AD AD1895

a
FEATURES
Automatically Senses Sample Frequencies
No Programming Required
Attenuates Sample Clock Jitter
3.3 V–5 V Input and 3.3 V Core Supply Voltages
Accepts 16-/18-/20-/24-Bit Data
Up to 192 kHz Sample Rate
Input/Output Sample Ratios from 7.75:1 to 1:8
Bypass Mode
Multiple AD1895 TDM Daisy-Chain Mode
128 dB Signal-to-Noise and Dynamic Range
(A-Weighted, 20 Hz–20 kHz BW)
Up to –122 dB THD + N
Linear Phase FIR Filter
Hardware Controllable Soft Mute
Supports 256 fS, 512 fS or 768 fS Master Mode
Clock
Flexible Three-Wire Serial Data Port with Left-Justified,
I2S, Right-Justified (16-, 18-, 20-, 24-Bits), and TDM
Serial Port Modes
Master/Slave Input and Output Modes
28-Lead SSOP Plastic Package
APPLICATIONS
Home Theater Systems, Automotive Audio Systems,
DVD, DVD-R, CD-R, Set-Top Boxes, Digital Audio
Effects Processors
PRODUCT OVERVIEW
The AD1895 is a 24-bit, high-performance, single-chip, secondgeneration asynchronous sample rate converter. Based upon
Analog Devices, Inc. experience with its first asynchronous
sample rate converter, the AD1890, the AD1895 offers improved
performance and additional features. This improved performance
includes a THD + N range of –115 dB to –122 dB depending
on sample rate and input frequency, 128 dB (A-Weighted)
dynamic range, 192 kHz sampling frequencies for both input and
output sample rates, improved jitter rejection, and 1:8 upsampling
and 7.75:1 downsampling ratios. Additional features include
more serial formats, a bypass mode, and better interfacing to
digital signal processors.
The AD1895 has a 3-wire interface for the serial input and
output ports that supports left-justified, I2S, and right-justified
(16-, 18-, 20-, 24-bit) modes. Additionally, the serial output
port supports TDM mode for daisy chaining multiple AD1895s to
192 kHz Stereo Asynchronous
Sample Rate Converter
AD1895*
FUNCTIONAL BLOCK DIAGRAM
RESET
VDD_IO VDD_CORE
AD1895
MUTE_I
SDATA_I
SCLK_I
LRCLK_I
SMODE_IN_0
SMODE_IN_1
SMODE_IN_2
FSOUT
FSIN
FIFO
SDATA_O
SCLK_O
LRCLK_O
SERIAL
INPUT
TDM_IN
DIGITAL
PLL
BYPASS
FIR
FILTER
SERIAL
OUTPUT
SMODE_O_0
SMODE_O_1
MUTE_O
CLOCK DIVIDER
MCLK_I
MSMODE_0
MCLK_O
ROM
WLNGTH_O_0
WLNGTH_O_1
MSMODE_2
MSMODE_1
a digital signal processor. The serial output data is dithered down
to 20, 18 or 16 bits when 20-, 18- or 16-bit output data is selected.
The AD1895 sample rate converts the data from the serial input
port to the sample rate of the serial output port. The sample rate
at the serial input port can be asynchronous with respect to the
output sample rate of the output serial port. The master clock to
the AD1895, MCLK, can be asynchronous to both the serial
input and output ports.
MCLK can either be generated off-chip or on-chip by the AD1895
master clock oscillator. Since MCLK can be asynchronous to the
input or output serial ports, a crystal can be used to generate
MCLK internally to reduce noise and EMI emissions on the
board. When MCLK is synchronous to either the output or input
serial port, the AD1895 can be configured in a master mode where
MCLK is divided down and used to generate the left/right
and bit clocks for the serial port that is synchronous to MCLK.
The AD1895 supports master modes of 256 × fS, 512 × fS, and
768 × fS for both input and output serial ports.
Conceptually, the AD1895 interpolates the serial input data by
a rate of 220 and samples the interpolated data stream by the
output sample rate. In practice, a 64-tap FIR filter with 220
polyphases, a FIFO, a digital servo loop that measures the time
difference between input and output samples within 5 ps, and a
digital circuit to track the sample rate ratio are used to perform
the interpolation and output sampling. Refer to the Theory of
Operation section. The digital servo loop and sample rate ratio
circuit automatically track the input and output sample rates.
(Continued on page 15)
*Patents pending.
REV. A
Information furnished by Analog Devices is believed to be accurate and
reliable. However, no responsibility is assumed by Analog Devices for its
use, nor for any infringements of patents or other rights of third parties that
may result from its use. No license is granted by implication or otherwise
under any patent or patent rights of Analog Devices.
One Technology Way, P.O. Box 9106, Norwood, MA 02062-9106, U.S.A.
Tel: 781/329-4700
www.analog.com
Fax: 781/326-8703
© Analog Devices, Inc., 2001
AD1895–SPECIFICATIONS
TEST CONDITIONS UNLESS OTHERWISE NOTED
Supply Voltages
VDD_CORE
VDD_IO
Ambient Temperature
Input Clock
Input Signal
Measurement Bandwidth
Word Width
Load Capacitance
Input Voltage HI
Input Voltage LO
3.3 V
5.0 V or 3.3 V
25°C
30.0 MHz
1.000 kHz, 0 dBFS
20 to fS_OUT/2 Hz
24 Bits
50 pF
2.4 V
0.8 V
Specifications subject to change without notice.
DIGITAL PERFORMANCE (VDD_CORE = 3.3 V 5%, VDD_IO = 5.0 V 10%)
Parameter
Min
Typ
Max
Resolution
Sample Rate @ MCLK_I = 30 MHz
Sample Rate (@ Other Master Clocks)1
Sample Rate Ratios
Upsampling
Downsampling
Dynamic Range2
(20 Hz to fS_OUT/2, 1 kHz, –60 dBFS Input) A-Weighted
44.1 kHz: 48 kHz
48 kHz: 44.1 kHz
48 kHz: 96 kHz
44.1 kHz: 192 kHz
96 kHz: 48 kHz
192 kHz: 32 kHz
(20 Hz to fS_OUT/2, 1 kHz, –60 dBFS Input) No Filter
44.1 kHz: 48 kHz
48 kHz: 44.1 kHz
48 kHz: 96 kHz
44.1 kHz: 192 kHz
96 kHz: 48 kHz
192 kHz: 32 kHz
Total Harmonic Distortion + Noise2
(20 Hz to fS_OUT/2, 1 kHz, 0 dBFS Input) No Filter
Worst-Case (48 kHz:96 kHz)3
44.1 kHz: 48 kHz
48 kHz: 44.1 kHz
48 kHz: 96 kHz
44.1 kHz: 192 kHz
96 kHz: 48 kHz
192 kHz: 32 kHz
Interchannel Gain Mismatch
Interchannel Phase Deviation
Mute Attenuation (24 Bits Word Width)
24
6
215
MCLK_I/5000 ≤ fS_OUT ≤ MCLK_I/138
Unit
Bits
kHz
kHz
1:8
7.75:1
128
128
128
128
127
127
dB
dB
dB
dB
dB
dB
125
125
125
125
124
124
dB
dB
dB
dB
dB
dB
–120
–119
–118
–120
–122
–122
0.0
0.0
–127
dB
dB
dB
dB
dB
dB
dB
dB
Degrees
dB
–115
NOTES
1
Lower sampling rates than given by this formula are possible, but the jitter rejection will decrease.
2
Refer to the Typical Performance Characteristics section for DNR and THD+N numbers over wide range of Input and Output Sample Rates.
3
For any other ratio, minimum THD+N will be better than –115 dB. Please refer to detailed performance plots.
