PDF Data Sheet Rev. B

Quad Input Multiservice Line Card Adaptive
Clock Translator with Frame Sync
AD9558
Data Sheet
FEATURES
Pin program function for easy frequency translation
configuration
Software controlled power-down
64-lead, 9 mm × 9 mm, LFCSP package
Supports GR-1244 Stratum 3 stability in holdover mode
Supports smooth reference switchover with virtually
no disturbance on output phase
Supports Telcordia GR-253 jitter generation, transfer, and
tolerance for SONET/SDH up to OC-192 systems
Supports ITU-T G.8262 synchronous Ethernet slave clocks
Supports ITU-T G.823, G.824, G.825, and G.8261
Auto/manual holdover and reference switchover
4 reference inputs (single-ended or differential)
Input reference frequencies: 2 kHz to 1250 MHz
Reference validation and frequency monitoring (1 ppm)
Programmable input reference switchover priority
20-bit programmable input reference divider
6 pairs of clock output pins with each pair configurable as
a single differential LVDS/HSTL output or as 2 single-ended
CMOS outputs
Output frequencies: 352 Hz to 1250 MHz
Programmable 17-bit integer and 24-bit fractional
feedback divider in digital PLL
Programmable digital loop filter covering loop bandwidths
from 0.1 Hz to 5 kHz (2 kHz maximum for <0.1 dB of
peaking)
Low noise system clock multiplier
Frame sync support
Adaptive clocking
Optional crystal resonator for system clock input
On-chip EEPROM to store multiple power-up profiles
APPLICATIONS
Network synchronization, including synchronous Ethernet
and SDH to OTN mapping/demapping
Cleanup of reference clock jitter
SONET/SDH clocks up to OC-192, including FEC
Stratum 3 holdover, jitter cleanup, and phase transient
control
Wireless base station controllers
Cable infrastructure
Data communications
GENERAL DESCRIPTION
The AD9558 is a low loop bandwidth clock multiplier that
provides jitter cleanup and synchronization for many systems,
including synchronous optical networks (SONET/SDH). The
AD9558 generates an output clock synchronized to up to four
external input references. The digital PLL allows for reduction
of input time jitter or phase noise associated with the external
references. The digitally controlled loop and holdover circuitry
of the AD9558 continuously generates a low jitter output clock
even when all reference inputs have failed.
The AD9558 operates over an industrial temperature range of
−40°C to +85°C. If a smaller package is required, refer to the
AD9557 for the two-input/two-output version of the same part.
FUNCTIONAL BLOCK DIAGRAM
STABLE
SOURCE
AD9558
CHANNEL 0
DIVIDER
CLOCK
MULTIPLIER
CHANNEL 1
DIVIDER
÷3 TO ÷11
HF DIVIDER 0
DIGITAL
PLL
ANALOG
PLL
÷3 TO ÷11
HF DIVIDER 1
REFERENCE INPUT
AND
MONITOR MUX
CHANNEL 3
DIVIDER
FRAME SYNC
EEPROM
STATUS AND
CONTROL PINS
09758-001
SERIAL INTERFACE
(SPI OR I2C)
CHANNEL 2
DIVIDER
Figure 1.
Rev. B
Document Feedback
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. Specifications subject to change without notice. No
license is granted by implication or otherwise under any patent or patent rights of Analog Devices.
Trademarks and registered trademarks are the property of their respective owners.
One Technology Way, P.O. Box 9106, Norwood, MA 02062-9106, U.S.A.
Tel: 781.329.4700 ©2011–2013 Analog Devices, Inc. All rights reserved.
Technical Support
www.analog.com
AD9558
Data Sheet
TABLE OF CONTENTS
Features ........................................................................................................ 1
Digital PLL (DPLL) Core ...................................................................31
Applications ................................................................................................ 1
Loop Control State Machine .............................................................34
General Description .................................................................................. 1
System Clock (SYSCLK) .........................................................................35
Functional Block Diagram ....................................................................... 1
System Clock Inputs............................................................................35
Revision History......................................................................................... 3
System Clock Multiplier .....................................................................35
Specifications .............................................................................................. 4
Output PLL (APLL).................................................................................37
Supply Voltage ........................................................................................ 4
Clock Distribution ...................................................................................38
Supply Current ....................................................................................... 4
Clock Dividers......................................................................................38
Power Dissipation.................................................................................. 5
Output Power-Down ..........................................................................38
Logic Inputs (SYNC, RESET, PINCONTROL, M7 to M0).......... 5
Output Enable ......................................................................................38
Logic Outputs (M7 to M0, IRQ) ........................................................ 6
Output Mode........................................................................................38
System Clock Inputs (XOA, XOB) ..................................................... 6
Clock Distribution Synchronization................................................38
Reference Inputs .................................................................................... 7
Frame Synchronization ...........................................................................40
Reference Monitors ............................................................................... 8
Reference Configuration in Frame Synchronization Mode ........40
Reference Switchover Specifications.................................................. 8
Clock Outputs in Frame Synchronization Mode ..........................40
Distribution Clock Outputs................................................................. 9
Control Registers for Frame Synchronization Mode....................40
Time Duration of Digital Functions ................................................ 10
Level Sensitive Mode and One-Shot Mode ....................................40
Digital PLL ............................................................................................ 11
M3b Divider/OUT5 Programming in Frame
Synchronization Mode .......................................................................41
E
Digital PLL Lock Detection............................................................... 11
Holdover Specifications...................................................................... 11
Serial Port Specifications—SPI Mode ............................................. 12
Serial Port Specifications—I2C Mode .............................................. 13
Jitter Generation .................................................................................. 13
Absolute Maximum Ratings .................................................................. 16
ESD Caution ......................................................................................... 16
Pin Configuration and Function Descriptions .................................. 17
Typical Performance Characteristics.................................................... 20
Input/Output Termination Recommendations ................................. 25
Getting Started.......................................................................................... 26
Chip Power Monitor and Startup ..................................................... 26
Multifunction Pins at Reset/Power-Up ........................................... 26
Device Register Programming When Using a Register
Setup File ............................................................................................... 26
Status and Control ...................................................................................42
Multifunction Pins (M7 to M0)........................................................42
Watchdog Timer ..................................................................................43
EEPROM...............................................................................................43
Serial Control Port ...................................................................................49
SPI/I2C Port Selection.........................................................................49
SPI Serial Port Operation...................................................................49
I2C Serial Port Operation ...................................................................53
Programming the I/O Registers ............................................................56
Buffered/Active Registers...................................................................56
Autoclear Registers..............................................................................56
Register Access Restrictions ..............................................................56
Thermal Performance .............................................................................57
Power Supply Partitions ..........................................................................58
Register Programming Overview .................................................... 26
Recommended Configuration for 3.3 V Switching Supply.........58
Theory of Operation ............................................................................... 29
Configuration for 1.8 V Supply.........................................................58
Overview ............................................................................................... 29
Pin Program Function Description ......................................................59
Reference Clock Inputs ...................................................................... 30
Overview of On-Chip ROM Features .............................................59
Reference Monitors ............................................................................. 30
Hard Pin Programming Mode..........................................................60
Reference Profiles ................................................................................ 30
Soft Pin Programming Overview .....................................................61
Reference Switchover .......................................................................... 30
Register Map .............................................................................................62
Rev. B | Page 2 of 104
Data Sheet
AD9558
Register Map Bit Descriptions............................................................... 72
Serial Port Configuration (Register 0x0000 to Register 0x0005)72
Silicon Revision (Register 0x000A) ................................................. 72
Clock Part Serial ID (Register 0x000C to Register 0x000D) ...... 72
System Clock (Register 0x0100 to Register 0x0108) .................... 73
General Configuration (Register 0x0200 to Register 0x0214) ... 74
Frame Synchronization (Register 0x0640 to
Register 0x0641) .......................................................................... 86
DPLL Profile Registers (Register 0x0700 to
Register 0x07E6) ......................................................................... 87
Operational Controls (Register 0x0A00 to Register
0x0A10) ........................................................................................ 89
IRQ Mask (Register 0x020A to Register 0x020F) ...................75
Quick In/Out Frequency Soft Pin Configuration (Register
0x0C00 to Register 0x0C08) ...................................................... 92
DPLL Configuration (Register 0x0300 to Register 0x032E) .76
Status ReadBack (Register 0x0D00 to Register 0x0D14)....... 93
Output PLL Configuration (Register 0x0400 to
Register 0x0408) ..........................................................................79
EEPROM Control (Register 0x0E00 to Register 0x0E03) ..... 97
EEPROM Storage Sequence (Register 0x0E10 to
Register 0x0E3C)......................................................................... 98
Output Clock Distribution (Register 0x0500 to
Register 0x0515) ..........................................................................81
Outline Dimensions ......................................................................104
Reference Inputs (Register 0x0500 to Register 0x0507) ........85
Ordering Guide .........................................................................104
REVISION HISTORY
6/13—Rev. A to Rev. B
Changes to Multifunction Pins (M7 to M0) Section .................. 42
Change to Address 0x0101, Table 35 ............................................ 62
Changes to Bit 4, Table 43 .............................................................. 73
Updated Outline Dimensions ...................................................... 104
4/12—Rev 0 to Rev. A
Changed 3 Hz to 352 kHz in Output Frequencies List Item,
Features Section ................................................................................. 1
Change to Output Frequency Range Parameter, Min; and System
Clock Input Doubler Duty Cycle Parameter Description, Table 6... 6
Changes to Test Conditions/Comments Column, Table 9 .......... 8
Changes to Output Frequency Parameters, Min, Table 10 .......... 9
Changes to Pin 4 and Pin 42, Table 20 ......................................... 17
Changes to Device Register Programming When Using a Register
Setup File and Register Programming Overview Sections ............. 26
Changed APLL VCO Lower Frequency and OUT5 Frequency
Range, Figure 35; Changed 225 MHz to 200 MHz and 3.45 GHz
to 3.35 GHz in Overview Section................................................... 29
Changes to Reference Profiles Section ......................................... 30
Changes to Programmable Digital Loop Filter Section ............. 32
Changes to System Clock Inputs Section ..................................... 35
Changes to Output PLL (APLL) Section; Changes to Figure 39 .... 37
Changes to Figure 40; Changed 1024 to 1023 in Clock Dividers
Section; Changes to Clock Distribution Synchronization
Section .............................................................................................. 38
Changes to Multifunction Pins (M7 to M0) and IRQ Pin
Sections ............................................................................................. 42
Changes to Figure 44 ...................................................................... 43
Changes to EEPROM Conditional Processing Section and
Figure 45 ............................................................................................ 46
Added Programming the EEPROM to Configure an M Pin
to Control Synchronization of Clock Distribution Section ......... 48
Changes to the Power Supply Partitions Section ........................ 58
Changed 89.5° to 88.5° in DPLL Phase Margin Section ............ 59
Changes to Address 0x0006, Address 0x0007, and
Address 0x000A, Table 35 .............................................................. 62
Changes to Address 0x0304, Table 35 .......................................... 63
Changes to Address 0x0405, Table 35 .......................................... 64
Changes to Address 0x071A and Address 0x071D, Table 35 .... 65
Changes to Address 0x0780, Address 0x0785 to
Address 0x078A, Address 0x079A, Address 0x079D, Table 35 ... 67
Changes to Address 0x07C0, Address 0x07DA, and
Address 0x07DD, Table 35 ............................................................. 68
Change to Address 0x0A01, Bit 7, Table 35 ................................. 69
Added Address 0x0E3D to Address 0x0E45, Table 35 ............... 71
Change to Table 38; Added Table 40, Renumbered Sequentially;
Changes to Table 41 ........................................................................ 72
Change to Bit 0, Address 0x0101, Table 43 .................................. 73
Changes to Address 0x0304, Table 55 .......................................... 76
Deleted Address 0x0305, Table 55 ................................................ 76
Changes to Table Title, Table 63; Changes to Address 0x0400
and Address 0x0403, Table 64........................................................ 79
Changes to Address 0x0405, Table 64 .......................................... 80
Changes to Descriptions, Address 0x0500, Table 67 .................. 81
Changes to Bit 0, Address 0x0501, Table 68 ................................ 82
Changes to Bits[6:4], Address 0x0505 and Changes to
Address 0x0506, Table 70 ............................................................... 83
Changes to Bits[6:4] and Bit 0, Address 0x050F, Table 73 ......... 84
Change to Address 0x0704, Table 78; Changes to Bits[3:0] in
Address 0x0707 and Address 070A, Table 79; and Changes to
Address 0x070E, Table 82................................................................ 87
Changes to Address 0x0710, Table 83; and Changes to Bits[3:0],
Address 0x0714, Table 84 ................................................................. 88
Changes to Bits[1:0], Address 0x0A01, Table 90........................... 89
Changes to Descriptions, Address 0x0A0B, Table 99................. 91
Changes to Bit 4, Address 0x0C06, Table 100 ............................. 93
Changes to Bit 6 and Bit 1, Address 0x0D01, Table 102 ............ 94
Changes to Table Summary, Table 114 ......................................... 98
Added Table 128 ............................................................................ 101
Changes to Table 129 .................................................................... 102
Changes to Table 130 .................................................................... 103
10/11—Revision 0: Initial Version
Rev. B | Page 3 of 104
AD9558
Data Sheet
SPECIFICATIONS
Minimum (min) and maximum (max) values apply for the full range of supply voltage and operating temperature variations. Typical (typ)
values apply for AVDD3 = DVDD_I/O = 3.3 V; AVDD = DVDD = 1.8 V; TA= 25°C, unless otherwise noted.
SUPPLY VOLTAGE
Table 1.
Parameter
SUPPLY VOLTAGE
DVDD3
DVDD
AVDD3
AVDD
Min
Typ
Max
Unit
3.135
1.71
3.135
1.71
3.30
1.80
3.30
1.80
3.465
1.89
3.465
1.89
V
V
V
V
Test Conditions/Comments
SUPPLY CURRENT
The test conditions for the maximum (max) supply current are the same as the test conditions for the All Blocks Running parameter of
Table 3. The test conditions for the typical (typ) supply current are the same as the test conditions for the Typical Configuration parameter
of Table 3.
Table 2.
Parameter
SUPPLY CURRENT FOR TYPICAL
CONFIGURATION
IDVDD3
IDVDD
IAVDD3
IAVDD
Min
Typ
Max
Unit
12
12
50
152
19
20
70
230
26
28
92
305
mA
mA
mA
mA
SUPPLY CURRENT FOR THE ALL BLOCKS
RUNNING CONFIGURATION
IDVDD3
IDVDD
IAVDD3
IAVDD
23
11
73
168
34
22
108
250
46
32
143
331
mA
mA
mA
mA
Rev. B | Page 4 of 104
Test Conditions/Comments
Typical numbers are for the typical configuration listed
in Table 3
Pin 45, Pin 46, Pin 51, Pin 52, Pin 64
Pin 6, Pin 55, Pin 56
Pin 25, Pin 26, Pin 31
Pin 7, Pin 10, Pin 12, Pin 17, Pin 22, Pin 29, Pin 30, Pin 35,
Pin 37, Pin 38
Maximum numbers are for all blocks running configuration
in Table 3
Pin 45, Pin 46, Pin 51, Pin 52, Pin 64
Pin 6, Pin 55, Pin 56
Pin 25, Pin 26, Pin 31
Pin 7, Pin 10, Pin 12, Pin 17, Pin 22, Pin 29, Pin 30, Pin 35,
Pin 37, Pin 38
Data Sheet
AD9558
POWER DISSIPATION
Table 3.
Parameter
POWER DISSIPATION
Typical Configuration
Min
Typ
Max
Unit
Test Conditions/Comments
0.47
0.74
1.02
W
All Blocks Running
0.6
1.0
1.32
W
44
125
mW
System clock: 49.152 MHz crystal; DPLL active;
both 19.44 MHz input references in differential mode;
one HSTL driver at 644.53125 MHz;
one 3.3 V CMOS driver at 161.1328125 MHz and 80 pF
capacitive load on CMOS output
System clock: 49.152 MHz crystal; DPLL active;
both input references in differential mode;
four HSTL drivers at 750 MHz;
four 3.3 V CMOS drivers at 250 MHz and 80 pF capacitive
load on CMOS outputs
Typical configuration with no external pull-up or pulldown resistors; about 2/3 of this power is on AVDD3
Conditions = typical configuration; table values show the
change in power due to the indicated operation
20
26
5
25
32
7
32
40
9
mW
mW
mW
Additional current draw is in the DVDD3 domain only
Additional current draw is in the DVDD3 domain only
Additional current draw is in the DVDD3 domain only
12
14
14
18
17
21
21
27
22
28
28
36
mW
mW
mW
mW
Additional current draw is in the AVDD domain only
Additional current draw is in the AVDD domain only
A single 1.8 V CMOS output with an 80 pF load
A single 3.3 V CMOS output with an 80 pF load
36
10
51
17
64
23
mW
mW
Additional current draw is in the AVDD domain only
Additional current draw is in the AVDD domain only
Max
Unit
Test Conditions/Comments
0.8
±100
V
V
µA
pF
2.2
0.6
±100
V
V
V
µA
pF
Full Power-Down
Incremental Power Dissipation
Input Reference On/Off
Differential Without Divide-by-2
Differential With Divide-by-2
Single-Ended (Without Divide-by-2)
Output Distribution Driver On/Off
LVDS (at 750 MHz)
HSTL (at 750 MHz)
1.8 V CMOS (at 250 MHz)
3.3 V CMOS (at 250 MHz)
Other Blocks On/Off
Second RF Divider
Channel Divider Bypassed
LOGIC INPUTS (SYNC, RESET, PINCONTROL, M7 TO M0)
E
A
A
Table 4.
Parameter
LOGIC INPUTS (SYNC, RESET, PINCONTROL)
Input High Voltage (VIH)
Input Low Voltage (VIL)
Input Current (IINH, IINL)
Input Capacitance (CIN)
LOGIC INPUTS (M7 to M0)
Input High Voltage (VIH)
Input ½ Level Voltage (VIM)
Input Low Voltage (VIL)
Input Current (IINH, IINL)
Input Capacitance (CIN)
Min
Typ
E
A
EA
A
A
2.1
±50
3
2.5
1.0
±60
3
Rev. B | Page 5 of 104
AD9558
Data Sheet
LOGIC OUTPUTS (M7 TO M0, IRQ)
Table 5.
Parameter
LOGIC OUTPUTS (M7 to M0, IRQ)
Output High Voltage (VOH)
Output Low Voltage (VOL)
IRQ Leakage Current
Active Low Output Mode
Active High Output Mode
Min
Typ
Max
Unit
Test Conditions/Comments
0.4
V
V
−200
100
μA
μA
IOH = 1 mA
IOL = 1 mA
Open-drain mode
VOH = 3.3 V
VOL = 0 V
Max
Unit
Test Conditions/Comments
750
805
MHz
The VCO range may place limitations on
nonstandard system clock input
frequencies
150
255
MHz
2
600
MHz
V/μs
DVDD3 − 0.4
SYSTEM CLOCK INPUTS (XOA, XOB)
Table 6.
Parameter
SYSTEM CLOCK MULTIPLIER
Output Frequency Range
Phase Frequency Detector (PFD) Rate
Frequency Multiplication Range
SYSTEM CLOCK REFERENCE INPUT PATH
Input Frequency Range
Minimum Input Slew Rate
Common-Mode Voltage
Differential Input Voltage Sensitivity
Min
Typ
10
20
Assumes valid system clock and PFD
rates
1.05
250
1.16
1.25
V
mV p-p
45
46
47
50
50
50
3
4.2
55
54
53
%
%
%
pF
kΩ
System Clock Input Doubler Duty Cycle
System Clock Input = 50 MHz
System Clock Input = 20 MHz
System Clock Input = 16 MHz to 20 MHz
Input Capacitance
Input Resistance
CRYSTAL RESONATOR PATH
Crystal Resonator Frequency Range
Maximum Crystal Motional Resistance
10
50
100
Rev. B | Page 6 of 104
MHz
Ω
Minimum limit imposed for jitter
performance
Internally generated
Minimum voltage across pins required
to ensure switching between logic
states; the instantaneous voltage on
either pin must not exceed the supply
rails; can accommodate single-ended
input by ac grounding of complementary
input; 1 V p-p recommended for optimal
jitter performance
This is the amount of duty cycle
variation that can be tolerated on the
system clock input to use the doubler
Single-ended, each pin
Fundamental mode, AT cut crystal
Data Sheet
AD9558
REFERENCE INPUTS
Table 7.
Parameter
DIFFERENTIAL OPERATION
Frequency Range
Sinusoidal Input
LVPECL Input
LVDS Input
Min
Max
Unit
10
0.002
750
1250
MHz
MHz
0.002
750
MHz
Minimum Input Slew Rate
Common-Mode Input Voltage
AC-Coupled
DC-Coupled
Differential Input Voltage Sensitivity
40
fIN < 800 MHz
fIN = 800 to 1050 MHz
fIN = 1050 to 1250 MHz
Differential Input Voltage Hysteresis
Input Resistance
Input Capacitance
240
320
400
Minimum Pulse Width High
LVPECL
LVDS
Minimum Pulse Width Low
LVPECL
LVDS
SINGLE-ENDED OPERATION
Frequency Range (CMOS)
Minimum Input Slew Rate
Input Voltage High (VIH)
1.2 V to 1.5 V Threshold Setting
1.8 V to 2.5 V Threshold Setting
3.0 V to 3.3 V Threshold Setting
Input Voltage Low (VIL)
1.2 V to 1.5 V Threshold Setting
1.8 V to 2.5 V Threshold Setting
3.0 V to 3.3 V Threshold Setting
Input Resistance
Input Capacitance
Minimum Pulse Width High
Minimum Pulse Width Low
1.9
1.0
Typ
V/μs
2
58
21
3
2.2
2.4
100
V
V
mV
ps
ps
390
640
ps
ps
300
1.0
1.4
2.0
47
3
1.5
1.5
MHz
V/μs
V
V
V
0.35
0.5
1.0
The reference input divide-by-2 block must be
engaged for fIN > 705 MHz
The reference input divide-by-2 block must be
engaged for fIN > 705 MHz
Minimum limit imposed for jitter performance
Internally generated
Minimum differential voltage across pins is required to
ensure switching between logic levels; instantaneous
voltage on either pin must not exceed the supply rails
mV
mV
mV
mV
kΩ
pF
390
640
0.002
40
Test Conditions/Comments
V
V
V
kΩ
pF
ns
ns
Rev. B | Page 7 of 104
Minimum limit imposed for jitter performance
AD9558
Data Sheet
REFERENCE MONITORS
Table 8.
Parameter
REFERENCE MONITORS
Reference Monitor
Loss of Reference Detection Time
Frequency Out-of-Range Limits
Validation Timer
1
Min
Typ
Max
Unit
Test Conditions/Comments
1.1
DPLL PFD
period
Δf/fREF
(ppm)
Nominal phase detector period = R/fREF 1
<2
105
0.001
65.535
sec
0F0F
Programmable (lower bound is subject to quality
of the system clock (SYSCLK)); SYSCLK accuracy
must be better than the lower bound
Programmable in 1 ms increments
fREF is the frequency of the active reference; R is the frequency division factor determined by the R-divider.
REFERENCE SWITCHOVER SPECIFICATIONS
Table 9.
Parameter
REFERENCE SWITCHOVER SPECIFICATIONS
Maximum Output Phase Perturbation
(Phase Build-Out Switchover)
50 Hz DPLL Loop Bandwidth
Peak
Steady State
2 kHz DPLL Loop Bandwidth
Peak
Steady State
Time Required to Switch to
a New Reference
Phase Build-Out Switchover
Min
Typ
Max
Unit
0
0
±100
±100
ps
ps
0
0
±250
±100
ps
ps
1.1
DPLL PFD
period
Test Conditions/Comments
Assumes a jitter-free reference; satisfies Telcordia
GR-1244-CORE requirements; select high PM base
loop filter bit (Register 0x070E, Bit 0) is set to 1 for
all active references
Valid for automatic and manual reference switching
Valid for automatic and manual reference switching
Rev. B | Page 8 of 104
Calculated using the nominal phase detector
period (NPDP = R/fREF); the total time required is
equal to the time plus the reference validation
time and the time required to lock to the new
reference
Data Sheet
AD9558
DISTRIBUTION CLOCK OUTPUTS
Table 10.
Parameter
HSTL MODE
Output Frequency
Min
Typ
Max
Unit
Test Conditions/Comments
1250
MHz
140
250
ps
OUT5 only; OUT0 to OUT4 minimum output
frequency is 360 kHz
100 Ω termination across output pins
48
48
43
950
870
52
53
1200
960
%
%
%
mV
mV
Magnitude of voltage across pins; output driver static
Output driver static
1250
MHz
185
280
ps
48
47
43
53
53
%
%
%
454
mV
50
mV
1.375
50
24
V
mV
mA
Output driver static
Voltage difference between pins; output driver static
Output driver static
0.000352
Rise/Fall Time (20% to 80%) 1
Duty Cycle
Up to fOUT = 700 MHz
Up to fOUT = 750 MHz
Up to fOUT = 1250 MHz
Differential Output Voltage Swing
Common-Mode Output Voltage
LVDS MODE
Output Frequency
1F
Rise/Fall Time (20% to 80%)1
Duty Cycle
Up to fOUT = 750 MHz
Up to fOUT = 800 MHz
Up to fOUT = 1250 MHz
Differential Output Voltage Swing
Balanced, VOD
45
42
700
700
0.000352
44
43
247
Unbalanced, ΔVOD
Offset Voltage
Common-Mode, VOS
Common-Mode Difference, ΔVOS
Short-Circuit Output Current
CMOS MODE
Output Frequency
1.8 V Supply
3.3 V Supply (OUT0 and OUT5)
Strong Drive Strength Setting
Weak Drive Strength Setting
Rise/Fall Time (20% to 80%)1
1.8 V Supply
3.3 V Supply
Strong Drive Strength Setting
Weak Drive Strength Setting
Duty Cycle
1.8 V Mode
3.3 V Strong Mode
3.3 V Weak Mode
Output Voltage High (VOH)
AVDD3 = 3.3 V, IOH = 10 mA
AVDD3 = 3.3 V, IOH = 1 mA
AVDD3 = 1.8 V, IOH = 1 mA
Output Voltage Low (VOL)
AVDD3 = 3.3 V, IOL = 10 mA
AVDD3 = 3.3 V, IOL = 1 mA
AVDD3 = 1.8 V, IOL = 1 mA
1.125
1.26
13
OUT5 only; OUT0 to OUT4 minimum output
frequency is 360 kHz
100 Ω termination across the output pair
Voltage swing between output pins; output driver
static
Absolute difference between voltage swing of
normal pin and inverted pin; output driver static
0.000352
150
MHz
OUT5 only; OUT0 to OUT4 minimum output
frequency is 360 kHz
10 pF load
0.000352
0.000352
250
25
MHz
MHz
10 pF load
10 pF load
1.5
3
ns
10 pF load
0.4
8
0.6
ns
ns
10 pF load
10 pF load
%
%
%
10 pF load
10 pF load
10 pF load
Output driver static; strong drive strength
50
47
51
AVDD3 − 0.3
AVDD3 − 0.1
AVDD − 0.2
V
V
V
Output driver static; strong drive strength
0.3
0.1
0.1
Rev. B | Page 9 of 104
V
V
V
AD9558
Parameter
OUTPUT TIMING SKEW
Between OUT0 and OUT1
Data Sheet
Typ
Max
Unit
10
70
ps
Between OUT0 and OUT3
105
222
ps
Between OUT0 and OUT5
1.39
1.76
ns
Between OUT1 and OUT2
(OUT1 and OUT2 Share the
Same Divider)
Between OUT3 and OUT4
(OUT3 and OUT4 Share the
Same Divider)
Across All OUT0 to OUT4 HSTL
1
12
ps
1
24
ps
HSTL mode on both drivers; rising edge only; any
divide value
105
235
ps
Across All OUT0 to OUT4 LVDS
100
235
ps
HSTL mode on all drivers; rising edge only; any
divide value
LVDS mode on all drivers; rising edge only; any
divide value
−5
+1
+5
ps
−5
0
+5
ps
3.53
3.55
3.56
4.84
3.59
3.65
3.68
5.1
ns
ns
ns
ns
Typ
Max
Unit
Test Conditions/Comments
13
20
ms
138
145
ms
Using default EEPROM storage sequence
(see Register 0x0E10 to Register 0x0E3F)
Using default EEPROM storage sequence
(see Register 0x0E10 to Register 0x0E3F)
Time from power-down exit to system clock
lock detect
Additional Delay on One Driver by
Changing Its Logic Type
HSTL to LVDS
HSTL to 1.8 V CMOS
Min
HSTL to 3.3 V CMOS, Strong Mode
OUT0 CMOS to OUT1 HSTL
OUT0 CMOS to OUT3 HSTL
OUT0 CMOS to OUT4 HSTL
OUT0 CMOS to OUT5 HSTL
1
Test Conditions/Comments
10 pF load
HSTL mode on both drivers; rising edge only; any
divide value
HSTL mode on both drivers; rising edge only; any
divide value
HSTL mode on both drivers; rising edge only; any
divide value
HSTL mode on both drivers; rising edge only; any
divide value
Positive value indicates that the LVDS edge is
delayed relative to HSTL
Positive value indicates that the CMOS edge is
delayed relative to HSTL
The CMOS edge is delayed relative to HSTL
The listed values are for the slower edge (rise or fall).
TIME DURATION OF DIGITAL FUNCTIONS
Table 11.
Parameter
TIME DURATION OF DIGITAL
FUNCTIONS
EEPROM-to-Register Download
Time
Register-to-EEPROM Upload Time
Minimum Power-Down Exit Time
Min
1
ms
Rev. B | Page 10 of 104
Data Sheet
AD9558
DIGITAL PLL
Table 12.
Parameter
DIGITAL PLL
Phase-Frequency Detector (PFD)
Input Frequency Range
Loop Bandwidth
Phase Margin
Closed-Loop Peaking
Reference Input (R) Division Factor
Integer Feedback (N1) Division Factor
Fractional Feedback Divide Ratio
Min
Typ
Max
Unit
2
100
kHz
0.1
30
<0.1
2000
89
Hz
Degrees
dB
1
180
0
220
217
0.999
Test Conditions/Comments
Programmable design parameter
Programmable design parameter
Programmable design parameter ;
part can be programmed for <0.1 dB peaking in
accordance with Telcordia GR-253 jitter transfer
1, 2, …, 1,048,576
180, 181, …, 131,072
Maximum value: 16,777,215/16,777,216
DIGITAL PLL LOCK DETECTION
Table 13.
Parameter
PHASE LOCK DETECTOR
Threshold Programming Range
Threshold Resolution
FREQUENCY LOCK DETECTOR
Threshold Programming Range
Threshold Resolution
Min
Typ
0.001
Max
Unit
65.5
ns
ps
16,700
ns
ps
Reference-to-feedback period difference
Max
Unit
Test Conditions/Comments
ppm
Excludes frequency drift of SYSCLK source;
excludes frequency drift of input reference prior
to entering holdover; compliant with GR-1244
Stratum 3
1
0.001
1
Test Conditions/Comments
HOLDOVER SPECIFICATIONS
Table 14.
Parameter
HOLDOVER SPECIFICATIONS
Initial Frequency Accuracy
Min
Typ
<0.01
Rev. B | Page 11 of 104
AD9558
Data Sheet
SERIAL PORT SPECIFICATIONS—SPI MODE
Table 15.
Parameter
CS
Input Logic 1 Voltage
Input Logic 0 Voltage
Input Logic 1 Current
Input Logic 0 Current
Input Capacitance
SCLK
Input Logic 1 Voltage
Input Logic 0 Voltage
Input Logic 1 Current
Input Logic 0 Current
Input Capacitance
SDIO
As an Input
Input Logic 1 Voltage
Input Logic 0 Voltage
Input Logic 1 Current
Input Logic 0 Current
Input Capacitance
As an Output
Output Logic 1 Voltage
Output Logic 0 Voltage
SDO
Output Logic 1 Voltage
Output Logic 0 Voltage
TIMING
SCLK
Clock Rate, 1/tCLK
Pulse Width High, tHIGH
Pulse Width Low, tLOW
SDIO to SCLK Setup, tDS
SCLK to SDIO Hold, tDH
SCLK to Valid SDIO and SDO, tDV
CS to SCLK Setup (tS)
CS to SCLK Hold (tC)
CS Minimum Pulse Width High
Min
Typ
Max
Unit
Test Conditions/Comments
E
A
E
A
A
E
A
A
E
A
A
2.0
0.8
60
100
2
V
V
µA
µA
pF
2.0
0.8
200
1
2
V
V
µA
µA
pF
2.0
0.8
1
1
2
V
V
µA
µA
pF
Internal 30 kΩ pull-down resistor
DVDD3 − 0.6
0.4
V
V
1 mA load current
1 mA load current
0.4
V
V
1 mA load current
1 mA load current
DVDD3 − 0.6
40
10
13
3
6
10
10
0
6
Rev. B | Page 12 of 104
MHz
ns
ns
ns
ns
ns
ns
ns
ns
Data Sheet
AD9558
SERIAL PORT SPECIFICATIONS—I2C MODE
Table 16.
Parameter
SDA, SCL (AS INPUT)
Input Logic 1 Voltage
Min
Typ
Pulse Width of Spikes That Must Be Suppressed by
the Input Filter, tSP
SDA (AS OUTPUT)
Output Logic 0 Voltage
Output Fall Time from VIHmin to VILmax
TIMING
SCL Clock Rate
Bus-Free Time Between a Stop and Start
Condition, tBUF
Repeated Start Condition Setup Time, tSU; STA
Repeated Hold Time Start Condition, tHD; STA
Stop Condition Setup Time, tSU; STO
Low Period of the SCL Clock, tLOW
High Period of the SCL Clock, tHIGH
SCL/SDA Rise Time, tR
SCL/SDA Fall Time, tF
Data Setup Time, tSU; DAT
Data Hold Time, tHD; DAT
Capacitive Load for Each Bus Line, Cb1
1
Unit
Test Conditions/Comments
V
0.7 ×
DVDD3
Input Logic 0 Voltage
Input Current
Hysteresis of Schmitt Trigger Inputs
Max
V
0.3 ×
DVDD3
+10
µA
50
ns
0.4
250
V
ns
400
1.3
kHz
µs
0.6
0.6
µs
µs
0.6
1.3
0.6
20 + 0.1 Cb1
20 + 0.1 Cb1
100
100
µs
µs
µs
ns
ns
ns
ns
pF
−10
0.015 ×
DVDD3
20 + 0.1 Cb 1
2F
300
300
400
For VIN = 10% to 90% DVDD3
IO = 3 mA
10 pF ≤ Cb ≤ 400 pF
After this period, the first clock pulse is
generated
Cb is the capacitance (pF) of a single bus line.
JITTER GENERATION
Jitter generation (random jitter) uses 49.152 MHz crystal for system clock input.
Table 17.
Parameter
JITTER GENERATION
fREF = 19.44 MHz; fOUT = 622.08 MHz; fLOOP = 50 Hz
HSTL Driver
Bandwidth: 5 kHz to 20 MHz
Bandwidth: 12 kHz to 20 MHz
Bandwidth: 20 kHz to 80 MHz
Bandwidth: 50 kHz to 80 MHz
Bandwidth: 16 MHz to 320 MHz
Min
Typ
Max
304
296
300
266
185
Rev. B | Page 13 of 104
Unit
fs rms
fs rms
fs rms
fs rms
fs rms
Test Conditions/Comments
System clock doubler enabled;
high phase margin mode enabled;
Register 0x0405 = 0x20; Register 0x0403 =
0x07; Register 0x0400 = 0x81;
in cases where multiple driver types are
listed, both driver types were tested at
those conditions, and the one with higher
jitter is quoted, although there is usually
not a significant jitter difference between
the driver types
AD9558
Parameter
fREF = 19.44 MHz; fOUT = 644.53 MHz; fLOOP = 50 Hz
HSTL and/or LVDS Driver
Bandwidth: 5 kHz to 20 MHz
Bandwidth: 12 kHz to 20 MHz
Bandwidth: 20 kHz to 80 MHz
Bandwidth: 50 kHz to 80 MHz
Bandwidth: 16 MHz to 320 MHz
fREF = 19.44 MHz; fOUT = 693.48 MHz; fLOOP = 50 Hz
HSTL Driver
Bandwidth: 5 kHz to 20 MHz
Bandwidth: 12 kHz to 20 MHz
Bandwidth: 20 kHz to 80 MHz
Bandwidth: 50 kHz to 80 MHz
Bandwidth: 16 MHz to 320 MHz
fREF = 19.44 MHz; fOUT = 174.703 MHz; fLOOP = 1 kHz
HSTL Driver
Bandwidth: 5 kHz to 20 MHz
Bandwidth: 12 kHz to 20 MHz
Bandwidth: 20 kHz to 80 MHz
Bandwidth: 50 kHz to 80 MHz
Bandwidth: 4 MHz to 80 MHz
fREF = 19.44 MHz; fOUT = 174.703 MHz; fLOOP = 100 Hz
LVDS and/or 3.3 V CMOS Driver
Bandwidth: 5 kHz to 20 MHz
Bandwidth: 12 kHz to 20 MHz
Bandwidth: 20 kHz to 80 MHz
Bandwidth: 50 kHz to 80 MHz
Bandwidth: 4 MHz to 80 MHz
fREF = 25 MHz; fOUT = 161.1328 MHz; fLOOP = 100 Hz
HSTL Driver
Bandwidth: 5 kHz to 20 MHz
Bandwidth: 12 kHz to 20 MHz
Bandwidth: 20 kHz to 80 MHz
Bandwidth: 50 kHz to 80 MHz
Bandwidth: 4 MHz to 80 MHz
fREF = 2 kHz; fOUT = 70.656 MHz; fLOOP = 100 Hz;
HSTL and/or 3.3 V CMOS Driver
Bandwidth: 10 Hz to 30 MHz
Bandwidth: 5 kHz to 20 MHz
Bandwidth: 12 kHz to 20 MHz
Bandwidth: 10 kHz to 400 kHz
Bandwidth: 100 kHz to 10 MHz
fREF = 25 MHz; fOUT = 1 GHz; fLOOP = 500 Hz
HSTL Driver
Bandwidth: 100 Hz to 500 MHz (Broadband)
Bandwidth: 12 kHz to 20 MHz
Bandwidth: 20 kHz to 80 MHz
Data Sheet
Min
Typ
Max
Unit
334
321
319
277
185
fs rms
fs rms
fs rms
fs rms
fs rms
298
285
286
252
183
fs rms
fs rms
fs rms
fs rms
fs rms
354
301
321
290
177
fs rms
fs rms
fs rms
fs rms
fs rms
306
293
313
283
166
fs rms
fs rms
fs rms
fs rms
fs rms
316
302
324
292
171
fs rms
fs rms
fs rms
fs rms
fs rms
3.22
ps
rms
fs rms
fs rms
fs rms
fs rms
338
324
278
210
1.71
343
338
Rev. B | Page 14 of 104
ps
rms
fs rms
fs rms
Test Conditions/Comments
Data Sheet
AD9558
Jitter generation (random jitter) uses 19.2 MHz TCXO for system clock input.
Table 18.
