AD AD6459

a
FEATURES
Fully Compliant with Standard and Enhanced GSM
Specification
–11 dBm Input 1 dB Compression Point
0 dBm Input Third Order Intercept
10 dB SSB Noise Figure (50 V)
DC-500 MHz RF and LO Bandwidths
Linear IF Amplifier
Linear-in-dB and Stable over Temperature
Voltage Gain Control
Quadrature Demodulator
On-Board Phase-Locked Quadrature Oscillator
Demodulates IFs from 5 MHz to 50 MHz
Low Power
8 mA at Midgain
2 mA Sleep Mode Operation
2.7 V to 5.5 V Operation
Interfaces to AD7013, AD7015 and AD6421 Baseband
Converters
20-Lead SSOP
GENERAL DESCRIPTION
The AD6459 is a 3 V, low power receiver IF subsystem for
operation at input frequencies as high as 500 MHz and IFs
from 5 MHz up to 50 MHz. It is optimized for operation in
GSM, DCS1800 and PCS1900 receivers. It consists of a mixer,
an IF amplifier, I and Q demodulators, a phase-locked quadrature oscillator, a precise AGC subsystem, and a biasing system
with external power-down.
The AD6459’s low noise, high intercept mixer is a doublybalanced Gilbert-Cell type. It has a nominal –11 dBm inputreferred 1 dB compression point and a 0 dBm input-referred
third-order intercept. The mixer section of the AD6459 also
includes a local oscillator (LO) preamplifier, which lowers the
required LO drive to –16 dBm.
The gain control input accepts an external gain-control voltage
input from an external AGC detector or a DAC. It provides an
80 dB gain range with 27 mV/dB gain scaling.
The I and Q demodulators provide in-phase and quadrature
baseband outputs to interface with Analog Devices’ AD7013
GSM 3 V Receiver IF Subsystem
AD6459
FUNCTIONAL BLOCK DIAGRAM
LO
I
RF
BPF
PLL
Q
GAIN
CONTROL
FREF
AD6459
(IS54, TETRA, MSAT) AD7015 and AD6421 (GSM,
DCS1800, PCS1900) baseband converters. An on-board
quadrature VCO that is externally phase-locked to the IF signal
drives the I and Q demodulators. This locked reference signal is
normally provided by an external VCTCXO under the control of
the radio’s digital processor. The AD6459 can also provide
demodulation of N-PSK and N-QAM in many non-TDMA
systems when used with external analog carrier recovery systems
such as the Costas Loop. Finally, the VCO can be phase-locked
to a frequency that is deliberately offset from the IF as in the
case of a Beat-Frequency oscillator (BFO) resulting in the
product detection of CW or SSB.
The AD6459 uses supply voltages from 2.7 V to 5.5 V over the
temperature range of –40°C to +85°C. Operation is enabled by a
CMOS logical level; response time is typically < 80 µs. When
disabled, the standby current is reduced to 2 µA.
The AD6459 comes in a 20-pin shrink small outline (SSOP)
surface mount package.
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
which may result from its use. No license is granted by implication or
otherwise under any patent or patent rights of Analog Devices.
One Technology Way, P.O. Box 9106, Norwood, MA 02062-9106, U.S.A.
Tel: 617/329-4700
World Wide Web Site: http://www.analog.com
Fax: 617/326-8703
© Analog Devices, Inc., 1996
AD6459–SPECIFICATIONS (@ T = +258C, V = 3.0 V, GREF = 1.2 V, unless otherwise noted)
A
P
Model
Parameter
Conditions
DYNAMIC PERFORMANCE
MIXER
Maximum RF and LO Frequency
AGC Conversion Gain Variation
Input 1 dB Compression Point
Input Third-Order Intercept
SSB Noise Figure1
Mixer Output Bandwidth at MXOP
0.2 V < VGAIN < 2.25 V
@ VGAIN = 0.2 V
@ VGAIN = 0.2 V
@ ZS = 50 Ω, FRF = 240 MHz, FLO = 229.3 MHz at –16 dBm
@ –3 dB
500
–3 to +16
–11
0
10
80
MHz
dB
dBm
dBm
dB
MHz
IF AMPLIFIERS
AGC Gain Variation
Input Referred Noise
Input Resistance
Bandwidth
0.2 V < VGAIN <2.25 V
AC Short Circuit Input
@ VGAIN = 0.2 V
@ –3 dB
–13 to +46
