EVAL-ADF7021-VDB1Z [ADI]
HIgh Performance Narrow-Band Transceiver IC; 高性能窄带收发器IC型号: | EVAL-ADF7021-VDB1Z |
厂家: | ADI |
描述: | HIgh Performance Narrow-Band Transceiver IC |
文件: | 总60页 (文件大小:1552K) |
中文: | 中文翻译 | 下载: | 下载PDF数据表文档文件 |
High Performance,
Narrow-Band Transceiver IC
ADF7021-V
On-chip fractional-N PLL
FEATURES
On-chip, 7-bit ADC and temperature sensor
Fully automatic frequency control (AFC) loop
Digital received signal strength indication (RSSI)
Integrated Tx/Rx switch
High performance, low power, narrow-band transceiver
Enhanced performance ADF7021-N with external VCO
Frequency bands using external VCO: 80 MHz to 960 MHz
Improved adjacent channel power (ACP) and adjacent
channel rejection (ACR) compared with the ADF7021-N
Programmable IF filter bandwidths: 9 kHz, 13.5 kHz,
and 18.5 kHz
Modulation schemes: 2FSK, 3FSK, 4FSK, MSK
Spectral shaping: Gaussian and raised cosine filtering
Data rates: 0.05 kbps to 24 kbps
Power supply: 2.3 V to 3.6 V
Programmable output power: −16 dBm to +13 dBm
in 63 steps
Automatic power amplifier (PA) ramp control
Receiver sensitivity
−125 dBm at 250 bps, 2FSK
Leakage current in power-down mode: 0.1 μA
APPLICATIONS
Narrow-band, short-range device (SRD) standards
ETSI EN 300 220
500 mW output power capability in 869 MHz g3 subband
with external PA
High performance receiver rejection, blocking, and
adjacent channel power (ACP)
FCC Part 90 (meets Emission Mask D requirements)
FCC Part 95
ARIB STD-T67
Wireless metering
Narrow-band wireless telemetry
−122 dBm at 1 kbps, 2FSK
Patent pending, on-chip image rejection calibration
FUNCTIONAL BLOCK DIAGRAM
CE
CREG[1:4]
MUXOUT
RSET
TEMP
SENSOR
MUX
7-BIT ADC
R
LDO[1:4]
LNA
TEST MUX
2FSK
3FSK
4FSK
LNA
TxRxCLK
CLOCK
AND DATA
RECOVERY
RFIN
RFIN
RSSI/
LOG AMP
Tx/Rx
CONTROL
TxRxDATA
IF FILTER
DEMODULATOR
SWD
GAIN
AGC
CONTROL
SLE
SDATA
SREAD
SCLK
SERIAL
PORT
ADF7021-V
AFC
CONTROL
PA RAMP
2FSK
3FSK
4FSK
GAUSSIAN/
RAISED COSINE
FILTER
Σ-Δ
MODULATOR
÷1/÷2
DIV P
N/N + 1
RFOUT
MOD CONTROL
÷2
3FSK
ENCODING
BUFFER
CP
PFD
CLK
DIV
DIV R
OSC
L2
CPOUT
OSC1 OSC2
CLKOUT
Figure 1.
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. 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 registeredtrademarks arethe property of their respective owners.
One Technology Way, P.O. Box 9106, Norwood, MA 02062-9106, U.S.A.
Tel: 781.329.4700
Fax: 781.461.3113
www.analog.com
©2010 Analog Devices, Inc. All rights reserved.
ADF7021-V
TABLE OF CONTENTS
Features .............................................................................................. 1
Demodulation, Detection, and CDR....................................... 30
Receiver Setup............................................................................. 32
FSK Demodulator Optimization.............................................. 33
AFC Operation........................................................................... 34
Automatic Sync Word Detection (SWD)................................ 35
Applications Information.............................................................. 36
IF Filter Bandwidth Calibration............................................... 36
LNA/PA Matching...................................................................... 37
Image Rejection Calibration..................................................... 38
Packet Structure and Coding.................................................... 39
Programming After Initial Power-Up ..................................... 39
Applications Circuit................................................................... 42
Serial Interface ................................................................................ 43
Readback Format........................................................................ 43
Interfacing to a Microcontroller/DSP ..................................... 44
Register 0—N Register............................................................... 45
Register 1—Oscillator Register................................................. 46
Register 2—Transmit Modulation Register ............................ 47
Register 3—Transmit/Receive Clock Register........................ 48
Register 4—Demodulator Setup Register ............................... 49
Register 5—IF Filter Setup Register......................................... 50
Register 6—IF Fine Calibration Setup Register ..................... 51
Register 7—Readback Setup Register...................................... 52
Register 8—Power-Down Test Register .................................. 53
Register 9—AGC Register......................................................... 54
Register 10—AFC Register ....................................................... 55
Register 11—Sync Word Detect Register................................ 56
Register 12—SWD/Threshold Setup Register........................ 56
Register 13—3FSK/4FSK Demodulation Register................. 57
Register 14—Test DAC Register............................................... 58
Register 15—Test Mode Register ............................................. 59
Outline Dimensions....................................................................... 60
Ordering Guide .......................................................................... 60
Applications....................................................................................... 1
Functional Block Diagram .............................................................. 1
Revision History ............................................................................... 2
General Description......................................................................... 3
Specifications..................................................................................... 4
RF and PLL Specifications........................................................... 4
Transmission Specifications........................................................ 5
Receiver Specifications ................................................................ 6
Digital Specifications ................................................................... 9
General Specifications ............................................................... 10
Timing Characteristics .............................................................. 10
Timing Diagrams........................................................................ 11
Absolute Maximum Ratings.......................................................... 14
ESD Caution................................................................................ 14
Pin Configuration and Function Descriptions........................... 15
Typical Performance Characteristics ........................................... 17
Frequency Synthesizer ................................................................... 21
Reference Input........................................................................... 21
MUXOUT.................................................................................... 22
Voltage Controlled Oscillator (VCO)...................................... 23
Choosing a VCO for Best System Performance..................... 23
Transmitter ...................................................................................... 24
RF Output Stage.......................................................................... 24
Modulation Schemes.................................................................. 24
Spectral Shaping ......................................................................... 26
Modulation and Filtering Options........................................... 27
Transmit Latency........................................................................ 27
Test Pattern Generator............................................................... 27
Receiver Section.............................................................................. 28
RF Front End............................................................................... 28
IF Filter......................................................................................... 28
RSSI/AGC.................................................................................... 28
REVISION HISTORY
4/10—Revision 0: Initial Version
Rev. 0 | Page 2 of 60
ADF7021-V
GENERAL DESCRIPTION
The ADF7021-V is a high performance, low power, narrow-band
RF transceiver based on the ADF7021-N. The architecture of
the ADF7021-V transceiver is similar to that of the ADF7021-N
except that an external VCO is used by the on-chip RF synthesizer
for applications that require improved phase noise performance.
The transmit section contains a low noise fractional-N PLL with
an output resolution of <1 ppm. The frequency-agile PLL allows
the ADF7021-V to be used in frequency-hopping spread spectrum
(FHSS) systems. The VCO is external, which provides better
phase noise and thus lower adjacent channel power (ACP) and
adjacent channel rejection (ACR) compared with the ADF7021-N.
The VCO tuning range extends from 0.2 V to 2 V, which should
be taken into account when choosing the external VCO.
The ADF7021-V is designed to operate in both the license-free
ISM bands and in the licensed bands from 80 MHz to 960 MHz.
To minimize RF feedthrough and spurious emissions, the
external VCO operates at 2× or 4× the desired RF frequency;
the ADF7021-V supports a maximum VCO frequency operation
of 1920 MHz. The 4× VCO operation is programmable by
enabling an additional on-chip divide-by-2 outside the RF
synthesizer loop and offers improved phase noise performance.
The transmitter output power is programmable in 63 steps from
−16 dBm to +13 dBm and has an automatic power amplifier ramp
control to prevent spectral splatter and help meet regulatory
standards. The transceiver RF frequency, channel spacing, and
modulation are programmable using a simple 3-wire interface.
The device operates with a power supply range of 2.3 V to 3.6 V
and can be powered down when not in use.
As with the ADF7021-N receiver, the IF filter bandwidths
of 9 kHz, 13.5 kHz, and 18.5 kHz are supported, making the
ADF7021-V ideally suited to worldwide narrow-band telemetry
applications.
A low IF architecture is used in the receiver (100 kHz), which
minimizes power consumption and the external component
count yet avoids dc offset and flicker noise at low frequencies.
The IF filter has programmable bandwidths of 9 kHz, 13.5 kHz,
and 18.5 kHz. The ADF7021-V supports a wide variety of pro-
grammable features, including Rx linearity, sensitivity, and IF
bandwidth, allowing the user to trade off receiver sensitivity
and selectivity against current consumption, depending on
the application. The receiver also features a patented automatic
frequency control (AFC) loop with programmable pull-in range
that allows the PLL to remove the frequency error in the
incoming signal.
The part has both Gaussian and raised cosine transmit data
filtering options to improve spectral efficiency for narrow-band
applications. It is suitable for circuit applications targeted at the
following:
•
•
•
•
•
European ETSI EN 300 220
North American FCC Part 15, Part 90, and Part 95
Japanese ARIB STD-T67
Korean short-range device regulations
Chinese short-range device regulations
The receiver achieves an image rejection performance of 50 dB
using a patent-pending IR calibration scheme that does not
require the use of an external RF source.
A complete transceiver can be built using a small number of
discrete external components, making the ADF7021-V very
suitable for area-sensitive, high performance driven applications.
An on-chip ADC provides readback of the integrated temper-
ature sensor, external analog input, battery voltage, and RSSI
signal, which can eliminate the need for an external ADC in
some applications. The temperature sensor is accurate to 10ꢀC
over the full operating temperature range of −40ꢀC to +85ꢀC.
This accuracy can be improved by performing a one-point cali-
bration at room temperature and storing the result in memory.
The range of on-chip FSK modulation and data filtering options
allows users greater flexibility in their choice of modulation
schemes while meeting the tight spectral efficiency requirements.
The ADF7021-V also supports protocols that dynamically switch
among 2FSK, 3FSK, and 4FSK to maximize communication
range and data throughput.
Rev. 0 | Page 3 of 60
ADF7021-V
SPECIFICATIONS
VDD = 2.3 V to 3.6 V, GND = 0 V, TA = TMIN to TMAX, unless otherwise noted. Typical specifications are at VDD = 3 V, TA = 25ꢀC. All
measurements are performed with the EVAL-ADF7021-VDBxZ using the PN9 data sequence, unless otherwise noted. The version
number of ETSI EN 300 200-1 is V2.3.1. LBW = loop bandwidth and IFBW = IF filter bandwidth.
RF AND PLL SPECIFICATIONS
Table 1.
Parameter
Min
Typ
Max
Unit
Test Conditions/Comments
RF CHARACTERISTICS
Phase Frequency Detector (PFD)
Frequency
RF/256
24
MHz
Maximum usable PFD at a particular RF frequency
is limited by the minimum N divider value
PHASE-LOCKED LOOP (PLL)
Normalized In-Band Phase Noise
Floor1
PLL Settling
−203
155
dBc/Hz
μs
Measured for a 100 kHz frequency step to within
5 ppm accuracy, PFD = 19.68 MHz, LBW = 8 kHz
EXTERNAL VCO
Tuning Range
0.2
0
2
V
dBm
Pin L2 Input Sensitivity
REFERENCE INPUT
Crystal Reference2
External Oscillator2, 3
Crystal Start-Up Time4
XTAL Bias = 20 μA
XTAL Bias = 35 μA
Input Level for External Oscillator
OSC1 Pin
VCO frequency < 1920 MHz
3.625
3.625
24
24
MHz
MHz
10 MHz XTAL, 33 pF load capacitors, VDD = 3.0 V
0.930
0.438
ms
ms
0.8
V p-p
V
Clipped sine wave
OSC2 Pin
CMOS levels
ADC PARAMETERS
Integral Nonlinearity (INL)
Differential Nonlinearity (DNL)
VDD = 2.3 V to 3.6 V, TA = 25°C
0.4
0.4
LSB
LSB
1 This value can be used to calculate the in-band phase noise for any operating frequency. Use the following equation to calculate the in-band phase noise performance
as seen at the power amplifier (PA) output: −203 + 10 log(fPFD) + 20 logN.
2 Guaranteed by design. Sample tested to ensure compliance.
3 A TCXO, VCXO, or OCXO can be used as an external oscillator.
4 Crystal start-up time is the time from chip enable (CE) being asserted to correct clock frequency on the CLKOUT pin.
Rev. 0 | Page 4 of 60
ADF7021-V
TRANSMISSION SPECIFICATIONS
LBW = loop bandwidth.
Table 2.
Parameter
DATA RATE
2FSK
3FSK
4FSK
Min
Typ
Max
Unit
Test Conditions/Comments
Limited by the loop bandwidth
LBW must be ≥1.25 × data rate for correct operation
LBW = 18.5 kHz
0.05
0.05
0.05
18.5
18.5
24
kbps
kbps
kbps
LBW = 18.5 kHz
MODULATION
Frequency Deviation (fDEV
)
0.056
0.306
56
28.26
156
kHz
kHz
Hz
PFD = 3.625 MHz
PFD = 20 MHz
PFD = 3.625 MHz
Frequency Deviation Resolution
Gaussian Filter Bandwidth Time (BT)
Raised Cosine Filter Alpha
TRANSMIT POWER
Maximum Transmit Power1
Transmit Power Variation vs.
Temperature
0.5
0.5/0.7
Programmable
13
1
dBm
dB
VDD = 3.0 V, TA = 25°C
TA = −40°C to +85°C
Transmit Power Variation vs. VDD
Transmit Power Flatness
Programmable Step Size
1
1
dB
dB
dB
VDD = 2.3 V to 3.6 V at 915 MHz, TA = 25°C
902 MHz to 928 MHz, VDD = 3 V, TA = 25°C
−16 dBm to +13 dBm
0.3125
ADJACENT CHANNEL POWER (ACP)
Gaussian 2FSK modulation, 13 dBm output power,
PFD = 19.68 MHz, LBW = 6 kHz
460 MHz
12.5 kHz Channel Spacing
−47
−53
dBm
dBm
Measured in a 8.5 kHz bandwidth at 12.5 kHz
offset, 2.4 kbps PN9 data, fDEV = 1.2 kHz
Measured in a 16 kHz bandwidth at 25 kHz
offset, 4.8 kbps PN9 data, fDEV = 2.4 kHz
25 kHz Channel Spacing
868 MHz
Compliant with ETSI EN 300 220
12.5 kHz Channel Spacing
−44
−49
dBm
dBm
Measured in a 8.5 kHz bandwidth at 12.5 kHz
offset, 2.4 kbps PN9 data, fDEV = 1.2 kHz
Measured in a 16 kHz bandwidth at 25 kHz
offset, 4.8 kbps PN9 data, fDEV = 2.4 kHz
25 kHz Channel Spacing
MODULATION BANDWIDTH
869.525 MHz, Gaussian 2FSK modulation, 4.8 kbps,
fDEV = 2.4 kHz, 10 dBm output power,2 compliant
with ETSI EN 300 220, LBW = 6 kHz
125 kHz Offset
125 kHz + 200 kHz
125 kHz + 400 kHz
125 kHz + 1 MHz
EMISSION MASK
−74.5
−79
−69.5
−62
dBm/1 kHz
dBm/1 kHz
dBm/10 kHz
dBm/100 kHz
FCC Part 90 Emission Mask D, 100 Hz resolution
bandwidth, Gaussian 2FSK modulation, LBW =
6 kHz, 10 dBm output power, 2.4 kbps PN9 data,
fDEV = 1.2 kHz
12.5 kHz Offset
460 MHz
−77
dBc
OCCUPIED BANDWIDTH
99.0% of total mean power, LBW = 6 kHz, 10 dBm
output power
2FSK, Gaussian Data Filtering
12.5 kHz Channel Spacing
25 kHz Channel Spacing
4.0
8.5
kHz
kHz
2.4 kbps PN9 data, fDEV = 1.2 kHz
4.8 kbps PN9 data, fDEV = 2.4 kHz
2FSK, Raised Cosine Data Filtering
12.5 kHz Channel Spacing
25 kHz Channel Spacing
4.5
9.6
kHz
kHz
2.4 kbps PN9 data, fDEV = 1.2 kHz
4.8 kbps PN9 data, fDEV = 2.4 kHz
Rev. 0 | Page 5 of 60
ADF7021-V
Parameter
Min
Typ
Max
Unit
Test Conditions/Comments
3FSK, Raised Cosine Filtering
12.5 kHz Channel Spacing
25 kHz Channel Spacing
4FSK, Raised Cosine Filtering
25 kHz Channel Spacing
SPURIOUS EMISSIONS
Reference Spurs
4.3
8.5
kHz
kHz
2.4 kbps PN9 data, fDEV = 1.2 kHz
4.8 kbps PN9 data, fDEV = 2.4 kHz
11.3
−65
kHz
dBc
9.6 kbps PN9 data, fDEV = 1.2 kHz
LBW = 8 kHz
HARMONICS3
13 dBm output power
Second Harmonic
Third Harmonic
All Other Harmonics
OPTIMUM PA LOAD IMPEDANCE
fRF = 915 MHz
fRF = 868 MHz
fRF = 470 MHz
fRF = 450 MHz
fRF = 426 MHz
−35/−52
−43/−60
−36/−65
dBc
dBc
dBc
Unfiltered conductive/filtered conductive
Unfiltered conductive/filtered conductive
Unfiltered conductive/filtered conductive
39 + j61
48 + j54
Ω
Ω
Ω
Ω
Ω
Ω
Ω
Ω
97.5 + j64.4
98 + j65
100 + j65
129 + j63
173 + j49
74.5 + j48.5
fRF = 315 MHz
fRF = 175 MHz
fRF = 169 MHz
1 Measured as maximum unmodulated power.
2 Suitable for ETSI 500 mW Tx requirements.
3 Conductive filtered harmonic emissions measured on the EVAL-ADF7021-VDBxZ, which includes a T-stage harmonic filter (two inductors and one capacitor).
RECEIVER SPECIFICATIONS
LBW = loop bandwidth and IFBW = IF filter bandwidth.
Table 3.