Specifications subject to change without notice.
–2–
REV. A
AD1895
DIGITAL TIMING (–40C < TA < +105C, VDD_CORE = 3.3 V 5%, VDD_IO = 5.0 V 10%)
Parameter1
Min
Max
MCLK_I Period
MCLK_I Frequency
MCLK_I Pulsewidth High
MCLK_I Pulsewidth Low
8
12
ns
MHz
ns
ns
Input Serial Port Timing
tLRIS
LRCLK_I Setup to SCLK_I
tSIH
SCLK_I Pulsewidth High
tSIL
SCLK_I Pulsewidth Low
SDATA_I Setup to SCLK_I Rising Edge
tDIS
tDIH
SDATA_I Hold from SCLK_I Rising Edge
8
8
8
8
3
ns
ns
ns
ns
ns
3
3
ns
ns
ns
ns
ns
ns
ns
ns
ns
tMCLKI
fMCLK
tMPWH
tMPWL
33.3
Unit
30.02, 3
Output Serial Port Timing
TDM_IN Setup to SCLK_O Falling Edge
tTDMS
tTDMH
TDM_IN Hold from SCLK_O Falling Edge
SDATA_O Propagation Delay from SCLK_O, LRCLK_O
tDOPD
SDATA_O Hold from SCLK_O
tDOH
tLROS
LRCLK_O Setup to SCLK_O (TDM Mode Only)
LRCLK_O Hold from SCLK_O (TDM Mode Only)
tLROH
tSOH
SCLK_O Pulsewidth High
tSOL
SCLK_O Pulsewidth Low
tRSTL
RESET Pulsewidth LO
20
3
5
3
10
5
200
NOTES
1
Refer to Timing Diagram Section.
2
The maximum possible sample rate is: FSMAX = fMCLK /138.
3
fMCLK of up to 34 MHz is possible under the following conditions: 0°C < TA < 70°C, 45/55 or better MCLK_I duty cycle.
Specifications subject to change without notice.
TIMING DIAGRAMS
MCLK I
LRCLK_I
t SIH
t LRIS
RESET
SCLK I
t DIS
t RSTL
t SIL
Figure 2. RESET Timing
SDATA I
t DIH
LRCLK O
t MPWH
tSOH
SCLK O
tSOL
tDOPD
t MPWL
SDATA O
Figure 3. MCLK_I Timing
tDOH
tLROS
LRCLK O
tLROH
SCLK O
tTDMS
TDM IN
tTDMH
Figure 1. Input and Output Serial Port Timing (SCLK I/O,
LRCLK I/O, SDATA I/O, TDM_IN)
REV. A
–3–
AD1895–SPECIFICATIONS
DIGITAL FILTERS (VDD_CORE = 3.3 V 5%, VDD_IO = 5.0 V 10%)
Parameter
Passband
Passband Ripple
Transition Band
Stop Band
Stop Band Attenuation
Group Delay
Min
Typ
0.4535 fS_OUT
0.5465 fS_OUT
Max
Unit
0.4535 fS_OUT
± 0.016
0.5465 fS_OUT
Hz
dB
Hz
Hz
dB
Max
Unit
0.8
2
–2
10
V
µA
µA
pF
V
V
mA
mA
–125
Refer to the Group Delay Equations Section
Specifications subject to change without notice.
DIGITAL I/O CHARACTERISTICS (VDD_CORE = 3.3 V 5%, VDD_IO = 5.0 V 10%)
Parameter
Min
Input Voltage HI (VIH)
Input Voltage LO (VIL)
Input Leakage (IIH @ VIH = 5 V)
Input Leakage (IIL @ VIL = 0 V)
Input Capacitance
Output Voltage HI (VOH @ IOH = –4 mA)
Output Voltage LO (VOL @ IOL = +4 mA)
Output Source Current HI (IOH)
Output Sink Current LO (IOL)
2.4
Typ
VDD_CORE – 0.5
5
VDD_CORE – 0.4
0.2
0.5
–4
+4
Specifications subject to change without notice.
POWER SUPPLIES
Parameter
Supply Voltage
VDD_CORE
VDD_IO*
Active Supply Current
I_ CORE_ACTIVE
48 kHz: 48 kHz
96 kHz: 96 kHz
192 kHz: 192 kHz
I_IO_ACTIVE
Power-Down Supply Current: (All Clocks Stopped)
I_CORE_PWRDN
I_IO_PWRDN
Min
Typ
Max
Unit
3.135
VDD_CORE
3.3
3.3/5.0
3.465
5.5
V
V
20
26
43
2
mA
mA
mA
mA
0.5
10
mA
µA
*For 3.3 V tolerant Inputs, VDD_IO supply should be set to 3.3 V; however, VDD_CORE supply voltage should not exceed VDD_IO.
Specifications subject to change without notice.
–4–
REV. A
AD1895
POWER SUPPLIES (VDD_CORE = 3.3 V 5%, VDD_IO = 5.0 V 10%)
Parameter
Min
Typ
Total Active Power Dissipation
48 kHz: 48 kHz
96 kHz: 96 kHz
192 kHz: 192 kHz
Total Power Down Dissipation: (RESET LO)
Max
Unit
65
85
132
2
mW
mW
mW
mW
Specifications subject to change without notice.
TEMPERATURE RANGE
Parameter
Min
Specifications Guaranteed
Functionality Guaranteed
Storage
Thermal Resistance, θJA (Junction-to-Ambient)
–40
–55
Typ
Max
Unit
+105
+150
°C
°C
°C
°C/W
Min
Max
Unit
–0.3
–0.3
+3.6
+6.0
V
V
DGND – 0.3
–40
± 10
VDD_IO + 0.3
+105
mA
V
°C
25
109
Specifications subject to change without notice.
ABSOLUTE MAXIMUM RATINGS*
Parameter
Power Supplies
VDD_CORE
VDD_IO
Digital Inputs
Input Current
Input Voltage
Ambient Temperature (Operating)
*Stresses greater than those listed under Absolute Maximum Ratings may cause permanent damage to the device. This is a stress rating only; functional operation of the
device at these or any other conditions above those indicated in the operational section of this specification is not implied. Exposure to absolute maximum rating conditions
for extended periods may affect device reliability.
ORDERING GUIDE
Model
Temperature Range
Package Description
Package Option
AD1895YRS
AD1895YRSRL
–40°C to +105°C
–40°C to +105°C
28-Lead SSOP
28-Lead SSOP
RS-28
RS-28 on 13" Reel
CAUTION
ESD (electrostatic discharge) sensitive device. Electrostatic charges as high as 4000 V readily
accumulate on the human body and test equipment and can discharge without detection. Although
the AD1895 features proprietary ESD protection circuitry, permanent damage may occur on
devices subjected to high-energy electrostatic discharges. Therefore, proper ESD precautions are
recommended to avoid performance degradation or loss of functionality.
REV. A
–5–
WARNING!
ESD SENSITIVE DEVICE
AD1895
PIN FUNCTION DESCRIPTIONS
Pin No.