3
5
50
dB
nV/√Hz
kΩ
MHz
I AND Q DEMODULATORS
Demodulation Gain
Output Voltage Range
Output Voltage Common-Mode Level
Output Offset Voltage
Error in Quadrature
Amplitude Match
I/Q Output Bandwidth
Output Resistance
Differential, IRXP, IRXN, QRXP, QRXN
(Not Power Supply Dependent)
Differential, VGAIN = GREF
Differential from I to Q, IF = 13 MHz
I to Q
CLOAD = 10 pF
Each Pin
GAIN CONTROL
Total Gain Control Range
Control Voltage Range at GAIN
Gain Scaling
Gain Law Conformance
Bias Current at GREF
Input Resistance at GAIN
PLL
Frequency Range
Phase Noise
Acquisition Time
Input Drive Level (FREF)
POWER-DOWN INTERFACE
Logical Threshold
Input Current for Logical High
Turn-On Response Time
Turn-Off Response time
Standby Current
POWER SUPPLY
Supply Range
Supply Current
OPERATING TEMPERATURE
TMIN to TMAX
Min
AD6459ARS
Typ
Max
17
0.3
1.5
–150
1.5
0.25
2
4.7
Mixer + IF + Demod, 0.2 V < VGAIN <2.25 V
dB
VP – 0.2 V
V
150
mV
3.5
Degree
dB
MHz
kΩ
76
0.2
23
27
± 0.5
0.5
20
5
2.4
32
50
0.5
80
IF = 19.5 MHz, Using Suggested Filter
100
Power Up on Logical High
VPOS
1.5
75
80
1
2
To Fully Meet Specifications (PLL Lock)
To 200 µA Supply Current
2.7
@ VGAIN = 1.2 V
–40
–25
dB
V
mV/dB
dB
µA
kΩ
MHz
Degree rms
µs
mV
V
µA
µs
µs
µA
5.5
V
mA
+85
+85
°C
°C
8
Operation to 3.3 V Minimum Supply Voltage
Operation to 2.7 V Minimum Supply Voltage
Units
NOTES
1
Including IF noise and using suggested filter, at V GAIN = 0.2 V.
Specifications subject to change without notice.
–2–
REV. 0
AD6459
ABSOLUTE MAXIMUM RATINGS 1
PIN CONNECTION
Supply Voltage VPS1, VPS2 to COM1, COM2 . . . . . +5.5 V
Internal Power Dissipation2 . . . . . . . . . . . . . . . . . . . 600 mW
Operating Temperature Range . . . . . . . . . . . –40°C to +85°C
Storage Temperature Range . . . . . . . . . . . . –65°C to +150°C
Lead Temperature, Soldering (60 sec) . . . . . . . . . . . . +300°C
20-Pin SSOP (RS-20)
NOTES
1
Stresses above those listed under “Absolute Maximum Ratings” may cause
permanent damage to the device. This is a stress rating only, and 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 rating conditions for extended periods
may affect device reliability.
2
Thermal Characteristics: 20-lead SSOP package: θJA = 126°C/W.
AD6459ARS
Temperature
Range
Package
Description
Package
Option
–25°C to +85°C
for 2.7 V to 5.5 V
–40°C to +85°C
for 3.3 V to 5.5 V
20-Pin Plastic
SSOP
RS-20
20 VPS1
19 FLTR
PRUP 3
18 VPS2
LOIP 4
RFLO 5
17 IRXP
AD6459
16 IRXN
TOP VIEW
RFHI 6 (Not to Scale) 15 QRXP
ORDERING GUIDE
Model
FREF 1
COM1 2
COM2 7
14 QRXN
GREF 8
13 GAIN
MXOP 9
12 IFIM
MXOM 10
11 IFIP
PIN DESCRIPTIONS
Pin
Pin
Label
Description
Function
1
FREF
Frequency Reference Input
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
COM1
PRUP
LOIP
RFLO
RFHI
COM2
GREF
MXOP
MXOM
IFIP
IFIM
GAIN
QRXN
QRXP
IRXN
IRXP
VPS2
FLTR
VPS1
Common 1
Power Up Input
Local Oscillator Input
RF “Low” Input
RF “High” Input
Common 2
Gain Reference Input
Mixer Output “Plus”
Mixer Output “Minus”
IF Input “Plus”
IF Input “Minus”
Gain Control Input
Q Output “Negative”
Q Output “Positive”
I Output “Negative”
I Output “Positive”
VPOS Supply 2
PLL Loop Filter
VPOS Supply 1
Demodulation LO Input. May either be 3 V CMOS input or >100 mV p-p.
AC-coupled for lowest stand by current.
Ground.
CMOS Compatible Power-Up Control; <1.5 V = OFF, >1.5 V = ON.
AC-Coupled LO Input. 50 mV p-p drive needed, 500 mV p-p max.
Mixer Differential Input. AC-coupled.
Mixer Differential Input. AC-coupled.
Ground.
High Impedance Input. Sets gain scaling, typically 1.2 V.
Differential Output of the Mixer. See Figure 22.
Differential Output of the Mixer. See Figure 22.
Differential Input of Variable Gain Amplifier. AC-coupled.