Parameter
DATA RATE
2FSK
Min
Typ
Max
Unit
Test Conditions/Comments
Limited by the IF filter bandwidth1
IFBW = 9 kHz
IFBW = 13.5 kHz
IFBW = 18.5 kHz
0.05
0.05
0.05
0.05
0.05
9.0
kbps
kbps
kbps
kbps
kbps
13.5
18.5
18.5
24
3FSK
4FSK
IFBW = 18.5 kHz
IFBW = 18.5 kHz
SENSITIVITY
2FSK
Bit error rate (BER) = 10−3
Sensitivity at 0.25 kbps
−125
−122
−119
−116
−114
dBm
dBm
dBm
dBm
dBm
fDEV = 1 kHz, high sensitivity mode, IFBW = 9.0 kHz
fDEV = 1 kHz, high sensitivity mode, IFBW = 9.0 kHz
fDEV = 1.2 kHz, high sensitivity mode, IFBW = 9.0 kHz
fDEV = 2.4 kHz, high sensitivity mode, IFBW = 9.0 kHz
fDEV = 4.8 kHz, high sensitivity mode, IFBW = 18.5 kHz
Sensitivity at 1 kbps
Sensitivity at 2.4 kbps
Sensitivity at 4.8 kbps
Sensitivity at 9.6 kbps
Gaussian 2FSK
Sensitivity at 0.25 kbps
Sensitivity at 1 kbps
Sensitivity at 2.4 kbps
Sensitivity at 4.8 kbps
Sensitivity at 9.6 kbps
GMSK
−125
−122
−120
−117
−114
dBm
dBm
dBm
dBm
dBm
fDEV = 1 kHz, high sensitivity mode, IFBW = 9.0 kHz
fDEV = 1 kHz, high sensitivity mode, IFBW = 9.0 kHz
fDEV = 1.2 kHz, high sensitivity mode, IFBW = 9.0 kHz
fDEV = 2.4 kHz, high sensitivity mode, IFBW = 9.0 kHz
fDEV = 4.8 kHz, high sensitivity mode, IFBW = 18.5 kHz
Sensitivity at 4.8 kbps
−114.5
dBm
fDEV = 1.2 kHz, high sensitivity mode, IFBW = 9.0 kHz
Rev. 0 | Page 6 of 60
ADF7021-V
Parameter
Min
Typ
Max
Unit
Test Conditions/Comments
Raised Cosine 2FSK
Sensitivity at 0.25 kbps
Sensitivity at 1 kbps
Sensitivity at 2.4 kbps
Sensitivity at 4.8 kbps
Sensitivity at 9.6 kbps
3FSK
−125
−121
−120
−115
−114
dBm
dBm
dBm
dBm
dBm
fDEV = 1 kHz, high sensitivity mode, IFBW = 9.0 kHz
fDEV = 1 kHz, high sensitivity mode, IFBW = 9.0 kHz
fDEV = 1.2 kHz, high sensitivity mode, IFBW = 9.0 kHz
fDEV = 2.4 kHz, high sensitivity mode, IFBW = 9.0 kHz
fDEV = 4.8 kHz, high sensitivity mode, IFBW = 18.5 kHz
Sensitivity at 4.8 kbps
−110
−110
−112
−109
dBm
dBm
dBm
dBm
fDEV = 2.4 kHz, high sensitivity mode, IFBW = 18.5 kHz,
Viterbi detection on
Raised Cosine 3FSK
Sensitivity at 4.8 kbps
fDEV = 2.4 kHz, high sensitivity mode, IFBW = 13.5 kHz,
alpha = 0.5, Viterbi detection on
4FSK
Sensitivity at 4.8 kbps
fDEV (inner)2 = 1.2 kHz, high sensitivity mode, IFBW = 13.5
kHz
Raised Cosine 4FSK
Sensitivity at 4.8 kbps
fDEV (inner)2 = 1.2 kHz, high sensitivity mode, IFBW = 13.5 kHz,
alpha = 0.5
INPUT IP3
Two-tone test, fLO = 860 MHz, f1 = fLO + 100 kHz,
f2 = fLO − 800 kHz
Low Gain, Enhanced Linearity
Mode
−3
dBm
LNA_GAIN = 3, MIXER_LINEARITY = 1
Medium Gain Mode
High Sensitivity Mode
−13.5
−24
dBm
dBm
LNA_GAIN = 10, MIXER_LINEARITY = 0
LNA_GAIN = 30, MIXER_LINEARITY = 0
ADJACENT CHANNEL REJECTION
(ACR)
868 MHz
Desired signal is 3 dB above the sensitivity point of
−109.5 dBm as per EN 300 220; rejection is measured as the
level of an unmodulated interferer to cause a BER of 10−2 for
the desired signal
12.5 kHz Channel Spacing
25 kHz Channel Spacing
12.5 kHz Channel Spacing
25 kHz Channel Spacing
12.5 kHz Channel Spacing
25 kHz Channel Spacing
12.5 kHz Channel Spacing
25 kHz Channel Spacing
25 kHz Channel Spacing
460 MHz
−60
−39
−60
−40
−59.5
−42
−63
−45
−57
dBm
dBm
dBm
dBm
dBm
dBm
dBm
dBm
dBm
IFBW = 9 kHz, data rate = 0.25 kbps, fDEV = 1 kHz, LBW = 6 kHz
IFBW = 9 kHz, data rate = 0.25 kbps, fDEV = 1 kHz, LBW = 6 kHz
IFBW = 9 kHz, data rate = 1 kbps, fDEV = 1 kHz, LBW = 6 kHz
IFBW = 9 kHz, data rate = 1 kbps, fDEV = 1 kHz, LBW = 6 kHz
IFBW = 9 kHz, data rate = 2.4 kbps, fDEV = 1.2 kHz, LBW = 6 kHz
IFBW = 9 kHz, data rate = 2.4 kbps, fDEV = 1.2 kHz, LBW = 6 kHz
IFBW = 9 kHz, data rate = 4.8 kbps, fDEV = 2.4 kHz, LBW = 6 kHz
IFBW = 9 kHz, data rate = 4.8 kbps, fDEV = 2.4 kHz, LBW = 6 kHz
IFBW = 18.5 kHz, data rate = 9.6 kbps, fDEV = 4.8 kHz, LBW = 6 kHz
Desired signal is at −106.5 dBm; rejection is measured as the
level of an unmodulated interferer to cause a BER of 10−2 for
the desired signal
12.5 kHz Channel Spacing
25 kHz Channel Spacing
12.5 kHz Channel Spacing
25 kHz Channel Spacing
12.5 kHz Channel Spacing
25 kHz Channel Spacing
12.5 kHz Channel Spacing
25 kHz Channel Spacing
25 kHz Channel Spacing
COCHANNEL REJECTION
−59.5
−37.5
−60
−41
−62
dBm
dBm
dBm
dBm
dBm
dBm
dBm
dBm
dBm
IFBW = 9 kHz, data rate = 0.25 kbps, fDEV = 1 kHz, LBW = 6 kHz
IFBW = 9 kHz, data rate = 0.25 kbps, fDEV = 1 kHz, LBW = 6 kHz
IFBW = 9 kHz, data rate = 1 kbps, fDEV = 1 kHz, LBW = 6 kHz
IFBW = 9 kHz, data rate = 1 kbps, fDEV = 1 kHz, LBW = 6 kHz
IFBW = 9 kHz, data rate = 2.4 kbps, fDEV = 1.2 kHz, LBW = 6 kHz
IFBW = 9 kHz, data rate = 2.4 kbps, fDEV = 1.2 kHz, LBW = 6 kHz
IFBW = 9 kHz, data rate = 4.8 kbps, fDEV = 2.4 kHz, LBW = 6 kHz
IFBW = 9 kHz, data rate = 4.8 kbps, fDEV = 2.4 kHz, LBW = 6 kHz
IFBW = 18.5 kHz, data rate = 9.6 kbps, fDEV = 4.8 kHz, LBW = 6 kHz
−43
−61.5
−44.5
−56
Desired signal is 3 dB above the sensitivity point of
−109.5 dBm; rejection is measured as the level of an
interferer to cause a BER of 10−2 for the desired signal
868 MHz
−5
dB
IFBW = 9 kHz, data rate = 4.8 kbps, fDEV = 2.4 kHz, LBW = 6 kHz
Rev. 0 | Page 7 of 60
ADF7021-V
Parameter
Min
Typ
Max
Unit
Test Conditions/Comments
IMAGE CHANNEL REJECTION
Desired signal (2FSK, 9.6 kbps, 4 kHz deviation) is 3 dB
above the sensitivity point (BER = 10−2); modulated inter-
ferer (2FSK, 9.6 kbps, 4 kHz deviation) is placed at the
image frequency of fRF − 200 kHz; the interferer level is
increased until BER = 10−2
868 MHz
460 MHz
26/39
29/50
dB
dB
Uncalibrated/calibrated,3 VDD = 3.0 V, TA = 25°C
Uncalibrated/calibrated,3 VDD = 3.0 V, TA = 25°C
BLOCKING
Desired signal is 3 dB above the sensitivity point of
−109.5 dBm; rejection is measured as the level of an
unmodulated interferer to cause a BER of 10−2 for the
desired signal; as per ETSI EN 300 220-1
1 MHz
2 MHz
5 MHz
10 MHz
−29.5
−26.5
−26
−25.5
12
dBm
dBm
dBm
dBm
dBm
SATURATION (MAXIMUM
INPUT LEVEL)
2FSK mode, BER = 10−3
RECEIVED SIGNAL STRENGTH
INDICATION (RSSI)
Input Power Range4
Linearity
Absolute Accuracy
Response Time
−120 to −47
2
3
dBm
dB
dB
Input power range = −100 dBm to −47 dBm
Input power range = −100 dBm to −47 dBm
As per AGC gain stage, AGC clock = 3 kHz
333
μs
AUTOMATIC FREQUENCY LOOP
(AFC)
Pull-In Range, Minimum
Pull-In Range, Maximum
0.5
1.5 × IF_
kHz
kHz
Range is programmable in Register 10 (Bits[DB31:DB24])
Range is programmable in Register 10 (Bits[DB31:DB24])
FILTER_BW
Response Time
Accuracy
96
0.5
Bits
kHz
Dependent on modulation index
Input power range = −100 dBm to +12 dBm
Rx SPURIOUS EMISSIONS5
External 920 MHz VCO
−54/−88
−45/−66
−85/−85
−39/−52
dBm
dBm
dBm
dBm
<1 GHz at antenna input, unfiltered conductive/filtered
conductive
>1 GHz at antenna input, unfiltered conductive/filtered
conductive
<1 GHz at antenna input, unfiltered conductive/filtered
conductive
>1 GHz at antenna input, unfiltered conductive/filtered
conductive
External 920 MHz VCO
External 1738 MHz VCO
External 1738 MHz VCO
LNA INPUT IMPEDANCE
RFIN to RFGND; refer to the AN-859 Application Note for
other frequencies
fRF = 915 MHz
fRF = 868 MHz
fRF = 470 MHz
fRF = 450 MHz
fRF = 426 MHz
fRF = 315 MHz
fRF = 175 MHz
fRF = 169 MHz
24 − j60
26 − j63
Ω
Ω
Ω
Ω
Ω
Ω
Ω
Ω
58 − j124
63 − j129
68 − j134
96 − j160
178 − j190
182.5 − j194
1 Using Gaussian or raised cosine filtering. The frequency deviation should be chosen to ensure that the transmit-occupied signal bandwidth is within the receiver
IF filter bandwidth.
2 4FSK fDEV is defined as the frequency spacing from the RF carrier to +fDEV or −fDEV. It is also equal to half the frequency spacing between adjacent symbols.
3 Calibration of the image rejection used an external RF source.
4 For received signal levels < −100 dBm, it is recommended that the RSSI readback value be averaged over a number of samples to improve RSSI accuracy at low input power.
5 Filtered conductive receive spurious emissions are measured on the EVAL-ADF7021-VDBxZ, which includes a T-stage harmonic filter (two inductors and one capacitor).
Rev. 0 | Page 8 of 60
ADF7021-V
DIGITAL SPECIFICATIONS
Table 4.
Parameter
Min
Typ
Max
Unit
Test Conditions/Comments
TIMING INFORMATION
Chip Enabled to Regulator
Ready
50
μs
CREG[1:4] = 100 nF
Chip Enabled to Tx Mode
TCXO Reference
XTAL
32-bit register write time = 50 μs
Depends on VCO settling
Depends on VCO settling
1
2
ms
ms
Chip Enabled to Rx Mode
32-bit register write time = 50 μs, IF filter coarse
calibration only
TCXO Reference
XTAL
Tx-to-Rx Turnaround Time
1.2
2.2
ms
ms
ms
Depends on VCO settling
Depends on VCO settling
Time to synchronized data output; includes AGC
settling (three AGC levels) and CDR synchronization;
tBIT = data bit period; AFC settling not included
AGC settling +
(5 × tBIT)
LOGIC INPUTS
Input High Voltage, VINH
Input Low Voltage, VINL
Input Current, IINH/IINL
Input Capacitance, CIN
Control Clock Input
LOGIC OUTPUTS
0.7 × VDD
V
V
μA
pF
MHz
0.2 × VDD
1
10
50
Output High Voltage, VOH
Output Low Voltage, VOL
CLKOUT Rise/Fall Time
CLKOUT Load
VDD2 − 0.4
V
V
ns
pF
IOH = 500 μA
IOL = 500 μA
0.4
5
10
Rev. 0 | Page 9 of 60
ADF7021-V
GENERAL SPECIFICATIONS
Table 5.
Parameter
Min
Typ
Max
Unit
Test Conditions/Comments
TEMPERATURE RANGE (TA)
POWER SUPPLIES
Voltage Supply, VDD
TRANSMIT CURRENT CONSUMPTION1, 2
868 MHz
−40
+85
°C
2.3
3.6
V
All VDDx pins must be tied together
VDD = 3.0 V, PA is matched into 50 Ω
0 dBm
5 dBm
10 dBm
17.6
20.8
27.1
mA
mA
mA
460 MHz
0 dBm
5 dBm
10 dBm
13.8
17
23
mA
mA
mA
RECEIVE CURRENT CONSUMPTION2
VDD = 3.0 V
868 MHz
Low Current Mode
High Sensitivity Mode
460 MHz
19.3
21.7
mA
mA
Low Current Mode
High Sensitivity Mode
POWER-DOWN CURRENT CONSUMPTION2
Low Power Sleep Mode
16.3
18.3
mA
mA
0.1
1
μA
CE low
1 The transmit current consumption tests used the same combined PA and LNA matching network as that used on the EVAL-ADF7021-VDBxZ evaluation boards.
Improved PA efficiency is achieved by using a separate PA matching network.
2 Device current only. VCO and TCXO currents are excluded.
TIMING CHARACTERISTICS
VDD = 3 V 10%, GND = 0 V, TA = 25ꢀC, unless otherwise noted. Guaranteed by design but not production tested.
Table 6.
Parameter
Limit at TMIN to TMAX
Unit
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns
μs
μs
Description
t1
t2
t3
t4
t5
t6
t8
t9
t10
t11
t12
t13
t14
t15
>10
>10
>25
>25
>10
>20
<25
<25
SDATA to SCLK setup time
SDATA to SCLK hold time
SCLK high duration
SCLK low duration
SCLK to SLE setup time
SLE pulse width
SCLK to SREAD data valid, readback
SREAD hold time after SCLK, readback
SCLK to SLE disable time, readback
TxRxCLK negative edge to SLE
TxRxDATA to TxRxCLK setup time (Tx mode)
TxRxCLK to TxRxDATA hold time (Tx mode)
TxRxCLK negative edge to SLE
SLE positive edge to positive edge of TxRxCLK (Rx mode)
>10
5 < t11 < (¼ × tBIT)
>5
>5
5 < t14 < (¼ × tBIT)
>¼ × tBIT
Rev. 0 | Page 10 of 60
ADF7021-V
TIMING DIAGRAMS
Serial Interface
t3
t4
SCLK
t1
t2
DB2
(CONTROL BIT C3)
DB1
DB0 (LSB)
(CONTROL BIT C1)
SDATA
SLE
DB31 (MSB)
DB30
(CONTROL BIT C2)
t6
t5
Figure 2. Serial Interface Timing Diagram
t1
t2
SCLK
SDATA
SLE
REG 7 DB0
(CONTROL BIT C1)
t3
t10
RV2
RV1
X
X
RV16
RV15
SREAD
t9
t8
Figure 3. Serial Interface Readback Timing Diagram
2FSK/3FSK Timing
±1 × DATA RATE/32
1/DATA RATE
TxRxCLK
TxRxDATA
DATA
Figure 4. TxRxDATA/TxRxCLK Timing Diagram in Receive Mode
1/DATA RATE
TxRxCLK
TxRxDATA
DATA
FETCH
SAMPLE
Figure 5. TxRxDATA/TxRxCLK Timing Diagram in Transmit Mode
Rev. 0 | Page 11 of 60
ADF7021-V
4FSK Timing
In 4FSK receive mode, MSB/LSB synchronization should be guaranteed by detection of the SWD pin in the receive bit stream.
REGISTER 0 WRITE
SWITCH FROM Rx TO Tx
tSYMBOL
t13
t12
t11
tBIT
SLE
TxRxCLK
Rx SYMBOL
LSB
Rx SYMBOL
LSB
Tx SYMBOL
MSB
Rx SYMBOL
MSB
Rx SYMBOL
MSB
Tx SYMBOL
LSB
Tx SYMBOL
MSB
TxRxDATA
Tx/Rx MODE
Rx MODE
Tx MODE
Figure 6. Receive-to-Transmit Timing Diagram in 4FSK Mode
REGISTER 0 WRITE
SWITCH FROM Tx TO Rx
t15
tSYMBOL
t14
tBIT
SLE
TxRxCLK
Tx SYMBOL
LSB
Tx SYMBOL
LSB
Tx SYMBOL
MSB
Tx SYMBOL
MSB
Rx SYMBOL
MSB
Rx SYMBOL
LSB
TxRxDATA
Tx/Rx MODE
Tx MODE
Rx MODE
Figure 7. Transmit-to-Receive Timing Diagram in 4FSK Mode
Rev. 0 | Page 12 of 60
ADF7021-V
UART/SPI Mode
UART mode is enabled by setting Register 0, Bit DB28 to 1. SPI mode is enabled by setting Register 0, Bit DB28 to 1 and setting Register 15,
Bits[DB19:DB17] to 0x7. The transmit/receive data clock is available on the CLKOUT pin.
tBIT
CLKOUT
(TRANSMIT/RECEIVE DATA
CLOCK IN SPI MODE.
NOT USED IN UART MODE.)
FETCH SAMPLE
TxRxCLK
(TRANSMIT DATA INPUT
IN UART/SPI MODE.)
Tx BIT
Tx BIT
Tx BIT
Tx BIT
Tx BIT
TxRxDATA
(RECEIVE DATA OUTPUT
IN UART/SPI MODE.)
HIGH-Z
Tx/Rx MODE
Tx MODE
Figure 8. Transmit Timing Diagram in UART/SPI Mode
tBIT
CLKOUT
(TRANSMIT/RECEIVE DATA
CLOCK IN SPI MODE.
FETCH SAMPLE
NOT USED IN UART MODE.)
TxRxCLK
(TRANSMIT DATA INPUT
IN UART/SPI MODE.)
HIGH-Z
TxRxDATA
(RECEIVE DATA OUTPUT
IN UART/SPI MODE.)
Rx BIT
Rx BIT
Rx BIT
Rx BIT
Rx BIT
Tx/Rx MODE
Rx MODE
Figure 9. Receive Timing Diagram in UART/SPI Mode
Rev. 0 | Page 13 of 60
ADF7021-V
ABSOLUTE MAXIMUM RATINGS
TA = 25ꢀC, unless otherwise noted.
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.
Table 7.
Parameter
VDD to GND1
Rating
−0.3 V to +5 V
Analog I/O Voltage to GND1
Digital I/O Voltage to GND1
Operating Temperature Range
Industrial (B Version)
Storage Temperature Range
Maximum Junction Temperature
MLF θJA Thermal Impedance
Reflow Soldering
−0.3 V to VDDx + 0.3 V
−0.3 V to VDDx + 0.3 V
This device is a high performance RF integrated circuit with an
ESD rating of <2 kV, and it is ESD sensitive. Proper precautions
should be taken for handling and assembly.
−40°C to +85°C
−65°C to +125°C
150°C
26°C/W
ESD CAUTION
Peak Temperature
Time at Peak Temperature
260°C
40 sec
1 GND = GND1 = GND2 = GND4 = RFGND = 0 V.
Rev. 0 | Page 14 of 60
ADF7021-V
PIN CONFIGURATION AND FUNCTION DESCRIPTIONS
PIN 1
INDICATOR
1
36
35
34
33
32
31
30
29
28
27
26
25
CLKOUT
TxRxCLK
TxRxDATA
SWD
VCOIN
2
CREG1
3
VDD1
4
RFOUT
5
VDD2
RFGND
ADF7021-V
TOP VIEW
(Not to Scale)
6
7
RFIN
RFIN
CREG2
ADCIN
GND2
8
R
LNA
9
SCLK
VDD4
RSET
10
11
12
SREAD
SDATA
SLE
CREG4
GND4
NOTES
1. THE EXPOSED PADDLE MUST BE CONNECTED
TO THE GROUND PLANE.
Figure 10. Pin Configuration
Table 8. Pin Function Descriptions
Pin No.
Mnemonic
Description
1
2
VCOIN
CREG1
Do not connect.
Regulator Voltage for PA Block. Place a series 3.9 Ω resistor and a 100 nF capacitor between this pin and
ground for regulator stability and noise rejection.
3
4
VDD1
Voltage Supply for PA Block. Place decoupling capacitors of 0.1 μF and 100 pF as close as possible to this
pin. Tie all VDDx pins together.
The modulated signal is available at this pin. Output power levels are from −16 dBm to +13 dBm. The
output should be impedance matched to the desired load using suitable components.
RFOUT
5
6
RFGND
RFIN
Ground for Output Stage of Transmitter. Tie all GND pins together.
LNA Input for Receiver Section. Input matching is required between the antenna and the differential LNA
input to ensure maximum power transfer.
7
8
9
RFIN
RLNA
VDD4
Complementary LNA Input.
External Bias Resistor for LNA. Optimum resistor is 1.1 kΩ with 5% tolerance.
Voltage Supply for LNA/Mixer Block. Decouple this pin to ground with a 10 nF capacitor. Tie all VDDx pins
together.
10
11
RSET
External Resistor. Sets charge pump current and some internal bias currents. Use a 3.6 kΩ resistor with
5% tolerance.
Regulator Voltage for LNA/Mixer Block. Place a 100 nF capacitor between this pin and ground for
regulator stability and noise rejection.
CREG4
12, 19, 22
13 to 16
GND4
MIX_I, MIX_I,
MIX_Q, MIX_Q
Ground for LNA/Mixer Block. Tie all GND pins together.
Signal Chain Test Pins. These pins are high impedance under normal conditions and should be left
unconnected.
17, 18, 20,
21
FILT_I, FILT_I,
FILT_Q, FILT_Q,
Signal Chain Test Pins. These pins are high impedance under normal conditions and should be left
unconnected.
23
24
TEST_A
CE
Signal Chain Test Pin. This pin is high impedance under normal conditions and should be left unconnected.
Chip Enable. Bringing CE low puts the ADF7021-V into complete power-down. Register values are lost
when CE is low, and the part must be reprogrammed after CE is brought high.
25
26
SLE
Load Enable, CMOS Input. When SLE goes high, the data stored in the shift registers is loaded into one of
the 16 latches. A latch is selected using the control bits.
Serial Data Input. The serial data is loaded MSB first with the four LSBs as the control bits. This pin is a
high impedance CMOS input.
SDATA
Rev. 0 | Page 15 of 60
ADF7021-V
Pin No.
Mnemonic
Description
27
SREAD
Serial Data Output. This pin is used to feed readback data from the ADF7021-V to the microcontroller. The
SCLK input is used to clock each readback bit (for example, AFC or ADC) from the SREAD pin.
28
SCLK
Serial Clock Input. The serial clock is used to clock in the serial data to the registers. The data is latched
into the 32-bit shift register on the SCLK rising edge. This pin is a digital CMOS input.
29
30
GND2
ADCIN
Ground for Digital Block. Tie all GND pins together.
Analog-to-Digital Converter Input. The internal 7-bit ADC can be accessed through this pin. Full scale is
0 V to 1.9 V. Readback is through the SREAD pin.
31
32
33
34
CREG2
VDD2
Regulator Voltage for Digital Block. Place a 100 nF capacitor between this pin and ground for regulator
stability and noise rejection.
Voltage Supply for Digital Block. Place a decoupling capacitor of 10 nF as close as possible to this pin. Tie
all VDDx pins together.
Sync Word Detect. The ADF7021-V asserts this pin when it finds a match for the sync word sequence.
This provides an interrupt for an external microcontroller, indicating that valid data is being received.
Transmit Data Input/Received Data Output. This is a digital pin, and normal CMOS levels apply. In UART/SPI
receive mode, this pin provides an output for the received data. In UART/SPI transmit mode, this pin is
high impedance.
SWD
TxRxDATA
35
TxRxCLK
Outputs the data clock in both receive and transmit modes. This is a digital pin, and normal CMOS levels
apply. The positive clock edge is matched to the center of the received data. In standard transmit mode,
this pin outputs an accurate clock to latch the data from the microcontroller into the transmit section at
the exact required data rate. In UART/SPI transmit mode, this pin is used to input the transmit data. In
UART/SPI receive mode, this pin is high impedance.
36
37
CLKOUT
Divided-Down Version of the Crystal Reference with Output Driver. The digital clock output can be used
to drive several other CMOS inputs, such as a microcontroller clock. The output has a 50:50 mark/space
ratio and is inverted with respect to the reference. Place a series 1 kΩ resistor as close as possible to the
pin in applications where the CLKOUT feature is used.
Provides the DIGITAL_LOCK_DETECT signal. This signal is used to determine whether the PLL is locked to
the correct frequency. It also provides other signals such as REGULATOR_READY, which is an indicator of
the status of the serial interface regulator.
MUXOUT
38
39
40
41
42
43
OSC2
OSC1
VDD3
CREG3
CPOUT
VDD
Connect the reference crystal between this pin and OSC1. A TCXO reference can be used by driving this
pin with CMOS levels and disabling the internal crystal oscillator.
Connect the reference crystal between this pin and OSC2. A TCXO reference can be used by driving this
pin with ac-coupled 0.8 V p-p levels and by enabling the internal crystal oscillator.
Voltage Supply for Charge Pump and PLL Dividers. Decouple this pin to ground with a 10 nF capacitor. Tie
all VDDx pins together.
Regulator Voltage for Charge Pump and PLL Dividers. Place a 100 nF capacitor between this pin and
ground for regulator stability and noise rejection.
Charge Pump Output. This output generates current pulses that are integrated in the loop filter. The
integrated current changes the control voltage on the input to the VCO.
Voltage Supply for RF Circuitry. Place a decoupling capacitor of 10 nF as close as possible to this pin. Tie
all VDDx pins together.
44
45
46
47
48
EP
L2
GND
L1
GND1
CVCO
Exposed Paddle
VCO Buffer Input.
Ground. Tie all GND pins together.
Do not connect.
Ground. Tie all GND pins together.
Do not connect.
The exposed paddle must be connected to the ground plane.
