AD AD9975ABST

a
FEATURES
Low Cost, 3.3 V-CMOS, Mixed Signal, Front End
Converter for Broadband Modems
10-Bit D/A Converter (TxDAC+®)
50 MSPS Input Word Rate
2ⴛ Interpolating Low-Pass Transmit Filter
100 MSPS DAC Output Update Rate
Wide (21 MHz) Transmit Bandwidth
Power-Down Modes
10-Bit, 50 MSPS A/D Converter
Fourth Order LPF with Selectable Cutoff Frequency
Dual Mode Programmable Gain Amplifier
Internal Clock Multiplier (PLL)
Two Auxiliary Clock Outputs
48-Lead LQFP Package
Broadband Modem
Mixed-Signal Front End
AD9975
FUNCTIONAL BLOCK DIAGRAM
AD9975
TXEN
10
TX–
RXCLK
CLK1
ADIO[9:0]
The AD9975 is a single-supply, broadband modem, mixed
signal, front end (MxFE™) IC. The device contains a transmit
path interpolation filter and DAC and a receive path PGA,
LPF, and ADC required for a variety of broadband modem
applications. Also on-chip is a PLL clock multiplier that provides all required clocks from a single crystal or clock input.
The TxDAC+ uses a digital 2× interpolation low-pass filter to
oversample the transmit data and ease the complexity of analog
reconstruction filtering. The transmit path bandwidth is 21 MHz
when sampled at 100 MSPS. The 10-bit DAC provides differential current outputs. The DAC full-scale current can be adjusted
from 2 to 20 mA by a single resistor, providing 20 dB of additional
gain range.
CLK2
CLK-GEN
OSCIN
XTAL
ADC
PGA
LPF
PGA
RX+
RX–
AGC[2:0]
SPORT
GENERAL DESCRIPTION
TxDAC+
TXCLK
10
APPLICATIONS
Powerline Networking
Home Phone Networking
TX+
10
K
RXEN
3
3
REGISTER
CONTROL
The filter cutoff frequency can also be tuned or bypassed where
filter requirements differ. The 10-bit ADC uses a multistage
differential pipeline architecture to achieve excellent dynamic
performance with low power consumption.
The digital transmit and receive ports are multiplexed onto a
10-bit databus and have individual TX/RX clocks and TX/RX
enable lines. This interface connects directly to Homelug 1.0
PHY/MAC chips from Intellon and Conexant.
The AD9975 is available in a space-saving 48-lead LQFP package. The device is specified over the commercial (–40°C to
+85°C) temperature range.
The receive path consists of a PGA, LPF, and ADC. The programmable gain amplifier (PGA) has two modes of operation. One
mode allows programming through the serial port and provides a
gain range from –6 dB to +36 dB in 2 dB steps. The other mode
allows the gain to be controlled through an asynchronous 3-pin
port and offers a gain range from 0 dB to 48 dB in 8 dB steps
with the use of an external gain stage. The receive path LPF
cutoff frequency can be selected to either 12 MHz or 26 MHz.
TxDAC+ is a registered trademark and MxFE is a trademark of Analog Devices, Inc.
REV. 0
Information furnished by Analog Devices is believed to be accurate and
reliable. However, no responsibility is assumed by Analog Devices for its
use, nor for any infringements of patents or other rights of third parties that
may result from its use. No license is granted by implication or otherwise
under any patent or patent rights of Analog Devices.
One Technology Way, P.O. Box 9106, Norwood, MA 02062-9106, U.S.A.
Tel: 781/329-4700
www.analog.com
Fax: 781/326-8703
© Analog Devices, Inc., 2002
AD9975–SPECIFICATIONS
(VS = 3.3 V ⴞ10%, FOSCIN = 50 MHz, FDAC = 100 MHz, Gain = –6 dB,
RSET = 4.02 k⍀, 100 ⍀ DAC Load.)
Parameter
Temp
Test
Level
OSC IN CHARACTERISTICS
Frequency Range
Duty Cycle
Input Capacitance
Input Impedance
Full
25°C
25°C
25°C
I
II
III
III
CLOCK OUTPUT CHARACTERISTICS
CLKA Jitter (FCLKA Derived from PLL)
CLKA Duty Cycle
25°C
25°C
II
III
14
50 ± 5
ps rms
%
Full
Full
Full
II
II
II
30
0.8
35
FDAC Cycles
dB
dB
Full
Full
Full
Full
25°C
25°C
25°C
25°C
25°C
25°C
II
II
II
II
II
II
III
II
III
III
25°C
25°C
25°C
II
III
III
25°C
25°C
III
III
N/A
Full
N/A
N/A
II
N/A
Full
Full
Full
Full
Full
TX CHARACTERISTICS
2× Interpolation Filter Characteristics
TX Path Latency, 2× Interpolation
Pass-Band Flatness 0 MHz to 20.7 MHz
Stop-Band Rejection @ 29.3 MHz
TxDAC
Resolution
Conversion Rate
Full-Scale Output Current
Voltage Compliance Range (TX+ or TX– AVSS)
Gain Error
Output Offset
Differential Nonlinearity
Integral Nonlinearity
Output Capacitance
Phase Noise @ 1 kHz Offset, 10 MHz Signal
Signal-to-Noise and Distortion (SINAD)
5 MHz Analog Out (20 MHz BW)
Wideband SFDR (to Nyquist, 50 MHz max)
5 MHz Analog Out
Narrowband SFDR (3 MHz Window)
5 MHz Analog Out
IMD (f1 = 6.25 MHz, f2 = 7.8125 MHz)
RX PATH CHARACTERISTICS (LFP Bypassed)
Resolution
Conversion Rate
Pipeline Delay, ADC Clock Cycles
Dynamic Performance (AIN = –0.5 dBFS, f = 5 MHz)
@ FOSCIN = 50 MHz, RX LPF Bypassed
Signal-to-Noise and Distortion Ratio (SINAD)
Effective Number of Bits (ENOB)
Signal-to-Noise Ratio (SNR)
Total Harmonic Distortion (THD)
Spurious-Free Dynamic Range (SFDR)
RX PATH GAIN/OFFSET
Minimum Programmable Gain
Maximum Programmable Gain
Narrow Band Rx LPF or Rx LPF Bypassed
Wideband Rx LPF
Gain Step Size
Gain Step Accuracy
Gain Range Error
Absolute Gain Error, PGA Gain = 0 dB
RX PATH INPUT CHARACTERISTICS
Input Voltage Range (Gain = –6 dB)
Input Capacitance
Differential Input Resistance
Input Bandwidth (–3 dB) (Rx LPF Bypassed)
Input Referred Noise (at +36 dB Gain with Filter)
Input Referred Noise (at –6 dB Gain with Filter)
Common-Mode Rejection
Min
10
40
Typ
50
3
100
Max
Unit
50
60
MHz
%
pF
MΩ
10
10
2
–0.5
–5.5
0
5
–100
Bits
MHz
mA
V
%FS
µA
LSB
LSB
pF
dBc/Hz
–60.6
dB
–76.2
dBc
–77.9
–77
dBc
dBFS
10
±2
2
0.5
100
20
+1.5
+5.0
TBD
1
5.5
Bits
MHz
Cycles
III
III
III
III
III
–56.6
9.1
–59.2
–60.1
–66
dB
Bits
dB
dB
dB
25°C
I
–6
dB
25°C
25°C
25°C
25°C
Full
Full
I
I
I
II
II
II
+36
+30
2
± 0.4
± 1.0
± 0.8
dB
dB
dB
dB
dB
dB
Full
25°C
25°C
25°C
25ºC
25ºC
25ºC
III
III
III
III
III
III
III
4
4
270
50
16
684
40
Vppd
pF
Ω
MHz
µV rms
µV rms
dB
–2–
10
10
50
REV. 0
AD9975
Parameter
Temp
Test
Level
RX PATH LPF (Low Cutoff Frequency)
Cutoff Frequency
Cutoff Frequency Variation
Attenuation @ 22 MHz
Pass-Band Ripple
Group Delay Variation
Settling Time (to 1% FS, Min to Max Gain Change)
Total Harmonic Distortion at Max Gain (THD)
Full
Full
Full