Differential Input of Variable Gain Amplifier. AC-coupled.
0.2 V–2.4 V Using 3 V Supply. Max gain at 0.2 V.
Differential Q Output. Output resistance 4.7 kΩ.
Differential Q Output. Output resistance 4.7 kΩ.
Differential I Output. Output resistance 4.7 kΩ.
Differential I Output. Output resistance 4.7 kΩ.
Supply Voltage.
Series RC Loop Filter. Connected to VPS2.
Supply Voltage.
CAUTION
ESD (electrostatic discharge) sensitive device. Electrostatic charges as high as 4000 V readily
accumulate on the human body and test equipment and can discharge without detection.
Although the AD6459 features proprietary ESD protection circuitry, permanent damage may
occur on devices subjected to high energy electrostatic discharges. Therefore, proper ESD
precautions are recommended to avoid performance degradation or loss of functionality.
REV. 0
–3–
WARNING!
ESD SENSITIVE DEVICE
AD6459
R1
20kΩ
FREF
C1
0.1µF
R8 1kΩ
PRUP
C2 1nF
R2
50Ω
R3
50Ω
C3
1nF
GREF
VPOS
OPEN
R4
OPEN
R5
FLTR 19
3 PRUP
VSP2 18 C10 1nF
C11
0.1µF
IRXP
IRXP 17
AD6459
5 RFLO
IRXN 16
6 RFHI
QRXP 15
7 COM2
QRXN 14
8 GREF
GAIN 13
9 MXOP
IFIM 12
10 MXOM
IFIP 11
IRXN
QRXP
C9
10nF
(BOTTOM)
C7
1nF
C13
10nF
MXOP
2 COM1
4 LOIP
C4 1nF
RFHI
C5
1nF
VPOS
VSP1 20
1 FREF
R9
50Ω
LOIP
VPOS
C12
1nF
C6
1nF
MXOM
R6
50Ω
QRXN
GAIN
C8
1nF
R7
50Ω
IFIM
IFIP
Figure 1. AD6459 Characterization Board
VPOS
FREF
1
VP 8
2
7
3
A=1 6
4
PRUP
PRUP
LOIP
LOIP
RFIP
RFIP
GREF
FREF
AD830
VN 5
VPOS
VP
R4
50Ω IOUT
C6
0.1µF
VN
C7
0.1µF
IRXP
IRXN
AD6459
CHARACTERIZATION
BOARD
GREF
MXOP MXOM
IFIP
1
VP 8
QRXP
2
7
QRXN
3
A=1 6
GAIN
4
AD830
VN 5
IFIN
VP
R3
50Ω IOUT
C4
0.1µF
C5
0.1µF
VN
GAIN
VP 8
1
2
7
3
A=1 6
4
AD830
VP
C3
0.1µF
VN 5
C2
0.1µF
VP 8
1
VN
2
7
3
A=1 6
4
AD830
R6
50Ω
C11
0.1µF
VN 5
C10
0.1µF
R2
50Ω
MXOP
VP
VN
VP 8
1
2
7
3
A=1 6
4
AD830
VP
R5
50Ω
C9
0.1µF
VN 5
C8
0.1µF
VN
R1
50Ω
IFIN
Figure 2. Characterization Test Set
–4–
REV. 0
AD6459
20
20
18
RIN = 50Ω, IF = 13MHz
14
RIN = 50Ω, F = 45MHz
10
GAIN – dB
16
SSB NF – dB
VGAIN = 0.2V
15
RIN = 50Ω, IF = 26MHz
RIN = 1kΩ, IF = 13MHz
12
VGAIN = 1.0V
5
0
10
VGAIN = 2.25V
–5
8
RIN = 400Ω, IF = 13MHz
6
50
–10
100
150
200
250
300
RF FREQUENCY – dB
350
450
400
6
Figure 3. Mixer Noise Figure vs. RF Frequency
2000
4.6
RESISTANCE – Ω
1400
4.4
1200
4.2
C SHUNT
VGAIN = 1.0V
1000
4.0
800
3.8
600
3.6
400
3.4
0
50
C SHUNT
VGAIN = 2.2V
100
150
200 250 300 350 400
RF FREQUENCY – MHz
450
500
46
AMP/DEMOD, VPOS = 5.5V
40
30
MIXER, VPOS = 2.7V
20
MIXER, VPOS = 5.5V
3.2
R SHUNT
VGAIN = 0.2V
42
50
10
200
38
60
GAIN – dB
1600
22
26
30
34
RF FREQUENCY – MHz
AMP/DEMOD, VPOS = 2.7V
4.8
C SHUNT
VGAIN = 0.2V
18
70
CAPACITANCE – pF
1800
14
Figure 6. Mixer Conversion Gain vs. IF Frequency,
TA = +25°C, VPOS = 2.7 V, VREF = 1.2 V, FRF = 250 MHz
5.0
R SHUNT
VGAIN = 2.2V