Rev. 0 | Page 16 of 60
ADF7021-V
TYPICAL PERFORMANCE CHARACTERISTICS
–80
16
12
PA_BIAS = 11µA
PA_BIAS = 9µA
RF FREQ = 460MHz
TCXO = 19.2MHz
–90
8
4
–100
–110
0
PA_BIAS = 5µA
–4
PA_BIAS = 7µA
–8
I
I
= 0.3mA
= 0.9mA
CP
CP
–120
–130
–140
–150
–160
–12
–16
–20
–24
–28
–32
–36
–40
0
4
8
12 16 20 24 28 32 36 40 44 48 52 56 60
PA SETTING
1
10
100
1k
10k
100k
FREQUENCY OFFSET (kHz)
Figure 11. Phase Noise Response at 460 MHz, VDD = 3 V
Figure 14. RF Output Power vs. PA Setting
–60
–70
20
0
RF FREQ = 868MHz
TCXO = 19.2MHz
–80
–90
–20
–40
–60
I
I
I
I
= 0.3mA
= 0.9mA
= 1.5mA
= 2.1mA
CP
CP
CP
CP
–100
–110
–120
–130
–140
–150
–160
–80
–100
1
10
100
1k
10k
300
800
1300
1800
2300
2800
FREQUENCY OFFSET (kHz)
FREQUENCY (MHz)
Figure 12. Phase Noise Response at 868 MHz, VDD = 2.3 V
Figure 15. PA Output Harmonic Response with T-Stage LC Filter
20
10
10
DEMODULATION = GFSK
FCC PART 90
EMISSION MASK D
DATA RATE = 9.6kbps
DATA = PRBS9
fDEV = 2.4kHz
RF FREQ = 868MHz
DATA RATE = 2.4kbps
fDEV = 1.2kHz
RF FREQ = 470MHz
IFBW = 4kHz
0
0
–10
–10
–20
–30
–20
2FSK
–30
GFSK
–40
–40
–50
–60
–70
–80
–50
–60
–70
–80
867.97
867.98
867.99
868.00
868.01
868.02
868.03
FREQUENCY (MHz)
FREQUENCY OFFSET FROM CARRIER (Hz)
Figure 16. Output Spectrum in 2FSK and GFSK Modes
Figure 13. Output Spectrum in FCC Part 90 Emission Mask D
and GFSK Modes
Rev. 0 | Page 17 of 60
ADF7021-V
10
RAMP RATE:
CW ONLY
256 CODES/BIT
128 CODES/BIT
64 CODES/BIT
32 CODES/BIT
TRACE = MAX HOLD
DATA RATE = 9.6kbps
DATA = PRBS9
fDEV = 2.4kHz
10
0
PA ON/OFF RATE = 3Hz
PA ON/OFF CYCLES = 10,000
0
V
= 3.0V
DD
RF FREQ = 868MHz
–10
–20
–30
–40
–50
–60
–10
–20
–30
–40
–50
–60
2FSK
RC2FSK
–70
–80
–100
–50
0
50
100
867.97
867.98
867.99
868.00
868.01
868.02
868.03
FREQUENCY (MHz)
FREQUENCY OFFSET (kHz)
Figure 17. Output Spectrum in 2FSK and Raised Cosine 2FSK Modes
Figure 20. Output Spectrum in Maximum Hold
for Various PA Ramp Rate Options
0
–1
–2
10
DATA RATE = 9.6kbps
DATA = PRBS9
fDEV = 2.4kHz
RF FREQ = 868MHz
DATA RATE = 2.4kbps
f
= 4.8Hz
0
DEV
RF FREQ = 868MHz
IFBW = 9kHz
–10
–20
–30
–3
–4
–5
–6
–7
–40°C, 2.3V
–40°C, 3V
–40°C, 3.6V
+25°C, 2.3V
+25°C, 3V
+25°C, 3.6V
+85°C, 2.3V
+85°C, 3V
3FSK
–40
–50
–60
RC3FSK
–70
–80
+85°C, 3.6V
–125 –123 –121 –119 –117 –115 –113 –111 –109 –107 –105
867.97
867.98
867.99
868.00
868.01
868.02
868.03
RF INPUT POWER (dBm)
FREQUENCY (MHz)
Figure 18. Output Spectrum in 3FSK and Raised Cosine 3FSK Modes
Figure 21. 2FSK Sensitivity vs. VDD and Temperature at 868 MHz
0
10
DATA RATE = 9.6kbps
DATA = PRBS9
fDEV = 2.4kHz
RF FREQ = 868MHz
DATA RATE = 1.2kbps
0
f
= 2.4Hz
DEV
–1
–2
RF FREQ = 460MHz
IFBW = 9kHz
–10
–20
–30
–3
–4
–5
–6
–7
4FSK
–40
–50
–60
–40°C, 2.3V
–40°C, 3V
–40°C, 3.6V
+25°C, 2.3V
+25°C, 3V
+25°C, 3.6V
+85°C, 2.3V
+85°C, 3V
RC4FSK
868.00
–70
–80
–90
+85°C, 3.6V
–127 –125 –123 –121 –119 –117 –115 –113 –111
867.94
867.96
867.98
868.02
868.04
868.06
RF INPUT POWER (dBm)
FREQUENCY (MHz)
Figure 19. Output Spectrum in 4FSK and Raised Cosine 4FSK Modes
Figure 22. 2FSK Sensitivity vs. VDD and Temperature at 460 MHz
Rev. 0 | Page 18 of 60
ADF7021-V
0.6
0.5
–20
–40
DATA RATE = 1.2kbps
= 2.4Hz
RF FREQ = 868MHz
IFBW = 9kHz
f
RSSI
READBACK LEVEL
DEV
–40°C, 2.3V
–40°C, 3V
–40°C, 3.6V
+25°C, 2.3V
+25°C, 3V
+25°C, 3.6V
+85°C, 2.3V
+85°C, 3V
–60
0.4
0.3
0.2
–80
–100
–120
–140
+85°C, 3.6V
ACTUAL RF INPUT LEVEL
0.1
0
–122.5 –112.5 –102.5 –92.5 –82.5 –72.5 –62.5 –52.5 –42.5
RF INPUT POWER (dBm)
–130
–125
–120
–115
–110
–105
RF INPUT POWER (dBm)
Figure 26. Digital RSSI Readback Linearity
Figure 23. 2FSK Sensitivity vs. VDD and Temperature at 868 MHz
0.6
80
DATA RATE = 1.2kbps
f
= 2.4Hz
DEV
70
60
50
RF FREQ = 460MHz
IFBW = 9kHz
0.5
0.4
0.3
0.2
–40°C, 2.3V
–40°C, 3V
–40°C, 3.6V
+25°C, 2.3V
+25°C, 3V
+25°C, 3.6V
+85°C, 2.3V
+85°C, 3V
40
30
20
10
0
CALIBRATED
UNCALIBRATED
0.1
0
+85°C, 3.6V
–10
–130
–125
–120
–115
–110
–105
RF INPUT POWER (dBm)
BLOCKER FREQUENCY (MHz)
Figure 24. 2FSK Sensitivity vs. VDD and Temperature at 460 MHz
Figure 27. Image Rejection, Uncalibrated vs. Calibrated
2.5
0
100
90
80
70
60
50
40
30
20
10
0
+90°C
–2.5
–5.0
–7.5
–10.0
–12.5
–15.0
–17.5
–20.0
–22.5
–25.0
–27.5
–30.0
–32.5
–35.0
–37.5
–40°C
–10
–20
–15
–10
–5
0
5
10
15
20
90
92
94
96
98
100 102 104 106 108 110
FREQUENCY OFFSET (MHz)
IF FREQUENCY (kHz)
Figure 25. Wideband Interference Rejection (Modulated Carrier Is Swept 20 MHz
Either Side of an 868 MHz Modulated GFSK 2.4 kHz/4.8 kbps Wanted Signal at the
Sensitivity Point (−106.5 dBm); the Power Level of the Blocker Is Adjusted to
Give a BER of 10−2; Interferer Is a GFSK PRBS15 4.8 kHz/2.4 kHz Signal)
Figure 28. Variation of IF Filter Response with Temperature
(IF_FILTER_BW = 9 kHz, Temperature Range Is −40°C to +90°C in 10° Steps)
Rev. 0 | Page 19 of 60
ADF7021-V
–70
–80
–100
–102
–104
–106
–108
–110
–112
–114
–116
RF FREQ = 860MHz
2FSK MODULATION
DATA RATE = 9.6kbps
IFBW = 25kHz
2FSK MODULATION
DATA RATE = 9.6kbps
fDEV = 4kHz
HIGH MIXER
LINEARITY
IFBW = 12.5kHz
DEMOD = CORRELATOR
SENSITIVITY @ BER = 10
V
= 3.0V
DD
–3
TEMPERATURE = 25°C
IP3 = –5dBm
–90
–100
–110
–120
–130
IP3 = –9dBm
DISCRIMINATOR BANDWIDTH =
2× FSK FREQUENCY DEVIATION
IP3 = –3dBm
IP3 = –20dBm
IP3 = –24dBm
DEFAULT
MIXER
IP3 = –13.5dBm
LINEARITY
DISCRIMINATOR BANDWIDTH =
1× FSK FREQUENCY DEVIATION
–118
0
0.2 0.4 0.6
0.8
1.0
1.2
3, 72
10, 72
30, 72
(HIGH GAIN MODE)
(LOW GAIN MODE)
(MEDIUM GAIN MODE)
MODULATION INDEX
LNA GAIN, FILTER GAIN
Figure 29. 2FSK Sensitivity vs. Modulation Index and Correlator
Discriminator Bandwidth
Figure 31. 2FSK Receiver Sensitivity vs. LNA Gain/IF Filter Gain and Mixer
Linearity Settings (Input IP3 at Each Setting Also Shown)
0
–1
THRESHOLD DETECTION
–2
VITERBI DETECTION
–3
–4
–5
3FSK MODULATION
DD
V
= 3.0V, TEMP = 25°C
DATA RATE = 9.6kbps
–6
fDEV = 2.4kHz
RF FREQ = 868MHz
IFBW = 18.75kHz
–7
–120 –118 –116 –114 –112 –110 –108 –106 –104 –102 –100
INPUT POWER (dBm)
Figure 30. 3FSK Receiver Sensitivity Using Viterbi Detection and
Threshold Detection
Rev. 0 | Page 20 of 60
ADF7021-V
FREQUENCY SYNTHESIZER
CLKOUT Divider and Buffer
REFERENCE INPUT
The CLKOUT circuit takes the reference clock signal from the
oscillator section, shown in Figure 32, and supplies a divided-
down, 50:50 mark/space signal to the CLKOUT pin. The
CLKOUT signal is inverted with respect to the reference clock.
An even divide from 2 to 30 is available; this divide number is
set in Register 1, Bits[DB10:DB7]. On power-up, the CLKOUT
defaults to divide-by-8.
The on-board crystal oscillator circuitry (see Figure 32) can use
a quartz crystal as the PLL reference. A quartz crystal with a fre-
quency tolerance of ≤10 ppm for narrow-band applications is
recommended. It is possible to use a quartz crystal with >10 ppm
tolerance, but compensation for the frequency error of the crystal
is necessary to comply with the absolute frequency error speci-
fications of narrow-band regulations (for example, ARIB STD-T67
and ETSI EN 300 220).
V
DD
CLKOUT
ENABLE BIT
The oscillator circuit is enabled by setting Bit DB12 in Register 1
high. It is enabled by default on power-up and is disabled by
bringing CE low. Errors in the crystal can be corrected using
the automatic frequency control (AFC) feature or by adjusting
the fractional-N value (see the N Counter section).
DIVIDER
1 TO 15
OSC1
÷2
CLKOUT
Figure 33. CLKOUT Stage
To disable CLKOUT, set the divide number to 0. The output
buffer can drive a load of up to 20 pF with a 10% rise time at
4.8 MHz. Faster edges can result in some spurious feedthrough
to the output. A series resistor (1 kΩ) can be used to slow the
clock edges to reduce these spurs at the CLKOUT frequency.
OSC1
OSC2
CP1
CP2
Figure 32. Crystal Oscillator Circuit on the ADF7021-V
R Counter
Two parallel resonant capacitors are required for oscillation at
the correct frequency. Their values are dependent on the crystal
specification. The resonant capacitors should be selected to
ensure that the series value of capacitance added to the PCB
track capacitance adds up to the specified load capacitance of
the crystal, usually 12 pF to 20 pF. Track capacitance values vary
from 2 pF to 5 pF, depending on board layout. When possible,
choose capacitors that have a very low temperature coefficient
to ensure stable frequency operation over all conditions.
The 3-bit R counter divides the reference input frequency by an
integer from 1 to 7. The divided-down signal is presented as the
reference clock to the phase frequency detector (PFD). The
divide ratio is set in Register 1, Bits[DB6:DB4]. Maximizing the
PFD frequency reduces the N value. This reduces the noise multi-
plied at a rate of 20 log(N) to the output and reduces occurrences
of spurious components.
Register 1 defaults to R = 1 on power-up.
PFD (Hz) = XTAL/R
Using a TCXO Reference
A single-ended reference (TCXO, VCXO, or OCXO) can also be
used with the ADF7021-V. This is recommended for applications
that have absolute frequency accuracy requirements of <10 ppm,
such as applications requiring compliance with ARIB STD-T67
or ETSI EN 300 220. The following are two options for inter-
facing the ADF7021-V to an external reference oscillator.
Loop Filter
The loop filter integrates the current pulses from the charge
pump to form a voltage that tunes the output of the VCO to the
desired frequency. It also attenuates spurious levels generated by
the PLL. A typical loop filter design is shown in Figure 34.
•
An oscillator with CMOS output levels can be applied to
OSC2. The internal oscillator circuit should be disabled by
setting Bit DB12 in Register 1 low.
CHARGE
VCO
PUMP OUT
•
An oscillator with 0.8 V p-p levels can be ac-coupled through
a 22 pF capacitor into OSC1. The internal oscillator circuit
should be enabled by setting Bit DB12 in Register 1 high.
Figure 34. Typical Loop Filter Configuration
The loop should be designed so that the loop bandwidth (LBW) is
approximately 6 kHz. This provides a good compromise between
in-band phase noise and out-of-band spurious rejection. Widening
the LBW excessively reduces the time spent jumping between
frequencies, but it can cause insufficient spurious attenuation.
The loop filter design on the EVAL-ADF7021-VDBxZ should
be used for optimum performance.
Programmable Crystal Bias Current
Bias current in the oscillator circuit can be configured from
20 μA to 35 μA by writing to the XTAL_BIAS bits (Register 1,
Bits[DB14:DB13]). Increasing the bias current allows the crystal
oscillator to power up faster.
Rev. 0 | Page 21 of 60
ADF7021-V
The free design tool ADIsimSRD™ Design Studio can also be
used to design loop filters for the ADF7021-V. See the ADIsimSRD
Design Studio website (www.analog.com/adisimsrd) for details).
The serial interface operates from a regulator supply. Therefore,
to write to the part, CE must be high and the regulator voltage
must be stabilized. Regulator status (CREG4) can be monitored
using the REGULATOR_READY signal from the MUXOUT pin.
N Counter
MUXOUT
The feedback divider in the ADF7021-V PLL consists of an
8-bit integer counter (set using Register 0, Bits[DB26:DB19])
and a 15-bit, Σ-Δ fractional-N divider (set using Register 0,
Bits[DB18:DB4]). The integer counter is the standard pulse-
swallow type that is common in PLLs. It sets the minimum
integer divide value to 23. The fractional divide value provides
very fine resolution at the output, where the output frequency
of the PLL is calculated as
The MUXOUT pin allows access to various digital points in the
ADF7021-V. The state of MUXOUT is controlled in Register 0,
Bits[DB31:DB29].
REGULATOR_READY
REGULATOR_READY is the default setting on MUXOUT after
the transceiver is powered up. The power-up time of the regulator
is typically 50 μs. Because the serial interface is powered from
the regulator, the regulator must be at its nominal voltage before
the ADF7021-V can be programmed. The regulator status can
be monitored at MUXOUT. When the regulator ready signal on
MUXOUT is high, programming of the ADF7021-V can begin.
FRACTIONAL_ N
⎛
⎞
XTAL
R
⎜
⎟
⎟
fOUT
=
× INTEGER _ N +
215
⎜
⎝
⎠
When RF_DIVIDE_BY_2 is enabled (see the Voltage
Controlled Oscillator (VCO) section), this formula becomes
V
DD
XTAL
R
FRACTIONAL _ N
⎛
⎝
⎞
⎟
⎠
fOUT
=
×0.5× INTEGER_N +
⎜
215
REGULATOR_READY (DEFAULT)
FILTER_CAL_COMPLETE
DIGITAL_LOCK_DETECT
RSSI_READY
The combination of INTEGER_N (maximum = 255) and
FRACTIONAL_N (maximum = 32,768/32,768) gives a
maximum N divider of 255 + 1. Therefore, the minimum
usable PFD is
MUX
CONTROL
MUXOUT
Tx_Rx
LOGIC_ZERO
TRISTATE
Maximum Required Output Frequency
LOGIC_ONE
PFDMIN (Hz) =
(
255 +1
)
For example, when operating in the European 868 MHz to
870 MHz band, PFDMIN = 3.4 MHz.
REFERENCE IN
GND
Figure 36. MUXOUT Circuit
PFD/
÷R
FILTER_CAL_COMPLETE
CHARGE
PUMP
VCO
MUXOUT can be set to FILTER_CAL_COMPLETE. This signal
goes low for the duration of both a coarse IF filter calibration
and a fine IF filter calibration. It can be used as an interrupt to
a microcontroller to signal the end of the IF filter calibration.
÷N
THIRD-ORDER
DIGITAL_LOCK_DETECT
Σ-Δ MODULATOR
DIGITAL_LOCK_DETECT indicates when the PLL has locked.
The lock detect circuit is located at the PFD. When the phase
error on five consecutive cycles is less than 15 ns, lock detect is
set high. Lock detect remains high until a 25 ns phase error is
detected at the PFD.
FRACTIONAL_N
INTEGER_N
Figure 35. Fractional-N PLL
Voltage Regulators
The ADF7021-V contains four regulators to supply stable
voltages to the part. The nominal regulator voltage is 2.3 V.
Regulator 1 requires a 3.9 ꢁ resistor and a 100 nF capacitor in
series between CREG1 and ground, whereas the other regula-
tors require a 100 nF capacitor connected between CREGx and
ground. When CE is high, the regulators and other associated
circuitry are powered on, drawing a total supply current of
2 mA. Bringing the CE pin low disables the regulators, reduces
the supply current to less than 1 μA, and erases all values held
in the registers.
RSSI_READY
MUXOUT can be set to RSSI_READY. This indicates that the
internal analog RSSI has settled and that a digital RSSI readback
can be performed.
Tx_Rx
Tx_Rx signifies whether the ADF7021-V is in transmit or
receive mode. When in transmit mode, this signal is low.
When in receive mode, this signal is high. It can be used to
control an external Tx/Rx switch.
Rev. 0 | Page 22 of 60
ADF7021-V
The VCO tuning voltage can be checked for a particular RF
VOLTAGE CONTROLLED OSCILLATOR (VCO)
output frequency by measuring the voltage on the CPOUT pin
when the part is fully powered up in transmit or receive mode.
The VCO tuning range of the external VCO must be 0.2 V to 2 V.
To minimize feedthrough and spurious emissions, the external
VCO must be chosen to operate at a minimum of twice the
required RF frequency. The VCO frequency is divided by 2 inside
the synthesizer loop, providing the required frequency for the
transmitter and for the local oscillator (LO) of the receiver. For
improved phase noise performance, an additional divide-by-2
can be enabled by setting the RF_DIVIDE_BY_2 bit (Bit DB18)
in Register 1.
The input impedance of the L2 pin is programmable and can
be selected to have a high impedance value or 50 Ω impedance,
depending on the VCO selected. The impedance of this pin can
be set using the BUFFER_IMPEDANCE bit (Bit DB17) in
Register 1.
As an example, for 80 MHz operation, a 160 MHz external VCO
could be used with the RF_DIVIDE_BY_2 bit disabled, or a
320 MHz VCO could be used with the RF_DIVIDE_BY_2 bit
enabled to support operation in the 80 MHz band. Assuming
that both VCOs have similar phase noise performance, the
320 MHz design using the additional divide-by-2 should result
in improved transmit ACP, as well as improved ACR, blocking,
and image rejection in the receiver.
CHOOSING A VCO FOR BEST SYSTEM
PERFORMANCE
The interaction between the RF VCO frequency and the refer-
ence frequency can lead to fractional spur creation. When the
synthesizer is in fractional mode (that is, the RF VCO and
reference frequencies are not integer related), spurs can appear
on the VCO output spectrum at an offset frequency that
corresponds to the difference frequency between an integer
multiple of the reference and the VCO frequency.
The maximum VCO frequency of operation supported on the
ADF7021-V is 1920 MHz, which results in a maximum RF
channel frequency of 960 MHz using a 2× VCO or 480 MHz
using a 4× VCO.
These spurs are attenuated by the loop filter. They are more
noticeable on channels close to integer multiples of the reference
where the difference frequency may be inside the loop bandwidth
(thus, the name integer boundary spurs). The occurrence of
these spurs is rare because the integer frequencies are around
multiples of the reference, which is typically >10 MHz. To avoid
having very small or very large values in the fractional register,
choose a suitable reference frequency.
EXTERNAL
COMPONENTS
LOOP
VCO
FILTER
REF
TCXO/XTAL
÷2
÷R
PFD/CP
In addition to spurious considerations, the selection of a high
performance VCO with very low phase noise is essential to
minimize the ACP performance of the transmitter and to
maximize the ACR and blocking resilience of the receiver.
MUX
TO
PA
÷N
÷2
ADF7021-V
SYNTH
FREQUENCY
Figure 37. Voltage Controlled Oscillator (VCO)
Rev. 0 | Page 23 of 60
ADF7021-V
TRANSMITTER
1
2
3
4
...
8
... 16
RF OUTPUT STAGE
DATA BITS
The power amplifier (PA) of the ADF7021-V is based on a
single-ended, controlled current, open-drain amplifier that has
been designed to deliver up to 13 dBm into a 50 Ω load at a
maximum frequency of 960 MHz.
PA RAMP 0
(NO RAMP)
PA RAMP 1
(256 CODES PER BIT)
PA RAMP 2
The PA output current and, consequently, the output power
are programmable over a wide range. The PA configuration is
shown in Figure 38. The output power is set using Register 2,
Bits[DB18:DB13].
(128 CODES PER BIT)
PA RAMP 3
(64 CODES PER BIT)
PA RAMP 4
(32 CODES PER BIT)
REGISTER 2,
BITS[DB12:DB11]
PA RAMP 5
(16 CODES PER BIT)
2
PA RAMP 6
(8 CODES PER BIT)
6
REGISTER 2,
IDAC
PA RAMP 7
(4 CODES PER BIT)
BITS[DB18:DB13]
Figure 39. PA Ramping Settings
RFOUT
PA Bias Currents
REGISTER 2, BIT DB7
+
REGISTER 0, BIT DB27
The PA_BIAS bits (Register 2, Bits[DB12:DB11]) facilitate an
adjustment of the PA bias current to further extend the output
power control range, if necessary. If this feature is not required,
the default value of 9 μA is recommended. If output power
greater than 10 dBm is required, a PA bias setting of 11 μA is
recommended. The output stage is powered down by resetting
Register 2, Bit DB7 to 0.
RFGND
FROM VCO
Figure 38. PA Configuration
The PA is equipped with overvoltage protection, which makes it
robust in severe mismatch conditions. Depending on the applica-
tion, users can design a matching network for the PA to exhibit
optimum efficiency at the desired radiated output power level for
a wide range of antennas, such as loop or monopole antennas.
See the LNA/PA Matching section for more information.
MODULATION SCHEMES
The ADF7021-V supports 2FSK, 3FSK, and 4FSK modulation.
The implementation of these modulation schemes is shown in
Figure 40.
