1.5 GHz to 2.4 GHz frequency range
Continuous magnitude control of −4.5 dB to −34.5 dB
Continuous phase control of 0° to 360°
Output third-order intercept 17.5 dBm
Output 1 dB compression point 8.5 dBm
Output noise floor −150.5 dBm/Hz @ full gain
Adjustable modulation bandwidth up to 230 MHz
Fast output power disable
4.75 V to 5.25 V single-supply voltage
APPLICATIONS
RF PA linearization/RF predistortion
Amplitude and phase modulation
Variable attenuators and phase shifters
CDMA2000, WCDMA, GSM/EDGE linear power amplifiers
Smart antennas
GENERAL DESCRIPTION
The AD8341 vector modulator performs arbitrary amplitude
and phase modulation of an RF signal. Since the RF signal path
is linear, the original modulation is preserved. This part can be
used as a general-purpose RF modulator, a variable attenuator/phase shifter, or a remodulator. The amplitude can be
controlled from a maximum of −4.5 dB to less than −34.5 dB,
and the phase can be shifted continuously over the entire 360°
range. For maximum gain, the AD8341 delivers an OP1dB of
8.5 dBm, an OIP3 of 17.5 dBm, and an output noise floor of
−150.5 dBm/Hz, independent of phase. It operates over a
frequency range of 1.5 GHz to 2.4 GHz.
The baseband inputs in Cartesian I and Q format control the
amplitude and phase modulation imposed on the RF input
signal. Both I and Q inputs are dc-coupled with a ±500 mV
differential full-scale range. The maximum modulation bandwidth is 230 MHz, which can be reduced by adding external
capacitors to limit the noise bandwidth on the control lines.
RF Vector Modulator
AD8341
FUNCTIONAL BLOCK DIAGRAM
VPRF
90°
RFIP
RFIM
0°
CMOP
Figure 1.
Both the RF inputs and outputs can be used differentially or
single-ended and must be ac-coupled. The RF input and output
impedances are nominally 50 Ω over the operating frequency
range. The DSOP pin allows the output stage to be disabled
quickly in order to protect subsequent stages from overdrive.
The AD8341 operates off supply voltages from 4.75 V to 5.25 V
while consuming approximately 125 mA.
The AD8341 is fabricated on Analog Devices’ proprietary, high
performance 25 GHz SOI complementary bipolar IC process. It
is available in a 24-lead, Pb-free LFCSP package and operates
over a −40°C to +85°C temperature range. Evaluation boards
are available.
VPS2QBBMQBBP
RFOP
RFOM
DSOPIBBMIBBP
04700-001
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
registered trademarks are the property of their respective owners.
VS = 5 V, TA = 25°C, ZO = 50 Ω, f = 1.9 GHz, single-ended, ac-coupled source drive to RFIP through 1.2 nH series inductor, RFIM
ac-coupled through 1.2 nH series inductor to common, differential-to-single-ended conversion at output using 1:1 balun.
Table 1.