Full
Full
25°C
Full
III
III
III
II
II
II
I
12
±7
20
± 1.0
30
150
–61
MHz
%
dB
dB
ns
ns
dBc
RX PATH LPF (High Cutoff Frequency)
Cutoff Frequency
Cutoff Frequency Variation
Attenuation @ 35 MHz
Pass-Band Ripple
Group Delay Variation
Settling Time (to 1% FS, Min to Max Gain Change)
Total Harmonic Distortion at Max Gain (THD)
Full
Full
Full
Full
Full
25°C
Full
III
III
III
II
II
II
I
26
±7
20
± 1.2
15
80
–61
MHz
%
dB
dB
ns
ns
dBc
1
6
fADC /400
Cycle
dB/Octave
Hz
Min
RX PATH DIGITAL HPF
Latency (ADC Clock Source Cycles)
Roll-Off in Stop Band
–3 dB Frequency
Typ
Max
Unit
POWER-DOWN/DISABLE TIMING
Power-Down Delay (Active-to-Power-Down)
DAC
Interpolator
Power-Up Delay (Power-Down-to-Active)
DAC
PLL
ADC
PGA
LPF
Interpolator
Minimum RESET Pulsewidth Low (tRL)
25°C
25°C
II
II
200
200
ns
ns
25°C
25°C
25°C
25°C
25°C
25°C
Full
II
II
II
II
II
II
III
10
10
1000
1
1
200
5
µs
µs
µs
µs
µs
ns
fOSCIN Cycles
ADIO PORT INTERFACE
Maximum Input Word Rate
TX-Data Setup Time (tSU)
TX-Hold Hold Time (tHD)
RX-Data Valid Time(tVT)
RX-Data Hold Time (tHT)
25°C
25°C
25°C
25°C
25°C
I
II
II
II
II
REV. 0
–3–
100
3.0
0
3.0
1.5
MHz
ns
ns
ns
ns
AD9975
SPECIFICATIONS (continued)
Parameter
Temp
Test
Level
SERIAL CONTROL BUS
Maximum SCLK Frequency (fSCLK)
Clock Pulsewidth High (tPWH)
Clock Pulsewidth Low (tPWL)
Clock Rise/Fall Time
Data/Chip-Select Setup Time (tDS)
Data Hold Time (tDH)
Data Valid Time (tDV)
Full
Full
Full
Full
Full
Full
Full
II
II
II
II
II
II
II
25
18
18
CMOS LOGIC INPUTS
Logic “1” Voltage
Logic “0” Voltage
Logic “1” Current
Logic “0” Current
Input Capacitance
25°C
25°C
25°C
25°C
25°C
II
II
II
II
III
VDRVDD – 0.7
CMOS LOGIC OUTPUTS (1 mA Load)
Logic “1” Voltage
Logic “0” Voltage
Digital Output Rise/Fall Time
Full
25°C
Full
II
II
II
VDRVDD – 0.6
25°C
25°C
25°C
25°C
I
III
III
III
210
22.5
5.5
182
25°C
25°C
25°C
25°C
25°C
25°C
III
III
III
III
III
III
110
55
2
20
18
22
25°C
25°C
25°C
25°C
I
III
III
III
21
10
0
11
POWER SUPPLY
All Blocks Powered Up
IS_TOTAL (Total Supply Current)
Digital Supply Current (IDRVDD + IDVDD)
Clock Supply Current (ICLKVDD)
Analog Supply Current (IAVDD)
Power Consumption of Functional Blocks
Rx LPF
ADC and SPGA
Rx Reference
Interpolator
DAC
PLL-A
All Blocks Powered Down
IS_TOTAL (Total Supply Current)
Digital Supply Current (IDRVDD + IDVDD)
Clock Supply Current (ICLKVDD)
Analog Supply Current (IAVDD)
Min
Typ
Max
Unit
10
MHz
ns
ns
µs
ns
ns
ns
25
0
20
0.4
12
12
3
0.4
2.5
1.5
227
V
V
µA
µA
pF
V
V
ns
mA
mA
mA
mA
mA
mA
mA
mA
mA
27
mA
mA
mA
mA
Specifications subject to change without notice.
ABSOLUTE MAXIMUM RATINGS*
EXPLANATION OF TEST LEVELS
Power Supply (VS) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.9 V
Digital Output Current . . . . . . . . . . . . . . . . . . . . . . . . . . 5 mA
Digital Inputs . . . . . . . . . . . . . . . . –0.3 V to DRVDD + 0.3 V
Analog Inputs . . . . . . . . . . . . . . . . . –0.3 V to AVDD + 0.3 V
Operating Temperature . . . . . . . . . . . . . . . . . . –40°C to +85°C
Maximum Junction Temperature . . . . . . . . . . . . . . . . . 150°C
Storage Temperature . . . . . . . . . . . . . . . . . . –65°C to +150°C
Lead Temperature (Soldering 10 sec) . . . . . . . . . . . . . . 300°C
I. Devices are 100% production tested at 25°C and guaranteed
by design and characterization testing for the commercial
operating temperature range (–40°C to +85°C).
II. Parameter is guaranteed by design and/or characterization
testing.
III. Parameter is a typical value only.
THERMAL CHARACTERISTICS
Thermal Resistance
*Absolute Maximum Ratings are limiting values to be applied individually and
beyond which the serviceability of the circuit may be impaired. Functional
operability under any of these conditions is not necessarily implied. Exposure to
absolute maximum rating conditions for extended periods of time may affect
device reliability.
48-Lead LQFP
θJA = 57ºC/W
θJC = 28ºC/W
–4–
REV. 0
AD9975
ORDERING GUIDE
Model
Temperature Range
Package Description
Package Option
AD9975ABST
AD9975ABSTEB
AD9975ABSTRL
–40ºC to +85ºC
–40ºC to +85ºC
–40ºC to +85ºC
48-Lead LQFP
AD9975 EVAL Board
AD9975ABST Reel
ST-48
CAUTION
ESD (electrostatic discharge) sensitive device. Electrostatic charges as high as 4000 V readily
accumulate on the human body and test equipment and can discharge without detection. Although the
AD9975 features proprietary ESD protection circuitry, permanent damage may occur on devices
subjected to high energy electrostatic discharges. Therefore, proper ESD precautions are recommended
to avoid performance degradation or loss of functionality.
35
36
37
40, 41
44
45
48
REV. 0
37 RESET
38 AVDD
39 AVSS
40 REFB
41 REFT
42 AVSS
43 AVSS
44 RX+
45 RX–
36 DRVSS
35 DRVDD
SCLK 3
34 RXBOOST/SDO
SDATA 4
AD9975
AVSS 6
48-PIN LQFP
TOP VIEW
(Not to Scale)
TX+ 7
TX– 8
–5–
32 RXCLK
31 TXCLK
30 TXEN
29 RXEN
ADIO4 24
25 ADIO3
ADIO5 23
26 ADIO2
CLKVDD 12
ADIO6 22
27 ADIO1
REFIO 11
ADIO7 21
28 ADIO0
FS ADJ 10
ADIO8 20
AVSS 9
ADIO9 19
Transmit DAC + Output
Transmit DAC – Output
DAC Full-Scale Output Current
Adjust with External Resistor
REFIO
DAC Band Gap Decoupling Node
CLKVDD
Power Supply for CLKOUT1
DVSS
Digital Ground
DVDD
Digital 3.3 V Power Supply
AGC[2:0]
AGC Control Inputs
CLKOUT1
Auxiliary Clock Output
ADIO[9:0]
Digital Data I/O Port
RXEN
ADIO Direction Control Input
TXEN
TX Path Enable
TXCLK
ADIO Sample Clock Input
RXCLK
ADIO Request Clock Input
CLKOUT2
Auxiliary Clock Output
RXBOOST/
External Gain Control Output/
SDO
Serial Data Output
DRVDD
Digital I/O 3.3 V Power Supply
DRVSS
Digital I/O Ground
RESET
Reset Input
REFB, REFT ADC Reference Decoupling Node
Rx+
Receive Path + Input
Rx–
Receive Path – Input
XTAL
Crystal Oscillator Inverter Output
CLKOUT1 18
Tx+
Tx–
FS ADJ
33 CLKOUT2
AVDD 5
DVSS 13
11
12
13
14
15–17
18
19–28
29
30
31
32
33
34
OSC IN 1
SENABLE 2
AGC0 17
Crystal Oscillator Inverter Input
Serial Bus Enable Input
Serial Bus Clock Input
Serial Bus Data I/O
Analog 3.3 V Power Supply
Analog Ground
46 AVSS
OSC IN
SENABLE
SCLK
SDATA
AVDD
AVSS
AGC1 16
1
2
3
4
5, 38, 47
6, 9, 39, 42,
43, 46
7
8
10
47 AVDD
Function
AGC2 15
Mnemonic
DVDD 14
Pin No.