PA Ramping
TO
PA STAGE
LOOP FILTER
PFD/
CHARGE
PUMP
REF
When the PA is switched on or off quickly, its changing input
impedance momentarily disturbs the VCO output frequency.
This process is called VCO pulling, and it manifests as spectral
splatter or spurs in the output spectrum around the desired
carrier frequency. Some radio emissions regulations place
limits on these PA transient-induced spurs (for example, the
ETSI EN 300 220 regulations). By gradually ramping the PA
on and off, PA transient spurs are minimized.
÷2
VCO
÷N
FRACTIONAL_N
THIRD-ORDER
Σ-Δ MODULATOR
INTEGER_N
Tx_FREQUENCY_
DEVIATION
The ADF7021-V has built-in PA ramping configurability. As
Figure 39 illustrates, there are eight ramp rate settings, defined
as a certain number of PA setting codes per one data bit period.
The PA steps through each of its 64 code levels but at different
speeds for each setting. The ramp rate is set by configuring
Bits[DB10:DB8] in Register 2.
2FSK
TxRxDATA
GAUSSIAN
2
3FSK
4FSK
PRE-
CODER
1 – D PR
SHAPING
OR
MUX
RAISED COSINE
FILTERING
4FSK
BIT SYMBOL
MAPPER
If the PA is enabled/disabled by the PA_ENABLE bit (Register 2,
Bit DB7), it ramps up and down. If it is enabled/disabled by the
Tx/Rx bit (Register 0, Bit DB27), it ramps up and turns hard off.
Figure 40. Transmit Modulation Implementation
Rev. 0 | Page 24 of 60
ADF7021-V
Setting the Transmit Data Rate
Three-Level Frequency Shift Keying (3FSK)
In all modulation modes except for oversampled 2FSK mode, an
accurate clock is provided on the TxRxCLK pin to latch the data
from the microcontroller into the transmit section at the required
data rate. The exact frequency of this clock is defined by
In three-level FSK modulation—3FSK, also known as modified
duobinary FSK and as partial response maximum likelihood
Class 4 (PRML4) signaling—the binary data (Logic 0 and Logic 1)
is mapped onto three distinct frequencies: the carrier frequency
(fC), the carrier frequency minus a deviation frequency (fC − fDEV),
and the carrier frequency plus the deviation frequency (fC + fDEV).
XTAL
DATA CLK =
DEMOD_CLK_DIVIDE×CDR_CLK_DIVIDE×32
A Logic 0 is mapped to the carrier frequency, whereas a Logic 1
is mapped onto either the fC − fDEV frequency or the fC + fDEV
frequency.
where:
XTAL is the crystal or TCXO frequency.
DEMOD_CLK_DIVIDE is the divider that sets the demodulator
clock rate (Register 3, Bits[DB9:DB6]).
CDR_CLK_DIVIDE is the divider that sets the CDR clock rate
(Register 3, Bits[DB17:DB10]).
0
+1
–1
fC
–
fDEV
fC
fC + fDEV
See the Register 3—Transmit/Receive Clock Register section for
more programming information.
RF FREQUENCY
Figure 41. 3FSK Symbol-to-Frequency Mapping
Setting the FSK Transmit Deviation Frequency
Compared with 2FSK, this bit-to-frequency mapping results
in a reduced transmission bandwidth because some energy is
removed from the RF sidebands and transferred to the carrier
frequency. At low modulation index, 3FSK improves the trans-
mit spectral efficiency by up to 25% when compared with 2FSK.
In all modulation modes, the deviation from the center
frequency is set using the Tx_FREQUENCY_DEVIATION
bits (Register 2, Bits[DB27:DB19]).
The deviation from the center frequency in Hz is as follows:
For direct RF output,
The bit-to-symbol mapping for 3FSK is implemented using a
linear convolutional encoder that also permits Viterbi detection
to be used in the receiver. A block diagram of the transmit hard-
ware used to realize this system is shown in Figure 42. The
convolutional encoder polynomial used to implement the
transmit spectral shaping is
PFD ×Tx_FREQUENCY_DEVIATION
fDEV (Hz)=
216
With RF_DIVIDE_BY_2 (Register 1, Bit DB18) enabled,
PFD ×Tx_FREQUENCY_DEVIATION
fDEV (Hz)= 0.5×
216
P(D) = 1 − D2
where Tx_FREQUENCY_DEVIATION is a number from 1 to
511 (Register 2, Bits[DB27:DB19]).
where:
P is the convolutional encoder polynomial.
In 4FSK modulation, the four symbols (00, 01, 11, 10) are
D is the unit delay operator.
transmitted as 3 × fDEV and 1 × fDEV
.
A digital precoder with transfer function 1/P(D) implements
an inverse modulo-2 operation of the 1 − D2 shaping filter in
the transmitter.
Binary Frequency Shift Keying (2FSK)
Binary frequency shift keying is implemented by setting the
N value for the center frequency and then toggling it with the
TxRxDATA line. The deviation from the center frequency is set
using the Tx_FREQUENCY_DEVIATION bits (Register 2,
Bits[DB27:DB19]).
Tx DATA
0, 1
0, 1
CONVOLUTIONAL
ENCODER
P(D)
PRECODER
1/P(D)
0, +1, –1
2FSK is selected by setting the MODULATION_SCHEME bits
(Register 2, Bits[DB6:DB4]) to 000.
fC
fC
fC
FSK MOD
CONTROL
AND
+
–
fDEV
fDEV
Minimum shift keying (MSK) or Gaussian minimum shift
keying (GMSK) is supported by selecting 2FSK modulation
and using a modulation index of 0.5. A modulation index of 0.5
is set by configuring Register 2, Bits[DB27:DB19] for an
fDEV = 0.25 × transmit data rate.
TO
N DIVIDER
DATA FILTERING
Figure 42. 3FSK Encoding
Rev. 0 | Page 25 of 60
ADF7021-V
The signal mapping of the input binary transmit data to the
three-level convolutional output is shown in Table 9. The
convolutional encoder restricts the maximum number of
sequential +1s or −1s to two and delivers an equal number of
+1s and −1s to the FSK modulator, thus ensuring equal spectral
energy in both RF sidebands.
The inner deviation frequencies (+fDEV and −fDEV) are set using
the Tx_FREQUENCY_DEVIATION bits (Bits[DB27:DB19] in
Register 2). The outer deviation frequencies are automatically
set to three times the inner deviation frequency.
The transmit clock from Pin TxRxCLK is available after writing
to Register 3 in the power-up sequence for receive mode. The
MSB of the first symbol should be clocked into the ADF7021-V
on the first transmit clock pulse from the ADF7021-V after
writing to Register 3. See Figure 6 and Figure 7 for more timing
information; see Figure 54 and Figure 55 for the power-up
sequences.
Table 9. Three-Level Signal Mapping of the Convolutional
Encoder
1
0
1
1
0
0
1
0
0
1
0
TxDATA
1
0
0
1
0
1
1
1
1
Precoder
Output
+1
0
−1 +1
0
0
+1
0
0
−1
Oversampled 2FSK
Encoder
Output
In oversampled 2FSK, there is no data clock from the TxRxCLK
pin. Instead, the transmit data at the TxRxDATA pin is sampled
at 32 times the programmed rate.
Another property of this encoding scheme is that the trans-
mitted symbol sequence is dc-free, which facilitates symbol
detection and frequency measurement in the receiver. In
addition, no code rate loss is associated with this three-level
convolutional encoder; that is, the transmitted symbol rate is
equal to the data rate presented at the transmit data input.
Oversampled 2FSK is the only modulation mode that can be
used with the UART mode interface for data transmission (see
the Interfacing to a Microcontroller/DSP section for more
information).
3FSK is selected by setting the MODULATION_SCHEME bits
(Register 2, Bits[DB6:DB4]) to 010. It can also be used with
raised cosine filtering to further increase the spectral efficiency
of the transmit signal.
SPECTRAL SHAPING
Gaussian or raised cosine filtering can be used to improve
transmit spectral efficiency. The ADF7021-V supports Gaussian
filtering (bandwidth time [BT] = 0.5) on 2FSK modulation.
Raised cosine filtering can be used with 2FSK, 3FSK, or 4FSK
modulation. The roll-off factor (alpha) of the raised cosine filter
has programmable options of 0.5 and 0.7. Both the Gaussian
and raised cosine filters are implemented using linear phase
digital filter architectures that deliver precise control over the
BT and alpha filter parameters, and guarantee a transmit spectrum
that is very stable over temperature and supply variation.
Four-Level Frequency Shift Keying (4FSK)
In 4FSK modulation, two bits per symbol spectral efficiency is
realized by mapping consecutive input bit-pairs in the Tx data
bit stream to one of four possible symbols (−3, −1, +1, +3). Thus,
the transmitted symbol rate is half the input bit rate. These
symbols are mapped to equally spaced discrete frequencies
centered on the RF carrier at
−3fDEV, −1fDEV, +1fDEV, and +3fDEV
Gaussian Frequency Shift Keying (GFSK)
where fDEV is programmed using the Tx_FREQUENCY_
DEVIATION bits (Bits[DB27:DB19] in Register 2) and is also
equal to half the frequency spacing between adjacent symbols.
Gaussian frequency shift keying reduces the bandwidth occupied
by the transmitted spectrum by digitally prefiltering the transmit
data. The BT product of the Gaussian filter used is 0.5.
By minimizing the separation between symbol frequencies,
4FSK can have high spectral efficiency. The bit-to-symbol
mapping for 4FSK is gray coded and is shown in Figure 43.
Gaussian filtering can be used only with 2FSK modulation. GFSK
is selected by setting Register 2, Bits[DB6:DB4] to 001.
Raised Cosine Filtering
Raised cosine filtering provides digital prefiltering of the transmit
data by using a raised cosine filter with a roll-off factor (alpha)
of either 0.5 or 0.7. The alpha is set to 0.5 by default, but the
raised cosine filter bandwidth can be increased to provide less
aggressive data filtering by using an alpha of 0.7 (set Register 2,
Bit DB30 to Logic 1). Raised cosine filtering can be used with
2FSK, 3FSK, and 4FSK modulation.
Tx DATA
0
0
0
1
1
0
1
1
f
+3fDEV
+fDEV
SYMBOL
FREQUENCIES
–fDEV
Raised cosine filtering is enabled by setting Register 2,
Bits[DB6:DB4] as shown in Table 10.
–3fDEV
t
Figure 43. 4FSK Bit-to-Symbol Mapping
Rev. 0 | Page 26 of 60
ADF7021-V
The figures for latency in Table 11 assume that the positive
MODULATION AND FILTERING OPTIONS
TxRxCLK edge is used to sample data (default). If the TxRxCLK
is inverted by setting Register 2, Bits[DB29:DB28], an additional
0.5 bit latency can be added to all values in Table 11.
The various modulation and data filtering options for the
ADF7021-V are described in Table 10.
Table 10. Modulation and Filtering Options
Table 11. Bit/Symbol Latency in Transmit Mode for Various
Modulation Schemes
Register 2,
Bits[DB6:DB4]
Modulation
Binary FSK
2FSK
Data Filtering
Modulation
Latency
2FSK
1 bit
None
None
None
000
000
000
MSK1
GFSK
4 bits
RC2FSK, alpha = 0.5
RC2FSK, alpha = 0.7
3FSK
5 bits
4 bits
OQPSK with Half Sine
Baseband Shaping2
1 bit
GFSK
Gaussian
Gaussian
Raised cosine
None
001
001
101
100
GMSK3
RC3FSK, alpha = 0.5
RC3FSK, alpha = 0.7
4FSK
RC4FSK, alpha = 0.5
RC4FSK, alpha = 0.7
5 bits
4 bits
RC2FSK
Oversampled 2FSK
Three-Level FSK
3FSK
1 symbol
5 symbols
4 symbols
None
010
110
RC3FSK
Raised cosine
TEST PATTERN GENERATOR
Four-Level FSK
4FSK
The ADF7021-V has a number of built-in test pattern generators
that can be used to facilitate radio link setup or RF measurement.
None
011
111
RC4FSK
Raised cosine
A full list of the supported test patterns is shown in Table 12.
The data rate for these test patterns is the programmed data rate
set in Register 3.
1 MSK is 2FSK modulation with a modulation index = 0.5.
2 Offset quadrature phase shift keying (OQPSK) with half sine baseband
shaping is spectrally equivalent to MSK.
3 GMSK is GFSK with a modulation index = 0.5.
The PN9 sequence is suitable for test modulation when carrying
out adjacent channel power (ACP) or occupied bandwidth
measurements.
TRANSMIT LATENCY
Transmit latency is the delay time from the sampling of a
bit/symbol by the TxRxCLK signal to when that bit/symbol
appears at the RF output. The latency without any data filtering
is 1 bit. The addition of data filtering adds a further latency as
indicated in Table 11.
Table 12. Transmit Test Pattern Generator Options
Register 15,
Bits[DB10:DB8]
Test Pattern
Normal
000
001
010
011
100
101
110
Transmit carrier only
Transmit +fDEV tone only
Transmit −fDEV tone only
Transmit 1010 pattern
Transmit PN9 sequence
Transmit SWD pattern repeatedly
It is important that the ADF7021-V be left in transmit mode
after the last data bit is sampled by the data clock to account for
this latency. The ADF7021-V should stay in transmit mode for
a time equal to the number of latency bit periods for the applied
modulation scheme. This ensures that all of the data sampled by
the TxRxCLK signal appears at RF.
Rev. 0 | Page 27 of 60
ADF7021-V
RECEIVER SECTION
If the AGC loop is disabled, the gain of the IF filter can be set to
one of three levels by using the FILTER_GAIN bits (Register 9,
Bits[DB23:DB22]). The filter gain is adjusted automatically if
the AGC loop is enabled.
RF FRONT END
The ADF7021-V is based on a fully integrated, low IF receiver
architecture. The low IF architecture facilitates a very low external
component count and does not suffer from powerline-induced
interference problems.
IF Filter Bandwidth and Center Frequency Calibration
To compensate for manufacturing tolerances, the IF filter should
be calibrated after power-up to ensure that the bandwidth and
center frequency are correct. Coarse and fine calibration schemes
are provided to offer a choice between fast calibration (coarse
calibration) and high filter centering accuracy (fine calibration).
Coarse calibration is enabled by setting Register 5, Bit DB4 high.
Fine calibration is enabled by setting Register 6, Bit DB4 high.
Figure 44 shows the structure of the receiver front end. The
many programming options allow users to trade off sensitivity,
linearity, and current consumption to best suit their application.
To achieve a high level of resilience against spurious reception,
the low noise amplifier (LNA) features a differential input.
Switch SW2 shorts the LNA input when transmit mode is
selected (Register 0, Bit DB27 = 0). This feature facilitates the
design of a combined LNA/PA matching network, avoiding the
need for an external Tx/Rx switch. See the LNA/PA Matching
section for details on the design of the matching network.
For details on when it is necessary to perform a filter calibration,
and in what applications to use either a coarse calibration or
fine calibration, see the IF Filter Bandwidth Calibration section.
I (TO FILTER)
RSSI/AGC
RFIN
Tx/Rx SELECT
(REG 0, BIT DB27)
The RSSI is implemented as a successive compression log amp
following the baseband (BB) channel filtering. The log amp
achieves 3 dB log linearity. It also doubles as a limiter to
convert the signal-to-digital levels for the FSK demodulator.
The offset correction circuit uses the BBOS_CLK_DIVIDE bits
(Bits DB5:DB4] in Register 3) and should be set between 1 MHz
and 2 MHz. The RSSI level is converted for user readback and
for digitally controlled AGC by an 80-level (7-bit) flash ADC.
This level can be converted to input power in dBm. By default,
the AGC is on when powered up in receive mode.
SW2 LNA
LO
RFIN
Q (TO FILTER)
LNA_MODE
(REG 9, BIT DB25)
MIXER_LINEARITY
(REG 9, BIT DB28)
LNA_BIAS
(REG 9, BITS[DB27:DB26])
LNA_GAIN
(REG 9, BITS[DB21:DB20])
LNA/MIXER_ENABLE
(REG 8, BIT DB6)
Figure 44. RF Front End
The LNA is followed by a quadrature downconversion mixer,
which converts the RF signal to the IF frequency of 100 kHz.
An important consideration is that the output frequency of the
synthesizer must be programmed to a value 100 kHz below the
center frequency of the received channel. The LNA has two
basic operating modes: high gain/low noise mode and low gain/
low power mode. To switch between these two modes, use the
LNA_MODE bit (Register 9, Bit DB25). The mixer is also config-
urable for either a low current mode or an enhanced linearity
mode using the MIXER_LINEARITY bit (Register 9, Bit DB28).
OFFSET
CORRECTION
FSK
DEMOD
1
A
A
A
LATCH
CLK
IFWR
IFWR
IFWR
IFWR
RSSI
ADC
R
Figure 45. RSSI Block Diagram
RSSI Thresholds
Based on the specific sensitivity and linearity requirements of
the application, it is recommended that the LNA_MODE bit and
the MIXER_LINEARITY bit be adjusted as shown in Table 14.
The gain of the LNA is configured by the LNA_GAIN bits
(Register 9, Bits[DB21:DB20]) and can be set by the user or by
the automatic gain control (AGC) logic.
When the RSSI is above AGC_HIGH_THRESHOLD (Register 9,
Bits[DB17:DB11]), the gain is reduced. When the RSSI is below
AGC_LOW_THRESHOLD (Register 9, Bits[DB10:DB4]), the
gain is increased. The thresholds default to 70 (high threshold)
and 30 (low threshold) on power-up in receive mode. A delay
(set by AGC_CLK_DIVIDE in Register 3, Bits[DB31:DB26]) is
programmed to allow for settling of the loop. A value of 33 is
recommended to give an AGC update rate of 3 kHz.
IF FILTER
IF Filter Settings
Out-of-band interference is rejected by means of a fifth-order
Butterworth polyphase IF filter centered on a frequency of
100 kHz. The bandwidth of the IF filter can be programmed to
9 kHz, 13.5 kHz, or 18.5 kHz in Register 4, Bits[DB31:DB30],
and should be chosen as a compromise between interference
rejection and attenuation of the desired signal.
The user has the option of changing the two threshold values
from the defaults of 70 and 30 (Register 9). The default AGC
setup values should be adequate for most applications. The
threshold values must be more than 30 apart for the AGC to
operate correctly.
Rev. 0 | Page 28 of 60
ADF7021-V
Offset Correction Clock
The total AFC settling time depends on the number of AGC
gain changes during reception of a packet. A total of five gain
changes gives a worst-case AGC settling time of 5 × 333 μs. To
allow for AGC settling, the preamble length should be adjusted
accordingly.
In Register 3, the user should set the BBOS_CLK_DIVIDE bits
(Bits[DB5:DB4]) to give a baseband offset clock (BBOS CLK)
frequency between 1 MHz and 2 MHz.
BBOS CLK (Hz) = XTAL/(BBOS_CLK_DIVIDE)
where BBOS_CLK_DIVIDE can be set to 4, 8, 16, or 32.
AGC Information and Timing
RSSI Formula (Converting to dBm)
The RSSI formula is
Input Power (dBm) = (−130 dBm + (Readback Code +
Gain Mode Correction)) × 0.5
AGC is selected by default and operates by setting the appro-
priate LNA and filter gain settings for the measured RSSI level.
To enter one of the LNA/mixer modes listed in Table 14, the
user can disable AGC by writing to Register 9. After each gain
change, the AGC loop waits for a programmed time to allow
transients to settle. This AGC update rate is set according to
where:
Readback Code is given by Bit RV7 to Bit RV1 in the readback
register (see Figure 57 and the Readback Format section).
Gain Mode Correction is given by the values in Table 13.
The LNA gain (LG2, LG1) and filter gain (FG2, FG1) values
are also obtained from the readback register, as part of an RSSI
readback.
SEQ _ CLK _ DIVIDE (Hz)
AGC Update Rate (Hz) =
AGC _CLK _ DIVIDE
where:
Table 13. Gain Mode Correction
SEQ_CLK_DIVIDE = 100 kHz (Register 3, Bits[DB25:DB18]).
AGC_CLK_DIVIDE is set by Register 3, Bits[DB31:DB26]. A
value of 33 is recommended.
LNA Gain
(LG2, LG1)
Filter Gain
(FG2, FG1)
Gain Mode
Correction
H (1, 0)
M (0, 1)
M (0, 1)
M (0, 1)
L (0, 0)
H (1, 0)
H (1, 0)
M (0, 1)
L (0, 0)
L (0, 0)
0
It is recommended that AGC_CLK_DIVIDE be set to a value of
33, which allows a settling time of 333 μs for each gain change.
By using the recommended setting for AGC_CLK_DIVIDE, the
total AGC settling time is
24
38
58
86
Number of AGC Gain Changes
AGC Settling Time (sec) =
An additional factor should be introduced to account for losses
in the front-end-matching network/antenna.
AGC Update Rate (Hz)
Table 14. LNA/Mixer Modes (Register 9 Settings)
Sensitivity (2FSK,
MIXER_LINEARITY Data Rate = 4.8 kbps, Rx Current Input IP3
LNA_MODE LNA_GAIN
(Bit DB25)
Receiver Mode
(Bits[DB21:DB20]) (Bit DB28)
fDEV = 4 kHz) (dBm)
Consumption (mA) (dBm)
High Sensitivity
Mode (Default)
0
30
30
0
1
−116.5
20.1
20.1
−24
−20
Enhanced Linearity,
High Gain
0
−113
Medium Gain
1
1
10
10
0
1
−108
−102
17.9
17.9
−13.5
−9
Enhanced Linearity,
Medium Gain
Low Gain
1
1
3
3
0
1
−99
−91
17.9
17.9
−5
−3
Enhanced Linearity,
Low Gain
Rev. 0 | Page 29 of 60
ADF7021-V
The quadrature outputs of the IF filter are first limited and then
fed to a digital frequency correlator that performs filtering and
frequency discrimination of the 2FSK/3FSK/4FSK spectrum.
DEMODULATION, DETECTION, AND CDR
System Overview
An overview of the demodulation, detection, and clock and
data recovery (CDR) of the received signal on the ADF7021-V
is shown in Figure 46.
For 2FSK modulation, data is recovered by comparing the
output levels from two correlators. The performance of this
frequency discriminator approximates that of a matched filter
detector, which is known to provide optimum detection in the
presence of additive white Gaussian noise (AWGN). This
method of FSK demodulation provides approximately 3 dB to
4 dB better sensitivity than a linear demodulator.
LIMITERS
I
CORRELATOR
DEMODULATOR
Q
MUX
LINEAR
DEMODULATOR
Linear Demodulator
Figure 48 shows a block diagram of the linear demodulator.
I
LEVEL
THRESHOLD
IF
DETECTION
+
TxRxDATA
TxRxCLK
2FSK/3FSK/4FSK
2FSK RxDATA
LIMITERS
Q
SLICER
2FSK
CLOCK
AND
DATA
MUX
FREQUENCY
RECOVERY
LINEAR
DISCRIMINATOR
VITERBI
DETECTION
3FSK
RxCLK
REG 4, BITS[DB29:DB20]
FREQUENCY
READBACK
AND AFC LOOP
Figure 46. Overview of Demodulation, Detection, and CDR Process
Figure 48. Block Diagram of Linear FSK Demodulator
The quadrature outputs of the IF filter are first limited and
then fed to either the correlator FSK demodulator or to the
linear FSK demodulator. The correlator demodulator is used
to demodulate 2FSK, 3FSK, and 4FSK. The linear demodulator
is used for frequency measurement and is enabled when the
AFC loop is active. The linear demodulator can also be used
to demodulate 2FSK.