Parameter
JITTER GENERATION
fREF = 19.44 MHz; fOUT = 644.53 MHz; fLOOP = 0.1 Hz
HSTL Driver
Bandwidth: 5 kHz to 20 MHz
Bandwidth: 12 kHz to 20 MHz
Bandwidth: 20 kHz to 80 MHz
Bandwidth: 50 kHz to 80 MHz
Bandwidth: 16 MHz to 320 MHz
fREF = 19.44 MHz; fOUT = 693.48 MHz; fLOOP = 0.1 Hz
HSTL Driver
Bandwidth: 5 kHz to 20 MHz
Bandwidth: 12 kHz to 20 MHz
Bandwidth: 20 kHz to 80 MHz
Bandwidth: 50 kHz to 80 MHz
Bandwidth: 16 MHz to 320 MHz
fREF = 19.44 MHz; fOUT = 312.5 MHz; fLOOP = 0.1 Hz
HSTL Driver
Bandwidth: 5 kHz to 20 MHz
Bandwidth: 12 kHz to 20 MHz
Bandwidth: 20 kHz to 80 MHz
Bandwidth: 50 kHz to 80 MHz
Bandwidth: 4 MHz to 80 MHz
fREF = 25 MHz; fOUT = 161.1328 MHz; fLOOP = 0.1 Hz
HSTL Driver
Bandwidth: 5 kHz to 20 MHz
Bandwidth: 12 kHz to 20 MHz
Bandwidth: 20 kHz to 80 MHz
Bandwidth: 50 kHz to 80 MHz
Bandwidth: 4 MHz to 80 MHz
fREF = 2 kHz; fOUT = 70.656 MHz; fLOOP = 0.1 Hz
HSTL and/or 3.3 V CMOS Driver
Bandwidth: 10 Hz to 30 MHz
Bandwidth: 12 kHz to 20 MHz
Bandwidth: 10 kHz to 400 kHz
Bandwidth: 100 kHz to 10 MHz
Min
Typ
Max
Unit
402
393
391
347
179
fs rms
fs rms
fs rms
fs rms
fs rms
379
371
371
335
175
fs rms
fs rms
fs rms
fs rms
fs rms
413
404
407
358
142
fs rms
fs rms
fs rms
fs rms
fs rms
399
391
414
376
190
fs rms
fs rms
fs rms
fs rms
fs rms
970
404
374
281
fs rms
fs rms
fs rms
fs rms
Rev. B | Page 15 of 104
Test Conditions/Comments
System clock doubler enabled; high phase
margin mode enabled; Register 0x0405 = 0x20;
Register 0x0403 = 0x07; Register 0x0400 = 0x81;
in cases where multiple driver types are listed,
both driver types were tested at those conditions,
and the one with higher jitter is quoted, although
there is usually not a significant jitter difference
between the driver types
AD9558
Data Sheet
ABSOLUTE MAXIMUM RATINGS
Table 19.
Parameter
Analog Supply Voltage (AVDD)
Digital Supply Voltage (DVDD)
Digital I/O Supply Voltage (DVDD3)
Analog Supply Voltage (AVDD3)
Maximum Digital Input Voltage
Storage Temperature Range
Operating Temperature Range
Lead Temperature
(Soldering 10 sec)
Junction Temperature
Rating
2V
2V
3.6 V
3.6 V
−0.5 V to DVDD3 + 0.5 V
−65°C to +150°C
−40°C to +85°C
300°C
Stresses above 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.
ESD CAUTION
150°C
Rev. B | Page 16 of 104
Data Sheet
AD9558
64
63
62
61
60
59
58
57
56
55
54
53
52
51
50
49
DVDD3
M6
M5
M4
M3
M2
M1
M0
DVDD
DVDD
REFB
REFB
DVDD3
DVDD3
REFD
REFD
PIN CONFIGURATION AND FUNCTION DESCRIPTIONS
PIN 1
INDICATOR
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
AD9558
TOP VIEW
(Not to Scale)
48
47
46
45
44
43
42
41
40
39
38
37
36
35
34
33
REFC
REFC
DVDD3
DVDD3
REFA
REFA
SYNC
M7
PINCONTROL
RESET
AVDD
AVDD
NC
AVDD
NC
LF_VCO2
NOTES
1. NC = NO CONNECT. DO NOT CONNECT TO THIS PIN.
2. THE EXPOSED PAD MUST BE CONNECTED TO GROUND (VSS).
09758-002
AVDD
OUT2
OUT2
OUT1
OUT1
AVDD
OUT5
OUT5
AVDD3
AVDD3
OUT0
OUT0
AVDD
AVDD
AVDD3
LDO_VCO2
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
IRQ
SCLK/SCL
SDIO/SDA
SDO
CS
DVDD
AVDD
XOA
XOB
AVDD
NC
AVDD
OUT4
OUT4
OUT3
OUT3
Figure 2. Pin Configuration
Table 20. Pin Function Descriptions
Pin No.
1
2
Mnemonic
IRQ
SCLK/SCL
Input/
Output
O
I
Pin Type
3.3 V CMOS
3.3 V CMOS
3
SDIO/SDA
I/O
3.3 V CMOS
4
SDO
O
3.3 V CMOS
CS
I
3.3 V CMOS
5
E
A
6, 55, 56
7
8
DVDD
AVDD
XOA
I
I
I
Power
Power
Differential
input
9
XOB
I
Differential
input
10
11
12, 17,
22, 29,
30
AVDD
NC
AVDD
I
I
I
Power
Power
Description
Interrupt Request Line.
Serial Programming Clock (SCLK) in SPI Mode. Data clock for serial programming.
Open-Collector Serial Clock Pin (SCL) in I2C Mode. Requires a pull-up resistor, usually
2.2 kΩ; the resistor size depends on the number of I2C devices on the bus.
Serial Data Input/Output (SDIO) in SPI Mode. When the device is in 4-wire SPI mode,
data is written via this pin. In 3-wire SPI mode, both data reads and writes occur on
this pin. There is no internal pull-up/pull-down resistor on this pin.
Open-Collector Serial Data Pin (SDA) in I2C Mode. Requires a pull-up resistor, usually
2.2 kΩ; the resistor size depends on the number of I2C devices on the bus.
Serial Data Output. Use this pin to read data in 4-wire mode. There is no internal pull-up/
pull-down resistor on this pin. This pin is high impedance in the default 3-wire mode.
Chip Select (SPI), Active Low. When programming a device, this pin must be held low.
In systems where more than one AD9558 is present, this pin enables individual
programming of each AD9558. This pin has an internal 10 kΩ pull-up resistor.
1.8 V Digital Supply.
1.8 V Analog (SYSCLK) Power Supply.
System Clock Input. XOA contains internal dc biasing and should be ac-coupled with
a 0.01 μF capacitor, except when using a crystal. If a crystal is used, connect the crystal
across XOA and XOB. Single-ended 1.8 V CMOS is also an option but can introduce
a spur if the duty cycle is not 50%. When using XOA as a single-ended input, connect
a 0.01 μF capacitor from XOB to ground.
Complementary System Clock Input. Complementary signal to XOA. XOB contains
internal dc biasing and should be ac-coupled with a 0.01 μF capacitor, except when
using a crystal. If a crystal is used, connect the crystal across XOA and XOB.
1.8 V Analog (VCO) Power Supply.
No Connection. Do not connect to this pin.
1.8 V Analog (Output Driver) Power Supply.
Rev. B | Page 17 of 104
AD9558
Data Sheet
Pin No.
13
Mnemonic
OUT4
Input/
Output
O
14
OUT4
O
OUT3
O
OUT3
O
OUT2
O
OUT2
O
OUT1
O
OUT1
O
OUT5
O
24
OUT5
O
25, 26
27
AVDD3
OUT0
I
O
28
OUT0
O
31
32
AVDD3
LDO_VCO2
I
I
Power
LDO bypass
33
LF_VCO2
I/O
Loop filter
34
35
36
37, 38
39
NC
AVDD
NC
AVDD
RESET
I
Power
I
I
Power
3.3 V CMOS
40
PINCONTROL
I
3.3 V CMOS
41
M7
I/O
3.3 V CMOS
I
3.3 V CMOS
E
15
A
E
16
A
18
E
19
A
20
E
21
A
23
42
A
E
A
E
A
E
SYNC
A
E
Pin Type
HSTL, LVDS, or
1.8 V CMOS
HSTL, LVDS,
or 1.8 V CMOS
HSTL, LVDS, or
1.8 V CMOS
HSTL, LVDS, or
1.8 V CMOS
HSTL, LVDS, or
1.8 V CMOS
HSTL, LVDS, or
1.8 V CMOS
HSTL, LVDS, or
1.8 V CMOS
HSTL, LVDS,
or 1.8 V CMOS
HSTL, LVDS,
1.8 V CMOS,
3.3 V CMOS
HSTL, LVDS,
1.8 V CMOS,
3.3 V CMOS
Power
HSTL, LVDS,
1.8 V CMOS,
3.3 V CMOS
HSTL, LVDS,
1.8 V CMOS,
3.3 V CMOS
Description
Complementary Output 4. This output can be configured as HSTL, LVDS, or single-ended
1.8 V CMOS.
Output 4. This output can be configured as HSTL, LVDS, or single-ended 1.8 V CMOS.
LVPECL levels can be achieved by ac coupling and using the Thevenin-equivalent
termination as described in the Input/Output Termination Recommendations section.
Complementary Output 3. This output can be configured as HSTL, LVDS,
or single-ended 1.8 V CMOS.
Output 3. This output can be configured as HSTL, LVDS, or single-ended 1.8 V CMOS.
LVPECL levels can be achieved by ac coupling and using the Thevenin-equivalent
termination as described in the Input/Output Termination Recommendations section.
Complementary Output 2. This output can be configured as HSTL, LVDS, or
single-ended 1.8 V CMOS.
Output 2. This output can be HSTL, LVDS, or single-ended 1.8 V CMOS.
LVPECL levels can be achieved by ac coupling and using the Thevenin-equivalent
termination as described in the Input/Output Termination Recommendations section.
Complementary Output 1. This output can be configured as HSTL, LVDS, or
single-ended 1.8 V CMOS.
Output 1. This output can be configured as HSTL, LVDS, or single-ended 1.8 V CMOS.
LVPECL levels can be achieved by ac coupling and using the Thevenin-equivalent
termination as described in the Input/Output Termination Recommendations section.
Complementary Output 5. This output can be configured as HSTL, LVDS, or single-ended
1.8 V or 3.3 V CMOS.
Output 5. This output can be configured as HSTL, LVDS, or single-ended 1.8 V or 3.3 V
CMOS. LVPECL levels can be achieved by ac coupling and by using the Theveninequivalent termination as described in the Input/Output Termination
Recommendations section.
3.3 V Analog (Output Driver) Power Supply.
Complementary Output 0. This output can be configured as HSTL, LVDS, or singleended 1.8 V or 3.3 V CMOS.
Output 0. This output can be configured as HSTL, LVDS, or single-ended 1.8 V or 3.3 V
CMOS. LVPECL levels can be achieved by ac coupling and by using the Theveninequivalent termination as described in the Input/Output Termination
Recommendations section.
3.3V Analog (VCO 2) Power Supply.
Output PLL Loop Filter Voltage Regulator. Connect a 0.47 μF capacitor from this pin
to ground. This pin is also the ac ground reference for the integrated output PLL
external loop filter.
Loop Filter Node for the Output PLL. Connect an external 6.8 nF capacitor from this
pin to Pin 32 (LDO_VCO2).
No Connect. There is no internal connection for this pin.
1.8 V Analog (APLL) Power Supply.
No Connect. There is no internal connection for this pin.
1.8 V Analog (DCO and TDC) Power Supplies.
Chip Reset. When this active low pin is asserted, the chip goes into reset. This pin has
an internal 50 kΩ pull-up resistor.
Pin Program Mode Enable Pin. When pulled high during startup, this pin enables pin
programming of the AD9558 configuration during startup. If this pin is low during
startup, the user must program the part via the serial port, or use values that are
stored in the EEPROM.
Configurable I/O Pin. Along with pins M6 through M0, this pin is configured through
the AD9558 register space.
Clock Distribution Synchronization Pin. When this pin is activated, output drivers are
held static and then synchronized on a low-to-high transition of this pin. This pin is
used to arm the frame sync function when frame sync mode is enabled and has an
internal 60 kΩ pull-up resistor.
Rev. B | Page 18 of 104
Data Sheet
Pin No.
43
Mnemonic
REFA
Input/
Output
I
REFA
I
44
A
Differential
input
Power
I
REFC
I
Differential
input
I
Differential
input
Differential
input
48
REFC
A
E
REFD
I
REFD
I
REFB
I
REFB
I
M0, M1, M2,
M3, M4, M5,
M6
I/O
Differential
input
3.3 V CMOS
DVDD3
VSS
I
O
Power
Exposed pad
49
50
A
E
53
57, 58,
59, 60,
61, 62,
63
64
EP
E
Pin Type
Differential
input
DVDD3
45, 46,
51, 52
47
54
AD9558
A
E
Differential
input
Differential
input
Description
Reference A Input. This internally biased input is typically ac-coupled and, when
configured as such, can accept any differential signal with single-ended swing up
to 3.3 V. If dc-coupled, input can be LVPECL, LVDS, or single-ended CMOS.
Complementary Reference A Input. This pin is the complementary input to Pin 43.
3.3 V Digital (Reference Input) Power Supply.
Reference C Input. This internally biased input is typically ac-coupled and, when
configured as such, can accept any differential signal with single-ended swing up
to 3.3 V. If dc-coupled, input can be LVPECL, LVDS, or single-ended CMOS.
Complementary Reference C Input. This pin is the complementary input to Pin 47.
Reference D Input. This internally biased input is typically ac-coupled and, when
configured as such, can accept any differential signal with single-ended swing up
to 3.3 V. If dc-coupled, input can be LVPECL, LVDS, or single-ended CMOS.
Complementary Reference D Input. This pin is the complementary input to Pin 49.
Reference B Input. This internally biased input is typically ac-coupled and, when
configured as such, can accept any differential signal with single-ended swing up
to 3.3 V. If dc-coupled, input can be LVPECL, LVDS, or single-ended CMOS.
Complementary Reference B Input. This pin is the complementary input to Pin 53.
Configurable I/O Pins. These pins are configured under program control. The M7 pin
(Pin 41) is the last pin of this group.
3.3 V Digital Supply.
The exposed pad must be connected to ground (VSS).
Rev. B | Page 19 of 104
AD9558
Data Sheet
TYPICAL PERFORMANCE CHARACTERISTICS
fR = input reference clock frequency; fOUT = output clock frequency; fSYS = SYSCLK input frequency; fS = internal system clock frequency;
LF = SYSCLK PLL internal loop filter used. AVDD, AVDD3, and DVDD at nominal supply voltage; fS = 786.432 MHz, unless otherwise noted.
–60
–60
INTEGRATED RMS JITTER (12kHz TO 20MHz): 285fs
–70
–80
–80
–90
–100
–110
–120
–130
–90
–100
–110
–120
–130
–140
–140
–150
–150
–160
100
1k
10k
100k
1M
10M
100M
FREQUENCY OFFSET (Hz)
–160
100
10k
100k
1M
10M
100M
FREQUENCY OFFSET (Hz)
Figure 3. Absolute Phase Noise (Output Driver = HSTL),
fR = 19.44 MHz, fOUT = 622.08 MHz,
DPLL Loop BW = 50 Hz, fSYS = 49.152 MHz Crystal
Figure 5. Absolute Phase Noise (Output Driver = HSTL),
fR = 19.44 MHz, fOUT = 693.482991 MHz,
DPLL Loop BW = 50 Hz, fSYS = 49.152 MHz Crystal
–60
–70
INTEGRATED RMS JITTER (12kHz TO 20MHz): 301fs
INTEGRATED RMS JITTER (12kHz TO 20MHz): 320fs
–70
–80
–80
PHASE NOISE (dBc/Hz)
–90
–90
–100
–110
–120
–130
–100
–110
–120
–130
–140
–140
–160
100
1k
10k
100k
1M
10M
100M
FREQUENCY OFFSET (Hz)
Figure 4. Absolute Phase Noise (Output Driver = HSTL),
fR = 19.44 MHz, fOUT = 644.53125 MHz,
DPLL Loop BW = 50 Hz, fSYS = 49.152 MHz Crystal
–160
100
1k
10k
100k
1M
10M
100M
FREQUENCY OFFSET (Hz)
Figure 6. Absolute Phase Noise (Output Driver = HSTL),
fR = 19.44 MHz, fOUT = 174.703 MHz,
DPLL Loop BW = 1 kHz, fSYS = 49.152 MHz Crystal
Rev. B | Page 20 of 104
09758-006
–150
–150
09758-004
PHASE NOISE (dBc/Hz)
1k
09758-005
PHASE NOISE (dBc/Hz)
–70
09758-003
PHASE NOISE (dBc/Hz)
INTEGRATED RMS JITTER (12kHz TO 20MHz): 296fs
Data Sheet
AD9558
–80
–60
INTEGRATED RMS JITTER (12kHz TO 20MHz): 302fs
INTEGRATED RMS JITTER (12kHz TO 20MHz): 393fs
–70
–90
PHASE NOISE (dBc/Hz)
PHASE NOISE (dBc/Hz)
–80
–100
–110
–120
–130
–140
–90
–100
–110
–120
–130
–140
–150
10k
100k
1M
10M
100M
FREQUENCY OFFSET (Hz)
–160
09758-007
1k
10
1M
10M
100M
INTEGRATED RMS JITTER (12kHz TO 20MHz): 371s
–70
–80
–80
–100
–110
–120
–130
–90
–100
–110
–120
–130
–140
–140
–150
–150
10k
100k
1M
10M
100M
FREQUENCY OFFSET (Hz)
–160
09758-008
1k
10
100
1k
10k
100k
1M
10M
100M
FREQUENCY OFFSET (Hz)
Figure 8. Absolute Phase Noise (Output Driver = HSTL),
fR = 2 kHz, fOUT = 125 MHz,
DPLL Loop BW = 100 Hz, fSYS = 49.152 MHz Crystal
09758-011
PHASE NOISE (dBc/Hz)
–90
Figure 11. Absolute Phase Noise (Output Driver = HSTL),
fR = 19.44 MHz, fOUT = 693.482991 MHz,
DPLL Loop BW = 0.1 Hz, fSYS = 19.2 MHz TCXO
–70
–60
INTEGRATED RMS JITTER (12kHz TO 20MHz): 404fs
INTEGRATED RMS JITTER (12kHz TO 20MHz): 343fs
–70
–80
–80
PHASE NOISE (dBc/Hz)
–90
–90
–100
–110
–120
–130
–100
–110
–120
–130
–140
–140
–150
1k
10k
100k
1M
10M
FREQUENCY OFFSET (Hz)
100M
09758-009
–150
–160
Figure 9. Absolute Phase Noise (Output Driver = HSTL),
fR = 25 MHz, fOUT = 1 GHz,
DPLL Loop BW = 500 Hz, fSYS = 49.152 MHz Crystal
10
100
1k
10k
100k
1M
10M
FREQUENCY OFFSET (Hz)
Figure 12. Absolute Phase Noise (Output Driver = HSTL),
fR = 19.44 MHz, fOUT = 312.5 MHz,
DPLL Loop BW = 0.1 Hz, fSYS = 19.2 MHz TCXO
Rev. B | Page 21 of 104
100M
09758-012
PHASE NOISE (dBc/Hz)
100k
–60
INTEGRATED RMS JITTER (12kHz TO 20MHz): 308fs
PHASE NOISE (dBc/Hz)
10k
Figure 10. Absolute Phase Noise (Output Driver = HSTL),
fR = 19.44 MHz, fOUT = 644.53 MHz,
DPLL Loop BW = 0.1 Hz, fSYS = 19.2 MHz TCXO
–70
–160
100
1k
FREQUENCY OFFSET (Hz)
Figure 7. Absolute Phase Noise (Output Driver = 3.3.V CMOS),
fR = 19.44 MHz, fOUT = 161.1328125 MHz,
DPLL Loop BW = 100 Hz, fSYS = 49.152 MHz Crystal
–160
100
100
09758-010
–150
–160
100
AD9558
Data Sheet
DIFFERENTIAL PEAK-TO-PEAK AMPLITUDE (V)
–70
INTEGRATED RMS JITTER (12kHz TO 20MHz): 391fs
–80
–100
–110
–120
–130
–140
–150
–80
PHASE NOISE (dBc/Hz)
–90
–100
–110
–120
–130
–140
1k
10k
100k
1M
10M
100M
FREQUENCY OFFSET (Hz)
1300
09758-116
0.9
LVDS BOOST MODE
0.8
0.7
LVDS DEFAULT
0.6
0.5
0.4
09758-014
–150
1.0
0
100
200
300
400
500
600
700
800
FREQUENCY (MHz)
09758-117
DIFFERENTIAL PEAK-TO-PEAK AMPLITUDE (V)
INTEGRATED RMS JITTER (12kHz TO 20MHz): 395fs
100
1100
Figure 16. Amplitude vs. Toggle Rate,
HSTL Mode (LVPECL-Compatible Mode)
–70
10
1200
FREQUENCY (MHz)
Figure 13. Absolute Phase Noise (Output Driver = 3.3 V CMOS),
fR = 19.44 MHz, fOUT =161.1328125 MHz,
DPLL Loop BW = 0.1 Hz, fSYS = 19.2 MHz TCXO
–160
900
1.0
1000
FREQUENCY OFFSET (Hz)
1.1
800
100M
1.2
700
10M
1.3
600
1M
1.4
500
100k
1.5
300
10k
1.6
400
1k
1.7
200
100
1.8
0
10
09758-013
–160
1.9
100
PHASE NOISE (dBc/Hz)
–90
2.0
Figure 17. Amplitude vs. Toggle Rate, LVDS
Figure 14. Absolute Phase Noise (Output Driver = 1.8 V CMOS),
fR = 2 kHz, fOUT = 70.656 MHz,
DPLL Loop BW = 0.1 Hz, fSYS = 19.2 MHz TCXO
3.5
–60
INTEGRATED RMS JITTER (12kHz TO 20MHz): 388fs
3.3V CMOS
PEAK-TO-PEAK AMPLITUDE (V)
–70
–90
–100
–110
–120
–130
–140
3.0
2.5
2.0
1.8V CMOS
1.5
–160
10
100
1k
10k
100k
1M
10M
FREQUENCY OFFSET (Hz)
100M
Figure 15. Absolute Phase Noise (Output Driver = HSTL),
fR = 19.44 MHz, fOUT = 644.53 MHz, fSYS = 19.2 MHz TCXO
Holdover Mode
1.0
0
50
100
150
200
250
FREQUENCY (MHz)
Figure 18. Amplitude vs. Toggle Rate with 10 pF Load,
3.3 V (Strong Mode) and 1.8 V CMOS
Rev. B | Page 22 of 104
300
09758-118
–150
09758-016
PHASE NOISE (dBc/Hz)
–80
AD9558
70
3.0
60
2.5
50
2.0
1.5
30
20
0.5
10
0
20
10
30
40
50
60
70
80
FREQUENCY (MHz)
Figure 19. Amplitude vs. Toggle Rate with 10 pF Load,
3.3 V (Weak Mode) CMOS
0
0
50
100
150
200
FREQUENCY (MHz)
Figure 22. Power Consumption vs. Frequency, One CMOS Driver on Output
Driver Power Supply Only (Pin 12, Pin 17, Pin 22, and Pin 29) for 1.8 V CMOS Mode,
or on Pin 25 and Pin 26 for 3.3 V CMOS Mode
1.0
110
105
100
95
90
85
80
75
70
65
60
55
50
45
40
35
30
DIFFERENTIAL AMPLITUDE (V)
0.8
OUTPUT CHANNEL 1 OR 2:
BOTH HSTL DRIVERS ENABLED
OUTPUT CHANNEL 1 OR 2:
ONE HSTL DRIVER ENABLED
OUTPUT CHANNEL 0 OR 3:
HSTL DRIVER ENABLED
0.6
0.4
0.2
0
–0.2
–0.4
–0.6
0
250
500
750
FREQUENCY (MHz)
1000
1250
–1.0
–1
0
1
3
2
4
5
TIME (ns)
09758-123
–0.8
09758-120
POWER (mW)
40
1.0
0
1.8V CMOS MODE
3.3V CMOS STRONG MODE
3.3V CMOS WEAK MODE
09758-122
POWER (mW)
3.5
09758-119
PEAK-TO-PEAK AMPLITUDE (V)
Data Sheet
Figure 23. Output Waveform, HSTL (400 MHz)
Figure 20. Power Consumption vs. Frequency,
HSTL Mode (Single Channel) on 1.8 V Output Driver Power Supply Only
(Pin 12, Pin 17, Pin 22, and Pin29)
0.4
65
60
0.3
DIFFERENTIAL AMPLITUDE (V)
OUTPUT CHANNEL 1 OR 2:
BOTH LVDS DRIVERS ENABLED
55
50
POWER (mW)
45
40
35
OUTPUT CHANNEL 1 OR 2:
ONE LVDS DRIVER ENABLED
30
25
OUTPUT CHANNEL 0 OR 3:
LVDS DRIVER ENABLED
20
15
10
0.2
0.1
0
–0.1
–0.2
–0.3
0
100
200
300
400
500
FREQUENCY (MHz)
600
700
800
Figure 21. LVDS Power Consumption vs. Frequency on 1.8 V Output Driver
Power Supply Only (Pin 12, Pin 17, Pin 22, and Pin 29)
Rev. B | Page 23 of 104
–0.4
–1
0
1
2
3
TIME (ns)
Figure 24. Output Waveform, LVDS (400 MHz)
4
09758-124
0
09758-121
5
AD9558
Data Sheet
3
3.4
0
3.0
–3
2.6
–6
1.8
1.4
1.0
0.2
0
1
2
3
4
5
6 7 8
TIME (ns)
9
10 11 12 13 14 15
1.9
LOOP BW = 2kHz;
HIGH PHASE MARGIN;
PEAKING: 0.097dB; –3dB: 1.23kHz
–27
LOOP BW = 5kHz;
HIGH PHASE MARGIN;
PEAKING: 0.14dB; –3dB: 4.27kHz
100
1k
10k
FREQUENCY OFFSET (Hz)
100k
3
0
1.5
–3
–6
LOOP GAIN (dB)
1.3
1.1
0.9
0.7
0.5
–9
–12
–15
–18
–21
2pF LOAD
10pF LOAD
0.3
LOOP BW = 100Hz;
NORMAL PHASE MARGIN;
PEAKING: 0.09dB; –3dB: 117Hz
LOOP BW = 2kHz;
NORMAL PHASE MARGIN;
PEAKING: 1.6dB; –3dB: 2.69kHz
–24
0.1
–27
0
1
2
3
4
5
6 7 8
TIME (ns)
9
10 11 12 13 14 15
–30
09758-127
–0.1
–1
10
Figure 26. Output Waveform, 1.8 V CMOS (100 MHz)
3.2
2.8
2.4
2.0
1.6
1.2
0.8
25
35
45
55
65
75
85
TIME (ns)
95
09758-128
0.4
15
1k
10k
FREQUENCY OFFSET (Hz)
100k
Figure 29. Closed-Loop Transfer Function for 100 Hz and 2 kHz Loop
Bandwidth Settings; Normal Phase Margin Loop Filter Setting
2pF LOAD
10pF LOAD
5
100
09758-230
AMPLITUDE (V)
–24
1.7
0
–5
LOOP BW = 100Hz;
HIGH PHASE MARGIN;
PEAKING: 0.06dB; –3dB: 69Hz
Figure 28. Closed-Loop Transfer Function for 100 Hz, 2 kHz, and 5 kHz Loop
Bandwidth Settings; High Phase Margin Loop Filter Setting
(This is compliant with Telcordia GR-253 jitter transfer test for loop
bandwidths < 2 kHz.)
Figure 25. Output Waveform,
3.3 V CMOS (100 MHz, Strong Mode)
AMPLITUDE (V)
–18
–30
10
09758-126
–0.2
–1
–15
–21
2pF LOAD
10pF LOAD
0.6
–9
–12
09758-129
LOOP GAIN (dB)
AMPLITUDE (V)
2.2
Figure 27. Output Waveform, 3.3 V CMOS (20 MHz, Weak Mode)
Rev. B | Page 24 of 104
Data Sheet
AD9558
INPUT/OUTPUT TERMINATION RECOMMENDATIONS
10pF
0.1µF
XOA
0.1µF
XOB
10pF
Figure 30. AC-Coupled LVDS or HSTL Output Driver
(100 Ω resistor can go on either side of decoupling capacitors and should be
as close as possible to the destination receiver.)
Figure 33. System Clock Input (XOA, XOB) in Crystal Mode
(The recommended CLOAD = 10 pF is shown. The values of the 10 pF shunt
capacitors shown here should equal the CLOAD of the crystal.)
Z0 = 50Ω
SINGLE-ENDED
(NOT COUPLED)
HSTL OR
LVDS
100Ω
3.3V
CMOS
TCXO
LVDS OR 1.8V HSTL
HIGH IMPEDANCE
DIFFERENTIAL
RECEIVER
82Ω
82Ω
3.3V
LVPECL
SINGLE-ENDED
(NOT COUPLED)
1.8V
HSTL
0.1µF
Z0 = 50Ω
127Ω
127Ω
09758-132
AD9557/
AD9558
XOA
150Ω
AD9557/
AD9558
XOB
VS = 3.3V
Z0 = 50Ω
0.1µF
Figure 34. System Clock Input (XOA, XOB) When Using a TCXO/OCXO with
3.3 V CMOS Output
Figure 31. DC-Coupled LVDS or HSTL Output Driver
0.1µF
300Ω
0.1µF
Z0 = 50Ω
09758-131
AD9557/
AD9558
AD9557/
AD9558
09758-134
HSTL OR
LVDS
10MHz TO 50MHz FUNDAMENTAL
AT CUT CRYSTAL WITH
10pF LOAD CAPACITANCE
09758-133
100Ω
09758-130
AD9557/
AD9558
DOWNSTREAM
DEVICE
WITH HIGH
IMPEDANCE
INPUT AND
INTERNAL
DC BIAS
Figure 32. Interfacing the HSTL Driver to a 3.3 V LVPECL Input
(This method incorporates impedance matching and dc biasing for bipolar
LVPECL receivers. If the receiver is self-biased, the termination scheme shown
in Figure 30 is recommended.)
Rev. B | Page 25 of 104
AD9558
Data Sheet
GETTING STARTED
CHIP POWER MONITOR AND STARTUP
The AD9558 monitors the voltage on the power supplies at
power-up. When DVDD3 is greater than 2.35 V ± 0.1 V and
DVDD and AVDD are greater than 1.4 V ± 0.05 V, the device
generates a 20 ms reset pulse. The power-up reset pulse is internal
and independent of the RESET pin. This internal power-up reset
sequence eliminates the need for the user to provide external power
supply sequencing. Within 45 ns after the leading edge of the
internal reset pulse, the M7 to M0 multifunction pins behave
as high impedance digital inputs and continue to do so until
programmed otherwise. The delay on the M7 to M0 pin function
change is 45 ns for pin reset or soft reset.
E
A
A
During a device reset (either via the power-up reset pulse or the
RESET pin), the multifunction pins (M7 to M0) behave as high
impedance inputs, but upon removal of the reset condition,
level-sensitive latches capture the logic pattern present on the
multifunction pins.
E
A
A
MULTIFUNCTION PINS AT RESET/POWER-UP
The AD9558 requires the user to supply the desired logic state
to the PINCONTROL pin, as well as to the M7 to M0 pins. If
PINCONTROL is high, the part is in hard pin programming
mode. See the Pin Program Function Description section for
details on hard pin programming.
At startup, there are three choices for the M7 to M0 pins: pullup, pull-down, and floating. If the PINCONTROL pin is low,
the M7 to M0 pins determine the following configurations:
•
•
Following a reset, the M1 and M0 pins determine whether
the serial port interface behaves according to the SPI or I2C
protocol. Specifically, M0 = M1 = low selects the SPI
interface, and any other value selects the I2C port. The 3level logic of M1 and M0 allows the user to select eight
possible I2C addresses (see Table 24).
The M3 and M2 pins select which of the eight possible
EEPROM profiles are loaded, or if the EEPROM loading is
bypassed. Leaving M3 and M2 floating at startup bypasses
the EEPROM loading, and the factory defaults are used
instead (see Table 22).
DEVICE REGISTER PROGRAMMING WHEN USING
A REGISTER SETUP FILE
The evaluation software contains a programming wizard and
a convenient graphical user interface that assists the user in
determining the optimal configuration for the DPLL, APLL,
and SYSCLK based on the desired input and output frequencies.
It generates a register setup file with a .STP extension that is
easily readable using a text editor.
After using the evaluation software to create the setup file, use
the following sequence to program the AD9557 once:
1.
2.
Register 0x0A01 = 0x20 (set user free run mode).
Register 0x0A02 = 0x02 (hold outputs in static SYNC).
(Skip this step if using SYNC on DPLL phase lock or SYNC
on DPLL frequency lock. See Register 0x0500[1:0].)
3. Register 0x0405 = 0x20 (clear APLL VCO calibration).
4. Write the register values in the STP file from Address 0x0000
to Address 0x032E.
5. Register 0x0005 = 0x01 (update all registers).
6. Write the rest of the registers in the STP file, starting at
Address 0x0400.
7. Register 0x0405 = 0x21 (calibrate APLLon next I/O update).
8. Register 0x0403 = 0x07 (configure APLL).
9. Register 0x0400 = 0x81 (configure APLL).
10. Register 0x0005 = 0x01 (update all registers).
11. Register 0x0A01[5] = 0b (clear user free run mode).
12. Register 0x0005 = 0x01 (update all registers).
REGISTER PROGRAMMING OVERVIEW
This section provides an overview of the register blocks in the
AD9558, describing what they do and why they are important.
Registers Differing from Defaults for Optimal Performance
Ensure that the following registers are programmed to the listed
values for optimal performance:
•
•
•
Register 0x0405[7:4] = 0x2
Register 0x0403 = 0x07
Register 0x0400 = 0x81
If the silicon revision (Register 0x000A) equals 0x21 or higher,
the values listed here are already the default values.
Rev. B | Page 26 of 104
Data Sheet
AD9558
Program the System Clock and Free Run Tuning Word
Program the Clock Distribution Outputs
The system clock multiplier (SYSCLK) parameters are at
Register 0x0100 to Register 0x0108, and the free run tuning
word is at Register 0x0300 to Register 0x0303. Use the following
steps for optimal performance:
The APLL output goes to the clock distribution block. The
clock distribution parameters reside in Register 0x0500 to
Register 0x0509. They include the following:
1.
2.
3.
4.
5.
Set the system clock PLL input type and divider values.
Set the system clock period.
It is essential to program the system clock period because
many of the AD9558 subsystems rely on this value.
Set the system clock stability timer.
It is highly recommended that the system clock stability
timer be programmed. This is especially important when
using the system clock multiplier and also applies when
using an external system clock source, especially if the
external source is not expected to be completely stable
when power is applied to the AD9558. The system clock
stability timer specifies the amount of time that the system
clock PLL must be locked before the part declares that the
system clock is stable. The default value is 50 ms.
Program the free run tuning word.
The free run frequency of the digital PLL (DPLL) determines
the frequency appearing at the APLL input when free run
mode is selected. The free run tuning word is at Register
0x0300 to Register 0x0303. The correct free run frequency
is required for the APLL to calibrate and lock correctly.
Set user free run mode (Register 0x0A01[5] = 1b).
Initialize and Calibrate the Output PLL (APLL)
The registers controlling the APLL are at Register 0x0400
to Register 0x0408. This low noise, integer-N PLL multiplies
the DPLL output (which is usually 175 MHz to 200 MHz) to a
frequency in the 3.35 GHz to 4.05 GHz range. After the system
clock is configured and the free run tuning word is set in
Register 0x0300 to Register 0x0303, the user can set the manual
APLL VCO calibration bit (Register 0x0405[0]) and issue an I/O
update (Register 0x0005[0]). This process performs the APLL
VCO calibration. VCO calibration ensures that, at the time of
calibration, the dc control voltage of the APLL VCO is centered in
the middle of its operating range. It is important to remember the
following points when calibrating the APLL VCO:
•
•
•
•
•
The system clock must be stable.
The APLL VCO must have the correct frequency from the
30-bit DCO (digitally controlled oscillator) during
calibration.
The APLL VCO must be recalibrated any time the APLL
frequency changes.
APLL VCO calibration occurs on the low-to-high transition
of the manual APLL VCO calibration bit, and this bit is not
autoclearing. Therefore, this bit must be cleared (and an I/O
update issued) before another APLL calibration is started.
The best way to monitor successful APLL calibration is to
monitor Bit 2 in Register 0x0D01 (APLL lock).
•
•
•
•
•
Output power-down control
Output enable (disabled by default)
Output synchronization
Output mode control
Output divider functionality
See the Clock Distribution section for more information.
Generate the Output Clock
If Register 0x0500[1:0] is programmed for automatic clock
distribution synchronization via the DPLL phase or frequency
lock, the synthesized output signal appears at the clock distribution
outputs. Otherwise, set and then clear the soft sync clock
distribution bit (Register 0x0A02, Bit 1), or use a multifunction
pin input (if programmed for use) to generate a clock
distribution sync pulse, which causes the synthesized output
signal to appear at the clock distribution outputs.
Program the Multifunction Pins (Optional)
This step is required only if the user intends to use any of the
multifunction pins for status or control. The multifunction pin
parameters are at Register 0x0200 to Register 0x0208.
Program the IRQ Functionality (Optional)
This step is required only if the user intends to use the IRQ feature.
The IRQ monitor registers are at Register 0x0D02 to Register
0x0D09. If the desired bits in the IRQ mask registers at Register
0x020A to Register 0x020F are set high, the appropriate IRQ
monitor bit at Register 0x0D02 to Register 0x0D07 is set high
when the indicated event occurs.
Individual IRQ events are cleared by using the IRQ clearing
registers at Register 0x0A04 to Register 0x0A09, or by setting
the clear all IRQs bit (Register 0x0A03[1]) to 1b.
The default values of the IRQ mask registers are such that
interrupts are not generated. The IRQ pin mode default is opendrain NMOS.
Program the Watchdog Timer (Optional)
This step is required only if the user intends to use the watchdog
timer. The watchdog timer control is in Register 0x0210 and
Register 0x0211 and is disabled by default.
The watchdog timer is useful for generating an IRQ after a fixed
amount of time. The timer is reset by setting the clear watchdog
timer bit (Register 0x0A03[0]) to 1b.
Rev. B | Page 27 of 104
AD9558
Data Sheet
Program the Digital Phase-Locked Loop (DPLL)
Program the Reference Profiles
The DPLL parameters reside in Register 0x0300 to
Register 0x032E. They include the following:
The reference profile parameters reside in Register 0x0700 to
Register 0x07E6. The AD9558 evaluation software contains a
wizard that calculates these values based on the user’s input
frequency. See the Reference Profiles section for details on
programming these functions. They include the following:
•
•
•
•
•
Free run frequency
DPLL pull-in range limits
DPLL closed-loop phase offset
Phase slew control (for hitless reference switching)
Tuning word history control (for holdover operation)
Program the Reference Inputs
The reference input parameters reside in Register 0x0600 to
Register 0x0602. See the Reference Clock Input section for
details on programming these functions. They include the
following:
•
•
•
Reference power-down
Reference logic family
Reference priority
•
•
•
•
•
•
•
•
Reference period
Reference period tolerance
Reference validation timer
Selection of high phase margin loop filter coefficients
DPLL loop bandwidth
Reference prescaler (R divider)
Feedback dividers (N1, N2, N3, FRAC1, and MOD1)
Phase and frequency lock detector controls
Generate the Reference Acquisition
After the registers are programmed, the user can clear the user
freerun bit (Register 0x0A01[5]) and issue an I/O update, using
Register 0x0005[0] to invoke all of the register settings that are
programmed up to this point.
After the registers are programmed, the DPLL locks to the first
available valid reference that has the highest priority.
Rev. B | Page 28 of 104
Data Sheet
AD9558
THEORY OF OPERATION
XO OR XTAL
SYNC
RESET
PINCONTROL
M0 M1 M2 M3 M4 M5 M6 M7 IRQ
XO FREQUENCIES
10MHz TO 180MHz
XTAL: 10MHz TO 50MHz
÷M0 TO ÷M3b ARE
10-BIT INTEGER
DIVIDERS
1
2kHz TO 1.25GHz
SPI/I2C
SPI/I2C
SERIAL PORT
EEPROM
REFA
REFA
÷2
REFB
REFB
÷2
REFC
REFC
÷2
REFD
REFD
÷2
REGISTER
SPACE
ROM
AND
FSM
÷M0
MULTIFUNCTION I/O PINS
(CONTROL AND STATUS
READBACK)
÷2
RF
DIVIDER 1
÷3 TO ÷11
R DIVIDER
(20-BIT)
17-BIT
INTEGER
FRAC1/
÷N1
MOD1
DIGITAL
LOOP
FILTER
OUT31
÷M2
÷M3
FRAME SYNC PULSE
OUT31
OUT41
OUT41
÷M3b
30-BIT
NCO
DIGITAL PLL (DPLL)
OUT11
OUT11
OUT21
×2
TUNING
WORD
CLAMP
AND
HISTORY
OUT01
OUT21
MAX 1.25GHz
RF
DIVIDER 2
÷3 TO ÷11
FREE RUN
TW
DPFD
FRAME SYNC
MODE ONLY
÷M1
SYSTEM
PFD/CP ÷N3
CLOCK
MULTIPLIER
LF
24b/24b
RESOLUTION
×2
OUT51
OUT51
÷N2
OUTPUT PLL (APLL)
REF MONITORING
AUTOMATIC
SWITCHING
AD9558
×2
OUT0
PFD/CP
LF
3.35GHz
TO
4.05GHz
2kHz TO 8kHz FRAME SYNC SIGNAL
LF_VCO2
09758-135
APLL
STATUS
1OUT0, OUT1, OUT2, OUT3, OUT4: 360kHz TO 1.25GHz; OUT5: 352Hz TO 1.25GHz
Figure 35. Detailed Block Diagram
OVERVIEW
The AD9558 provides clocking outputs that are directly related
in phase and frequency to the selected (active) reference, but
with jitter characteristics that are governed by the system clock,
the DCO, and the output PLL (APLL). The AD9558 supports
up to four reference inputs and input frequencies ranging from
2 kHz to 1250 MHz. The core of this product is a digital phaselocked loop (DPLL). The DPLL has a programmable digital loop
filter that greatly reduces jitter that is transferred from the active
reference to the output. The AD9558 supports both manual and
automatic holdover. While in holdover, the AD9558 continues to
provide an output as long as the system clock is present. The
holdover output frequency is a time average of the output
frequency history just prior to the transition to the holdover
condition. The device offers manual and automatic reference
switchover capability if the active reference is degraded or fails
completely. The AD9558 also has adaptive clocking capability
that allows the DPLL divider ratios to be changed while the DPLL
is locked.
The AD9558 has a system clock multiplier, a digital PLL (DPLL),
and an analog PLL (APLL). The input signal goes first to the DPLL,
which performs the jitter cleaning and most of the frequency
translation. The DPLL features a 30-bit digitally controlled
oscillator (DCO) output that generates a signal in the 175 MHz
to 200 MHz range. The DPLL output goes to an analog integer-N
PLL (APLL), which multiplies the signal up to the 3.35 GHz to
4.05 GHz range. That signal is then sent to the clock distribution
section, which has two divide-by-3 to divide-by-11 RF dividers
that are cascaded with 10-bit integer (divide-by-1 to divide-by1024) channel dividers.
The XOA and XOB pins provide the input for the system clock.
These pins accept a reference clock in the 10 MHz to 600 MHz
range, or a 10 MHz to 50 MHz crystal connected directly across
the XOA and XOB pins. The system clock provides the clocks to
the frequency monitors, the DPLL, and internal switching logic.
The AD9558 has six output drivers, arranged into four channels.
Each channel has a dedicated 10-bit programmable post divider.
Channel 0 and Channel 3 have one driver each, and Channel 1
and Channel 2 have two drivers each. Each driver is programmable
either as a single differential or dual single-ended CMOS output.
The clock distribution section operates at up to 1250 MHz.
In differential mode, the output drivers run on a 1.8 V power
supply to offer very high performance with minimal power
consumption. There are two differential modes: LVDS and 1.8 V
HSTL. In 1.8 V HSTL mode, the voltage swing is compatible
with LVPECL. If LVPECL signal levels are required, the designer
can ac-couple the AD9558 output and use Thevenin-equivalent
termination at the destination to drive the LVPECL inputs.
In single-ended mode, each differential output driver can
operate as two single-ended CMOS outputs. OUT0 and OUT5
support either 1.8 V or 3.3 V CMOS operation. OUT1 through
OUT4 support only 1.8 V operation.
Rev. B | Page 29 of 104
AD9558
Data Sheet
REFERENCE CLOCK INPUTS
Reference Validation Timer
Four pairs of pins provide access to the reference clock receivers.
To accommodate input signals with slow rising and falling edges,
both the differential and single-ended input receivers employ
hysteresis. Hysteresis also ensures that a disconnected or
floating input does not cause the receiver to oscillate.
Each reference input has a dedicated validation timer. The
validation timer establishes the amount of time that a previously
faulted reference must remain unfaulted before the AD9558
declares it valid. The timeout period of the validation timer is
programmable via a 16-bit register. The 16-bit number stored
in the validation register represents units of milliseconds (ms),
which yields a maximum timeout period of 65,535 ms.
When configured for differential operation, the input receivers
accommodate either ac- or dc-coupled input signals. The input
receivers are capable of accepting dc-coupled LVDS and 2.5 V
and 3.3 V LVPECL signals. The receiver is internally dc biased
to handle ac-coupled operation, but there is no internal 50 Ω or
100 Ω termination.
When configured for single-ended operation, the input
receivers exhibit a pull-down load of 45 kΩ (typical). Three
user-programmable threshold voltage ranges are available for
each single-ended receiver.
REFERENCE MONITORS
The accuracy of the input reference monitors depends on a
known and accurate system clock period. Therefore, the
functioning of the reference monitors is not operable until the
system clock is stable.
Reference Period Monitor
Each reference input has a dedicated monitor that repeatedly
measures the reference period. The AD9558 uses the reference
period measurements to determine the validity of the reference
based on a set of user-provided parameters in the profile
register area of the register map.
The monitor works by comparing the measured period of a
particular reference input with the parameters stored in the
profile register assigned to that same reference input. The
parameters include the reference period, an inner tolerance,
and an outer tolerance. A 40-bit number defines the reference
period in units of femtoseconds (fs). The 40-bit range allows for
a reference period entry of up to 1.1 ms. A 20-bit number defines
the inner and outer tolerances. The value stored in the register
is the reciprocal of the tolerance specification. For example, a
tolerance specification of 50 ppm yields a register value of
1/(50 ppm) = 1/0.000050 = 20,000 (0x04E20).
The use of two tolerance values provides hysteresis for the
monitor decision logic. The inner tolerance applies to a
previously faulted reference and specifies the largest period
tolerance that a previously faulted reference can exhibit before it
qualifies as nonfaulted. The outer tolerance applies to an already
nonfaulted reference. It specifies the largest period tolerance
that a nonfaulted reference can exhibit before being faulted.
To produce decision hysteresis, the inner tolerance must be less
than the outer tolerance. That is, a faulted reference must meet
tighter requirements to become nonfaulted than a nonfaulted
reference must meet to become faulted.
It is possible to disable the validation timer by programming the
validation timer to 0b. With the validation timer disabled, the
user must validate a reference manually via the manual reference
validation override controls register (Address 0x0A0B).
Reference Validation Override Control
The user also has the ability to override the reference validation
logic and can either force an invalid reference to be treated as
valid, or force a valid reference to be treated as an invalid
reference. These controls are in Register 0x0A0B to Register
0x0A0D.
REFERENCE PROFILES
The AD9558 has an independent profile for each reference input.
A profile consists of a set of device parameters such as the R divider
and N divider, among others The profiles allow the user to
prescribe the specific device functionality that should take effect
when one of the input references becomes the active reference.
The AD9558 evaluation software includes a frequency planning
wizard that can configure the profile parameters, given the
input and output frequencies.
The user should not change a profile that is currently in use
because unpredictable behavior may result. The user can either
select free run or holdover mode or invalidate the reference
input prior to changing it.
REFERENCE SWITCHOVER
An attractive feature of the AD9558 is its versatile reference
switchover capability. The flexibility of the reference switchover
functionality resides in a sophisticated prioritization algorithm
that is coupled with register-based controls. This scheme
provides the user with maximum control over the state machine
that handles reference switchover.
The main reference switchover control resides in the loop
mode register (Address 0x0A01). The REF switchover mode
bits (Register 0x0A01, Bits[4:2]) allow the user to select one of
the five operating modes of the reference switchover state machine,
as follows:
•
•
•
•
•
Rev. B | Page 30 of 104
Automatic revertive mode
Automatic non-revertive mode
Manual with automatic fallback mode
Manual with holdover mode
Full manual mode (without auto-holdover)
Data Sheet
AD9558
In the automatic modes, a fully automatic priority-based algorithm
selects which reference is the active reference. When programmed
for an automatic mode, the device chooses the highest priority
valid reference. When both references have the same priority,
REFA gets preference over REFB. However, the reference position
is used only as a tie-breaker and does not initiate a reference switch.
The following list gives an overview of the five operating modes:





Automatic revertive mode. The device selects the highest
priority valid reference and switches to a higher priority
reference if it becomes available, even if the reference in use
is still valid. In this mode, the user reference is ignored.
Automatic non-revertive mode. The device stays with the
currently selected reference as long as it is valid, even if
a higher priority reference becomes available. The user
reference is ignored in this mode.
Manual with automatic fallback mode. The device uses the
user reference for as long as it is valid. If it becomes invalid,
the reference input with the highest priority is chosen in
accordance with the priority-based algorithm.
Manual with holdover mode. The user reference is the
active reference until it becomes invalid. At that point,
the device automatically goes into holdover.
Manual mode without holdover. The user reference is the
active reference, regardless of whether or not it is valid.
The user also has the option to force the device directly into
holdover or free run operation via the user holdover and user
freerun bits. In free run mode, the free run frequency tuning
word register defines the free run output frequency. In holdover
mode, the output frequency depends on the holdover control
settings (see the Holdover section).
Phase Build-Out Reference Switching
The AD9558 supports phase build-out reference switching,
which is the term given to a reference switchover that
completely masks any phase difference between the previous
reference and the new reference. That is, there is virtually no
phase change detectable at the output when a phase build-out
switchover occurs.
A diagram of the DPLL core of the AD9558 appears in Figure 36.
The phase/frequency detector, feedback path, lock detectors,
phase offset, and phase slew rate limiting that comprise this
second generation DPLL are all digital implementations.
The start of the DPLL signal chain is the reference signal, fR,
which is the frequency of the reference input. A reference
prescaler reduces the frequency of this signal by an integer factor,
R + 1, where R is the 20-bit value stored in the appropriate
profile register and 0 ≤ R ≤ 1,048,575. Therefore, the frequency
at the output of the R-divider (or the input to the time-to-digital
converter (TDC)) is
fTDC 
A TDC samples the output of the R-divider. The TDC/PFD
produces a time series of digital words and delivers them to the
digital loop filter. The digital loop filter offers the following
advantages:




The DPLL includes a feedback divider that causes the digital
loop to operate at an integer-plus-fractional multiple. The
output of the DPLL is
FRAC1 
f OUT _ DPLL  f TDC  (N1  1) 
MOD1 

where N1 is the 17-bit value stored in the appropriate profile
registers (Register 0x0715 to Register 0x0717 for REFA). FRAC1
and MOD1 are the 24-bit numerators and denominators of the
fractional feedback divider block. The fractional portion of the
feedback divider can be bypassed by setting FRAC1 to 0, but
MOD1 should never be 0.
DPLL Overview
SYSTEM
CLOCK
+
TUNING
WORD
CLAMP
AND
HISTORY
17-BIT
24-BIT/24-BIT
INTEGER RESOLUTION
The DPLL output frequency is usually 175 MHz to 200 MHz for
optimal performance.
TO APLL
FROM APLL
09758-136
DIGITAL
LOOP
FILTER
×2
30-BIT NCO
FRAC1/
MOD1
DPFD
÷N1
FREE RUN
TW
R DIVIDER
(20-BIT)
Determination of the filter response by numeric
coefficients rather than by discrete component values
The absence of analog components (R/L/C), which
eliminates tolerance variations due to aging
The absence of thermal noise associated with analog
components
The absence of control node leakage current associated
with analog components (a source of reference feedthrough spurs in the output spectrum of a traditional
analog PLL)
The digital loop filter produces a time series of digital words at
its output and delivers them to the frequency tuning input of a
sigma-delta (Σ-Δ) modulator (SDM). The digital words from
the loop filter steer the DCO frequency toward frequency and
phase lock with the input signal (fTDC).
DIGITAL PLL (DPLL) CORE
FROM
REF
INPUT
MUX
fR
R 1
Figure 36. Digital PLL Core
Rev. B | Page 31 of 104
AD9558
Data Sheet
TDC/PFD
DPLL Digitally Controlled Oscillator Free Run Frequency
The phase-frequency detector (PFD) is an all-digital block. It
compares the digital output from the TDC (which relates to the
active reference edge) with the digital word from the feedback
block. It uses a digital code pump and digital integrator (rather
than a conventional charge pump and capacitor) to generate the
error signal that steers the DCO frequency toward phase lock.
The AD9558 uses a Σ-Δ modulator (SDM) as a digitally controlled
oscillator (DCO). The DCO free run frequency can be calculated
from the following equation:
Programmable Digital Loop Filter
where FTW0 is the value in Register 0x0300 to Register 0x0303,
and fSYS is the system clock frequency. See the System Clock
section for information on calculating the system clock frequency.
The AD9558 loop filter is a third order digital IIR filter that is
analogous to the third order analog loop shown in Figure 37.
C2
C3
09758-015
R2
2
FTW0
8  30
2
Adaptive Clocking
R3
C1
f dco _ freerun  f SYS 
Figure 37. Third Order Analog Loop Filter
The AD9558 loop filter block features a simplified architecture
in which the user enters the desired loop characteristics directly
into the profile registers. This architecture makes the calculation
of individual coefficients unnecessary in most cases, while still
offering complete flexibility.
The AD9558 has two preset digital loop filters: high (88.5°) phase
margin and normal (70°) phase margin. The loop filter coefficients
are stored in Register 0x0317 to Register 0x0322 for high phase
margin and Register 0x0323 to Register 0x032E for normal phase
margin. The high phase margin loop filter is intended for
applications in which the closed-loop transfer function must
not have greater than 0.1 dB of peaking.
Bit 0 of the following registers selects which filter is used for
each profile: Register 0x070E for Profile A, Register 0x074E for
Profile B, Register 0x078E for Profile C, and Register 0x07CE
for Profile D.
The loop bandwidth for each profile is set in the following
registers: Register 0x070F to Register 0x0711 for Profile A,
Register 0x074F to Register 0x0751 for Profile B, Register 0x078F
to Register 0x0791 for Profile C, and Register 0x07CF to
Register 0x07D1 for Profile D.
The two preset conditions should cover all of the intended
applications for the AD9558. For special cases where these
conditions must be modified, the tools for calculating these
coefficients are available by contacting Analog Devices directly.
The AD9558 can support adaptive clocking applications such as
asynchronous mapping and demapping. In these applications,
the output frequency can be dynamically adjusted by up to
±100 ppm from the nominal output frequency without manually
breaking the DPLL loop and reprogramming the part. This
function is supported for REFA only, not REFB.
The following registers are used in this function:



Register 0x0717 (DPLL N1 divider)
Register 0x0718 to Register 0x071A (DPLL FRAC1 divider)
Register 0x071B to Register 0x071D (DPLL MOD1
divider)
Writing to these registers requires an I/O update by writing
0x01 to Register 0x0005 before the new values take effect.
To make small adjustments to the output frequency, the user
can vary the FRAC1 and issue an I/O update. The advantage to
using only FRAC1 to adjust the output frequency is that the
DPLL does not briefly enter holdover. Therefore, the FRAC1 bit
can be updated as fast as the phase detector frequency of
the DPLL.
Writing to the N1 and MOD1 dividers allows for larger changes
to the output frequency. When the AD9558 detects that the N1
or MOD1 values have changed, it automatically enters and exits
holdover for a brief instant without any disturbance in the
output frequency. This limits how quickly the output frequency
can be adapted.
It is important to realize that the amount of frequency adjustment
is limited to ±100 ppm before the output PLL (APLL) needs a
recalibration. Variations that are larger than ±100 ppm are
possible, but the ability of the AD9558 to maintain lock over
temperature extremes may be compromised.
It is also important to remember that the rate of change in
output frequency depends on the DPLL loop bandwidth.
Rev. B | Page 32 of 104
Data Sheet
AD9558
DPLL Phase Lock Detector
The DPLL contains an all-digital phase lock detector. The user
controls the threshold sensitivity and hysteresis of the phase
detector via the profile registers.
The phase lock detector behaves in a manner analogous to
water in a tub (see Figure 38). The total capacity of the tub is
4096 units with −2048 denoting empty, 0 denoting the 50%
point, and +2048 denoting full. The tub also has a safeguard to
prevent overflow. Furthermore, the tub has a low water mark at
−1024 and a high water mark at +1024. To change the water
level, the user adds water with a fill bucket or removes water
with a drain bucket. The user specifies the size of the fill and
drain buckets via the 8-bit fill rate and drain rate values in the
profile registers.
PREVIOUS
STATE
LOCKED
UNLOCKED
2048
DRAIN
RATE
UNLOCK LEVEL
–1024
09758-017
0
FILL
RATE
–2048
Figure 38. Lock Detector Diagram
The water level in the tub is what the lock detector uses to
determine the lock and unlock conditions. When the water level
is below the low water mark (−1024), the detector indicates an
unlock condition. Conversely, when the water level is above the
high water mark (+1024), the detector indicates a lock condition.
When the water level is between the marks, the detector simply
holds its last condition. This concept appears graphically in
Figure 38, with an overlay of an example of the instantaneous
water level (vertical) vs. time (horizontal) and the resulting
lock/unlock states.
During any given PFD cycle, the detector either adds water with
the fill bucket or removes water with the drain bucket (one or
the other but not both). The decision of whether to add or
remove water depends on the threshold level specified by the
user. The phase lock threshold value is a 16-bit number stored
in the profile registers and is expressed in picoseconds (ps).
Thus, the phase lock threshold extends from 0 ns to ±65.535 ns
and represents the magnitude of the phase error at the output of
the PFD.
The phase lock detector compares each phase error sample at
the output of the PFD to the programmed phase threshold value.
If the absolute value of the phase error sample is less than or
equal to the programmed phase threshold value, the detector
control logic dumps one fill bucket into the tub. Otherwise,
Note that when the AD9558 enters the free run or holdover
mode, the DPLL phase lock detector indicates an unlocked
state. However, when the AD9558 performs a reference switch,
the lock detector state prior to the switch is preserved during
the transition period.
DPLL Frequency Lock Detector
LOCK LEVEL
1024
it removes one drain bucket from the tub. Note that it is not the
polarity of the phase error sample, but its magnitude relative to
the phase threshold value, that determines whether to fill or drain.
If more filling is taking place than draining, the water level in
the tub eventually rises above the high water mark (+1024), which
causes the phase lock detector to indicate lock. If more draining
is taking place than filling, then the water level in the tub eventually
falls below the low water mark (−1024), which causes the phase
lock detector to indicate unlock. The ability to specify the threshold
level, fill rate, and drain rate enables the user to tailor the operation
of the phase lock detector to the statistics of the timing jitter
associated with the input reference signal.
The operation of the frequency lock detector is identical to that
of the phase lock detector. The only difference is that the fill or
drain decision is based on the period deviation between the
reference and feedback signals of the DPLL instead of the phase
error at the output of the PFD.
The frequency lock detector uses a 24-bit frequency threshold
register specified in units of picoseconds (ps). Thus, the frequency
threshold value extends from 0 μs to ±16.777215 μs. It represents
the magnitude of the difference in period between the reference
and feedback signals at the input to the DPLL. For example, if
the reference signal is 1.25 MHz and the feedback signal is
1.38 MHz, the period difference is approximately 75.36 ns
(|1/1,250,000 − 1/1,380,000| ≈ 75.36 ns).
Frequency Clamp
The AD9558 DPLL features a digital tuning word clamp that
ensures that the DPLL output frequency stays within a defined
range. This feature is very useful to eliminate undesirable
behavior in cases where the reference input clocks may be
unpredictable. The tuning word clamp is also useful to
guarantee that the APLL never loses lock by ensuring that the
APLL VCO frequency stays within its tuning range.
Frequency Tuning Word History
The AD9558 has the ability to track the history of the tuning
word samples generated by the DPLL digital loop filter output.
It does so by periodically computing the average tuning word
value over a user-specified interval. This average tuning word is
used during holdover mode to maintain the average frequency
when no input references are present.
Rev. B | Page 33 of 104
AD9558
Data Sheet
LOOP CONTROL STATE MACHINE
Switchover
Switchover occurs when the loop controller switches directly
from one input reference to another. The AD9558 handles a
reference switchover by briefly entering holdover mode, loading
the new DPLL parameters, and then immediately recovering.
During the switchover event, however, the AD9558 preserves
the status of the lock detectors to avoid phantom unlock
indications.
Holdover
The holdover state of the DPLL is typically used when none of
the input references are present, although the user can also
manually engage holdover mode. In holdover mode, the output
frequency remains constant. The accuracy of the AD9558 in
holdover mode is dependent on the device programming and
availability of tuning word history.
Recovery from Holdover
When in holdover mode and a valid reference becomes
available, the device exits holdover operation. The loop state
machine restores the DPLL to closed-loop operation, locks to
the selected reference, and sequences the recovery of all the
loop parameters based on the profile settings for the active
reference.
Note that, if the user holdover bit is set, the device does not
automatically exit holdover when a valid reference is available.
However, automatic recovery can occur after clearing the user
holdover bit (Bit 6 in Register 0x0A01).
Rev. B | Page 34 of 104
Data Sheet
AD9558
SYSTEM CLOCK (SYSCLK)
SYSTEM CLOCK INPUTS
Functional Description
The SYSCLK circuit provides a low jitter, stable, high frequency
clock for use by the rest of the chip. The XOA and XOB pins
connect to the internal SYSCLK multiplier. The SYSCLK multiplier
can synthesize the system clock by connecting a crystal resonator
across the XOA and XOB input pins or by connecting a low
frequency clock source. The optimal signal for the system clock
input is either a crystal in the 50 MHz range or an ac-coupled
square wave with a 1 V p-p amplitude.
System Clock Period
For the AD9558 to accurately measure the frequency of incoming
reference signals, the user must enter the system clock period
into the nominal system clock period registers (Register 0x0103
to Register 0x0105). The SYSCLK period is entered in units of
nanoseconds (ns).
System Clock Details
There are two internal paths for the SYSCLK input signal: low
frequency non-xtal (LF) and crystal resonator (XTAL).
Using a TCXO for the system clock is a common use for the LF
path. Applications requiring DPLL loop bandwidths of less
than 50 Hz or high stability in holdover require a TCXO. As an
alternative to the 49.152 MHz crystal for these applications, the
AD9558 reference design uses a 19.2 MHz TCXO, which offers
excellent holdover stability and a good combination of low jitter
and low spurious content.
The 1.8 V differential receiver connected to the XOA and XOB
pins is self-biased to a dc level of ~1 V, and ac coupling is strongly
recommended. When a 3.3 V CMOS oscillator is in use, it is
important to use a voltage divider to reduce the input high
voltage to a maximum of 1.8 V. See Figure 34 for details on
connecting a 3.3 V CMOS TCXO to the system clock input.
The non-xtal input path permits the user to provide an LVPECL,
LVDS, 1.8 V CMOS, or sinusoidal low frequency clock for
multiplication by the integrated SYSCLK PLL. The LF path handles
input frequencies from 3.5 MHz up to 100 MHz. However,
when using a sinusoidal input signal, it is best to use a frequency
that is in excess of 20 MHz. Otherwise, the resulting low slew
rate can lead to substandard noise performance. Note that the
non-xtal path includes an optional 2× frequency multiplier to
double the rate at the input to the SYSCLK PLL and potentially
reduce the PLL in-band noise. However, to avoid exceeding the
maximum PFD rate of 150 MHz, the 2× frequency multiplier is
only for input frequencies that are below 75 MHz.
The XTAL path enables the connection of a crystal resonator
(typically 10 MHz to 50 MHz) across the XOA and XOB input
pins. An internal amplifier provides the negative resistance
required to induce oscillation. The internal amplifier expects an
AT cut, fundamental mode crystal with a maximum motional
resistance of 100 Ω. The following crystals, listed in alphabetical
order, may meet these criteria. Analog Devices, Inc., does not
guarantee their operation with the AD9558 nor does Analog
Devices endorse one crystal supplier over another. The AD9558
reference design uses a 49.152 MHz crystal, which is high
performance, low spurious content, and readily available.







AVX/Kyocera CX3225SB
ECS ECX-32
Epson/Toyocom TSX-3225
Fox FX3225BS
NDK NX3225SA
Siward SX-3225
Suntsu SCM10B48-49.152 MHz
SYSTEM CLOCK MULTIPLIER
The SYSCLK PLL multiplier is an integer-N design with an
integrated VCO. It provides a means to convert a low frequency
clock input to the desired system clock frequency, fSYS (750 MHz
to 805 MHz). The SYSCLK PLL multiplier accepts input signals
of between 3.5 MHz and 600 MHz, but frequencies that are in
excess of 150 MHz require the system clock P-divider to ensure
compliance with the maximum PFD rate (150 MHz). The PLL
contains a feedback divider (N) that is programmable for divide
values between 4 and 255.
f SYS  f OSC 
sysclk _ Ndiv
sysclk _ Pdiv
where:
fOSC is the frequency at the XOA and XOB pins.
sysclk_Ndiv is the value stored in Register 0x0100.
sysclk_Pdiv is the system clock P divider that is determined by
the setting of Register 0x0101[2:1].
If the system clock doubler is used, the value of sysclk_Ndiv
should be half of its original value.
The system clock multiplier features a simple lock detector that
compares the time difference between the reference and feedback
edges. The most common cause of the SYSCLK multiplier not
locking is a non-50% duty cycle at the SYSCLK input while the
system clock doubler is enabled.
The non-xtal path also includes an input divider (M) that is
programmable for divide-by-1, -2, -4, or -8. The purpose of the
divider is to limit the frequency at the input to the PLL to less
than 150 MHz (the maximum PFD rate).
Rev. B | Page 35 of 104
AD9558
Data Sheet
System Clock Stability Timer
Because the reference monitors depend on the system clock
being at a known frequency, it is important that the system
clock be stable before activating the monitors. At initial powerup, the system clock status is not known, and, therefore, it is
reported as being unstable. After the part has been programmed,
the system clock PLL (if enabled) eventually locks. When a
stable operating condition is detected, a timer is run for the
duration that is stored in the system clock stability period
registers. If at any time during this waiting period, the condition
is violated, the timer is reset and halted until a stable condition
is reestablished. After the specified period elapses, the AD9558
reports the system clock as stable.
Rev. B | Page 36 of 104
Data Sheet
AD9558
OUTPUT PLL (APLL)
A diagram of the output PLL (APLL) is shown in Figure 39.
Calibration of the APLL must be performed at startup and
when the nominal input frequency to the APLL changes by
more than ±100 ppm, although the APLL maintains lock
over voltage and temperature extremes without recalibration.
Calibration centers the dc operating voltage at the input to the
APLL VCO.
INTEGER DIVIDER
÷N2
OUTPUT PLL DIVIDER (APLL)
FROM DPLL
PFD
CP
LF
TO CLOCK
DISTRIBUTION
VCO2
3.35GHz TO 4.05GHz
LF CAP
09758-138
LF_VCO2
Figure 39. Output PLL Block Diagram
The APLL provides the frequency upconversion from the DPLL
output to the 3.35 GHz to 4.05 GHz range, while also providing
noise filtering on the DPLL output. The APLL reference input is
the output of the DPLL. The feedback divider is an integer divider.
The loop filter is partially integrated with the one external 6.8 nF
capacitor. The nominal loop bandwidth for this PLL is 250 kHz,
with 68 degrees of phase margin.
The frequency wizard that is included in the evaluation software
configures the APLL, and the user should not need to make
changes to the APLL settings. However, there may be special cases
where the user may wish to adjust the APLL loop bandwidth to
meet a specific phase noise requirement. The easiest way to change
the APLL loop BW is to adjust the APLL charge pump current
in Register 0x0400. There is sufficient stability (68⁰ of phase
margin) in the APLL default settings to permit a broad range of
adjustment without causing the APLL to be unstable. The user
should contact Analog Devices directly if more detail is needed.
APLL calibration at startup can be accomplished during initial
register loading by following the instructions in the Device
Register Programming When Using a Register Setup File
section of this datasheet.
To recalibrate the APLL VCO after the chip has been running,
the user should first input the new settings (if any). Ensure that
the system clock is still locked and stable, and that the DPLL is
in free run mode with the free run tuning word set to the same
output frequency that is used when the DPLL is locked.
Use the following steps to calibrate the APLL VCO:
1.
2.
3.
4.
5.
6.
Ensure that the system clock is locked and stable.
Ensure that the DPLL is in user free run mode
(Register 0x0A01[5] = 1b), and the free run tuning word is set.
Write Register 0x0405 = 0x20.
Write Register 0x0005 = 0x01.
Write Register 0x0405 = 0x21.
Write Register 0x0005 = 0x01.
Monitor the APLL status using Bit 2 in Register 0x0D01.
Rev. B | Page 37 of 104
AD9558
Data Sheet
CLOCK DISTRIBUTION
÷M0
RF
DIVIDER 1
÷3 TO ÷11
MAX
1.25GHz
FRAME SYNC
MODE ONLY
÷M1
OUT01
OUT01
OUT11
OUT11
OUT21
FROM APLL
(3.35GHz TO 4.05GHz)
OUT21
RF
DIVIDER 2
÷3 TO ÷11
MAX
1.25GHz
÷M2
OUT31
OUT31
OUT41
OUT41
SYNC SIGNAL TO
M0 TO M3 DIVIDERS
CHANNEL
SYNC
BLOCK
÷M3
FRAME SYNC
FRAME SYNC ENGAGED SIGNAL
Fsync_ALIGN_METHOD
SELECTED INPUT FRAME PULSE
÷M3b
FRAME
SYNC
BLOCK
×2
OUT51
OUT51
FRAME
SYNC
MONITOR
OUT0, OUT1, OUT2, OUT3, OUT4: 360kHz TO 1.25GHz; OUT5: 352Hz TO 1.25GHz.
09758-139
CHIP RESET
SYNC/SOFT_SYNC
Figure 40. Clock Distribution Block Diagram
CLOCK DIVIDERS
The channel divider blocks, M0, M1, M2, M3, and M3b, are
10-bit integer dividers with a divide range of 1 to 1023. The
channel divider block contains duty cycle correction that
guarantees 50% duty cycle for both even and odd divide ratios.
All CMOS drivers feature a CMOS drive strength that allows
the user to choose between a strong, high performance CMOS
driver, or a lower power setting with less EMI and crosstalk.
The best setting is application dependent.
The output drivers can be individually powered down.
For applications where LVPECL levels are required, the user
should choose the HSTL mode, and ac-couple the output signal.
See the Input/Output Termination Recommendations section
for recommended termination schemes.
OUTPUT ENABLE
CLOCK DISTRIBUTION SYNCHRONIZATION
Each of the output channels offers independent control of enable/
disable functionality via the distribution enable register. The
distribution outputs use synchronization logic to control
enable/disable activity to avoid the production of runt pulses
and ensure that outputs with the same divide ratios become
active/inactive in unison.
Divider Synchronization
OUTPUT POWER-DOWN
OUTPUT MODE
The user has independent control of the operating mode of each of
the four output channels via the output clock distribution registers
(Address 0x0500 to Address 0x0515). The operating mode
control includes
•
•
•
•
•
Logic family and pin functionality
Output drive strength
Output polarity
Divide ratio
Phase of each output channel
Channel 0 and Channel 3 provide 3.3 V CMOS, in addition
to 1.8 V CMOS modes. Channel 1 and Channel 2 have 1.8 V
CMOS, LVDS, and HSTL modes.
The dividers in the clock distribution channels can be
synchronized with each other.
At power-up, the clock dividers are held static until a sync signal
is initiated by the channel SYNC block. The following are
possible sources of a SYNC signal, and these settings are found
in Register 0x0500:
•
•
•
•
•
•
Direct sync via Bit 2 of Register 0x0500
Direct sync via a sync op code (0xA1) in the EEPROM
storage sequence during EEPROM loading
DPLL phase or frequency lock
A rising edge of the selected reference input
The SYNC pin
A multifunction pin configured for the SYNC signal
E
A
A
The APLL lock detect signal gates the SYNC signal from the
channel sync block shown in Figure 40. The channel dividers
receive a SYNC signal from the channel SYNC block only if the
APLL is calibrated and locked, unless the APLL locked
controlled sync bit (Register 0x0405[3]) is set.
Rev. B | Page 38 of 104
Data Sheet
AD9558
A channel can be programmed to ignore the sync function by
setting the mask Channel x sync bits in Register 0x0500[7:4].
When programmed to ignore the sync, the channel ignores
both the user initiated sync signal and the zero delay initiated
sync signals, and the channel divider starts toggling, provided
that the APLL is calibrated and locked, or if the APLL locked
controlled sync bit (Register 0x0405[3]) is set.
If the output SYNC function is to be controlled using an M pin,
use the following steps:
1.
2.
3.
First, enable the M pins by writing Register 0x0200 = 0x01.
Issue an I/O update (Register 0x0005 = 0x01).
Set the appropriate M pin function.
If this process is not followed, a SYNC pulse is issued
automatically.
Rev. B | Page 39 of 104
AD9558
Data Sheet
FRAME SYNCHRONIZATION
The AD9558 provides frame synchronization function/mode.
With this function, the AD9558 can take a pair consisting of a
reference clock and a 2 kHz or 8 kHz frame pulse as input signals
and generate a pair consisting of a synchronized output clock
and an output frame pulse while the output frame pulse is
synchronized with the input frame pulse also. The reference
clock is used to synthesize the output clock and output frame
pulse through DPLL/output PLL/distribution and the input frame
pulse is used to control the phase of the output frame pulse.
CLOCK OUTPUTS IN FRAME SYNCHRONIZATION
MODE
Frame synchronization is not supported in the soft or hard pin
control mode.
The AD9558 has six outputs (OUT0 to OUT5). In frame sync
mode, the OUT0 and OUT5 form the pair of output clock
(OUT0) and output frame pulse (OUT5). The frequency of
OUT0 is required to have the integer relation with the frequency of
the OUT5 (fOUT0 = M × fOUT5). The rest of the outputs (OUT1 to
OUT4) do not participate in frame synchronization mode and
are programmed and synchronized with each other the same as
in normal mode (except for SYNC function). However, OUT1
to OUT4 should not be synchronized with the OUT0/OUT5.
REFERENCE CONFIGURATION IN FRAME
SYNCHRONIZATION MODE
CONTROL REGISTERS FOR FRAME
SYNCHRONIZATION MODE
In frame synchronization mode, four AD9558 reference inputs
(REFA/REFB/REFC/REFD) are arranged into two pairs of signals:
REFA and REFC form a pair of input clock/input frame pulse
with REFA as the input clock, and REFC as the frame pulse.
REFB and REFD form the second pair of input clock/ input
frame pulse with REFB as the input clock and REFD is the frame
pulse. During reference switchover, only two input clocks, REFA
and REFB, are assigned with a priority index. The two frame
pulses, REFC and REFD, are not assigned with the priority index
(the priority register bits in the profiles associated with input
frame pulse are ignored). Each pair of input clock/ frame pulse
participates in the reference selection as a group, and the valid
state and priority of the pair are used in determining the reference
selection. The priority of the pair is indicated by the priority
index of the input clock in the pair.
The frame synchronization function is enabled by setting
Register 0x0640[0] to 1b. When Register 0x0640[0] = 1b, the
following occurs:
Users have the option to either include or exclude the valid state
of input frame pulse in the reference selection by programming
the validate Fsync reference bit (Register 0x0641[3]). When
Register 0x0641[3] is programmed to 1b, the valid state of the
input frame pulse and the valid state of the paired input clock
are logically ANDed and the result is used to indicate the valid
state of the pair. When validate Fsync ref is programmed to 0b,
the valid state of the input frame pulse is excluded in reference
selection and only the valid state of the input clock in the pair is
used to indicate the valid state of the pair. The valid pair with the
higher priority index is selected as the DPLL reference and input
frame pulse to control the phase of the output frame pulse. If no
pair is valid, the selection does not change and DPLL is switched
to either holdover or free run mode, and the phase of the output
frame pulse is not controlled by any of the input frame pulses.
The five reference switchover modes for frame synchronization
mode is the same as for normal mode.
E
A
A
•
The frame synchronization control bits (Register
0x0641[3:0]) are enabled.
The SYNC pin switches from SYNC function to frame
SYNC function. In frame synchronization mode, SYNC
cannot be used as clock distribution synchronization
function as it is in normal mode. Instead it is used as the
frame synchronization arm function.
•
E
E
A
A
A
A
E
E
A
A
A
A
When the AD9558 is in frame synchronization mode, the frame
synchronization function can be armed by either the SYNC pin
or the arm soft Fsync bit (Register 0x0641[0]), which is selected
by the Fsync arm method bit (Register 0x0641[1]. A value of 0b
(which is the default) selects the SYNC pin as the arm method.
If SYNC is selected as arm method, SYNC = low means armed;
if the register is selected as arm method, Register 0x0641[0] = 1b
means armed. ARM means that once armed, the output frame
pulse is edge aligned with the paired output clock edge after the
rising edge of the input frame pulse.
E
A
A
E
A
A
E
A
E
A
A
A
LEVEL SENSITIVE MODE AND ONE-SHOT MODE
The frame synchronization can operate in level sensitive or oneshot mode as determined by the Fsync one shot bit (Register
0x0641[2]). When in level sensitive mode (Register 0x0641[2] =
0b) and the frame sync ARM signal is high, each rising edge of
the selected input frame pulse signal is used to control the
phase of the output frame pulse. When in one-shot mode, after
the frame sync ARM signal is high, only the immediate next
rising edge of the selected input frame pulse signal is used to
control the phase of the output frame pulse (one time phase
alignment). After that, the phase of the output frame pulse is
not controlled by the selected input frame pulse. Instead, it
follows the phase of the input clock of M3 divider. In either
alignment control mode, the resolution of the phase realignment
between the input frame pulse and the output frame pulse is
one clock cycle of the paired clock output.
Rev. B | Page 40 of 104
Data Sheet
AD9558
M3b DIVIDER/OUT5 PROGRAMMING IN FRAME
SYNCHRONIZATION MODE
This means that in frame synchronization mode, the total divide
ratio between the RF divider and OUT5 is M0 × M3 × M3b.
In frame synchronization mode, the clock distribution signal path
for OUT5 is changed, as follows: the OUT5 signal goes from the RF
divider to the M0 divider, and then to the M3 and M3b dividers.
The other important change is that the sync signal for the M3b
divider is no longer the standard clock distribution sync. It is
controlled by a signal derived from the input frame pulse.
ACTIVE SYNC
ACTIVE SYNC
OUPUT CLOCK
(OUT0)
FRAME CLOCK INPUT
(REFC/REFD)
ENABLE Fsync
(REGISTER 0x0640[0] = 1b)
FRAME SYNC ARM
(REGISTER OR SYNC PIN)
NOTES
1. AFTER THE FRAME SYNC IS ARMED, THE FRAME CLOCK OUTPUT IS SYNCHRONIZED TO THE FRAME CLOCK INPUT.
THE SKEW BETWEEN THE FRAME CLOCK INPUT AND FRAME CLOCK OUTPUT IS 15ns (NOMINAL) PLUS A DELAY OF 0.5 TO 1.5 OUTPUT CLOCK CYCLES.
09758-141
FRAME CLOCK
(OUT5)
Figure 41. Frame Synchronization in Level Sensitive Mode
ACTIVE SYNC
OUPUT CLOCK
(OUT0)
FRAME CLOCK INPUT
(REFC/REFD)
ENABLE Fsync
(REGISTER 0x0640[0] = 1b)
FRAME SYNC ARM
(REGISTER OR SYNC PIN)
NOTES
1. AFTER THE FRAME SYNC IS ARMED, THE FRAME CLOCK OUTPUT IS SYNCHRONIZED TO THE FRAME CLOCK INPUT.
THE SKEW BETWEEN THE FRAME CLOCK INPUT AND FRAME CLOCK OUTPUT IS 15ns (NOMINAL) PLUS 0.5 TO 1.5 OUTPUT CLOCK CYCLES.
Figure 42. Frame Synchronization in One Shot Mode
ENABLE Fsync
M3 DIVIDER
FROM
APLL
M3b DIVIDER
OUT5
ENABLE
Fsync
RF
DIVIDER 1
LOGIC 0
OUT0
M0 DIVIDER
ENABLE Fsync
CLOCK
DISTRIBUTION
SYNC LOGIC
DELAY
OUTPUT
SYNC
LEVEL
(RETIMED)
ENABLE Fsync
OUTPUT
SYNC
PULSE
FRAME SYNC
PULSE GENERATOR
FRAME SYNC PULSE INPUT (ON REFC OR REFD)
NOTES
1. THE ENABLE Fsync FUNCTION IN THE DIAGRAM IS CONTROLLED BY REGISTER 0x0640[0].
Figure 43. Frame Synchronization Implementation
Rev. B | Page 41 of 104
09758-143
Fsync_arm (FROM EITHER THE SYNC PIN OR THE Arm Soft Fsync BIT IN REGISTER 0x0641[0])
09758-142
FRAME CLOCK
(OUT5)
AD9558
Data Sheet
STATUS AND CONTROL
MULTIFUNCTION PINS (M7 TO M0)
The AD9558 has eight digital CMOS I/O pins (M7 to M0)
that are configurable for a variety of uses. To use these
functions, the user must enable them by writing a 0x01 to
Register 0x0200. The function of these pins is programmable via
the register map. Each pin can control or monitor an assortment
of internal functions based on the contents of Register 0x0201 to
Register 0x0208.
To monitor an internal function with a multifunction pin, write
a Logic 1 to the most significant bit of the register associated
with the desired multifunction pin. The value of the seven least
significant bits of the register defines the control function, as
shown in Table 129.
To control an internal function with a multifunction pin, write a
Logic 0 to the most significant bit of the register associated with
the desired multifunction pin. The monitored function depends
on the value of the seven least significant bits of the register, as
shown in Table 130.
If more than one multifunction pin operates on the same
control signal, then internal priority logic ensures that only one
multifunction pin serves as the signal source. The selected pin is
the one with the lowest numeric suffix. For example, if both M0
and M3 operate on the same control signal, then M0 is used as
the signal source and the redundant pins are ignored.
At power-up, the multifunction pins can be used to force the
device into certain configurations as defined in the initial pin
programming section. This functionality, however, is valid only
during power-up or following a reset, after which the pins can
be reconfigured via the serial programming port or via the
EEPROM.
If the output SYNC function is to be controlled using an M pin,
1.
2.
3.
First, enable the M pins by writing Register 0x0200 = 0x01.
Issue an I/O update (Register 0x0005 = 0x01).
Set the appropriate M pin function.
If this process is not followed, a SYNC pulse is issued automatically.
IRQ Pin
The AD9558 has a dedicated interrupt request (IRQ) pin.
Bits[1:0] of the IRQ pin output mode register (Register 0x0209)
control how the IRQ pin asserts an interrupt based on the value
of the two bits, as follows:
00—The IRQ pin is high impedance when deasserted and active
low when asserted and requires an external pull-up resistor.
01—The IRQ pin is high impedance when deasserted and active
high when asserted and requires an external pull-down
resistor.
10—The IRQ pin is Logic 0 when deasserted and Logic 1 when
asserted.
11—The IRQ pin is Logic 1 when deasserted and Logic 0 when
asserted. (This is the default operating mode.)
The AD9558 asserts the IRQ pin when any bit in the IRQ monitor
register (Address 0x0D02 to Address 0x0D07) is a Logic 1. Each
bit in this register is associated with an internal function that is
capable of producing an interrupt. Furthermore, each bit of the
IRQ monitor register is the result of a logical AND of the associated
internal interrupt signal and the corresponding bit in the IRQ
mask register (Address 0x020A to Address 0x020E). That is, the
bits in the IRQ mask register have a one-to-one correspondence
with the bits in the IRQ monitor register. When an internal
function produces an interrupt signal and the associated IRQ
mask bit is set, the corresponding bit in the IRQ monitor register
is set. The user should be aware that clearing a bit in the IRQ
mask register removes only the mask associated with the
internal interrupt signal. It does not clear the corresponding bit
in the IRQ monitor register.
Note that the IRQ function detects a state change in the function
that is being monitored. However, if IRQs are cleared (or if they
are enabled for the first time), they do not generate for a preexisting condition. The state must change after the IRQs are
enabled. For example, if REFA is already invalid before the REFA
invalid IRQ is enabled, the IRQ does not generate.
The IRQ pin is the result of a logical OR of all the IRQ monitor
register bits. Thus, the AD9558 asserts the IRQ pin as long as
any IRQ monitor register bit is a Logic 1. Note that it is possible
to have multiple bits set in the IRQ monitor register. Therefore,
when the AD9558 asserts the IRQ pin, it may indicate an interrupt
from several different internal functions. The IRQ monitor
register provides the user with a means to interrogate the
AD9558 to determine which internal function produced the
interrupt.
Typically, when the IRQ pin is asserted, the user interrogates
the IRQ monitor register to identify the source of the interrupt
request. After servicing an indicated interrupt, the user should
clear the associated IRQ monitor register bit via the IRQ
clearing register (Address 0x0A04 to Address 0x0A09). The bits
in the IRQ clearing register have a one-to-one correspondence with
the bits in the IRQ monitor register. Note that the IRQ clearing
register is autoclearing. The IRQ pin remains asserted until the
user clears all of the bits in the IRQ monitor register that
indicate an interrupt.
It is also possible to collectively clear all of the IRQ monitor
register bits by setting the clear all IRQs bit in the reset function
register (Register 0x0A03, Bit 1). Note that this is an autoclearing
bit. Setting this bit results in deassertion of the IRQ pin.
Alternatively, the user can program any of the multifunction
pins to clear all IRQs. This allows the user to clear all IRQs by
means of a hardware pin rather than by using a serial I/O port
operation.
Rev. B | Page 42 of 104
Data Sheet
AD9558
WATCHDOG TIMER
EEPROM
The watchdog timer is a general purpose programmable timer.
To set the timeout period, the user writes to the 16-bit watchdog
timer register (Address 0x0210 to Address 0x0211). A value of
0b in this register disables the timer. A nonzero value sets the
timeout period in milliseconds (ms), giving the watchdog timer
a range of 1 ms to 65.535 sec. The relative accuracy of the timer
is approximately 0.1% with an uncertainty of 0.5 ms.
EEPROM Overview
If enabled, the timer runs continuously and generates a timeout
event when the timeout period expires. The user has access to
the watchdog timer status via the IRQ mechanism and the
multifunction pins (M7 to M0). In the case of the multifunction
pins, the timeout event of the watchdog timer is a pulse that lasts
32 system clock periods.
There are two ways to reset the watchdog timer (thereby
preventing it from causing a timeout event). The first is by
writing a Logic 1 to the autoclearing clear watchdog bit in the
reset functions register (Register 0x0A03, Bit 0). Alternatively,
the user can program any of the multifunction pins to reset the
watchdog timer. This allows the user to reset the timer by means
of a hardware pin rather than by using a serial I/O port operation.
The AD9558 contains an integrated 2048-byte, electrically
erasable, programmable read-only memory (EEPROM).
The AD9558 can be configured to perform a download at
power-up via the multifunction pins (M3 and M2), but uploads
and downloads can also be done on demand via the EEPROM
control registers (Address 0x0E00 to Address 0x0E03).
The EEPROM provides the ability to upload and download
configuration settings to and from the register map. Figure 44
shows a functional diagram of the EEPROM.
Register 0x0E10 to Register 0x0E3F represent a 53-byte EEPROM
storage sequence area (referred to as the “scratch pad” in this
section) that enables the user to store a sequence of instructions
for transferring data to the EEPROM from the device settings
portion of the register map. Note that the default values for
these registers provide a sample sequence for saving/retrieving
all of the AD9558 EEPROM-accessible registers. Figure 44 shows
the connectivity between the EEPROM and the controller that
manages data transfer between the EEPROM and the register map.
The controller oversees the process of transferring EEPROM data
to and from the register map. There are two modes of operation
handled by the controller: saving data to the EEPROM (upload
mode) or retrieving data from the EEPROM (download mode).
In either case, the controller relies on a specific instruction set.
DATA
DATA
EEPROM
(0x000 TO 0x1FF)
SCRATCH PAD
ADDRESS
POINTER
SCRATCH PAD
(EEPROM STORAGE SEQUENCE)
(0x0E10 TO 0x0E45)
REGISTER MAP
Figure 44. EEPROM Functional Diagram
Rev. B | Page 43 of 104
SERIAL
INPUT/OUTPUT
PORT
09758-024
DEVICE
SETTINGS
(0x0004 TO 0x0A0D)
CONDITION
(0E01 [3:0])
DATA
DEVICE
SETTINGS
ADDRESS
POINTER
EEPROM
ADDRESS
POINTER
EEPROM
CONTROLLER
M3
M2
AD9558
Data Sheet
Table 21. EEPROM Controller Instruction Set
Instruction
Value (Hex)
0x00 to 0x7F
Instruction Type
Data
Bytes
Required
3
0x80
I/O update
1
0xA0
Calibrate
1
0xA1
Distribution sync
1
0xB0 to 0xCF
Condition
1
0xFE
Pause
1
0xFF
End
1
Description
A data instruction tells the controller to transfer data to or from the device settings
part of the register map. A data instruction requires two additional bytes that
together, indicate a starting address in the register map. Encoded in the data
instruction is the number of bytes to transfer, which is one more than the
instruction value.
When the controller encounters this instruction while downloading from the
EEPROM, it issues a soft I/O update.
When the controller encounters this instruction while downloading from the
EEPROM, it initiates a system clock calibration sequence.
When the controller encounters this instruction while downloading from the
EEPROM, it issues a sync pulse to the output distribution synchronization.
B1 to CF are condition instructions and correspond to Condition 1 through
Condition 31, respectively. B0 is the null condition instruction. See the EEPROM
Conditional Processing section for details.
When the controller encounters this instruction in the EEPROM storage sequence
area while uploading to the EEPROM, it holds both the scratch pad address pointer
and the EEPROM address pointer at its last value. This allows storage of more than
one instruction sequence in the EEPROM. Note that the controller does not copy
this instruction to the EEPROM during upload.
When the controller encounters this instruction in the the EEPROM storage
sequence area while uploading to the EEPROM, it resets both the register area
address pointer and the EEPROM address pointer and then enters an idle state.
When the controller encounters this instruction while downloading from the
EEPROM, it resets the EEPROM address pointer and then enters an idle state.
EEPROM Instructions
Table 21 lists the EEPROM controller instruction set. The
controller recognizes all instruction types whether it is in
upload or download mode, except for the pause instruction,
which is recognized only in upload mode.
The I/O update, calibrate, distribution sync, and end instructions are mostly self-explanatory. The others, however, warrant
further detail, as described in the following paragraphs.
Data instructions are those that have a value from 0x000 to
0x7FF. A data instruction tells the controller to transfer data
between the EEPROM and the register map. The controller
requires the following two parameters to carry out the data
transfer:
•
•
The number of bytes to transfer
The register map target address
The controller decodes the number of bytes to transfer directly
from the data instruction itself by adding one to the value of the
instruction. For example, the 1A data instruction has a decimal
value of 26; therefore, the controller knows to transfer 27 bytes
(one more than the value of the instruction). When the
controller encounters a data instruction, it knows to read the
next two bytes in the scratch pad because these contain the
register map target address.
Note that, in the EEPROM scratch pad, the two registers that
comprise the address portion of a data instruction have the
MSB of the address in the D7 position of the lower register
address. The bit weight increases from left to right, from the
lower register address to the higher register address. Furthermore,
the starting address always indicates the lowest numbered
register map address in the range of bytes to transfer. That is,
the controller always starts at the register map target address
and counts upward regardless of whether the serial I/O port is
operating in I2C, SPI LSB-first, or SPI MSB-first mode.
As part of the data transfer process during an EEPROM upload,
the controller calculates a 1-byte checksum and stores it as the
final byte of the data transfer. As part of the data transfer process
during an EEPROM download, however, the controller again
calculates a 1-byte checksum value but compares the newly
calculated checksum with the one that was stored during the
upload process. If an upload/download checksum pair does not
match, the controller sets the EEPROM fault status bit. If the
upload/download checksums match for all data instructions
encountered during a download sequence, the controller sets
the EEPROM complete status bit.
Condition instructions are those that have a value from B0 to
CF. The B1 to CF condition instructions represent Condition 1
to Condition 31, respectively. The B0 condition instruction is
special because it represents the null condition (see the
EEPROM Conditional Processing section).
Rev. B | Page 44 of 104
Data Sheet
AD9558
A pause instruction, like an end instruction, is stored at the
end of a sequence of instructions in the scratch pad. When the
controller encounters a pause instruction during an upload
sequence, it keeps the EEPROM address pointer at its last value.
This way the user can store a new instruction sequence in the
scratch pad and upload the new sequence to the EEPROM. The
new sequence is stored in the EEPROM address locations
immediately following the previously saved sequence. This
process is repeatable until an upload sequence contains an end
instruction. The pause instruction is also useful when used in
conjunction with condition processing. It allows the EEPROM
to contain multiple occurrences of the same registers, with each
occurrence linked to a set of conditions (see the EEPROM
Conditional Processing section).
EEPROM Upload
To upload data to the EEPROM, the user must first ensure that
the write enable bit (Register 0x0E00, Bit 0) is set. Then, on
setting the autoclearing save to EEPROM bit (Register 0x0E02,
Bit 0), the controller initiates the EEPROM data storage process.
Uploading EEPROM data requires that the user first write an
instruction sequence into the scratch pad registers. During the
upload process, the controller reads the scratch pad data byteby-byte, starting at Register 0x0E10 and incrementing the scratch
pad address pointer, as it goes, until it reaches a pause or end
instruction.
As the controller reads the scratch pad data, it transfers the
data from the scratch pad to the EEPROM (byte-by-byte) and
increments the EEPROM address pointer accordingly, unless
it encounters a data instruction. A data instruction tells the
controller to transfer data from the device settings portion of
the register map to the EEPROM. The number of bytes to transfer
is encoded within the data instruction, and the starting address
for the transfer appears in the next two bytes in the scratch pad.
When the controller encounters a data instruction, it stores the
instruction in the EEPROM, increments the EEPROM address
pointer, decodes the number of bytes to be transferred, and
increments the scratch pad address pointer. Then it retrieves the
next two bytes from the scratch pad (the target address) and
increments the scratch pad address pointer by 2. Next, the
controller transfers the specified number of bytes from the
register map (beginning at the target address) to the EEPROM.
When it completes the data transfer, the controller stores an extra
byte in the EEPROM to serve as a checksum for the transferred
block of data. To account for the checksum byte, the controller
increments the EEPROM address pointer by one more than the
number of bytes transferred. Note that, when the controller
transfers data associated with an active register, it actually
transfers the buffered contents of the register (refer to the
Buffered/Active Registers section for details on the difference
between buffered and active registers). This allows for the transfer
of nonzero autoclearing register contents.
Note that conditional processing (see the EEPROM Conditional
Processing section) does not occur during an upload sequence.
EEPROM Download
An EEPROM download results in data transfer from the
EEPROM to the device register map. To download data,
the user sets the autoclearing load from the EEPROM bit
(Register 0x0E03, Bit 1). This commands the controller to
initiate the EEPROM download process. During download, the
controller reads the EEPROM data byte-by-byte, incrementing
the EEPROM address pointer as it goes, until it reaches an end
instruction. As the controller reads the EEPROM data, it
executes the stored instructions, which includes transferring
stored data to the device settings portion of the register map
when it encounters a data instruction.
Note that conditional processing (see the EEPROM Conditional
Processing section) is applicable only when downloading.
Automatic EEPROM Download
Following a power-up, an assertion of the RESET pin, or a soft
reset (Register 0x0000, Bit 5 = 1), if the PINCONTROL pin is
low, and M3 and M2 are either high or low (see Table 22), the
instruction sequence stored in the EEPROM executes automatically
with one of eight conditions. If M3 and M2 are left floating and
the PINCONTROL pin is low, the EEPROM is bypassed and
the factory defaults are used. In this way, a previously stored set
of register values downloads automatically on power-up or with
a hard or soft reset. See the EEPROM Conditional Processing
section for details regarding conditional processing and the way
it modifies the download process.
Table 22. EEPROM Setup
M3
Low
Low
Low
Open
Open
Open
High
High
High
Rev. B | Page 45 of 104
M2
Low
Open
High
Low
Open
High
Low
Open
High
ID
1
2
3
4
0
5
6
7
8
EEPROM Download?
Yes, EEPROM Condition 1
Yes, EEPROM Condition 2
Yes, EEPROM Condition 3
Yes, EEPROM Condition 4
No
Yes, EEPROM Condition 5
Yes, EEPROM Condition 6
Yes, EEPROM Condition 7
Yes, EEPROM Condition 8
AD9558
Data Sheet
EEPROM Conditional Processing
The condition tag board is a table maintained by the EEPROM
controller. When the controller encounters a condition instruction, it decodes the B1 through CF instructions as Condition = 1
through Condition = 8, respectively, and tags that particular
condition in the condition tag board. However, the B0 condition
instruction decodes as the null condition, for which the controller
clears the condition tag board; and subsequent download
instructions execute unconditionally (until the controller
encounters a new condition instruction).
The condition instructions allow conditional execution of
EEPROM instructions during a download sequence. During
an upload sequence, however, they are stored as is and have
no effect on the upload process.
Note that, during EEPROM downloads, the condition instructions
themselves and the end instruction always execute unconditionally.
Conditional processing makes use of two elements: the condition
(from Condition 1 to Condition 8) and the condition tag board.
The relationships among the condition, the condition tag board,
and the EEPROM controller appear schematically in Figure 45.
During download, the EEPROM controller executes or skips
instructions depending on the value of the condition and the
contents of the condition tag board. Note, however, that the
condition instructions and the end instruction always execute
unconditionally during download. If condition = 0, then all
instructions during download execute unconditionally. If
Condition ≠ 0 and there are any tagged conditions in the condition
tag board, then the controller executes instructions only if the
condition is tagged. If the condition is not tagged, then the
controller skips instructions until it encounters a condition
instruction that decodes as a tagged condition. Note that the
condition tag board allows for multiple conditions to be tagged
at any given moment. This conditional processing mechanism
enables the user to have one download instruction sequence
with many possible outcomes depending on the value of the
condition and the order in which the controller encounters
condition instructions.
The condition is a 4-bit value with 16 possibilities. Condition = 0
is the null condition. When the null condition is in effect, the
EEPROM controller executes all instructions unconditionally.
Condition 9 through Condition 15 are not accessible using the
M pins. The remaining eight possibilities (that is, Condition = 1
through Condition = 8) modify the EEPROM controller’s
handling of a download sequence. The condition originates
from one of two sources (see Figure 45), as follows:
•
FncInit, Bits[3:0], which is the state of the M2 and M3
multifunction pins at power-up (see Table 22)
Register 0x0E01, Bits[3:0]
If Register 0x0E01, Bits[3:0] ≠ 0, then the condition is the value
that is stored in Register 0x0E01, Bits[3:0]; otherwise, the condition
is FncInit, Bits[3:0]. Note that a nonzero condition that is present
in Register 0x0E01, Bits[3:0] takes precedence over FncInit,
Bits[3:0].
M3 AND M2 PINS
(3-LEVEL LOGIC)
CONDITION
TAG BOARD
EXAMPLE
CONDITION 3 AND
CONDITION 13
ARE TAGGED
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
29
30
31
4
IF B1 ≤ INSTRUCTION ≤ CF,
THEN TAG DECODED CONDITION
WATCH FOR
OCCURRENCE OF
CONDITION
INSTRUCTIONS
DURING
DOWNLOAD.
UPLOAD
PROCEDURE
4
4
COND ITION
CONDITION
HANDLER
SCRATCH
PAD
FncInit, BITS[3:0]
IF {0E01, BITS[3:0] ≠ 0}
CONDITION = 0E01, BITS[3:0]
ELSE
CONDITION = FncInit, BITS[3:0]
ENDIF
IF INSTRUCTION = B0,
THEN CLEAR ALL TAGS
EEPROM
STORE CONDITION
INSTRUCTIONS AS
THEY ARE READ FROM
THE SCRATCH PAD.
REGISTER
0x0E01, BITS[3:0]
EXECUTE/SKIP
INSTRUCTION(S)
DOWNLOAD
PROCEDURE
IF {NO TAGS} OR {CONDITION = 0}
EXECUTE INSTRUCTIONS
ELSE
IF {CONDITION IS TAGGED}
EXECUTE INSTRUCTIONS
ELSE
SKIP INSTRUCTIONS
ENDIF
ENDIF
EEPROM CONTROLLER
Figure 45. EEPROM Conditional Processing
Rev. B | Page 46 of 104
09758-025
•
Data Sheet
AD9558
Table 23 lists a sample EEPROM download instruction sequence.
It illustrates the use of condition instructions and how they alter
the download sequence. The table begins with the assumption
that no conditions are in effect. That is, the most recently executed
condition instruction is either B0 or no conditional instructions
have been processed.
Table 23. EEPROM Conditional Processing Example
Instruction
0x08
0x01
0x00
0xB1
0x19
0x04
0x00
0xB2
0xB3
0x07
0x05
0x00
0x0A
0xB0
0x80
0x0A
4.
Tag Condition 1.
Transfer the clock distribution register contents
only if tag condition = 1.
Tag Condition 2.
Tag Condition 3.
Transfer the reference input register contents only
if tag condition = 1, 2, or 3.
Calibrate the system clock only if tag condition =
1, 2, or 3.
Clear the tag condition board.
Execute an I/O update, regardless of the value of
the tag condition.
Calibrate the system clock, regardless of the value
of the tag condition.
Conditional processing makes it possible to create a number of
different device setups, store them in EEPROM, and download
a specific setup on demand. To do so, first program the device
control registers for a specific setup. Then, store an upload
sequence in the EEPROM scratch pad with the following
general form:
2.
3.
1.
2.
3.
Action
Transfer the system clock register contents,
regardless of the current condition.
Storing Multiple Device Setups in EEPROM
1.
Reprogram the device control registers for the next desired
setup. Then store a new upload sequence in the EEPROM
scratch pad with the following general form:
Condition instruction (B1 to CF) to identify the setup with
a specific condition (1 to 31)
Data instructions (to save the register contents) along with
any required calibrate and/or I/O update instructions
Pause instruction (FE)
Condition instruction (B0)
The next desired condition instruction (B1 to CF, but
different from the one used during the previous upload to
identify a new setup)
Data instructions (to save the register contents) along with
any required calibrate and/or I/O update instructions
Pause instruction (FE)
With the upload sequence written to the scratch pad, perform
an EEPROM upload (Register 0x0E02, Bit 0).
Repeat the process of programming the device control registers
for a new setup, storing a new upload sequence in the EEPROM
scratch pad (Step 1 through Step 4), and executing an EEPROM
upload (Register 0x0E02, Bit 0) until all of the desired setups
have been uploaded to the EEPROM.
Note that, on the final upload sequence stored in the scratch
pad, the pause instruction (FE) must be replaced with an end
instruction (FF).
To download a specific setup on demand, first store the condition
associated with the desired setup in Register 0x0E01, Bits[4:0].
Then perform an EEPROM download (Register 0x0E03, Bit 1).
Alternatively, to download a specific setup at power-up, apply
the required logic levels necessary to encode the desired condition
on the M2 and M3 multifunction pins. Then power up the device;
an automatic EEPROM download occurs. The condition (as
established by the M2 and M3 multifunction pins) guides the
download sequence and results in a specific setup.
Keep in mind that the number of setups that can be stored
in the EEPROM is limited. The EEPROM can hold a total of
2048 bytes. Each nondata instruction requires one byte of
storage. Each data instruction, however, requires N + 4 bytes of
storage, where N is the number of transferred register bytes and
the other four bytes include the data instruction itself (one
byte), the target address (two bytes), and the checksum
calculated by the EEPROM controller during the upload
sequence (one byte).
With the upload sequence written to the scratch pad, perform
an EEPROM upload (Register 0x0E02, Bit 0).
Rev. B | Page 47 of 104
AD9558
Data Sheet
Programming the EEPROM to Configure an M Pin to
Control Synchronization of Clock Distribution
The default EEPROM loading sequence from Register 0x0E10 to
Register 0x0E16 is unchanged. The following steps must be
inserted into the EEPROM storage sequence:
A special EEPROM loading sequence is required to use the
EEPROM to load the registers and to use an M pin to
enable/disable outputs.
To control the output sync function by using an M pin, perform
the following steps:
1.
2.
3.
Enable the M pins by writing Register 0x0200 = 0x01.
Issue an I/O update (Register 0x0005 = 0x01).
Set the appropriate M pin function (see the Clock
Distribution Synchronization section for details).
1.
2.
3.
4.
5.
6.
If this sequence is not performed, a SYNC pulse is issued
automatically.
The following changes write Register 0x0200 first and then issue
an I/O update before writing the remaining M pin configuration
registers in Register 0x0201 to Register 0x0208.
R0x0E17 = 0x00 # Write one byte
R0x0E18 = 0x02 # at Register 0x0200
R0x0E19 = 0x00 #
R0x0E1A = 0x80 # Op code for I/O
Update R0x0E1B = 0x10 # Transfer 17 instead of 18 bytes
R0x0E1C = 0x02 # Transfer starts at Register address
R0x0E1D = 0x01 # 0x0201 instead of 0x0200
The rest of the EEPROM loading sequence is the same as the
default EEPROM loading sequence, except that the register
address of the EEPROM storage sequence is shifted down four
bytes from the default. For example,
•
•
•
•
•
Rev. B | Page 48 of 104
R0x0E1E = default value of Register 0x0E1A = 0x2E
R0x0E1F = default value of Register 0x0E1B = 0x03
R0x0E20 = default value of Register 0x0E1C = 0x00
…
R0x0E40 = default value of Register 0x0E1C = 0x3C = 0xFF
(end of data)
Data Sheet
AD9558
SERIAL CONTROL PORT
The AD9558 serial control port is a flexible, synchronous serial
communications port that provides a convenient interface to
many industry-standard microcontrollers and microprocessors.
The AD9558 serial control port is compatible with most
synchronous transfer formats, including I²C, Motorola SPI, and
Intel SSR protocols. The serial control port allows read/write
access to the AD9558 register map.
In SPI mode, single or multiple byte transfers are supported.
The SPI port configuration is programmable via Register 0x0000.
This register is integrated into the SPI control logic rather than
in the register map and is distinct from the I2C Register 0x0000.
It is also inaccessible to the EEPROM controller.
Although the AD9558 supports both the SPI and I2C serial port
protocols, only one or the other is active following power-up (as
determined by the M0 and M1 multifunction pins during the
startup sequence). That is, the only way to change the serial port
protocol is to reset the device (or cycle the device power supply).
The CS (chip select) pin is an active low control that gates read
and write operations. This pin is internally connected to a 30 kΩ
pull-up resistor. When CS is high, the SDO and SDIO pins go
into a high impedance state.
E
A
E
A
SPI Mode Operation
The SPI port supports both 3-wire (bidirectional) and 4-wire
(unidirectional) hardware configurations and both MSB-first
and LSB-first data formats. Both the hardware configuration
and data format features are programmable. By default, the
AD9558 uses the bidirectional MSB-first mode. The reason that
bidirectional is the default mode is so that the user can still
write to the device, if it is wired for unidirectional operation, to
switch to unidirectional mode.
Assertion (active low) of the CS pin initiates a write or read
operation to the AD9558 SPI port. For data transfers of three
bytes or fewer (excluding the instruction word), the device
supports the CS stalled high mode (see Table 25). In this mode,
the CS pin can be temporarily deasserted on any byte boundary,
allowing time for the system controller to process the next byte.
CS can be deasserted only on byte boundaries, however. This
applies to both the instruction and data portions of the transfer.
E
E
Table 24. SPI/I²C Serial Port Setup
M0
Low
Open
High
Low
Open
High
Low
Open
High
A
E
A
Because the AD9558 supports both SPI and I²C protocols, the
active serial port protocol depends on the logic state of the
PINCONTROL, M1, and M0 pins. The PINCONTROL pin
must be low, and the state of the M0 and M1 pins determines
the I2C address, or if SPI mode is enabled. See Table 24 for the
I2C address assignments.
M1
Low
Low
Low
Open
Open
Open
High
High
High
A
A
A
SPI/I²C PORT SELECTION
A
A
E
A
A
During stall high periods, the serial control port state machine
enters a wait state until all data is sent. If the system controller
decides to abort a transfer midstream, then the state machine must
be reset either by completing the transfer or by asserting the CS
pin for at least one complete SCLK cycle (but less than eight
SCLK cycles). Deasserting the CS pin on a nonbyte boundary
terminates the serial transfer and flushes the buffer.
E
A
A
SPI/I²C
SPI
I²C, 1101000
I²C, 1101001
I²C, 1101010
I²C, 1101011
I²C, 1101100
I²C, 1101101
I²C, 1101110
I²C, 1101111
E
A
A
In streaming mode (see Table 25), any number of data bytes can
be transferred in a continuous stream. The register address is
automatically incremented or decremented. CS must be deasserted
at the end of the last byte that is transferred, thereby ending the
streaming mode.
E
A
A
Table 25. Byte Transfer Count
SPI SERIAL PORT OPERATION
Pin Descriptions
The SCLK (serial clock) pin serves as the serial shift clock. This
pin is an input. SCLK synchronizes serial control port read and
write operations. The rising edge SCLK registers write data bits,
and the falling edge registers read data bits. The SCLK pin
supports a maximum clock rate of 40 MHz.
The SDIO (serial data input/output) pin is a dual-purpose pin
and acts as either an input only (unidirectional mode) or as
both an input and an output (bidirectional mode). The AD9558
default SPI mode is bidirectional.
The SDO (serial data output) pin is useful only in unidirectional
I/O mode. It serves as the data output pin for read operations.
W1
0
0
1
1
W0
0
1
0
1
Bytes to Transfer
1
2
3
Streaming mode
Communication Cycle—Instruction Plus Data
The SPI protocol consists of a two-part communication cycle.
The first part is a 16-bit instruction word that is coincident with
the first 16 SCLK rising edges and a payload. The instruction
word provides the AD9558 serial control port with information
regarding the payload. The instruction word includes the R/W
bit that indicates the direction of the payload transfer (that is, a
read or write operation). The instruction word also indicates
the number of bytes in the payload and the starting register
address of the first payload byte.
Rev. B | Page 49 of 104
E
A
A
AD9558
Data Sheet
Write
SPI Instruction Word (16 Bits)
If the instruction word indicates a write operation, the payload
is written into the serial control port buffer of the AD9558. Data
bits are registered on the rising edge of SCLK. The length of the
transfer (1, 2, or 3 bytes or streaming mode) depends on the W0
and W1 bits (see Table 25) in the instruction byte. When not
streaming, CS can be deasserted after each sequence of eight
bits to stall the bus (except after the last byte, where it ends the
cycle). When the bus is stalled, the serial transfer resumes when
CS is asserted. Deasserting the CS pin on a nonbyte boundary
resets the serial control port. Reserved or blank registers are not
skipped over automatically during a write sequence. Therefore,
the user must know what bit pattern to write to the reserved
registers to preserve proper operation of the part. Generally, it
does not matter what data is written to blank registers, but it is
customary to write 0s.
The MSB of the 16-bit instruction word is R/W, which indicates
whether the instruction is a read or a write. The next two bits, W1
and W0, indicate the number of bytes in the transfer (see Table 25).
E
A
The final 13 bits are the register address (A12 to A0), which
indicates the starting register address of the read/write operation
(see Table 27).
E
A
A
E
SPI MSB-/LSB-First Transfers
The AD9558 instruction word and payload can be MSB first or
LSB first. The default for the AD9558 is MSB first. The LSB-first
mode can be set by writing a 1 to Register 0x0000, Bit 6.
Immediately after the LSB-first bit is set, subsequent serial
control port operations are LSB first.
E
A
A
A
A
A
When MSB-first mode is active, the instruction and data bytes
must be written from MSB to LSB. Multibyte data transfers in
MSB-first format start with an instruction byte that includes the
register address of the most significant payload byte. Subsequent
data bytes must follow in order from high address to low
address. In MSB-first mode, the serial control port internal
address generator decrements for each data byte of the multibyte transfer cycle.
Most of the serial port registers are buffered (refer to the
Buffered/Active Registers section for details on the difference
between buffered and active registers). Therefore, data written into
buffered registers does not take effect immediately. An
additional operation is required to transfer buffered serial
control port contents to the registers that actually control the
device. This is accomplished with an I/O update operation,
which is performed in one of two ways. One is by writing a
Logic 1 to Register 0x0005, Bit 0 (this bit is autoclearing). The
other is to use an external signal via an appropriately
programmed multifunction pin. The user can change as many
register bits as desired before executing an I/O update. The I/O
update operation transfers the buffer register contents to their
active register counterparts.
When Register 0x0000, Bit 6 = 1 (LSB first), the instruction and
data bytes must be written from LSB to MSB. Multibyte data
transfers in LSB-first format start with an instruction byte that
includes the register address of the least significant payload byte
followed by multiple data bytes. The serial control port internal
byte address generator increments for each byte of the multibyte
transfer cycle.
For multibyte MSB-first (default) I/O operations, the serial
control port register address decrements from the specified
starting address toward Address 0x0000. For multibyte LSB-first
I/O operations, the serial control port register address
increments from the starting address toward Address 0x1FFF.
Reserved addresses are not skipped during multibyte I/O
operations; therefore, the user should write the default value to
a reserved register and 0s to unmapped registers. Note that it is
more efficient to issue a new write command than to write the
default value to more than two consecutive reserved (or
unmapped) registers.
Read
The AD9558 supports the long instruction mode only. If the
instruction word indicates a read operation, the next N × 8
SCLK cycles clock out the data from the address specified in the
instruction word. N is the number of data bytes read and
depends on the W0 and W1 bits of the instruction word. The
readback data is valid on the falling edge of SCLK. Blank
registers are not skipped over during readback.
A readback operation takes data from either the serial control
port buffer registers or the active registers, as determined by
Register 0x0004, Bit 0.
Table 26. Streaming Mode (No Addresses Are Skipped)
Write Mode
LSB First
MSB First
Address Direction
Increment
Decrement
Stop Sequence
0x0000 ... 0x1FFF
0x1FFF ... 0x0000
Table 27. Serial Control Port, 16-Bit Instruction Word, MSB First
MSB
I15
I14
I13
I12
I11
I10
I9
I8
I7
I6
I5
I4
I3
I2
I1
LSB
I0
R/W
W1
W0
A12
A11
A10
A9
A8
A7
A6
A5
A4
A3
A2
A1
A0
A
E
Rev. B | Page 50 of 104
Data Sheet
AD9558
CS
SCLK DON'T CARE
R/W W1 W0 A12 A11 A10 A9
SDIO DON'T CARE
A8
A7
A6 A5
A4 A3 A2
A1 A0
D7 D6 D5
16-BIT INSTRUCTION HEADER
D4 D3
D2 D1
D0
D7
D6 D5
REGISTER (N) DATA
D4 D3 D2
D1 D0
DON'T CARE
REGISTER (N – 1) DATA
09758-029
DON'T CARE
Figure 46. Serial Control Port Write—MSB First, 16-Bit Instruction, Two Bytes of Data
CS
SCLK
DON'T CARE
SDIO
DON'T CARE
SDO DON'T CARE
D7 D6 D5 D4 D3 D2 D1 D0 D7 D6 D5 D4 D3 D2 D1 D0 D7 D6 D5 D4 D3 D2 D1 D0 D7 D6 D5 D4 D3 D2 D1 D0
REGISTER (N) DATA
16-BIT INSTRUCTION HEADER
REGISTER (N – 1) DATA
REGISTER (N – 2) DATA
REGISTER (N – 3) DATA
DON'T
CARE
09758-030
R/W W1 W0 A12 A11 A10 A9 A8 A7 A6 A5 A4 A3 A2 A1 A0
Figure 47. Serial Control Port Read—MSB First, 16-Bit Instruction, Four Bytes of Data
tDS
tHIGH
tS
tLOW
CS
DON'T CARE
SDIO
DON'T CARE
DON'T CARE
R/W
W1
W0
A12
A11
A10
A9
A8
A7
A6
A5
D4
D3
D2
D1
D0
DON'T CARE
09758-031
SCLK
tC
tCLK
tDH
Figure 48. Serial Control Port Write—MSB First, 16-Bit Instruction, Timing Measurements
CS
SCLK
DATA BIT N
09758-032
tDV
SDIO
SDO
DATA BIT N – 1
Figure 49. Timing Diagram for Serial Control Port Register Read
CS
SCLK DON'T CARE
A4
A5 A6
A7
A8
A9 A10 A11 A12 W0 W1 R/W D0 D1 D2 D3 D4
16-BIT INSTRUCTION HEADER
D5 D6
REGISTER (N) DATA
D7
D0
D1 D2
D6
REGISTER (N + 1) DATA
Figure 50. Serial Control Port Write—LSB First, 16-Bit Instruction, Two Bytes of Data
Rev. B | Page 51 of 104
D3 D4 D5
D7
DON'T CARE
09758-033
SDIO DON'T CARE
DON'T CARE
A0 A1 A2 A3
AD9558
Data Sheet
CS
tS
tC
tCLK
tHIGH
tLOW
tDS
SCLK
SDIO
BIT N
BIT N + 1
Figure 51. Serial Control Port Timing—Write
Table 28. Serial Control Port Timing
Parameter
tDS
tDH
tCLK
tS
tC
tHIGH
tLOW
tDV
Description
Setup time between data and the rising edge of SCLK
Hold time between data and the rising edge of SCLK
Period of the clock
Setup time between the CS falling edge and the SCLK rising edge (start of the communication cycle)
Setup time between the SCLK rising edge and CS rising edge (end of the communication cycle)
Minimum period that SCLK should be in a logic high state
Minimum period that SCLK should be in a logic low state
SCLK to valid SDIO and SDO (see Figure 49)
E
A
A
E
A
A
Rev. B | Page 52 of 104
09758-034
tDH
Data Sheet
AD9558
The transfer of data is shown in Figure 52. One clock pulse is
generated for each data bit transferred. The data on the SDA
line must be stable during the high period of the clock. The
high or low state of the data line can change only when the
clock signal on the SCL line is low.
I²C SERIAL PORT OPERATION
The I2C interface has the advantage of requiring only two control
pins and is a de facto standard throughout the I2C industry.
However, its disadvantage is programming speed, which is
400 kbps maximum. The AD9558 I²C port design is based on the
I²C fast mode standard; therefore, it supports both the 100 kHz
standard mode and 400 kHz fast mode. Fast mode imposes a
glitch tolerance requirement on the control signals. That is, the
input receivers ignore pulses of less than 50 ns duration.
SDA
SCL
DATA LINE
STABLE;
DATA VALID
Figure 52. Valid Bit Transfer
Start/stop functionality is shown in Figure 53. The start
condition is characterized by a high-to-low transition on the
SDA line while SCL is high. The start condition is always
generated by the master to initialize a data transfer. The stop
condition is characterized by a low-to-high transition on the
SDA line while SCL is high. The stop condition is always
generated by the master to terminate a data transfer. Every byte
on the SDA line must be eight bits long. Each byte must be
followed by an acknowledge bit; bytes are sent MSB first.
The AD9558 allows up to seven unique slave devices to occupy
the I2C bus. These are accessed via a 7-bit slave address that is
transmitted as part of an I2C packet. Only the device that has a
matching slave address responds to subsequent I2C commands.
Table 24 lists the supported device slave addresses.
I2C Bus Characteristics
A summary of the various I2C protocols appears in Table 29.
The acknowledge bit (A) is the ninth bit attached to any 8-bit
data byte. An acknowledge bit is always generated by the
receiving device (receiver) to inform the transmitter that the
byte has been received. It is done by pulling the SDA line low
during the ninth clock pulse after each 8-bit data byte.
Table 29. I2C Bus Abbreviation Definitions
Abbreviation
S
Sr
P
A
A
W
R
Definition
Start
Repeated start
Stop
Acknowledge
Nonacknowledge
Write
Read
E
E
A
The nonacknowledge bit (A) is the ninth bit attached to any 8bit data byte. A nonacknowledge bit is always generated by the
receiving device (receiver) to inform the transmitter that the
byte has not been received. It is done by leaving the SDA line
high during the ninth clock pulse after each 8-bit data byte.
E
A
A
SDA
SCL
S
P
START CONDITION
09758-036
STOP CONDITION
Figure 53. Start and Stop Conditions
MSB
SDA
ACK FROM
SLAVE RECEIVER
1
SCL
2
3 TO 7
8
9
ACK FROM
SLAVE RECEIVER
1
S
Figure 54. Acknowledge Bit
Rev. B | Page 53 of 104
2
3 TO 7
8
9
10
P
09758-037
A
CHANGE
OF DATA
ALLOWED
09758-035
The AD9558 I²C port consists of a serial data line (SDA) and a
serial clock line (SCL). In an I²C bus system, the AD9558 is
connected to the serial bus (data bus SDA and clock bus SCL)
as a slave device; that is, no clock is generated by the AD9558.
The AD9558 uses direct 16-bit memory addressing instead of
traditional 8-bit memory addressing.
AD9558
Data Sheet
Data Transfer Process
per transfer is unrestricted. In write mode, the first two data
The master initiates data transfer by asserting a start condition.
This indicates that a data stream follows. All I²C slave devices
connected to the serial bus respond to the start condition.
bytes immediately after the slave address byte are the internal
memory (control registers) address bytes, with the high address
byte first. This addressing scheme gives a memory address of up
to 216 − 1 = 65,535. The data bytes after these two memory address
bytes are register data written to or read from the control registers.
In read mode, the data bytes after the slave address byte are
register data written to or read from the control registers.
The master then sends an 8-bit address byte over the SDA line,
consisting of a 7-bit slave address (MSB first) plus an R/W bit.
This bit determines the direction of the data transfer, that is,
whether data is written to or read from the slave device (0 =
write, 1 = read).
E
A
A
When all data bytes are read or written, stop conditions are
established. In write mode, the master (transmitter) asserts
a stop condition to end data transfer during the 10th clock pulse
following the acknowledge bit for the last data byte from the slave
device (receiver). In read mode, the master device (receiver)
receives the last data byte from the slave device (transmitter) but
does not pull SDA low during the ninth clock pulse. This is known
as a nonacknowledge bit. By receiving the nonacknowledge bit,
the slave device knows that the data transfer is finished and
enters idle mode. The master then takes the data line low
during the low period before the 10th clock pulse, and high
during the 10th clock pulse to assert a stop condition.
The peripheral whose address corresponds to the transmitted
address responds by sending an acknowledge bit. All other
devices on the bus remain idle while the selected device waits
for data to be read from or written to it. If the R/W bit is 0, the
master (transmitter) writes to the slave device (receiver). If the
R/W bit is 1, the master (receiver) reads from the slave device
(transmitter).
E
A
A
E
A
The format for these commands is described in the Data Transfer
Format section.
Data is then sent over the serial bus in the format of nine clock
pulses, one data byte (eight bits) from either master (write mode)
or slave (read mode) followed by an acknowledge bit from the
receiving device. The number of bytes that can be transmitted
A start condition can be used in place of a stop condition.
Furthermore, a start or stop condition can occur at any time,
and partially transferred bytes are discarded.
MSB
ACK FROM
SLAVE RECEIVER
1
SCL
2
3 TO 7
8
9
ACK FROM
SLAVE RECEIVER
1
2
3 TO 7
8
9
S
10
P
09758-038
SDA
Figure 55. Data Transfer Process (Master Write Mode, 2-Byte Transfer)
SDA
ACK FROM
MASTER RECEIVER
SCL
1
2
3 TO 7
8
9
NON-ACK FROM
MASTER RECEIVER
1
2
3 TO 7
S
8
9
10
P
Figure 56. Data Transfer Process (Master Read Mode, 2-Byte Transfer)
Rev. B | Page 54 of 104
09758-039
A
Data Sheet
AD9558
Data Transfer Format
Write byte format—the write byte protocol is used to write a register address to the RAM starting from the specified RAM address.
S
Slave
address
A
W
E
A
RAM address
high byte
A
RAM address
low byte
A
RAM
Data 0
A
RAM
Data 1
A
RAM
Data 2
A
P
Send byte format—the send byte protocol is used to set up the register address for subsequent reads.
S
Slave address
A
W
E
A
RAM address high byte
A
RAM address low byte
A
P
Receive byte format—the receive byte protocol is used to read the data byte(s) from RAM starting from the current address.
S
Slave address
R
A
RAM Data 0
A
RAM Data 1
A
RAM Data 2
P
A
A
E
Read byte format—the combined format of the send byte and the receive byte.
S
Slave
Address
W
E
A
A
RAM
Address
High Byte
A
RAM
Address
Low Byte
A
Sr
Slave
Address
R
A
RAM
Data 0
A
RAM
Data 1
A
RAM
Data 2
A
A
I²C Serial Port Timing
SDA
tLOW
tF
tSU; DAT
tR
tHD; STA
tSP
tBUF
tR
tF
tHD; STA
S
tHD; DAT
tHIGH
tSU; STO
tSU; STA
Sr
Figure 57. I²C Serial Port Timing
Table 30. I²C Timing Definitions
Parameter
fSCL
tBUF
tHD; STA
tSU; STA
tSU; STO
tHD; DAT
tSU; DAT
tLOW
tHIGH
tR
tF
tSP
Description
Serial clock
Bus free time between stop and start conditions
Repeated hold time start condition
Repeated start condition setup time
Stop condition setup time
Data hold time
Date setup time
SCL clock low period
SCL clock high period
Minimum/maximum receive SCL and SDA rise time
Minimum/maximum receive SCL and SDA fall time
Pulse width of voltage spikes that must be suppressed by the input filter
Rev. B | Page 55 of 104
P
S
09758-040
SCL
E
P
AD9558
Data Sheet
PROGRAMMING THE I/O REGISTERS
The register map spans an address range from 0x0000 through
0x0E3C. Each address provides access to 1 byte (eight bits) of
data. Each individual register is identified by its four-digit
hexadecimal address (for example, Register 0x0A10). In some
cases, a group of addresses collectively defines a register.
In general, when a group of registers defines a control
parameter, the LSB of the value resides in the D0 position of the
register with the lowest address. The bit weight increases right
to left, from the lowest register address to the highest register
address.
Note that the EEPROM storage sequence registers (Address
0x0E10 to Address 0x0E3C) are an exception to the above
convention (see the EEPROM Instructions section).
REGISTER ACCESS RESTRICTIONS
Read and write access to the register map may be restricted
depending on the register in question, the source and direction
of access, and the current state of the device. Each register can
be classified into one or more access types. When more than
one type applies, the most restrictive condition is the one that
applies.
When access is denied to a register, all attempts to read the
register return a 0 byte, and all attempts to write to the register
are ignored. Access to nonexistent registers is handled in the
same way as for a denied register.
Regular Access
BUFFERED/ACTIVE REGISTERS
There are two copies of most registers: buffered and active. The
value in the active registers is the one that is in use. The buffered
registers are the ones that take effect the next time the user
writes 0x01 to the I/O update register (Register 0x0005).
Buffering the registers allows the user to update a group of
registers (like the digital loop filter coefficients) at the same
time, which avoids the potential of unpredictable behavior in
the part. Registers with an L in the option column are live,
meaning that they take effect the moment the serial port
transfers that data byte.
AUTOCLEAR REGISTERS
An A in the option column of the register map identifies an
autoclear register. Typically, the active value for an autoclear
register takes effect following an I/O update. The bit is cleared
by the internal device logic upon completion of the prescribed
action.
Registers with regular access do not fall into any other category.
Both read and write access to registers of this type can be from
either the serial ports or the EEPROM controller. However, only
one of these sources can have access to a register at any given
time (access is mutually exclusive). When the EEPROM controller
is active, in either load or store mode, it has exclusive access to
these registers.
Read-Only Access
An R in the option column of the register map identifies readonly registers. Access is available at all times, including when
the EEPROM controller is active. Note that read-only registers
(R) are inaccessible to the EEPROM, as well.
Exclusion from EEPROM Access
An E in the option column of the register map identifies a
register with contents that are inaccessible to the EEPROM.
That is, the contents of this type of register cannot be
transferred directly to the EEPROM or vice versa. Note that
read-only registers (R) are inaccessible to the EEPROM, as well.
Rev. B | Page 56 of 104
Data Sheet
AD9558
THERMAL PERFORMANCE
Table 31. Thermal Parameters for the 64-Lead LFCSP Package
Symbol
θJA
θJMA
θJMA
θJB
θJC
ΨJT
1
2
Thermal Characteristic Using a JEDEC51-7 Plus JEDEC51-5 2S2P Test Board1
Junction-to-ambient thermal resistance, 0.0 m/sec airflow per JEDEC JESD51-2 (still air)
Junction-to-ambient thermal resistance, 1.0 m/sec airflow per JEDEC JESD51-6 (moving air)
Junction-to-ambient thermal resistance, 2.5 m/sec airflow per JEDEC JESD51-6 (moving air)
Junction-to-board thermal resistance, 0.0 m/sec airflow per JEDEC JESD51-8 (still air)
Junction-to-case thermal resistance (die-to-heat sink) per MIL-Std 883, Method 1012.1
Junction-to-top-of-package characterization parameter, 0 m/sec airflow per JEDEC JESD51-2 (still air)
Value2
21.7
18.9
16.9
11.3
1.2
0.1
Unit
°C/W
°C/W
°C/W
°C/W
°C/W
°C/W
The exposed pad on the bottom of the package must be soldered to ground to achieve the specified thermal performance.
Results are from simulations. The PCB is a JEDEC multilayer type. Thermal performance for actual applications requires careful inspection of the conditions in the
application to determine if they are similar to those assumed in these calculations.
The AD9558 is specified for a case temperature (TCASE). To ensure
that TCASE is not exceeded, an airflow source can be used. Use
the following equation to determine the junction temperature
on the application PCB:
TJ = TCASE + (ΨJT × PD)
Values of θJA are provided for package comparison and PCB
design considerations. θJA can be used for a first order approximation of TJ by the equation
TJ = TA + (θJA × PD)
where TA is the ambient temperature (°C).
where:
TJ is the junction temperature (°C).
TCASE is the case temperature (°C) measured by the customer at
the top center of the package.
ΨJT is the value as indicated in Table 31.
PD is the power dissipation (see Table 3).
Values of θJC are provided for package comparison and PCB
design considerations when an external heat sink is required.
Values of θJB are provided for package comparison and PCB
design considerations.
Rev. B | Page 57 of 104
AD9558
Data Sheet
POWER SUPPLY PARTITIONS
The AD9558 power supplies are divided into four groups:
DVDD3, DVDD, AVDD3, and AVDD. All power and ground
pins should be connected, even if certain blocks of the chip are
powered down.
RECOMMENDED CONFIGURATION FOR 3.3 V
SWITCHING SUPPLY
The ADP7104 is another good choice for converting 3.3 V to
1.8 V. The close-in noise of the ADP7104 is lower than that of
the ADP222; therefore, it may be better suited for applications
where close-in phase noise is critical and the AD9557 DPLL loop
bandwidth is <50 Hz. In such cases, all 1.8 V supplies can be
connected to one ADP7104.
Use of Ferrite Beads on 1.8 V Supplies
A popular power supply arrangement is to power the AD9558
with the output of a 3.3 V switching power supply.
CONFIGURATION FOR 1.8 V SUPPLY
To ensure the very best output-to-output isolation, one ferrite
bead should be used instead of a bypass capacitor for each of
the following AVDD pins: Pin 12, Pin 22, Pin 29, and Pin 30. The
ferrite beads should be placed in between the 1.8 V LDO output
and each pin listed above. Ferrite beads that have low (<0.7 Ω)
dc resistance and approximately 600 Ω impedance at 100 MHz
are suitable for this application.
When 1.8 V supplies are preferred, it is recommended that an
LDO regulator, such as the ADP222, be used to generate the
1.8 V supply from the 3.3 V supply.
See Table 2 for the current consumed by each group. Refer to
Figure 20, Figure 21, and Figure 22 for information on the
power consumption vs. output frequency.
When the AD9558 is powered using 3.3 V switching power
supplies, all of the 3.3 V supplies can be connected to the 3.3 V
switcher output, and a 0.1 µF bypass capacitor should be placed
adjacent to each 3.3 V power supply pin.
The ADP222 offers excellent power supply rejection in a small
(2 mm × 2 mm) package. It has two 1.8 V outputs. One output
can be used for the DVDD pins (Pin 6, Pin 34, and Pin 35), and
the other output can drive the AVDD pins.
Rev. B | Page 58 of 104
Data Sheet
AD9558
PIN PROGRAM FUNCTION DESCRIPTION
The AD9558 supports both hard pin and soft pin program
function with the on-chip ROM containing the predefined
configurations. When a pin program function is enabled and
initiated, the selected predefined configuration is transferred
from the ROM to the corresponding registers to configure the
part into the desired state.
OVERVIEW OF ON-CHIP ROM FEATURES
Input/Output Frequency Translation Configuration
The AD9558 has one on-chip ROM that contains a total of 256
different input-output frequency translation configurations for
independent selection of 16 input frequencies and 16 output
frequencies. Each input/output frequency translation configuration
assumes that all input frequencies are the same and all the output
frequencies are the same. Each configuration reprograms the
following registers/parameters:
•
•
•
•
•
Reference input period register
Reference divider R register
Digital PLL feedback divider register (Fractional Part
FRAC1, Modulus Part MOD1 and Integer Part N1) free run
Tuning word register
Output PLL feedback divider N2 register
•
•
RF divider register
Clock distribution channel divider register
All configurations are set to support one single system clock
frequency as 786.432 MHz (16× the default 49.152 MHz system
clock reference frequency).
Four Different System Clock PLL Configurations
•
•
•
•
REF = 49.152 MHz XO (×2 on, N = 8)
REF = 49.152 MHz XTAL (×2 on, N = 8)
REF = 24.756 MHz XTAL (×2 on, N = 16)
REF = 98.304 MHz XO (×2 off, N = 8)
Four Different DPLL Loop Bandwidths
•
1 Hz, 10 Hz, 50 Hz, 100 Hz
DPLL Phase Margin
•
•
Normal phase margin (70°)
High phase margin (88.5°)
The ROM also contains an APLL VCO calibration bit. This bit
is used to program Register 0x0405[0] (from 0) to 1 to generate
a low-high transition to automatically initiate APLL VCO cal.
Table 32. Preset Input Frequencies for Hard Pin and Soft Pin Programming
Freq ID
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Frequency (MHz)
0.008
19.44
25
125
156.7072
622.08
625
644.53125
657.421875
660.184152
669.3266
672.1627
690.569
693.48299
693.482991
698.81236
Frequency Description
8 kHz
19.44 MHz
25 MHz
125 MHz
156..25 MHz × 1027/1024
622.08 MHz
625 MHz
625 MHz × 33/32
657.421875 MHz
657.421875 MHz × 239/238
622.08 MHz × 255/238
622.08 MHz × 255/236
622.08 MHz × 255/236
644.53125 MHz × 255/238
644.53125 MHz × 255/237
622.08 × 255/237
Hard Pin Program
PINCONTROL = High
M4 Pin
M0 Pin
M5 Pin
0
0
0
0
0
½
0
0
1
0
½
0
0
½
½
0
½
1
0
1
0
0
1
½
0
1
1
½
0
0
½
0
½
½
0
1
½
½
0
½
½
½
½
½
1
½
1
0
Rev. B | Page 59 of 104
Soft Pin Program
PINCONTROL = Low
Register 0x0C01[3:0]
B3
B2
B1
B0
0
0
0
0
0
0
0
1
0
0
1
0
0
0
1
1
0
1
0
0
0
1
0
1
0
1
1
0
0
1
1
0
1
0
0
1
1
0
0
0
1
0
1
1
1
0
1
0
1
1
0
1
1
1
0
0
1
1
1
1
1
1
1
0
AD9558
Data Sheet
Table 33. Preset Output Frequencies for Hard Pin and Soft Pin Programming
Freq ID
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Frequency (MHz)
19.44
25
125
156.7071
622.08
625
644.53125
657.421875
660.184152
666.5143
669.3266
672.1627
690.5692
693.4830
698.8124
704.380580
Hard Pin Program
PINCONTROL = High
M3
M2
M1
0
0
0
0
0
½
0
0
1
0
½
0
0
½
½
0
½
1
0
1
0
0
1
½
0
1
1
½
0
0
½
0
½
½
0
1
½
½
0
½
½
½
½
½
1
½
1
0
Frequency Description
19.44 MHz
25 MHz
125 MHz
156.25 MHz × 1027/1024
622.08 MHz
625 MHz
625 MHz × 33/32
657.421875 MHz
657.421875 MHz × 239/238
622.08 MHz × 255/238
622.08 MHz × 255/237
622.08 MHz × 255/236
644.53125 MHz × 255/238
644.53125 MHz × 255/237
622.08 MHz × 255/237
657.421875 MHz × 255/238
B7
0
0
0
0
0
0
0
1
1
1
1
1
1
1
1
Soft Pin Program
PINCONTROL = Low,
Register 0x0C01[3:0]
B6
B5
0
0
0
0
0
1
0
1
1
0
1
0
1
1
1
1
0
0
0
0
0
1
0
1
1
0
1
0
1
1
1
1
B4
0
1
0
1
0
1
0
1
0
1
0
1
0
1
0
1
Table 34. System Clock Configuration in Hard Pin and Soft Pin Programming Modes
Freq ID
0
1
2
3
Frequency (MHz)
49.152
49.152
24.576
98.304
System Clock Configuration
XTAL mode, doubler on, N = 8
XTAL mode off, doubler on, N = 8
XTAL mode, doubler on, N = 16
XTAL mode off, doubler off, N = 8
HARD PIN PROGRAMMING MODE
The state of the PINCONTROL pin at power-up controls whether
or not the chip is in hard pin programming mode. Setting the
PINCONTROL pin high disables the I2C protocol, although the
register map can be accessed via the SPI protocol.
The M0, M5, and M4 pins select one of 16 input frequencies, and
the M3 to M1 pins select one of 16 possible output frequencies. See
Table 32 and Table 33 for details.
Hard Pin Program
PINCONTROL = High,
IRQ Pin
IRQ Pin
0
½
1
N/A
Soft Pin Program
PINCONTROL = Low,
Register 0x0C02[1:0]
Bit 1
Bit 0
0
0
0
1
1
0
1
1
Equivalent
System Clock
PLL Register
Settings
0001, 0000, 1000
The system clock configuration is controlled by the state of the
IRQ pin at startup (see Table 34 for details). The digital PLL
loop bandwidth, reference input frequency accuracy tolerance
ranges, and DPLL phase margin selection are not available in
hard pin programming mode unless the user uses the serial port
to change their default values.
When in hard pin programming mode, the user must set
Register 0x0200[0] = 1 to activate the IRQ, REF status, and PLL
lock status signals at the multifunction pins.
Rev. B | Page 60 of 104
Data Sheet
AD9558
•
SOFT PIN PROGRAMMING OVERVIEW
The soft pin program function is controlled by a dedicated
register section (Address 0x0C00 to Address 0x0C08). The
purpose of soft pin program is to use the register bits to mimic
the hard pins for the configuration section. When in soft pin
program mode, both SPI and I2C port are available.
•
•
•
•
•
•
Address 0x0C00[0] enables accessibility to Address 0x0C01
and Address 0x0C02 (Soft Pin Section 1). This bit must be
set in soft pin mode.
Address 0x0C03[0] enables accessibility to Address 0x0C04 to
Address 0x0C06 (Soft Pin Section 2). This bit must be set
in soft pin mode.
Address 0x0C01[3:0] select one of 16 input frequencies.
Address 0x0C01[7:4] select one of 16 output frequencies.
Address 0x0C02[1:0] select the system clock configuration.
Address 0x0C06[1:0] select one of four input frequency
tolerance ranges.
•
•
•
Rev. B | Page 61 of 104
Address 0x0C06[3:2] select one of four DPLL loop
bandwidths.
Address 0x0C06[4] selects the DPLL phase margin.
Address 0x0C04[3:0] scales the REFA/REFB/REFC/REFD
input frequency down by divide-by-1, divide-by-4, divideby-8, divide-by-16 independently. For example, when
Address 0x0C01[3:0] = 0101 to select input frequency as
622.08 MHz for all REFA/REFB/REFC/REFD, setting
Address 0x0C04[1:0] = 0x01 scales down the REFA input
frequency to 155.52 MHz (= 622.08 MHz/4). This is done
by internally scaling the R divider for REFA up by 4× and
the REFA period up by 4×.
Address 0x0C05[3:0] scales the Channel 0/1/2/3 output
frequency down by divide-by-1, divide-by-4, divide-by-8,
or divide-by-16. (Channel 2 and Channel 3 are available
only for the AD9558).
AD9558
Data Sheet
REGISTER MAP
Register addresses that are not listed in Table 35 are not used, and writing to those registers has no effect. The user should write the
default value to sections of registers marked reserved. R = read-only. A = autoclear. E = excluded from EEPROM loading. L = live (I/O
update not required for register to take effect or for a read-only register to be updated).
Table 35. Register Map
Reg
Addr
(Hex)
Opt
Name
D7
D6
Serial Control Port Configuration and Part Identification
0x0000
L, E
SPI control
SDO enable
LSB first/
increment
address
0x0000
L
I²C control
Reserved
0x0004
0x0005
A, L
0x0006
L
0x0007
L
0x000A
R, L
0x000B
R, L
0x000C
R, L
0x000D
R, L
System Clock
0x0100
0x0101
0x0102
0x0103
0x0104
0x0105
0x0106
0x0107
0x0108
Readback
control
I/O update
User scratch
pad
D5
D4
D3
D2
Reserved
Reserved
SYSCLK period
Reserved
SYSCLK
stability
A
Reserved
General Configuration
0x0200
EN_MPIN
Def
Reserved
00
Soft reset
Reserved
00
Reserved
User scratch pad[7:0]
User scratch pad[15:8]
Silicon revision[7:0]
Reserved
Clock part family ID[7:0]
Clock part family ID[15:8]
SYSCLK config
PLL feedback
divider
D0
Soft reset
Reserved
Silicon rev
Reserved
Part ID
D1
Read buffer
register
I/O update
System clock N divider[7:0]
Load from
SYSCLK XTAL
SYSCLK P divider[1:0]
SYSCLK
ROM (readenable
doubler
only)
enable
Reserved
Nominal system clock period (fs), Bits[7:0] (1 ns at 1 ppm accuracy)
Nominal system clock period (fs), Bits[15:8] (1 ns at 1 ppm accuracy)
Nominal system clock period (fs), Bits[20:16]
System clock stability period (ms), Bits[7:0]
System clock stability period (ms), Bits[15:8]
Reset
System clock stability period (ms), Bits[19:16]
SYSCLK
(not autoclearing)
stab timer
(autoclear)
Reserved
Enable M pins
and IRQ pin
function
00
00
00
00
21
0F
01
00
08
09
or
19
00
0E
67
13
32
00
00
00
0x0201
M0FUNC
Output/inputE
Function[6:0]
B0
0x0202
M1FUNC
Output/AinputE
Function[6:0]
B1
0x0203
M2FUNC
Output/AinputE
Function[6:0]
C0
0x0204
M3FUNC
Output/AinputE
Function[6:0]
C1
0x0205
M4FUNC
Output/AinputE
Function[6:0]
B2
0x0206
M5func
Output/AinputE
Function[6:0]
B3
0x0207
M6FUNC
Output/AinputE
Function[6:0]
C2
0x0208
M7FUNC
Output/AinputE
Function[6:0]
C3
A
E
A
E
A
E
A
E
A
E
A
E
A
A
E
E
Rev. B | Page 62 of 104
Data Sheet
Reg
Addr
(Hex)
0x0209
Opt
0x020A
Name
IRQ pin
output mode
AD9558
D7
IRQ mask
D6
Reserved
Reserved
0x020B
D4
SYSCLK
unlocked
SYSCLK
locked
Freerun
Pin
program
end
Holdover
Reserved
0x020C
Switching
0x020D
Closed
Reserved
0x020E
Reserved
REFB
validated
0x020F
Reserved
REFD
validated
0x0210
0x0211
0x0300
0x0301
0x0302
0x0303
0x0304
REFB
fault
cleared
REFD
fault
cleared
Watchdog
Timer 1
Freerun
frequency TW
Reserved
Reserved
Digital
oscillator
control
0x0305
0x0306
0x0307
0x0308
0x0309
0x030A
0x030B
0x030C
0x030D
0x030E
0x030F
0x0310
0x0311
0x0312
0x0313
0x0314
0x0315
0x0316
0x0317
0x0318
0x0319
0x031A
0x031B
0x031C
0x031D
0x031E
0x031F
0x0320
0x0321
0x0322
D5
DCO
4-level
output
D3
Status signal
at IRQ pin[1:0]
History
updated
REFB
fault
REFD fault
D1
APLL
unlocked
D2
Use IRQ pin
for status
signal
APLL
locked
D0
IRQ pin driver type[1:0]
APLL cal
complete
APLL cal
started
00
Sync
distribution
Watchdog
timer
EEPROM fault
EEPROM
complete
00
Frequency
unlocked
Frequency
unclamped
Reserved
Frequency
locked
Frequency
clamped
REFA
validated
Phase unlocked
Phase locked
00
Phase slew
unlimited
REFA fault
cleared
Phase slew
limited
REFA fault
00
Reserved
REFC
validated
REFC fault
cleared
REFC fault
00
Reset SDM
00
00
11
15
64
1B
10
Watchdog timer (ms), Bits[7:0]
Watchdog timer (ms), Bits[15:8]
30-bit free run frequency tuning word[7:0]
30-bit free run frequency tuning word[15:8]
30-bit free run frequency tuning word[23:16]
30-bit free run frequency tuning word[29:24]
Reserved
Reserved
Reserved
Lower limit of pull-in range[19:16]
Upper limit of pull-in range[7:0]
Upper limit of pull-in range[15:8]
Reserved
Closed-loop
phase lock
offset
[±0.5 ms]
Phase slew
rate limit
Holdover
history
History mode
L
L
L
L
L
L
L
L
L
L
L
L
Base Loop
Filter A
coefficient set
(high phase
margin)
Reserved
Reserved
Reserved
Reserved
00
(must be 1b)
Reserved
Lower limit of pull-in range[7:0]
Lower limit of pull-in range[15:8]
DPLL
frequency
clamp
Def
1F
Upper limit of pull-in range[19:16]
Fixed phase lock offset (signed; ps), Bits[7:0]
Fixed phase lock offset (signed; ps), Bits[15:8]
Fixed phase lock offset (signed; ps), Bits[23:16]
Reserved
Fixed phase lock offset (signed; ps), Bits[29:24]
Incremental phase lock offset step size (ps/step), Bits[7:0] (up to 65.5 ns/step)
Incremental phase lock offset step size (ps/step), Bits[15:8] (up to 65.5 ns/step)
Phase slew rate limit (µs/sec), Bits[7:0] (315 µs/sec up to 65.536 ms/sec)
Phase slew rate limit (µs/sec), Bits[15:8] (315 µs/sec up to 65.536 ms/sec)
History accumulation timer (ms), Bits[7:0] (up to 65 seconds)
History accumulation timer (ms), Bits[15:8] (up to 65 seconds)
Reserved
Single
Persistent
Incremental average
sample
history
fallback
HPM Alpha-0[7:0]
HPM Alpha-0[15:8]
HPM Alpha-1[6:0]
HPM Beta-0[7:0]
HPM Beta-0[15:8]
HPM Beta-1[6:0]
HPM Gamma-0[7:0]
HPM Gamma-0[15:8]
HPM Gamma-1[6:0]
HPM Delta-0[7:0]
HPM Delta-0[15:8]
HPM Delta-1[6:0]
Rev. B | Page 63 of 104
00
51
B8
02
3E
0A
0B
00
00
00
00
00
00
00
00
0A
00
00
8C
AD
4C
F5
CB
73
24
D8
59
D2
8D
5A
AD9558
Reg
Addr
(Hex)
0x0323
0x0324
0x0325
0x0326
0x0327
0x0328
0x0329
0x032A
0x032B
0x032C
0x032D
0x032E
Opt
L
L
L
L
L
L
L
L
L
L
L
L
Data Sheet
Name
Base loop
Filter A
coefficient set
(normal phase
margin of 70º)
D7
D6
D5
D4
Reserved
Reserved
Reserved
Reserved
Output PLL (APLL)
0x0400
APLL charge
pump
0x0401
APLL N divider
0x0402
0x0403
APLL loop
filter control
0x0404
0x0405
0x0406
0x0407
0x0408
APLL VCO
control
0x0501
Channel 0
0x0502
0x0503
0x0504
0x0505
Channel 1
0x0507
0x0508
Channel 2
0x050B
0x050C
0x050D
0x050E
0x050F
0x0510
0x0511
0x0512
0x0513
0x0514
0x0515
Mask
Channel 3
sync
Enable 3.3 V
CMOS driver
Mask
Channel 2
sync
Channel 3
D0
Def
24
8C
49
55
C9
7B
9C
FA
55
EA
E2
57
81
Output PLL (APLL) feedback N divider[7:0]
Reserved
APLL loop filter control[7:0]
Reserved
14
00
07
00
APLL locked
controlled
sync disable
Reserved
RF divider
startup
mode
Mask
Mask
Channel 1
Channel 0
sync
sync
OUT0 format[2:0]
Reserved
Reserved
Bypass
internal Rzero
Manual APLL
VCO (not
autoclearing)
Reserved
RF Divider 1[3:0]
PD RF Divider 2
Sync
source
selection
OUT0 polarity[1:0]
PD RF Divider 1
Automatic sync mode
OUT0 drive
strength
Enable OUT0
Channel 0 (M0) division ratio[7:0]
Channel 0 PD
Select RF
Channel 0 (M0) division ratio[9:8]
Divider 2
Reserved
Channel 0 divider phase[5:0]
Reserved
OUT1 format[2:0]
OUT1 polarity[1:0]
OUT1 drive
Enable OUT1
strength
Reserved
OUT2 format[2:0]
OUT2 polarity[1:0]
OUT2 drive
Enable OUT2
strength
Channel 1 (M1) division ratio[7:0]
Reserved
Channel 1 PD
Select RF
Channel 1 (M1) division ratio[9:8]
Divider 2
Reserved
Channel 1 divider phase[5:0]
Reserved
OUT3 format[2:0]
OUT3 polarity[1:0]
OUT3 drive
Enable OUT3
strength
Reserved
OUT4 format[2:0]
OUT4 polarity[1:0]
OUT4 drive
Enable OUT4
strength
Channel 2 (M2) division ratio[7:0]
Reserved
Channel 2 PD
Select RF
Channel 2 (M2) division ratio[9:8]
Divider 2
Reserved
Channel 2 divider phase[5:0]
Enable 3.3 V
OUT5 format[2:0]
OUT5 polarity[1:0]
OUT5 drive
Enable OUT5
CMOS driver
strength
Channel 3, Divider 1 (M3) division ratio[7:0]
Reserved
Channel 3, Divider 1 (M3)
division ratio[9:8]
Channel 3, Divider 2 (M3b) division ratio[7:0]
Reserved
Enable
Channnel 3
Select RF
Channel 3, Divider 2 (M3b)
Channel 3
PD
Divider 2
division ratio[9:8]
doubler
Reserved
Channel 3, Divider 1 phase[5:0]
Reserved
Channel 3, Divider 2 phase[5:0]
Reserved
0x0506
0x0509
0x050A
RF Divider 2[3:0]
Reserved
D1
Output PLL (APLL) charge pump[7:0]
Reserved (default: 0x2)
RF divider
Output Clock Distribution
0x0500
Distribution
output sync
D3
D2
NPM Alpha-0[7:0]
NPM Alpha-0[15:8]
NPM Alpha-1[6:0]
NPM Beta-0[7:0]
NPM Beta-0[15:8]
NPM Beta-1[6:0]
NPM Gamma-0[7:0]
NPM Gamma-0[15:8]
NPM Gamma-1[6:0]
NPM Delta-0[7:0]
NPM Delta-0[15:8]
NPM Delta-1[6:0]
Rev. B | Page 64 of 104
20
00
44
02
02
10
00
00
00
10
10
03
00
00
10
10
00
00
00
10
03
00
00
00
00
00
Data Sheet
Reg
Addr
(Hex)
Opt
Name
Reference Inputs
0x0600
Reference
power-down
0x0601
Reference
logic type
0x0602
Reference
priority
0x0603
Frame Synchronization Mode
0x0640
En frame sync
0x0641
Frame sync
options
Profile A (for REFA)
0x0700
L
Reference
period (up to
0x0701
L
1.1 ms)
0x0702
L
0x0703
L
0x0704
L
0x0705
L
Frequency
tolerance
0x0706
L
0x0707
0x0708
L
L
0x0709
L
0x070A
0x070B
0x070C
0x070D
0x070E
L
L
L
L
L
0x070F
0x0710
0x0711
L
L
L
DPLL loop BW
0x0712
0x0713
0x0714
L
L
L
DPLL
R divider
(20 bits)
0x071D
0x071E
0x071F
0x0720
0x0721
0x0722
0x0723
0x0724
0x0725
0x0726
L
L
L
L
L
L
L
L
L
D4
D3
D2
Reserved
D1
D0
Reference power-down[3:0]
REFD logic type[1:0]
REFC logic type[1:0]
REFD priority[1:0]
REFC priority[1:0]
REFB logic type[1:0]
Fsync one
shot
00
00
REFA priority[1:0]
00
00
Reserved
Validate
Fsync ref
Fsync
arm
method
Enable Fsync
Arm soft Fsync
Nominal reference period (fs), Bits[7:0] (default: 51.44 ns =1/(19.44 MHz) for default system clock setting)
Nominal period (fs), Bits[15:8]
Nominal period (fs), Bits[23:16]
Nominal period (fs), Bits[31:24]
Nominal period (fs), Bits[39:32]
00
Inner tolerance, Bits[19:16]
Outer tolerance, Bits[19:16]
Validation timer (ms), Bits[7:0] (up to 65.5 seconds)
Validation timer (ms), Bits[15:8] (up to 65.5 seconds)
Reserved
Reserved
Digital PLL loop BW scaling factor[7:0] (default: 0x01F4 = 50 Hz)
Digital PLL loop BW scaling factor[15:8]
Reserved
Reserved
R divider[7:0]
R divider[15:8]
Enable REFA
R divider[19:16]
divide-by-2
Digital PLL feedback divider—Integer Part N1[7:0]
Digital PLL feedback divider—Integer Part N1[15:8]
Reserved
00
00
C9
EA
10
03
00
14
Inner tolerance (1 ppm), Bits[7:0] (for reference invalid to valid; 50% down to 1 ppm) (default: 5%)
Inner tolerance (1 ppm), Bits[15:8] (for reference invalid to valid; 50% down to 1 ppm)
Select base
loop filter
Def
REFA logic type[1:0]
REFB priority[1:0]
Reserved
Reserved
Validation
DPLL
fractional
feedback
divider
(24 bits)
DPLL
fractional
feedback
divider
modulus
(24 bits)
Lock detectors
0x071C
D5
Reserved
0x0718
0x071B
D6
00
0A
Outer tolerance (1 ppm), Bits[7:0] (for reference valid to invalid; 50% down to 1 ppm) (default: 10%)
Outer tolerance (1 ppm), Bits[15:8] (for reference valid to invalid; 50% down to 1 ppm)
DPLL
N divider
(17 bits)
0x071A
D7
Reserved
0x0715
0x0716
0x0717
0x0719
AD9558
00
Sel high PM
base loop
filter
BW scaling
factor[16]
00
0A
00
00
00
F4
01
00
C5
00
00
Digital PLL
feedback
divider—
Integer Part
N1[16]
6B
07
00
Digital PLL fractional feedback divider—FRAC1[7:0]
04
Digital PLL fractional feedback divider—FRAC1[15:8]
00
Digital PLL fractional feedback divider—FRAC1[23:16]
00
Digital PLL feedback divider modulus—MOD1[7:0]
05
Digital PLL feedback divider modulus—MOD1[15:8]
00
Digital PLL feedback divider modulus—MOD1[23:16]
00
Phase lock threshold (ps), Bits[7:0]
Phase lock threshold (ps), Bits[15:8]
Phase lock fill rate[7:0]
Phase lock drain rate[7:0]
Frequency lock threshold[7:0]
Frequency lock threshold[15:8]
Frequency lock threshold[23:16]
Frequency lock fill rate[7:0]
Frequency lock drain rate[7:0]
BC
02
0A
0A
BC
02
00
0A
0A
Rev. B | Page 65 of 104
AD9558
Data Sheet
Reg
Addr
(Hex)
Opt
Name
Profile B (for REFB)
0x0740
L
Reference
period (up to
0x0741
L
1.1 ms)
0x0742
L
0x0743
L
0x0744
L
0x0745
L
Frequency
tolerance
0x0746
L
0x0747
L
0x0748
L
0x0749
L
0x074A
L
0x074B
L
0x074C
L
0x074D
0x074E
L
L
0x074F
L
0x0750
0x0751
L
L
0x0752
0x0753
0x0754
L
L
L
Validation
0x075D
0x075E
0x075F
0x0760
0x0761
0x0762
0x0763
0x0764
0x0765
0x0766
L
L
L
L
L
L
L
L
L
D3
D2
D1
D0
Inner tolerance (1 ppm), Bits[7:0] (for reference invalid to valid; 50% down to 1 ppm) (default: 5%)
Inner tolerance (1 ppm), Bits[15:8] (for reference invalid to valid; 50% down to 1 ppm)
Reserved
Inner tolerance, Bits[19:16]
Outer tolerance (1 ppm), Bits[7:0] (for reference valid to invalid; 50% down to 1 ppm] (default: 10%)
Outer tolerance (1 ppm), Bits[15:8] (for reference valid to invalid; 50% down to 1 ppm)
Reserved
Outer tolerance, Bits[19:16]
Validation timer (ms), Bits[7:0] (up to 65.5 seconds)
Validation timer (ms), Bits[15:8] (up to 65.5 seconds)
Reserved
DPLL
R divider
(20 bits)
Sel high PM
base loop
filter
Digital PLL loop BW scaling factor[7:0] (default: 0x01F4 = 50 Hz)
Digital PLL loop BW scaling factor[15:8]
Reserved
DPLL
fractional
feedback
divider
(24 bits)
DPLL
fractional
feedback
divider
modulus
(24 bits)
Lock detectors
0x075C
D4
Nominal period (fs), Bits[7:0] (default: 125 µs = 1/(8 kHz) for default system clock setting)
Nominal period (fs), Bits[15:8]
Nominal period (fs), Bits[23:16]
Nominal period (fs), Bits[31:24]
Nominal period (fs), Bits[39:32]
DPLL loop BW
0x0758
0x075B
D5
Select base
loop filter
DPLL
N divider
(17 bits)
0x075A
D6
Reserved
0x0755
0x0756
0x0757
0x0759
D7
Reserved
R divider[7:0]
R divider[15:8]
Enable REFB
R divider[19:16]
divide-by-2
Digital PLL feedback divider—Integer Part N1[7:0]
Digital PLL feedback divider—Integer Part N1[15:8]
Reserved
Def
00
A2
94
1A
1D
14
00
00
0A
00
00
0A
00
00
00
F4
BW scaling
factor[16]
01
00
00
00
00
Digital PLL
feedback
divider—
Integer Part
N1[16]
1F
5B
00
Digital PLL fractional feedback divider—FRAC1[7:0]
00
Digital PLL fractional feedback divider—FRAC1[15:8]
00
Digital PLL fractional feedback divider—FRAC1[23:16]
00
Digital PLL feedback divider modulus—MOD1[7:0]
01
Digital PLL feedback divider modulus—MOD1[15:8]
00
Digital PLL feedback divider modulus—MOD1[23:16]
00
Phase lock threshold (ps), Bits[7:0]
Phase lock threshold(ps), Bits[15:8]
Phase lock fill rate[7:0]
Phase lock drain rate[7:0]
Frequency lock threshold[7:0]
Frequency lock threshold[15:8]
Frequency lock threshold[23:16]
Frequency lock fill rate[7:0]
Frequency lock drain rate[7:0]
BC
02
0A
0A
BC
02
00
0A
0A
Rev. B | Page 66 of 104
Data Sheet
Reg
Addr
(Hex)
Opt
Name
Profile C (for REFC)
0x0780
L
Reference
period (up to
0x0781
L
1.1 ms)
0x0782
L
0x0783
L
0x0784
L
0x0785
L
0x0786
L
0x0787
L
0x0788
L
0x0789
L
0x078A
L
0x078B
L
0x078C
L
0x078D
0x078E
L
L
0x078F
0x0790
0x0791
L
L
L
DPLL loop BW
0x0792
0x0793
0x0794
L
L
L
DPLL
R divider
(20 bits)
Frequency
tolerance
Validation
DPLL
N divider
(17 bits)
0x0798
0x0799
0x079A
DPLL
fractional
feedback
divider
(24 bits)
DPLL
fractional
feedback
divider
modulus
(24 bits)
Lock detectors
0x079E
0x079F
0x07A0
0x07A1
0x07A2
0x07A3
0x07A4
0x07A5
0x07A6
L
L
L
L
L
L
L
L
L
D7
D6
D5
D4
D3
D2
D1
D0
Def
Nominal period (fs), Bits[7:0] (default: 125 µs = 1/(8 kHz) for default system clock setting)
Nominal period (fs), Bits[15:8]
Nominal period (fs), Bits[23:16]
Nominal period (fs), Bits[31:24]
Nominal period (fs), Bits[39:32]
Inner tolerance (1 ppm), Bits[7:0] (for reference invalid to valid; 50% down to 1 ppm) (default: 5%)
Inner tolerance (1 ppm), Bits[15:8] (for reference invalid to valid; 50% down to 1 ppm)
Reserved
Inner tolerance, Bits[19:16]
Outer tolerance (1 ppm), Bits[7:0] (for reference valid to invalid; 50% down to 1 ppm] (default: 10%)
Outer tolerance (1 ppm), Bits[15:8] (for reference valid to invalid; 50% down to 1 ppm)
Reserved
Outer tolerance, Bits[19:16]
Validation timer (ms), Bits[7:0] (up to 65.5 seconds)
Validation timer (ms), Bits[15:8] (up to 65.5 seconds)
C9
Reserved
00
00
Select base
loop filter
0x0795
0x0796
0x0797
0x079B
0x079C
0x079D
AD9558
Reserved
Digital PLL loop BW scaling factor[7:0] (default: 0x01F4 = 50 Hz)
Digital PLL loop BW scaling factor[15:8]
Reserved
Reserved
R divider[7:0]
R divider[15:8]
Enable REFC
R divider[19:16]
divide-by-2
Digital PLL feedback divider—Integer Part N1[7:0]
Digital PLL feedback divider—Integer Part N1[15:8]
Reserved
Sel high PM
base loop filt
BW scaling
factor[16]
EA
10
03
00
14
00
00
0A
00
00
0A
00
F4
01
00
C5
00
00
Digital PLL
feedback
divider—
Integer Part
N1[16]
6B
07
00
Digital PLL fractional feedback divider—FRAC1[7:0]
Digital PLL fractional feedback divider—FRAC1[15:8]
Digital PLL fractional feedback divider—FRAC1[23:16]
04
00
00
Digital PLL feedback divider modulus—MOD1[7:0]
Digital PLL feedback divider modulus—MOD1[15:8]
Digital PLL feedback divider modulus—MOD1[23:16]
05
00
00
Phase lock threshold (ps), Bits[7:0]
Phase lock threshold (ps), Bits[15:8]
Phase lock fill rate[7:0]
Phase lock drain rate[7:0]
Frequency lock threshold[7:0]
Frequency lock threshold[15:8]
Frequency lock threshold[23:16]
Frequency lock fill rate[7:0]
Frequency lock drain rate[7:0]
BC
02
0A
0A
BC
02
00
0A
0A
Rev. B | Page 67 of 104
AD9558
Data Sheet
Reg
Addr
(Hex)
Opt
Name
DPLL Profile D (for REFD)
0x07C0
L
Reference
period (up to
0x07C1
L
1.1 ms)
0x07C2
L
0x07C3
L
0x07C4
L
0x07C5
L
0x07C6
L
0x07C7
L
0x07C8
L
0x07C9
L
0x07CA
L
0x07CB
L
0x07CC
L
0x07CD
0x07CE
L
L
0x07CF
0x07D0
0x07D1
L
L
L
DPLL loop BW
0x07D2
0x07D3
0x07D4
L
L
L
DPLL
R divider
(20 bits)
Tolerance
Validation
DPLL
N divider
(17 bits)
0x07D8
0x07D9
0x07DA
DPLL
fractional
feedback
divider
(24 bits)
DPLL
fractional
feedback
divider
modulus
(24 bits)
Lock detectors
0x07DE
0x07DF
0x07E0
0x07E1
0x07E2
0x07E3
0x07E4
0x07E5
0x07E6
L
L
L
L
L
L
L
L
L
D6
D5
D4
D3
D2
D1
D0
Def
Nominal reference period (fs), Bits[7:0] (default: 51.44 ns =1/(19.44 MHz) for default system clock setting)
Nominal period (fs), Bits[15:8]
Nominal period (fs), Bits[23:16]
Nominal period (fs), Bits[31:24]
Nominal period (fs), Bits[39:32]
Inner tolerance (1 ppm), Bits[7:0] (for reference invalid to valid; 50% down to 1 ppm) (default: 5%)
Inner tolerance (1 ppm), Bits[15:8] (for reference invalid to valid; 50% down to 1 ppm)
Reserved
Inner tolerance, Bits[19:16]
Outer tolerance (1 ppm), Bits[7:0] (for reference valid to invalid; 50% down to 1 ppm) (default: 10%)
Outer tolerance (1 ppm), Bits[15:8] (for reference valid to invalid; 50% down to 1 ppm)
Reserved
Outer tolerance, Bits[19:16]
Validation timer (ms), Bits[7:0] (up to 65.5 seconds)
Validation timer (ms), Bits[15:8] (up to 65.5 seconds]
00
Reserved
00
00
Select base
loop filter
0x07D5
0x07D6
0x07D7
0x07DB
0x07DC
0x07DD
D7
Reserved
Digital PLL loop BW scaling factor[7:0] (default: 0x01F4 = 50 Hz)
Digital PLL loop bandwidth BW scaling factor[15:8]
Reserved
Sel high PM
base loop
filter
BW scaling
factor[16]
R divider[7:0]
R divider[15:8]
Reserved
Enable
R divider[19:16]
REFD
divide-by-2
Digital PLL feedback divider—Integer Part N1[7:0]
Digital PLL feedback divider—Integer Part N1[15:8]
Reserved
A2
94
1A
1D
14
00
00
0A
00
00
0A
00
F4
01
00
00
00
00
Digital PLL
feedback
divider—
Integer Part
N1[16]
1F
5B
00
Digital PLL fractional feedback divider—FRAC1[7:0]
Digital PLL fractional feedback divider—FRAC1[15:8]
Digital PLL fractional feedback divider—FRAC1[23:16]
00
00
00
Digital PLL feedback divider modulus—MOD1[7:0]
Digital PLL feedback divider modulus—MOD1[15:8]
Digital PLL feedback divider modulus—MOD1[23:16]
01
00
00
Phase lock threshold (ps), Bits[7:0]
Phase lock threshold(ps), Bits[15:8]
Phase lock fill rate[7:0]
Phase lock drain rate[7:0]
Frequency lock threshold[7:0]
Frequency lock threshold[15:8]
Frequency lock threshold[23:16]
Frequency lock fill rate[7:0]
Frequency lock drain rate[7:0]
BC
02
0A
0A
BC
02
00
0A
0A
Rev. B | Page 68 of 104
Data Sheet
AD9558
Reg
Addr
(Hex)
D7
D6
D5
D4
D3
D2
D1
D0
Def
DCO PD
SYSCLK
PD
Ref input
PD
TDC PD
APLL PD
Clock dist PD
Full PD
00
0x0A01
Loop mode
Soft reset
exclude
regmap
Reserved
User
holdover
User
freerun
User ref in manual
switchover mode
00
0x0A02
Cal/sync
Opt
Name
Operational Controls
0x0A00
Power-down
Reserved
0x0A03
A
0x0A04
A
0x0A05
A
0x0A06
A
0x0A07
A
0x0A08
A
Reserved
0x0A09
A
Reserved
0x0A0A
A
Increment
phase offset
0x0A0B
A
Manual
reference
validation
Manual
reference
invalidation
Static
reference
validation
0x0A0C
0x0A0D
Clear/reset
functions
IRQ clearing
Reserved
Clear LF
Clear CCI
Reserved
Reserved
Reserved
SYSCLK
unlocked
SYSCLK
locked
Pin
program
end
Holdover
Reserved
Switching
L, E
0x0C04
L, E
0x0C05
L, E
0x0C06
L, E
0x0C07
L, A,
E
L, E
0x0C08
Closed
Freerun
Reserved
Quick In-Out Frequency Soft Pin Configuration
0x0C00
L, E
Enable
Soft Pin
Section 1
0x0C01
L, E
Soft Pin
Section 1
0x0C02
L, E
0x0C03
Enable
Soft Pin
Section 2
Soft Pin
Section 2
REF switchover mode[2:0]
REFB
validated
REFD
validated
REFB fault
cleared
REFD
fault
cleared
Reserved
Soft sync
clock dist
Clear all IRQs
Reserved
00
Clear watchdog
00
APLL cal
started
EEPROM
complete
00
Clear auto
sync
APLL
unlocked
Sync clock
dist
Clear TW
history
APLL
locked
Watchdog
timer
Reserved
Phase
unlocked
Phase slew
unlimited
REFA fault
cleared
REFC fault
cleared
Phase
locked
Phase slew
limited
REFA fault
REFD fault
Frequency
locked
Frequency
clamped
REFA
validated
REFC
validated
00
History
updated
REFB fault
Frequency
unlocked
Frequency
unclamped
Reserved
REFC fault
00
Decrement
phase offset
Increment
phase offset
00
Force
Timeout B
Force
Timeout A
00
APLL cal ended
EEPROM fault
00
00
00
Reserved
Force
Timeout D
Reset
phase
offset
Force
Timeout C
Reserved
REF Mon
Override D
REF Mon
Override C
REF Mon
Override B
REF Mon
Override A
00
Reserved
REF Mon
Bypass D
REF Mon
Bypass C
REF Mon
Bypass B
REF Mon
Bypass A
00
En Soft Pin
Section 1
00
Reserved
Output frequency selection[3:0]
Reserved
Reserved
REFD frequency scale[1:0]
Channel 3 output frequency
scale[1:0]
Reserved
REFC frequency
scale[1:0]
Channel 2 output
frequency scale[1:0]
Sel high
PM base
loop filter
Reserved
Reserved
Rev. B | Page 69 of 104
Input frequency selection[3:0]
SYSCLK PLL REF selection[1:0]
En Soft Pin
Section 2
00
00
00
REFB frequency scale[1:0]
REFA frequency scale[1:0]
00
Channel 1 output frequency
scale[1:0]
DPLL BW[1:0]
Channel 0 output frequency
scale[1:0]
REF tolerance[1:0]
00
Soft pin start
transfer
Soft pin reset
00
00
00
AD9558
Data Sheet
Reg
Addr
(Hex)
Opt
Name
D7
D6
D5
Read-Only Status (Accessible During EEPROM Transactions)
0x0D00
R, L
EEPROM
Reserved
D4
D3
D2
D1
D0
Def
Fault
detected
Load in
progress
Save in
progress
N/A
0x0D01
R, L
SYSCLK and
PLL status
All PLLs
locked
APLL VCO
status
Pin program
ROM load
process
APLL cal
in progress
APLL lock
SYSCLK stable
SYSCLK
lock detect
N/A
0x0D02
R, L
IRQ monitor
events
SYSCLK
unlocked
SYSCLK
locked
APLL
unlocked
APLL
lock detect
APLL cal ended
APLL cal
started
N/A
0x0D03
R, L
Output dist
sync
Watchdog
timer
EEPROM
fault
EEPROM
complete
N/A
0x0D04
R, L
Freerun
Pin
program
end
Holdover
R, L
0x0D06
R, L
Reserved
R, L
Reserved
Phase
unlocked
Phase slew
unlimited
REFA fault
cleared
REFC
fault cleared
Phase
locked
Phase slew
limited
REFA fault
0x0D07
Frequency
locked
Frequency
clamped
REFA
validated
REFC
validated
N/A
0x0D05
Frequency
unlocked
Frequency
unclamped
Reserved
REFC
fault
N/A
0x0D08
R
Holdover
Active
Freerun
N/A
0x0D09
R
0x0D0A
0x0D0B
R
R
REFA/REFB
B valid
B fault
B fast
B slow
Reserved
A valid
A fault
A fast
A slow
N/A
N/A
0x0D0C
R
REFC/REFD
D valid
D fault
D fast
D slow
C valid
C fault
C fast
C slow
N/A
Reserved
DPLL_
APLL_
Lock
Reserved
Reserved
Switching
Closed
Reserved
DPLL
Reserved
0x0D0D
R
Holdover
history
0x0D0E
R
0x0D0F
R
0x0D10
R
0x0D11
R
Lock detector
phase tub
0x0D12
R
0x0D13
R
Lock detector
frequency tub
0x0D14
R
Nonvolatile Memory (EEPROM) Control
0x0E00
E
Write protect
0x0E01
0x0E02
E
A,E
Condition
Save
0x0E03
A,E
Load
REFB
validated
REFD
validated
Offset slew
limiting
Reserved
REFB fault
cleared
REFD
fault
cleared
Frequency
lock
Frequency
clamped
History
updated
REFB fault
REFD
fault
Reserved
Phase lock
Loop
switching
Reserved
History
available
Active
reference
priority
Current active reference
Tuning word readback[31:0]
Phase tub[11:8]
Frequency tub[7:0]
Reserved
Frequency tub[11:8]
Reserved
Reserved
Write enable
Conditional value[3:0]
Reserved
Reserved
Rev. B | Page 70 of 104
Save to
EEPROM
Load from
EEPROM
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
Phase tub[7:0]
Reserved
N/A
00
00
00
00
Data Sheet
Reg
Addr
(Hex)
Opt
Name
EEPROM Storage Sequence
0x0E10
E
EEPROM ID
0x0E11
E
0x0E12
E
0x0E13
E
System
clock
0x0E14
E
0x0E15
E
0x0E16
E
I/O update
0x0E17
E
General
0x0E18
E
0x0E19
E
0x0E1A
E
DPLL
0x0E1B
E
0x0E1C
E
0x0E1D
E
APLL
0x0E1E
E
0x0E1F
E
0x0E20
E
Clock dist
0x0E21
E
0x0E22
E
0x0E23
E
I/O update
0x0E24
E
Reference
inputs
0x0E25
E
0x0E26
E
0x0E27
E
0x0E28
E
0x0E29
E
0x0E2A
E
Profile REFA
0x0E2B
E
0x0E2C
E
0x0E2D
E
Profile REFB
0x0E2E
E
0x0E2F
E
0x0E30
E
Profile REFC
0x0E31
E
0x0E32
E
0x0E33
E
Profile REFD
0x0E34
E
0x0E35
E
0x0E36
E
I/O update
0x0E37
E
Operational
controls
0x0E38
E
0x0E39
E
0x0E3A
E
Calibrate APLL
0x0E3B
E
I/O update
0x0E3C
E
End of data
0x0E3D
E
Unused
to
0xE45
AD9558
D7
D6
D5
D4
D3
D2
Data: two bytes
Address: 0x0006
Data: nine bytes
Address: 0x0100
Action: I/O update
Data: 18 bytes
Address: 0x0200
Data: 47 bytes
Address: 0x0300
Data: nine bytes
Address: 0x0400
Data: 22 bytes
Address: 0x0500
Action: I/O update
Data: four bytes
Address: 0x0600
Data: two bytes
Address: 0x0640
Data: 39 bytes
Address: 0x0700
Data: 39 bytes
Address: 0x0740
Data: 39 bytes
Address: 0x0780
Data: 39 bytes
Address: 0x07C0
Action: I/O update
Data: 14 bytes
Address: 0x0A00
Action: calibrate output PLL
Action: I/O update
Action: End of data
Unused
(available for additional EEPROM instructions)
Rev. B | Page 71 of 104
D1
D0
Def
01
00
06
08
01
00
80
11
02
00
2E
03
00
08
04
00
15
05
00
80
03
06
00
01
06
40
26
07
00
26
07
40
26
07
80
26
07
C0
80
0D
0A
00
A0
80
FF
00
AD9558
Data Sheet
REGISTER MAP BIT DESCRIPTIONS
SERIAL PORT CONFIGURATION (REGISTER 0x0000 TO REGISTER 0x0005)
Table 36. Serial Configuration (Note that the contents of Register 0x0000 are not stored to the EEPROM.)
Address
0x0000
Bits
7
Bit Name
SDO enable
6
LSB first/increment
address
5
Soft reset
[4:0]
Reserved
Description
Enables SPI port SDO pin.
1 = 4-wire (SDO pin enabled).
0 (default) = 3-wire.
Bit order for SPI port.
1 = least significant bit and byte first.
Register addresses are automatically incremented in multibyte transfers.
0 (default) = most significant bit and byte first.
Register addresses are automatically decremented in multibyte transfers.
Device reset (invokes an EEPROM download or pin program ROM download if EEPROM or
pin program is enabled. See the EEPROM and Pin Configuration and Function Descriptions
sections for details.
Reserved.
Table 37. Readback Control
Address
0x0004
Bits
[7:1]
Bit Name
Reserved
Description
Reserved.
0
Read buffer register
For buffered registers, serial port readback reads from actual (active) registers instead of the
buffer.
1 = reads buffered values that take effect on next assertion of I/O update.
0 (default) = reads values currently applied to the device’s internal logic.
Table 38. Soft I/O Update
Address
0x0005
Bits
[7:1]
0
Bit Name
Reserved
I/O update
Description
Reserved.
Writing a 1 to this bit transfers the data in the serial I/O buffer registers to the device’s
internal control registers. Unless a register is marked as live (as indicated by an L in the Opt
column of the register map), the user must write to this bit before any register settings can
take effect and before a read-only register can be updated with the most current value.
This is an autoclearing bit.
Table 39. User Scratch Pad
Address
0x0006
0x0007
Bits
[7:0]
[7:0]
Bit Name
User scratch pad[7:0]
User scratch pad[15:8]
Description
User programmable EEPROM ID registers. These registers enable users to write a unique
code of their choosing to keep track of revisions to the EEPROM register loading. It has no
effect on part operation.
0 = default.
SILICON REVISION (REGISTER 0x000A)
Table 40. Silicon Revision
Address
0x000A
Bits
[7:0]
Bit Name
Silicon revision
Description
This read-only register identifies the revision level of the AD9558.
CLOCK PART SERIAL ID (REGISTER 0x000C TO REGISTER 0x000D)
Table 41. Clock Part Family ID
Address
0x000C
Bits
[7:0]
Bit Name
Clock part family ID[7:0]
0x000D
[7:0]
Clock part family ID[15:8]
Description
This read-only register (along with Register 0x000D) uniquely identifies an AD9557 or
AD9558. No other part in the ADI AD95xx family has a value of 0x0001 in these two registers.
Default: 0x01 for the AD9557 and AD9558.
This register is a continuation of Register 0x000C.
Default: 0x00 for the AD9557 and AD9558.
Rev. B | Page 72 of 104
Data Sheet
AD9558
SYSTEM CLOCK (REGISTER 0x0100 TO REGISTER 0x0108)
Table 42. System Clock PLL Feedback Divider (N3 Divider)
Address
0x0100
Bits
[7:0]
Bit Name
SYSCLK N3 divider
Description
System clock PLL feedback divider value: 4 ≤ N3 ≤ 255 (default: 0x08).
Table 43. SYSCLK Configuration
Address
0x0101
Bits
[7:5]
4
Bit Name
Reserved
Load from ROM
(read-only)
3
SYSCLK XTAL enable
[2:1]
SYSCLK P divider
0
SYSCLK doubler enable
Description
Reserved.
This read-only bit is set if the PINCONTROL pin was high during the last power-on.
0 = The PINCONTROL pin was low at power-on (or reset).
1 = The PINCONTROL pin was high at power-on (or reset).
Enables the crystal maintaining amplifier for the system clock input.
1 (default) = crystal mode (crystal maintaining amplifier enabled).
0 = external XO or other system clock source.
System clock input divider.
00 (default) = 1.
01 = 2.
10 = 4.
11 = 8.
Enable clock doubler on system clock input to reduce noise.
0 = disable.
1 (default) = enable.
Table 44. Nominal System Clock Period1
Address
0x0103
Bits
[7:0]
0x0104
[7:0]
0x0105
[7:5]
[4:0]
1
Bit Name
Nominal system clock period (fs)
Reserved
Nominal system clock period (fs)
Description
System clock period, Bits[7:0].
Default: 0x0E.
System clock period, Bits[15:8].
Default: 0x67.
Reserved.
System clock period, Bits[20:16].
Default: 0x13.
Note that the default setting for system clock period is 1.271566 ns, which is the period of 786.432 MHz (= 49.152 MHz × 16).
Table 45. System Clock Stability Period
Address
0x0106
Bits
[7:0]
0x0107
[7:0]
0x0108
[7:5]
4
[3:0]
Bit Name
System clock stability period (ms)
Reserved
Reset SYSCLK stability timer
System clock stability period
Description
System clock period, Bits[7:0].
Default: 0x32 (0x000032 = 50 ms).
System clock period, Bits[15:8].
Default: 0x00.
Reserved.
This autoclearing bit resets the system clock stability timer.
System clock period, Bits[19:16].
Default: 0x00.
Rev. B | Page 73 of 104
AD9558
Data Sheet
GENERAL CONFIGURATION (REGISTER 0x0200 TO REGISTER 0x0214)
Multifunction Pin Control (M0 to M7) and IRQ Pin Control (Register 0x0200 to Register 0x0209)
Note that the default setting for the M0 to M5 multifunction pins and the IRQ is that of a 3-level logic input; M6 and M7 are unused inputs at
startup. After startup, M0 to M7 are 2-level inputs or 2-level outputs, based on the settings of Register 0x0200 to Register 0x0208. Setting
Bit 0 in Register 0x0200 to 1 enables M0 to M7 pin functionality.
Table 46. Multifunction Pin (M0 to M7) Control
Address
0x0200
Bits
[7:1]
0
Bit Name
Reserved
Enable M pins and IRQ pin
function
0x0201
7
M0 output/input
0x0202
[6:0]
7
M0 function
M1 output/input
0x0203
[6:0]
7
M1 function
M2 output/input
0x0204
[6:0]
7
M2 function
M3 output/input
0x0205
[6:0]
7
M3 function
M4 output/input
0x0206
[6:0]
7
M4 function
M5 output/input
0x0207
[6:0]
7
M5 function
M6 output/input
0x0208
[6:0]
7
M6 function
M7 output/input
[6:0]
M7 function
A
A
E
E
See Table 129 and Table 130. Default: 0xB1 = REFB valid.
In/out control for M2 pin (same as M0).
A
E
See Table 129 and Table 130. Default: 0xC0 = REFA active.
In/out control for M3 pin (same as M0).
A
E
See Table 129 and Table 130. Default: 0xC1 = REFB active.
In/out control for M3 pin (same as M0).
A
E
See Table 129 and Table 130. Default: 0xB2 = REFC valid.
In/out control for M3 pin (same as M0).
A
E
See Table 129 and Table 130. Default: 0xB3 = REFD valid.
In/out control for M3 pin (same as M0).
A
A
Description
Reserved.
0 (default) = disables the function of the M pins and IRQ pin control register (Address
0x0201 to Address 0x0209) and the M pins and IRQ pin are in 3-level logic input state.
1 = the M pins and IRQ pin are out of 3-level logic input state. Enables the function of the
M pins and IRQ pin control register (Address 0x0201 to Address 0x0209).
In/out control for M0 pin.
0 = input (2-level logic control pin).
1 (default) = output (2-level logic status pin).
See Table 129 and Table 130. Default: 0xB0 = REFA valid.
In/out control for M1 pin (same as M0).
E
E
See Table 129 and Table 130. Default: 0xC2 = REFC active.
In/out control for M3 pin (same as M0).
See Table 129 and Table 130. Default: 0xC3 = REFD active.
Table 47. IRQ Pin Output Mode
Address
0x0209
Bits
[7:5]
[4:3]
Bit Name
Reserved
Status signal at IRQ pin
2
Use IRQ pin for status signal
[1:0]
IRQ pin driver type
Description
Default: 000b
This selection is valid only when Address 0x0209[2] = 1
00 = DPLL phase locked
01 = DPLL frequency locked
10 = system clock PLL locked
11 (default) = (DPLL phase locked) AND (system clock PLL locked) AND (APLL locked)
0 = uses IRQ pin to monitor IRQ event
1 (default) = uses IRQ pin to monitor internal status signals
Select the output mode of the IRQ pin
00 = NMOS, open drain (requires an external pull-up resistor)
01 = PMOS, open drain (requires an external pull-down resistor)
10 = CMOS, active high
11 (default) = CMOS, active low
Rev. B | Page 74 of 104
Data Sheet
AD9558
IRQ MASK (REGISTER 0x020A TO REGISTER 0x020F)
The IRQ mask register bits form a one-to-one correspondence with the bits of the IRQ monitor register (0x0D02 to 0x0D09). When set to
Logic 1, the IRQ mask bits enable the corresponding IRQ monitor bits to indicate an IRQ event. The default for all IRQ mask bits is Logic 0,
which prevents the IRQ monitor from detecting any internal interrupts.
Table 48. IRQ Mask for SYSCLK
Address
0x020A
Bits
[7:6]
5
4
3
2
1
0
Bit Name
Reserved
SYSCLK unlocked
SYSCLK locked
APLL unlocked
APLL locked
APLL cal complete
APLL cal started
Description
Reserved
Enables IRQ for indicating a SYSCLK PLL state transition from locked to unlocked
Enables IRQ for indicating a SYSCLK PLL state transition from unlocked to locked
Enables IRQ for indicating an APLL state transition from locked to unlocked
Enables IRQ for indicating an APLL state transition from unlocked to locked
Enables IRQ for indicating that APLL (LCVCO) calibration has completed
Enables IRQ for indicating that APLL (LCVCO) calibration has begun
Table 49. IRQ Mask for Distribution Sync, Watchdog Timer, and EEPROM
Address
0x020B
Bits
[7:5]
4
3
2
1
0
Bit Name
Reserved
Pin program end
Sync distribution
Watchdog timer
EEPROM fault
EEPROM complete
Description
Reserved
Enables IRQ for indicating successful completion of a pin program ROM load
Enables IRQ for indicating a distribution sync event
Enables IRQ for indicating expiration of the watchdog timer
Enables IRQ for indicating a fault during an EEPROM load or save operation
Enables IRQ for indicating successful completion of an EEPROM load or save operation
Table 50. IRQ Mask for the Digital PLL
Address
0x020C
Bits
7
6
5
4
3
2
1
0
Bit Name
Switching
Closed
Freerun
Holdover
Frequency unlocked
Frequency locked
Phase unlocked
Phase locked
Description
Enables IRQ for indicating that the DPLL is switching to a new reference
Enables IRQ for indicating that the DPLL has entered closed-loop operation
Enables IRQ for indicating that the DPLL has entered free run mode
Enables IRQ for indicating that the DPLL has entered holdover mode
Enables IRQ for indicating that the DPLL lost frequency lock
Enables IRQ for indicating that the DPLL has acquired frequency lock
Enables IRQ for indicating that the DPLL lost phase lock
Enables IRQ for indicating that the DPLL has acquired phase lock
Table 51. IRQ Mask for History Update, Frequency Limit and Phase Slew Limit
Address
0x020D
Bits
[7:5]
4
3
2
1
Bit Name
Reserved
History updated
Frequency unclamped
Frequency clamped
Phase slew unlimited
0
Phase slew limited
Description
Reserved
Enables IRQ for indicating the occurrence of a tuning word history update
Enables IRQ for indicating a frequency limit state transition from clamped to unclamped
Enables IRQ for indicating a state transition of the frequency limiter from unclamped to clamped
Enables IRQ for indicating a state transition of the phase slew limiter from slew limiting to not
slew limiting
Enables IRQ for indicating a state transition of the phase slew limiter from not slew limiting to
slew limiting
Rev. B | Page 75 of 104
AD9558
Data Sheet
Table 52. IRQ Mask for Reference Inputs
Address
0x020E
0x020F
Bits
7
6
5
4
3
2
1
0
[7:0]
Bit Name
Reserved
REFB validated
REFB fault cleared
REFB fault
Reserved
REFA validated
REFA fault cleared
REFA fault
Reserved
Description
Reserved
Enables IRQ for indicating that REFB has been validated
Enables IRQ for indicating that REFB has been cleared of a previous fault
Enables IRQ for indicating that REFB has been faulted
Reserved
Enables IRQ for indicating that REFA has been validated
Enables IRQ for indicating that REFA has been cleared of a previous fault
Enables IRQ for indicating that REFA has been faulted
Reserved
Table 53. Watchdog Timer 11
Address
0x0210
0x0211
1
Bits
[7:0]
[7:0]
Bit Name
Watchdog timer (ms)
Description
Watchdog timer, Bits[7:0]; default: 0x00
Watchdog timer, Bits[15:8]; default: 0x00
Note that the watchdog timer is expressed in units of milliseconds (ms). The default value is 0 (disabled).
DPLL CONFIGURATION (REGISTER 0x0300 TO REGISTER 0x032E)
Table 54. Free Run Frequency Tuning Word1
Address
0x0300
0x0301
0x0302
0x0303
1
Bits
[7:0]
[7:0]
[7:0]
[7:6]
[5:0]
Bit Name
30-bit free run frequency tuning word
Reserved
30-bit free run frequency word
Description
Free run frequency tuning word, Bits[7:0]; default: 0x11
Free run frequency tuning word, Bits[15:8]; default: 0x15
Free run frequency tuning word, Bits[23:9]; default: 0x64
Reserved
Free run frequency tuning word, Bits[29:24]; default: 0x1B
Note that the default free run tuning word is 0x1B641511, which is used for 8 kHz/19.44 MHz = 622.08 MHz translation.
Table 55. Digital Oscillator Control
Address
0x0304
Bits
[7:6]
5
Bit Name
Reserved
DCO 4-level output
4
[3:0]
Reserved
Reserved
Description
Default: 00b
0 (default) = DCO 3-level output mode
1 = enables DCO 4-level output mode
Reserved (must be set to 1b)
Reserved (default: 0x0)
Table 56. DPLL Frequency Clamp
Address
0x0306
Bits
[7:0]
Bit Name
Lower limit of pull-in range (expressed
as a 20-bit frequency tuning word)
0x0307
[7:0]
0x0308
[7:4]
[3:0]
Reserved
Lower limit of pull-in range
0x0309
[7:0]
0x030A
[7:0]
Upper limit of pull-in range
(expressed as a 20-bit frequency
tuning word)
0x030B
[7:4]
[3:0]
Reserved
Upper limit of pull-in range
Description
Lower limit pull-in range, Bits[7:0].
Default: 0x51.
Lower limit pull-in range, Bits[15:8].
Default: 0xB8.
Default: 0x0.
Lower limit pull-in range, Bits[19:16].
Default: 0x2.
Upper limit pull-in range, Bits[7:0].
Default: 0x3E.
Upper limit pull-in range, Bits[15:8].
Default: 0x0A.
Default: 0x0.
Upper limit pull-in range, Bits[19:16].
Default: 0xB.
Rev. B | Page 76 of 104
Data Sheet
AD9558
Table 57. Fixed Closed-Loop Phase Lock Offset
Address
0x030C
Bits
[7:0]
0x030D
[7:0]
0x030E
[7:0]
0x030F
[7:6]
[5:0]
Bit Name
Fixed phase lock offset (signed; ps)
Reserved
Fixed phase lock offset (signed; ps)
Description
Fixed phase lock offset, Bits[7:0].
Default: 0x00.
Fixed phase lock offset, Bits[15:8].
Default 0x00.
Fixed phase lock offset, Bits[23:16].
Default: 0x00.
Reserved; default: 0x0.
Fixed phase lock offset, Bits[29:24].
Default: 0x00.
Table 58. Incremental Closed-Loop Phase Lock Offset Step Size1
Address
0x0310
Bits
[7:0]
0x0311
[7:0]
1
Bit Name
Incremental phase lock offset
step size (ps)
Description
Incremental phase lock offset step size, Bits[7:0].
Default: 0x00.
This controls the static phase offset of the DPLL while it is locked.
Incremental phase lock offset step size, Bits[15:8].
Default: 0x00.
This controls the static phase offset of the DPLL while it is locked.
Note that the default incremental closed-loop phase lock offset step size value is 0x0000 = 0 (0 ns).
Table 59. Phase Slew Rate Limit
Address
0x0312
Bits
[7:0]
0x0313
[7:0]
Bit Name
Phase slew rate limit (µs/sec)
Description
Phase slew rate limit, Bits[7:0].
Default: 0x00.
This register controls the maximum allowable phase slewing during transients
and reference switching.
The default phase slew rate limit is 0, or disabled. Minimum useful value is 310 µs/sec.
Phase slew rate limit, Bits[15:8].
Default: 0x00.
Table 60. History Accumulation Timer
Address
0x0314
Bits
[7:0]
0x0315
[7:0]
Bit Name
History accumulation timer (ms)
Description
History accumulation timer bits[7:0].
Default: 0x0A. For Register 0x0314 and Register 0x0315, 0x000A = 10 ms.
Maximum is 65 sec. This register controls the amount of tuning word averaging used
to determine the tuning word used in holdover. Never program a timer value of
zero. The default value is 0x000A = 10 decimal, which equates to 10 ms
History accumulation timer bits[15:8].
Default: 0x00.
Rev. B | Page 77 of 104
AD9558
Data Sheet
Table 61. History Mode
Address
0x0316
Bits
[7:5]
4
Bit Name
Reserved
Single sample fallback
3
Persistent history
[2:0]
Incremental average
Description
Reserved.
Controls holdover history. If tuning word history is not available for the reference
that was active just prior to holdover, then:
0 (default) = uses the free run frequency tuning word register value.
1 = uses the last tuning word from the DPLL.
Controls holdover history initialization. When switching to a new reference:
0 (default) = clears the tuning word history.
1 = retain the previous tuning word history.
History mode value from 0 to 7 (default: 0).
When set to non-zero, causes the first history accumulation to update prior to the
first complete averaging period. After the first full interval, updates occur only at the
full period.
0 (default) = update only after the full interval has elapsed.
1 = update at 1/2 the full interval.
2 = update at 1/4 and 1/2 of the full interval.
3 = update at 1/8, 1/4, and 1/2 of the full interval.
...
7 = update at 1/256, 1/128, 1/64, 1/32, 1/16, 1/8, 1/4, and 1/2 of the full interval.
Table 62. Base Digital Loop Filter with High Phase Margin (PM = 88.5°, BW = 0.1 Hz, Third Pole Frequency = 10 Hz, N1 = 1)1
Address
0x0317
0x0318
0x0319
0x031A
0x031B
0x031C
0x031D
0x031E
0x031F
0x0320
0x0321
0x0322
1
Bits
[7:0]
[7:0]
7
[6:0]
[7:0]
[7:0]
7
[6:0]
[7:0]
[7:0]
7
[6:0]
[7:0]
[7:0]
7
[6:0]
Bit Name
HPM Alpha-0 linear
Reserved
HPM Alpha-1 exponent
HPM Beta-0 linear
Reserved
HPM Beta-1 exponent
HPM Gamma-0 linear
Reserved
HPM Gamma-1 exponent
HPM Delta-0 linear
Reserved
HPM Delta-1 exponent
Description
Alpha-0 coefficient linear bits[7:0]. Default: 0x8C
Alpha-0 coefficient linear bits[15:8]
Reserved
Alpha-1 coefficient exponent bits[6:0]
Beta-0 coefficient linear bits[7:0]
Beta-0 coefficient linear bits[15:8]
Reserved
Beta-1 coefficient exponent bits[6:0]
Gamma-0 coefficient linear bits[7:0]
Gamma-0 coefficient linear bits[15:8]
Reserved
Gamma-1 coefficient exponent bits[6:0]
Delta-0 coefficient linear bits[7:0]
Delta-0 coefficient linear bits[15:8]
Reserved
Delta-1 coefficient exponent bits[6:0]
Note that the base digital loop filter coefficients (α, β, γ, and δ) have the following general form: x(2y), where x is the linear component and y is the exponential
component of the coefficient. The value of the linear component (x) constitutes a fraction, where 0 ≤ x ≤ 1. The exponential component (y) is a signed integer.