ESD SENSITIVE DEVICE
PIN CONFIGURATION
48 XTAL
PIN FUNCTION DESCRIPTION
WARNING!
AD9975
DEFINITIONS OF SPECIFICATIONS
Clock Jitter
Offset Error
First transition should occur for an analog value 1/2 LSB above
negative full scale. Offset error is defined as the deviation of the
actual transition from that point.
The clock jitter is a measure of the intrinsic jitter of the PLL
generated clocks. It is a measure of the jitter from one rising
edge of the clock with respect to another edge of the clock nine
cycles later.
Gain Error
An ideal converter exhibits code transitions that are exactly 1 LSB
apart. DNL is the deviation from this ideal value. Guaranteed no
missing codes to 10-bit resolution indicate that all 1024 codes,
respectively, must be present over all operating ranges.
The first code transition should occur at an analog value 1/2 LSB
above negative full scale. The last transition should occur for an
analog value 1 1/2 LSB below the nominal full scale. Gain error
is the deviation of the actual difference between first and last
code transitions and the ideal difference between first and last
code transitions.
Integral Nonlinearity Error (INL)
Input Referred Noise
Linearity error refers to the deviation of each individual code from
a line drawn from “negative full scale” through “positive full
scale.” The point used as negative full scale occurs 1/2 LSB
before the first code transition. Positive full scale is defined as a
level 1 1/2 LSB beyond the last code transition. The deviation is
measured from the middle of each particular code to the true
straight line.
The rms output noise is measured using histogram techniques.
The ADC output code’s standard deviation is calculated in LSB
and converted to an equivalent voltage. This results in a
noise figure that can directly be referred to the RX input of
the AD9975.
Differential Nonlinearity Error (DNL, No Missing Codes)
Signal-to-Noise and Distortion Ratio (SINAD)
SINAD is the ratio of the rms value of the measured input signal
to the rms sum of all other spectral components below the Nyquist
frequency, including harmonics but excluding dc. The value for
SINAD is expressed in decibels.
Phase Noise
Single-sideband phase noise power density is specified relative
to the carrier (dBc/Hz) at a given frequency offset (1 kHz) from
the carrier. Phase noise can be measured directly on a generated
single tone with a spectrum analyzer that supports noise marker
measurements. It detects the relative power between the carrier
and the offset (1 kHz) sideband noise and takes the resolution
bandwidth (rbw) into account by subtracting 10 log(rbw). It also
adds a correction factor that compensates for the implementation
of the resolution bandwidth, log display, and detector characteristic.
Effective Number of Bits (ENOB)
For a sine wave, SINAD can be expressed in terms of the number
of bits. Using the following formula,
N = ( SINAD – 1.76 ) dB / 6.02
it is possible to get a measure of performance expressed as N,
the effective number of bits.
Output Compliance Range
Signal-to-Noise Ratio (SNR)
The range of allowable voltage at the output of a current-output
DAC. Operation beyond the maximum compliance limits may
cause either output stage saturation, resulting in nonlinear performance, or breakdown.
SNR is the ratio of the rms value of the measured input signal to
the rms sum of all other spectral components below the Nyquist
frequency, excluding harmonics and dc. The value for SNR is
expressed in decibels.
Spurious-Free Dynamic Range (SFDR)
Total Harmonic Distortion (THD)
The difference, in dB, between the rms amplitude of the DAC’s
output signal (or ADC’s input signal) and the peak spurious
signal over the specified bandwidth (Nyquist bandwidth, unless
otherwise noted).
THD is the ratio of the rms sum of the first six harmonic components to the rms value of the measured input signal and is
expressed as a percentage or in decibels.
Pipeline Delay (Latency)
Power supply rejection specifies the converter’s maximum
full-scale change when the supplies are varied from nominal to
minimum and maximum specified voltages.
Power Supply Rejection
The number of clock cycles between conversion initiation and
the associated output data being made available.
–6–
REV. 0
Typical Performance Characteristics–AD9975
10
0
INTERPOLATION FILTER
0
–10
–10
–20
MAGNITUDE – dBc
MAGNITUDE – dB
–20
INCLUDING
SIN(X)/X
–30
–40
–50
–60
–70
–30
–40
–50
–60
–70
–80
–80
–90
–90
–100
–100
0
0.1
0.2
0.3
0.4
0.5
0.6
FSAMPLE
0.7
0.8
0.9
0
1.0
5
10
15
20
25
30
35
40
45
FREQUENCY – MHz
TPC 1. 2 × Low-Pass Interpolation Filter
TPC 4. Single Tone Spectral Plot @ fDATA = 50 MSPS,
fOUT = 11 MHz, 2 × LPF
10
10
0
–10
–10
INCLUDING
SIN(X)/X
–20
–20
–30
–30
–40
–40
dBc
MAGNITUDE – dB
INTERPOLATION FILTER
0
–50
–60
–60
–70
–70
–80
–80
–90
–90
–100
10.0
–100
0
0.1
0.2
0.3
0.4
0.6
0.5
FS – MHz
0.7
0.8
0.9
1.0
TPC 2. 2 × Band-Pass Interpolation Filter, FS /2
Modulation, Adjacent Image Preserved
0
–70
–10
–72
–20
–74
–30
–40
–50
–60
–70
10.4
10.6
10.8 11.0 11.2 11.4
FREQUENCY – MHz
11.6
11.8
12.0
–76
–80
–82
–84
–86
–90
–88
–100
THIRD ORDER
HARMONIC
–78
–80
SECOND ORDER
HARMONIC
–90
0
5
10
15
20
25
30
35
40
45
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
FREQUENCY – MHz
FREQUENCY – MHz
TPC 3. Single Tone Spectral Plot @ fDATA = 50 MSPS,
fOUT = 5 MHz, 2 × LPF
REV. 0
1.02
TPC 5. Dual Tone Spectral Plot @ fDATA = 50 MSPS,
fOUT = 6.7 MHz and 7.3 MHz, 2 × LPF
MAGNITUDE – dBc
MAGNITUDE – dBc
–50
TPC 6. Harmonic Distortion vs. fOUT @fDATA = 50 MSPS
–7–
AD9975
10
35
0
–10
FREQUENCY – MHz
–20
dBc
–30
–40
–50
–60
30
25
–70
–80
–90
20
192
–100
10.003 10.004 10.005 10.006 10.007 10.008 10.009 10.010 10.011 10.012 10.013
208
FREQUENCY – MHz
TPC 7. Phase Noise Plot @fDATA 50 MSPS,
fOUT = 10 MHz, 2 × LPF
224
240
TUNING TARGET – Decimal
255
TPC 10. FC vs. Tuning Target, FADC = 50 MHz,
LPF = Wideband Rx Filter
10
20
0
18
FREQUENCY – MHz
–10
dBc
–20
–30
–40
16
14
–50
12
–60
10
192
–70
3
5
7
9
11
13
15
17
FREQUENCY – MHz
19
21
23
208
224
240
TUNING TARGET – Decimal
255
TPC 11. FC vs. Tuning Target, FADC = 50 MHz,
LPF = Narrowband Rx Filter
TPC 8. “In-Band” Multitone Spectral Plot
@fDATA = 50 MSPS, fOUT = k × 195 kHz, 2 × LPF
10
0.60
0
0.40
GAIN ACCURACY – dB
–10
dBc
–20
–30
–40
–50
0.20
0
–0.20
–0.40
–60
–0.60
–70
1
11
21
31
41
51
61
71
FREQUENCY – MHz
81
91
101
–0.80
–6 –4 –2 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36
VGA GAIN – dB
TPC 9. “Wide-Band” Multitone Spectral Plot
@fDATA = 50 MSPS, fOUT = k × 195 kHz, 2 × LPF
TPC 12. PGA Gain Error vs. Gain
–8–
REV. 0
AD9975
2.5
5dB/Div
LOG MAG
REF 0dB
–3.0dB
24.1MHz
