GS1582-IBE3 View Datasheet(PDF) - Semtech Corporation

Part Name

Description

Manufacturer

GS1582-IBE3

Multi-Rate Serializer with Cable Driver, Audio Multiplexer and ClockCleaner™

Semtech Corporation 

4.7.20 Audio Word Clock

The GS1582 uses two word clocks, WCLK_1 and WCLK_2. By default, Audio group 1 will

use the clock at the WCLK_1 input, and Audio group 2 will use the word clock at

WCLK_2. The word clock assignment for each audio group can be changed by

programming the GP1_WCLK_SRC[1:0] and GP2_WCLK_SRC[1:0] registers.

In AES mode, the audio clocks are extracted from the input audio data. By default, audio

clock for group 1 is extracted from channel pair 1/2, and audio clock group 2 is extracted

from channel pair 5/6. The default can be changed by programming the

GP1_WCLK_SRC[1:0] and GP2_WCLK_SRC[1:0] register. Table 4-20: Audio Clock

Selection Host Interface Settings shows the audio clock source for each setting of the

registers.

Each audio group consists of 4 channels, which share the same word clock. Therefore,

the audio data applied to each channel within the group must be the same format and

have identical word clock requirements.

NOTE: In AES mode, by default, word clock is extracted from channels 1/2 for Audio

group 1 and channels 5/6 for Audio group 2. If audio is applied only to 3/4 or 7/8 only,

then no audio is embedded until the word clock source is changed from channels 1/2 or

5/6, to channels 3/4 or 7/8.

Table 4-20: Audio Clock Selection Host Interface Settings

GP_WCLK_SRC[1:0]

00b

01b

10b

11b

Word Clock Extraction Source

(AES Mode)

Channels 1/2

Channels 3/4

Channels 5/6

Channels 7/8

WCLK Source (Serial

Audio Mode)

WCLK_1

WCLK_1

WCLK_2

WCLK_2

4.7.21 GS1582 SD Audio FIFO Block

The GS1582 SD audio FIFO block contains the audio sample buffers. There is a buffer per

audio channel, which are 52 audio samples deep. At power up or reset, the read pointer

is held at the zero position until 26 samples have been written into the FIFO. Once audio

is being multiplexed, the offset between the audio sample buffer read and write pointers

is maintained at an average of 26 samples.

The position of the write pointer with respect to the read pointer is checked constantly.

If the write pointer is less than 6 samples ahead of the read pointer, a sample is repeated

from the read-side of the buffer. If the write pointer is less than 6 samples behind the

read pointer, a sample is dropped. This scheme avoids buffer underflow/overflow

conditions.

The repeat or drop sample operation is performed up to a maximum of 28 consecutive

times. After 28 repeat/drops, the GS1582 will mute (null audio packets are embedded).

GS1582 Multi-Rate Serializer with Cable Driver, Audio

Multiplexer and ClockCleanerTM

Data Sheet

40117 - 4

December 2011

59 of 115

Share Link: