PDF Data Sheet Rev. H

Low Cost, DC to 150 MHz,
Variable Gain Amplifier
AD8330
Data Sheet
FUNCTIONAL BLOCK DIAGRAM
Fully differential signal path, also used with single-sided signals
Inputs from 0.3 mV to 1 V rms, rail-to-rail outputs
Differential RIN = 1 kΩ; ROUT (each output) 75 Ω
Automatic offset compensation (optional)
Linear-in-dB and linear-in-magnitude gain modes
0 dB to 50 dB, for 0 V < VDBS < 1.5 V (30 mV/dB)
Inverted gain mode: 50 dB to 0 dB at −30 mV/dB
×0.03 to ×10 nominal gain for 15 mV < VMAG < 5 V
Constant bandwidth: 150 MHz at all gains
Low noise: 5 nV/√Hz typical at maximum gain
Low distortion: ≤−62 dBc typical
Low power: 20 mA typical at VS of 2.7 V to 6 V
Available in a space-saving, 3 mm × 3 mm LFCSP package
ENBL
OFST CNTR
BIAS AND VREF
CM AND
OFFSET
CONTROL
VGA CORE
OUTPUT
STAGES
INHI
OPLO
INLO
MODE
OPHI
GAIN INTERFACE
OUTPUT
CONTROL
VDBS CMGN COMM
VMAG
CMOP
03217-101
FEATURES
Figure 1.
APPLICATIONS
Pre-ADC signal conditioning
75 Ω cable driving adjust
AGC amplifiers
GENERAL DESCRIPTION
The AD8330 is a wideband variable gain amplifier for applications
requiring a fully differential signal path, low noise, well-defined
gain, and moderately low distortion, from dc to 150 MHz. The
input pins can also be driven from a single-ended source. The
peak differential input is ±2 V, allowing sine wave operation at
1 V rms with generous headroom. The output pins can drive
single-sided loads essentially rail-to-rail. The differential output
resistance is 150 Ω. The output swing is a linear function of the
voltage applied to the VMAG pin that internally defaults to 0.5 V,
providing a peak output of ±2 V. This can be raised to 10 V p-p,
limited by the supply voltage.
The basic gain function is linear-in-dB, controlled by the voltage
applied to Pin VDBS. The gain ranges from 0 dB to 50 dB for
control voltages between 0 V and 1.5 V—a slope of 30 mV/dB.
The gain linearity is typically within ±0.1 dB. By changing the
logic level on Pin MODE, the gain decreases over the same range,
with an opposite slope. A second gain control port is provided
at the VMAG pin and allows the user to vary the numeric gain
from a factor of 0.03 to 10. All the parameters of the AD8330
have low sensitivities to temperature and supply voltages. Using
VMAG, the basic 0 dB to 50 dB range can be repositioned to
any value from 20 dB higher (that is, 20 dB to 70 dB) to at least
Rev. H
30 dB lower (that is, –30 dB to +20 dB) to suit the application,
thereby providing an unprecedented gain range of over 100 dB.
A unique aspect of the AD8330 is that its bandwidth and pulse
response are essentially constant for all gains, over both the
basic 50 dB linear-in-dB range, but also when using the linearin-magnitude function. The exceptional stability of the HF
response over the gain range is of particular value in those VGA
applications where it is essential to maintain accurate gain lawconformance at high frequencies.
An external capacitor at Pin OFST sets the high-pass corner of
an offset reduction loop, whose frequency can be as low as 5 Hz.
When this pin is grounded, the signal path becomes dc-coupled.
When used to drive an ADC, an external common-mode control
voltage at Pin CNTR can be driven to within 0.5 V of either ground
or VS to accommodate a wide variety of requirements. By default,
the two outputs are positioned at the midpoint of the supply, VS/2.
Other features, such as two levels of power-down (fully off and
a hibernate mode), further extend the practical value of this
exceptionally versatile VGA.
The AD8330 is available in 16-lead LFCSP and 16-lead QSOP
packages and is specified for operation from −40°C to +85°C.
Document Feedback
Information furnished by Analog Devices is believed to be accurate and reliable. However, no
responsibility is assumed by Analog Devices for its use, nor for any infringements of patents or other
rights of third parties that may result from its use. Specifications subject to change without notice. No
license is granted by implication or otherwise under any patent or patent rights of Analog Devices.
Trademarks and registered trademarks are the property of their respective owners.
One Technology Way, P.O. Box 9106, Norwood, MA 02062-9106, U.S.A.
Tel: 781.329.4700 ©2002–2016 Analog Devices, Inc. All rights reserved.
Technical Support
www.analog.com
AD8330
Data Sheet
TABLE OF CONTENTS
Features .............................................................................................. 1 Applications Information .............................................................. 26 Applications ....................................................................................... 1 ADC Driving............................................................................... 26 Functional Block Diagram .............................................................. 1 Simple AGC Amplifier .............................................................. 26 General Description ......................................................................... 1 Wide Range True RMS Voltmeter............................................ 27 Revision History ............................................................................... 2 Evaluation Board ............................................................................ 29 Specifications..................................................................................... 3 General Description ................................................................... 29 Absolute Maximum Ratings............................................................ 5 Basic Operation .......................................................................... 29 ESD Caution .................................................................................. 5 Options ........................................................................................ 30 Pin Configurations and Function Descriptions ........................... 6 Measurement Setup.................................................................... 30 Typical Performance Characteristics ............................................. 8 AD8330-EVALZ Board Design................................................. 30 Theory of Operation ...................................................................... 15 Outline Dimensions ....................................................................... 32 Circuit Description..................................................................... 15 Ordering Guide .......................................................................... 32 Using the AD8330 ...................................................................... 21 REVISION HISTORY
5/2016—Rev. G to Rev. H
Changes to Figure 2 and Table 3 ..................................................... 6
Moved Figure 3 ................................................................................. 7
Changes to Table 4 ............................................................................ 7
Change to Figure 45 ....................................................................... 15
Changes to Simple AGC Amplifier Section ................................ 27
Updated Outline Dimensions ....................................................... 32
Changes to Ordering Guide .......................................................... 32
5/2014—Rev. F to Rev. G
Changes to Table 1 ............................................................................ 3
11/2012—Rev. E to Rev. F
Changes to Figure 1 .......................................................................... 1
Changes to Output (Input) Common-Mode Control ............... 20
Updated Outline Dimensions ....................................................... 31
Changes to Ordering Guide .......................................................... 31
3/2010—Rev. D to Rev. E
Changes to Figure 2 and Table 3 ..................................................... 6
Changes to Figure 69 ...................................................................... 28
Changes to Figure 71 ...................................................................... 29
Changes to Figure 72 ...................................................................... 30
Deleted Table 7; Renumbered Sequentially ................................ 31
Changes to Ordering Guide .......................................................... 32
1/2008—Rev. C to Rev. D
Changes to Figure 28 and Figure 29............................................. 12
Added Evaluation Board Section ................................................. 28
Changes to Ordering Guide .......................................................... 33
6/2006—Rev. B to Rev. C
Updated Format .................................................................. Universal
Changes to Figure 1 ...........................................................................1
Deleted Figure 2; Renumbered Sequentially .................................1
Changes to Specifications Section ...................................................3
Change to Absolute Maximum Ratings .........................................5
Changes to Typical Performance Characteristics Summary
Statement ............................................................................................7
Changes to Figure 14 and Figure 15................................................8
Changes to Figure 31 and Figure 32............................................. 11
Updated Outline Dimensions ....................................................... 28
10/2004—Rev. A to Rev. B
Changes to Absolute Maximum Ratings Section and Ordering
Guide Section .....................................................................................4
Change to TPC 14 .............................................................................8
Note Added to CP-16 Package ...................................................... 26
4/2003—Rev. 0 to Rev. A
Updated Outline Dimensions ....................................................... 26
10/2002—Revision 0: Initial Version
Rev. H | Page 2 of 32
Data Sheet
AD8330
SPECIFICATIONS
VS = 5 V, TA = 25°C, CL = 12 pF on OPHI and OPLO, RL = ∞, VDBS = 0.75 V, VMODE = high, VMAG = Pin VMAG open circuit (0.5 V),
VOFST = 0 V, differential operation, unless otherwise noted.
Table 1.
Parameter
INPUT INTERFACE
Full-Scale Input
Input Resistance
Input Capacitance
Voltage Noise Spectral Density
Common-Mode Voltage Level
Input Offset
Drift
Permissible CM Range 1
Common-Mode AC Rejection
OUTPUT INTERFACE
Small Signal –3 dB Bandwidth
Peak Slew Rate
Peak-to-Peak Output Swing
Common-Mode Voltage
Offset Voltage
Offset Correction Enabled
Offset Correction Disabled
Voltage Noise Spectral Density
Differential Output Impedance
HD2 2
HD32
OUTPUT OFFSET CONTROL
AC-Coupled Offset
High-Pass Corner Frequency
COMMON-MODE CONTROL
Usable Voltage Range
Input Resistance
DECIBEL GAIN CONTROL
Normal Voltage Range
Elevated Range
Gain Scaling
Gain Linearity Error
Absolute Gain Error
Bias Current
Incremental Resistance
Gain Settling Time to 0.5 dB Error
Mode Up/Down
Mode Up Logic Level
Mode Down Logic Level
Test Conditions/Comments
Pin INHI, Pin INLO
VDBS = 0 V, differential drive
VDBS = 1.5 V
Pin-to-pin
Either pin to COMM
f = 1 MHz, VDBS = 1.5 V; inputs ac-shorted
Min
Typ
±1.4
±4.5
800
±2
±6.3
1k
4
5
3.0
1
2
2.75
Pin OFST connected to Pin COMM
VMAG ≥ 2 V (peaks are supply limited)
Pin CNTR O/C
Pin OFST connected to ground
f = 1 MHz, VDBS = 0 V
Pin-to-pin
VOUT = 1 V p-p, f = 10 MHz, RL = 1 kΩ
VOUT = 1 V p-p, f = 10 MHz, RL = 1 kΩ
Pin OFST
CHPF on Pin OFST (0 V < VDBS < 1.5 V)
CHPF = 3.3 nF, from OFST to CNTR (scales as 1/CHPF)
Pin CNTR
±1.8
±4
2.4
120
−60
−55
150
1500
±2
±4.5
2.5
MHz
V/µs
V
V
V
VDBS stepped from 0.05 V to 1.45 V or 1.45 V to 0.05 V
Pin MODE
Gain increases with VDBS, MODE = O/C
Gain decreases with VDBS
Rev. H | Page 3 of 32
1.2 k
3.25
VS
1
8
62
150
−62
−53
±2.2
2.6
±5
±40
180
10
100
0.5
From Pin CNTR to VS/2
VDBS, CMGN, and MODE pins
CMGN connected to COMM
CMGN O/C (VCMGN rises to 0.2 V)
Mode high or low
0.3 V ≤ VDBS ≤ 1.2 V
VDBS = 0 V
Flows out of Pin VDBS
27
−0.35
−2
Unit
V
mV
Ω
pF
nV/√Hz
V
mV rms
µV/°C
V
dB
dB
0
f = 1 MHz, 0.1 V rms
f = 50 MHz
Pin OPHI, Pin OPLO
0 V < VDBS < 1.5 V
VDBS = 0 V
Max
mV
mV
nV/√Hz
Ω
dBc
dBc
mV rms
kHz
4
4.5
V
kΩ
0 to 1.5
0.2 to 1.7
30
±0.1
±0.5
100
100
250
V
V
mV/dB
dB
dB
nA
MΩ
ns
33
+0.35
+2
1.5
0.5
V
V
AD8330
Parameter
LINEAR GAIN INTERFACE
Peak Output Scaling, Gain vs. VMAG
Gain Multiplication Factor vs. VMAG
Usable Input Range
Default Voltage
Incremental Resistance
Bandwidth
CHIP ENABLE
Logic Voltage for Full Shutdown
Logic Voltage for Hibernate Mode
Logic Voltage for Full Operation
Current in Full Shutdown
Current in Hibernate Mode
Minimum Time Delay 3
POWER SUPPLY
Supply Voltage
Quiescent Current
Data Sheet
Test Conditions/Comments
Pin VMAG, Pin CMGN
See the Circuit Description section
Gain is nominal when VMAG = 0.5 V
Min
Typ
Max
Unit
3.8
4.0
×2
4.2
V/V
5
0.52
V
V
kΩ
MHz
0.5
1.7
V
V
V
µA
mA
µs
0
0.48
VMAG O/C
For VMAG ≥ 0.1 V
Pin ENBL
Output pins remain at CNTR
1.3
2.3
0.5
4
150
1.5
20
1.5
1.7
100
VPSI, VPOS, VPSO, COMM, and CMOP pins
2.7
VDBS = 0.75 V
20
6
27
V
mA
The use of an input common-mode voltage significantly different from the internally set value is not recommended due to its effect on noise performance. See Figure 56.
See the Typical Performance Characteristics section for more detailed information on distortion in a variety of operating conditions.
3
For minimum sized coupling capacitors.
1
2
Rev. H | Page 4 of 32
Data Sheet
AD8330
ABSOLUTE MAXIMUM RATINGS
Table 2.
Parameter
Supply Voltage
Power Dissipation
16-Lead QSOP Package1
16-Lead LFCSP Package
Input Voltage at Any Pin
Storage Temperature Range
θJA
16-Lead QSOP Package
16-Lead LFCSP Package
θJC
16-Lead QSOP Package
Operating Temperature Range
Lead Temperature (Soldering 60 sec)
1
Rating
6V
0.62 W
1.67 W
VS + 200 mV
−65°C to +150°C
Stresses at or above those listed under Absolute Maximum
Ratings may cause permanent damage to the product. This is a
stress rating only; functional operation of the product at these
or any other conditions above those indicated in the operational
section of this specification is not implied. Operation beyond
the maximum operating conditions for extended periods may
affect product reliability.
ESD CAUTION
105.4°C/W
60°C/W
39°C/W
−40°C to +85°C
300°C
4-layer JEDEC Board (252P).
Rev. H | Page 5 of 32
AD8330
Data Sheet
13 CNTR
14 VPOS
16 ENBL
15 OFST
PIN CONFIGURATIONS AND FUNCTION DESCRIPTIONS
VPSI 1
INLO 3
12 VPSO
AD8330
TOP VIEW
(Not to Scale)
11 OPHI
10 OPLO
9
CMOP
VMAG 8
COMM 7
VDBS 5
CMGN 6
MODE 4
NOTES
1. THE EXPOSED PAD IS NOT CONNECTED INTERNALLY.
FOR INCREASED RELIABILITY OF THE SOLDER JOINTS
AND MAXIMUM THERMAL CAPABILITY, IT IS RECOMMENDED
THAT THE PAD BE SOLDERED TO THE GROUND PLANE.
03217-003
INHI 2
Figure 2. 16-Lead LFCSP Pin Configuration
Table 3. 16-Lead LFCSP Pin Function Descriptions
Pin No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Mnemonic
VPSI
INHI
INLO
MODE
VDBS
CMGN
COMM
VMAG
CMOP
OPLO
OPHI
VPSO
CNTR
VPOS
OFST
16
ENBL
EPAD
Description
Positive Supply for Input Stages.