10
0
–50 –40 –30 –20 –10
3.0
550
0
10
20 30
40
50
60
70 80
90
TEMPERATURE – °C
Figure 4. Mixer Input Impedance vs. RF Frequency,
VPOS = 2.7 V, TA = +25°C
Figure 7. Mixer Conversion Gain and IF Amplifier/
Demodulator Gain vs. Temperature, VGAIN = 0.2 V,
VREF = 1.2 V , FIF = 26 MHz, FRF = 250 MHz
–9
20
VPOS = 5.5V
TA = +85°C
INPUT 1dB COMPRESSION POINT
REFERED TO 50Ω – dBm
15
VGAIN = 0.2V
GAIN – dB
10
VGAIN = 1.0V
5
0
VGAIN = 2.25V
–5
–10
50
–11
100
150
200
250
300
RF FREQUENCY – MHz
350
400
450
VPOS = 5.5V
TA = +25°C
VPOS = 2.7V
TA = +25°C
–12
VPOS = 2.7V
TA = –25°C
–13
–14
–15
Figure 5. Mixer Conversion Gain vs. RF Frequency,
TA = +25°C, VPOS = 2.7 V, VREF = 1.2 V, FIF = 26 MHz
REV. 0
–10
VPOS = 5.5V
TA = –45°C
0
0.5
1
1.5
GAIN VOLTAGE – Volts
2
2.5
Figure 8. Mixer Input 1 dB Compression Point vs.
VGAIN, VREF = 1.2 V, FRF = 250 MHz, FIF = 26 MHz
–5–
AD6459
1
70
VGAIN = 0.2V
0.8
60
50
0.4
ERROR – dB
IF AMP/DEMOD GAIN – dB
0.6
VGAIN = 1.0V
40
30
VGAIN = 1.5V
MIXER
0.2
0
–0.2
IF AMP/DEMOD
–0.4
20
–0.6
10
0
–0.8
VGAIN = 2.25V
–1.0
5
10
15
20
25
30
35
INTERMEDIATE FREQUENCY – dB
40
0
45
RESISTANCE – Ω
C SHUNT, VGAIN= 1.0V
8000
6000
R SHUNT, VGAIN= 1.0V
2000
20
30
40
50
60
70
IF FREQUENCY – MHz
2.0
1.0
0
10
2.5
1.5
R SHUNT, VGAINS= 0.2V
0
3.0
2.5
C SHUNT, VGAIN = 2.2V
4000
3.0
QUADRATURE ERROR – Degrees
R SHUNT, VGAIN = 2.2V
10000
3.5
CAPACITANCE – pF
C SHUNT, VGAIN = 0.2V
80
90
1
1.5
GAIN VOLTAGE – Volts
2
2.5
Figure 12. AD6459 Gain Error vs. Gain Control
Voltage, Representative Part
Figure 9. IF Amplifier and Demodulator Gain vs.
Frequency, TA = +25°C, VPOS = 2.7 V, VREF = 1.2 V
12000
0.5
2.0
1.5
1.0
0.5
0
0.5
100
Figure 10. IF Amplifier Input Impedance vs.
Frequency, TA = +25°C, VPOS = 2.7 V, VREF = 1.2 V
5
10
15
20
25
30
35
FREF FREQUENCY – MHz
40
45
Figure 13. Demodulator Quadrature Error vs.
FREF Frequency, TA = +25°C, VPOS = 2.7 V
–90
–5
–95
–15
–20
PHASE NOISE – dBc
INPUT 1dB COMPRESSION POINT
REFERED TO 50Ω – dBm
–10
–25
–30
–35
–40
–45
–100
–105
–110
–115
–50
–55
0
0.5
1
1.5
GAIN VOLTAGE – Volts
2
–120
0.1
2.5
Figure 11. IF Amplifier/Demodulator Input 1 dB
Compression Point vs. VGAIN , FIF = 19.5 MHz,
VREF = 1.2 V, TA = +25°C, VPOS = 2.7 V
1
10
100
CARRIER FREQUENCY – kHz
1k
10k
Figure 14. PLL Phase Noise vs. Frequency,
VPOS = 3 V, C10 = 1 nF, FREF = 13 MHz
–6–
REV. 0
–0.1
0
–0.3
–10
INPUT IP3 REFERED TO 50Ω – dBm
FLTR PIN VOLTAGE
REFERENCED TO VPOS – Volts
AD6459
–0.5
–0.7
–0.9
–1.1
–1.3
5
10
15
20
25
30
35
40
PLL FREQUENCY – MHz
45
50
–40
–50
–60
55
0.5
1.0
1.5
GAIN VOLTAGE – Volts
2.0
2.5
Figure 17. System (Mixer + IF LC Filter + I F Amplifier +
Demodulator) IP3 vs. Gain, TA = +25°C, VPOS = 2.7 V,
IF = 13 MHz, VREF = 1.2 V
Figure 15. PLL Loop Voltage at FLTR Pin (KVCO) vs.