Parameter Conditions Min Typ Max Unit
OVERALL FUNCTION
Frequency Range 1.5 2.4 GHz
Maximum Gain Maximum gain setpoint for all phase setpoints −4.5 dB
Minimum Gain
= V
V
BBI
= 0 V differential
BBQ
(at recommended common-mode level)
Gain Control Range Relative to maximum gain 30 dB
Phase Control Range Over 30 dB control range 360 Degrees
Gain Flatness Over any 60 MHz bandwidth 0.5 dB
Group Delay Flatness Over any 60 MHz bandwidth 50 ps
RF INPUT STAGE RFIM, RFIP (Pins 21 and 22)
Input Return Loss From RFIP to CMRF (with 1.2 nH series inductors) 12 dB
CARTESIAN CONTROL INTERFACE (I AND Q) IBBP, IBBM, QBBP, QBBM (Pins 16, 15, 3, 4)
Gain Maximum gain setpoint −4.5 dB
Output Noise Floor Maximum gain setpoint, no input −150.5 dBm/Hz
P
= 0 dBm, frequency offset = 20 MHz −149 dBm/Hz
IN
Output IP3 f1 = 1900 MHz, f2 = 1897.5 MHz, maximum gain setpoint 17.5 dBm
Adjacent Channel Power
CDMA2000, single carrier, P
= -4 dBm,
OUT
maximum gain, phase setpoint = 45° (See Figure 35)
Output 1 dB Compression Point Maximum gain 8.5 dBm
POWER SUPPLY
VPS2 (Pins 5, 6, and 14), VPRF (Pins 19 and 24),
RFOP, RFOM (Pins 9 and 10)
Positive Supply Voltage 4.75 5 5.25 V
Total Supply Current Includes load current 105 125 145 mA
OUTPUT DISABLE DSOP (Pin 13)
Disable Threshold (See Figure 24) Vs/2 V
Attenuation DSOP = 5 V 33 dB
Enable Response Time
Delay following high-to-low transition until
RF output amplitude is within 10% of final value.
Disable Response Time
Delay following low-to-high transition until
device produces full attenuation
−34.5 dB
45 ns
45 ns
−76 dBm
30 ns
15 ns
Rev. 0 | Page 3 of 20
AD8341
ABSOLUTE MAXIMUM RATINGS
Table 2.
Parameters Rating
Supply Voltage VPRF, VPS2 5.5 V
DSOP 5.5 V
IBBP, IBBM, QBBP, QBBM 2.5 V
RFOP, RFOM 5.5V
RF Input Power at Maximum Gain 13 dBm, re: 50 Ω
(RFIP or RFIM, Single-Ended Drive)
Equivalent Voltage 2.8 V p-p
Internal Power Dissipation 825 mW
θJA (With Pad Soldered to Board) 59 °C/W
Maximum Junction Temperature 125°C
Operating Temperature Range −40°C to +85°C
Storage Temperature Range −65°C to +150°C
Lead Temperature Range (Soldering 60 sec) 300°C
ESD 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 this product 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.
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.
Rev. 0 | Page 4 of 20
AD8341
PIN CONFIGURATION AND FUNCTION DESCRIPTIONS
VPRF
CMRF
RFIP
RFIM
CMRF
2322212019
PIN 1
INDICATOR
11
10
RFOP
RFOM
CMOP
CMOP
VPRF
12
CMOP
18
17
16
15
14
13
IFLP
IFLM
IBBP
IBBM
VPS2
DSOP
04700-002
24
QFLP
1
QFLM
2
QBBP
3
4
5
6
AD8341
TOP VIEW
(Not to Scale)
789
CMOP
QBBM
VPS2
VPS2
Figure 2. 24-Lead Lead Frame Chip Scale Package (LFCSP)
The AD8341 is a linear RF vector modulator with Cartesian
baseband controls. In the simplified block diagram given in
Figure 26, the RF signal propagates from the left to the right
while baseband controls are placed above and below. The RF
input is first split into in-phase (I) and quadrature (Q) components. The variable attenuators independently scale the I and Q
components of the RF input. The attenuator outputs are then
summed and buffered to the output.
By controlling the relative amounts of I and Q components that
are summed, continuous magnitude and phase control of the
gain is possible. Consider the vector gain representation of the
AD8341 expressed in polar form in Figure 27. The attenuation
factors for the I and Q signal components are represented on
the x- and y-axis, respectively, by the baseband inputs, V
. The resultant of their vector sum represents the vector
V
BBQ
gain, which can also be expressed as a magnitude and phase. By
applying different combinations of baseband inputs, any vector
gain within the unit circle can be programmed.
or V
A change in sign of V
BBI
can be viewed as a change in
BBQ
sign of the gain or as a 180° phase change. The outermost
circle represents the maximum gain magnitude of unity. The
circle origin implies, in theory, a gain of 0. In practice, circuit
mismatches and unavoidable signal feedthrough limit the
minimum gain to approximately −34.5 dB. The phase angle
between the resultant gain vector and the positive x-axis is defined as the phase shift. Note that there is a nominal, systematic
insertion phase through the AD8341 to which the phase shift is
added. In the following discussions, the systematic insertion
phase is normalized to 0°.