Differential Signal Input, Positive Polarity.
Differential Signal Input, Negative Polarity.
Logic Input: Selects Gain Slope. High = gain up vs. VDBS.
Input for Linear-in-dB Gain Control Voltage, VDBS.
Common Baseline for Gain Control Interfaces.
Ground for Input and Gain Control Bias Circuitry.
Input for Gain/Amplitude Control, VMAG.
Ground for Output Stages.
Differential Signal Output, Negative Polarity.
Differential Signal Output, Positive Polarity.
Positive Supply for Output Stages.
Common-Mode Output Voltage Control.
Positive Supply for Inner Stages.
Internal Offset Compensation Feature. When the OFST pin is unconnected, this feature is enabled. When the
OFST pin is grounded, this feature is disabled. See the Offset Compensation section.
Power Enable, Active High.
Exposed Pad. The exposed pad is not connected internally. For increased reliability of the solder joints and
maximum thermal capability, it is recommended that the pad be soldered to the ground plane.
Rev. H | Page 6 of 32
Data Sheet
AD8330
O FS T 1
16
VPOS
E N BL 2
15
CN T R
14
VPSO
VPSI 3
AD8330
TOP VIEW 13 OPHI
(Not to Scale)
12 OPLO
I N LO 5
MODE 6
11
CMOP
VDBS 7
10
VMAG
CMGN 8
9
COMM
03217-004
INHI 4
Figure 3. 16-Lead QSOP Pin Configuration
Table 4. 16-Lead QSOP Pin Function Descriptions
Pin No.
1
Mnemonic
OFST
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
ENBL
VPSI
INHI
INLO
MODE
VDBS
CMGN
COMM
VMAG
CMOP
OPLO
OPHI
VPSO
CNTR
VPOS
Description
Internal Offset Compensation Feature. When the OFST pin is unconnected, this feature is enabled.
When the OFST pin is grounded, this feature is disabled. See the Offset Compensation section.
Power Enable, Active High.
Positive Supply for Input Stages.
Differential Signal Input, Positive Polarity.
Differential Signal Input, Negative Polarity.
Logic Input: Selects Gain Slope. High = gain up vs. VDBS.
Input for linear-in-dB Gain Control Voltage, VDBS.
Common Baseline for Gain Control Interfaces.
Ground for Input and Gain Control Bias Circuitry.
Input for Gain/Amplitude Control, VMAG.
Ground for Output Stages.
Differential Signal Output, Negative Polarity.
Differential Signal Output, Positive Polarity.
Positive Supply for Output Stages.
Common-Mode Output Voltage Control.
Positive Supply for Inner Stages.
Rev. H | Page 7 of 32
AD8330
Data Sheet
TYPICAL PERFORMANCE CHARACTERISTICS
VS = 5 V, TA = 25°C, CL = 12 pF, VDBS = 0.75 V, VMODE = high (or O/C) VMAG = O/C (0.5 V), RL = ∞, VOFST = 0, differential operation, unless
otherwise noted.
2.0
50
45
LO MODE
40
HI MODE
1.0
GAIN ERROR (dB)
35
30
25
20
0.5
0
–0.5
0
0.25
0.75
VDBS (V)
0.50
1.00
1.25
1.50
–2.0
03217-005
0
0
0.2
Figure 4. Gain vs. VDBS
20
9
0.6
0.8
1.0
VDBS (V)
1.2
1.4
1.6
2340 UNITS
MODE = LO
15
8
10
% OF UNITS
6
5
4
3
5
0
–30.6 –30.5 –30.4 –30.3 –30.2 –30.1 –30.0 –29.9 –29.8 –29.7 –29.6 –29.5 –29.4 –29.3 –29.2 –29.1 –29.0
20
2
10
1
5
0
2
1
3
4
5
VMAG (V)
MODE = HI
15
0
29.1 29.2 29.3 29.4 29.5 29.6 29.7 29.8 29.9 30.0 30.1 30.2 30.3 30.4 30.5 30.6
GAIN SCALING (mV/dB)
Figure 5. Linear Gain Multiplication Factor vs. VMAG
Figure 8. Gain Slope Histogram
1.0
60
0.8
50
1.2V
0.9V
30
0.4
0.6V
GAIN (dB)
20
T = –40°C
0
–0.2
0V
0
–20
T = +25°C
–0.6
0.3V
10
–10
T = +85°C
–0.4
–30
–0.8
–40
0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
VDBS (V)
Figure 6. Gain Linearity Error Normalized at 25°C vs. VDBS,