A digital frequency discriminator provides an output signal that
is linearly proportional to the frequency of the limiter outputs.
The discriminator output is filtered and averaged using a com-
bined averaging filter and envelope detector. The demodulated
2FSK data from the postdemodulator filter is recovered by
slicing against the output of the envelope detector, as shown in
Figure 48. This method of demodulation corrects for frequency
errors between the transmitter and receiver when the received
spectrum is close to or within the IF bandwidth. This envelope
detector output is also used for AFC readback and provides the
frequency estimate for the AFC control loop.
Following the demodulator, a digital postdemodulator filter
removes excess noise from the demodulator signal output.
Threshold/slicer detection is used for data recovery of 2FSK and
4FSK. Data recovery of 3FSK can be implemented using either
threshold detection or Viterbi detection.
Postdemodulator Filter
An on-chip CDR PLL is used to resynchronize the received bit
stream to a local clock. It outputs the retimed data and clock on
the TxRxDATA and TxRxCLK pins, respectively.
A second-order, digital low-pass filter removes excess noise from
the demodulated bit stream at the output of the discriminator.
The bandwidth of this postdemodulator filter is programmable
and must be optimized for the user’s data rate and the received
modulation type. If the bandwidth is too narrow, performance
degrades due to intersymbol interference (ISI). If the bandwidth
is too wide, excess noise degrades the performance of the receiver.
The POST_DEMOD_BW bits (Register 4, Bits[DB29:DB20])
set the bandwidth of this filter.
Correlator Demodulator
The correlator demodulator can be used for 2FSK, 3FSK, and
4FSK demodulation. Figure 47 shows the operation of the
correlator demodulator for 2FSK.
FREQUENCY CORRELATOR
DISCRIM BW
I
OUTPUT LEVELS:
2FSK = +1, –1
3FSK = +1, 0, –1
LIMITERS
Q
4FSK = +3, +1, –1, –3
IF
IF – fDEV
IF + fDEV
REG 4, BITS[DB9:DB8]
Rx_INVERT
REG 4, BITS[DB19:DB10]
DISCRIMINATOR_BW
REG 4, BIT DB7
DOT_PRODUCT
Figure 47. 2FSK Correlator FSK Demodulator Operation
Rev. 0 | Page 30 of 60
ADF7021-V
2FSK Bit Slicer/Threshold Detection
When used with Viterbi detection, the receiver sensitivity
for 3FSK is typically 3 dB greater than that obtained using
threshold detection. When the Viterbi detector is enabled,
however, the receiver bit latency is increased by twice the
Viterbi path memory length.
2FSK demodulation can be implemented using the correlator
FSK demodulator or the linear FSK demodulator. In both cases,
threshold detection is used for data recovery at the output of the
postdemodulator filter.
Clock and Data Recovery (CDR)
The output signal levels of the correlator demodulator are always
centered about 0. Therefore, the slicer threshold level can be
fixed at 0, and the demodulator performance is independent of
the run-length constraints of the transmit data bit stream. This
results in robust data recovery that does not suffer from the
classic baseline wander problems that exist in more traditional
FSK demodulators.
An oversampled digital clock and data recovery (CDR) PLL is
used to resynchronize the received bit stream to a local clock in
all modulation modes. The oversampled clock rate of the PLL
(CDR CLK) must be set at 32 times the symbol rate (see the
Register 3—Transmit/Receive Clock Register section). The maxi-
mum data/symbol rate tolerance of the CDR PLL is determined
by the number of zero-crossing symbol transitions in the trans-
mitted packet. For example, if using 2FSK with a 101010 preamble,
a maximum tolerance of 3.0% of the data rate is achieved.
However, this tolerance is reduced during recovery of the
remainder of the packet, where symbol transitions may not be
guaranteed to occur at regular intervals. To maximize the data
rate tolerance of the CDR, some form of encoding and/or data
scrambling is recommended that guarantees a number of
transitions at regular intervals.
When the linear demodulator is used for 2FSK demodulation,
the output of the envelope detector is used as the slicer threshold,
and this output tracks frequency errors that are within the IF
filter bandwidth.
3FSK and 4FSK Threshold Detection
4FSK demodulation is implemented using the correlator
demodulator followed by the postdemodulator filter and
threshold detection. The output of the postdemodulator filter
is a four-level signal that represents the transmitted symbols
(−3, −1, +1, +3). Threshold detection of 4FSK requires three
threshold settings: one that is always fixed at 0 and two that are
programmable and are symmetrically placed above and below 0
using the 3FSK/4FSK_SLICER_THRESHOLD bits (Register 13,
Bits[DB10:DB4]).
For example, using 2FSK with Manchester-encoded data
achieves a data rate tolerance of 2.0%.
The CDR PLL is designed for fast acquisition of the recovered
symbols during preamble and typically achieves bit synchro-
nization within five-symbol transitions of preamble.
In 4FSK modulation, the tolerance using the +3, −3, +3, −3
preamble is 3% of the symbol rate (or 1.5% of the data rate).
However, this tolerance is reduced during recovery of the
remainder of the packet, where symbol transitions may not
be guaranteed to occur at regular intervals. To maximize the
symbol/data rate tolerance of the CDR, the remainder of the
4FSK packet should be constructed so that the transmitted
symbols retain close to dc-free properties by using data scram-
bling and/or by inserting specific dc-balancing symbols into the
transmitted bit stream at regular intervals, such as after every
8 or 16 symbols.
3FSK demodulation is implemented using the correlator demod-
ulator, followed by a postdemodulator filter. The output of the
postdemodulator filter is a three-level signal that represents the
transmitted symbols (−1, 0, +1). Data recovery of 3FSK can be
implemented using threshold detection or Viterbi detection.
Threshold detection is implemented using two thresholds that
are programmable and are symmetrically placed above and
below 0 using the 3FSK/4FSK_SLICER_THRESHOLD bits
(Register 13, Bits[DB10:DB4]).
3FSK Viterbi Detection
Viterbi detection of 3FSK operates on a four-state trellis and is
implemented using two interleaved Viterbi detectors operating
at half the symbol rate. The Viterbi detector is enabled by
Register 13, Bit DB11.
In 3FSK modulation, the linear convolutional encoder scheme
guarantees that the transmitted symbol sequence is dc-free,
facilitating symbol detection. However, Tx data scrambling is
recommended to limit the run length of 0 symbols in the
transmit bit stream. Using 3FSK, the CDR data rate tolerance is
typically 0.5%.
To facilitate different run-length constraints in the transmitted
bit stream, the Viterbi path memory length is programmable in
steps of 4 bits, 6 bits, 8 bits, or 32 bits by setting the VITERBI_
PATH_MEMORY bits (Register 13, Bits[DB14:DB13]). This
value should be set equal to or greater than the maximum
number of consecutive 0s in the interleaved transmit bit stream.
Rev. 0 | Page 31 of 60
ADF7021-V
Table 17. Assignment of Correlator K Value for 4FSK
RECEIVER SETUP
Correlator Demodulator Setup
K
Register 4, Bit DB7 Register 4, Bits[DB9:DB8]
Even
Odd
0
1
00
00
To enable the correlator for various modulation modes, see
Table 15.
Linear Demodulator Setup
Table 15. Enabling the Correlator Demodulator
The linear demodulator can be used for 2FSK demodulation. To
enable the linear demodulator, set the DEMOD_SCHEME bits
(Register 4, Bits[DB6:DB4]) to 000.
DEMOD_SCHEME
(Register 4, Bits[DB6:DB4])
Received Modulation
2FSK
3FSK
4FSK
001
010
011
Postdemodulator Filter Setup
The 3 dB bandwidth of the postdemodulator filter should be
set according to the received modulation type and data rate.
The bandwidth is controlled by Register 4, Bits[DB29:DB20]
and is given by
To optimize receiver sensitivity, the correlator bandwidth must
be optimized for the specific deviation frequency and modula-
tion used by the transmitter. The discriminator bandwidth is
controlled by Register 4, Bits[DB19:DB10], and is defined as
2
11 ×π× fCUTOFF
POST _ DEMOD _ BW =
DEMOD CLK ×K
DEMOD CLK
DISCRIMINATOR_ BW =
400 ×103
where fCUTOFF is the target 3 dB bandwidth in Hz of the post-
demodulator filter.
where:
DEMOD CLK is as defined in the Register 3—Transmit/Receive
Clock Register section.
K is set for each modulation mode as follows:
Table 18. Postdemodulator Filter Bandwidth Settings for
2FSK/3FSK/4FSK Modulation Schemes
Received
Modulation
Postdemodulator Filter Bandwidth,
fCUTOFF (Hz)
For 2FSK,
3
2FSK
3FSK
4FSK
0.75 × data rate
1 × data rate
1.6 × symbol rate (0.8 × data rate)
⎛
⎜
⎜
⎝
⎞
⎟
⎟
⎠
100×10
K = Round
For 3FSK,
K = Round
fDEV
3FSK Viterbi Detector Setup
3
⎛
⎜
⎜
⎝
⎞
⎟
⎟
⎠
100×10
The Viterbi detector can be used for 3FSK data detection; it is
activated by setting Register 13, Bit DB11, to Logic 1.
2× fDEV
The Viterbi path memory length is programmable in steps of
4, 6, 8, or 32 bits (VITERBI_PATH_MEMORY, Register 13,
Bits[DB14:DB13]). The path memory length should be set
equal to or greater than the maximum number of consecutive
0s in the interleaved transmit bit stream.
For 4FSK,
K = Round4FSK
where:
3
⎛
⎜
⎜
⎝
⎞
⎟
⎟
⎠
100 ×10
4 × fDEV
Round is rounded to the nearest integer.
Round4FSK is rounded to the nearest of the following integers:
32, 31, 28, 27, 24, 23, 20, 19, 16, 15, 12, 11, 8, 7, 4, 3.
The Viterbi detector also uses threshold levels to implement the
maximum likelihood detection algorithm. These thresholds are
programmable via the 3FSK/4FSK_SLICER_THRESHOLD bits
(Register 13, Bits[DB10:DB4]).
f
DEV is the transmit frequency deviation in Hz. For 4FSK, fDEV
is the frequency deviation used for the 1 symbols (that is, the
inner frequency deviations).
These bits are assigned as follows:
3FSK/4FSK_SLICER_THRESHOLD =
To optimize the coefficients of the correlator, Register 4, Bit DB7
and Register 4, Bits[DB9:DB8] must also be assigned. The value
of these bits depends on whether K is odd or even. These bits
are assigned according to Table 16 and Table 17.
⎛
⎜
⎜
⎝
⎞
⎟
⎟
⎠
Tx _ FREQUENCY _ DEVIATION × K
100 × 103
57 ×
where K is the value calculated for correlator discriminator
bandwidth.
Table 16. Assignment of Correlator K Value for 2FSK and 3FSK
Register 4, Register 4,
(K + 1)/2 Bit DB7
3FSK Threshold Detector Setup
K
K/2
Bits[DB9:DB8]
To activate threshold detection of 3FSK, Register 13, Bit DB11,
should be set to Logic 0. The 3FSK/4FSK_SLICER_THRESHOLD
bits (Register 13, Bits[DB10:DB4]) should be set as described in
the 3FSK Viterbi Detector Setup section.
Even
Even
Odd
Odd
Even
Odd
N/A
N/A
N/A
N/A
Even
Odd
0
0
1
1
00
10
00
10
Rev. 0 | Page 32 of 60
ADF7021-V
3FSK CDR Setup
11001100…) can also be used, but result in a longer synchro-
nization time of the received bit stream in the receiver. The
preamble must allow enough bits for AGC settling of the
receiver and CDR acquisition (see Table 20).
In 3FSK, a transmit preamble of at least 40 bits of continuous
1s is recommended to ensure a maximum number of symbol
transitions for the CDR to acquire lock.
The remaining fields that follow the preamble do not need to
use dc-free coding. For these fields, the ADF7021-V can accom-
modate coding schemes with a run length of greater than eight
bits without any performance degradation. Refer to the AN-915
Application Note for more information.
The clock and data recovery for 3FSK requires a number of
parameters in Register 13 to be set (see Table 19).
4FSK Threshold Detector Setup
The threshold for the 4FSK detector is set using the
3FSK/4FSK_SLICER_THRESHOLD bits (Register 13,
Bits[DB10:DB4]). The threshold should be set as follows:
4FSK Preamble and Data Coding
The recommended preamble bit pattern for 4FSK is a repeating
00100010… bit sequence. This two-level sequence of repeating
−3, +3, −3, +3 symbols is dc-free and maximizes the symbol
timing performance and data recovery of the 4FSK preamble
in the receiver. The minimum recommended length of the
preamble is 32 bits (16 symbols).
3FSK/4FSK_SLICER_THRESHOLD =
⎛
⎜
⎜
⎝
⎞
⎟
⎟
⎠
4FSK Outer Tx Deviation × K
100 × 103
78 ×
where K is the value calculated for correlator discriminator
bandwidth.
The remainder of the 4FSK packet should be constructed so
that the transmitted symbols retain close to a dc-free balance by
using data scrambling and/or by inserting specific dc-balancing
symbols in the transmitted bit stream at regular intervals, such
as after every 8 or 16 symbols.
FSK DEMODULATOR OPTIMIZATION
2FSK Preamble
The recommended preamble bit pattern for 2FSK, GFSK, and
RC2FSK is a dc-free pattern (such as a 10101010… pattern).
Preamble patterns with longer run-length constraints (such as
Table 19. 3FSK CDR Settings
Parameter (Register 13)
Recommended Setting
Purpose
PHASE_CORRECTION (Bit DB12)
3FSK_CDR_THRESHOLD (Bits[DB21:DB15])
1
Phase correction is on
Sets CDR decision threshold levels
⎛
⎜
⎜
⎝
⎞
⎟
⎟
⎠
Tx _ FREQUENCY _ DEVIATION × K
100 ×103
62×
where K is the value calculated for correlator
discriminator bandwidth.
3FSK_PREAMBLE_TIME_VALIDATE (Bits[DB25:DB22])
15
Preamble detector time qualifier
Table 20. Preamble Bit Length for 2FSK Modulation
Sensitivity Degradation
from Specifications
Rx Frequency Error
Tolerance (1% PER)
Demodulator
Correlator (AFC off)
Mod index = 2
Mod index = 1
Mod index = 0.5
Linear (AFC off)
Minimum Preamble (Bits)
0 dB
0 dB
0 dB
30% × fDEV
25% × fDEV
20% × fDEV
16
16
16
f
f
f
DEV = 4.2 kHz
DEV = 2.2 kHz
DEV = 1.6 kHz
3 dB
3 dB
3 dB
0.5 × IFBW1
64
112
128
0.5 × IFBW1
0.5 × IFBW1
Correlator (AFC on)
Linear (AFC on)
Correlator + bypass CDR (AFC off)
2 dB
3 dB
2 dB to 3 dB4
AFC pull-in range2
AFC pull-in range2
50% × fDEV
96 to 1283
96 to 128
8
5
1 This value is generally true; however, some sensitivity degradation may occur close to the edge of the IF filter.
2 Limited to 0.5 × IFBW or AFC pull-in range, whichever is less.
3 Dependent on modulation index and fDEV. At higher modulation indexes (1.0 or greater) and higher fDEV (>4.0 kHz), the minimum preamble length is 96 bits. The
minimum preamble length increases as the modulation index and fDEV are reduced.
4 Dependent on the performance of the symbol timing recovery module on the external microcontroller.
5 Depends on the pulse width mark/space ratio of Logic 1 to Logic 0 that the symbol timing recovery scheme on the external microcontroller can tolerate. In this mode,
the mark/space ratio of the recovered bit stream increases with frequency error. In the absence of frequency error, the mark/space ratio is 50:50, that is, the width of a
Logic 1 is the same as the width of a Logic 0.
Rev. 0 | Page 33 of 60
ADF7021-V
Correlator Demodulator and Low Modulation Indexes
Internal AFC
The modulation index in 2FSK is defined as
The ADF7021-V supports a real-time, internal, automatic
frequency control loop. In this mode, an internal control loop
automatically monitors the frequency error and adjusts the
synthesizer-N divider using an internal proportional integral
(PI) control loop.
2× fDEV
Modulation Index =
Data Rate
The receiver sensitivity performance and receiver frequency
tolerance can be maximized at low modulation indexes by
increasing the discriminator bandwidth of the correlator
demodulator. For modulation indexes of less than 0.4, it is
recommended that the correlator bandwidth be doubled by
calculating K as follows:
The internal AFC control loop parameters are controlled in
Register 10. The internal AFC loop is activated by setting
Bit DB4 in Register 10 to 1. A scaling coefficient must also be
entered, based on the crystal frequency in use. This is set up
using Bits[DB16:DB5] in Register 10 and should be calculated
as follows:
1003
⎛
⎞
⎜
⎜
⎟
⎟
K = Round
24
2× fDEV
⎛
⎜
⎜
⎝
⎞
⎟
⎟
⎠
⎝
⎠
2
× 500
AFC _ SCALING _ FACTOR = Round
XTAL
The DISCRIMINATOR_BW value in Register 4 should be
recalculated using the new K value. Figure 29 illustrates the
improved sensitivity that can be achieved for 2FSK modulation,
at low modulation indexes, by doubling the correlator bandwidth.
Maximum AFC Range
The maximum frequency correction range of the AFC loop
is programmable using Register 10, Bits[DB31:DB24]. The
maximum AFC correction range is the difference in frequency
between the upper and lower limits of the AFC tuning range.
For example, if the maximum AFC correction range is set to
10 kHz, the AFC can adjust the receiver LO within the fLO
5 kHz range.
AFC OPERATION
The ADF7021-V also supports a real-time AFC loop that is
used to remove frequency errors due to mismatches between
the transmit and receive crystals/TCXOs. The AFC loop uses
the linear frequency discriminator block to estimate frequency
errors. The linear FSK discriminator output is filtered and
averaged to remove the FSK frequency modulation using a
combined averaging filter and envelope detector. In receive
mode, the output of the envelope detector provides an estimate
of the average IF frequency.
However, when RF_DIVIDE_BY_2 (Register 1, Bit DB18) is
enabled, the programmed range is halved. The user should
account for this halving by doubling the programmed maxi-
mum AFC range.
The recommended maximum AFC correction range should be
≤1.5 × IF filter bandwidth. If the maximum frequency correction
range is set to be >1.5 × IF filter bandwidth, the attenuation of
the IF filter can degrade the AFC loop sensitivity.
The two methods of AFC supported on the ADF7021-V are
external AFC and internal AFC.
External AFC
With external AFC, the user reads back the frequency infor-
mation through the ADF7021-V serial port and applies a
frequency correction value to the synthesizer-N divider.
The adjacent channel rejection (ACR) performance of the receiver
can be degraded when AFC is enabled and the AFC correction
range is close to or greater than the IF filter bandwidth. However,
because the AFC correction range is programmable, the user
can trade off AFC correction range and ACR performance of
the receiver.
The frequency information is obtained by reading the
signed, 16-bit AFC readback value, as described in the
Readback Format section, and by applying the following
formula:
When AFC errors are removed using either the internal or
external AFC, further improvement in receiver sensitivity can
be obtained by reducing the IF filter bandwidth using the
IF_FILTER_BW bits (Register 4, Bits[DB31:DB30]).
Frequency Readback (Hz) = (AFC READBACK ×
DEMOD CLK)/218
Although the AFC readback value is a signed number, under
normal operating conditions, it is positive. In the absence of
frequency errors, the frequency readback value is equal to the
IF frequency of 100 kHz.
Rev. 0 | Page 34 of 60
ADF7021-V
The SWD pin status can be configured by setting Bits[DB7:DB6]
in Register 12. Bits[DB5:DB4] in Register 11 are used to set the
length of the sync/ID word, which can be 12, 16, 20, or 24 bits
long. A value of 24 bits is recommended to minimize false sync
word detection in the receiver that can occur during recovery of
the remainder of the packet or when a noise/no signal is present
at the receiver input. The transmitter must transmit the sync
byte MSB first, LSB last to ensure proper alignment in the
receiver sync-byte-detection hardware.
AUTOMATIC SYNC WORD DETECTION (SWD)
The ADF7021-V also supports automatic detection of the sync
or ID fields. To activate this mode, the sync (or ID) word must
be preprogrammed into the ADF7021-V. In receive mode, this
preprogrammed word is compared to the received bit stream.
When a valid match is identified, the external SWD pin is
asserted by the ADF7021-V on the next Rx clock pulse.
This feature can be used to alert the microprocessor that a
valid channel has been detected. It relaxes the computational
requirements of the microprocessor and reduces the overall
power consumption.
An error tolerance parameter can also be programmed that
accepts a valid match when up to three bits of the word are
incorrect. The error tolerance value is assigned in Register 11,
Bits[DB7:DB6].
The SWD signal can also be used to frame the received packet
by staying high for a preprogrammed number of bytes. The data
packet length can be set in Register 12, Bits[DB15:DB8].
Rev. 0 | Page 35 of 60
ADF7021-V
APPLICATIONS INFORMATION
Lower Tone Frequency (kHz) =
IF FILTER BANDWIDTH CALIBRATION
XTAL
The IF filter should be calibrated on every power-up in receive
mode to correct for errors in the bandwidth and filter center
frequency due to process variations. The automatic calibration
requires no external intervention when it is initiated by a write
to Register 5. Depending on numerous factors, such as IF filter
bandwidth, received signal bandwidth, and temperature varia-
tion, the user must determine whether to carry out a coarse
calibration or a fine calibration.
IF_CAL_LOWER_TONE_DIVIDE × 2
Upper Tone Frequency (kHz) =
XTAL
IF_CAL_UPPER_TONE_DIVIDE × 2
It is recommended that the lower tone and the upper tone be set
as shown in Table 22.
The performance of both calibration methods is shown in
Table 21.
Table 22. IF Filter Fine Calibration Tone Frequencies
IF Filter
Bandwidth (kHz)
Lower Tone
Frequency (kHz)
Upper Tone
Frequency (kHz)
Table 21. IF Filter Calibration Specifications
Filter Calibration
Method
Center Frequency
Calibration
Time (Typ)
9
13.5
18.5
78.1
79.4
78.1
116.3
116.3
119
Accuracy1
Coarse Calibration
Fine Calibration
100 kHz 2.5 kHz
100 kHz 0.6 kHz
200 μs
8.2 ms
Because the filter attenuation is slightly asymmetrical, it is
necessary to have a small offset in the filter center frequency to
provide near equal rejection at the upper and lower adjacent
channels. The calibration tones listed in Table 22 provide this
small positive offset in the IF filter center frequency.
1 After calibration.
Calibration Setup
IF filter calibration is initiated by writing to Register 5 and
setting the IF_CAL_COARSE bit (Bit DB4). This initiates a
coarse filter calibration. If the IF_FINE_CAL bit (Register 6,
Bit DB4) has already been set high, the coarse calibration is
followed by a fine calibration; otherwise, the calibration ends.