The correspondence between the desired gain and phase setpoints, Gain
V
, is given by simple trigonometric identities
BBQ
and PhaseSP, and the Cartesian inputs, V
SP
[]
()
SP
BBI
2
()
VVPhase/arctan=
BBI
BBQSP
//
VVVVGain+=
OBBQO
2
where:
V
is the baseband scaling constant (500 mV).
O
V
and V
BBI
are the differential I and Q baseband voltages,
BBQ
respectively.
Note that when evaluating the arctangent function, the proper
phase quadrant must be selected. For example, if the principal
value of the arctangent (known as the Arctangent(x)) is used,
quadrants 2 and 3 could be interpreted mistakenly as quadrants
4 and 1, respectively. In general, both V
and V
BBI
BBQ
in concert to modulate the gain and the phase.
and
BBI
and
BBI
are needed
Pure amplitude modulation is represented by radial movement
of the gain vector tip at a fixed angle, while pure phase modulation is represented by rotation of the tip around the circle at a
fixed radius. Unlike traditional I-Q modulators, the AD8341 is
designed to have a linear RF signal path from input to output.
Traditional I-Q modulators provide a limited LO carrier path
through which any amplitude information is removed.
BBI
LINEAR
ATTENUATOR
LINEAR
ATTENUATOR
VBBQ
V
q
+0.5
|A|
–0.5
θ
A
I-V
OUTPUT
DISABLE
SINGLE-ENDED OR
DIFFERENTIAL
50Ω OUTPUT
V
i
+0.5–0.5
04700-027
SINGLE-ENDED OR
DIFFERENTIAL
50Ω INPUT Z
Figure 26. Simplified Architecture of the AD8341
I CHANNEL INPUT
V-I
0°/90°
V-I
Q CHANNEL INPUT
MAX GAIN
MIN GAIN
Figure 27. Vector Gain Representation
RF QUADRATURE GENERATOR
The RF input is directly coupled differentially or single-ended
to the quadrature generator, which consists of a multistage RC
polyphase network tuned over the operating frequency range of
1.5 GHz to 2.4 GHz. The recycling nature of the polyphase network generates two replicas of the input signal, which are in
precise quadrature, i.e., 90°, to each other. Since the passive
network is perfectly linear, the amplitude and phase information
contained in the RF input is transmitted faithfully to both channels. The quadrature outputs are then separately buffered to
drive the respective attenuators. The characteristic impedance
of the polyphase network is used to set the input impedance of
the AD8341.
04700-026
Rev. 0 | Page 10 of 20
AD8341
I-Q ATT ENUATOR S AN D BA SEBAND AMPLIFIERS
The proprietary linear-responding attenuator structure is an
active solution with differential inputs and outputs that offer
excellent linearity, low noise, and greater immunity from mismatches than other variable attenuator methods. The gain, in
linear terms, of the I and Q channels is proportional to its control
voltage with a scaling factor designed to be 2/V, i.e., a full-scale
gain setpoint of 1.0 (−4.5 dB) for a V
(or a V
BBI
) of 500 mV. The
BBQ
control voltages can be driven differentially or single-ended. The
combination of the baseband amplifiers and attenuators allows
for maximum modulation bandwidths in excess of 200 MHz.
OUTPUT AMPLIFIER
The output amplifier accepts the sum of the attenuator outputs
and delivers a differential output signal into the external load.