2.4
0
2.3
OFFSET – LSBs
2.2
2.1
2.0
1.9
1.8
1.7
1.6
1.5
–6 –4 –2 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36
VGA GAIN – dB
1MHz
TPC 13. PGA Gain Step vs. Gain
LOG MAG
5dB/Div REF 0dB
10MHz
100MHz
TPC 16. Rx LPF Group Delay, LPF = Narrowband
Rx Filter, FADC = 50 MHz, Tuning Target = 255
5 ns/Div
DELAY
–3.0dB
REF 0s
33.9ns
22.1MHz
12.1MHz
0
0
1MHz
10MHz
1MHz
100MHz
10ns/Div
REF 0s
100MHz
TPC 17. Rx LPF Group Delay, LPF = Wideband Rx
Filter, FADC = 50 MHz, Tuning Target = 255
TPC 14. Rx LPF Frequency Response, LPF =
Narrowband Rx Filter, FADC = 50 MHz,
Tuning Target = 255
DELAY
10MHz
70.1ns
5dB/Div
LOG MAG
9.9MHz
REF 0dB
–3.00dB
123.5 kHz
0
0
1MHz
10MHz
100MHz
10kHz
TPC 15. Rx LPF Frequency Response, LPF =
Wideband Rx Filter, FADC = 50 MHz,
Tuning Target = 255
REV. 0
100kHz
1MHz
TPC 18. Rx HPF Frequency Response, FADC = 50 MHz
–9–
AD9975
60
4000
3800
59
MAGNITUDE – dB
ADC OUTPUT CODE
3600
3400
3200
3000
58
57
2800
56
2600
55
2400
1
5
9
13
17
21
25
29
ADC CLOCK CYCLES
33
20
37
TPC 19. Rx Path Settling, 1/2 Scale Rising Step with
Gain Change, LPF FC = 26 MHz, FADC = 50 MHz
25
30
35
FS – MHz
45
40
50
TPC 22. Rx Path SNR vs. FADC, FIN = 5 MHz,
Gain = –6 dB, Rx LPF Bypassed
–55
4000
3800
MAGNITUDE – dB
ADC OUTPUT CODE
3600
3400
3200
3000
–60
2800
2600
–65
2400
1
5
9
13
17
21
25
29
ADC CLOCK CYCLES
33
20
37
TPC 20. Rx Path Settling, 1/2 Scale Falling Step with
Gain Change, LPF FC = 26 MHz, FADC = 50 MHz
25
30
35
FS – MHz
45
40
50
TPC 23. Rx Path THD vs. FADC, FIN = 5 MHz,
Gain = –6 dB, Rx LPF Bypassed
9.5
9.50
9.3
9.25
ENOBs
ENOBs
9.1
9.00
8.9
8.75
8.7
8.5
8.50
20
25
30
35
FS – MHz
40
45
0
50
2
4
6
8
10
12
FIN – MHz
14
16
18
20
TPC 24. Rx Path ENOB vs. FIN, FADC = 50 MHz,
Gain = –6 dB, Rx LPF Bypassed
TPC 21. Rx Path ENOB vs. FADC, FIN = 5 MHz,
Gain = –6 dB, Rx LPF Bypassed
–10–
REV. 0
AD9975
–50
58
57
–55
MAGNITUDE – dB
MAGNITUDE – dB
56
55
54
53
–60
–65
52
–70
51
50
–75
0
2
4
6
8
10
12
FIN – MHz
14
16
18
20
0
TPC 25. Rx Path SNR vs. FIN, FADC = 50 MHz,
Gain = –6 dB, Rx LPF Bypassed
REV. 0
2
4
6
8
10
12
FIN – MHz
14
16
18
20
TPC 26. Rx Path THD vs. FIN, FADC = 50 MHz,
Gain = –6 dB, Rx LPF Bypassed
–11–
AD9975
The AD9975 transmit path consists of a digital interface port,
a bypassable 2× interpolation filter, and a transmit DAC. The
clock signals required by these blocks are generated by the internal PLL. The block diagram below shows the interconnection
between the major functional components of the transmit path.
of L is programmed through the serial interface port and can be
set to 1, 2, 4, or 8. The transmit path expects a new input sample
at the ADIO interface at a rate of fDAC/2 if the interpolation
filter is being used. If the interpolation filter is bypassed, the
transmit path expects a new input sample at the ADIO interface
at a rate of fDAC.
INTERPOLATION FILTER
D/A CONVERTER
The interpolation filter can be programmed to run at a 2×
upsampling ratio in either a low-pass filter or band-pass filter
mode. The transfer functions of these two modes are shown in
TPC 1 and TPC 2, respectively. The y-axes of the figures show
the magnitude response of the filters in dB, and the x-axes show
the frequency normalized to FDAC. The top trace of the plot shows
the discrete time transfer function of the interpolation filter. The
bottom trace shows the TX path transfer function including the
sin(x)/x transfer function of the DAC. In addition to the two
upsampling modes, the interpolation filter can be programmed
into a pass-through mode if no interpolation filtering is desired.
The AD9975 DAC provides differential output current on the
TX+ and TX– pins. The values of the output currents are complementary, meaning they will always sum to IFS, the full-scale
current of the DAC. For example, when the current from TX+
is at full scale, the current from TX– is zero. The two currents
will typically drive a resistive load that will convert the output
currents to a voltage. The TX+ and TX– output currents are
inherently ground seeking and should each be connected to
matching resistors, RL, that are tied directly to AGND.
TRANSMIT PATH
The table below shows the following parameters as a function of
the mode in which it is programmed.
Latency – The number of clock cycles from the time a digital
impulse is written to the DAC until the peak value is output at
the TX+ and TX– Pins.
The full-scale output current of the DAC is set by the value of
the resistor placed from the FS ADJ pin to AGND. The relationship between the resistor, RSET, and the full-scale output
current is governed by the following equation:
IFS = 39.4 / RSET
Flush – The number of clock cycles from the time a digital
impulse is written to the DAC until the output at the TX+ and
TX– Pins settles to zero.
The full-scale current can be set from 2 to 20 mA. Generally, there
is a trade-off between DAC performance and power consumption.
The best DAC performance will be realized at an IFS of 20 mA.
However, the value of IFS adds directly to the overall current
consumption of the device.
FPASS – The frequency band over which the pass-band ripple is
less than the stated magnitude (i.e., 0.1 dB or 1.0 dB).
The single-ended voltage outputs appearing at the TX+ and
TX– nodes are:
FSTOP – The frequency band over which the stop-band attenuation
is greater than the stated magnitude (i.e., 40 dB or 50 dB).
VTX + = ITX + × RL
VTX – = ITX – × RL
Table I. Interpolation Filters vs. Mode
Note that the full-scale voltage of VTX+ and VTX– should not
exceed the maximum output compliance range of 1.5 V to prevent signal compression. To maintain optimum distortion and
linearity performance, the maximum voltages at VTX+ and VTX–
should not exceed ± 0.5 V.
Register 7[7:4]
0x1
0x5
Mode
Latency, FDAC Clock Cycles
Flush, FDAC Clock Cycles
FPASS, 0.1 dB
FPASS, 1.0 dB
FSTOP, 40 dB
FSTOP, 50 dB
2× LPF
30
48
<0.204
<0.207
<0.296
<0.302
2× BPF, Adj. Image
30
48
>0.296, <0.704
>0.293, <0.707
>0.204, <0.796
<0.198, >0.802
DPLL-A CLOCK DISTRIBUTION
Figure 1 shows the clock signals used in the transmit path. The
DAC sampling clock, fDAC, is generated by DPLL-A. fDAC has a
frequency equal to L × fOSCIN, where L is the PLL clock multiplier
value and fOSCIN is the frequency of the input to PLL-A. The value
The single-ended full-scale voltage at either output node will be:
VFS = IFS × RL
The differential voltage, VDIFF, appearing across VTX+ and VTX– is:
VDIFF = ( ITX + – ITX – ) × RL
and
VDIFF _ FS = IFS × RL
It should be noted that the differential output impedance of the
DAC is 2 × RL and any load connected across the two output
resistors will load down the output voltage accordingly.