Frequency
–10
18
–20
16
VPOS = 2.7V, TA = +85°C
SUPPLY CURRENT – mA
INPUT 1dB COMPRESSION POINT
REFERED TO 50Ω – dBm
–30
–70
–1.5
–30
–40
–50
–60
14
8
–80
4
1.0
1.5
GAIN VOLTAGE – Volts
2.0
2.5
VPOS = 5.5V, TA = +85°C
10
6
0.5
VPOS = 2.7V, TA = +25°C
12
–70
VPOS = 5.5V, TA = +25°C
VPOS = 5.5V, TA = –40°C
0
0.5
1
1.5
GAIN VOLTAGE – Volts
2
2.5
Figure 18. Power Supply Current vs. Gain Control
Voltage, VREF = 1.2 V
Figure 16. System (Mixer + IF LC Filter +IF Amplifier +
Demodulator) 1 dB Compression Point vs. Gain,
TA = +25°C, VPOS = 2.7 V, FIF = 13 MHz, VREF = 1.2 V
REV. 0
–20
–7–
AD6459
PRODUCT OVERVIEW
mixer is determined, at the upper end, by the maximum input
signal level of ± 90 mV (–11 dBm in 50 Ω between RFHI and
RFLO) up to which the mixer remains essentially linear, and at
the lower end, by the noise level. It is customary to define the
linearity of a mixer in terms of its 1 dB gain-compression point
and third-order intercept, which for the AD6459 are –11 dBm
and 0 dBm, respectively, in a 50 Ω system.
The AD6459 provides most of the active circuitry required to
realize a complete low power, single-conversion superheterodyne receiver, or the latter part of a double-conversion receiver,
at input frequencies up to 500 MHz, with an IF from 5 MHz to
50 MHz. The internal I/Q demodulators, and their associated
phase-locked loop, support a wide variety of modulation modes,
including n-PSK, n-QAM and GMSK. A single positive supply
voltage of 3 V is required (2.7 V minimum, 5.5 V maximum) at
a typical supply current of 8 mA at midgain. In the following
discussion, VPOS will be used to denote the power supply voltage,
which will be normally assumed to be 3 V.
The mixer’s RF input port is differential; that is, pin RFLO is
functionally identical to RFHI, and these nodes are internally
biased. The RF port can be modeled as a parallel RC circuit as
shown in Figure 19.
Figure 20 shows the main sections of the AD6459. It consists of
a variable-gain UHF mixer and a linear two-stage IF strip,
which together provide a calibrated voltage-controlled gain range
of more than 76 dB, followed by dual quadrature demodulators.
These are driven by inphase and quadrature clocks that are
generated by a Phase-Locked Loop (PLL), which is locked to a
corrected external reference. A CMOS-compatible power-down
interface completes the AD6459.
RFHI
CSH
RSH
RFLO
Figure 19. Mixer Port Modeled as a Parallel RC Network
The local oscillator (LO) input is internally biased at VP–0.8 V
and must be ac coupled. The LO interface includes a preamplifier that minimizes the drive requirements, thus simplifying the
oscillator design and reducing LO leakage from the RF port.
The LO requires a single-sided drive of ± 50 mV, or –16 dBm
in a 50 Ω system. For operation above 300 MHz, noise figure
can be improved by increasing the LO level.
Mixer
The UHF mixer is an improved Gilbert-cell design and can
operate from low frequencies (it is internally dc-coupled) up to
an RF input of 500 MHz. The dynamic range at the input of the
LOIP
4
4.7kΩ
17 IRXP
16 IRXN
RFHI
6
RFLO 5
MXOP
MXOM
9
10
LC
BANDPASS
FILTER
11
0°
IFIP
4.7kΩ
1 FREF
+
PLL
12
–
19 FLTR
IFIM
50°
4.7kΩ
15 QRXP
14 QRXN
4.7kΩ
AGC VOLTAGE
GAIN TO
COMPENSATION
13 GAIN
VPS1 20
VPS2 18
AD6459
BIAS
CIRCUIT
8 GREF
PRUP 3
2
7
COM1
COM2
Figure 20. Functional Block Diagram
–8–
REV. 0
AD6459
The output of the mixer is differential. The nominal conversion
gain is specified for operation into a 19.5 MHz LC IF bandpass
filter as shown in Figure 21 and Table I.
The conversion gain is measured between the mixer input and
the input of this filter and varies between –5 dB and +15 dB.