The output pins must be pulled up to an external supply,
preferably through RF chokes. When the 50 Ω load is taken
differentially, an output P1dB and IP3 of 8.5 dBm and 17.5 dBm
is achieved, respectively, at 1.9 GHz. The output can be taken in
single-ended fashion, albeit at lower performance levels.
NOISE AND DISTORTION
The output noise floor and distortion levels vary with the gain
magnitude but do not vary significantly with the phase. At the
higher gain magnitude setpoints, the OIP3 and the noise floor
vary in direct proportion with the gain. At lower gain magnitude setpoints, the noise floor levels off while the OIP3
continues to vary with the gain.
GAIN AND PHASE ACCURACY
There are numerous ways to express the accuracy of the
AD8341. Ideally, the gain and phase should precisely follow the
setpoints. Figure 4 illustrates the gain error in dB from a best fit
line, normalized to the gain measured at the gain setpoint = 1.0,
for the different phase setpoints. Figure 6 shows the gain error
in a different form, normalized to the gain measured at phase
setpoint = 0°; the phase setpoint is swept from 0° to 360° for
different gain setpoints. Figure 8 and Figure 22 show analogous
errors for the phase error as a function of gain and phase
setpoints. The accuracy clearly depends on the region of operation within the vector gain unit circle. Operation very close to
the origin generally results in larger errors as the relative
accuracy of the I and Q vectors degrades.
RF FREQUENCY RANGE
The frequency range on the RF input is limited by the internal
polyphase quadrature phase-splitter. The phase-splitter splits
the incoming RF input into two signals, 90° out of phase, as
previously described in the RF Quadrature Generator section.
This polyphase network has been designed to ensure robust
quadrature accuracy over standard fabrication process
parameter variations for the 1.5 GHz to 2.4 GHz specified RF
frequency range. Using the AD8341 as a single-sideband modulator and measuring the resulting sideband suppression is a
good gauge of how well the quadrature accuracy is maintained
over RF frequency. A typical plot of sideband suppression from
1.1 GHz to 2.7 GHz is shown in Figure 28. The level of sideband
suppression degradation outside the 1.5 GHz to 2.4 GHz specified range will be subject to manufacturing process variations.
–15
–20
–25
–30
–35
SIDEBAND SUPPRESSION (dBc)
–40
–45
0.72.72.52.32.11.91.71.51.30.9 1.1
Figure 28. Sideband Suppression vs. Frequency
FREQUENCY (GHz)
04700-028
Rev. 0 | Page 11 of 20
AD8341
APPLICATIONS
USING THE AD8341
The AD8341 is designed to operate in a 50 Ω impedance
system. Figure 30 illustrates an example where the RF input is
driven in a single-ended fashion while the differential RF output is converted to a single-ended output with an RF balun. The
baseband controls for the I and Q channels are typically driven
from differential DAC outputs. The power supplies, VPRF and
VPS2, should be bypassed appropriately with 0.1 µF and 100 pF
capacitors. Low inductance grounding of the CMOP and CMRF
common pins is essential to prevent unintentional peaking of
the gain.
RF INPUT AND MATCHING
The input impedance of the AD8341 is defined by the characteristics of the polyphase network. The capacitive component of
the network causes its impedance to roll-off with frequency
albeit at a rate slower than 6 dB/octave. By using matching
inductors on the order of 1.2 nH in series with each of the RF
inputs, RFIP and RFIM, a 50 Ω match is achieved with a return
loss of >10 dB over the operating frequency range. Different
matching inductors can improve matching over a narrower
frequency range. The single-ended and differential input
impedances are exactly the same.
1.2nH
RF
50Ω
100pF
100pF
1.2nH
RFIM
~1VDC
RFIP
PHASE
Figure 29. RF Input Interface to the AD8341 Showing
Coupling Capacitors and Matching Inductors
The RFIP and RFIM should be ac-coupled through low loss
series capacitors as shown in Figure 29. The internal dc levels
are at approximately 1 V. For single-ended operation, one input
is driven by the RF signal while the other input is ac grounded.