In some applications, an offset may not be required, and the
user may wish to center the IF filter at 100 kHz exactly. In this
case, the user can alter the tone frequencies from those given in
Table 22 to adjust the fine calibration result.
When initiated by writing to the part, calibration is performed
automatically without user intervention. The calibration time is
200 μs for coarse calibration and 8.2 ms for fine calibration, during
which time the ADF7021-V should not be accessed. The IF
filter calibration logic requires that the IF_FILTER_DIVIDER
bits (Register 5, Bits[DB13:DB5]) be set such that
The calibration algorithm adjusts the filter center frequency and
measures the RSSI 10 times during the calibration. The time for
an adjustment plus RSSI measurement is given by
IF_CAL_DWELL_TIME
IF Tone Calibration Time =
SEQ CLK
XTAL(Hz)
= 50 kHz
It is recommended that the IF tone calibration time be at least
800 μs. The total time for the IF filter fine calibration is given by
IF _ FILTER _ DIVIDER
The fine calibration uses two internally generated tones at certain
offsets around the IF filter. The two tones are attenuated by the
IF filter, and the level of this attenuation is measured using the
RSSI. The filter center frequency is adjusted to allow equal
attenuation of both tones. The attenuation of the two test tones
is then remeasured. This process continues for a maximum of
10 RSSI measurements, at which point the calibration algorithm
sets the IF filter center frequency to within 0.6 kHz of 100 kHz.
IF Filter Fine Calibration Time = IF Tone Calibration Time × 10
When to Use Coarse Calibration
It is recommended that a coarse calibration be performed on
every power-up in receive mode. This calibration typically takes
200 μs. The FILTER_CAL_COMPLETE signal from MUXOUT
(set using Bits[DB31:DB29] in Register 0) can be used to monitor
the filter calibration duration or to signal the end of calibration.
The ADF7021-V should not be accessed during calibration.
The frequency of these tones is set in Register 6 by the
IF_CAL_LOWER_TONE_DIVIDE bits (Bits[DB12:DB5]) and
the IF_CAL_UPPER_TONE_DIVIDE bits (Bits[DB20:DB13]),
as shown in the following equations.
Rev. 0 | Page 36 of 60
ADF7021-V
When to Use Fine Calibration
LNA/PA MATCHING
In cases where the receive signal bandwidth is very close to the
bandwidth of the IF filter, it is recommended that a fine filter
calibration be performed every time that the unit powers up in
receive mode.
The ADF7021-V exhibits optimum performance in terms of
sensitivity, transmit power, and current consumption only if its
RF input and output ports are properly matched to the antenna
impedance. For cost-sensitive applications, the ADF7021-V is
equipped with an internal Tx/Rx switch that facilitates the use
of a simple, combined passive LNA/PA matching network.
Alternatively, an external Tx/Rx switch such as the ADG919 can
be used, which yields a slightly improved receiver sensitivity
and lower transmitter power consumption.
A fine calibration should be performed if
OBW + Coarse Calibration Variation > IF_FILTER_BW
where:
OBW is the 99% occupied bandwidth of the transmit signal.
Coarse Calibration Variation is 2.5 kHz.
IF_FILTER_BW is set by Register 4, Bits[DB31:DB30].
Internal Tx/Rx Switch
Figure 49 shows the ADF7021-V in a configuration where
the internal Tx/Rx switch is used with a combined LNA/PA
matching network. This is the configuration used on the
EVAL-ADF7021-VDBxZ evaluation board. For most applica-
tions, the slight performance degradation of 1 dB to 2 dB
caused by the internal Tx/Rx switch is acceptable, allowing
the user to take advantage of the cost-saving potential of this
solution. The design of the combined matching network must
compensate for the reactance presented by the networks in the
Tx and the Rx paths, taking the state of the Tx/Rx switch into
consideration.
The FILTER_CAL_COMPLETE signal from MUXOUT (set by
Register 0, Bits[DB31:DB29]) can be used to monitor the filter
calibration duration or to signal the end of calibration. A coarse
filter calibration is automatically performed prior to a fine filter
calibration.
When to Use Single Fine Calibration
In applications where the receiver powers up numerous times
in a short period, it is necessary to perform fine calibration only
once, on the initial power-up in receive mode.
After the initial coarse calibration and fine calibration, the result
of the fine calibration can be read back through the serial inter-
face using the FILTER_CAL_READBACK result (see the Filter
Bandwidth Calibration Readback section). On subsequent
power-ups in receive mode, the filter is manually adjusted using
the previous fine filter calibration result. This manual adjustment
is performed using the IF_FILTER_ADJUST bits (Register 5,
Bits[DB19:DB14]).
V
BAT
ADF7021-V
L1
C1
RFOUT
PA
ANTENNA
Z _PA
OPT
OPTIONAL
BPF OR LPF
Z
_RFIN
IN
C
A
RFIN
L
LNA
A
This method should only be used if the successive power-ups in
receive mode are over a short duration, during which time there
is little variation in temperature (<15ꢀC).
RFIN
Z
_RFIN
IN
C
B
IF Filter Variation with Temperature
When calibrated, the filter center frequency can vary with changes
in temperature. If the ADF7021-V is used in an application where
it remains in receive mode for a considerable length of time, the
user must consider this variation of filter center frequency with
temperature. This variation is typically 1 kHz per 20ꢀC, which
means that if a coarse filter calibration and fine filter calibration
are performed at 25ꢀC, the initial maximum error is 0.5 kHz,
and the maximum possible change in the filter center frequency
over temperature (−40ꢀC to +85ꢀC) is 3.25 kHz. This gives a
total error of 3.75 kHz.
Figure 49. ADF7021-V with Internal Tx/Rx Switch
The procedure typically requires several iterations until an accept-
able compromise is reached. The successful implementation of
a combined LNA/PA matching network for the ADF7021-V is
critically dependent on the availability of an accurate electrical
model for the PCB. In this context, the use of a suitable CAD
package is strongly recommended. To avoid this effort, a small
form-factor reference design for the ADF7021-V is provided,
including matching and harmonic filter components. The design is
on a 4-layer PCB. Gerber files are available at www.analog.com.
If the receive signal occupied bandwidth is considerably
narrower than the IF filter bandwidth, the variation of filter
center frequency over the operating temperature range may
not be an issue. However, if the IF filter bandwidth is not wide
enough to tolerate the variation with temperature, a periodic
filter calibration can be performed or, alternatively, the on-chip
temperature sensor can be used to determine when a filter cali-
bration is necessary by monitoring for changes in temperature.
Rev. 0 | Page 37 of 60
ADF7021-V
External Tx/Rx Switch
performance of the receiver is dependent on how well matched
the I and Q signals are in amplitude and how well matched the
quadrature is between them (that is, how close to 90ꢀ apart they
are). The uncalibrated image rejection performance is approxi-
mately 29 dB (at 460 MHz). However, it is possible to improve
this performance by as much as 20 dB by finding the optimum
I/Q gain and phase adjust settings.
Figure 50 shows a configuration using an external Tx/Rx switch.
This configuration allows independent optimization of the
matching and filter network in the transmit and receive paths.
Therefore, it is more flexible and less difficult to design than the
configuration using the internal Tx/Rx switch. The PA is biased
through Inductor L1, whereas C1 blocks dc current. Together,
L1 and C1 form the matching network that transforms the source
impedance into the optimum PA load impedance, ZOPT_PA.
Calibration Using Internal RF Source
With the LNA powered off, an on-chip generated, low level RF
tone is applied to the mixer inputs. The LO is adjusted to make
the tone fall at the image frequency where it is attenuated by the
image rejection of the IF filter. The power level of this tone is
then measured using the RSSI readback. The I/Q gain and phase
adjust DACs (Register 5, Bits[DB31:DB20]) are adjusted and the
RSSI is remeasured. This process is repeated until the optimum
values for the gain and phase adjust are found that provide the
lowest RSSI readback level, thereby maximizing the image
rejection performance of the receiver.
V
BAT
ADF7021-V
L1
ADG919
C1
RFOUT
OPTIONAL
LPF
PA
ANTENNA
Z
_PA
OPT
Z
_RFIN
IN
C
A
OPTIONAL
BPF
RFIN
RFIN
(SAW)
L
LNA
A
Using the internal RF source, the RF frequencies that can be
used for image calibration are programmable and are odd
multiples of the reference frequency.
Z
_RFIN
IN
C
B
Rx/Tx – SELECT
Figure 50. ADF7021-V with External Tx/Rx Switch
Calibration Using External RF Source
ZOPT_PA depends on various factors, such as the required out-
put power, the frequency range, the supply voltage range, and
the temperature range. Selecting an appropriate ZOPT_PA helps
to minimize the Tx current consumption in the application.
The AN-764 Application Note and the AN-859 Application
Note contain a number of ZOPT_PA values for representative
conditions. Under certain conditions, however, it is recom-
mended that a suitable ZOPT_PA value be obtained by means
of a load-pull measurement.
IR calibration can also be implemented using an external RF
source. The IR calibration procedure is the same as that used for
the internal RF source, except that an RF tone is applied to the
LNA input.
Calibration Procedure and Setup
The IR calibration algorithm available from Analog Devices, Inc., is
based on a low complexity, 2D optimization algorithm that can
be implemented in an external microprocessor or microcontroller.
Due to the differential LNA input, the LNA matching network
must be designed to provide both a single-ended-to-differential
conversion and a complex, conjugate impedance match. The net-
work with the lowest component count that can satisfy these
requirements is the configuration shown in Figure 50, consist-
ing of two capacitors and one inductor.
To enable the internal RF source, the IR_CAL_SOURCE_DRIVE_
LEVEL bits (Register 6, Bits[DB29:DB28]) should be set to the
maximum level. The LNA should be set to its minimum gain
setting, and the AGC should be disabled if the internal RF source
is being used. Alternatively, an external RF source can be used.
The magnitude of the phase adjust is set using the IR_PHASE_
ADJUST_MAG bits (Register 5, Bits[DB23:DB20]). This correc-
tion can be applied to either the I or Q channel, depending on
the value of the IR_PHASE_ADJUST_DIRECTION bit
(Register 5, Bit DB24).
Depending on the antenna configuration, the user may need a
harmonic filter at the PA output to satisfy the spurious emission
requirement of the applicable government regulations. The har-
monic filter can be implemented in various ways, for example, a
discrete LC pi or T-stage filter. The immunity of the ADF7021-V
to strong out-of-band interference can be improved by adding a
band-pass filter in the Rx path. Alternatively, the ADF7021-V
blocking performance can be improved by selecting one of the
enhanced linearity modes, as described in Table 14.
The magnitude of the I/Q gain is adjusted using the IR_GAIN_
ADJUST_MAG bits (Register 5, Bits[DB29:DB25]). This correction
can be applied to either the I or Q channel, depending on the
value of the IR_GAIN_ADJUST_I/Q bit (Register 5, Bit DB30),
whereas the IR_GAIN_ADJUST_UP/DN bit (Register 5,
Bit DB31) sets whether the gain adjustment defines a gain or
an attenuation adjust.
IMAGE REJECTION CALIBRATION
The image channel in the ADF7021-V is 200 kHz below the
desired signal. The polyphase filter rejects this image with an
asymmetric frequency response. The image rejection (IR)
Rev. 0 | Page 38 of 60
ADF7021-V
ADF7021-V
RFIN
RFIN
LNA
POLYPHASE
IF FILTER
RSSI/
MUX
LOG AMP
INTERNAL
SIGNAL
SOURCE
7-BIT
ADC
PHASE ADJUST
Q
I
FROM LO
SERIAL
INTERFACE
4
PHASE ADJUST
REGISTER 5
RSSI READBACK
4
GAIN ADJUST
REGISTER 5
MICROCONTROLLER
I/Q GAIN/PHASE ADJUST AND
RSSI MEASUREMENT
ALGORITHM
Figure 51. Image Rejection Calibration Using the Internal Calibration Source and a Microcontroller
PACKET STRUCTURE AND CODING
The calibration results are valid over changes in the ADF7021-V
supply voltage. However, there is some variation with temperature.
A typical plot of variation in image rejection over temperature
after initial calibrations at −40ꢀC, +25ꢀC, and +85ꢀC is shown in
Figure 52. The internal temperature sensor on the ADF7021-V
can be used to determine whether a new IR calibration is required.
The suggested packet structure to use with the ADF7021-V is
shown in Figure 53.
SYNC
ID
PREAMBLE
DATA FIELD
CRC
WORD
FIELD
Figure 53. Typical Format of a Transmit Protocol
60
CAL AT +25°C
See the Receiver Setup section for information about the required
preamble structure and length for the various modulation schemes.
50
PROGRAMMING AFTER INITIAL POWER-UP
40
30
20
10
0
CAL AT +85°C
CAL AT –40°C
Table 23 lists the minimum number of writes needed to set up
the ADF7021-V in either Tx or Rx mode after CE is brought
high for a minimum of 100 μs before programming any register.
Additional registers can also be written to tailor the part to a
particular application, such as setting up sync byte detection
or enabling AFC. When going from Tx to Rx or vice versa, the
user needs to toggle the Tx/Rx bit and write only to Register 0
to alter the LO by 100 kHz.
V
= 3.0V
IFBW = 25kHz
DD
INTERFERER SIGNAL:
WANTED SIGNAL:
RF FREQ = 430MHz
MODULATION = 2FSK
DATA RATE = 9.6kbps,
DATA = PRBS9
fDEV = 4kHz
LEVEL= –100dBm
RF FREQ = 429.8MHz
MODULATION = 2FSK
DATA RATE = 9.6kbps,
DATA = PRBS11
fDEV = 4kHz
–60
–40
–20
0
20
40
60
80
100
TEMPERATURE (°C)
Table 23. Minimum Register Writes Required for Tx/Rx Setup
Figure 52. Image Rejection vs. Temperature After Initial Calibrations
at −40°C, +25°C, and +85°C
Mode
Required Register Writes
Reg 1, Reg 3, Reg 0, Reg 2
Reg 1, Reg 3, Reg 5, Reg 0, Reg 4
Reg 0
Tx
Rx
Tx to Rx and Rx to Tx
Rev. 0 | Page 39 of 60
ADF7021-V
The recommended programming sequences for transmit and
receive are shown in Figure 54 and Figure 55, respectively.
The difference in the power-up routine for a TCXO and XTAL
reference is shown in these figures.
TCXO
REFERENCE
POWER-DOWN
CE LOW
TURN ON EXTERNAL VCO AND
ALLOW ADEQUATE SETTLING
XTAL
REFERENCE
CE HIGH
WAIT 50µs + 1ms
(REGULATOR POWER-UP + TYPICAL XTAL SETTLING)
CHECK FOR REGULATOR READY
CE HIGH
WAIT 50µs (REGULATOR POWER-UP)
CHECK FOR REGULATOR READY
WRITE TO REGISTER 1
WRITE TO REGISTER 3 (TURNS ON Tx/Rx CLOCKS)
WRITE TO REGISTER 0 (TURNS ON PLL)
WAIT 40µs (TYPICAL PLL SETTLING)
WRITE TO REGISTER 2 (TURNS ON PA)
WAIT FOR PA TO RAMP UP (ONLY IF PA RAMP ENABLED)
Tx MODE
WAIT FOR Tx LATENCY NUMBER OF BITS
(REFER TO TABLE 11)
WRITE TO REGISTER 2 (TURNS OFF PA)
WAIT FOR PA TO RAMP DOWN
CE LOW
POWER-DOWN
OPTIONAL. ONLY NECESSARY IF PA
RAMP-DOWN IS REQUIRED.
Figure 54. Power-Up Sequence for Transmit Mode
Rev. 0 | Page 40 of 60
ADF7021-V
TCXO
REFERENCE
XTAL
REFERENCE
POWER-DOWN
CE LOW
TURN ON EXTERNAL VCO AND
ALLOW ADEQUATE SETTLING
CE HIGH
WAIT 50µs + 1ms
(REGULATOR POWER-UP + TYPICAL XTAL SETTLING)
CHECK FOR REGULATOR READY
CE HIGH
WAIT 50µs (REGULATOR POWER-UP)
CHECK FOR REGULATOR READY
WRITE TO REGISTER 1
WRITE TO REGISTER 3 (TURNS ON Tx/Rx CLOCKS)
OPTIONAL:
ONLY NECESSARY IF
IF FILTER FINE CALIBRATION
IS REQUIRED.
WRITE TO REGISTER 6 (SETS UP IF FILTER FINE CALIBRATION)
WRITE TO REGISTER 5 (STARTS IF FILTER CALIBRATION)
WAIT 0.2ms (COARSE CAL) OR WAIT 8.2ms
(COARSE CALIBRATION + FINE CALIBRATION)
OPTIONAL:
ONLY NECESSARY IF
SWD IS REQUIRED.
WRITE TO REGISTER 11 (SET UP SWD)
WRITE TO REGISTER 12 (ENABLE SWD)
WRITE TO REGISTER 0 (TURNS ON PLL)
WAIT 40µs (TYPICAL PLL SETTLING)
WRITE TO REGISTER 4 (TURNS ON DEMOD)
WRITE TO REGISTER 10 (TURNS ON AFC)
Rx MODE
OPTIONAL:
ONLY NECESSARY IF
AFC IS REQUIRED.
CE LOW
POWER-DOWN
OPTIONAL.
Figure 55. Power-Up Sequence for Receive Mode
Rev. 0 | Page 41 of 60
ADF7021-V
For recommended component values, see the ADF7021-V
evaluation board data sheet and the AN-859 Application Note,
accessible from the ADF7021-V product page. Follow the refer-
ence design schematic closely to ensure optimum performance
in narrow-band applications.
APPLICATIONS CIRCUIT
The ADF7021-V requires very few external components for
operation. Figure 56 shows the recommended application
circuit. Note that the power supply decoupling and regulator
capacitors are omitted for clarity.
EXTERNAL VCO
LOOP FILTER
RF
V
TUNE
OUT
VDD
TCXO
REFERENCE
VDD
VDD
1
2
3
4
5
6
36
CLKOUT
VCOIN
CREG1
VDD1
MATCHING
35
TO
TxRxCLK
TxRxDATA
SWD
MICROCONTROLLER
Tx/Rx SIGNAL
INTERFACE
34
33
32
VDD
T-STAGE LC
FILTER
RFOUT
RFGND
RFIN
ANTENNA
CONNECTION
VDD2
VDD
31
CREG2
ADCIN
GND2
ADF7021-V
30
29
7
8
RFIN
R
LNA
28
27
26
25
9
VDD4
RSET
CREG4
GND4
SCLK
VDD
TO
10
11
12
SREAD
SDATA
SLE
MICROCONTROLLER
CONFIGURATION
INTERFACE
R
LNA
RESISTOR
CHIP ENABLE
TO MICROCONTROLLER
RSET
RESISTOR
NOTES
1. PINS[13:18], PINS[20:21], AND PIN 23 ARE TEST PINS AND ARE NOT USED IN NORMAL OPERATION.
Figure 56. Typical Application Circuit (Regulator Capacitors and Power Supply Decoupling Not Shown)
Rev. 0 | Page 42 of 60
ADF7021-V
SERIAL INTERFACE
The serial interface allows the user to program the 16 32-bit
registers using a 3-wire interface (SCLK, SDATA, and SLE). It
consists of a level shifter, 32-bit shift register, and 16 latches.
Signals should be CMOS compatible. The serial interface is
powered by the regulator and, therefore, is inactive when CE is low.
RSSI Readback
The format of the RSSI readback word is shown in Figure 57. It
comprises the RSSI-level information (Bit RV7 to Bit RV1), the
current filter gain (FG2, FG1), and the current LNA gain (LG2,
LG1) setting. The filter and LNA gain are coded in accordance
with the definitions in the Register 9—AGC Register section. For
signal levels below −100 dBm, averaging the measured RSSI values
improves accuracy. The input power can be calculated from the
RSSI readback value as described in the RSSI/AGC section.
Data is clocked into the register, MSB first, on the rising edge of
each clock (SCLK). Data is transferred to one of 16 latches on the
rising edge of SLE. The destination latch is determined by the
value of the four control bits (C4 to C1); these bits are the four
LSBs, DB3 to DB0, as shown in Figure 2. Data can also be read
back on the SREAD pin.
Readback with AFC or Linear Demodulation On
To perform any readback with the AFC on, the AGC must first be
locked. To lock the AGC, use the LOCK_THRESHOLD_MODE
bits (Bits[DB5:DB4] in Register 12) for packet reception. The lock
threshold mode locks the threshold of the envelope detector, as
well as the AFC and AGC circuits. It can be set to lock on recep-
tion of a valid SWD and remain locked until it is released by a
subsequent SPI command (LOCK_THRESHOLD_MODE = 1).
It can also be set to lock on reception of a valid SWD for a specified
number of bytes by setting LOCK_THRESHOLD_MODE = 2;
or it can be locked at any time by setting LOCK_THRESHOLD_
MODE = 3. After the threshold is locked, a readback can be
performed. The AGC/AFC lock is released by setting
READBACK FORMAT
The readback operation is initiated by writing a valid control word
to the readback setup register and enabling the READBACK_
SELECT bit (Register 7, Bit DB8 = 1). The readback can begin
after the control word has been latched with the SLE signal. SLE
must be kept high while the data is being read out. Each active
edge at the SCLK pin successively clocks the readback word out
at the SREAD pin, MSB first (see Figure 57). The data appearing
at the first clock cycle following the latch operation must be
ignored. An extra clock cycle is needed after the 16th readback
bit to return the SREAD pin to tristate. Therefore, 18 total clock
cycles are needed for each readback. After the 18th clock cycle,
SLE should be brought low.
LOCK_THRESHOLD_MODE = 0.
Battery Voltage/ADCIN/Temperature Sensor Readback
AFC Readback
The battery voltage is measured at Pin VDD4. The readback
information is contained in Bit RV7 to Bit RV1. This also
applies to the readback of the voltage at the ADCIN pin and the
temperature sensor. From the readback information, the battery
or ADCIN voltage can be determined as follows:
The AFC readback is valid only during the reception of FSK
signals with either the linear or correlator demodulator active.
The AFC readback value is formatted as a signed, 16-bit integer
comprising Bit RV16 to Bit RV1 and is scaled according to the
following formula:
V
V
BATTERY = (BATTERY VOLTAGE READBACK)/21.1
ADCIN = (ADCIN VOLTAGE READBACK)/42.1
FREQ RB (Hz) = (AFC READBACK × DEMOD CLK)/218
In the absence of frequency errors, FREQ RB is equal to the IF
frequency of 100 kHz. Note that, for the AFC readback to yield
a valid result, the downconverted input signal must not fall outside
the bandwidth of the analog IF filter. At low input signal levels, the
variation in the readback value can be improved by averaging.