Rev. B | Page 78 of 104
Data Sheet
AD9558
Table 63. Base Digital Loop Filter with Normal Phase Margin (PM = 70°, BW = 0.1 Hz, Third Pole Frequency = 2 Hz, N1 = 1)1
Address
0x0323
0x0324
0x0325
0x0326
0x0327
0x0328
0x0329
0x032A
0x032B
0x032C
0x032D
0x032E
1
Bits
[7:0]
[7:0]
7
[6:0]
[7:0]
[7:0]
7
[6:0]
[7:0]
[7:0]
7
[6:0]
[7:0]
[7:0]
7
[6:0]
Bit Name
NPM Alpha-0 linear
Reserved
NPM Alpha-1 exponent
NPM Beta-0 linear
Reserved
NPM Beta-1 exponent
NPM Gamma-0 linear
Reserved
NPM Gamma-1 exponent
NPM Delta-0 linear
Reserved
NPM Delta-1 exponent
Description
Alpha-0 coefficient linear, Bits[7:0]
Alpha-0 coefficient linear, Bits[15:8]
Reserved
Alpha-1 coefficient exponent, Bits[6:0]
Beta-0 coefficient linear, Bits[7:0]
Beta-0 coefficient linear, Bits[15:8]
Reserved
Beta-1 coefficient exponent, Bits[6:0]
Gamma-0 coefficient linear, Bits[7:0]
Gamma-0 coefficient linear, Bits[15:8]
Reserved
Gamma-1 coefficient exponent, Bits[6:0]
Delta-0 coefficient linear, Bits[7:0]
Delta-0 coefficient linear, Bits[15:8]
Reserved
Delta-1 coefficient exponent, Bits[6:0]
Note that the digital loop filter base coefficients (α, β, γ, and δ) have the following general form: x(2y), where x is the linear component and y is the exponential
component of the coefficient. The value of the linear component (x) constitutes a fraction, where 0 ≤ x ≤ 1. The exponential component (y) is a signed integer.
OUTPUT PLL CONFIGURATION (REGISTER 0x0400 TO REGISTER 0x0408)
Table 64. Output PLL Setting1
Address
0x0400
Bits
[7:0]
Bit Name
Output PLL (APLL)
charge pump current
0x0401
[7:0]
APLL N divider
0x0402
0x0403
[7:0]
[7:6]
Reserved
APLL loop filter control
[5:3]
Description
LSB = 3.5 µA
00000001b = 1 × LSB; 00000010b = 2 × LSB; 11111111b = 255 × LSB
Default: 0x81 = 451 µA CP current
Division = 14 to 255
Default: 0x14 = divide-by-20
Reserved; default: 0x00
Pole 2 resistor Rp2; default: 0x07
Rp2 (Ω)
Bit 7
Bit 6
500 (default)
0
0
333
0
1
250
1
0
200
1
1
Zero resistor, Rzero
Rzero (Ω)
Bit 5
Bit 4
Bit 3
1500 (default)
0
0
0
1250
0
0
1
1000
0
1
0
930
0
1
1
1250
1
0
0
1000
1
0
1
750
1
1
0
680
1
1
1
Rev. B | Page 79 of 104
AD9558
Data Sheet
Address
Bits
[2:0]
Bit Name
0x0404
[7:1]
0
Reserved
Bypass internal Rzero
0x0405
[7:4]
3
Reserved
APLL locked controlled
sync
[2:1]
0
Reserved
Manual APLL
VCO calibration
1
Description
Pole 1 Cp1.
Cp1 (pF)
Bit 2
Bit 1
Bit 0
0
0
0
0
20
0
0
1
80
0
1
0
100
0
1
1
20
1
0
0
40
1
0
1
100
1
1
0
120 (default)
1
1
1
Default: 0x00.
0 (default) = uses the internal Rzero resistor.
1 = bypasses the internal Rzero resistor (makes Rzero = 0 and requires the use of a series
external zero resistor).
Default: 0x2.
0 (default) = the clock distribution sync function is not enabled until the output PLL (APLL) is
calibrated and locked. After APLL calibration and lock, the output clock distribution sync is
armed, and the sync function for the clock outputs is under the control of Register 0x0500.
1 = overrides the lock detector state of the output PLL; allows Register 0x0500 to control
the output sync function, regardless of the APLL lock status.
Default: 00b.
1 = initiates VCO calibration. (Calibration occurs on low-to-high transition.)
0 (default) = does nothing. This is not an autoclearing bit.
Note that the default APLL loop BW is 180 KHz.
Table 65. Reserved
Address
0x0406
Bits
[7:0]
Bit Name
Reserved
Description
Default: 0x00
Table 66. RF Divider Setting
Address
0x0407
0x0408
Bits
[7:4]
Bit Name
RF Divider 2 division
[3:0]
RF Divider 1 division
[7:5]
4
Reserved
RF divider start-up mode
[3:2]
1
Reserved
PD RF Divider 2
0
PD RF Divider 1
Description
0000/0001 = 3
0010 = 4
0011 = 5
0100 = 6 (default)
0101 = 7
0110 = 8
0111 = 9
1000 = 10
1001 = 11
0000/0001 = 3
0010 = 4
0011 = 5
0100 = 6 (default)
0101 = 7
0110 = 8
0111 = 9
1000 = 10
1001 = 11
Reserved. Default: 000b
0 (default) = RF dividers are held in power-down until the APLL feedback divider is
detected, which ensures proper RF divider operation when exiting full power-down
1 = RF dividers are not held in power-down until the APLL feedback divider is detected
Reserved. Default: 00b
0 = enables RF Divider 2
1 (default) = powers down RF Divider 2
0 (default) = enables RF Divider 1
1 = powers down RF Divider 1
Rev. B | Page 80 of 104
Data Sheet
AD9558
OUTPUT CLOCK DISTRIBUTION (REGISTER 0x0500 TO REGISTER 0x0515)
Table 67. Clock Distribution Output Synchronization Settings
Address
0x0500
Bits
7
Bit Name
Mask Channel 3 sync
6
Mask Channel 2 sync
5
Mask Channel 1 sync
4
Mask Channel 0 sync
3
2
Reserved
Sync source selection
[1:0]
Automatic sync mode
Description
Masks the synchronous reset to the Channel 3 (M3) divider.
0 (default) = unmasked. The output drivers do not toggle until a SYNC pulse occurs.
1 = masked. Setting this bit asynchronously releases Channel 3 from the static SYNC state,
thus allowing the Channel 3 divider to toggle. Channel 3 ignores all SYNC events while
this bit is set. Setting this bit does not enable the output drivers connected to this channel. In
addition, the output distribution sync also depends on the setting of Register 0x0405[3].
Masks the synchronous reset to the Channel 2 (M2) divider.
0 (default) = unmasked. The output drivers do not toggle until a SYNC pulse occurs.
1 = masked. Setting this bit asynchronously releases Channel 2 from the static SYNC state,
thus allowing the Channel 2 divider to toggle. Channel 2 ignores all SYNC events while
this bit is set. Setting this bit does not enable the output drivers connected to this channel. In
addition, the output distribution sync also depends on the setting of Register 0x0405[3].
Masks the synchronous reset to the Channel 1 (M2) divider.
0 (default) = unmasked. The output drivers do not toggle until a SYNC pulse occurs.
1 = masked. Setting this bit asynchronously releases Channel 1 from the static SYNC state,
thus allowing the Channel 1 divider to toggle. Channel 1 ignores all SYNC events while
this bit is set. Setting this bit does not enable the output drivers connected to this
channel. In addition, the output distribution sync also depends on the setting of Register
0x0405[3].
Masks the synchronous reset to the Channel 0 (M0) divider.
0 (default) = unmasked. The output drivers do not toggle until a SYNC pulse occurs.
1 = masked. Setting this bit asynchronously releases Channel 0 from the static SYNC state,
thus allowing the Channel 0 divider to toggle. Channel 0 ignores all SYNC events while
this bit is set. Setting this bit does not enable the output drivers connected to this channel. In
addition, the output distribution sync also depends on the setting of Register 0x0405[3].
Default: 0b.
Selects the sync source for the clock distribution output channels.
0 (default) = direct. The sync pulse happens on the next I/O update.
1 = active reference.
Note that the output distribution sync also depends on the APLL being calibrated and
locked unless Register 0x0405[3] = 1b.
Autosync mode.
00 = disabled. A sync command must be issued manually, or by using the mask sync bits
in this register (Bits[7:4]).
01 = sync on DPLL frequency lock.
10 (default) = sync on DPLL phase lock.
11 = reserved.
Rev. B | Page 81 of 104
AD9558
Data Sheet
Table 68. Distribution OUT0 Setting
Address
0x0501
Bits
7
Bit Name
Enable 3.3 V CMOS driver
[6:4]
OUT0 format
[3:2]
OUT0 polarity
1
OUT0 drive strength
0
Enable OUT0
Description
0 (default) = disables 3.3 V CMOS driver; the OUT5 1.8 V logic is controlled by Register 0x0501[6:4].
1 = enables 3.3 V CMOS driver as operating mode of OUT0.
This bit should be enabled only if Bits[6:4] are in CMOS mode.
This control is valid when Register 0x0501[7] = 0.
When Register 0x0501[7] = 1, OUT5 is in 3.3 V CMOS mode and these bits are ignored.
Select the operating mode of OUT0.
000 = PD, tristate.
001 (default) = HSTL.
010 = LVDS.
011 = reserved.
100 = CMOS, both outputs active.
101 = CMOS, P output active, N output power-down.
110 = CMOS, N output active, P output power-down.
111 = reserved.
Control the OUT0 polarity.
00 (default) = positive, negative.
01 = positive, positive.
10 = negative, positive.
11 = negative, nevative.
Controls the output drive capability of OUT0.
0 (default) = CMOS: low drive strength; LVDS: 3.5 mA nominal.
1 = CMOS: normal drive strength; LVDS: 4.5 mA nominal (LVDS boost mode).
Note that this is only in 3.3 V CMOS mode for CMOS strength.
1.8 V CMOS has only the low drive strength.
Enables/disables (1b/0b) OUT0 1.8 V driver (default is disabled).
This bit does not enable/disable OUT0 if Bit 7 of this register is set.
Table 69. Distribution Channel 0 Divider Setting
Address
0x0502
Bits
[7:0]
Bit Name
Channel 0 (M0)
division ratio
0x0503
[7:4]
3
Reserved
Channel 0 PD
2
Select RF Divider 2
[1:0]
Channel 0 (M0)
division ratio
Reserved
Channel 0 divider phase
0x0504
[7:6]
[5:0]
Description
10-bit channel divider bits[7:0] (LSB).
Division equals channel divider, Bits[9:0] + 1.
(Bits[9:0] = 0 is divide-by-1, Bits[9:0] = 1 is divide-by-2…Bits[9:0] = 1023 is divide-by-1024)
Reserved.
0 (default) = normal operation.
1 = powers down Channel 0.
1 = selects RF Divider 2 as prescaler for Channel 0 divider.
0 (default) = selects RF Divider 1 as prescaler for Channel 0 divider.
10-bit channel divider, Bits[9:8] (MSB)
Reserved. Default: 00b
Divider initial phase after sync relative to the divider input clock (from the RF divider
output). LSB is ½ of a period of the divider input clock. Default: 00000b.
Phase = 0 is no phase offset.
Phase = 1 is ½ a period offset.
Rev. B | Page 82 of 104
Data Sheet
AD9558
Table 70. Distribution OUT1 Setting
Address
0x0505
0x0506
Bits
7
[6:4]
Bit Name
Reserved
OUT1 format
[3:2]
OUT1 polarity
1
OUT1 drive strength
0
7
[6:4]
Enable OUT1
Reserved
OUT2 format
[3:2]
OUT2 polarity
1
OUT2 drive strength
0
Enable OUT2
Description
Reserved; default: 0b
Select the operating mode of OUT1
000 = PD, tristate
001 (default) = HSTL
010 = LVDS
011 = reserved
100 = CMOS, both outputs active
101 = CMOS, P output active, N output power-down
110 = CMOS, N output active, P output power-down
111 = reserved
Configure the OUT1 polarity in CMOS mode and are active in CMOS mode only
00 (default) = positive, negative
01 = positive, positive
10 = negative, positive
11 = negative, negative
Controls the output drive capability of OUT1
0 (default) = LVDS: 3.5 mA nominal
1 = LVDS: 4.5 mA nominal (LVDS boost mode)
No CMOS control because OUT1 is 1.8 V CMOS only
Setting this bit enables the OUT1 driver (default is disabled)
Reserved; default: 0b
Select the operating mode of OUT2
000 = PD, tristate
001 (default) = HSTL
010 = LVDS
011 = reserved
100 = CMOS, both outputs active
101 = CMOS, P output active, N output power-down
110 = CMOS, N output active, P output power-down
111 = reserved
Configure the OUT2 polarity in CMOS mode and are active in CMOS mode only
00 (default) = positive, negative
01 = positive, positive
10 = negative, positive
11 = negative, negative
Controls the output drive capability of OUT2
0 (default) = LVDS: 3.5 mA nominal
1 = LVDS: 4.5 mA nominal (LVDS boost mode)
No CMOS control because OUT2 is 1.8 V CMOS only
Setting this bit enables the OUT2 driver (default is disabled)
Table 71. Distribution Channel 1 Divider Setting
Address
0x0507
0x0508
0x0509
Bits
[7:0]
[7:0]
[7:0]
Bit Name
Channel 1 divider
Channel 1 divider
Channel 1 divider
Description
The same control for Channel 1 divider as in Register 0x0502 for Channel 0 divider
The same control for Channel 1 divider as in Register 0x0503 for Channel 0 divider
The same control for Channel 1 divider as in Register 0x0504 for Channel 0 divider
Table 72. Clock Distribution Channel 2 and OUT3, OUT4 Driver Settings
Address
0x050A
0x050B
0x050C
0x050D
0x050E
Bits
[7:0]
[7:0]
[7:0]
[7:0]
[7:0]
Bit Name
OUT3
OUT4
Channel 2 divider
Channel 2 divider
Channel 2 divider
Description
The same control for OUT3 as in Register 0x0505 for OUT1
The same control for OUT4 as in Register 0x0505 for OUT1
The same control for Channel 2 divider as in Register 0x0502 for Channel 0 divider
The same control for Channel 2 divider as in Register 0x0503 for Channel 0 divider
The same control for Channel 2 divider as in Register 0x0504 for Channel 0 divider
Rev. B | Page 83 of 104
AD9558
Data Sheet
Table 73. Clock Distribution Channel 3 and OUT5 Driver Settings
Address
0x050F
Bits
7
Bit Name
Enable 3.3 V CMOS driver
[6:4]
OUT5 format
[3:2]
OUT5 polarity
1
OUT5 drive strength
0
OUT5 enable
0x0510
[7:0]
Channel 3, Divider 1 (M3)
division ratio
0x0511
[7:2]
[1:0]
0x0512
[7:0]
Reserved
Channel 3, Divider 1 (M3)
division ratio
Channel 3, Divider 2
(M3b) division ratio
0x0513
[7:4]
4
Reserved
Channel 3 doubler
3
PD Channel 3
2
Select RF divider
for Channel 2
[1:0]
Channel 3, Divider 2
(M3b) division ratio
Channel 3, Divider 1 (M3)
phase
Channel 3, Divider 2
(M3b) phase
0x0514
[7:0]
0x0515
[7:0]
Description
0 (default) = disable 3.3 V CMOS driver; the OUT5 1.8 V logic is controlled by Register 0x050F[6:4].
1 = enable 3.3 V CMOS driver as operating mode of OUT5.
This bit should be enabled only if Bits[6:4] are in CMOS mode.
This control is valid when Register 0x050F[7] = 0; when Register 0x050F[7] = 1, OUT5 is in 3.3 V
CMOS mode and these bits are ignored.
Select the operating mode of OUT5.
000 = PD, tristate.
001 (default) = HSTL.
010 = LVDS.
011 = reserved.
100 = CMOS, both outputs active.
101 = CMOS, P output active, N output power-down.
110 = CMOS, N output active, P output power-down.
111 = reserved.
Control the OUT5 polarity.
00 (default) = positive, negative.
01 = positive, positive.
10 = negative, positive.
11 = negative, negative.
Controls the output drive capability of OUT5.
0 (default) = CMOS: low drive strength, LVDS: 3.5 mA nominal.
1 = CMOS: normal drive strength, LVDS: 4.5 mA nominal (LVDS boost mode).
Note that this is only in 3.3 V CMOS mode for CMOS strength.
1.8 V CMOS has only the low drive strength.
Enables/disables (1b/0b) OUT5 1.8 V driver (default is disabled).
This bit does not enable/disable OUT5 if Bit 7 of this register is set.
10-bit channel divider, Bits[7:0] (LSB) (default: 0x03).
Division equals channel divider, Bits[9:0] + 1.
(Bits[9:0] = 0 is divide-by-1, Bits[9:0] = 1 is divide-by-2…Bits[9:0] = 1023 is divide-by-1024).
Reserved.
10-bit channel divider, Bits[9:8] (MSB) (default: 0x00).
10-bit channel divider, Bits[7:0] (LSB).
Division equals channel divider bits[9:0] + 1 .
(Bits[9:0] = 0 is divide-by-1, Bits[9:0] = 1 is divide-by-2…Bits[9:0] = 1023 is divide-by-1024).
Reserved.
0 (default) = normal operation.
1 = enables Channel 3 clock doubler. This bit activates an internal clock doubler that doubles
the frequency of the Channel 3 divider. In this mode, Channel 3, Divider 2 is bypassed.
0 (default) = normal operation.
1 = powers down Channel 3.
1 = selects RF Divider 2 as prescaler for the Channel 3 divider.
0 (default) = selects RF Divider 1 as a prescaler for the Channel 3 divider.
10-bit channel divider, Bits[9:8] (MSB).
The same control for Channel 3, Divider 1 phase as found in Register 0x0504 for Channel 0
divider phase (default: 0x00).
The same control for Channel 3, Divider 2 phase as found in Register 0x0504 for Channel 0
divider phase (default: 0x00).
Rev. B | Page 84 of 104
Data Sheet
AD9558
REFERENCE INPUTS (REGISTER 0x0500 TO REGISTER 0x0507)
Table 74. Reference Power-Down1
Address
0x0600
1
Bits
[7:4]
3
2
1
0
Bit Name
Reserved
REFD power-down
REFC power-down
REFB power-down
REFA power-down
Description
Default: 0x0
Power down REFD input receiver (default: not powered down).
Power down REFC input receiver (default: not powered down).
Power down REFB input receiver (default: not powered down).
Power down REFA input receiver (default: not powered down).
When all bits are set the reference receiver section enters a deep sleep mode.
Table 75. Reference Logic Family
Address
0x0601
Bits
[7:6]
Bit Name
REFD logic family
[5:4]
[3:2]
[1:0]
REFC logic family
REFB logic family
REFA logic family
Description
Select logic family for REFD input receiver; only REFD_P is used in CMOS mode
00 (default) = differential
01 = 1.2 V to 1.5 V CMOS
10 = 1.8 V to 2.5 V CMOS
11 = 3.0 V to 3.3 V CMOS
Same as Register 0x0601[7:6] for REFC
Same as Register 0x0601[7:6] for REFB
Same as Register 0x0601[7:6] for REFA
Table 76. Reference Priority Setting
Address
0x0602
Bits
[7:6]
Bit Name
REFD priority family
[5:4]
[3:2]
[1:0]
REFC priority family
REFB priority family
REFA priority family
Description
User-assigned priority level (0 to 3) of the reference associated with REFB, which ranks that
reference relative to the others.
00 (default) = 0.
01 = 1.
10 = 2.
11 = 3.
Same as Register 0x0602[7:6] for REFC.
Same as Register 0x0602[7:6] for REFB.
Same as Register 0x0602[7:6] for REFA.
Rev. B | Page 85 of 104
AD9558
Data Sheet
FRAME SYNCHRONIZATION (REGISTER 0x0640 TO REGISTER 0x0641)
Table 77. Frame Sync Setting
Address
0x0640
Bit(s)
[7:1]
0
Bit Name
Reserved
Enable Fsync
0x0641
[7:4]
3
Reserved
Validate Fsync ref
2
Fsync one shot
1
Fsync arm method
0
Arm soft Fsync
Description
Reserved; default: 0x00.
Enable frame synchronization.
0 (default) = frame synchronization disabled.
1 = frame synchronization enabled.
Reserved; default: 0x00.
Setting this bit forces the reference validation logic to declare REFA valid only if the
REFB (the sync pulse) input is also valid. This bit can be thought as a logical AND of
REFA VALID and REFB VALID signals. If REFC is selected, this bit requires that REFD
(the sync pulse) input also be valid before declaring REFC valid.
0 (default) = only the selected reference input must be valid.
1 = the sync pulse input must also be valid to validate the selected input.
Selects one-shot or level-sensitive frame sync function.
0 (default) = use level-sensitive frame sync. Frame sync occurs on every edge of the
frame pulse.
1 = use one-shot frame sync. Frame sync occurs only on the first frame sync pulse
(on REFB or REFD). User must re-arm by raising the SYNC pin high and then low, or
by clearing and resetting the arm soft Fsync bit. As with all buffered registers, an
I/O update is required (Register 0x0005[0] = 0x01) after writing this register.
Selects which signal is used to arm the frame sync
0 (default) = use SYNC pin.
1 = use arm soft Fsync (Register 0x0641[0]).
Arms frame sync after I/O update. Next pulse on REFB or REFD is the sync pulse. The
Fsync arm method bit must also be set for this bit to take effect.
0 = (default); frame sync unarmed.
1 = frame sync armed.
Rev. B | Page 86 of 104
Data Sheet
AD9558
DPLL PROFILE REGISTERS (REGISTER 0x0700 TO REGISTER 0x07E6)
Note that the default values of the REFA and REFC profiles are as follows: input frequency=19.44 MHz, output frequency = 622.08 MHz/
155.52 MHz, loop bandwidth = 400 Hz, normal phase margin, inner tolerance = 5%, and outer tolerance = 10%.
The default values of the REFB and REFD profiles are as follows: input frequency = 8 kHz, output frequency = 622.08 MHz/155.52 MHz,
loop bandwidth = 100 Hz, normal phase margin, inner tolerance = 5%, and outer tolerance = 10%.
REFA Profile (Register 0x0700 to Register 0x0726)
Table 78. Reference Period—REFA Profile
Address
0x0700
0x0701
0x0702
0x0703
0x0704
Bits
[7:0]
[7:0]
[7:0]
[7:0]
[7:0]
Bit Name
Nominal reference period (fs)
Description
Nominal reference period bits[7:0] (default: 0xC9).
Nominal reference period bits[15:8] (default: 0xEA).
Nominal reference period bits[23:16] (default: 0x10).
Nominal reference period bits[31:24] (default: 0x03).
Nominal reference period bits[39:32] (default: 0x00).
Default for Register 0x0700 to Register 0x0704 = 0x000310EAC9 = 51.44 ns (1/19.44 MHz).
Table 79. Reference Period Tolerance—REFA Profile
Address
0x0705
0x0706
0x0707
Bits
[7:0]
[7:0]
[7:4]
[3:0]
Bit Name
Inner tolerance
0x0708
0x0709
0x070A
[7:0]
[7:0]
[7:4]
[3:0]
Outer tolerance
Reserved
Inner tolerance
Reserved
Outer tolerance
Description
Input reference frequency monitor inner tolerance, Bits[7:0] (default: 0x14).
Input reference frequency monitor inner tolerance, Bits[15:8] (default: 0x00).
Default: 0x0.
Input reference frequency monitor inner tolerance, Bits[19:16].
Default for Register 0x0705 to Register 0x707 = 0x000014 = 20 (5% or 50,000 ppm).
The Stratum 3 clock requires inner tolerance of ±9.2 ppm and outer tolerance of
±12 ppm; an SMC clock requires an outer tolerance of ±48 ppm.
The allowable range for the inner tolerance is 0x00A (10%) to 0x8FF (1 ppm).
The tolerance of the input frequency monitor is only as accurate as the system clock
frequency.
Input reference frequency monitor outer tolerance, Bits[7:0] (default: 0x0A).
Input reference frequency monitor outer tolerance, Bits[15:8] (default: 0x00).
Reserved.
Input reference frequency monitor outer tolerance, Bits[19:16].
Default for Register 0x0708 to Register 0x70A = 0x00000A = 10 (10% or 100,000 ppm).
The Stratum 3 clock requires an inner tolerance of ±9.2 ppm and outer tolerance of
±12 ppm; an SMC clock requires outer tolerance of ±48 ppm.
The outer tolerance register setting should always be smaller than the inner tolerance.
Table 80. Reference Validation Timer—REFA Profile
Address
0x070B
Bits
[7:0]
0x070C
[7:0]
Bit Name
Validation timer (ms)
Description
Validation timer, Bits[7:0] (default: 0x0A).
This is the amount of time a reference input must be valid before it is declared valid by
the reference input monitor (default: 10 ms).
Validation timer, Bits[15:8] (default: 0x00).
Table 81. Reserved Register
Address
0x070D
Bits
[7:0]
Bit Name
Reserved
Description
Reserved. Default: 0x00.
Table 82. DPLL Base Loop Filter Selection—REFA Profile
Address
0x070E
Bits
[7:1]
0
Bit Name
Reserved
Sel high PM base loop filter
Description
Reserved. Default: 0x00.
0 = base loop filter with normal (70°) phase margin (default).
1 = base loop filter with high (88.5°) phase margin.
(≤0.1 dB peaking in the closed-loop transfer function for loop BWs ≤ 2 kHz. Setting this
bit is also recommended for loop BW > 2kHz.)
Rev. B | Page 87 of 104
AD9558
Data Sheet
Table 83. DPLL Loop BW Scaling Factor—REFA Profile1
Address
0x070F
0x0710
Bits
[7:0]
[7:0]
Bit Name
DPLL loop BW scaling factor
(unit of 0.1 Hz)
0x0711
[7:1]
0
Reserved
BW scaling factor
1
Description
Digital PLL loop bandwidth scaling factor, Bits[7:0] (default: 0xF4).
Digital PLL loop bandwidth scaling factor, Bits[15:8] (default: 0x01).
The default for Register 0x070F to Register 0x0710 = 0x01F4 = 500 (50 Hz loop bandwidth).
The loop bandwidth should always be less than the DPLL phase detector frequency
divided by 20.
Default: 0x00.
Digital PLL loop bandwidth scaling factor, Bit 16 (default: 0b).
Note that the default DPLL loop BW is 50.4 Hz.
Table 84. R-Divider—REFA Profile1
Address
0x0712
0x0713
0x0714
1
Bits
[7:0]
[7:0]
[7:5]
4
Bit Name
R divider
[3:0]
R divider
Reserved
Enable REFA div2
Description
DPLL integer reference divider (minus 1), Bits[7:0] (default: 0xC5)
DPLL integer reference divider, Bits[15:8] (default: 0x00)
Reserved. Default: 0x0
Enables the reference input divide-by-2 for REFA
0 = bypass the divide-by-2 (default)
1 = enable the divide-by-2
DPLL integer reference divider, Bits[19:16] (default: 0x0).
The default for Register 0x0712 to Register 0x0714 = 0x000C5 = 197 (which equals R = 198).
Note that the value stored in the R-divider register yields an actual divide ratio of one more than the programmed value.
Table 85. Integer Part of Fractional Feedback Divider N1—REFA Profile1
Address
0x0715
0x0716
0x0717
1
Bits
[7:0]
[7:0]
[7:1]
0
Bit Name
Integer Part N1
Reserved
Integer Part N1
Description
DPLL integer feedback divider (minus 1), Bits[7:0] (default: 0x6B).
DPLL integer feedback divider, Bits[15:8] (default: 0x07).
Reserved. Default: 0x00.
DPLL integer feedback divider, Bit 16 (default: 0b).
The default for Register 0x0715 to Register 0x717 = 0x0076B (which equals N1 = 1900).
Note that the value stored in the N1-divider register yields an actual divide ratio of one more than the programmed value.
Table 86. Fractional Part of Fractional Feedback Divider FRAC1—REFA Profile
Address
0x0718
0x0719
0x071A
Bits
[7:0]
[7:0]
[7:0]
Bit Name
Digital PLL fractional feedback
divider—FRAC1
Description
The numerator of the fractional-N feedback divider, Bits[7:0] (default: 0x04)
The numerator of the fractional-N feedback divider, Bits[15:8] (default: 0x00)
The numerator of the fractional-N feedback divider, Bits[23:16] (default: 0x00)
Table 87. Modulus of Fractional Feedback Divider MOD1—REFA Profile
Address
0x071B
0x071C
0x071D
Bits
[7:0]
[7:0]
[7:0]
Bit Name
Digital PLL feedback divider
modulus—MOD1
Description
The denominator of the fractional-N feedback divider, Bits[7:0] (default: 0x05)
The denominator of the fractional-N feedback divider, Bits[15:8] (default: 0x00)
The denominator of the fractional-N feedback divider, Bits[23:16] (default: 0x00)
Table 88. Phase and Frequency Lock Detector Controls—REFA Profile
Address
0x071E
0x071F
0x0720
0x0721
0x0722
0x0723
0x0724
0x0725
0x0726
Bits
[7:0]
[7:0]
[7:0]
[7:0]
[7:0]
[7:0]
[7:0]
[7:0]
[7:0]
Bit Name
Phase lock threshold
Phase lock fill rate
Phase lock drain rate
Frequency lock threshold
Frequency lock fill rate
Frequency lock drain rate
Description
Phase lock threshold, Bits[7:0] (default: 0xBC); default of 0x02BC = 700 ps
Phase lock threshold, Bits[15:8] (default: 0x02)
Phase lock fill rate, Bits[7:0] (default:0x0A = 10 code/PFD cycle)
Phase lock drain rate, Bits[7:0] (default: 0x0A = 10 code/PFD cycle)
Frequency lock threshold, Bits[7:0] (default: 0xBC); default of 0x02BC = 700 ps
Frequency lock threshold, Bits[15:8] (default: 0x02)
Frequency lock threshold, Bits[23:16] (default: 0x00)
Frequency lock fill rate, Bits[7:0] (default: 0x0A = 10 code/PFD cycle)
Frequency lock drain rate bits[7:0] (default: 0x0A = 10 code/PFD cycle)
Rev. B | Page 88 of 104
Data Sheet
AD9558
REFB Profile (Register 0x0740 to Register 0x0766)
REFB Profile Register 0x0740 to Register 0x0766 are identical to REFA Profile Register 0x0700 to Register 0x0726.
REFC Profile (Register 0x0780 to Register 0x07A6)
REFC Profile Register 0x0780 to Register 0x07A6 are identical to REFA Profile Register 0x0700 Register 0x0726.
REFD Profile (Register 0x07C0 to Register 0x07E6)
REFD Profile Register 0x07C0 to Register 0x07E6 are identical to REFA Profile Register 0x0700 to Register 0x0726.
OPERATIONAL CONTROLS (REGISTER 0x0A00 TO REGISTER 0x0A10)
Table 89. General Power-Down
Address
0x0A00
Bits
7
6
5
5
3
2
1
0
Bit Name
Soft reset exclude regmap
DCO power-down
SYSCLK power-down
Reference input power-down
TDC power-down
APLL power-down
Clock dist power-down
Full power-down
Description
Resets device but retains programmed register values (default: not reset)
Places DCO in deep sleep mode (default: not powered down)
Places SYSCLK input and PLL in deep sleep mode (default: not powered down)
Places reference clock inputs in deep sleep mode (default: not powered down)
Places the time-to-digital converter in deep sleep mode (default: not powered down)
Places the output PLL in deep sleep mode (default: not powered down)
Places the clock distribution outputs in deep sleep mode (default: not powered down)
Places the entire device in deep sleep mode (default: not powered down)
Table 90. Loop Mode
Address
0x0A01
Bits
7
6
Bit Name
Reserved
User holdover
5
User freerun
[4:2]
REF switchover mode
[1:0]
User reference in manual
switchover mode
Description
Reserved.
Forces the device into holdover mode (default: not forced holdover mode).
If a tuning word history is available, then the history tuning word specifies the DCO
output frequency. Otherwise, the free run frequency tuning word register specifies the
DCO output frequency.
The phase and frequency lock detectors are forced into the unlocked state.
Forces the device into user free run mode (default is not forced user free run mode).
The free run frequency tuning word register specifies the DCO output frequency. When
the user freerun bit is set, it overrides the user holdover bit (Address 0x0A01, Bit 6).
Selects the operating mode of the reference switching state machine.
Reference Switchover Mode, Bits[2:0]
Register 0x0A01[4:2]
Reference Selection Mode
000 (default)
Automatic revertive mode
001
Automatic nonrevertive mode
010
Manual reference select
(with automatic fallback mode)
011
Manual reference select mode
(with auto-holdover)
100
Full manual mode (no auto-holdover)
101
Not used
110
Not used
111
Not used
Input reference when REF switchover mode bits (Register 0x0A01, Bits[4:2]) = 010, 011, or 100.
00 (default) = Input Reference A.
01 = Input Reference B.
10 = Input Reference C.
11 = Input Reference D.
Table 91. Cal/Sync Distribution
Address
0x0A02
Bits
[7:2]
1
Bit Name
Reserved
Soft sync clock distribution
0
Reserved
Description
Default: 0x00
Setting this bit initiates synchronization of the clock distribution output (default: 0b).
Nonmasked outputs stall when value is 1b; restart is initialized on 1b to 0b transition.
Default: 0b.
Rev. B | Page 89 of 104
AD9558
Data Sheet
Reset Functions (Register 0x0A03)
Table 92. Reset Functions1
Address
0x0A03
(autoclear)
1
Bits
7
6
5
4
3
2
1
Bit Name
Reserved
Clear LF
Clear CCI
Reserved
Clear auto sync
Clear TW history
Clear all IRQs
0
Clear watchdog timer
Description
Default: 0b.
Setting this bit clears the digital loop filter (intended as a debug tool).
Setting this bit clears the CCI filter (intended as a debug tool).
Default: 0b.
Setting this bit resets the automatic synchronization logic (see Register 0x0500).
Setting this bit resets the tuning word history logic (part of holdover functionality).
Setting this bit clears the entire IRQ monitor register (Register 0x0D02 to Register 0x0D07).
It is the equivalent of setting all the bits of the IRQ clearing register (Register 0x0A04 to
Register 0x0A0D).
Setting this bit resets the watchdog timer (see Register 0x0211 and Register 0x0212). If the
timer times out, it starts a new timing cycle. If the timer has not yet timed out, it restarts at
time zero without causing a timeout event. Continuously resetting the watchdog timer at
intervals less than its timeout period prevents the watchdog timer from generating a timeout
event.
Note that all bits in this register are autoclearing.
IRQ Clearing (Register 0x0A04 to Register 0x0A09)
The IRQ clearing registers are identical in format to the IRQ monitor registers (Register 0x0D02 to Register 0x0D09). When set to logic 1,
an IRQ clearing bit resets the corresponding IRQ monitor bit, thereby canceling the interrupt request for the indicated event. The IRQ
clearing register is an autoclearing register.
Table 93. IRQ Clearing for SYSCLK
Address
0x0A04
Bits
[7:6]
5
4
3
2
1
0
Bit Name
Reserved
SYSCLK unlocked
SYSCLK locked
APLL unlocked
APLL locked
APLL Cal ended
APLL Cal started
Description
Reserved
Clears SYSCLK unlocked IRQ
Clears SYSCLK locked IRQ
Clears Output PLL unlocked IRQ
Clears Output PLL locked IRQ
Clears APLL calibration complete IRQ
Clears APLL calibration started IRQ
Table 94. IRQ Clearing for Distribution Sync, Watchdog Timer, and EEPROM
Address
0x0A05
Bits
[7:5]
4
3
2
1
0
Bit Name
Reserved
Pin program end
Sync clock distribution
Watchdog timer
EEPROM fault
EEPROM complete
Description
Reserved
Clears pin program end IRQ
Clears distribution sync IRQ
Clears watchdog timer IRQ
Clears EEPROM fault IRQ
Clears EEPROM complete IRQ
Table 95. IRQ Clearing for the Digital PLL
Address
0x0A06
Bits
7
6
5
4
3
2
1
0
Bit Name
Switching
Closed
Freerun
Holdover
Frequency unlocked
Frequency locked
Phase unlocked
Phase locked
Description
Clears switching IRQ
Clears closed IRQ
Clears free run IRQ
Clears holdover IRQ
Clears frequency unlocked IRQ
Clears frequency locked IRQ
Clears phase unlocked IRQ
Clears phase locked IRQ
Rev. B | Page 90 of 104
Data Sheet
AD9558
Table 96. IRQ Clearing for History Update, Frequency Limit, and Phase Slew Limit
Address
0x0A07
Bits
[7:5]
4
3
2
1
0
Bit Name
Reserved
History updated
Frequency unclamped
Frequency clamped
Phase slew unlimited
Phase slew limited
Description
Reserved
Clears history updated IRQ
Clears frequency unclamped IRQ
Clears frequency clamped IRQ
Clears phase slew unlimited IRQ
Clears phase slew limited IRQ
Table 97. IRQ Clearing for Reference Inputs
Address
0x0A08
0x0A09
Bits
7
6
5
4
3
2
1
0
7
6
5
4
3
2
1
0
Bit Name
Reserved
REFB validated
REFB fault cleared
REFB fault
Reserved
REFA validated
REFA fault cleared
REFA fault
Reserved
REFD validated
REFD fault cleared
REFD fault
Reserved
REFC validated
REFC fault cleared
REFC fault
Description
Reserved
Clears REFB validated IRQ
Clears REFB fault cleared IRQ
Clears REFB fault IRQ
Reserved
Clears REFA validated IRQ
Clears REFA fault cleared IRQ
Clears REFA fault IRQ
Reserved
Clears REFD validated IRQ
Clears REFD fault cleared IRQ
Clears REFD fault IRQ
Reserved
Clears REFC validated IRQ
Clears REFC fault cleared IRQ
Clears REFC fault IRQ
Table 98. Incremental Phase Offset Control
Address
0x0A0A
Bits
[7:3]
2
Bit Name
Reserved
Reset phase offset
1
Decrement phase offset
0
Increment phase offset
Description
Reserved.
Resets the incremental phase offset to zero.
This is an autoclearing bit.
Decrements the incremental phase offset by the amount specified in the incremental
phase lock offset step size registers (Register 0x0312 to Register 0x0313).
This is an autoclearing bit.
Increments the incremental phase offset by the amount specified in the incremental
phase lock offset step size registers (Register 0x0312 to Register 0x0313).
This is an autoclearing bit.
Table 99. Reference Validation Override Controls
Address
0x0A0B
0x0A0C
Bits
[7:4]
3
Bit Name
Reserved
Force Timeout D
2
Force Timeout C
1
Force Timeout B
0
Force Timeout A
[7:4]
3
2
1
0
Reserved
Ref Mon Override D
Ref Mon Override C
Ref Mon Override B
Ref Mon Override A
Description
Reserved.
Setting this autoclearing bit emulates timeout of the validation timer for Reference D
and allows the user to make REFD valid immediately.
Setting this autoclearing bit emulates timeout of the validation timer for Reference C
and allows the user to make REFC valid immediately.
