Microprocessor control can also be used to link several

recording time. In this application both the speaker and

microphone signals can be connected in parallel. The micro-

processor will then control which device currently drives the

speaker by enabling or disabling each device using their

respective /CE pins. A continuous message can not be

recorded in multiple devices however because the transition

from one device to the next will incur a delay that is notice-

able upon playback. For this reason it is recommended that

message boundaries and device boundaries always coincide.

Signal Storage

instantaneous voltage samples in non-volatile FLASH mem-

ory cells. Each memory cell can support voltage ranges from

0 to 256 levels. These 256 discrete voltage levels are the

playback the stored signals are retrieved from memory,

smoothed to form a continuous signal, and then amplified

before being fed to an external speaker.

Sampling Rate & Voice Quality

According to the Shannon’s sampling theorem, the highest

possible frequency component introduced to the input of a

sampling system must be equal to or less than half the sam-

pling frequency if aliasing errors are to be eliminated. The

APR9600

automatically filters its input, based on the selected

sampling frequency, to meet this requirement.

Higher sampling rates increase the bandwidth and hence the

voice quality, but they also use more memory cells for the

same length of recording time. Lower sampling rates use

fewer memory cells and effectively increase the duration

capabilities of the device, but they also reduce incoming sig-

nal bandwidth. The APR9600accommodates sampling rates

as high as 8 kHz and as low a 4 kHz. You can control the

quality/duration trade off by controlling the sampling fre-

quency.

Oscillator frequency can be changed by changing the resis-

tance from the OscR pin to GND. Table 2 summarizes resis-

tance values and the corresponding sampling frequencies, as

well as the resulting input bandwidth and duration.

Table 2

Reference Rosc Value & Sampling Frequency

Ref Rosc

Sampling

Frequency

Input

Bandwidth

Duration

84 K

W

4.2 kHz

2.1 kHz

60 sec

38 K

W

6.4 kHz

3.2 kHz

40 sec

24 K

W

8.0 kHz

4.0 kHz

32 sec

APR9600

APLUS

NOTICE: Rosc table above is for reference only,different

lots ICs will have somewhat different Rosc value

performance..