IN/OUT
Mnemonic
Description
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
IN
IN
OUT
IN
IN/OUT
IN/OUT
IN
IN
IN
IN
IN
IN
IN
IN
OUT
IN
IN
IN
IN
IN
IN
IN
OUT
IN/OUT
IN/OUT
IN
IN
IN
NC
MCLK_IN
MCLK_OUT
SDATA_I
SCLK_I
LRCLK_I
VDD_IO
DGND
BYPASS
SMODE_IN_0
SMODE_IN_1
SMODE_IN_2
RESET
MUTE_IN
MUTE_OUT
WLNGTH_OUT_1
WLNGTH_OUT_0
SMODE_OUT_1
SMODE_OUT_0
TDM_IN
DGND
VDD_CORE
SDATA_O
LRCLK_O
SCLK_O
MMODE_0
MMODE_1
MMODE_2
No Connect
Master Clock or Crystal Input
Master Clock Output or Crystal Output
Input Serial Data (at Input Sample Rate)
Master/Slave Input Serial Bit Clock
Master/Slave Input Left/Right Clock
3.3 V/5 V Input/Output Digital Supply Pin
Digital Ground Pin
ASRC Bypass Mode, Active High
Input Port Serial Interface Mode Select Pin 0
Input Port Serial Interface Mode Select Pin 1
Input Port Serial Interface Mode Select Pin 2
Reset Pin, Active Low
Mute Input Pin— Active HI Normally Connected to MUTE_OUT
Output Mute Control – Active HI
Hardware Selectable Output Wordlength—Select Pin 1
Hardware Selectable Output Wordlength—Select Pin 0
Output Port Serial Interface Mode Select Pin 1
Output Port Serial Interface Mode Select Pin 0
Serial Data Input* (Only for Daisy-Chain Mode). Ground when not used.
Digital Ground Pin
3.3 V Digital Supply Pin
Output Serial Data (at Output Sample Rate)
Master/Slave Output Left/Right Clock
Master/Slave Output Serial Bit Clock
Master/Slave Clock Ratio Mode Select Pin 0
Master/Slave Clock Ratio Mode Select Pin 1
Master/Slave Clock Ratio Mode Select Pin 2
*Also used to input matched-phase mode data.
PIN CONFIGURATION
NC 1
28 MMODE_2
MCLK_IN 2
27 MMODE_1
MCLK_OUT 3
26 MMODE_0
SDATA_I 4
SCLK_I 5
LRCLK_I 6
AD1895
25 SCLK_O
TOP VIEW
24 LRCLK_O
(NOT TO SCALE)
23 SDATA_O
VDD_IO 7
22 VDD_CORE
DGND 8
21 DGND
BYPASS 9
20 TDM_IN
SMODE_IN_0 10
19 SMODE_OUT_0
SMODE_IN_1 11
18 SMODE_OUT_1
SMODE_IN_2 12
17 WLNGTH_OUT_0
RESET 13
16 WLNGTH_OUT_1
MUTE_IN 14
15 MUTE_OUT
NC = NO CONNECT
–6–
REV. A
0
0
–20
–20
–40
–40
–60
–60
–80
–80
dBFS
dBFS
Typical Performance Characteristics–AD1895
–100
–120
–140
–140
–160
–160
–180
–180
–200
0
2.5
5.0
7.5
10.0 12.5 15.0
FREQUENCY – kHz
–200
17.5
20.0
22.5
0
0
0
–20
–20
–40
–40
–60
–60
–80
–80
–100
–120
–140
–140
–160
–160
–180
–180
–200
–200
2.5
5.0
7.5
10.0 12.5 15.0
FREQUENCY – kHz
17.5
20.0
22.5
TPC 2. Wideband FFT Plot (16k Points) 0 dBFS 1 kHz Tone,
44.1 kHz:48 kHz (Asynchronous)
30
0
–20
–40
–40
–60
–60
–80
–80
dBFS
0
–20
–120
–140
–140
–160
–160
–180
–180
0
2.5
5.0
7.5
10.0
12.5 15.0
FREQUENCY – kHz
–200
–200
5
10
15
20
25
30
FREQUENCY – kHz
35
40
70
80
90
17.5
20.0
–100
–120
0
40
50
60
FREQUENCY – kHz
TPC 5. Wideband FFT Plot (16k Points) 48 kHz:44.1 kHz,
0 dBFS 1 kHz Tone
–100
0
45
TPC 3. Wideband FFT Plot (16k Points) 48 kHz:96 kHz,
0 dBFS 1 kHz Tone
REV. A
20
–100
–120
0
10
TPC 4. Wideband FFT Plot (16k Points) 44.1 kHz:192 kHz,
0 dBFS 1 kHz Tone
dBFS
dBFS
TPC 1. Wideband FFT Plot (16k Points) 0 dBFS 1 kHz Tone,
48 kHz:48 kHz (Asynchronous)
dBFS
–100
–120
2.5
5.0
7.5
10.0 12.5 15.0
FREQUENCY – kHz
17.5
20.0
22.5
TPC 6. Wideband FFT Plot (16k Points) 96 kHz:48 kHz,
0 dBFS 1 kHz Tone
–7–
AD1895
0
–50
–60
–20
–70
–40
–80
–90
–60
–100
–110
dBFS
dBFS
–80
–100
–120
–120
–130
–140
–150
–140
–160
–160
–170
–180
–180
–190
–200
0
2.5
5.0
7.5
10.0 12.5 15.0
FREQUENCY – kHz
–200
17.5
20.0
22.5
0
TPC 7. Wideband FFT Plot (16k Points) 192 kHz:48 kHz,
0 dBFS 1 kHz Tone
15
20
25
30
FREQUENCY – kHz
35
40
45
–50
–60
–60
–70
–70
–80
–80
–90
–90
–100
–110
–110
–100
–120
dBFS
dBFS
10
TPC 10. Wideband FFT Plot (16k Points) 48 kHz:96 kHz,
–60 dBFS 1 kHz Tone
–50
–130
–120
–130
–140
–140
–150
–150
–160
–160
–170
–170
–180
–180
–190
–190
–200
–200
0
2.5
5.0
7.5
10.0 12.5 15.0
FREQUENCY – kHz
17.5
20.0
22.5
0
TPC 8. Wideband FFT Plot (16k Points) –60 dBFS 1 kHz
Tone, 48 kHz:48 kHz (Asynchronous)
10
20
30
40
50
60
FREQUENCY – kHz
70
80
90
TPC 11. Wideband FFT Plot (16k Points) 44.1 kHz:192 kHz,
–60 dBFS 1 kHz Tone
–50
–50
–60
–60
–70
–70
–80
–80
–90
–90
–100
–110
–100
–110
dBFS
dBFS
5
–120
–130
–140
–120
–130
–140
–150
–150
–160
–160
–170
–170
–180
–180
–190
–190
–200
–200
0
2.5
5.0
7.5
10.0 12.5 15.0
FREQUENCY – kHz
17.5
20.0
0
22.5
2.5
5.0
7.5
10.0 12.5 15.0
FREQUENCY – kHz
17.5
20.0
TPC 12. Wideband FFT Plot (16k Points) 48 kHz:44.1 kHz,
–60 dBFS 1 kHz Tone
TPC 9. Wideband FFT Plot (16k Points) 44.1 kHz:48 kHz,
–60 dBFS 1 kHz Tone
–8–
REV. A
AD1895
–50
0
–60
–20
–70
–40
–80
–90
–60
–100
–80
dBFS
dBFS
–110
–120
–130
–150
–140
–160
–160
–170
–180
–180
–190
–200
–200
0
2.5
5.0
7.5
10.0 12.5 15.0
FREQUENCY – kHz
17.5
20.0
0
22.5
TPC 13. Wideband FFT Plot (16k Points) 96 kHz:48 kHz,
–60 dBFS 1 kHz Tone
7.5
10.0 12.5 15.0
FREQUENCY – kHz
17.5
20.0
22.5
–20
–70
–80
–40
–90
–60
–100
–80
dBFS
–110
dBFS
5.0
0
–60
–120
–130
–140
–100
–120
–150
–140
–160
–160
–170
–180
–180
–190
–200
0
2.5
5.0
7.5
10.0 12.5 15.0
FREQUENCY – kHz
17.5
20.0
–200
22.5
TPC 14. Wideband FFT Plot (16k Points) 192 kHz:48 kHz,
–60 dBFS 1 kHz Tone
0
2.5
5.0
7.5
10.0 12.5 15.0
FREQUENCY – kHz
17.5
20.0
TPC 17. IMD, 10 kHz and 11 kHz 0 dBFS Tone
48 kHz:44.1 kHz
0
0
–20
–20
–40
–40
–60
–60
–80