C1
MXOP
IFIP
MXOM
The AD6459’s overall gain, expressed in decibels, is linear with
respect to the AGC voltage VGAIN at pin GAIN. The gain of all
sections is maximum when VGAIN is 0.2 V and falls off as the
bias is increased to VGAIN = 2.25 V. The gain is independent
of the power supply voltage. The gain of all stages changes
simultaneously. The AD6459’s gain scaling is also temperature compensated.
Note that GAIN pin of the AD6459 is an input driven by an
external low impedance voltage source, normally a DAC, under
the control of the radio’s digital processor.
L1
C2
Gain Scaling
IFIM
C1
Figure 21. Suggested IF Filter Inserted Between the
Mixer’s Output Port and the Amplifier’s Input Port
Table I. Filter Component Values for Selected Frequencies
Frequency
C1
L1
C2
13 MHz
19.5 MHz
26 MHz
40 MHz
27 pF
27 pF
22 pF
22 pF
0.82 µH
0.56 µH
0.39 µH
0.12 µH
180 pF
110 pF
82 pF
100 pF
The maximum permissible signal level between MXOP and
MXOM is determined by the maximum gain control voltage.
The mixer output port, having pull-up resistors of 250 Ω to
VPOS, is shown in Figure 22.
250Ω
MXOP
VPOS
250Ω
The gain-control scaling is directly proportional to the reference
voltage applied to the pin GREF and is independent of the
power supply voltage. When this input is set to the nominal
value of 1.2 V, the scale is nominally 27 mV/dB (37 dB/V).
Under these conditions, 76 dB of gain range (mixer plus IF)
corresponds to a control voltage of 0.2 V ≤ VGAIN % 2.25 V.
The final centering of this 2.05 V range depends on the insertion losses of the IF filters used.
Pin GREF can be tied to an external voltage reference (VREF)
provided, for example, by an AD1580 (1.21 V) voltage reference.
When using the Analog Devices AD7013 (IS54, TETRA, and
satellite receiver applications) and AD7015 or AD6421 (GSM,
DCS1800, PCS1900) baseband converters, the external reference may also be provided by the reference output of the
baseband converters. The interface between the AD6459 and
the AD6421 baseband converter is shown in Figure 24. The
AD7015 baseband converter provides a VR of 1.23 V. An auxiliary DAC in the AD7015 can be used to generate the AGC
voltage. Since it uses the same reference voltage, the numerical
input to this DAC provides an accurate RSSI value in digital
form, no longer requiring the reference voltage to have high
absolute accuracy.
MXOM
AD6459
Figure 22. Mixer Output Port
AD6421
IRXP
IRXP
IF Amplifier
100pF
100pF
IRXP
QRXP
IRXN
QRXP
100pF
100pF
BREFOUT
GREF
The IF input is differential, at IFIP and IFIM. Figure 23 shows
a simplified schematic of the IF interface modeled as parallel
RC network.
BREFCAP
0.1µF
160Ω
The IF’s small-signal bandwidth is approximately 50 MHz from
IFIP and IFIM through the demodulator.
AGC DAC
GAIN
1nF
IFHI
FREF
CSH
IRXP
IRXN
Most of the gain in the AD6459 is provided by the IF amplifier
strip, which comprises two stages. Both are fully differential and
each has a gain span of 26 dB for the AGC voltage range of 0.2
V to 2.25 V. Thus, in conjunction with the variable gain of the
mixer, the total gain span is 76 dB. The overall IF gain varies
from –13 dB to 45 dB for the nominal AGC voltage of 0.2 V to
2.25 V. Maximum gain is at VGAIN = 0.2 V.
VCTCXO
AFC DAC
RSH
IFLO
Figure 24. Interfacing the AD6459 to the AD6421
Baseband Converter
Figure 23. IF Amplifier Port Modeled as a Parallel RC
Network
REV. 0
–9–
AD6459
I/Q Demodulators
Both demodulators (I and Q) receive their inputs internally
from the IF amplifiers. Each demodulator comprises a full-wave
synchronous detector followed by an 8 MHz, two-pole low-pass
filter, producing differential outputs at pins IRXP and IRXN,
and QRXP and QRXN. Using the I and Q demodulators for
IFs above 50 MHz is precluded by the 5 MHz to 50 MHz range
of the PLL used in the demodulator section.
the VFQO always provides quadrature between its own I and Q
outputs, but the phasing between it and the reference carrier
will swing around the final value during the PLL’s settling time.
Bias System
The AD6459 operates from a single supply (VPOS) usually 3 V,
at a typical supply current of 8 mA at midgain and TA = +25°C,
corresponding to a power consumption of 24 mW. Any voltage
from 2.7 V to 5.5 V may be used.