–12–
REV. 0
AD9975
There are two pass band settings for the LPF. Within each pass
band, the filters are tunable over about a ± 15% frequency range.
The formula for the cutoff frequency is:
RECEIVE PATH DESCRIPTION
The receive path consists of a two stage PGA, a continuous
time, 4-pole LPF, an ADC, and a digital HPF. Also working
in conjunction with the receive path is an offset correction
circuit and a digital phase-locked loop. Each of these blocks
will be discussed in detail in the following sections.
FC = FADC × 64 / (64 + Target )
Where Target is the decimal value programmed as the tuning
target in Register 5.
PROGRAMMABLE GAIN AMPLIFIER
The PGA has a programmable gain range from –6 dB to +36 dB
if the narrower (approximately 12 MHz) LPF bandwidth is selected,
or if the LPF is bypassed. If the wider (approximately 29 MHz)
LPF bandwidth is selected, the gain range is –6 dB to +30 dB.
The PGA is comprised of two sections, a continuous time PGA
(CPGA), and a switched capacitor PGA (SPGA). The CPGA
has possible gain settings of 0, 6, 12, 18, 24 and 30. The SPGA
has possible gain settings of –6 dB, –4 dB, –2 dB, 0 dB, +2 dB,
+4 dB, and +6 dB. Table II shows how the gain is distributed
for each programmed gain setting.
The CPGA input appears at the device RX+ and RX– input pins.
The input impedance of this stage is nominally 270 Ω differential
and is not gain dependent. It is best to ac-couple the input signal
to this stage and let the inputs self-bias. This will lower the
offset voltage of the input signal, which is important at higher
gains, since any offset will lower the output compliance range of
the CPGA output. When the inputs are driven by direct coupling,
the dc level should be AVDD/2. However, this could lead to
larger dc offsets and reduce the dynamic range of the RX path.
There are two modes for selecting the RX path gain. The first
mode is to program the PGA through the serial port. A 5-bit
word determines the gain with a resolution of 2 dB per step.
More detailed information about this mode is included in the
Register Programming Definitions section of this data sheet.
This filter may also be bypassed. In this case, the bandwidth of
the RX path will be gain dependent and will be around 50 MHz
at the highest gain settings.
ADC
The AD9975’s analog-to-digital converter implements pipelined
multistage architecture to achieve high sample rates while consuming low power. The ADC distributes the conversion over several
smaller A/D subblocks, refining the conversion with progressively
higher accuracy as it passes the results from stage to stage. As a
consequence of the distributed conversion, ADCs require a small
fraction of the 2n comparators used in a traditional n-bit flashtype A/D. A sample-and-hold function within each of the stages
permits the first stage to operate on a new input sample while the
remaining stages operate on preceding samples. Each stage of
the pipeline, excluding the last, consists of a low resolution
Flash A/D connected to a switched capacitor DAC and interstage
residue amplifier (MDAC). The residue amplifier amplifies the
difference between the reconstructed DAC output and the flash
input for the next stage in the pipeline. One bit of redundancy is
used in each one of the stages to facilitate digital correction of
flash errors. The last stage simply consists of a Flash A/D.
SHA
GAIN
SHA
GAIN
AINP
A/D
The second mode sets the gain through the asynchronous
AGC[2:0] pins. These three pins set the PGA gain and state of
the RXBOOST pin according to Table II.
AINN
A/D
Table II. AGC[2:0] Gain Mapping
A/D
D/A
CORRECTION LOGIC
AGC
[2:0]
Rx
Path
Gain
CPGA
Gain
SPGA
Gain
RXBOOST
0x00
0x01
0x02
0x03
0x04
0x05
0x06
0x07
–6
–6
2
10
2
10
18
26
–6
–6
–6
0
–6
0
12
18
0
0
8
10
8
10
6
8
0
0
0
0
1
1
1
1
Figure 1. ADC Theory of Operation
LOW-PASS FILTER
The low-pass filter (LPF) is a programmable, three-stage, fourth
order low-pass filter. The first real pole is implemented within
the CPGA. The second filter stage implements a complex pair
of poles. The last real pole is implemented in a buffer stage that
drives the SPGA.
REV. 0
D/A
The digital data outputs of the ADC are represented in straight
binary format. They saturate to full scale or zero when the input
signal exceeds the input voltage range.
The maximum value will be output from the ADC when the
RX+ input is 1 V or more greater than the RX– input. The minimum value will be output from the ADC when the RX– input is
1 V or more greater than the RX+ input. This results in a full-scale
ADC voltage of 2 Vppd.
The data can be translated to straight binary data format by
simply inverting the most significant bit.
The timing of the interface is fully described in the Digital Interface Port Timing section.
–13–
AD9975
DIGITAL HPF
AGC PROGRAMMING
Following the ADC, there is a bypassable digital HPF. The
response is a single pole IIR HPF. The transfer function is
approximately:
The gain in the receive path can be programmed in two ways.
The default method is through the AGC[2:0] pins. In this mode,
the gain is achieved using a combination of internal and external
gain. The external gain is controlled by the RXBOOST output
pin, which is determined by the decode of the 3-bit AGC gain value.
H ( z ) = ( Z – 0.99994 ) / ( Z – 0.98466 )
where the sampling period is equal to the ADC clock period.
This results in a 3 dB frequency approximately 1/400th of the
ADC sampling rate. The transfer function of the digital HPF with
an ADC sample rate of 50 MSPS is plotted in TPC 23.
DIGITAL INTERFACE PORT OPERATION
The digital HPF introduces a 1 ADC clock cycle latency. If the
HPF function is not desired, the HPF can be bypassed and the
latency will not be incurred.
CLOCK AND OSCILLATOR CIRCUITRY
The AD9975 generates all internally required clocks from a single
clock source. This source can be supplied in one of two ways.
The first method uses the on-chip oscillator by connecting a
fundamental frequency quartz crystal between the OSC IN (Pin 1)
and XTAL (Pin 48) with parallel resonant load capacitors as
specified by the crystal manufacturer. Alternatively, a TTL-level
clock applied to OCS IN with the XTAL pin left unconnected
can overdrive the internal oscillator circuit.
The PLL has a frequency capture range between 10 MHz and
50 MHz.
AGC TIMING CONSIDERATIONS
When implementing the AGC timing loop, it is important to
consider the delay and settling time of the RX path in response
to a change in gain. Figure 2 shows the delay the receive signal
experiences through the blocks of the RX path. Whether the gain
is programmed through the serial port or via the AGC[2:0] pins,
the gain takes effect immediately with the delays shown in Figure 2.
When gain changes do not involve the CPGA, the new gain will
be evident in samples after about 7 ADC clock cycles. When the
gain change does involve the CPGA, it takes an additional 45 ns
to 70 ns due to the propagation delays of the buffer, LPF and PGA.
Table VI in the Register Programming section details the PGA
programming map.
GAIN REGISTER
5ns
DECODE LOGIC
The digital interface port is a 10-bit bidirectional bus shared in
burst fashion between the transmit path and receive path. The
MxFE acts as a slave to the digital ASIC, accepting two input
enable signals, TXEN and RXEN, as well as two input clock
signals, TXCLK and RXCLK. Because the sampling clocks for
the DAC and ADC are derived internally from the OSC IN
signal, it is required that the TXCLK and RXCLK signals are
exactly the same frequency as the OSC IN signal. The phase
relationships between the TXCLK, RXCLK, and OSC IN signal
are arbitrary.
In order to add flexibility to the digital interface port, there are
several programming options available. The data input format is
straight binary by default. It is possible to independently change
the data format of the transmit path and receive path to twos
complement. Also, the clock timing can be independently changed
on the transmit and receive paths by selecting either the rising
or falling clock edge as the validating/sampling edge of the clock.
The digital interface port can also be programmed into a threestate output mode allowing it to be connected onto a shared bus.
The timing of the interface is fully described in the Digital Interface Port Timing section.
CLOCK DISTRIBUTION
The DAC sampling clock, fDAC, is generated by the internal
digital phase-locked loop (DPLL). fDAC has a frequency equal to
L × fOSCIN, where fOSCIN is the internal signal generated either by
the crystal oscillator when a crystal is connected between the
OSC IN and XTAL pins or by the clock that is fed into the
OSC IN pin, and L is the multiplier programmed through the
serial port. L can have the values of 1, 2, 4, or 8.