dBFS
–80
–100
–100
–120
–120
–140
–140
–160
–160
–180
–180
–200
–200
0
2.5
5.0
7.5
10.0 12.5 15.0
FREQUENCY – kHz
17.5
20.0
22.5
0
TPC 15. IMD, 10 kHz and 11 kHz 0 dBFS Tone
44:1 kHz:48 kHz
REV. A
2.5
TPC 16. IMD, 10 kHz and 11 kHz 0 dBFS Tone
96 kHz:48 kHz
–50
dBFS
–100
–120
–140
2.5
5.0
7.5
10.0 12.5 15.0
FREQUENCY – kHz
17.5
20.0
22.5
TPC 18. Wideband FFT Plot (16k Points) 44.1 kHz:48 kHz,
0 dBFS 20 kHz Tone
–9–
0
0
–20
–20
–40
–40
–60
–60
–80
–80
dBFS
dBFS
AD1895
–100
–120
–140
–140
–160
–160
–180
–180
–200
–200
0
10
20
30
40
50
60
FREQUENCY – kHz
70
80
0
90
TPC 19. Wideband FFT Plot (16k Points) 192 kHz:192 kHz,
0 dBFS 80 kHz Tone
0
0
–20
–20
–40
–40
–60
–60
–80
–80
–100
–120
–140
–140
–160
–160
–180
–180
–200
–200
2.5
5.0
7.5
10.0 12.5 15.0
FREQUENCY – kHz
17.5
20.0
10
15
20
25
30
FREQUENCY – kHz
35
40
45
–100
–120
0
5
TPC 22. Wideband FFT Plot (16k Points) 48 kHz:96 kHz,
0 dBFS 20 kHz Tone
dBFS
dBFS
–100
–120
22.5
0
TPC 20. Wideband FFT Plot (16k Points) 48 kHz:48 kHz,
0 dBFS 20 kHz Tone
2.5
5.0
7.5
10.0 12.5 15.0
FREQUENCY – kHz
17.5
20.0
22.5
TPC 23. Wideband FFT Plot (16k Points) 96 kHz:48 kHz,
0 dBFS 20 kHz Tone
0
–119
–20
–121
–60
–123
THD + N – dBFS
–40
dBFS
–80
–100
–120
–125
–127
–129
–140
–131
–160
–133
–180
–200
0
2.5
5.0
7.5
10.0 12.5 15.0
FREQUENCY – kHz
17.5
–135
30000
20.0
TPC 21. Wideband FFT Plot (16k Points) 48 kHz:44:1 kHz,
0 dBFS 20 kHz Tone
55000
80000 105000 130000 155000
OUTPUT SAMPLE RATE – Hz
180000
TPC 24. THD + N vs. Output Sample Rate, fS_IN = 192 kHz,
0 dBFS 1 kHz Tone
–10–
REV. A
–119
–119
–121
–121
–123
–123
THD + N – dBFS
THD + N – dBFS
AD1895
–125
–127
–129
–127
–129
–131
–131
–133
–133
–135
30000
55000
80000 105000 130000 155000
OUTPUT SAMPLE RATE – Hz
–135
30000
180000
TPC 25. THD + N vs. Output Sample Rate, fS_IN = 48 kHz,
0 dBFS 1 kHz Tone
–119
–119
–121
–121
–123
–123
–125
–127
–131
–131
–133
–133
80000 105000 130000 155000
OUTPUT SAMPLE RATE – Hz
180000
–127
–129
55000
80000 105000 130000 155000
OUTPUT SAMPLE RATE – Hz
–125
–129
–135
30000
55000
TPC 28. THD + N vs. Output Sample Rate, fS_IN = 96 kHz,
0 dBFS 1 kHz Tone
DNR – dBFS
THD + N – dBFS
–125
–135
30000
180000
TPC 26. THD + N vs. Output Sample Rate, fS_IN =
44.1 kHz, 0 dBFS 1 kHz Tone
55000
80000 105000 130000 155000
OUTPUT SAMPLE RATE – Hz
180000
TPC 29. DNR vs. Output Sample Rate, fS_IN = 192 kHz,
–60 dBFS 1 kHz Tone
–119
–119
–121
–121
–123
DNR – dBFS
THD + N – dBFS
–123
–125
–127
–129
–125
–127
–129
–131
–131
–133
–133
–135
30000
55000
–135
30000
80000 105000 130000 155000 180000
OUTPUT SAMPLE RATE – Hz
80000 105000 130000 155000
OUTPUT SAMPLE RATE – Hz
180000
TPC 30. DNR vs. Output Sample Rate, fS_IN = 32 kHz,
–60 dBFS 1 kHz Tone
TPC 27. THD + N vs. Output Sample Rate, fS_IN = 32 kHz,
0 dBFS 1 kHz Tone
REV. A
55000
–11–
–119
–119
–121
–121
–123
–123
DNR – dBFS
DNR – dBFS
AD1895
–125
–127
–125
–127
–129
–129
–131
–131
–133
–133
–135
30000
55000
80000 105000 130000 155000
OUTPUT SAMPLE RATE – Hz
–135
30000
180000
55000
80000 105000 130000 155000
OUTPUT SAMPLE RATE – Hz
180000
TPC 34. DNR vs. Output Sample Rate, fS_IN = 44.1 kHz,
–60 dBFS 1 kHz Tone
TPC 31. DNR vs. Output Sample Rate, fS_IN = 96 kHz,
–60 dBFS 1 kHz Tone
0.00
5
–0.01
–15
–0.02
192 kHz:96 kHz
–0.03
–0.04
192 kHz:48 kHz
–55
dBFS
dBFS
–35
–75
–0.05
–0.06
192 kHz:32 kHz
–0.07
–95
–0.08
–115
–0.09
–0.10
–135
0
10000
20000
30000
40000
FREQUENCY – Hz
50000
0
60000
4000
8000
12000
16000
FREQUENCY – Hz
20000
24000
TPC 35. Passband Ripple, 192 kHz:48 kHz
TPC 32. Digital Filter Frequency Response
5
–119
4
–121
LINEARITY ERROR – dBr
3
DNR – dBFS
–123
–125
–127
–129
–131
2
1
0
–1
–2
–3
–133
–135
30000
–4
55000
80000 105000 130000 155000
OUTPUT SAMPLE RATE – Hz
–5
–140
180000
–120
–100
–80
–60
–40
INPUT LEVEL – dBFS
–20
0
TPC 36. Linearity Error, 48 kHz:48 kHz, 0 dBFS to
–140 dBFS Input, 200 Hz Tone
TPC 33. DNR vs. Output Sample Rate, fS_IN = 48 kHz,
–60 dBFS 1 kHz Tone
–12–
REV. A
5
5
4
4
3
3
LINEARITY ERROR – dBr
LINEARITY ERROR – dBr
AD1895
2
1
0
–1
–2
–2
–4
–4
–120
–100
–80
–60
–40
INPUT LEVEL – dBFS
–20
–5
–140
0
5
5
4
4
3
3
2
1
0
–1
–2
–20
–2
–5
–140
0
5
4
3
3
LINEARITY ERROR – dBr
5
4
2
1
0
–1
–2
0
0
–2
–4
–20
–20
–1
–4
–80
–60
–40
INPUT LEVEL – dBFS
–80
–60
–40
INPUT LEVEL – dBFS
1
–3
–100
–100
2
–3
–120
–120
TPC 41. Linearity Error, 44.1 kHz:192 kHz, 0 dBFS to
–140 dBFS Input, 200 Hz Tone
TPC 38. Linearity Error, 96 kHz:48 kHz, 0 dBFS to
–140 dBFS Input, 200 Hz Tone
–5
–140
0
–1
–4
–80
–60
–40
INPUT LEVEL – dBFS
–20
0
–3
–100
–80
–60
–40
INPUT LEVEL – dBFS
2
–4
–120
–100
1
–3
–5
–140
–120
TPC 40. Linearity Error, 48 kHz:96 kHz, 0 dBFS to
–140 dBFS Input, 200 Hz Tone
LINEARITY ERROR – dBr
LINEARITY ERROR – dBr
0
–1
–3
TPC 37. Linearity Error, 48 kHz:44.1 kHz, 0 dBFS to
–140 dBFS Input, 200 Hz Tone
LINEARITY ERROR – dBr
1
–3
–5
–140
–5
–140
0
–120
–100
–80
–60
–40
INPUT LEVEL – dBFS
–20
0
TPC 42. Linearity Error, 192 kHz:44:1 kHz, 0 dBFS to
–140 dBFS Input, 200 Hz Tone
TPC 39. Linearity Error, 44.1 kHz:48 kHz, 0 dBFS to
–140 dBFS Input, 200 Hz Tone
REV. A
2
–13–
AD1895
–110
–112
–112
–114
–114
–116
–116
–118
–118
–120
–120
–122