at Three Temperatures, f = 1 MHz
–50
100k
03217-007
–1.0
VDBS = 1.5V
40
0.6
0.2
03217-009
7
03217-006
GAIN MULTIPLICATION FACTOR
0.4
Figure 7. Gain Error vs. VDBS at Various Frequencies
10
GAIN ERROR (dB)
10MHz
100MHz
–1.5
5
1MHz
1MHz
–1.0
10
03217-008
15
0
50MHz
100MHz
10MHz, 50MHz
1M
10M
FREQUENCY (Hz)
100M
500M
03217-010
GAIN (dB)
NORMALIZED @ VDBS = 0.75V
1.5
Figure 9. Frequency Response in 10 dB Steps for Various Values of VDBS
Rev. H | Page 8 of 32
Data Sheet
50
AD8330
40
1048 UNITS
ENABLE MODE
1.52V
30
20
0.48V
20
0.15V
10
0.048V
0
0.015V
% OF UNITS
GAIN (dB)
25
VMAG = 4.8V
15
10
–10
–20
5
1.0
03217-014
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0
0.1
–0.1
–0.2
–0.3
Figure 13. Differential Input Offset Histogram
10
VDBS = 0.1V
0
OUTPUT BALANCE ERROR (dB)
8
GROUP DELAY (ns)
–0.4
DIFFERENTIAL OFFSET (mV)
Figure 10. Frequency Response for Various Values of VMAG,
VDBS = 0.75 V
10
–0.5
0
–0.6
500M
–0.7
100M
10M
FREQUENCY (Hz)
–0.8
1M
03217-011
–40
100k
–0.9
–30
6
4
2
–10
–20
–30
–40
–50
–60
–70
10M
FREQUENCY (Hz)
100M
300M
–90
100k
03217-012
1M
Figure 11. Group Delay vs. Frequency
1M
10M
FREQUENCY (Hz)
100M
03217-015
–80
0
100k
Figure 14. Output Balance Error vs. Frequency for a Representative Part
0
200
190
–1
OUTPUT IMPEDANCE (Ω)
T = –40°C
–3
–4
T = +25°C
–5
160
150
140
130
120
–6
T = +85°C
0
0.2
0.4
0.6
0.8
1.0
VDBS (V)
1.2
1.4
110
1.6
03217-013
–7
170
100
100k
Figure 12. Differential Output Offset vs. VDBS for Three Temperatures,
for a Representative Part
1M
10M
FREQUENCY (Hz)
100M
Figure 15. Output Impedance vs. Frequency
Rev. H | Page 9 of 32
300M
03217-016
OFFSET VOLTAGE (mV)
180
–2
AD8330
90
Data Sheet
6000
VDBS = 1.5V
80
VDBS = 0.75V
70
5000
NOISE (nV/√Hz)
60
CMRR (dB)
50
VDBS = 0V
40
VDBS = 1.5V
f = 1MHz
OFST: ENABLED
DISABLED
30
20
10
4000
3000
2000
1000
1M
10M
FREQUENCY (Hz)
0
03217-017
–10
50k 100k
100M
80
2.5
2.0
VMAG = 0.5V
f = 1MHz
70
T = +25°C
1200
1.5
Figure 19. Output Referred Noise vs. VMAG
T = +85°C
f = 1MHz
VMAG = 0.5V
1 .0
VMAG (V)
Figure 16. CMRR vs. Frequency
1500
0.5
0
03217-020
0
T = +85°C
NOISE (nV/√Hz)
NOISE (nV/√Hz)
60
900
T = –40°C
600
50
40
T = +25°C
30
T = –40°C
20
300
0
0.2
0.4
0.6
0.8
1.0
VDBS (V)
1.2
1.4
0
03217-018
0
1.6
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
VDBS (V)
Figure 17. Output Referred Noise vs. VDBS for Three Temperatures
Figure 20. Input Referred Noise vs. VDBS for Three Temperatures
700
180
f = 1MHz
f = 1MHz
160
600
140
NOISE (nV/√Hz)
500
400
300
200
VMAG = 0.125V
120
100
VMAG = 0.5V
80
60
40
0
20
0
0.5
1.0
1.5
2.0
VMAG (V)
2.5
0
VMAG = 2V
0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
VDBS (V)
Figure 18. Output Referred Noise vs. VMAG, VDBS = 0.75 V
Figure 21. Input Referred Noise vs. VDBS for Three Values of VMAG
Rev. H | Page 10 of 32
03217-022
100
03217-019
NOISE (nV/√Hz)
0
03217-021
10
Data Sheet
7
AD8330
0
VDBS = 1.5V
–20
DISTORTION (dBc)
5
4
3
2
–30
–40
HD3, RL = 1kΩ
–50
–60
HD2, RL = 1kΩ
–70
1M
100M
10M
FREQUENCY (Hz)
–80
03217-023
0
100k
0
Figure 22. Input Referred Noise vs. Frequency
f = 10MHz
–10
–20
–20
DISTORTION (dBc)
–30
–40
HD3
–50
HD2
–60
HD2 AND HD3, RL = 150Ω1
–30
–40
HD3, RL = 1kΩ
–50
–60
–70
HD2, RL = 1kΩ
–70
1M
10M
100M
FREQUENCY (Hz)
–80
03217-024
–80
100k
1OUTPUT
Figure 23. Harmonic Distortion vs. Frequency
0
4
2
3
VOUT (V p-p)
5
Figure 26. Harmonic Distortion vs. VOUT , VMAG = 2.0 V