When the interpolation filter is enabled (either 2× LPF or 2× BPF
is selected), the data rate is upsampled by a factor of two. In this
case, the transmit path expects a new data input word at the rate
of fDAC/2. When the interpolation filter is bypassed, the transmit
path expects a new input word at the same frequency as DAC
sampling clock, fDAC. Therefore, in terms of fOSCIN, the TXCLK
frequency should be:
fTXCLK = L × fOSCIN / K
DIGITAL HPF
1 CLK CYCLE
A/D
SHA
5 CLK CYCLE 1/2 CLK CYCLE
BUFFER
LPF
PGA
10ns
25ns OR 50ns
10ns
Figure 2. AGC Loop Timing
where K is the interpolation factor. The interpolation factor, K, is
equal to 2 when the interpolator is enabled and is equal to 1 when
the interpolator is bypassed.
The ADC sampling clock is derived from fOSCIN and a new
output sample is available every fOSCIN clock cycle. The ADC
sampling lock can be programmed to be equal to fOSCIN if desired.
The timing of the digital interface port is illustrated in the
Figures 3 and 4.
–14–
REV. 0
AD9975
DIGITAL INTERFACE PORT TIMING
Table III. Instruction Byte Bit Definitions
The ADIO[9:0] bus accepts input data-words into the transmit
path when the TXEN pin is high, the RXEN pin is low, and a
clock is present on the TXCLK pin. Figure 3 illustrates the
transmit path input timing.
MSB
LSB
I7
I6
I5
I4
I3
I2
I1
I0
R/W
N1
N0
A4
A3
A2
A1
A0
tSU
Bit I7 – R/W
This bit determines whether a read or a write data transfer will
occur after the instruction byte write. Logic high indicates a read
operation, and Logic 0 indicates a write operation.
TXCLK
TXEN
tHD
ADIO[9:0]
TX0
TX1
TX2
TX3
TX4
Bits I6:I5 – N1:N0
These two bits determine the number of bytes to be transferred
during the data transfer cycle. The bit decodes are shown in
Table IV.
RXEN
Figure 3. Transmit Data Input Timing Diagram
It should be noted that to clear the transmit path input buffers,
an additional six clock cycles on the TXCLK input are required
after TXEN goes low. The interpolation filters will be “flushed”
with zeros if the clock signal into the TXCLK pin is present for
48 clock cycles after TXEN goes low (the data on the ADIO
bus being irrelevant over this interval).
The output from the receive path will be driven onto the
ADIO[9:0] bus when the RXEN pin is high, and a clock is present
on the RXCLK pin. When both TXEN and RXEN are low, the
ADIO[9:0] bus is three-stated. Figure 4 illustrates the receive
path output timing.
N1:N0
Description
0:0
0:1
1:0
1:1
Transfer 1 Byte
Transfer 2 Bytes
Transfer 3 Bytes
Transfer 4 Bytes
Bits I4:I0 - A4:A0
These bits determine which register is accessed during the data
transfer portion of the communications cycle. For multibyte
transfers, this address is the starting byte address. The remaining
register addresses are generated by the AD9975.
tVT
RXCLK
Serial Interface Port Pin Description
SCLK—Serial Clock
The serial clock pin is used to synchronize data transfers to and
from the AD9975 and to run the internal state machines. SCLK
maximum frequency is 25 MHz. All data transmitted to the
AD9975 is sampled on the rising edge of SCLK. All data read
from the AD9975 is validated on the rising edge of SCLK and is
updated on the falling edge.
tHT
RXEN
ADIO[9:0]
Table IV. N1:N0 Bit Map
RX0
RX1
RX2
RX3
Figure 4. Receive Data Output Timing Diagram
SERIAL INTERFACE FOR REGISTER CONTROL
The serial port is a 3-wire serial communications port consisting
of a clock (SCLK), chip select (SENABLE), and a bidirectional
data (SDATA) signal. The interface allows read/write access to
all registers that configure the AD9975 internal parameters.
Single or multiple byte transfers are supported as well as MSB
first or LSB first transfer formats.
General Operation of the Serial Interface
Serial communication over the serial interface can be from 1 byte
to 5 bytes in length. The first byte is always the instruction byte.
The instruction byte establishes whether the communication is
going to be a read or write access, the number of data bytes to be
transferred, and the address of the first register to be accessed.
The instruction byte transfer is complete immediately upon the
eighth rising edge of SCLK after SENABLE is asserted. Likewise,
the data registers change immediately upon writing to the eighth
bit of each data byte.
SENABLE—Serial Interface Enable
The SENABLE Pin is active low. It enables the serial communication to the device. SENABLE select should stay low during
the entire communication cycle. All input on the serial port is
ignored when SENABLE is inactive.
SDATA—Serial Data I/O
The signal on this line is sampled on the first eight rising edges
of SCLK after SENABLE goes active. Data is then read from or
written to the AD9975 depending on what was read.
Figures 5 and 6 show the timing relationships between the three
SPI signals.
tDS
tSCLK
SENABLE
tPWH
tPWL
SCLK
Instruction Byte
tHT
tDS
The instruction byte contains the information shown in Table III.
SDATA
INSTRUCTION BIT 7
INSTRUCTION BIT 6
Figure 5. Timing Diagram Register Write to AD9975
REV. 0
–15–
AD9975
Figures 7a and 7b show how the serial port words are built for
each of these modes.
SCLK
tDV
SDATA
INSTRUCTION BIT n
INSTRUCTION BIT n–1
INSTRUCTION CYCLE
SENABLE
DATA TRANSFER CYCLE
SCLK
Figure 6. Timing Diagram Register Read from AD9975
SDATA
MSB/LSB Transfers
The AD9975 serial port can support both most significant bit
(MSB) first or least significant bit (LSB) first data formats. The
bit order is controlled by the SPI LSB First Bit (Register 0, Bit 6).
The default value is 0, MSB first. Multibyte data transfers in
MSB format can be completed by writing an instruction byte
that includes the register address of the last address to be accessed.
The AD9975 will automatically decrement the address for each
successive byte required for the multibyte communication cycle.
When the SPI LSB First Bit (Register 0, Bit 6) is set high, the
serial port interprets both instruction and data bytes LSB first.
Multibyte data transfers in LSB format can be completed by
writing an instruction byte that includes the register address of
the first address to be accessed. The AD9975 will automatically
increment the address for each successive byte required for the
multibyte communication cycle.
R/W
I6(n) I5(n)
I4
I3
I2
I1
I0
D7n
D6n
D6n
D7n
Figure 7a. Serial Register Interface Timing MSB First
INSTRUCTION CYCLE
SENABLE
DATA TRANSFER CYCLE
SCLK
SDATA
I0
I1
I2
I3
I4
I5(n) I6(n)
R/W D00
D10
D20
D6n
D7n
Figure 7b. Serial Register Interface Timing LSB First
Notes on Serial Port Operation
The serial port is disabled and all registers are set to their default
values during a hardware reset. During a software reset, all registers except Register 0 are set to their default values. Register 0
will remain at the last value sent, with the exception that the
Software Reset Bit will be set to 0.
Table V. Register Layout
Address
(hex)
Bit 7
00
Bit 6
Select 4-Wire LSB/MSB
SPORT
First
01
Bit 5
Bit 4
Bit 3
Bit 2
Bit 1
Bit 0
Software
Reset
PowerDown
PLL-A
PowerDown
DAC
00
R/W
Power-Down Power-Down PowerPower-Down 00
Interpolators RX Reference Down ADC Receive Filter
and SPGA and CPGA
R/W
00
R/W
01
R/W
01
R/W
80
R/W
00
R/W
Sample TX TX Data
10
on Falling Input Twos
TXCLKIN Complement
R/W
ADC Output RX Data
00
on Falling Output Twos
RXCLK
Complement
R/W
02
03
ADC Clock
Source
OSC IN/2
RX LPF
Tuning
Update
Disable
05
RX LPF Filter Tuning Target [7:0]
06
RXBOOST
Active Low
07
08
RX LPF
Tuning
Update in
Progress
PLL-A (xL)
Multiplier [1:0]
04
RX Path RX Digital Fast ADC
DC Offset HPF
Sampling
Correction Bypass
RXBOOST PGA Gain
Setting
through
Register
Wideband RX Enable
LPF
1-Pole
RX LPF
RX PGA Gain [4:0]
TX Interpolation Filter Select
CLK-B Equal CLK-A
CLK-B
to OSC IN/4 Equal
Output
to OSC IN Disable
CLK-A
Output
Disable
Default
(hex) Type
Three-State
RX Port
0F
Version [3:0]
–16–
RX LPF
Bypass
00
R
REV. 0
AD9975
The serial port is operated by an internal state machine and is
dependent on the number of SCLK cycles since the last time
SENABLE went active. On every eighth rising edge of SCLK, a
byte is transferred over the SPI. During a multibyte write cycle,
this means the registers of the AD9975 are not simultaneously
updated but occur sequentially. For this reason, it is recommended that single byte transfers be used when changing the
SPI configuration or performing a software reset.