–122
–124
–124
dBr
dBr
–110
–126
–128
–130
–130
–132
–132
–134
–134
–136
–136
–138
–138
–140
–140
–120
–100
–80
–60
–40
INPUT LEVEL – dBFS
–20
–140
–140
0
TPC 43. THD + N vs. Input Amplitude, 48 kHz:44.1 kHz,
1 kHz Tone
–114
–116
–116
–118
–118
–120
–120
–122
–122
–124
–124
dBr
dBr
–112
–114
–126
–80
–60
–40
INPUT LEVEL – dBFS
–20
0
–126
–128
–128
–130
–130
–132
–132
–134
–134
–136
–136
–138
–138
–120
–100
–80
–60
–40
INPUT LEVEL – dBFS
–20
–140
–140
0
TPC 44. THD + N vs. Input Amplitude, 96 kHz:48 kHz,
1 kHz Tone
–120
–100
–80
–60
–40
INPUT LEVEL – dBFS
–20
0
TPC 47. THD + N vs. Input Amplitude, 44.1 kHz:192 kHz,
1 kHz Tone
–110
–110
–112
–112
–114
–114
–116
–116
–118
–118
–120
–120
–122
–124
–124
dBr
–122
–126
–126
–128
–128
–130
–130
–132
–132
–134
–134
–136
–136
–138
–138
–140
–140
–100
–110
–112
–140
–140
–120
TPC 46. THD + N vs. Input Amplitude, 48 kHz:96 kHz,
1 kHz Tone
–110
dBr
–126
–128
–120
–100
–80
–60
–40
INPUT LEVEL – dBFS
–20
–140
–140
0
TPC 45. THD + N vs. Input Amplitude, 44.1 kHz:48 kHz,
1 kHz Tone
–120
–100
–80
–60
–40
INPUT LEVEL – dBFS
–20
0
TPC 48. THD + N vs. Input Amplitude, 192 kHz:48 kHz,
1 kHz Tone
–14–
REV. A
AD1895
–110
–115
–115
–120
–120
–125
–125
–130
–130
–135
–135
–140
–140
dBr
dBr
–110
–145
–150
–155
–155
–160
–160
–165
–165
–170
–170
–175
–175
–180
–180
2.5
5.0
7.5
10.0
12.5
FREQUENCY – kHz
15.0
17.5
20.0
TPC 49. THD + N vs. Frequency Input, 48 kHz:44.1 kHz,
0 dBFS
–115
–120
–120
–125
–125
–130
–130
–135
–140
dBr
–135
–140
–145
–150
7.5
10.0
12.5
FREQUENCY – kHz
15.0
17.5
20.0
–145
–150
–155
–155
–160
–160
–165
–165
–170
–170
–175
–175
2.5
5.0
7.5
10.0
12.5
FREQUENCY – kHz
15.0
17.5
–180
20.0
TPC 50. THD + N vs. Frequency Input, 44.1 kHz:48 kHz,
00 dBFS
(Continued from page 1)
The digital servo loop measures the time difference between
input and output sample rates within 5 ps. This is necessary in
order to select the correct polyphase filter coefficient The digital
servo loop has excellent jitter rejection for both input and output sample rates as well as the master clock. The jitter rejection
begins at less than 1 Hz. This requires a long settling time
whenever RESET is deasserted or when the input or output
sample rate changes. To reduce the settling time, upon deassertion of RESET or a change in a sample rate, the digital servo loop
enters the fast settling mode. When the digital servo loop has
adequately settled in the fast mode, it switches into the normal
or slow settling mode and continues to settle until the time
difference measurement between input and output sample
rates is within 5 ps. During fast mode, the MUTE_OUT signal
REV. A
5.0
–110
–115
–180
2.5
TPC 51. THD + N vs. Frequency Input, 48 kHz:96 kHz,
0 dBFS
–110
dBr
–145
–150
2.5
5.0
7.5
10.0
12.5
FREQUENCY – kHz
15.0
17.5
20.0
TPC 52. THD + N vs. Frequency Input, 96 kHz:48 kHz,
0 dBFS
is asserted high. Normally, the MUTE_OUT is connected to the
MUTE_IN pin. The MUTE_IN signal is used to softly mute
the AD1895 upon assertion and softly unmute the AD1895
when it is deasserted.
The sample rate converter of the AD1895 can be bypassed
altogether using the bypass mode. In bypass mode, the AD1895’s
serial input data is directly passed to the serial output port without any dithering. This is useful for passing through nonaudio
data or when the input and output sample rates are synchronous
to one another and the sample rate ratio is exactly 1 to 1.
The AD1895 is a 3.3 V, 5 V input tolerant part and is available
in a 28-lead SSOP SMD package. The AD1895 is 5 V inputtolerant only when the VDD_IO supply pin is supplied with 5 V.
–15–
AD1895
ASRC FUNCTIONAL OVERVIEW
THEORY OF OPERATION
Asynchronous sample rate conversion is converting data from
one clock source at some sample rate to another clock source at
the same or different sample rate. The simplest approach to
asynchronous sample rate conversion is the use of a zero-order
hold between two samplers shown in Figure 4. In an asynchronous system T2 is never equal to T1 nor is the ratio between T2
and T1 rational. As a result, samples at fS_OUT will be repeated or
dropped producing an error in the resampling process. The
frequency domain shows the wide side lobes that result from
this error when the sampling of fS_OUT is convolved with the
attenuated images from the sin(x)/x nature of the zero-order
hold. The images at fS_IN, dc signal images, of the zero-order hold
are infinitely attenuated. Since the ratio of T2 to T1 is an irrational number, the error resulting from the resampling at fS_OUT
can never be eliminated. However, the error can be significantly reduced through interpolation of the input data at fS_IN.
The AD1895 is conceptually interpolated by a factor of 220.
IN
ZERO-ORDER
HOLD
fS_IN = 1/T1
between each fS_IN sample and convolving this interpolated
signal with a digital low-pass filter to suppress the images. In the
time domain it can be seen that fS_OUT selects the closest fS_IN × 220
sample from the zero-order hold as opposed to the nearest fS_IN
sample in the case of no interpolation. This significantly reduces
the resampling error.