0
VDBS = 0.75V
VOUT = 1V p-p
RL = 1kΩ
–10
AMPLITUDE HARD LIMITED
1
0
03217-027
DISTORTION (dBc)
1.5
1.2
0
VDBS = 0.75V
VOUT = 1V p-p
–10
RL = 1kΩ
f = 10MHz
VOUT = 1V p-p
–10 RL = 1kΩ
–20
DISTORTION (dBc)
–20
–30
–40
HD3
–50
–60
HD2
–70
0
10
20
30
CLOAD (pF)
–30
HD3
–40
–50
HD2
–60
40
50
03217-025
DISTORTION (dBc)
0.6
0.9
VOUT (V p-p)
Figure 25. Harmonic Distortion vs. VOUT , VMAG = 0.5 V
0
–80
0.3
03217-026
1
Figure 24. Harmonic Distortion vs. CLOAD
–70
0
0.2
0.4
0.6
0.8
1.0
VDBS (V)
1.2
Figure 27. Harmonic Distortion vs. VDBS
Rev. H | Page 11 of 32
1.4
1.6
03217-028
NOISE (nV/√Hz)
f = 10MHz
–10
6
AD8330
Data Sheet
33
30
23
25
28
–10
3
20
23
–20
–7
–30
–17
–40
0.4
0.6
0.8
1.0
1.2
1.4
OIP3 (dBV rms)
10
13
–27
5
8
–37
1.6
0
VDBS (V)
0
0.4
0.2
Figure 28. Input V1dB Compression vs. VDBS
33
10
23
0
13
35
3
–20
–7
–30
–17
3
VMAG (V)
4
5
6
–27
3
33
28
20
23
f = 50MHz
15
18
10
13
5
8
0
0.2
Figure 29. Output V1dB Compression vs. VMAG
0.4
0.6
0.8
VMAG (V)
1.0
1.2
1.4
3
1.6
Figure 32. OIP3 vs. VMAG
0
1.5
VDBS = 0.75V
–10 VOUT = 1V p-p
1.0
–20
–30
0.5
VOUT (V)
–40
–50
–60
VDBS = 0V
0
–0.5
–70
–90
1M
10M
FREQUENCY (Hz)
100M
Figure 30. IMD3 Distortion vs. Frequency
–1.5
–50
–25
0
25
TIME (ns)
50
75
Figure 33. Full-Scale Transient Response, VDBS = 0 V
Rev. H | Page 12 of 32
100
03217-034
–1.0
–80
03217-031
IMD3 (dBc)
1.6
38
f = 10MHz
25
0
03217-030
2
OIP3 (dBV rms)
P1dB (REF 50Ω)
–10
1
1.4
43
30
0
1.2
40
f = 10MHz
–40
0.8
1.0
VDBS (V)
Figure 31. OIP3 vs. VDBS
20
INPUT V1dB COMPRESSION (dBV rms)
0.6
03217-032
0.2
18
OIP3 (dBm)
0
f = 50MHz
15
03217-033
–50
P1dB (REF 50Ω)
13
0
OIP3 (dBm)
f = 10MHz
f = 10MHz
03217-029
INPUT V1dB COMPRESSION (dBV rms)
10
Data Sheet
AD8330
1.5
1V
1.0
VOUT (V)
0.5
VDBS = 0.75V
0
–0.5
–25
0
25
TIME (ns)
50
75
100
1V
400ns
03217-038
–1.5
–50
03217-035
–1.0
Figure 37. VDBS Interface Response, Top: VDBS, Bottom: VOUT
Figure 34. Full-Scale Transient Response, VDBS = 0.75 V,
f = 1 MHz, VOUT = 2 V p-p
1.5
2V
1.0
VOUT (V)
0.5
VDBS = 1.5V
0
–1.0
–25
0
25
TIME (ns)
50
75
100
400ns
03217-036
1mV
–1.5
–50
03217-039
–0.5
Figure 38. VMAG Interface Response, Top: VMAG, Bottom: VOUT
Figure 35. Full-Scale Transient Response, VDBS = 1.5 V,
f = 1 MHz, VOUT = 2 V p-p
1V
500mV
CL = 12pF
VMAG = 5V
VMAG = 0.5V
CL = 54pF
CL = 24pF
100mV
Figure 36. Transient Response vs. Various Load Capacitances, G = 25 dB
Rev. H | Page 13 of 32
12.5ns
Figure 39. Transient Response vs. VMAG
03217-040
12.5ns
03217-037
VMAG = 0.05V
AD8330
Data Sheet
26
4.00V
OUTPUT
INPUT
25ns
+85°C
20
+25°C
18
–40°C
16
03217-041
50mV
22
14
Figure 40. Overdrive Response, VDBS = 1.5 V, VMAG = 0.5 V, 18.5 dB Overdrive
0
0.2
0.4
0.6
0.8
1.0
VDBS (V)
1.2
1.4
1.6
Figure 43. Supply Current vs. VDBS at Three Temperatures
2V
3.125V
2.5V
1.875V
3.125V
400ns
1.875V
03217-042
1V
100ns
Figure 41. ENBL Interface Response. Top: VENBL; Bottom: VOUT, f = 10 MHz
–10
–20
VDBS = 0.75V
–30
PSRR (dB)
–40
VPSI
–50
–60
VPOS
–70
VPSO
–80
–90
–110
1M
10M
FREQUENCY (Hz)
100M
200M
03217-043
–100
Figure 42. PSRR vs. Frequency
Rev. H | Page 14 of 32
Figure 44. CNTR Transient Response, Top: Input to CNTR,
Bottom: VOUT Single-Ended
03217-045
2.5V
03217-044
SUPPLY CURRENT (mA)
24
Data Sheet
AD8330
INPUT IS xlD
CIRCUIT DESCRIPTION
Many monolithic variable gain amplifiers use techniques that
share common principles that are broadly classified as translinear.
