9

Encoder and Decoder Control

Analogue Input and Output Switching

The Control Register, Byte 0 – bits 0 to 5, are used, in conjunction with the codec Powersave Bit (Byte 1 – bit 3) to control codec

input/output conditions and sample rates. Figure 3 shows the codec functional situation.

AUDIO IN

AUDIO OUT

OUTPUT

BIAS

INPUT

BIAS

AUDIO

BYPASS

MOD

DEMOD

500k

Ω

(nom)

CVSD Codec

200k

Ω

(nom)

Fig.3 Analogue Control – with reference to Fig.1

Control Register

Circuit Switches

Codec

Decoder

Powersave

Control

Audio

Audio

Output

Bit

“5”

“4”

“3”

By-Pass

Out

Bias

0

000

OFF

ON

OFF

0

001

ON

OFF

OFF

0

010

OFF

OFF

OFF

0

011

OFF

ON

OFF

–

–––

–––

0

111

OFF

ON

OFF

1

000

OFF

OFF

ON

1

001

ON

OFF

OFF

1

010

OFF

OFF

ON

1

011

OFF

OFF

ON

–

–––

–––

1

111

OFF

OFF

ON

Encoder

Control

Input

“2”

“1”

“0”

Bias

0

000

ON

0

001

OFF

0

010

OFF

0

011

OFF

–

–––

–

0

111

OFF

1

000

ON

–

–––

–

1

111

ON

Table 5 Analogue Control – with reference to Fig.3

OFF

=

Switch Open

ON

=

Switch Closed

Note

1

1

2

Decoder ‘idling’ fed with “1010101

....” pattern at 32kb/s.

Decoder running at the selected

sampling rate.

Decoder circuits Powersaved

Encoder running at 32kb/s but

Encoder Data Output forced to ‘idle’

pattern “01010 ...”

Encoder running at selected

Sampling Rate

Encoder circuits Powersaved

Notes

1. If the Delta Codec is in the Direct Access mode, these

sampling rates will be as provided by the externally

applied clock.

2. The Input Bias switch is operated by the Control Register

Codec Powersave’ and ‘Encoder Control’ bits to provide a

relatively low impedance path for V

coupling capacitor whenever the codec is powersaved, or the

Encoder control bits are set to “0,” so that input bias can be

established quickly prior to operation.

Time Compression of Speech

The 25kb/s and 50kb/s sampling rate options are provided

for time compression (and subsequent expansion) of speech

signals.

For example, 1.0 second of speech stored at 50kb/s may

be transmitted in 0.8 seconds if played out at 64kb/s, and

finally restored to its original speed at the receiver by storing

at 64kb/s and playing out at 50kb/s. A similar result (with a

degraded SINAD) may be achieved by using 25kb/s and

32kb/s sampling rates.

However, the speech frequencies are raised by time

compression, and since the signal transmitted to air must be

band limited to 3400Hz, the effective end-to-end bandwidth

is 0.8 x 3400Hz, which is approximately 2700Hz.