Bit 1,0: PLL-A Multiplier
Bits 1 and 0 determine the multiplication factor (L) for PLL-A
and the DAC sampling clock frequency, FDAC. FDAC = L × FCLKIN.
Bit 1,0
0,0: L = 1
0,1: L = 2
1,0: L = 4
1,1: L = 8
Bit 6: ADC Clock Source OSC IN/2
REGISTER PROGRAMMING DEFINITIONS
Register 0, RESET/SPI Configuration
Setting Bit 6 high selects the the OSC IN clock signal divided by
2 as the ADC sampling clock source. Setting Bit 6 low selects
the OSC IN clock to be used directly as the ADC sampling
clock source. The best ADC performance is achieved by using
an external crystal or by driving the OSC IN pin with a low jitter
clock source.
Bit 5: Software Reset
Setting this bit high resets the chip. The PLLs will relock to the
input clock and all registers (except Register 0x0, Bit 6) revert
to their default values. Upon completion of the reset, Bit 5 is
reset to 0.
Register 4, Receive Filter Selection
The content of the interpolator stage is not cleared by software
or hardware resets. It is recommended to “flush” the transmit
path with zeros before transmitting data.
The AD9975 receive path has a continuous time 4-pole LPF
and a 1-pole digital HPF. The 4-pole LPF has two selectable
cutoff frequencies. Additionally, the filter can be tuned around
those two cutoff frequencies. These filters can also be bypassed
to different degrees as described below.
Bit 6: LSB/MSB First
Setting this bit high causes the serial port to send and receive
data least significant bit (LSB) first. The default low state configures the serial port to send and receive data most significant
bit (MSB) first.
Bit 7: Select 4-Wire SPORT
Setting this bit high puts the serial port into a four-line mode.
The SCLK and SENABLE retain their normal functions,
SDATA becomes an input only line, and the RXBOOST/SDO
pin becomes the serial port output. When in 4-wire mode, the
data on the RXBOOST/SDO pin will change on the falling edge
of SCLK and should be sampled on the rising edge of SCLK.
Register 1, Power-Down
Bit 0: Power-Down Receive Filter and CPGA
Setting this bit high powers down and bypasses the RX LPF
and continuous time programmable gain amplifier.
Fcutoff _low = FADC × 64 / (64 + Target )
Fcutoff _ high = FADC × 158 / (64 + Target )
Bit 0: RX LPF Bypass
Setting this bit high bypasses the 4-pole LPF. The filter is automatically powered down when this bit is set.
Bit 1: Enable 1-Pole RX LPF
The AD9975 can be configured with a 1-pole filter when the
4-pole receive low-pass filter is bypassed. The 1-pole filter is
untrimmed and subject to cutoff frequency variations of ± 20%.
Bit 1: Power-Down ADC and SPGA
Setting this bit high powers down the ADC and the switched
capacitor programmable gain amplifier (SPGA).
Bit 2: Wideband RX LPF
This bit selects the nominal cutoff frequency of the 4-pole LPF.
Setting this bit high selects a nominal cutoff frequency of 28.8 MHz.
When the wideband filter is selected, the RX path gain is limited
to 30 dB.
Bit 2: Power-Down RX Reference
Setting this bit high powers down the ADC reference. This bit
should be set if an external reference is applied.
Bit 3: Fast ADC Sampling
Setting this bit increases the quiescent current in the SVGA block.
This may provide some performance improvement when the ADC
sampling frequency is greater than 40 MSPS.
Bit 3: Power-Down Interpolators
Setting this bit high powers down the transmit digital interpolator.
It does not clear the content of the data path.
Bit 4: Power-Down DAC
Setting this bit high powers down the transmit DAC.
Bit 5: Power-Down PLL-A
Setting this bit high powers down the on-chip phase-locked loop
that generates the transmit path clocks and the auxiliary clock
CLK-A. When powered down, the CLK-A output goes to a high
impedance state.
Register 3, Clock Source Configuration
The AD9975 contains a programmable PLL referred to as PLL-A.
The output of the PLL is used to generate the internal clocks for
the TX path and the auxiliary clock, CLK-A.
REV. 0
The continuous time 4-pole low-pass filter is automatically
calibrated to one of two selectable cutoff frequencies. The cutoff
frequency, Fcutoff, is described as a function of the ADC sampling
frequency FADC and can be influenced ± 15% by the RX filter
tuning target word in Register 5.
Bit 4: RX Digital HPF Bypass
Setting this bit high bypasses the 1-pole digital HPF that follows
the ADC. The digital filter must be bypassed for ADC sampling
above 50 MSPS.
Bit 5: RX Path DC Offset Correction
Writing a 1 to this bit triggers an immediate receive path offset
correction and reads back 0 after the completion of the offset
correction.
Bit 6: RX LPF Tuning Update in Progress
This bit indicates when receive filter calibration is in progress.
The duration of a receive filter calibration is about 500 µs.
Writing to this bit has no effect.
–17–
AD9975
Bit 7: RX LPF Tuning Update Disable
Setting this bit high disables the automatic background receive
filter calibration. The AD9975 automatically calibrates the
receive filter on reset and every few (~2) seconds thereafter to
compensate for process and temperature variation, power supply,
and long term drift. Programming a 1 to this bit disables this
function. Programming a 0 triggers an immediate first calibration
and enables the periodic update.
Table VI. PGA Programming Map
Register 5, Receive Filter Tuning Target
This register sets the filter tuning target as a function of FOSCIN.
See Register 4 description.
Register 6, RX Path Gain Adjust
The AD9975 uses a combination of a continuous time PGA
(CPGA) and a switched capacitor PGA (SPGA) for a gain range
of –6 dB to +36 dB with a resolution of 2 dB. The RX path gain
can be programmed over the serial interface by writing to the
RX path Gain Adjust Register or directly using the GAIN and
MSB aligned TX[5:1] Bits. The register default value is 0x00 for
the lowest gain setting (–6 dB). The register always reads back the
actual gain setting irrespective of which of the two programming
modes was used.
Bits [4:0]: RX PGA Gain
Table VI describes the gains and how they are achieved as a
function of the RX path adjust bits. It should be noted that the
value of these bits will read back the actual gain value to which
the PGA is set. If Bit 5 of this register is low, then the value read
back will be that set by the AGC[2:0] Pins.
Bit 5: PGA Gain Set through Register
Setting this bit high will result in the RX path gain being set by
writing to the PGA Gain Control Register. Default is zero, which
selects writing the gain through the AGC[2:0] pins in conjunction
with the RXBOOST pin.
Bit 6: RXBOOST
This bit is read-only. It reflects the level of the RXBOOST pin.
Bit 7: RXBOOST Active Low
Setting this bit high results in the value mapped to the RXBOOST
pin by the AGC inputs being inverted.
RX Path
Gain [4:0]
RX Path
Gain
CPGA
Gain
SPGA
Gain
0x00
0x01
0x02
0x03
0x04
0x05
0x06
0x07
0x08
0x09
0x0A
0x0B
0x0C
0x0D
0x0E
0x0F
0x10
0x11
0x12*
0x13*
0x14*
0x15*
–6
–4
–2
0
2
4
6
8
10
12
14
16
18
20
22
24
26
28
30/30
30/32
30/34
30/36
0
0
0
0
0
0
6
6
6
12
12
12
18
18
18
24
24
24
24/30
24/30
24/30
24/30
–6
–4
–2
0
2
4
0
2
4
0
2
4
0
2
4
0
2
4
6/0
6/2
6/4
6/6
*When the wideband RX filter bit is set high, the RX path gain is limited to 30 dB.