This term refers to circuit cells whose functions depend directly
on the very predictable properties of bipolar junction transistors,
notably the linear dependence of their transconductance on collector
current. Since the discovery of these cells in 1967, and their
commercial exploitation in products developed during the early
1970s, accurate wide bandwidth analog multipliers, dividers,
and variable gain amplifiers have invariably employed translinear
principles.
By varying IN, the overall function is that of a two-quadrant
analog multiplier, exhibiting a linear relationship to both the
signal modulation factor (x) and this numerator current. On
the other hand, by varying ID, a two-quadrant analog divider
is realized, having a hyperbolic gain function with respect to
the input factor, x, controlled by this denominator current. The
AD8330 exploits both modes of operation. However, because a
hyperbolic gain function is generally of less value than one in
which the decibel gain is a linear function of a control input, a
special interface is included to provide either increasing or
decreasing exponential control of ID.
(1–x) IN
2
+
Q4
Q2
ID
(1+x) IN
2
–
Q1
Q3
NUMERATOR
DENOMINATOR
BIAS CURRENT BIAS CURRENT
IN
Figure 45. Basic Core
ENBL
OFST
VPSI
BIAS AND
VREF
INHI
AD8330
VPOS
CNTR
VPSO
CM MODE AND
OFFSET CONTROL
OPHI
OUTPUT
STAGES
MODE
VGA CORE
GAIN INTERFACE
VDBS
CMGN
OPLO
OUTPUT
CONTROL CMOP
COMM
VMAG
03217-047
INLO
Figure 45 shows a basic representative cell comprising just four
transistors. This, or a very closely related form, is at the heart of
most translinear multipliers, dividers, and VGAs. The key concepts
are as follows:
In practice, the realization of the full potential of this circuit
involves many other factors, but these three elementary ideas
remain essential.
LOOP
AMPLIFIER
(1–x) ID
2
(1+x) ID
2
Although these techniques are well understood, the realization
of a high performance variable gain amplifier (VGA) requires
special technologies and attention to many subtle details in its
design. The AD8330 is fabricated on a proprietary silicon-oninsulator, complementary bipolar IC process and draws on
decades of experience in developing many leading edge products
using translinear principles to provide an unprecedented level of
versatility.
First, the ratio of the currents in the left-hand and right-hand
pairs of transistors is identical, represented by the modulation
factor, x, with values between −1 and +1. Second, the input
signal is arranged to modulate the fixed tail current, ID, to cause
the variable value of x, introduced in the left-hand pair, to be
replicated in the right-hand pair, and, thus, generate the output
by modulating its nominally fixed tail current, IN. Third, the
current gain of this cell is exactly G = IN/ID over many decades
of variable bias current.
OUTPUT IS xlN
G = IN/ID
03217-046
THEORY OF OPERATION
Figure 46. Block Schematic
Overall Structure
Figure 46 shows a block schematic of the AD8330 locating the
key sections. More detailed descriptions of its structure and
features are provided throughout the Theory of Operation
section; however, Figure 46 provides a general overview of its
capabilities.
The VGA core contains a more elaborate version of the cell
shown in Figure 45. The current, ID, is controlled exponentially
(linear-in-decibels) through the decibel gain interface at
Pin VDBS and its local common, Pin CMGN. The gain span
(that is, the decibel difference between maximum and minimum
values) provided by this control function is slightly more than
50 dB. The absolute gain from input to output is a function of
source and load impedance, and depends on the voltage on a
second gain control pin (VMAG), explained in the Normal
Operating Conditions section.
Rev. H | Page 15 of 32
AD8330
Data Sheet
To minimize confusion, normal operating conditions are
defined as follows:
•
•
•
•
•
The input pins are voltage driven (the source impedance is
assumed to be zero).
The output pins are open circuited (the load impedance is
assumed to be infinite).
Pin VMAG is unconnected setting up the output bias
current (IN in the four-transistor gain cell) to its nominal
value.
Pin CMGN is grounded.
MODE is either tied to a logic high or left unconnected, to
set the up gain mode.
The effects of other operating conditions are considered
separately.
Throughout this data sheet, the end-to-end voltage gain for the
normal operating conditions is referred to as the basic gain.
Under these conditions, it runs from 0 dB when VDBS = 0 V
(where this voltage is more exactly measured with reference to
Pin CMGN, which is not necessarily tied to ground) up to 50 dB
for VDBS = 1.5 V. The gain does not fold over when the VDBS
pin is driven below ground or above its nominal full-scale value.
The input is accepted at the INHI/INLO differential port. These
pins are internally biased to roughly the midpoint of the supply,
VS (it is actually ~2.75 V for VS = 5 V, VDBS = 0 V, and 1.5 V for
VS = 3 V), but the AD8330 is able to accept a forced commonmode value, from zero to VS, with certain limitations. This interface
provides good common-mode rejection up to high frequencies
(see Figure 16) and, thus, can be driven in either a single-sided
or differential manner. However, operation using a differential
drive is preferable, and this is assumed in the specifications,
unless otherwise stated.