IN
INTERPOLATE
BY N
LOW-PASS
FILTER
OUT
ZERO-ORDER
HOLD
fS_IN
fS_OUT
TIME DOMAIN OF fS_IN SAMPLES
TIME DOMAIN OUTPUT OF THE LOW-PASS FILTER
OUT
TIME DOMAIN OF fS_OUT RESAMPLING
fS_OUT = 1/T2
ORIGINAL SIGNAL
SAMPLED AT fS_IN
TIME DOMAIN OF THE ZERO-ORDER HOLD OUTPUT
SIN(X)/X OF ZERO-ORDER HOLD
Figure 5. Time Domain of the Interpolation and
Resampling
SPECTRUM OF ZERO-ORDER HOLD OUTPUT
In the frequency domain shown in Figure 6 the interpolation
expands the frequency axis of the zero-order hold. The images
from the interpolation can be sufficiently attenuated by a good
low-pass filter. The images from the zero-order hold are now
pushed by a factor of 220 closer to the infinite attenuation point
of the zero-order hold, which is fS_IN × 220. The images at the
zero-order hold are the determining factor for the fidelity of the
output at fS_OUT. The worst-case images can be computed from
the zero-order hold frequency response, maximum image =
sin (π × F/fS_INTERP)/(π × F/fS_INTERP). F is the frequency of the
worst-case image which would be 220 × fS_IN ± fS_IN/2 , and
fS_INTERP is fS_IN × 220.
SPECTRUM OF fS_OUT SAMPLING
2 fS_OUT
fS_OUT
FREQUENCY RESPONSE OF fS_OUT CONVOLVED WITH ZERO-ORDER
HOLD SPECTRUM
Figure 4. Zero-Order Hold Being Used by fS_OUT to
Resample Data from fS_IN
THE CONCEPTUAL HIGH INTERPOLATION MODEL
Interpolation of the input data by a factor of 220 involves placing
(220 –1) samples between each fS_IN sample. Figure 5 shows
both the time domain and the frequency domain of interpolation
by a factor of 220. Conceptually, interpolation by 220 would
involve the steps of zero-stuffing (220 –1) number of samples
The following worst-case images would appear for fS_IN =
192 kHz:
–16–
Image at fS_INTERP – 96 kHz = –125.1 dB
Image at fS_INTERP + 96 kHz = –125.1 dB
REV. A
AD1895
IN
INTERPOLATE
BY N
LOW-PASS
FILTER
ZERO-ORDER
HOLD
fS_IN
OUT
fS_OUT
FREQUENCY DOMAIN OF SAMPLES AT fS_IN
FREQUENCY DOMAIN OF THE INTERPOLATION
fS_IN
220 fS_IN
SIN(X)/X OF ZERO-ORDER HOLD
FREQUENCY DOMAIN OF fS_OUT RESAMPLING
FREQUENCY DOMAIN AFTER
RESAMPLING
220 fS_IN
220 fS_IN
Figure 6. Frequency Domain of the Interpolation and
Resampling
rate, fS_IN, the ROM starting address, input data and the length
of the convolution must be scaled. As the input sample rate
rises over the output sample rate, the antialiasing filter’s cutoff
frequency has to be lowered because the Nyquist frequency of
the output samples is less than the Nyquist frequency of the
input samples. To move the cutoff frequency of the antialiasing
filter, the coefficients are dynamically altered and the length
of the convolution is increased by a factor of (fS_IN/fS_OUT).
This technique is supported by the Fourier transform property
that if f(t) is F(ω), then f(k × t) is F(ω/k). Thus, the range of
decimation is simply limited by the size of the RAM.
THE SAMPLE RATE CONVERTER ARCHITECTURE
The architecture of the sample rate converter is shown in Figure
7. The sample rate converter’s FIFO block adjusts the left and
right input samples and stores them for the FIR filter’s convolution cycle. The fS_IN counter provides the write address to the
FIFO block and the ramp input to the digital servo loop. The
ROM stores the coefficients for the FIR filter convolution and
performs a high-order interpolation between the stored coefficients. The sample rate ratio block measures the sample rate for
dynamically altering the ROM coefficients and scaling of the
FIR filter length as well as the input data. The digital servo loop
automatically tracks the fS_IN and fS_OUT sample rates and provides the RAM and ROM start addresses for the start of the FIR
filter convolution.
HARDWARE MODEL
RIGHT DATA IN
LEFT DATA IN
The output rate of the low-pass filter of Figure 5 would be the
interpolation rate, 220 × 192000 kHz = 201.3 GHz. Sampling at
a rate of 201.3 GHz is clearly impractical, not to mention the
number of taps required to calculate each interpolated sample.
However, since interpolation by 220 involves zero-stuffing 220–1
samples between each fS_IN sample, most of the multiplies in
the low-pass FIR filter are by zero. A further reduction can be
realized by the fact that since only one interpolated sample is
taken at the output at the fS_OUT rate, only one convolution
needs to be performed per fS_OUT period instead of 220 convolutions. A 64-tap FIR filter for each fS_OUT sample is sufficient
to suppress the images caused by the interpolation.
The difficulty with the above approach is that the correct interpolated sample needs to be selected upon the arrival of fS_OUT.
Since there are 220 possible convolutions per fS_OUT period, the
arrival of the fS_OUT clock must be measured with an accuracy
of 1/201.3 GHz = 4.96 ps. Measuring the fS_OUT period with a
clock of 201.3 GHz frequency is clearly impossible; instead,
several coarse measurements of the fS_OUT clock period are made
and averaged over time.
Another difficulty with the above approach is the number of
coefficients required. Since there are 220 possible convolutions
with a 64-tap FIR filter, there need to be 220 polyphase coefficients for each tap, which requires a total of 226 coefficients. To
reduce the number of coefficients in ROM, the AD1895 stores a
small subset of coefficients and performs a high-order interpolation between the stored coefficients. So far the above approach
works for the case of fS_OUT > fS_IN. However, in the case when
the output sample rate, fS_OUT, is less than the input sample
REV. A
FIFO
ROM A
HIGH
ROM B
ORDER
ROM C
INTERP
ROM D
fS_IN
COUNTER
DIGITAL
SERVO LOOP
SAMPLE RATE RATIO
FIR FILTER
fS_IN
fS_OUT
SAMPLE RATE
RATIO
L/R DATA OUT
EXTERNAL
RATIO
Figure 7. Architecture of the Sample Rate Converter
The FIFO receives the left and right input data and adjusts the
amplitude of the data for both the soft muting of the sample rate
converter and the scaling of the input data by the sample rate
ratio before storing the samples in the RAM. The input data
is scaled by the sample rate ratio because as the FIR filter length
of the convolution increases, so does the amplitude of the convolution output. To keep the output of the FIR filter from saturating,
the input data is scaled down by multiplying it by (fS_OUT/fS_IN)
when fS_OUT < fS_IN. The FIFO also scales the input data for
muting and unmuting of the AD1895.
The RAM in the FIFO is 512 words deep for both left and right
channels. A small offset of 16 is added to the write address
provided by the fS_IN counter to prevent the RAM read pointer
from ever overlapping the write address. The maximum decimation rate can be calculated from the RAM word depth as
(512–16)/64 taps = 7.75 and a small offset.