The temperature can be calculated as follows:
Temperature (ꢀC) = −40 + [(68.4 − TEMP READBACK) × 9.32]
READBACK MODE
READBACK VALUE
DB15 DB14 DB13 DB12 DB11 DB10 DB9
DB8
DB7
RV8
FG1
DB6
RV7
RV7
DB5
RV6
RV6
DB4
RV5
RV5
DB3
RV4
RV4
DB2
RV3
RV3
DB1
RV2
RV2
DB0
RV1
RV1
AFC READBACK
RSSI READBACK
RV16 RV15 RV14 RV13 RV12 RV11 RV10 RV9
X
X
X
X
X
X
X
X
X
X
LG2
X
LG1
X
FG2
X
BATTERY VOLTAGE/ADCIN/
TEMP. SENSOR READBACK
X
RV7
RV7
RV7
RV6
RV6
RV6
RV5
RV5
RV5
RV4
RV4
RV4
RV3
RV3
RV3
RV2
RV2
RV2
RV1
RV1
RV1
SILICON REVISION
RV16 RV15 RV14 RV13 RV12 RV11 RV10 RV9
RV8
RV8
FILTER CAL READBACK
0
0
0
0
0
0
0
0
Figure 57. Readback Value Table
Rev. 0 | Page 43 of 60
ADF7021-V
Silicon Revision Readback
UART Mode
In UART mode, the TxRxCLK pin is configured to input trans-
mit data in transmit mode. In receive mode, the receive data is
available on the TxRxDATA pin, thus providing an asynchronous
data interface. The UART mode can only be used with oversampled
2FSK modulation. Figure 59 shows a possible interface to a micro-
controller using the UART mode of the ADF7021-V. To enable the
UART interface mode, set Bit DB28 in Register 0 high. Figure 8
and Figure 9 show the relevant timing diagrams for UART mode.
The silicon revision readback word is valid without setting any
other registers. The silicon revision word is coded with four
quartets in BCD format. The product code (PC) is coded with
three quartets extending from Bit RV16 to Bit RV5. The revision
code (RC) is coded with one quartet extending from Bit RV4 to
Bit RV1. The product code for the ADF7021-V should read back as
PC = 0x212. The current revision code should read as RC = 0x0.
Filter Bandwidth Calibration Readback
MICROCONTROLLER
ADF7021-V
The filter calibration readback word is contained in Bit RV8 to
Bit RV1 (see Figure 57). This readback can be used for manual
filter adjustment, thereby avoiding the need to do an IF filter
calibration in some instances. The manual adjust value is
programmed using Register 5, Bits[DB19:DB14]. To calculate
the manual adjustment based on a filter calibration readback,
use the following formula:
TxRxCLK
TxDATA
RxDATA
UART
TxRxDATA
CE
SWD
SREAD
SLE
GPIO
SDATA
SCLK
IF_FILTER_ADJUST = FILTER_CAL_READBACK − 128
Figure 59. ADF7021-V (UART Mode) to Asynchronous Microcontroller Interface
The result should be programmed into Register 5, Bits[DB19:DB14]
as described in the Register 5—IF Filter Setup Register section.
SPI Mode
In SPI mode, the TxRxCLK pin is configured to input transmit
data in transmit mode. In receive mode, the receive data is avail-
able on the TxRxDATA pin. The data clock in both transmit and
receive modes is available on the CLKOUT pin. In transmit mode,
data is clocked into the ADF7021-V on the rising edge of
CLKOUT. In receive mode, the TxRxDATA data pin should be
sampled by the microcontroller on the rising edge of CLKOUT.
INTERFACING TO A MICROCONTROLLER/DSP
Standard Transmit/Receive Data Interface
The standard transmit/receive signal and configuration interface
to a microcontroller is shown in Figure 58. In transmit mode, the
ADF7021-V provides the data clock on the TxRxCLK pin, and
the TxRxDATA pin is used as the data input. The transmit data
is clocked into the ADF7021-V on the rising edge of TxRxCLK.
To enable SPI interface mode, set Bit DB28 in Register 0 high
and set Bits[DB19:DB17] in Register 15 to 0x7. Figure 8 and
Figure 9 show the relevant timing diagrams for SPI mode;
Figure 60 shows the recommended interface to a micro-
controller using the SPI mode of the ADF7021-V.
ADuC84x
MISO
ADF7021-V
TxRxDATA
MOSI
SCLOCK
SS
TxRxCLK
P3.7
CE
MICROCONTROLLER
ADF7021-V
P3.2/INT0
P2.4
SWD
SREAD
SLE
TxRxCLK
MISO
P2.5
TxRxDATA
CLKOUT
MOSI
SCLK
SPI
GPIO
P2.6
P2.7
SDATA
SCLK
CE
Figure 58. ADuC84x to ADF7021-V Connection Diagram
SWD
SREAD
SLE
GPIO
In receive mode, the ADF7021-V provides the synchronized
data clock on the TxRxCLK pin. The received data is available
on the TxRxDATA pin. The rising edge of TxRxCLK should be
used to clock the receive data into the microcontroller. See
Figure 4 and Figure 5 for the relevant timing diagrams.
SDATA
SCLK
Figure 60. ADF7021-V (SPI Mode) to Microcontroller Interface
ADSP-BF533 Interface
In 4FSK transmit mode, the MSB of the transmit symbol is
clocked into the ADF7021-V on the first rising edge of the data
clock from the TxRxCLK pin. In 4FSK receive mode, the MSB
of the first payload symbol is clocked out on the first falling edge
of the data clock after the SWD and should be clocked into the
microcontroller on the following rising edge. See Figure 6 and
Figure 7 for the relevant timing diagrams.
The suggested method of interfacing to the Blackfin®
ADSP-BF533 is shown in Figure 61.
ADSP-BF533
SCK
ADF7021-V
SCLK
MOSI
MISO
SDATA
SREAD
SLE
PF5
RSCLK1
DT1PRI
DR1PRI
RFS1
TxRxCLK
TxRxDATA
SWD
CE
PF6
Figure 61. ADSP-BF533 to ADF7021-V Connection Diagram
Rev. 0 | Page 44 of 60
ADF7021-V
REGISTER 0—N REGISTER
ADDRESS
MUXOUT
INTEGER_N
FRACTIONAL_N
BITS
FRACTIONAL_N
DIVIDE RATIO
TR1
Tx/Rx
M15 M14 M13 ...
M3
M2
M1
0
1
TRANSMIT
RECEIVE
0
0
0
.
.
.
1
1
1
1
0
0
0
.
.
.
1
1
1
1
0
0
0
.
.
.
1
1
1
1
...
...
...
...
...
...
...
...
...
...
0
0
0
.
.
.
1
1
1
1
0
0
1
.
.
.
0
0
1
1
0
1
0
.
.
.
0
1
0
1
0
1
2
.
.
.
U1
UART_MODE
0
1
DISABLED
ENABLED
32,764
32,765
32,766
32,767
M3
M2
M1
MUXOUT
0
0
0
0
1
1
1
1
0
0
1
1
0
0
1
1
0
1
0
1
0
1
0
1
REGULATOR_READY (DEFAULT)
FILTER_CAL_COMPLETE
DIGITAL_LOCK_DETECT
RSSI_READY
Tx_Rx
LOGIC_ZERO
TRISTATE
LOGIC_ONE
INTEGER_N
N8
N7
N6
N5
N4
N3
N2
N1
DIVIDE RATIO
0
0
.
0
0
.
0
0
.
1
1
.
0
1
.
1
0
.
1
0
.
1
0
.
23
24
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
1
1
1
1
1
1
0
1
253
1
1
1
1
1
1
1
1
1
1
1
1
1
1
0
1
254
255
Figure 62. Register 0—N Register Map
•
The RF output frequency is calculated as follows:
For direct output,
•
In the MUXOUT map (Bits[DB31:DB29]), FILTER_CAL_
COMPLETE indicates when a coarse or coarse plus fine IF
filter calibration has finished. DIGITAL_LOCK_DETECT
indicates when the PLL has locked. RSSI_READY indicates
that the RSSI signal has settled and an RSSI readback can
be performed. Tx_Rx gives the status of Bit DB27 in this
register, which can be used to control an external Tx/Rx
switch.
FRACTIONAL _ N
⎛
⎝
⎞
⎟
⎠
RFOUT = PFD × INTEGER _ N +
⎜
215
With RF_DIVIDE_BY_2 (Register 1, Bit DB18) enabled,
FRACTIONAL _ N
⎛
⎝
⎞
⎟
⎠
RFOUT = PFD × 0.5 × INTEGER _ N +
⎜
215
•
In UART/SPI mode, the TxRxCLK pin is used to input
the transmitted data. The received data is available on the
TxRxDATA pin.
Rev. 0 | Page 45 of 60
ADF7021-V
REGISTER 1—OSCILLATOR REGISTER
XTAL_
BIAS
ADDRESS
BITS
CLKOUT_
DIVIDE
RESERVED
R_COUNTER
RF R_COUNTER
R3 R2 R1 DIVIDE RATIO
RF_DIVIDE_BY_2
RFD1
0
0
.
0
1
.
1
0
.
1
2
.
0
1
OFF
ON
.
.
.
.
.
.
.
.
1
1
1
7
CLKOUT_
DIVIDE RATIO
OFF
CL4
CL3
CL2
CL1
0
0
0
.
0
0
0
.
0
0
1
.
0
1
0
.
2
4
.
BUFFER_
VE1 IMPEDANCE
.
.
.
.
.
.
.
.
.
.
0
1
50Ω
30
1
1
1
1
HIGH IMPEDANCE
XTAL_
D1
0
1
DOUBLER
DISABLED
ENABLED
RSET = 3.6kΩ
CP2 CP1
XOSC_ENABLE
X1
0
1
I
(mA)
CP
OFF
ON
0
0
1
1
0
1
0
1
0.3
0.9
1.5
2.1
XB2 XB1 XTAL_BIAS
0
0
1
1
0
1
0
1
20µA
25µA
30µA
35µA
Figure 63. Register 1—Oscillator Register Map
•
The R_COUNTER and XTAL_DOUBLER relationship is
as follows:
•
•
CLKOUT_DIVIDE is a divided-down and inverted version
of the XTAL and is available on Pin 36 (CLKOUT).
Set XOSC_ENABLE high when using an external crystal.
If using an external oscillator (such as TCXO) with CMOS
level outputs into Pin OSC2, set XOSC_ENABLE low. If
using an external oscillator with a 0.8 V p-p clipped sine
wave output into Pin OSC1, set XOSC_ENABLE high.
If XTAL_DOUBLER = 0,
XTAL
R _COUNTER
PFD =
If XTAL_DOUBLER = 1,
XTAL × 2
PFD =
R _COUNTER
Rev. 0 | Page 46 of 60
ADF7021-V
REGISTER 2—TRANSMIT MODULATION REGISTER
TxDATA_
INVERT
MODULATION_
SCHEME
ADDRESS
BITS
Tx_FREQUENCY_DEVIATION
POWER_AMPLIFIER
PA_BIAS PA_RAMP
PE1 PA_ENABLE
OFF
ON
PA2 PA1 PA_BIAS
0
1
0
0
1
1
0
1
0
1
5µA
7µA
9µA
11µA
DI2 DI1 TxDATA_INVERT
PR3 PR2 PR1 PA_RAMP RATE
0
0
1
1
0
1
0
1
NORMAL
INVERT CLK
INVERT DATA
INV CLK AND DATA
0
0
0
0
1
1
1
1
0
0
1
1
0
0
1
1
0
1
0
1
0
1
0
1
NO RAMP
256 CODES/BIT
128 CODES/BIT
64 CODES/BIT
32 CODES/BIT
16 CODES/BIT
8 CODES/BIT
4 CODES/BIT
fDEV
TFD9 ... TFD3 TFD2 TFD1
0
0
0
0
.
...
...
...
...
...
...
0
0
0
0
.
0
0
1
1
.
0
1
0
1
.
0
1
2
3
S3
S2
S1
MODULATION_SCHEME
0
0
0
0
1
1
1
1
0
0
1
1
0
0
1
1
0
1
0
1
0
1
0
1
2FSK
GAUSSIAN 2FSK
3FSK
.
1
1
1
1
511
4FSK
OVERSAMPLED 2FSK
RAISED COSINE 2FSK
RAISED COSINE
RAISED COSINE
3FSK
4FSK
POWER_
NRC1 R-COSINE_ALPHA
P6
P5
...
AMPLIFIER
P2
P1
0.5 (DEFAULT)
0.7
0
1
0
0
0
0
.
0
0
0
0
.
...
...
...
...
...
...
...
0
0
1
1
.
0
1
0
1
.
0 (PA OFF)
1 (–16.0dBm)
2
3
.
.
.
.
.
1
.
1
1
1
63 (+13dBm)
Figure 64. Register 2—Transmit Modulation Register Map
•
The 2FSK/3FSK/4FSK frequency deviation is expressed as
follows:
•
The power amplifier (PA) ramps at the programmed rate
(Bits[DB10:DB8]) until it reaches its programmed level
(Bits[DB18:DB13]). If the PA is enabled/disabled by the
PA_ENABLE bit (Bit DB7), it ramps up and down. If it is
enabled/disabled by the Tx/Rx bit (Register 0, Bit DB27),
it ramps up and turns hard off.
R-COSINE_ALPHA sets the roll-off factor (alpha) of the
raised cosine data filter to either 0.5 or 0.7. The alpha is set
to 0.5 by default, but the raised cosine filter bandwidth can
be increased to provide less aggressive data filtering by
using an alpha of 0.7.
For direct RF output,
Frequency Deviation (Hz) =
Tx_FREQUENCY_DEVIATION × PFD
216
•
With RF_DIVIDE_BY_2 (Register 1, Bit DB18) enabled,
Frequency Deviation (Hz) =
Tx_FREQUENCY_DEVIATION × PFD
0.5 ×
216
where:
Tx_FREQUENCY_DEVIATION is set by Bits[DB27:DB19].
PFD is the PFD frequency.
•
In the case of 4FSK, there are tones at 3 × the frequency
deviation and at 1 × the frequency deviation.
Rev. 0 | Page 47 of 60
ADF7021-V
REGISTER 3—TRANSMIT/RECEIVE CLOCK REGISTER
DEMOD_CLK_
DIVIDE
ADDRESS
BITS
AGC_CLK_DIVIDE
SEQ_CLK_DIVIDE
CDR_CLK_DIVIDE
SK8 SK7 ...
...
SK3 SK2 SK1 SEQ_CLK_DIVIDE
BK2 BK1 BBOS_CLK_DIVIDE
0
0
.
0
0
.
0
0
.
0
1
.
1
0
.
1
2
.
0
0
1
0
1
4
...
...
...
...
0
1
1
8
16
32
1
1
1
1
1
1
1
1
0
1
254
255
OK4 OK3 OK2 OK1 DEMOD_CLK_DIVIDE
GD6
GD5
GD4
GD3
GD2
GD1 AGC_CLK_DIVIDE
0
0
0
0
INVALID
0
...
1
0
...
1
0
...
1
1
...
1
1
...
15
0
0
...
1
0
0
...
1
0
0
...
1
0
0
...
1
0
0
...
1
0
1
...
1
INVALID
1
...
63
FS8
FS7
...
FS3
FS2
FS1 CDR_CLK_ DIVIDE
0
0
.
1
1
0
0
.
1
1
...
...
...
...
...
0
0
.
1
1
0
1
.
1
1
1
0
.
0
1
1
2
.
254
255
Figure 65. Register 3—Transmit/Receive Clock Register Map
•
•
•
Baseband offset clock frequency (BBOS CLK) must be
greater than 1 MHz and less than 2 MHz, where
•
•
The sequencer clock (SEQ CLK) supplies the clock to the
digital receive block. It should be as close to 100 kHz as
possible.
BBOS CLK = (XTAL/BBOS_CLK_DIVIDE)
SEQ CLK = (XTAL/SEQ_CLK_DIVIDE)
Set the demodulator clock (DEMOD CLK) such that
2 MHz ≤ DEMOD CLK ≤ 15 MHz, where
The time allowed for each AGC step to settle is determined
by the AGC update rate. It should be set close to 3 kHz.
DEMOD CLK = (XTAL/DEMOD_CLK_DIVIDE)
AGC Update Rate (Hz) = (SEQ CLK/AGC_CLK_DIVIDE)
For 2FSK/3FSK, the clock/data recovery frequency (CDR
CLK) must be within 2% of (32 × data rate). For 4FSK, the
CDR CLK must be within 2% of (32 × symbol rate).
CDR CLK = (DEMOD CLK/CDR_CLK_DIVIDE)
Rev. 0 | Page 48 of 60
ADF7021-V
REGISTER 4—DEMODULATOR SETUP REGISTER
Rx_
INVERT
DEMOD_
SCHEME
ADDRESS
BITS
POST_DEMOD_BW
DISCRIMINATOR_BW
IF_FILTER _
BW
DP1
DOT_PRODUCT
IFB2 IFB1
0
1
CROSS_PRODUCT
DOT_PRODUCT
0
0
1
1
0
1
0
1
9kHz
13.5kHz
18.5kHz
INVALID
RI2 RI1
Rx_INVERT
0
1
0
1
0
0
1
1
NORMAL
INVERT CLK
INVERT DATA
INVERT CLK/DATA
DEMOD_SCHEME
DS3 DS2 DS1
0
0
0
0
1
1
1
1
0
0
1
1
0
0
1
1
0
1
0
1
0
1
0
1
DW10 .
DW6 DW5 DW4 DW3 DW2 DW1
2FSK LINEAR DEMODULATOR
2FSK CORRELATOR DEMODULATOR
3FSK DEMODULATOR
4FSK DEMODULATOR
RESERVED
RESERVED
RESERVED
RESERVED
POST_DEMOD_BW
0
0
.
.
.
.
.
.
.
.
0
0
.
0
0
.
0
0
.
0
0
.
0
1
.
1
0
.
1
2
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
1
.
.
.
.
.
.
1
1
1
1
1
1
1023
TD10 .
TD6 TD5 TD4 TD3 TD2 TD1
DISCRIMINATOR_BW
0
0
.
.
.
.
.
.
.
.
.
.
0
0
.
.
.
0
0
.
.
.
0
0
.
.
.
0
0
.
.
.
0
1
.
.
.
1
0
.
.
.
1
2
.
.
.
.
.
1
.
0
.
1
.
0
.
1
.
0
.
0
660
Figure 66. Register 4—Demodulator Setup Register Map
where:
•
To solve for DISCRIMINATOR_BW, (Bits[DB19:DB10]),
use the following equation:
Round is rounded to the nearest integer.
Round4FSK is rounded to the nearest of the following integers:
32, 31, 28, 27, 24, 23, 20, 19, 16, 15, 12, 11, 8, 7, 4, 3.
⎛
⎜
⎜
⎝
⎞
⎟
⎟
⎠
DEMOD CLK × K
400 × 103
DISCRIMINATOR _ BW =
f
f
DEV is the transmit frequency deviation in Hz. For 4FSK,
DEV is the frequency deviation used for the 1 symbols
where the maximum value = 660.
(that is, the inner frequency deviations).
For 2FSK,
•
•
Rx_INVERT (Bits[DB9:DB8]) and DOT_PRODUCT
(Bit DB7) must be set as indicated in Table 16 and Table 17.
3
⎛
⎜
⎜
⎝
⎞
⎟
⎟
⎠
100 × 10
K = Round
For 3FSK,
K = Round
For 4FSK,
fDEV
POST_DEMOD_BW (Bits[DB29:DB20]) sets the
bandwidth of the postdemodulator filter. To solve for
POST_DEMOD_BW, use the following equation:
3
⎛
⎜
⎜
⎝
⎞
⎟
⎟
⎠
100 × 10
211 ×π × fCUTOFF
POST_DEMOD_BW =
2 × fDEV
DEMOD CLK
where fCUTOFF (the cutoff frequency of the postdemodulator
filter) should typically be set equal to 0.75 × the data rate in
2FSK. In 3FSK, it should be set equal to the data rate, whereas
in 4FSK, it should be set equal to 1.6 × the symbol rate.
3
⎛
⎜
⎜
⎝
⎞
⎟
⎟
⎠
100 × 10
K = Round4FSK
4 × fDEV
Rev. 0 | Page 49 of 60
ADF7021-V
REGISTER 5—IF FILTER SETUP REGISTER
IR_GAIN_
ADJUST_MAG
IR_PHASE_
ADJUST_MAG
ADDRESS
BITS
IF_FILTER_ADJUST
IF_FILTER_DIVIDER
CC1 IF_CAL_COARSE
0
1
DISABLED
ENABLED
IR_PHASE_
PM4 PM3 PM2 PM1
ADJUST_MAG
0
0
0
.
0
1
2
...
15
IF_FILTER_
0
0
0
.
0
0
1
.
0
1
0
.
IFD9
.
IFD6 IFD5 IFD4 IFD3 IFD2 IFD1
DIVIDER
0
0
.
.
.
.
.
.
.
.
.
.
0
0
.
0
0
.
0
0
.
0
0
.
0
1
.
1
0
.
1
2
.
.
.
.
1
1
1
1
.
.
.
.
.
.
.
.
.
.
.
.
.
1
.
.
.
.
.
.
PD1
IR_PHASE_ADJUST_DIRECTION
1
1
1
1
1
1
511
0
1
ADJUST I CH
ADJUST Q CH
IR_GAIN_
ADJUST_MAG
GM5 GM4 GM3 GM2 GM1
0
0
0
.
0
0
0
.
0
1
2
...
31
0
0
0
.
0
0
1
.
0
1
0
.
IFA6 IFA5
IFA2 IFA1 IF_FILTER_ADJUST
...
...
...
...
...
...
...
...
...
...
...
0
0
0
..
0
1
1
1
1
1
0
0
0
..
1
0
0
0
.
0
0
1
..
1
0
0
1
.
0
1
0
..
1
0
1
0
.
0
+1
+2
...
+31
0
–1
–2
...
1
1
1
1
1
GQ1 IR_GAIN_ADJUST_I/Q
0
1
ADJUST I CH
ADJUST Q CH
GA1 IR_GAIN_ADJUST_UP/DN
1
1
–31
1
0
1
GAIN
ATTENUATE
Figure 67. Register 5—IF Filter Setup Register Map
•
•
A coarse IF filter calibration is performed when the
IF_CAL_COARSE bit (Bit DB4) is set. If the IF_FINE_
CAL bit (Register 6, Bit DB4) has been previously set, a
fine IF filter calibration is automatically performed after
the coarse calibration.
•
IF_FILTER_ADJUST allows the IF fine filter calibration
result to be programmed directly on subsequent receiver
power-ups, thereby eliminating the need to redo a fine filter
calibration in some instances. See the Filter Bandwidth
Calibration Readback section for information about using
the IF_FILTER_ADJUST bits.
Set IF_FILTER_DIVIDER such that
•
Bits[DB31:DB20] are used for image rejection calibration.
See the Image Rejection Calibration section for information
about how to program these parameters.
XTAL
= 50 kHz
IF _ FILTER _ DIVIDER
Rev. 0 | Page 50 of 60
ADF7021-V
REGISTER 6—IF FINE CALIBRATION SETUP REGISTER
ADDRESS
BITS
IF_CAL_DWELL_TIME
IF_CAL_UPPER_TONE_DIVIDE
IF_CAL_LOWER_TONE_DIVIDE
IRD1 IR_CAL_SOURCE ÷2
0
1
SOURCE ÷2 OFF
SOURCE ÷2 ON
IF_CAL_UPPER_
TONE_DIVIDE
FC1 IF_FINE_CAL
UT3 UT2 UT1
...
UT8
UT7
0
1
DISABLED
ENABLED
0
0
0
.
0
0
0
.
0
1
1
.
1
0
1
.
0
0
0
.
...
...
...
...
...
1
2
3
.
IR_CAL_SOURCE_
IRC2 IRC1 DRIVE_LEVEL
0
0
1
1
0
1
0
1
OFF
.