The I and Q outputs are differential and can swing up to
2.2 V p-p at the low supply voltage of 2.7 V. They are nominally
centered at 1.5 V, independent of power supply. They can
therefore directly drive the RX ADCs in the AD7015 baseband
converter, which require an amplitude of 1.23 V to fully load
them when driven by a differential signal. The conversion gain
of the I and Q demodulators is 17 dB.
The bias system includes a fast-acting active high CMOScompatible power-up switch, allowing the part to idle at 2 µA
when disabled. Biasing is generally proportional-to-absolutetemperature (PTAT) to ensure stable gain with temperature.
Other special biasing techniques are used to ensure very
accurate gain, stable over the full temperature range.
For IFs of less than 8 MHz, the on-chip low-pass filters (8 MHz
cutoff) do not adequately attenuate the IF or feedthrough
products; thus, the maximum input voltage must be limited to
allow sufficient headroom at the I and Q outputs for not only
the desired baseband signal but also the unattenuated higherorder demodulation products. These products can be removed
by an external low-pass filter. A simple 1-pole RC filter with its
corner above the modulation bandwidth is sufficient to attenuate undesired outputs. The design of the RC filter is eased by
the 4.7 kΩ resistor integrated at each I and Q output pin.
USING THE AD6459
Phase-Locked Loop
The demodulators are driven by quadrature signals that are
provided by a variable-frequency quadrature oscillator (VFQO),
phase-locked to a reference signal applied to pin FREF. When
this signal is at the IF, inphase and quadrature baseband
outputs are generated at the I output (IRXP and IRXN) and Q
output (QRXP and QRXN), respectively. The quadrature
accuracy of this VFQO is typically within ± 1.5° at 19.5 MHz. A
simplified diagram of the FREF input is shown in Figure 25.
VPOS
In this section, we will focus on a few areas of special importance and include a few general application tips. As with any
wideband high gain component, great care is needed in PC
board layout. The location of the particular grounding points
must be considered with due regard to the possibility of
unwanted signal coupling.
The high sensitivity of the AD6459 leads to the possibility that
unwanted local EM signals may have an effect on the performance. During system development, carefully-shielded test
assemblies should be used. The best solution is to use a fully
enclosed box enclosing all components with the minimum
number of needed signal connectors (RF, LO, I and Q outputs)
in miniature coax form.
Gain Distribution
As with all receivers, the most critical decisions in effectively
using the AD6459 relate to the partitioning of gain between the
various subsections (Mixer, IF Amplifier/Demodulator) and the
placement of filters to achieve the highest overall signal-to-noise
ratio and lowest intermodulation distortion.
Figure 26 shows an example of the main RF/IF signal path at
maximum and minimum signal levels.
5kΩ
20kΩ
FREF
SIGNAL LEVEL
IN dBm
5kΩ
–10
–20
50µA PTAT
–30
–15dBm
–19dBm
–19dBm
–22dBm
–16dBm
35mV
–19dBm
–36dBm
–40
–50
Figure 25. Simplified Schematic of the FREF interface
–60
–70
The VFQO operates from 5 MHz to 50 MHz and is controlled
by the voltage between VPOS and FLTR. In normal operation a
series RC network, forming the PLL loop filter, is connected
from FLTR to VPOS. The use of an integral sample-hold system
ensures that the frequency-control voltage on pin FLTR remains
held during power-down, so reacquisition of the carrier occurs
in less than 80 µs.
In practice, the probability of a phase mismatch at power-up is
high, so the worst case linear settling period to full lock needs to
be considered in making filter choices. This is typically < 80 µs for
a quadrature phase error of ± 3° at an IF of 19.5 MHz. Note that
–79dBm
–80
–90
–100
–95dBm
–82dBm
–76dBm
–79dBm
–99dBm
MIXER
CONVERSION
GAIN
IF INPUT
250 MHz
3dB
FILTER GAIN
IF
GAIN
DEMOD. CONSTANT
CONV. BASEBAND
OUTPUT
GAIN
I
Q
Figure 26. Signal Levels and Gain, Showing 76 dB Typical
and 80 dB Maximum Range in an Example Application
–10–
REV. 0
AD6459
R1
20kΩ
VPOS
FREF
R9
50Ω
C2
1nF
LOIP
C4 1nF
R2
50Ω
RFHI
C3