–17–
AD1895
10
0
–10
–20
–30
–40
SLOW MODE
–50
FAST MODE
–60
–70
–80
–90
–100
–110
–120
–130
–140
–150
–160
–170
–180
–190
–200
–210
–220
0.01
0.1
1
10
100
1e3
1e4
1e5
FREQUENCY – Hz
Figure 8. Frequency Response of the Digital Servo Loop. fS_IN is the X-Axis, fS_OUT = 192 kHz, Master Clock
Frequency Is 30 MHz
The digital servo loop is essentially a ramp filter that provides
the initial pointer to the address in RAM and ROM for the start
of the FIR convolution. The RAM pointer is the integer output
of the ramp filter while the ROM is the fractional part. The
digital servo loop must be able to provide excellent rejection of
jitter on the fS_IN and fS_OUT clocks as well as measure the arrival
of the fS_OUT clock within 4.97 ps. The digital servo loop will
also divide the fractional part of the ramp output by the ratio of
fS_IN/fS_OUT for the case when fS_IN > fS_OUT, to dynamically alter
the ROM coefficients.
The digital servo loop is implemented with a multirate filter. To
settle the digital servo loop filter quicker upon start-up or a change
in the sample rate, a “fast mode” was added to the filter. When
the digital servo loop starts up or the sample rate is changed, the
digital servo loop kicks into “fast mode” to adjust and settle
on the new sample rate. Upon sensing the digital servo loop
settling down to some reasonable value, the digital servo loop
will kick into “normal” or “slow mode.” During “fast mode”
the MUTE_OUT signal of the sample rate converter is asserted
to let the user know that they should mute the sample rate
converter to avoid any clicks or pops. The frequency response of
the digital servo loop for “fast mode” and “slow mode” are
shown in Figure 8.
The FIR filter is a 64-tap filter in the case of fS_OUT ≥ fS_IN and is
(fS_IN/fS_OUT) × 64 taps for the case when fS_IN > fS_OUT. The FIR
filter performs its convolution by loading in the starting address
of the RAM address pointer and the ROM address pointer
from the digital servo loop at the start of the fS_OUT period.
The FIR filter then steps through the RAM by decrementing its
address by 1 for each tap, and the ROM pointer increments its
address by the (fS_OUT/fS_IN) × 220 ratio for fS_IN > fS_OUT or 220
for fS_OUT ≥ fS_IN. Once the ROM address rolls over, the convolution is completed. The convolution is performed for both
the left and right channels, and the multiply accumulate circuit
used for the convolution is shared between the channels.
The fS_IN/fS_OUT sample rate ratio circuit is used to dynamically
alter the coefficients in the ROM for the case when fS_IN >
fS_OUT. The ratio is calculated by comparing the output of an
fS_OUT counter to the output of an fS_IN counter. If fS_OUT >
fS_IN, the ratio is held at one. If fS_IN > fS_OUT, the sample rate
ratio is updated if it is different by more than two fS_OUT periods
from the previous fS_OUT to fS_IN comparison. This is done to
provide some hysteresis to prevent the filter length from oscillating and causing distortion.
–18–
REV. A
AD1895
OPERATING FEATURES
RESET and Power Down
Upon RESET, or a change in the sample rate between LRCLK_I
and LRCLK_O, the MUTE_OUT pin will be asserted high. The
MUTE_OUT pin will remain asserted high until the digital
servo loop’s internal fast settling mode has completed. When
the digital servo loop has switched to slow settling mode, the
MUTE_OUT pin will deassert. While MUTE_OUT is asserted,
the MUTE_IN pin should be asserted as well to prevent any
major distortion in the audio output samples.
When RESET is asserted low, the AD1895 will turn off the
master clock input to the AD1895, MCLK_I, initialize all of its
internal registers to their default values, and three-state all of the
I/O pins. While RESET is active low, the AD1895 is consuming
minimum power. For the lowest possible power consumption
while RESET is active low, all of the input pins to the AD1895
should be static.
Master Clock
When RESET is deasserted, the AD1895 begins its initialization
routine where all locations in the FIFO are initialized to zero,
MUTE_OUT is asserted high, and any I/O pins configured as
outputs are enabled. The mute control counter, which controls
the soft mute attenuation of the input samples, is initialized to
maximum attenuation, –127 dB (see Mute Control section).
When asserting RESET and deasserting RESET, the RESET
should be held low for a minimum of 5 MCLK_I cycles. During
power-up the RESET should be held low until the power supplies have stabilized.
Power Supply and Voltage Reference
The AD1895 is designed for three-volt operation with five-volt
input tolerance on the input pins. VDD_CORE is the three-volt
supply that is used to power the core logic of the AD1895 and
to drive the output pins. VDD_IO is used to set the input voltage tolerance of the input pins. In order for the input pins to be
five-volt input tolerant, VDD_IO must be connected to a fivevolt supply. If the input pins do not have to be five-volt input
tolerant, then VDD_IO can be connected to VDD_CORE.
VDD_IO should never be less than VDD_CORE. VDD_CORE
and VDD_IO should be bypassed with 100 nF ceramic chip
capacitors, as close to the pins as possible, to minimize power
supply and ground bounce caused by inductance in the traces. A
bulk aluminium electrolytic capacitor of 47 µF should also be
provided on the same PC board as the AD1895.
A digital clock connected to the MCLK_I pin or a fundamental or
third overtone crystal connected between MCLK_I and MCLK_O
can be used to generate the master clock, MCLK_I. The MCLK_I
pin can be five-volt input-tolerant just like any of the other
AD1895 input pins. A fundamental mode crystal can be inserted
between MCLK_I and MCLK_O for master clock frequency
generation up to 27 MHz. For master clock frequency generation with a crystal beyond 27 MHz it is recommended that the
user use a third overtone crystal and to add an LC filter at the
output of MCLK_O to filter out the fundamental, do not notch
filter the fundamental. Please refer to your quartz crystal supplier
for values for external capacitors and inductor components.
AD1895
MCLK_I
MCLK_O
R = 45
C1
C2
Figure 9a. Fundamental-Mode Circuit Configuration
Digital Filter Group Delay
The filter group delay is given by the equation:
AD1895
MCLK_I
GD =
GD =
16
fS _ IN
16
fS _ IN
+
32
fS _ IN
MCLK_O
seconds for fS _ OUT > fS _ IN
R = 45
 32   fS _ IN 
+
 seconds for fS _ OUT < fS _ IN
 ×
 fS _ IN   fS _ OUT 
1nF
C1
C2
L1
Mute Control
When the MUTE_IN pin is asserted high, the MUTE_IN control
will perform a soft mute by linearly decreasing the input data
to the AD1895 FIFO to almost zero, –127 dB attenuation. When
MUTE_IN is deasserted low, the MUTE_IN control will linearly
decrease the attenuation of the input data to 0 dB. A 12-bit
counter, clocked by LRCLK_I is used to control the mute attenuation. Therefore, the time it will take from the assertion of
MUTE_IN to –127 dB full mute attenuation is 4096/LRCLK_I
seconds. Likewise, the time it will take to reach 0 dB mute
attenuation from the deassertion of MUTE_IN is 4096/
LRCLK_I seconds.
REV. A
Figure 9b. Third-Overtone Circuit Configuration
There are, of course, maximum and minimum operating frequencies for the AD1895 master clock. The maximum master
clock frequency at which the AD1895 is guaranteed to operate is
30 MHz. 30 MHz is more than sufficient to sample rate convert
sampling frequencies of 192 kHz + 12%. The minimum required
frequency for the master clock generation for the AD1895 depends
upon the input and output sample rates. The master clock has
to be at least 138 times greater than the maximum input or
output sample rate.
–19–
AD1895
Serial Data Ports—Data Format
24 bits, dither is added to the truncated bits. The Right Justified
serial data out mode assumes 64 SCLK_O cycles per frame,
divided evenly for left and right.
The serial data input port mode is set by the logic levels on the
SMODE_IN_0/1/2 pins. The serial data input port modes available are Left Justified, I2S and Right Justified (RJ), 16, 18, 20,
or 24 bits as defined in Table I.