VP
RC
04700-029
IBBM
IBBP
INPUT
QBBP
QBBM
VP
RF
VP
0.1µF
C3
C6
100pF
C5
100pF
C8
0.1µF
100pF
L3
1.2nH
L4
1.2nH
C4
C7
100pF
C12
(SEE TEXT)
IFLP
VPRF
CMRF
RFIM
RFIP
CMRF
VPRF
P
L
Q
F
C11
(SEE TEXT)
100pF
IFLM
AD8341
QFLM
C2
IBBP
QBBP
IBBM
QBBM
DSOP
VPS2
VPS2
VPS2
C9
100pF
C1
0.1µF
CMOP
CMOP
RFOM
RFOP
CMOP
CMOP
C10
0.1µF
120nH
VP
L1
A
OUTPUT
DISABLE
B
L2
120nH
C14
0.1µF
C17
100pF
C18
100pF
VP
ETC1-1-13
RF
OUTPUT
04700-030
Figure 30. Basic Connections
Rev. 0 | Page 12 of 20
AD8341
–
RF OUTPUT AND MATCHING
The RF outputs of the AD8341, RFOP, and RFOM, are open
collectors of a transimpedance amplifier which need to be
pulled up to the positive supply, preferably with RF chokes as
shown in Figure 31. The nominal output impedance looking
into each individual output pin is 25 Ω. Consequently, the
differential output impedance is 50 Ω.
V
P
RFOP
120nH
100pF
100pF
50Ω
DIFFERENTIAL
1:1
RF
OUTPUT
04700-031
R
T
RFOM
±I
SIG
G
M
R
T
Figure 31. RF Output Interface to the AD8341 Showing
Coupling Capacitors, Pull-Up RF Chokes, and Balun
Since the output dc levels are at the positive supply, ac coupling
capacitors will usually be needed between the AD8341 outputs
and the next stage in the system.
A 1:1 RF broadband output balun, such as the ETC1-1-13
(M/A-COM), converts the differential output of the AD8341
into a single-ended signal. Note that the loss and balance of the
balun directly impact the apparent output power, noise floor,
and gain/phase errors of the AD8341. In critical applications,
narrow-band baluns with low loss and superior balance are
recommended.
If the output is taken in a single-ended fashion directly into a
50 Ω load through a coupling capacitor, there will be an impedance mismatch. This can be resolved with a 1:2 balun to convert
the single-ended 25 Ω output impedance to 50 Ω. If loss of
signal swing is not critical, a 25 Ω back termination in series
with the output pin can also be used. The unused output pin
must still be pulled up to the positive supply. The user may load
it through a coupling capacitor with a dummy load to preserve
balance. The gain of the AD8341 when the output is singleended varies slightly with dummy load value as shown in Figure 32.
2.5
–3.0
–3.5
–4.0
–4.5
–5.0
–5.5
–6.0
GAIN (dB)
–6.5
–7.0
–7.5
–8.0
–8.5
Figure 32. Gain of the AD8341 Using a Single-Ended Output with Different
Dummy Loads, R
RL2 = SHORT
RL2 = 50
Ω
RL2 = OPEN
FREQUENCY (GHz)
, on the Unused Output
L2
RL = 50
Ω
04700-032
3.01.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8
The RF output signal can be disabled by raising the DSOP pin
to the positive supply. The output disable function provides
>30 dB attenuation of the input signal even at full gain. The
interface to DSOP is high impedance and the shutdown and
turn-on response times are <100 ns. If the disable function is
not needed, the DSOP pin should be tied to ground.
DRIVING THE I-Q BASEBAND CONTROLS
The I and Q inputs to the AD8341 set the gain and phase between input and output. These inputs are differential and should
normally have a common-mode level of 0.5 V. However, when
differentially driven, the common mode can vary from 250 mV
to 750 mV while still allowing full gain control. Each input pair
has a nominal input swing of ±0.5 V differential around the
common-mode level. The maximum gain of unity is achieved if
the differential voltage is equal to +500 mV or −500 mV. So
with a common-mode level of 500 mV, IBBP and IBBM will
each swing between 250 mV and 750 mV.