The first of the two values refers to the mode when the lower RX LPF cutoff
frequency is chosen, or when the RX LPF filter is bypassed.
Register 7, Transmit Path Settings
Bit 0: TX Data Input Twos Complement
Setting this bit high changes the TX path input data format to
twos complement. When this bit is low, the TX data format is
straight binary.
Bit 1: Sample TX Data on Falling TXCLKIN
If Bit 1 is set high, the TX path data will be sampled on the
falling edge of TXCLKIN. When this bit is low, the data will be
sampled on the rising edge of TXCLKIN.
Bit 4 to Bit 7: Interpolation Filter Select
Bits 4 to 7 define the interpolation filter characteristic and interpolation rate.
Bits 7:4;
0x1; see TPC 1. 2⫻ Interpolation, LPF.
0x2; Interpolation Bypass.
0x5; see TPC 2. 2⫻ Interpolation, BPF, Adj image.
The interpolation factor has a direct influence on the rate at
which the TX path will read the input data-words from the input
buffer. When the interpolation filter has been bypassed, the data
will be read out of the buffer at a rate of FOSCIN ⫻ L. When the
interpolator is configured to run in either of the 2⫻ interpolation modes, the data will be read out of the buffer at a rate of
0.5 ⫻ FOSCIN ⫻ L.
Register 8, Receiver and Clock Output Settings
Bit 0: Rx Data Output Twos Complement
Setting this bit high changes the RX path input data format to
twos complement. When this bit is low, the RX data format is
straight binary.
–18–
REV. 0
AD9975
Bit 1: ADC Output on Falling RXCLK
If Bit 1 is set high, the TX path data will be sampled on the falling
edge of RXCLK. When this bit is low, the data will be sampled on
the rising edge of RXCLK.
Bit 3: Three-State RX Port
This bit sets the receive output RX[5:0] into a high impedance
three-state mode. It allows for sharing the bus with other devices.
Bit 4: CLK-A Output Disable
Setting Bit 4 high fixes the CLK-A output to a Logic 0 output level.
Bit 5: CLK-B Output Disable
Setting Bit 5 high fixes the CLK-A output to a Logic 0 output level.
Bit 6: CLK-A Equal to OSC IN
Setting Bit 6 high sets the CLK-A output signal frequency equal
to the OSC IN signal frequency. Otherwise, the CLK-A output
frequency is equal to FOSCIN × L.
Bit 7: CLK-B Equal to OSC IN/4
Setting Bit 7 high sets the CLKB output signal frequency equal
to the OSC IN/4 signal frequency. Otherwise, the CLKB output
frequency is equal to OSC IN/2.
Register F, Die Revision
This register stores the die revision of the chip. It is a read-only
register.
PCB DESIGN CONSIDERATIONS
Although the AD9975 is a mixed signal device, the part should
be treated as an analog component. The digital circuitry on-chip
has been specially designed to minimize the impact that the
digital switching noise will have on the operation of the analog
circuits. Following the power, grounding, and layout recommendations in this section will help you get the best performance
from the MxFE.
Component Placement
If the three following guidelines of component placement are
followed, chances for getting the best performance from the
MxFE are greatly increased. First, manage the path of return currents flowing in the ground plane so that high frequency switching
currents from the digital circuits do not flow on the ground plane
under the MxFE or analog circuits. Second, keep noisy digital
signal paths and sensitive receive signal paths as short as possible.
Third, keep digital (noise generating) and analog (noise susceptible) circuits as far away from each other as possible.
In order to best manage the return currents, pure digital circuits
that generate high switching currents should be closest to the
power supply entry. This will keep the highest frequency return
current paths short and prevent them from traveling over the
sensitive MxFE and analog portions of the ground plane. Also,
these circuits should be generously bypassed at each device that
will further reduce the high frequency ground currents. The
MxFE should be placed adjacent to the digital circuits such that
the ground return currents from the digital sections will not flow
in the ground plane under the MxFE. The analog circuits should
be placed furthest from the power supply.
Power Planes and Decoupling
The AD9975 evaluation board demonstrates a good power supply
distribution and decoupling strategy. The board has four layers;
two signal layers, one ground plane, and one power plane. The
power plane is split into a 3VDD section, which is used for the
3 V digital logic circuits; a DVDD section, which is used to supply
the digital supply pins of the AD9975; an AVDD section, which
is used to supply the analog supply pins of the AD9975; and a
VANLG section, which supplies the higher voltage analog components on the board. The 3VDD section will typically have the
highest frequency currents on the power plane and should be
kept the furthest from the MxFE and analog sections of the board.
The DVDD portion of the plane brings the current used to power
the digital portion of the MxFE to the device. This should be
treated similar to the 3VDD power plane and be kept from
going underneath the MxFE or analog components. The MxFE
should largely sit on the AVDD portion of the power plane.
The AVDD and DVDD power planes may be fed from the same
low noise voltage source; however, they should be decoupled
from each other to prevent the noise generated in the DVDD
portion of the MxFE from corrupting the AVDD supply. This
can be done by using ferrite beads between the voltage source and
DVDD and between the source and AVDD. Both DVDD and
AVDD should have a low ESR, bulk decoupling capacitor on the
MxFE side of the ferrite as well as a low ESR, ESL decoupling
capacitors on each supply pin (i.e., the AD9975 requires five
power supply decoupling caps, one each on Pins 5, 38, 47, 14, and
35). The decoupling caps should be placed as close to the MxFE
supply pins as possible. An example of the proper decoupling is
shown in the AD9975 evaluation board schematic.
Ground Planes
In general, if the component placing guidelines discussed earlier
can be implemented, it is best to have at least one continuous
ground plane for the entire board. All ground connections should
be made as short as possible. This will result in the lowest impedance return paths and the quietest ground connections.
If the components cannot be placed in a manner that would keep
the high frequency ground currents from traversing under the
MxFE and analog components, it may be necessary to put current
steering channels into the ground plane to route the high frequency
currents around these sensitive areas. These current steering
channels should be made only when and where necessary.
Signal Routing
The digital RX and TX signal paths should be kept as short as
possible. Also, the impedance of these traces should have a controlled impedance of about 50 Ω. This will prevent poor signal
integrity and the high currents that can occur during undershoot
or overshoot caused by ringing. If the signal traces cannot be kept
shorter than about 1.5 inches, then series termination resistors
(33 Ω to 47 Ω) should be placed close to all signal sources. It is
a good idea to series terminate all clock signals at their source
regardless of trace length.
The receive RX+/RX– signals are the most sensitive signals on the
entire board. Careful routing of these signals is essential for good
The AD9975 has several pins that are used to decouple sensitive
receive path performance. The RX+/RX– signals form a differential
internal nodes. These pins are REFIO, REFB, and REFT. The decoupair and should be routed together as a pair. By keeping the traces
pling capacitors connected to these points should have low ESR
adjacent to each other, noise coupled onto the signals will appear
and ESL. These capacitors should be placed as close to the MxFE
as common mode and will be largely rejected by the MxFE receive
as possible and be connected directly to the analog ground plane.
input. Keeping the driving point impedance of the receive signal low
The resistor connected to the FS ADJ pin should also be placed
and placing any low-pass filtering of the signals close to the MxFE
close to the device and connected directly to the analog ground plane.
will further reduce the possibility of noise corrupting these signals.
REV. 0
–19–
AD9975
OUTLINE DIMENSIONS
48-Lead Plastic Quad Flatpack [LQFP]
1.4 mm Thick
(ST-48)
1.60 MAX
0.75
0.60
0.45
PIN 1
INDICATOR
9.00 BSC
37
48
36
1
1.45
1.40
1.35
SEATING
PLANE
SEATING
PLANE
7.00
BSC
TOP VIEW
(PINS DOWN)
7ⴗ
3.5ⴗ
0ⴗ
0.08 MAX
COPLANARITY
VIEW A
25
12
13
0.50
BSC
VIEW A
ROTATED 90ⴗ CCW
24
0.27
0.22
0.17
COMPLIANT TO JEDEC STANDARDS MS-026BBC
PRINTED IN U.S.A.
0.15
0.05
0.20
0.09
C03061–0–8/02(0)
Dimensions shown in millimeters
–20–
REV. 0