.
.
.
.
.
LOW
MED
HIGH
1
1
1
1
0
...
127
IF_CAL_LOWER_
TONE_DIVIDE
LT7
LT3 LT2 LT1
LT8
...
0
0
0
.
0
1
1
.
1
0
1
.
0
0
0
.
0
0
0
.
...
...
...
...
...
1
2
3
.
IF_CAL_
DWELL_TIME
CD3 CD2 CD1
CD7
...
0
0
0
.
0
1
1
.
1
0
1
.
0
0
0
.
...
...
...
...
...
1
2
3
.
.
.
.
.
.
.
1
1
1
1
1
...
255
.
.
.
.
.
1
1
1
1
...
127
Figure 68. Register 6—IF Fine Calibration Setup Register Map
•
•
A fine IF filter calibration is set by enabling the IF_FINE_
CAL bit (Bit DB4). A fine calibration is performed only
when Register 5 is written to and Register 5, Bit DB4 is set.
Lower Tone Frequency (kHz) =
•
The IF tone calibration time is the amount of time that
is spent at an IF calibration tone. It is dependent on the
sequencer clock. It is recommended that the IF tone
calibration time be at least 800 μs.
IF_CAL_DWELL_TIME
IF Tone Calibration Time =
SEQ CLK
XTAL
IF_CAL_LOWER_TONE_DIVIDE × 2
The total time for a fine IF filter calibration is
Upper Tone Frequency (kHz) =
IF Tone Calibration Time × 10
XTAL
IF_CAL_UPPER_TONE_DIVIDE × 2
•
Bits[DB30:DB28] control the internal source for the image
rejection (IR) calibration. The IR_CAL_SOURCE_DRIVE_
LEVEL bits (Bits[DB29:DB28]) set the drive strength of the
source, whereas the IR_CAL_SOURCE ÷2 bit (Bit DB30)
allows the frequency of the internal signal source to be
divided by 2.
It is recommended that the lower tone and the upper tone
be set as shown in Table 24.
Table 24. IF Filter Fine Calibration Tone Frequencies
IF Filter
Bandwidth (kHz)
Lower Tone
Frequency (kHz)
Upper Tone
Frequency (kHz)
9
13.5
18.5
78.1
79.4
78.1
116.3
116.3
119
Rev. 0 | Page 51 of 60
ADF7021-V
REGISTER 7—READBACK SETUP REGISTER
READBACK_
SELECT
ADC_
MODE
CONTROL
BITS
DB1
DB0
DB7
RB2
DB8
RB3
DB6
RB1
DB5
AD2
DB4
AD1
DB3
DB2
C4 (0) C3 (1) C2 (1) C1 (1)
RB3 READBACK_SELECT
AD2 AD1 ADC_MODE
0
1
DISABLED
ENABLED
0
0
1
1
0
1
0
1
MEASURE RSSI
BATTERY VOLTAGE
TEMP SENSOR
TO EXTERNAL PIN
RB2 RB1 READBACK MODE
0
0
1
1
0
1
0
1
AFC WORD
ADC OUTPUT
FILTER CAL
SILICON REV
Figure 69. Register 7—Readback Setup Register Map
•
•
Readback of the measured RSSI value is valid only in Rx
mode. Readback of the battery voltage, temperature sensor,
or voltage at the external ADCIN pin is not valid in Rx mode.
To read back the battery voltage, the temperature sensor, or
the voltage at the external ADCIN pin in Tx mode, the user
should first power up the ADC using Register 8, Bit DB8
because it is turned off by default in Tx mode to save power.
•
For AFC readback, use the following equations (see the
Readback Format section):
FREQ RB (Hz) = (AFC READBACK × DEMOD CLK)/218
V
BATTERY = BATTERY VOLTAGE READBACK/21.1
ADCIN = ADCIN VOLTAGE READBACK/42.1
V
Temperature (ꢀC) = −40 + [(68.4 − TEMP READBACK) × 9.32]
Rev. 0 | Page 52 of 60
ADF7021-V
REGISTER 8—POWER-DOWN TEST REGISTER
CONTROL
BITS
Rx_RESET
DB10 DB9
LE1 PD6
DB1
DB2
DB0
DB7
PD4
DB15 DB14 DB13 DB12 DB11
DB8
PD5
DB6
PD3
DB5
RES
DB4
PD1
DB3
PD7
SW1
C4 (1) C3 (0) C2 (0) C1 (0)
CR1
RR2
RR1
CR1 COUNTER_RESET
PD1 SYNTH_ENABLE
0
1
NORMAL
RESET
0
1
SYNTH OFF
SYNTH ON
RR2 CDR_RESET
0
1
NORMAL
RESET
RR1 DEMOD_RESET
0
1
NORMAL
RESET
PD3 LNA/MIXER_ENABLE
0
1
LNA/MIXER OFF
LNA/MIXER ON
PD7 PA_ENABLE_Rx_MODE
0
1
PA OFF
PA ON
PD4 FILTER_ENABLE
0
1
FILTER OFF
FILTER ON
SW1 Tx/Rx_SWITCH_ENABLE
0
1
DEFAULT (ON)
OFF
PD5 ADC_ENABLE
0
1
ADC OFF
ADC ON
LE1 LOG_AMP_ENABLE
0
1
LOG AMP OFF
LOG AMP ON
PD6 DEMOD_ENABLE
0
1
DEMOD OFF
DEMOD ON
Figure 70. Register 8—Power-Down Test Register Map
It is not necessary to write to this register under normal
operating conditions.
For a combined LNA/PA matching network, Bit DB11 should
always be set to 0, which enables the internal Tx/Rx switch. This
is the power-up default condition.
Rev. 0 | Page 53 of 60
ADF7021-V
REGISTER 9—AGC REGISTER
LNA_
BIAS
FILTER_
GAIN
LNA_
GAIN
AGC_
MODE
ADDRESS
BITS
AGC_HIGH_THRESHOLD
AGC_LOW_THRESHOLD
ML1 MIXER_LINEARITY
AGC_LOW_
THRESHOLD
AGC_MODE
GM2 GM1
GL7 GL6 GL5 GL4 GL3 GL2 GL1
0
1
DEFAULT
HIGH
0
0
1
1
0
1
0
1
AUTO AGC
0
0
0
0
.
0
0
0
0
.
0
0
0
0
.
0
0
0
0
.
0
0
0
1
.
0
1
1
0
.
1
0
1
0
.
1
2
3
4
.
.
MANUAL AGC
FREEZE AGC
RESERVED
LI2 LI1 LNA_BIAS
0
0
800µA (DEFAULT)
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
0
1
1
1
0
1
61
62
63
LM1 LNA_MODE
0
1
DEFAULT
REDUCED GAIN
AGC_HIGH_
THRESHOLD
GH7 GH6 GH5 GH4 GH3 GH2 GH1
FI1 FILTER_CURRENT
0
0
0
0
.
.
.
1
1
1
0
0
0
0
.
.
.
0
0
0
0
0
0
0
.
.
.
0
0
1
0
0
0
0
.
.
.
1
1
0
0
0
0
1
.
.
.
1
1
0
0
1
1
0
.
.
.
1
1
0
1
0
1
0
.
.
.
0
1
0
1
2
3
4
.
.
.
78
79
80
0
1
LOW
HIGH
FG2 FG1 FILTER_GAIN
0
0
1
1
0
1
0
1
8
24
72
INVALID
LG2 LG1 LNA_GAIN
0
0
1
1
0
1
0
1
3
10
30
INVALID
Figure 71. Register 9—AGC Register Map
•
•
It is necessary to program this register only if AGC settings
other than the defaults are required.
In receive mode, AGC is set to automatic AGC by default
on power-up. The default thresholds are AGC_LOW_
THRESHOLD = 30 and AGC_HIGH_THRESHOLD = 70.
See the RSSI/AGC section for details.
•
•
AGC high and low threshold values must be more than
30 apart to ensure correct operation.
An LNA gain of 30 is available only if LNA_MODE
(Bit DB25) is set to 0.
Rev. 0 | Page 54 of 60
ADF7021-V
REGISTER 10—AFC REGISTER
ADDRESS
BITS
MAX_AFC_RANGE
KP
KI
AFC_SCALING_FACTOR
AE1 AFC_EN
KP3 KP2 KP1 KP
KI4 KI3 KI2 KI1 KI
0
0
.
0
0
.
0
0
.
0
0
.
0
1
.
2^0
2^1
...
0
0
.
0
1
.
2^0
2^1
...
0
1
AFC OFF
AFC ON
1
1
1
1
1
2^7
1
1
2^15
AFC_SCALING_
FACTOR
MAX_AFC_RANGE
MA3 MA2 MA1
...
MA8
...
M3
M2
M1
M12
0
0
0
0
.
.
.
1
1
1
...
...
...
...
...
...
...
...
...
...
0
1
1
0
.
.
.
0
1
1
0
0
0
1
.
.
.
1
1
1
1
0
1
0
.
.
.
1
0
1
1
2
3
4
.
.
.
253
254
255
0
0
0
0
.
.
.
1
1
1
...
...
...
...
...
...
...
...
...
...
0
0
0
1
.
.
.
1
1
1
0
1
1
0
.
.
.
0
1
1
1
0
1
0
.
.
.
1
0
1
1
2
3
4
.
.
.
4093
4094
4095
Figure 72. Register 10—AFC Register Map
•
The AFC_SCALING_FACTOR can be expressed as
•
When RF_DIVIDE_BY_2 (Register 1, Bit DB18) is enabled,
the programmed AFC correction range is halved. The user
must account for this halving by doubling the programmed
MAX_AFC_RANGE value.
24
⎛
⎞
2
×500
⎜
⎜
⎟
⎟
AFC _ SCALING _ FACTOR = Round
XTAL
⎝
⎠
•
Signals that are within the AFC pull-in range but outside
the IF filter bandwidth are attenuated by the IF filter. As a
result, the signal can be below the sensitivity point of the
receiver and, therefore, not detectable by the AFC.
•
•
The settings for KI and KP affect the AFC settling time and
AFC accuracy. The allowable range for each parameter is
KI > 6 and KP < 7.
The recommended settings for optimal AFC performance
are KI = 11 and KP = 4. To trade off between AFC settling
time and AFC accuracy, the KI and KP parameters can be
adjusted from the recommended settings (staying within
the allowable range) such that
AFC Correction Range = MAX_AFC_RANGE × 500 Hz
Rev. 0 | Page 55 of 60
ADF7021-V
REGISTER 11—SYNC WORD DETECT REGISTER
CONTROL
BITS
SYNC_BYTE_SEQUENCE
SYNC_BYTE_
PL2 PL1 LENGTH
0
0
1
1
0
1
0
1
12 BITS
16 BITS
20 BITS
24 BITS
MATCHING_
MT2 MT1 TOLERANCE
0
0
1
1
0
1
0
1
ACCEPT 0 ERRORS
ACCEPT 1 ERROR
ACCEPT 2 ERRORS
ACCEPT 3 ERRORS
Figure 73. Register 11—Sync Word Detect Register Map
REGISTER 12—SWD/THRESHOLD SETUP REGISTER
CONTROL
BITS
DATA_PACKET_LENGTH
DPx DATA_PACKET_LENGTH
0
1
...
INVALID
1 BYTE
...
255 255 BYTES
ILx SWD_MODE
0
1
2
SWD PIN LOW
SWD PIN HIGH AFTER NEXT SYNC WORD
SWD PIN HIGH AFTER NEXT SYNC WORD
FOR DATA PACKET LENGTH NUMBER OF BYTES
SWD PIN HIGH
3
LMx LOCK_THRESHOLD_MODE
0
1
2
THRESHOLD FREE RUNNING
LOCK THRESHOLD AFTER NEXT SYNC WORD
LOCK THRESHOLD AFTER NEXT SYNC WORD
FOR DATA PACKET LENGTH NUMBER OF BYTES
LOCK THRESHOLD
3
Figure 74. Register 12—SWD/Threshold Setup Register Map
Lock threshold locks the threshold of the envelope detector.
This has the effect of locking the slicer in linear demodulation
and locking the AFC and AGC loops when using linear or
correlator demodulation.
Rev. 0 | Page 56 of 60
ADF7021-V
REGISTER 13—3FSK/4FSK DEMODULATION REGISTER
See the Receiver Setup section for information about programming these settings.
3FSK_PREAMBLE_
TIME_VALIDATE
3FSK/4FSK_
SLICER_THRESHOLD
CONTROL
BITS
3FSK_CDR_THRESHOLD
3FSK_VITERBI_
DETECTOR
VD1
3FSK_CDR_
VT3 VT2 VT1
VT7
...
THRESHOLD
0
1
DISABLED
ENABLED
0
0
0
0
.
0
0
1
1
.
0
0
0
0
.
...
...
...
...
...
...
0
1
0
1
.
OFF
1
2
3
.
PHASE_
CORRECTION
DISABLED
ENABLED
PC1
0
1
.
.
.
.
.
1
1
1
1
...
127
3FSK/4FSK_SLICER_
THRESHOLD
ST3 ST2 ST1
ST7
...
VITERBI_PATH _
MEMORY
0
0
0
0
.
0
0
1
1
.
0
0
0
0
.
...
...
...
...
...
...
0
1
0
1
.
OFF
VM2 VM1
1
2
3
.
0
0
1
1
0
1
0
1
4 BITS
6 BITS
8 BITS
32 BITS
.
.
.
.
.
1
1
1
1
...
127
3FSK_PREMABLE_
TIME_VALIDATE
PTV4 PTV3 PTV2 PTV1
0
0
0
0
.
0
0
0
0
.
0
0
1
1
.
0
1
0
1
.
0
1
2
3
.
.
.
.
.
.
1
1
1
1
15
Figure 75. Register 13—3FSK/4FSK Demodulation Register Map
Rev. 0 | Page 57 of 60
ADF7021-V
REGISTER 14—TEST DAC REGISTER
ADDRESS
BITS
TEST_DAC_GAIN
TEST_DAC_OFFSET
EFx ED_LEAK_FACTOR
LEAKAGE =
ERx PULSE_EXTENSION
TGx TEST_DAC_GAIN
TE1 TEST_TDAC_EN
0
1
2
3
NO PULSE EXTENSION
EXTENDED BY 1
EXTENDED BY 2
0
NO GAIN
0
1
TEST DAC DISABLED
TEST DAC ENABLED
0
1
2
3
4
5
6
7
2^–8
2^–9
1
...
15
× 2^1
...
× 2^15
2^–10
2^–11
2^–12
2^–13
2^–14
2^–15
EXTENDED BY 3
PEx ED_PEAK_RESPONSE
0
1
2
3
FULL RESPONSE TO PEAK
0.5 RESPONSE TO PEAK
0.25 RESPONSE TO PEAK
0.125 RESPONSE TO PEAK
Figure 76. Register 14—Test DAC Register Map
The demodulator tuning parameters, PULSE_EXTENSION,
ED_LEAK_FACTOR, and ED_PEAK_RESPONSE, can be
enabled only by setting Register 15, Bits[DB7:DB4] to 0x9.
Whereas the correlators and filters are clocked by DEMOD CLK,
the test DAC is clocked by CDR CLK. Note that although the test
DAC functions in regular user mode, the best performance is
achieved when CDR CLK is increased to or above the frequency
of DEMOD CLK. The CDR block does not function when this
condition exists.
Using the On-Chip Test DAC
The on-chip test DAC can be used to implement analog
demodulation or to provide access for measurement of FSK
demodulator output SNR or CNR. For detailed information
about using the test DAC, see the AN-852 Application Note.
Programming Register 14 enables the test DAC. Both the linear
and correlator demodulator outputs can be multiplexed into
the DAC.
The test DAC allows the postdemodulator filter output for both
linear and correlator demodulators to be viewed externally. The
test DAC also takes the 16-bit filter output and converts it to a
high frequency, single-bit output using a second-order, error
feedback Σ-Δ converter. The output can be viewed on the SWD
pin. This signal, when filtered appropriately, can then be used to
do the following:
Register 14 allows a fixed offset term to be removed from the
signal (to remove the IF component in the linear demodulator
case). It also has a signal gain term to allow the usage of the
maximum dynamic range of the DAC.
•
Monitor the signals at the FSK postdemodulator filter
output. This allows the demodulator output SNR to be
measured. Eye diagrams of the received bit stream can
also be constructed to measure the received signal quality.
Provide analog FM demodulation.
•
Rev. 0 | Page 58 of 60
ADF7021-V
REGISTER 15—TEST MODE REGISTER
ANALOG_TEST_
MODES
PLL_TEST_
MODES
PFD/CP_
CLK_MUX TEST_MODES
Σ-Δ_TEST_
MODES
Tx_TEST_
MODES
Rx_TEST_
MODES
ADDRESS
BITS
COx CAL_OVERRIDE
PCx PFD/CP_TEST_MODES
0
1
2
3
AUTO CAL
OVERRIDE GAIN
OVERRIDE BW
0
1
2
3
4
5
6
7
DEFAULT, NO BLEED
(+VE) CONSTANT BLEED
(–VE) CONSTANT BLEED
(–VE) PULSED BLEED
(–VE) PULSE BLD, DELAY UP
CP PUMP UP
OVERRIDE BW AND GAIN
RD1 REG1_PD
CP TRISTATE
CP PUMP DN
0
1
NORMAL
POWER-DOWN
FH1 FORCE_LD_HIGH
SDx Σ-Δ_TEST_MODES
0
1
NORMAL
FORCE
0
1
2
3
4
5
6
7
DEFAULT, 3RD-ORDER Σ-Δ, NO DITHER
1ST-ORDER Σ-Δ
2ND-ORDER Σ-Δ
DITHER TO FIRST STAGE
DITHER TO SECOND STAGE
DITHERTO THIRD STAGE
DITHER × 8
AMx ANALOG_TEST_MODES
0
1
2
3
4
5
6
7
BAND GAP VOLTAGE
40µA CURRENT FROM REG4
FILTER I CHANNEL: STAGE 1
FILTER I CHANNEL: STAGE 2
FILTER I CHANNEL: STAGE 1
FILTER Q CHANNEL: STAGE 1
FILTER Q CHANNEL: STAGE 2
FILTER Q CHANNEL: STAGE 1
ADC REFERENCE VOLTAGE
BIAS CURRENT FROM RSSI 5µA
FILTER COARSE CAL OSCILLATOR OUTPUT
ANALOG RSSI ILCHANNE
OFFSET LOOP +VE FBACK V (I CH)
SUMMED OUTPUT OF RSSI RECTIFIER+
SUMMED OUTPUT OF RSSI RECTIFIER–
BIAS CURRENT FROM BB FILTER
DITHER × 32
TMx Tx_TEST_MODES
0
1
2
3
4
5
6
NORMAL OPERATION
Tx CARRIER ONLY
Tx +fDEV TONE ONLY
Tx –fDEV TONE ONLY
Tx "1010" PATTERN
Tx PN9 DATA SEQUENCE
8
9
10
11
12
13
14
15
Tx SWD PATTERN REPEATEDLY
RTx Rx_TEST_MODES
0
1
2
3
4
5
6
7
NORMAL
SCLK, SDATA
REVERSE I,Q
I,Q TO TxRxCLK, TxRxDATA
3FSK SLICER ON TxRxDATA
CORRELATOR SLICER ON TxRxDATA
LINEAR SLICER ON TxRxDATA
SDATA TO CDR
ADDITIONAL FILTERING ON I,Q
ENABLE REG 14 DEMOD PARAMETERS
POWER DOWN DDT AND ED IN T/4 MODE
ENVELOPE DETECTOR WATCHDOG DISABLED
RESERVED
PROHIBIT CAL ACTIVE
FORCE CAL ACTIVE
PMx PLL_TEST_MODES
I,Q
0
NORMAL OPERATION
1
R DIV
2
N DIV
3
RCNTR/2 ON MUXOUT
4
NCNTR/2 ON MUXOUT
5
ACNTR TO MUXOUT
6
7
8
9
10
11
12
PFD PUMP UP TO MUXOUT
PFD PUMP DNTO MUXOUT
S DATA TO MUXOUT (OR SREAD)
ANALOG LOCK DETECT ON MUXOUT
END OF COARSE CAL ON MUXOUT
END OF FINE CAL ON MUXOUT
FORCE NEW PRESCALER CONFIG
8
9
10
11
12
13
14
15
FOR ALL N
ENABLE DEMOD DURING CAL
13
14
15
TEST MUX SELECTS DATA
LOCK DETECT PRECISION
RESERVED
CMx CLK_MUX ON CLKOUT PIN
0
1
2
3
4
5
6
7
NORMAL, NO OUTPUT
DEMOD CLK
CDR CLK
SEQ CLK
BB OFFSET CLK
Σ-Δ CLK
ADC CLK
TxRxCLK
Figure 77. Register 15—Test Mode Register Map
•
•
Analog RSSI can be viewed on the TEST_A pin by setting
ANALOG_TEST_MODES (Bits[DB27:DB24]) to 11.
Tx_TEST_MODES can be used to enable modulation test.
•
The CDR block can be bypassed by setting Rx_TEST_
MODES to 4, 5, or 6, depending on the demodulator used.
Rev. 0 | Page 59 of 60
ADF7021-V
OUTLINE DIMENSIONS
0.30
0.23
0.18
7.00
BSC SQ
0.60 MAX
0.60 MAX
PIN 1
INDICATOR
37
36
48
1
PIN 1
INDICATOR
EXPOSED
PAD
(BOTTOM VIEW)
4.25
4.10 SQ
3.95
TOP
VIEW
6.75
BSC SQ
0.50
0.40
0.30
25
24
12
13
0.25 MIN
5.50
REF
0.80 MAX
0.65 TYP
1.00
0.85
0.80
12° MAX
FOR PROPER CONNECTION OF
THE EXPOSED PAD, REFER TO
THE PIN CONFIGURATION AND
FUNCTION DESCRIPTIONS
0.05 MAX
0.02 NOM
COPLANARITY
0.08
0.50 BSC
SECTION OF THIS DATA SHEET.
0.20 REF
SEATING
PLANE
COMPLIANT TO JEDEC STANDARDS MO-220-VKKD-2
Figure 78. 48-Lead Lead Frame Chip Scale Package [LFCSP_VQ]
7 mm × 7 mm Body, Very Thin Quad
(CP-48-3)
Dimensions shown in millimeters
ORDERING GUIDE
Model1
ADF7021-VBCPZ
ADF7021-VBCPZ-RL
EVAL-ADF70XXMBZ2
EVAL-ADF7021-VDB1Z
EVAL-ADF7021-VDB2Z
Temperature Range
−40°C to +85°C
−40°C to +85°C
Package Description
Package Option
CP-48-3
CP-48-3
48-Lead Lead Frame Chip Scale Package [LFCSP_VQ]
48-Lead Lead Frame Chip Scale Package [LFCSP_VQ]
Evaluation Platform Mother Board
450 MHz to 470 MHz Daughter Board
868 MHz to 870 MHz Daughter Board
1 Z = RoHS Compliant Part.
©2010 Analog Devices, Inc. All rights reserved. Trademarks and
registered trademarks are the property of their respective owners.
D08635-0-4/10(0)
Rev. 0 | Page 60 of 60
相关型号:
©2020 ICPDF网 联系我们和版权申明