Table II. Serial Data Output Port Mode
SMODE_OUT_[0:2]
Table I. Serial Data Input Port Mode
SMODE_IN_[0:2]
Interface Format
2
1
0
0
0
0
0
1
1
1
1
0
0
1
1
0
0
1
1
0
1
0
1
0
1
0
1
Left Justified
I2 S
Undefined
Undefined
Right Justified, 16 Bits
Right Justified, 18 Bits
Right Justified, 20 Bits
Right Justified, 24 Bits
0
0
0
1
1
0
1
0
1
1
0
0
0
1
1
0
1
0
1
Word Width
24 Bits
20 Bits
18 Bits
16 Bits
The following timing diagrams show the serial mode formats.
RIGHT CHANNEL
MSB
LSB
MSB
MSB
Left Justified (LJ)
I2 S
TDM Mode
Right Justified (RJ)
WLNGTH_OUT_[0:1]
LEFT CHANNEL
LRCLK
SCLK
Interface Format
Table III. Word Width
The serial data output port mode is set by the logic levels on the
SMODE_OUT_0/1 and WLNGTH_OUT_0/1 pins. The serial
mode can be changed to Left Justified, I2S, Right Justified or
TDM as defined in the following table. The output word width
can be set by using the WLNGTH_OUT_0/1 pins as shown in
the Word Width table. When the output word width is less than
SDATA
1
LSB
LEFT JUSTIFIED MODE – 16 TO 24 BITS PER CHANNEL
LRCLK
LEFT CHANNEL
RIGHT CHANNEL
SCLK
SDATA
MSB
LSB
MSB
MSB
LSB
I2S MODE – 16 TO 24 BITS PER CHANNEL
RIGHT CHANNEL
LEFT CHANNEL
LRCLK
SCLK
MSB
SDATA
LSB
MSB
LSB
RIGHT JUSTIFIED MODE – SELECT NUMBER OF BITS PER CHANNEL
LRCLK
SCLK
SDATA
MSB
MSB
LSB
LSB
TDM MODE – 16 TO 24 BITS PER CHANNEL
1/fs
NOTES:
1. LRCLK NORMALLY OPERATES AT ASSOCIATIVE INPUT OR OUTPUT SAMPLE FREQUENCY (fs)
2. SCLK FREQUENCY IS NORMALLY 64 LRCLK EXCEPT FOR TDM MODE WHICH IS N 64 fs,
WHERE N = NUMBER OF STEREO CHANNELS IN THE TDM CHAIN, IN MASTER MODE N = 4
Figure 10. Input/Output Serial Data Formats
–20–
REV. A
AD1895
TDM MODE APPLICATION
In TDM mode, several AD1895s can be daisy-chained together
and connected to the serial input port of a SHARC® DSP. The
AD1895 contains a 64-bit parallel load shift register. When the
LRCLK_O pulse arrives, each AD1895 parallel loads its left and
right data into the 64-bit shift register. The input to the shift
register is connected to TDM_IN while the output is connected
to SDATA_O. By connecting the SDATA_O to the TDM_IN
of the next AD1895, a large shift register is created which is
clocked by SCLK_O.
The number of AD1895s that can be daisy-chained together is
limited by the maximum frequency of SCLK_O, which is about
25 MHz. For example, if the output sample rate, fS, is 48 kHz,
up to eight AD1895s could be connected since 512 × fS is less
than 25 MHz. In Master/TDM Mode, the number of AD1895s
that can be daisy-chained is fixed to four.
LRCLK
SCLK
AD1895
AD1895
SDATA_O
DR0
LRCLK_O
LRCLK_O
LRCLK_O
RFS0
SCLK_O
SCLK_O
SCLK_O
SDATA_O
TDM_IN
SHARC
DSP
AD1895
SDATA_O
TDM_IN
SLAVE-1
TDM_IN
SLAVE-2
RCLK0
SLAVE-n
M2
M1
M0
M2
M1
M0
M2
M1
M0
0
0
0
0
0
0
0
0
0
STANDARD MODE
Figure 11. Daisy-Chain Configuration for TDM Mode (All AD1895s Being Clock-Slaves)
AD1895
AD1895
SDATA_O
DR0
LRCLK_O
LRCLK_O
LRCLK_O
RFS0
SCLK_O
SCLK_O
SCLK_O
SDATA_O
TDM_IN
SHARC
DSP
AD1895
SDATA_O
TDM_IN
CLOCK-MASTER
TDM_IN
RCLK0
SLAVE-n
SLAVE-1
M2
M1
M0
M2
M1
M0
M2
M1
M0
0
1
1
0
0
0
0
0
0
STANDARD MODE
Figure 12. Daisy-Chain Configuration for TDM Mode (First AD1895 Being Clock-Master)
SHARC is a registered trademark of Analog Devices, Inc.
REV. A
–21–
AD1895
Serial Data Port Master Clock Modes
Bypass Mode
Either of the AD1895 serial ports can be configured as a master
serial data port. However, only one serial port can be a master
while the other has to be a slave. In master mode, the AD1895
requires a 256 × fS, 512 fS or 768 × fS master clock (MCLK_I).
For a maximum master clock frequency of 30 MHz, the maximum sample rate is limited to 96 kHz. In slave mode, sample
rates up to 192 kHz can be handled.
When the BYPASS pin is asserted high, the input data bypasses
the sample rate converter and is sent directly to the serial output
port. Dithering of the output data when the word length is set
to less than 24 bits is disabled. This mode is ideal when the
input and output sample rates are the same and LRCLK_I and
LRCLK_O are synchronous with respect to each other. This
mode can also be used for passing through non-AUDIO data
since no processing is performed on the input data in this mode.
When either of the serial ports is operated in master mode, the
master clock is divided down to derive the associated left/
right subframe clock (LRCLK) and serial bit clock (SCLK).
The master clock frequency can be selected for 256, 512, or 768
times the input or output sample rate. Both the input and output serial ports will support master mode LRCLK and SCLK
generation for all serial modes, Left Justified, I2S, Right Justified,
and TDM for the output serial port.
Table IV. Serial Data Port Clock Modes
MMODE_0/1/2
2
1
0
0
0
0
0
1
1
1
1
0
0
1
1
0
0
1
1
0
1
0
1
0
1
0
1
Interface Format
Both Serial Ports are in Slave Mode
Output Serial Port is Master with 768 × fS_OUT
Output Serial Port is Master with 512 × fS_OUT
Output Serial Port is Master with 256 × fS_OUT
Undefined
Input Serial Port is Master with 768 × fS_IN
Input Serial Port is Master with 512 × fS_IN
Input Serial Port is Master with 256 × fS_IN
–22–
REV. A
AD1895
OUTLINE DIMENSIONS
Dimensions shown in inches and (mm).
28-Lead Shrink Small Outline Package (SSOP)
(RS-28)
0.413 (10.50)
0.402 (10.20)
0.390 (9.90)
28
15
PIN 1
14
1
0.220 (5.60)
0.209 (5.30)
0.197 (5.00)
0.323 (8.20)
0.307 (7.80)
0.291 (7.40)
0.073 (1.85)
0.069 (1.75)
0.065 (1.65)
0.079
(2.00)
MAX
0.002 (0.05)
MIN
0.026
(0.65)
BSC
0.015 (0.38) SEATING
0.009 (0.22) PLANE 0.010 (0.25)
0.004 (0.09)
8
4
0
0.037 (0.95)
0.030 (0.75)
0.022 (0.55)
AD1895–Revision History
Location
Page
Data Sheet changed from REV. 0 to REV. A.
Changes to Specifications table . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
REV. A
–23–
–24–
PRINTED IN U.S.A.
C00758–1.5–6/01(A)