Setting this autoclearing bit emulates timeout of the validation timer for Reference B
and allows the user to make REFB valid immediately.
Setting this autoclearing bit emulates timeout of the validation timer for Reference A
and allows the user to make REFA valid immediately.
Reserved.
Overrides the reference monitor REF FAULT signal for Reference D (default: 0).
Overrides the reference monitor REF FAULT signal for Reference C (default: 0).
Overrides the reference monitor REF FAULT signal for Reference B (default: 0).
Overrides the reference monitor REF FAULT signal for Reference A (default: 0).
Rev. B | Page 91 of 104
AD9558
Address
0x0A0D
Bits
[7:4]
3
2
1
0
Data Sheet
Bit Name
Reserved
Ref Mon Bypass D
Ref Mon Bypass C
Ref Mon Bypass B
Ref Mon Bypass A
Description
Reserved.
Bypasses the reference monitor for Reference D (default: 0).
Bypasses the reference monitor for Reference C (default: 0).
Bypasses the reference monitor for Reference B (default: 0).
Bypasses the reference monitor for Reference A (default: 0).
QUICK IN/OUT FREQUENCY SOFT PIN CONFIGURATION (REGISTER 0x0C00 TO REGISTER 0x0C08)
Table 100. Soft Pin Program Setting1
Address
0x0C00
Bits
[7:1]
0
Bit Name
Reserved
Enable Soft Pin Section 1
0x0C01
[7:4]
Output frequency selection
[3:0]
Input frequency selection
0x0C02
[7:2]
[1:0]
Reserved
System clock PLL ref
selection
0x0C03
[7:1]
0
Reserved
Enable Soft Pin Section 2
0x0C04
[7:4]
[3:2]
Reserved
REFB frequency scale
[1:0]
REFA frequency scale
Description
Reserved.
0 (default) = disables the function of soft pin registers in Soft Pin Section 1 (Register
0x0C01 and Register 0x0C02).
1 = enables the function of soft pin registers in Soft Pin Section 1 (Register 0x0C01 to
Register 0x0C02) when the PINCONTROL pin is low at startup and/or reset.
The register in Soft Pin Section 1 configures the part into one of 256 preconfigured
input-to-output frequency translations stored in the on-chip ROM.
The registers in Soft Pin Section 1 (Register 0x0C00 to Register 0x0C02) are ignored
when the PINCONTROL pin is high at power-up and/or reset (which means the hard pin
program is enabled).
Selects one of 16 predefined output frequencies as ouptut freqeuncy of the desired
frequency translation and reprograms the free run TW, N2, RF divider, and M0 to M3b
dividers with the value stored in the ROM.
Selects one of 16 predefined input frequencies as the input frequency of the desired
frequency translation and reprograms the reference period, R divider, N1, FRAC1, and
MOD1 in four REF profiles with the value stored in the ROM.
Reserved.
Selects one of the four predefined system PLL references for the desired frequency
translation and reprograms the system PLL configuration with the value stored in the
ROM. To load values from the ROM, the user must write Register 0x0C07[0] = 1 after writing
this.
Equivalent System Clock PLL Settings,
Register 0x0C02[1:0] Register 0x0100 to Register 0x101[3:0]
System PLL Ref
Bit 1
Bit 0
12 Bits
1
0
0
24.576 MHz XTAL, ×2 on, N = 8
2
0
1
49.152 MHz XTAL, ×2 on, N = 8
3
1
0
24.576 MHz XO, ×2 off, N = 16
4
1
1
49.152 MHz XO, ×2 off, N = 8
Reserved.
0 (default) = disables the function of soft pin registers in Soft Pin Section 2 (Register
0x0C04 to Register 0x0C06).
1 = enables the function of soft pin registers in Soft Pin Section 2 (Register 0x0C04 to
Register 0x0C06) when PINCONTROL pin is low.
Reserved.
Scales the the selected input frequency (defined by Register 0x0C01[3:0]) for REFB.
00 (default) = divide-by-1.
01 = divide-by-4.
10 = divide-by-8.
11 = divide-by-16.
For example, if the selected input frequency is 622.08 MHz and Register 0x0C04[3:2] =
11, the new input frequency should be 622.08 MHz/16 = 38.8 MHz
Scales the the selected input frequency (defined by Register 0x0C01[3:0]) for REFA.
00 (default) = divide-by-1.
01 = divide-by-4.
10 = divide-by-8.
11 = divide-by-16.
Rev. B | Page 92 of 104
Data Sheet
Address
0x0C05
AD9558
Bits
[7:4]
[3:2]
Bit Name
Reserved
Channel 1 output frequency
scale
[1:0]
Channel 0 output frequency
scale
[7:5]
4
Reserved
Sel high PM base loop filter
[3:2]
DPLL loop BW
[1:0]
Reference input frequency
tolerance
0x0C07
[7:1]
0
Reserved
Soft pin start transfer
0x0C08
[7:1]
0
Reserved
Soft pin reset
0x0C06
1
Description
Reserved.
Scales the selected output frequency (defined by Register 0x0C01[7:4]) for the Channel
Divider 1 output.
00 (default) = divide-by-1.
01 = divide-by-4.
10 = divide-by-8.
11 = divide-by-16.
Scale the selected output frequency (defined by Register 0x0C01[7:4]) for the Channel
Divider 0 output.
00 (default) = divide-by-1.
01 = divide-by-4.
10 = divide-by-8.
11 = divide-by-16.
Reserved.
0 = base loop filter with normal (70°) phase margin (default).
1 = base loop filter with high (88.5°) phase margin.
(< 0.1 dB peaking in closed-loop transfer function).
Scale the DPLL loop BW while in soft pin mode.
00 (default) = 50 Hz.
01 = 1 Hz.
10 = 10 Hz.
11 =100 Hz.
Scales the input frequency tolerance while in soft pin mode.
00 (default) = outer tolerance: 10%; inner tolerance: 8% (for general conditions).
01 = outer tolerance: 12 ppm; inner tolerance: 9.6 ppm (for Stratum 3).
10 = outer tolerance: 48 ppm; inner tolerance: 38 ppm (for SMC clock standard).
11 = outer tolerance: 200 ppm; inner tolerance: 160 ppm (for XTAL system clock).
Reserved.
Autoclearing register. 1 = initiates ROM download without resetting the part/register map.
After ROM download is complete, this register is reset.
Reserved.
Autoclearing register; resets the part like soft reset (Register 0x0000[5]), except that this
reset function initiates a soft pin ROM download without resetting the part/register map.
After ROM download is complete, this register is pulled back to zero.
All bits in Register 0x0C00 to Register 0x0C06 take effect only with either a soft pin start transfer (Register 0x0C07) or soft pin reset (Register 0x0C08).
STATUS READBACK (REGISTER 0x0D00 TO REGISTER 0x0D14)
All bits in Register 0x0D00 to Register 0x0D14 are read only. To show the latest status, these registers require an I/O update (Register 0x0005 =
0x01) immediately before being read.
Table 101. EEPROM Status
Address
0x0D00
Bits
[7:4]
3
2
1
0
Bit Name
Reserved
Pin program ROM load process
Fault detected
Load in progress
Save in progress
Description
Reserved.
The control logic sets this bit when data is being read from the ROM.
An error occurred while saving data to or loading data from the EEPROM.
The control logic sets this bit while data is being read from the EEPROM.
The control logic sets this bit while data is being written to the EEPROM.
Rev. B | Page 93 of 104
AD9558
Data Sheet
Table 102. SYSCLK Status
Address
0x0D01
Bits
7
6
Bit Name
Reserved
DPLL_APLL_Lock
5
All PLLs locked
4
APLL VCO status
3
2
APLL cal in process
APLL lock
1
System clock stable
0
SYSCLK lock detect
Description
Reserved.
Indicates the status of the DPLL and APLL.
0 = either the DPLL or the APLL is unlocked.
1 = both the DPLL and APLL are locked.
Indicates the status of the system clock PLL, APLL, and DPLL.
0 = system clock PLL or APLL or DPLL is unlocked.
1 = all three PLLs (system clock PLL, APLL, and DPLL) are locked.
1 = OK.
0 = off/clocks are missing.
The control logic holds this bit set while the amplitude calibration of the APLL VCO is in progress.
Indicates the status of the APLL.
0 = unlocked.
1 = locked.
The control logic sets this bit when the device considers the system clock to be stable (see
the System Clock Stability Timer section).
0 = not stable (the system clock stability timer has not expired yet).
1 = stable (the system clock stability timer has expired).
Indicates the status of the system clock PLL.
0 = unlocked.
1 = locked.
IRQ Monitor (Register 0x0D02 to Register 0x0D09)
If not masked via the IRQ mask registers (Register 0x0209 to Register 0x020F), the appropriate IRQ monitor bit is set to a Logic 1 when
the indicated event occurs. These bits can be cleared only via the IRQ clearing registers (Register 0x0A04 to Register 0x0A09), the reset all
IRQs bit (Register 0x0A03[1]), or a device reset.
Table 103. IRQ Monitor for SYSCLK
Address
0x0D02
Bits
[7:6]
5
4
3
2
1
0
Bit Name
Reserved
SYSCLK unlocked
SYSCLK locked
APLL unlocked
APLL locked
APLL cal ended
APLL cal started
Description
Reserved
Indicates a SYSCLK PLL state transition from locked to unlocked
Indicates a SYSCLK PLL state transition from unlocked to locked
Indicates an output PLL state transition from locked to unlocked
Indicates an output PLL state transition from unlocked to locked
Indicates that APLL calibration is complete
Indicates that APLL in APLL calibration has begun
Table 104. IRQ Monitor for Distribution Sync, Watchdog Timer, and EEPROM
Address
0x0D03
Bit
[7:5]
4
3
2
1
0
Bit Name
Reserved
Pin program end
Output distribution sync
Watchdog timer
EEPROM fault
EEPROM complete
Description
Reserved
Indicates successful completion of a ROM load operation
Indicates a distribution sync event
Indicates expiration of the watchdog timer
Indicates a fault during an EEPROM load or save operation
Indicates successful completion of an EEPROM load or save operation
Rev. B | Page 94 of 104
Data Sheet
AD9558
Table 105. IRQ Monitor for the Digital PLL
Address
0x0D04
Bits
7
6
5
4
3
2
1
0
Bit Name
Switching
Closed
Freerun
Holdover
Frequency unlocked
Frequency locked
Phase unlocked
Phase locked
Description
Indicates that the DPLL is switching to a new reference
Indicates that the DPLL has entered closed-loop operation
Indicates that the DPLL has entered free run mode
Indicates that the DPLL has entered holdover mode
Indicates that the DPLL has lost frequency lock
Indicates that the DPLL has acquired frequency lock
Indicates that the DPLL has lost phase lock
Indicates that the DPLL has acquired phase lock
Table 106. IRQ Monitor for History Update, Frequency Limit, and Phase Slew Limit
Address
0x0D05
Bits
[7:5]
4
3
2
1
0
Bit Name
Reserved
History updated
Frequency unclamped
Frequency clamped
Phase slew unlimited
Phase slew limited
Description
Reserved
Indicates the occurrence of a tuning word history update
Indicates a frequency limiter state transition from clamped to unclamped
Indicates a frequency limiter state transition from unclamped to clamped
Indicates a phase slew limiter state transition from slew limiting to not slew limiting
Indicates a phase slew limiter state transition from not slew limiting to slew limiting
Table 107. IRQ Monitor for Reference Inputs
Address
0x0D06
0x0D07
Bits
7
6
5
4
3
2
1
0
7
6
5
4
3
2
1
0
Bit Name
Reserved
REFB validated
REFB fault cleared
REFB fault
Reserved
REFA validated
REFA fault cleared
REFA fault
Reserved
REFD validated
REFD fault cleared
REFD fault
Reserved
REFC validated
REFC fault cleared
REFC fault
Description
Reserved
Indicates that REFB has been validated
Indicates that REFB has been cleared of a previous fault
Indicates that REFB has been faulted
Reserved
Indicates that REFA has been validated
Indicates that REFA has been cleared of a previous fault
Indicates that REFA has been faulted
Reserved
Indicates that REFD has been validated
Indicates that REFD has been cleared of a previous fault
Indicates that REFD has been faulted
Indicates that REFC has been validated
Indicates that REFC has been cleared of a previous fault
Indicates that REFC has been faulted
Rev. B | Page 95 of 104
AD9558
Data Sheet
DPLL Status, Input Reference Status, Holdover History, and DPLL Lock Detect Tub Levels (Register 0x0D08 to Register 0x0D14)
Table 108. DPLL Status1
Address
0x0D08
0x0D09
1
Bits
7
6
5
4
3
2
1
0
[7:6]
5
4
[3:2]
Bit Name
Reserved
Offset slew limiting
Frequency lock
Phase lock
Loop switching
Holdover
Active
Freerun
Reserved
Frequency clamped
History available
Active reference priority
[1:0]
Current active reference
Description
Reserved
The current closed-loop phase offset is rate limited
The DPLL has achieved frequency lock
The DPLL has achieved phase lock
The DPLL is in the process of a reference switchover
The DPLL is in holdover mode
The DPLL is active (that is, operating in a closed-loop condition)
The DPLL is free run (that is, operating in an open-loop condition)
Default: 0b
The upper or lower frequency tuning word clamp is in effect
There is sufficient tuning word history available for holdover operation
Priority value of the currently active reference
00 = highest priority
…
11 = lowest priority
Index of the currently active reference
00 = Reference A
01 = Reference B
10 = Reference C
11 = Reference D
Note that the user must issue an I/O update by writing 0x01 to Register 0x0005 to update the status of these registers.
Table 109. Reserved Register
Address
0x0D0A
Bits
[7:0]
Bit Name
Reserved
Description
Reserved.
Table 110. Input Reference Status1
Address
0x0D0B
0x0D0C
1
Bits
7
6
5
4
3
2
1
0
[7:4]
[3:0]
Bit Name
B valid
B fault
B fast
B slow
A valid
A fault
A fast
A slow
Description
REFB is valid for use. (It is unfaulted, and its validation timer has expired.)
REFB is not valid for use.
This bit indicates that the frequency of REFB is higher than allowed by its profile settings.
This bit indicates that the frequency of REFB is lower than allowed by its profile settings.
REFA is valid for use. (It is unfaulted, and its validation timer has expired.)
REFA is not valid for use.
This bit indicates that the frequency of REFA is higher than allowed by its profile settings.
This bit indicates that the frequency of REFA is lower than allowed by its profile settings.
Same as Register 0xDOB[7:4] but for REFD instead of REFB .
Same as Register 0xDOB[3:0] but for REFC instead of REFA.
Note that the user must issue an I/O update by writing 0x01 to Register 0x0005 to update the status of these registers.
Table 111. Holdover History1
Address
0x0D0D
0x0D0E
0x0D0F
0x0D10
1
Bits
[7:0]
[7:0]
[7:0]
[7:0]
Bit Name
Holdover history
Description
Tuning word readback bits[7:0]
Tuning word readback bits[15:8]
Tuning word readback bits[23:9]
Tuning word readback bits[29:24]
Note that these registers contain the current 30-bit DCO frequency tuning word generated by the tuning word history logic.
Rev. B | Page 96 of 104
Data Sheet
AD9558
Table 112. Digital PLL Lock Detect Tub Levels1
Address
0x0D11
Bits
[7:0]
Bit Name
Phase tub
0x0D12
[7:4]
[3:0]
0x0D13
[7:0]
Frequency tub
0x0D14
[7:4]
[3:0]
Reserved
Frequency tub
Description
Read-only digital PLL lock detect bathtub level, Bits[7:0]; see the DPLL Frequency Lock Detector
section for details.
Reserved.
Read-only digital PLL lock detect bathtub level, Bits[11:8]; see the DPLL Frequency Lock
Detector section for details.
Read-only digital PLL lock detect bathtub level, Bits[7:0]; see the DPLL Phase Lock Detector
section for details.
Reserved.
Read-only digital PLL lock detect bathtub level, Bits[11:8]; see the DPLL Phase Lock Detector
section for details.
These registers contain the current digital PLL lock detection bathtub levels.
1
EEPROM CONTROL (REGISTER 0x0E00 TO REGISTER 0x0E03)
Table 113. EEPROM Control
Address
0x0E00
Bits
[7:1]
0
Bit Name
Reserved
Write enable
0x0E01
[7:4]
[3:0]
[7:1]
0
Reserved
Conditional value
Reserved
Save to EEPROM
[7:2]
1
0
Reserved
Load from EPROM
Reserved
0x0E02
0x0E03
Description
Reserved.
EEPROM write enable/protect.
0 (default) = EEPROM write protected.
1 = EEPROM write enabled.
Reserved.
When set to a non-zero value, it establishes the condition for EEPROM downloads. Default: 0b.
Reserved.
Upload data to the EEPROM based on the EEPROM Storage Sequence (Register 0x0E10 to
Register 0x0E3C) section.
Reserved.
Downloads data from the EEPROM.
Reserved.
Rev. B | Page 97 of 104
AD9558
Data Sheet
EEPROM STORAGE SEQUENCE (REGISTER 0x0E10 TO REGISTER 0x0E3C)
The default settings of Register 0x0E10 to Register 0x0E3C contain the default EEPROM instruction sequence. The tables in this section
provide descriptions of the register defaults, based on the assumption that the controller has been instructed to carry out an EEPROM
storage sequence in which all of the registers are stored and loaded by the EEPROM.
Table 114. EEPROM Storage Sequence for System Clock Settings
Address
0x0E10
Bits
[7:0]
0x0E11
[7:0]
0x0E12
[7:0]
0x0E13
[7:0]
0x0E14
[7:0]
0x0E15
[7:0]
0x0E16
[7:0]
Bit Name
EEPROM ID
System clock
I/O update
Description
The default value of this register is 0x01, which the controller interprets as a data instruction. Its decimal
value is 1, so this tells the controller to transfer two bytes of data (1 + 1), beginning at the address that is
specified by the next two bytes. The controller stores 0x01 in the EEPROM and increments the EEPROM
address pointer.
The default value of these two registers is 0x0006. Note that Register 0x0E11 and Register 0x0E12 are
the most significant and least significant bytes of the target address, respectively. Because the previous
register contains a data instruction, these two registers define a starting address (in this case, 0x0006).
The controller stores 0x0006 in the EEPROM and increments the EEPROM pointer by 2. It then transfers
two bytes from the register map (beginning at Address 0x0006) to the EEPROM and increments the
EEPROM address pointer by 3 (two data bytes and one checksum byte). The two bytes transferred
correspond to the system clock parameters in the register map.
The default value of this register is 0x08, which the controller interprets as a data instruction. Its decimal
value is 8, so this tells the controller to transfer nine bytes of data (8 + 1), beginning at the address that
is specified by the next two bytes. The controller stores 0x08 in the EEPROM and increments the
EEPROM address pointer.
The default value of these two registers is 0x0100. Note that Register 0x0E14 and Register 0x0E15 are
the most significant and least significant bytes of the target address, respectively. Because the previous
register contains a data instruction, these two registers define a starting address (in this case, 0x0100).
The controller stores 0x0100 in the EEPROM and increments the EEPROM pointer by 2. It then transfers
nine bytes from the register map (beginning at Address 0x0100) to the EEPROM and increments the
EEPROM address pointer by 10 (nine data bytes and one checksum byte). The nine bytes transferred
correspond to the system clock parameters in the register map.
The default value of this register is 0x80, which the controller interprets as an I/O update instruction.
The controller stores 0x80 in the EEPROM and increments the EEPROM address pointer.
Table 115. EEPROM Storage Sequence for General Configuration Settings
Address
0x0E17
Bits
[7:0]
0x0E18
[7:0]
0x0E19
[7:0]
Bit Name
General
Description
The default value of this register is 0x11, which the controller interprets as a data instruction. Its decimal
value is 17, which tells the controller to transfer 18 bytes of data (17 + 1), beginning at the address that
is specified by the next two bytes. The controller stores 0x11 in the EEPROM and increments the
EEPROM address pointer.
The default value of these two registers is 0x0200. Note that Register 0x0E18 and Register 0x0E19 are
the most significant and least significant bytes of the target address, respectively. Because the previous
register contains a data instruction, these two registers define a starting address (in this case, 0x0200).
The controller stores 0x0200 in the EEPROM and increments the EEPROM pointer by 2. It then transfers
18 bytes from the register map (beginning at Address 0x0200) to the EEPROM and increments the EEPROM
address pointer by 19 (18 data bytes and one checksum byte). The 18 bytes transferred correspond to
the general configuration parameters in the register map.
Table 116. EEPROM Storage Sequence for DPLL Settings
Address
0x0E1A
Bits
[7:0]
0x0E1B
[7:0]
0x0E1C
[7:0]
Bit Name
DPLL
Description
The default value of this register is 0x2E, which the controller interprets as a data instruction. Its decimal
value is 46, which tells the controller to transfer 47 bytes of data (46 + 1), beginning at the address that
is specified by the next two bytes. The controller stores 0x2E in the EEPROM and increments the EEPROM
address pointer.
The default value of these two registers is 0x03. Note that Register 0x0E1B and Register 0x0E1C are the
most significant and least significant bytes of the target address, respectively. Because the previous
register contains a data instruction, these two registers define a starting address (in this case, 0x0300).
The controller stores 0x0300 in the EEPROM and increments the EEPROM pointer by 2. It then transfers
47 bytes from the register map (beginning at Address 0x0300) to the EEPROM and increments the EEPROM
address pointer by 48 (47 data bytes and one checksum byte). The 47 bytes transferred correspond to
the DPLL parameters in the register map.
Rev. B | Page 98 of 104
Data Sheet
AD9558
Table 117. EEPROM Storage Sequence for Output PLL Settings
Address
0x0E1D
Bits
[7:0]
0x0E1E
[7:0]
0x0E1F
[7:0]
Bit Name
APLL
Description
The default value of this register is 0x08, which the controller interprets as a data instruction. Its
decimal value is 8, which tells the controller to transfer nine bytes of data (8 + 1), beginning at the
address that is specified by the next two bytes. The controller stores 0x08 in the EEPROM and
increments the EEPROM address pointer.
The default value of these two registers is 0x0400. Note that Register 0x0E1E and Register 0x0E1F
are the most significant and least significant bytes of the target address, respectively. Because the
previous register contains a data instruction, these two registers define a starting address (in this
case, 0x0400). The controller stores 0x0400 in the EEPROM and increments the EEPROM pointer by 2.
It then transfers nine bytes from the register map (beginning at Address 0x0400) to the EEPROM
and increments the EEPROM address pointer by 10 (nine data bytes and one checksum byte). The
nine bytes transferred correspond to APLL parameters in the register map.
Table 118. EEPROM Storage Sequence for Clock Distribution Settings
Address
0x0E20
Bits
[7:0]
0x0E21
[7:0]
0x0E22
[7:0]
0x0E23
[7:0]
Bit Name
Clock distribution
I/O update
Description
The default value of this register is 0x15, which the controller interprets as a data instruction. Its
decimal value is 21, which tells the controller to transfer 22 bytes of data (21+1), beginning at the
address that is specified by the next two bytes. The controller stores 0x15 in the EEPROM and
increments the EEPROM address pointer.
The default value of these two registers is 0x0500. Note that Register 0x0E21 and Register 0x0E22
are the most significant and least significant bytes of the target address, respectively. Because the
previous register contains a data instruction, these two registers define a starting address (in this
case, 0x0500). The controller stores 0x0500 in the EEPROM and increments the EEPROM pointer by 2.
It then transfers 22 bytes from the register map (beginning at Address 0x0500) to the EEPROM and
increments the EEPROM address pointer by 23 (22 data bytes and one checksum byte). The
22 bytes transferred correspond to the clock distribution parameters in the register map.
The default value of this register is 0x80, which the controller interprets as an I/O update
instruction. The controller stores 0x80 in the EEPROM and increments the EEPROM address
pointer.
Table 119. EEPROM Storage Sequence for Reference Input Settings
Address
0x0E24
Bits
[7:0]
0x0E25
[7:0]
0x0E26
[7:0]
Bit Name
Reference inputs
Description
The default value of this register is 0x03, which the controller interprets as a data instruction. Its
decimal value is 3, so this tells the controller to transfer four bytes of data (3 + 1), beginning at the
address that is specified by the next two bytes. The controller stores 0x03 in the EEPROM and
increments the EEPROM address pointer.
The default value of these two registers is 0x0600. Note that Register 0x0E25 and Register 0x0E26
are the most significant and least significant bytes of the target address, respectively. Because the
previous register contains a data instruction, these two registers define a starting address (in this
case, 0x0600). The controller stores 0x0600 in the EEPROM and increments the EEPROM pointer by 2.
It then transfers four bytes from the register map (beginning at Address 0x0600) to the EEPROM
and increments the EEPROM address pointer by 5 (four data bytes and one checksum byte). The
four bytes that are transferred correspond to the reference inputs parameters in the register map.
Table 120. EEPROM Storage Sequence for Frame Sync Settings
Address
0x0E27
Bit
[7:0]
0x0E28
[7:0]
0x0E29
[7:0]
Bit Name
Frame sync
Description
The default value of this register is 0x01, which the controller interprets as a data instruction. Its
decimal value is 1, which tells the controller to transfer two bytes of data (1 + 1), beginning at the
address specified by the next two bytes. The controller stores 0x01 in the EEPROM and increments
the EEPROM address pointer.
The default value of these two registers is 0x0640. Note that Register 0x0E28 and Register 0x0E29
are the most significant and least significant bytes of the target address, respectively. Because the
previous register contains a data instruction, these two registers define a starting address (in this
case, 0x0640). The controller stores 0x0640 in the EEPROM and increments the EEPROM pointer by 2.
It then transfers two bytes from the register map (beginning at Address 0x0640) to the EEPROM
and increments the EEPROM address pointer by 3 (two data bytes and one checksum byte). The
two bytes transferred correspond to the reference inputs parameters in the register map.
Rev. B | Page 99 of 104
AD9558
Data Sheet
Table 121. EEPROM Storage Sequence for REFA Profile Settings
Address
0x0E2A
Bits
[7:0]
0x0E2B
[7:0]
0x0E2C
[7:0]
Bit Name
REFA Profile
Description
The default value of this register is 0x26, which the controller interprets as a data instruction. Its decimal
value is 38, so this tells the controller to transfer 39 bytes of data (38 + 1), beginning at the address that
is specified by the next two bytes. The controller stores 0x26 in the EEPROM and increments the EEPROM
address pointer.
The default value of these two registers is 0x0700. Note that Register 0x0E2B and Register 0x0E2C are
the most significant and least significant bytes of the target address, respectively. Because the previous
register contains a data instruction, these two registers define a starting address (in this case, 0x0700).
The controller stores 0x0700 in the EEPROM and increments the EEPROM pointer by 2. It then transfers
39 bytes from the register map (beginning at Address 0x0700) to the EEPROM and increments the
EEPROM address pointer by 40 (39 data bytes and one checksum byte). The 39 bytes transferred
correspond to the REFA profile parameters in the register map.
Table 122. EEPROM Storage Sequence for REFB Profile Settings
Address
0x0E2D
Bits
[7:0]
0x0E2E
[7:0]
0x0E2F
[7:0]
Bit Name
REFB profile
Description
The default value of this register is 0x26, which the controller interprets as a data instruction. Its decimal
value is 38, so this tells the controller to transfer 39 bytes of data (38 + 1), beginning at the address that is
specified by the next two bytes. The controller stores 0x26 in the EEPROM and increments the EEPROM
address pointer.
The default value of these two registers is 0x0740. Note that Register 0x0E2E and Register 0x0E2F are the
most significant and least significant bytes of the target address, respectively. Because the previous register
contains a data instruction, these two registers define a starting address (in this case, 0x0740). The controller
stores 0x0740 in the EEPROM and increments the EEPROM pointer by 2. It then transfers 39 bytes from the
register map (beginning at Address 0x0740) to the EEPROM and increments the EEPROM address pointer
by 40 (39 data bytes and one checksum byte). The 39 bytes transferred correspond to the REFB profile
parameters in the register map.
Table 123. EEPROM Storage Sequence for REFC Profile Settings
Address
0x0E30
Bits
[7:0]
0x0E31
[7:0]
0x0E32
[7:0]
Bit Name
REFC profile
Description
The default value of this register is 0x26, which the controller interprets as a data instruction. Its decimal
value is 38, so this tells the controller to transfer 39 bytes of data (38 + 1), beginning at the address that is
specified by the next two bytes. The controller stores 0x26 in the EEPROM and increments the EEPROM
address pointer.
The default value of these two registers is 0x0780. Note that Register 0x0E31 and Register 0x0E32 are
the most significant and least significant bytes of the target address, respectively. Because the previous
register contains a data instruction, these two registers define a starting address (in this case, 0x0780).
The controller stores 0x0780 in the EEPROM and increments the EEPROM pointer by 2. It then transfers
39 bytes from the register map (beginning at Address 0x0780) to the EEPROM and increments the
EEPROM address pointer by 40 (39 data bytes and one checksum byte). The 39 bytes transferred
correspond to the REFC profile parameters in the register map.
Table 124. EEPROM Storage Sequence for REFD Profile Settings
Address
0x0E33
Bits
[7:0]
0x0E34
[7:0]
0x0E35
[7:0]
Bit Name
REFD profile
Description
The default value of this register is 0x26, which the controller interprets as a data instruction. Its decimal
value is 38, so this tells the controller to transfer 39 bytes of data (38 + 1), beginning at the address that is
specified by the next two bytes. The controller stores 0x26 in the EEPROM and increments the EEPROM
address pointer.
The default value of these two registers is 0x07C0. Note that Register 0x0E34 and Register 0x0E35 are
the most significant and least significant bytes of the target address, respectively. Because the previous
register contains a data instruction, these two registers define a starting address (in this case, 0x07C0).
The controller stores 0x07C0 in the EEPROM and increments the EEPROM pointer by 2. It then transfers
39 bytes from the register map (beginning at Address 0x07C0) to the EEPROM and increments the EEPROM
address pointer by 40 (39 data bytes and one checksum byte). The 39 bytes transferred correspond to
the REFD profile parameters in the register map.
Rev. B | Page 100 of 104
Data Sheet
AD9558
Table 125. EEPROM Storage Sequence for Operational Control Settings
Address
0x0E37
Bits
[7:0]
0x0E38
[7:0]
0x0E39
[7:0]
Bit Name
Operational controls
Description
The default value of this register is 0x0D, which the controller interprets as a data instruction.
Its decimal value is 13, so this tells the controller to transfer 14 bytes of data (13 + 1), beginning
at the address specified by the next two bytes. The controller stores 0x0D in the EEPROM and
increments the EEPROM address pointer.
The default value of these two registers is 0x0A00. Note that Register 0x0E38 and Register 0x0E39
are the most significant and least significant bytes of the target address, respectively. Because
the previous register contains a data instruction, these two registers define a starting address
(in this case, 0x0A00). The controller stores 0x0A00 in the EEPROM and increments the EEPROM
pointer by 2. It then transfers 14 bytes from the register map (beginning at Address 0x0A00) to
the EEPROM and increments the EEPROM address pointer by 15 (14 data bytes and one checksum
byte). The 14 bytes transferred correspond to the operational controls parameters in the register map.
Table 126. EEPROM Storage Sequence for APLL Calibration
Address
0x0E3A
Bits
[7:0]
Bit Name
Calibrate APLL
0x0E3B
[7:0]
I/O update
Description
The default value of this register is 0xA0, which the controller interprets as a calibrate instruction. The
controller stores 0xA0 in the EEPROM and increments the EEPROM address pointer.
The default value of this register is 0x80, which the controller interprets as an I/O update instruction.
The controller stores 0x80 in the EEPROM and increments the EEPROM address pointer.
Table 127. EEPROM Storage Sequence for End of Data
Address
0x0E3C
Bits
[7:0]
Bit Name
End of data
Description
The default value of this register is 0xFF, which the controller interprets as an end instruction. The
controller stores this instruction in the EEPROM, resets the EEPROM address pointer, and enters
an idle state.
Note that if this is a pause rather than an end instruction, the controller actions are the same
except that the controller increments the EEPROM address pointer rather than resetting it.
Table 128. Available for Additional EEPROM Instructions
Address
0x0E3D
to 0xE45
Bits
[7:0]
Bit Name
Unused
Description
This area is available for additional EEPROM instructions.
Rev. B | Page 101 of 104
AD9558
Data Sheet
Table 129. Multifunction Pin Output Functions (D7 = 1)
Register Value
0x80
0x81
0x82
0x83
0x84
0x85
0x86
0x87
0x88
0x89
0x8A
0x8B
0x8C
0x8D
0x8E
0x8F
0x90
0x91
0x92
0x93
0x94
0x95
0x96
0x97
0x98
0x99
0x9A to 0x9F
0xA0
0xA1
0xA2
0xA3
0xA4 to Ax2F
0xB0
0xB1
0xB2
0xB3
0xB4 to 0xBF
0xC0
0xC1
0xC2
0xC3
0xC4 to 0xCF
0xD0
0xD1
0xD2 to 0FF
Output Function
Static Logic 0
Static Logic 1
System clock divided by 32
Watchdog timer output
EEPROM upload in progress
EEPROM download in progress
EEPROM fault detected
SYSCLK PLL lock detected
SYSCLK PLL stable
Output PLL locked
APLL calibration in process
APLL input reference present
All PLLs locked
(DPLL phase lock) and (APLL lock) and (sys PLL lock)
(DPLL phase lock) and (APLL lock)
Reserved
Reserved
DPLL free run
DPLL active
DPLL in holdover
DPLL in reference switchover
DPLL phase locked
DPLL frequency locked
DPLL phase slew limited
DPLL frequency clamped
Tuning word history available
Tuning word history updated
Reserved
Reference A fault
Reference B fault
Reference C fault
Reference D fault
Reserved
Reference A valid
Reference B valid
Reference C valid
Reference D valid
Reserved
Reference A active
Reference B active
Reference C active
Reference D active
Reserved
Clock distribution sync pulse
Soft pin configuration in progress
Reserved
Rev. B | Page 102 of 104
Equivalent Status Register
None
None
None
None
Register 0x0D00, Bit 0
Register 0x0D00, Bit 1
Register 0x0D00, Bit 2
Register 0x0D01, Bit 0
Register 0x0D01, Bit 1
Register 0x0D01, Bit 2
Register 0x0D01, Bit 3
Register 0x0D01, Bit 4
Register 0x0D01, Bit 5
Register 0x0D01, Bit 6
Register 0x0D08, Bit 0
Register 0x0D08, Bit 1
Register 0x0D08, Bit 2
Register 0x0D08, Bit 3
Register 0x0D08, Bit 4
Register 0x0D08, Bit 5
Register 0x0D08, Bit 6
Register 0x0D09, Bit 5
Register 0x0D09, Bit 4
Register 0x0D05, Bit 4
Register 0x0D0B, Bit 2
Register 0x0D0B, Bit 6
Register 0x0D0C, Bit 2
Register 0x0D0C, Bit 6
Register 0x0D0B, Bit 3
Register 0x0D0B, Bit 7
Register 0x0D0C, Bit 3
Register 0x0D0C, Bit 7
Register 0x0D09, Bits[1:0]
Register 0x0D09, Bits[1:0]
Register 0x0D09, Bits[1:0]
Register 0x0D09, Bits[1:0]
Register 0x0D03, Bit 3
Register 0x0D03, Bit 4
Data Sheet
AD9558
Table 130. Multifunction Pin Input Functions (D7 = 0)
Register Value
0x00
0x01
0x02
0x03
0x04
0x05
0x06 to 0x0E
0x10
0x11
0x12
0x13
0x14
0x15 to 0x1F
0x20
0x21
0x22
0x23
0x24 to 0x2F
0x30
0x31
0x32
0x33
0x34 to 0x3F
0x40
0x41
0x42
0x43
0x44
0x45
0x46
0x47
0x48 to 0x7F
Input Function
Reserved, high-Z input
I/O update
Full power-down
Clear watchdog
Clear all IRQs
Tuning word history reset
Reserved
User holdover
User free run
Reset incremental phase offset
Increment incremental phase offset
Decrement incremental phase offset
Reserved
Override Reference Monitor A
Override Reference Monitor B
Override Reference Monitor C
Override Reference Monitor D
Reserved
Force Validation Timeout A
Force Validation Timeout B
Force Validation Timeout C
Force Validation Timeout D
Reserved
Enable OUT0
Enable OUT1
Enable OUT2
Enable OUT3
Enable OUT4
Enable OUT5
Enable OUT0, OUT1, OUT2, OUT3, OUT4, OUT5
Soft sync clock distribution outputs
Reserved
Equivalent Control Register
Register 0x0005, Bit 0
Register 0x0A00, Bit 0
Register 0x0A03, Bit 0
Register 0x0A03, Bit 1
Register 0x0A03, Bit 2
Register 0x0A01, Bit 6
Register 0x0A01, Bit 5
Register 0x0A0A, Bit 2
Register 0x0A0A, Bit 0
Register 0x0A0A, Bit 1
Register 0x0A0C, Bit 0
Register 0x0A0C, Bit 1
Register 0x0A0C, Bit 2
Register 0x0A0C, Bit 3
Register 0x0A0B, Bit 0
Register 0x0A0B, Bit 1
Register 0x0A0B, Bit 2
Register 0x0A0B, Bit 3
Register 0x0501, Bit 0
Register 0x0505, Bit 0
Register 0x0506, Bit 0
Register 0x050A, Bit 0
Register 0x050B, Bit 0
Register 0x050F, Bit 0
Register 0x0501/0x0505/0x0506/0x050A/0x050B/0x050F, Bit 0
Register 0x0A02, Bit 1
Rev. B | Page 103 of 104
AD9558
Data Sheet
OUTLINE DIMENSIONS
9.10
9.00 SQ
8.90
0.30
0.23
0.18
0.60 MAX
0.60
MAX
64 1
49
48
PIN 1
INDICATOR
PIN 1
INDICATOR
8.85
8.75 SQ
8.65
0.50
BSC
0.50
0.40
0.30
33
32
0.25 MIN
7.50 REF
0.80 MAX
0.65 TYP
0.05 MAX
0.02 NOM
SEATING
PLANE
16
0.20 REF
FOR PROPER CONNECTION OF
THE EXPOSED PAD, REFER TO
THE PIN CONFIGURATION AND
FUNCTION DESCRIPTIONS
SECTION OF THIS DATA SHEET.
COMPLIANT TO JEDEC STANDARDS MO-220-VMMD-4
06-14-2012-A
12° MAX
17
BOTTOM VIEW
TOP VIEW
1.00
0.85
0.80
5.25
5.10 SQ
4.95
EXPOSED
PAD
Figure 58. 64-Lead Lead Frame Chip Scale Package [LFCSP_VQ]
9 mm × 9 mm Body, Very Thin Quad
(CP-64-5)
Dimensions shown in millimeters
ORDERING GUIDE
Model1
AD9558BCPZ
AD9558BCPZ-REEL7
AD9558/PCBZ
1
Temperature Range
−40°C to +85°C
−40°C to +85°C
−40°C to +85°C
Package Description
64-Lead Lead Frame Chip Scale Package [LFCSP_VQ]
64-Lead Lead Frame Chip Scale Package [LFCSP_VQ]
Evaluation Board
Z = RoHS Compliant Part.
I2C refers to a communications protocol originally developed by Philips Semiconductors (now NXP Semiconductors).
©2011–2013 Analog Devices, Inc. All rights reserved. Trademarks and
registered trademarks are the property of their respective owners.
D09758-0-6/13(B)
Rev. B | Page 104 of 104
Package Option
CP-64-5
CP-64-5
CP-64-5