The I and Q inputs can also be driven with a single-ended
signal. In this case, one side of each input should be tied to a
low noise 0.5 V voltage source (a 0.1 µF decoupling capacitor
located close to the pin is recommended), while the other input
swings from 0 V to 1 V. Differential drive generally offers superior
even-order distortion and lower noise than single-ended drive.
The bandwidth of the baseband controls exceeds 200 MHz even
at full-scale baseband drive. This allows for very fast gain and
phase modulation of the RF input signal. In cases where lower
modulation bandwidths are acceptable or desired, external filter
capacitors can be connected across Pins IFLP to IFLM and
QFLP to QFLM to reduce the ingress of baseband noise and
spurious signal into the control path.
Rev. 0 | Page 13 of 20
AD8341
The 3 dB bandwidth is set by choosing C
according to the
FLT
following equation:
nF10kHz45
f
≈
3dB
This equation has been verified for values of C
×
pF0.5
+
C
FLT
from 10 pF to
FLT
0.1 µF (bandwidth settings of approximately 4.5 kHz to 43 MHz).
INTERFACING TO HIGH SPEED DACs
The AD977x family of dual DACs is well suited to driving the I
and Q vector controls of the AD8341. While these inputs can in
general be driven by any DAC, the differential outputs and bias
level of the ADI TxDAC® family allows for a direct connection
between DAC and modulator.
The AD977x family of dual DACs has differential current outputs. The full-scale current is user programmable and is usually
set to 20 mA, that is, each output swings from 0 mA to 20 mA.
The basic interface between the AD9777 DAC outputs and the
AD8341 I and Q inputs is shown in Figure 33. The Resistors R1
and R2 set the dc bias level according to the equation:
Bias Level = Average Output Current × R1
For example, if the full-scale current from each output is 20 mA,
each output will have an average current of 10 mA. Therefore to
set the bias level to the recommended 0.5 V, R1 and R2 should
be set to 50 Ω each. R1 and R2 should always be equal.
If R3 is omitted, this will result in an available swing from
the DAC of 2 V p-p differential, which is twice the maximum
voltage range required by the AD8341. DAC resolution can be
maximized by adding R3, which scales down this voltage
according to the following equation:
Figure 34 shows the relationship between the value of R3 and
the peak baseband voltage with R1 and R2 equal to 50 Ω.
From Figure 34, it can be seen that a value of 100 Ω for R3 will
provide a peak-to-peak swing of 1 V p-p differential into the
AD8341’s I and Q inputs.
When using a DAC, low-pass image reject filters are typically
used to eliminate the Nyquist images produced by the DAC.
They also provide the added benefit of eliminating broadband
noise that might feed into the modulator from the DAC.
CDMA2000 APPLICATION
To test the compliance to the CDMA2000 base station standard,
a single-carrier CDMA2000 test model signal (forward pilot,
sync, paging, and six traffic as per 3GPP2 C.S0010-B, Table
6.5.2.1) was applied to the AD8341 at 1960 MHz. A cavity tuned
filter was used to reduce noise from the signal source being
applied to the device. The 6.8 MHz pass band of this filter is
apparent in the subsequent spectral plots.
Figure 35 shows a plot of the spectrum of the output signal
under nominal conditions. P
= 0.353 V, i.e., V
V
BBQ
IBBP
is equal to −4 dBm and V
OUT
− V
= V
QBBP
− V
IBBM
= 0.353 V.
QBBM
Noise and distortion is measured in a 1 MHz bandwidth at
±2.25 MHz carrier offset (30 kHz measurement bandwidth).
