adau1373 Analog Devices, Inc., adau1373 Datasheet - Page 55

• Manufacturer Part Number: adau1373
• Description: Low Power Codec With Speaker And Headphone Amplifier Adau1373
• Manufacturer: Analog Devices, Inc.

ASYNCHRONOUS SAMPLE RATE CONVERTER
The ADAU1373 includes three bidirectional ASRCs to convert
the sample rate for the selected digital audio interface port. The
ASRCs can be set in automatic ratio detect mode to calculate the
required ratio between the input and internal sample rate. This
mode writes the value in Register 0x4A and Register 0x4B for
Digital Audio Interface A, Register 0x4C and Register 0x4D for
Digital Audio Interface B, and Register 0x4E and Register 0x4F for
Digital Audio Interface C. The automatic ratio detect can be
disabled, and the fractional value can be written into the same
registers for manual setting. The maximum sample rate is 48 kHz
for the ASRC, as well as for the internal DSP. This sets the limi-
tation on the ASRC such that the interface port sample rate
must be equal to or less than the internal core sample rate.
The output-to-input sample rate ratio number is split into the
integer part and the fractional part. The integer part is three bits
wide, and the fractional part is 12 bits wide. The total available
range is 1:8 to 8:1.
The ASRC is bidirectional and converts the sample rate at the
port side to the DSP side and vice versa. The DSP side of the
ASRC always works at the core sample rate and should be the
highest sample rate of all the ports.
Register 0x4A to Register 0x4F can be used to set the ASRC ratio
manually for ASRCA, ASRCB, and ASRCC. The following
section explains the manual setting of the sample rate
conversion ratio.
Manual Setting of Sample Rate Conversion Ratio
Bits[6:4] in Register 0x4A contain the SRCAINT bits for the
integer portion; and the SRCARFRE_HI bits (Bits[3:0]), along
with the SRCAFRE_LOW bits (Bits[7:0] in Register 0x4B), form
the total 12-bit fractional portion.
Bits[6:4] in Register 0x4C contain the SRCBINT bits for the
integer portion; and the SRCBRFRE_HI bits (Bits[3:0]), along
with the SRCBRFRE_LOW bits (Bits[7:0] in Register 0x4D),
form the total 12-bit fractional portion.
Bits[6:4] in Register 0x4E contain the SRCCINT bits for the
integer portion; and the SRCCRFRE_HI bits (Bits[3:0]), along
with the SRCCRFRE_LOW bits (Bits[7:0] in Register 0x4F),
form the total 12-bit fractional portion.
The ratio can be calculated using the following steps:
1.
2.
3.
4.
5.
Calculate the ratio.
Split the ratio number into integer and fractional parts.
Set M as the integer part of f
and set N as the fractional part.
Integer Part M can be set from 1 to 8 using Register 0x4A,
Register 0x4C, and Register 0x4E, Bits[6:4] (SRCxINT),
for ASRCA, ASRCB, and ASRCC, respectively.
Round Fractional Part N, using the following equation:
ROUND (N × 2
Convert the number to hexadecimal format.
12
) = ROUND (N × 4096)
S_DSP
/f
S_x_INT
or f
S_DSP
/f
S_x_EXT
,
Rev. 0 | Page 55 of 296
The fractional part is 12 bits wide with an upper nibble
(Bits[11:8]) and a lower byte (Bits[7:0]). The upper bits are set
using Register 0x4A, Register 0x4C, and Register 0x4E, Bits[3:0]
(SRCxRFRE_HI) and the lower byte is set using Register 0x4B,
Register 0x4D, and Register 0x4F, Bits[7:0] (SRCxRFRE_LOW)
for Digital Audio Interface A, Digital Audio Interface B, and
Digital Audio Interface C, respectively.
Example A
If the target sample rate is 48 kHz and the source sample rate is
44.1 kHz, then
Separate the integer portion, which is 1, and the fractional
portion, which is 0.088435.
For the integer value of the ratio in hexadecimal format, set
SRCxINT, Bits[2:0] = 0x1 (hexadecimal).
Next, to enter the fractional value, first convert the number to
a 12-bit integer and then to hexadecimal format.
The upper nibble is 0x1 (hexadecimal), whereas the lower byte
is 0x6A (hexadecimal).
That is, SRCxRFRE_HI, Bits[3:0] = 0x1, and SRCxRFRE_LOW,
Bits[7:0] = 0x6A.
Example B
If the target sample rate is 44.1 kHz and the source sample rate
is 8 kHz, then
Separate the integer portion, which is 5, and the fractional
portion, which is 0.5125.
For the integer value of the ratio in hexadecimal format, set
SRCxINT, Bits[2:0] = 0x5 (hexadecimal).
Next, to enter the fractional value, first convert the number to
a 12-bit integer and then to hexadecimal format.
The upper nibble is 0x0 (hexadecimal), whereas the lower byte
is 0x83 (hexadecimal).That is, SRCxRFRE_HI, Bits[3:0] = 0x0,
and SRCxRFRE_LOW, Bits[7:0] = 0x83.
48/44.1 = 1.088435
0.088435 × 4096 = 362.231 ≈ 362 = 0x16A (hex)
44.1/8 = 5.5125
0.5125 × 4096 = 2099.2 ≈ 131 = 0x83 (hex)
ADAU1373