SHORT
R3
50Ω
C7
1nF
GREF
VPOS
C5
0.1µF
R6
24.9kΩ
GND
C1
0.1µF
C12
1nF
1
FREF
VPS1 20
2
COM1
FLTR 19
R8 1kΩ
3
PRUP
VPS2 18
4
LOIP
IRXP 17
AD6459
5
RFLO
IRXN 16
6
RFHI
QRXP 15
7
COM2
QRXN 14
8
GREF
GAIN 13
9 MXOP
IFIM 12
10 MXOM
IFIP 11
JUMPER
L3
SHORT
C16
22pF
L4
SHORT
C17
22pF
R7
16.9kΩ
C10 1nF
C15
110pF
C11
0.1µF
PRUP
VPOS
IRXP
IRXN
QRXP
QRXN
GAIN
C9
10nF
GREF
L2
0.56µH
Figure 27. Evaluation Board as Received with 19.5 MHz Filter
Table II. AD6459 Evaluation Board Input and Output Connection
Reference
Designation
Connector
Type
Description
Coupling
Approximate
Signal Level
RFHI
SMA
RF Input
AC
–11 dBm max
LOIP
SMA
LO Input
AC
500 mV p-p max
FREF
SMA
AC
100 mV p-p min
MXOP
SMA
Demodulator Reference
Input
Mixer Output
NA
NA
IFIP
SMA
IF Input
NA
NA
J1
GREF
Jumper
J2-1
On-Board GREF Bias
External Reference Input
DC
DC
0.4 VPOS
1.2 V dc
GAIN
QRXN
QRXP
IRXN
IRXP
VPOS
J2-2
J2-3
J2-4
J2-5
J2-6
J2-7
DC
DC–2 MHz
DC–2 MHz
DC-2 MHz
DC-2 MHz
DC
0.2 V to 2.4 V dc
NA
NA
NA
NA
2.7 V to 5.5 V
PRUP
J2-8
Gain Bias Input
Q-Negative Output
Q-Positive Output
I-Negative Output
I-Positive Output
Power Supply
Positive Input
Power Up
DC-2 MHz
CMOS
GND
GND
J2-J9
J2-10
Ground
Ground
DC
DC
0V
0V
REV. 0
–11–
Comments
Input Is Terminated
in 50 Ω
Input Is Terminated
in 50 Ω
Input Is Terminated
in 50 Ω
Not Connected
for Unbalanced Output
Use XFMR
Not Connected
for Unbalanced Output
Use XFMR
Two Resistors Divider
Gain Scaling Reference
from External ADC
Maw Gain at 0.2 V
Z Series = 4.7 kΩ
Z Series = 4.7 kΩ
Z Series = 4.7 kΩ
Z Series = 4.7 kΩ
Supply Voltage
If Left Unconnected,
Board Is Active
NA
NA
AD6459 EVALUATION BOARD
Full Path Configuration
The AD6459 evaluation board (Figure 27) consists of a
AD6459, ground plane, I/O connectors, and a 19.5 MHz band
pass filter. The RF, LO and FREF ports are terminated in 50 Ω
to provide a broadband match or external signal generators.
As received, the board is configured for full-path evaluation
from RFHI to the I and Q outputs. The one-pole LC resonant
circuit provided represents a simple, yet balanced, IF bandpass
filtering approach. The filter supplied is centered at 19.5 MHz,
a common GSM intermediate frequency. Table I highlights the
filter component values for other IF frequencies. RFHI and
RFLO are true differential inputs, however for testing convenience, the RFLO terminal of the AD6459 is ac referenced to
ground on the evaluation board. The GAIN bias input, which is
bypassed with a 10 nF capacitor, is brought out to the interface
connector. The PRUP input is provided with a 20 kΩ pull up
resistor to VPOS that activates the board.
The board provides SMA connectors for the RF, LO, demodulator reference, mixer output and IF input signals. The MXOP
and IFIP connectors are left unconnected and are provided as a
testing convenience. Footprints for broadband matching transformers and matching components are also provided to aid in
stage breakout testing.
The remaining low frequency signals, including the I and Q
interface, bias and power connections are made via a dual row
pin header that acts as an Interface Connector located along the
edges of the board. An on-board gain-reference 1.2 V biasing
option is provided via a single jumper, J1. The evaluation board
will not function without this jumper unless an external bias
GREF is provided from an external reference that is normally
provided by the associated ADC.
C2204–12–10/96
AD6459
The four differential I and Q outputs are brought out unconditioned, directly to the interface connector. A high impedance,
high bandwidth FET-type probe should be used when measuring the I and Q ports. Excessive capacitive or resistive loading of
these ports will severely limit the video bandwidth and signal
swing. The demodulator PLL filter installed on the evaluation
board (R8, C10) can accommodate the full VFQO lock range
specified.
OUTLINE DIMENSIONS
Dimensions shown in inches and (mm).
0.295 (7.50)
0.271 (6.90)
11
1
10
0.311 (7.9)
0.301 (7.64)
0.212 (5.38)
0.205 (5.21)
20
0.07 (1.78)
0.066 (1.67)
0.008 (0.203)
0.002 (0.050)
0.0256
(0.65)
BSC
SEATING 0.009 (0.229)
PLANE
0.005 (0.127)
8°
0°
0.037 (0.94)
0.022 (0.559)
PRINTED IN U.S.A.
0.078 (1.98) PIN 1
0.068 (1.73)
–12–
REV. 0