Measured at ±2.25 MHz Carrier Offset in 1 MHz BW Input Signal Filtered
Using a Cavity Tuned Filter (Pass Band = 6.8 MHz)
Holding the differential I and Q control voltages steady at
0.353 V, input power was swept. Figure 36 shows variation in
spurious content, again measured at ±2.25 MHz carrier offset in
a 1 MHz bandwidth, as defined by the 3GPP2 specification.
–70
–72
–74
–76
–78
–80
–82
–84
–86
–88
ACP @ 2.25MHz OFFSET (dBm, 1MHz, BW)
–90
–20 –18 –16 –14 –12 –10 –8–6–4–20
OUTPUT POWER (dBm)
04700-036
Figure 36. Adjacent Channel Power vs. Output Power,
CDMA2000 Single Carrier @ 1960 MHz; ACP Measured at
= V
±2.25 MHz Carrier Offset (1 MHz BW ); V
= 0.353 V
BBI
BBQ
With a fixed input power of 2.4 dBm, the output power was
again swept by exercising the I and Q inputs. V
and V
BBI
BBQ
were
kept equal and were swept from 100 mV to 500 mV. The resulting output power and ACP are shown in Figure 37.
0
–5
–10
–15
–20
OUTPUT POWER (dBm)
–25
–30
00.10.20.30.40.5
IQ CONTROL VOLTAGE
–60
–65
–70
–75
–80
–85
ACP dBm (1MHz BW) @ 2.25MHz OFFSET
–90
Figure 37. Output Power and ACP vs. I and Q Control Voltages,
= V
CDMA2000 Test Model, V
BBI
BBQ
, ACP Measured at
±2.25 MHz Carrier Offset in 1 MHz BW
Figure 37 shows that for a fixed input power, the ACP (measured in
dBm) tracks the output power as the gain is changed.
WCDMA APPLICATION
Figure 38 shows a plot of the output spectrum of the AD8341
transmitting a single-carrier WCDMA signal (Test Model 1-64
at 2140 MHz). The carrier power is approximately −9 dBm. The
differential I and Q control voltages are both equal to 0.353 V,
that is, the vector is sitting on the unit circle at 45°. At this
power level, an adjacent channel power ratio of −61 dBc is
achieved. The alternate channel power ratio of −72 dBc is
dominated by the noise floor of the AD8341.
MARKER 1 [T1 ]
–28.39dBm
2.14050000GHz
C11
C0
C11
–24
–30
–40
–50
–60
–70
–80
–90
100
–110
120
124
REF LVL
–24dBm
OFFSET 1dB
C12C12
Figure 38. AD8341 Single-Carrier WCDMA Spectrum at 2140 MHz
Figure 39 shows how ACPR and noise vary with varying input
power (differential I and Q control voltages are held at 0.353 V).
At high power levels, both adjacent and alternate channel power
ratios increase sharply. As output power drops, adjacent and
alternate channel power ratios both reach minimums before the
measurement becomes dominated by the noise floor of the
AD8341. At this point, adjacent and alternate channel power
ratios become approximately equal.
Rev. 0 | Page 15 of 20
AD8341
)
As the output power drops, the noise floor, measured in dBm in
1 MHz BW at 50 MHz carrier offset, drops slightly.
–30
–35
–40
–45
–50
–55
–60
–65
–70
–75
–80
–30–25–20–15–10–505
ADJACENT/ALTERNATE CHANNEL POWER RATIO (dBc
NOISE –50MHz OFFSET
ACPR 5MHz OFFSET
ACPR 10MHz OFFSET
OUTPUT POWER (dBm)
Figure 39. AD8341 ACPR and Noise vs. Output Power;
Single-Carrier WCDMA ( Test Model 1-64 at 2140 MHz)
–50
–55
–60
–65
–70
–75
–80
–85
–90
–95
–100
NOISE dBm @ 50MHz CARRIER OFFSET (1MHz BW)
Figure 40 shows how output power, ACPR, and noise vary with
the differential I and Q control voltages. V
and V
BBI
are tied
BBQ
together and are varied from 0.5 V to 50 mV.
04700-039
0
–5
–10
–15
–20
–25
–30
–35
OUTPUT POWER (dBm)
–40
–45
–50
00.10.20.30.40.5
OUTPUT POWER dBm
ACPR 5MHz OFFSET
ACPR 10MHz OFFSET
NOISE –50MHz OFFSET
IQ CONTROL VOLTAGE
Figure 40. AD8341 Output Power, ACPR and Noise vs. V
–40
–45
–50
–55
–60
–65
–70
ACPR (dBc)
–75
–80
–85
–90
.
IQ
Single-Carrier WCDMA ( Test Model 1-64 at 2140 MHz)
In this case, adjacent channel power ratio remains constant as
the (noise dominated) alternate channel power degrades
roughly 1-for-1 with output power. As the I and Q control voltage drops, the noise floor again drops slowly.
NOISE dBm @ 50MHz OFFSET (1MHz BW)
04700-040
Rev. 0 | Page 16 of 20
AD8341
EVALUATION BOARD
The evaluation board circuit schematic for the AD8341 is
shown in Figure 41.
The evaluation board is configured to be driven from a
single-ended 50 Ω source. Although the input of the AD8341 is
differential, it may be driven single-ended, with no loss of performance.
The low-pass corner frequency of the baseband I and Q channels can be reduced by installing capacitors in the C11 and C12
positions. The low-pass corner frequency for either channel is
approximated by
nF10kHz45
f
3dB
≈
×
+
C
FLT
pF0.5
On this evaluation board, the I and Q baseband circuits are
identical to each other, so the following description applies
equally to each. The connections and circuit configuration for
the Q baseband inputs are described in Table 4.
The baseband input of the AD8341 requires a differential voltage drive. The evaluation board is set up to allow such a drive by
connecting the differential voltage source to QBBP and QBBM.
The common-mode voltage should be maintained at approximately 0.5 V. For this configuration, Jumpers W1 through W4
should be removed.
The baseband input of the evaluation board may also be driven
with a single-ended voltage. In this case, a bias level is provided
to the unused input from Potentiometer R10 by installing either
W1 or W2.
Setting SW1 in Position B disables the AD8341 output amplifier.
With SW1 set to Position A, the output amplifier is enabled.
With SW1 set to Position A, an external voltage signal, such as a
pulse, can be applied to the DSOP SMA connector to exercise
the output amplifier enable/disable function.
I Channel Baseband Interface. Resistors R7 and R9 may be installed to accommodate a
baseband source that requires a specific terminating impedance. Capacitors C15 and C19
are bypass capacitors.
For single-ended baseband drive, the Potentiometer R11 can be used to provide a bias level
to the unused input (install either W3 or W4).
Q Channel Baseband Interface. See the I Channel Baseband Interface section.
Baseband Low-Pass Filtering. By adding Capacitor C11 between QFLP and QFLM, and C12
between IFLP and IFLM, the 3 dB low-pass corner frequency of the baseband interface can
be reduced from 230 MHz (nominal). See equation in text.
Output Interface. The 1:1 balun transformer, T1, converts the 50 Ω differential output to 50
Ω single-ended. C17 and C18 are dc blocks. L1 and L2 provide dc bias for the output.
Input Interface. The input impedance of the AD8341 requires 1.2 nH inductors in series
with RFIP and RFIM for optimum return loss when driven by a single-ended 50 Ω line. C5
and C6 are dc blocks.
Output Disable Interface. The output stage of the AD8341 is disabled by applying a high
voltage to the DSOP pin by moving SW1 to Position B. The output stage is enabled moving
SW1 to Position A. The output disable function can also be exercised by applying an external high or low voltage to the DSOP SMA connector with SW1 in Position A.