The pin-to-pin input resistance is specified as 950 Ω ± 20%. The
driving-point impedance of the signal source can range from
zero up to values considerably in excess of this resistance, with a
corresponding variation in noise figure (see Figure 53). In most
cases, the input is coupled via two capacitors, chosen to provide
adequate low frequency transmission. This results in the minimum
input noise that increases when some other common-mode voltage is forced onto these pins. The short-circuit, input-referred
noise at maximum gain is approximately 5 nV/√Hz.
Output Pin OPHI and Output Pin OPLO operate at a commonmode voltage at the midpoint of the supply, VS/2, within a few
millivolts. This ensures that an analog-to-digital converter (ADC)
attached to these outputs operates within the often narrow
range permitted by their design. When a common-mode voltage
other than VS/2 is required at this interface, it can easily be forced
by applying an externally provided voltage to the output centering
pin, CNTR. This voltage can run from zero to the full supply,
though the use of such extreme values leaves only a small range
for the differential output signal swing.
The differential impedance measured between OPHI and OPLO is
150 Ω ± 20%. It follows that both the gain and the full-scale
voltage swing depend on the load impedance; both are nominally
halved when this is also 150 Ω. A fixed impedance output
interface, rather than an op amp style voltage-mode output, is
preferable in high speed applications because the effects of complex
reactive loads on the gain and phase can be better controlled.
The top end of the AD8330 ac response is optimally flat for a 12 pF
load on each pin, but this is not critical, and the system remains
stable for any value of load capacitance including zero.
Another useful feature of this VGA in connection with the
driving of an ADC is that the peak output magnitude can be
precisely controlled by the voltage on Pin VMAG. Usually, this
voltage is internally preset to 500 mV, and the peak differential
unloaded output swing is ±2 V ± 3%. However, any voltage from
zero to at least 5 V can be applied to this pin to alter the peak
output in an exactly proportional way. Because either output
pin can swing rail-to-rail, which in practice means down to at
least 0.35 V and to within the same voltage below the supply, the
peak-to-peak output between these pins can be as high as 10 V
using VS = 6 V.
CM MODE
FEEDBACK
VPSI
VPSO
TRANSIMPEDANCE
OUTPUT STAGE
INHI 500Ω
OPHI
ΔV = 0
ROUT = 150Ω
ΔV = 0
OPLO
INLO 500Ω
O/P CM-MODE
NORMALLY
AT VP/2
CNTR
LINEAR-IN-dB
INTERFACE
MAGNITUDE
INTERFACE
MODE
VMAG
VDBS
VDBS
COMM
100µA
12.65µA–4mA OR
4mA–12.65µA
VMAG
5kΩ
COMM
Figure 47. Schematic of Key Components
Linear-in-dB Gain Control (VDBS)
All Analog Devices, Inc., VGAs featuring a linear-in-dB gain
law, such as the X-AMP® family, provide exact, constant gain
scaling over the fully specified gain range, and the deviation
from the ideal response is within a small fraction of a dB. For
the AD8330, the scaling of both of its gain interfaces is
substantially independent of process, supply voltage, or
temperature. The basic gain, GB, is simply
GB (dB ) =
VDBS
30 mV
where VDBS is in volts.
Rev. H | Page 16 of 32
(1)
03217-048
Normal Operating Conditions
Data Sheet
AD8330
Alternatively, this can be expressed as a numerical gain
magnitude
provided by VMAG. The latter modifies the basic numerical gain
GBN to generate a total gain, expressed here in magnitude terms.
VDBS
GBN = 10 0.6 V
GT = GBN
(2)
The gain can be increased or decreased by changing the voltage,
VMAG, applied to the VMAG pin. The internally set default value
of 500 mV is derived from the same band gap reference that
determines the decibel scaling. The tolerance on this voltage,
and mismatches in certain on-chip resistors, cause small gain
errors (see the Specifications section). Though not all applications
of VGAs demand accurate gain calibration, it is a valuable asset
in many situations, for example, in reducing design tolerances.
Figure 47 shows the core circuit in more detail. The range and
scaling of VDBS is independent of the supply voltage, and the gain
control pin, VDBS, presents a high incremental input resistance
(~100 MΩ) with a low bias current (~100 nA), making the
AD8330 easy to drive from a variety of gain control sources.
VOUT = 2 × GIN × VIN × VMAG
GAIN (dB)
(4)
from which it is apparent that the AD8330 implements a linear,
two-quadrant multiplier with a bipolar VIN and a unipolar VMAG.
Because the AD8330 is a dc-coupled system, it can be used in
many applications where a wideband two-quadrant multiplier
function is required, from dc up to about 100 MHz from either
input (VIN or VMAG).
As VMAG is varied, so also is the peak output magnitude, up to a
point where this is limited by the absolute output limit imposed
by the supply voltage. In the absence of the latter effect, the
peak output into an open-circuited load is just
VOUT_PK = ±4 VMAG
The AD8330 supports many features that further extend the
versatility of this VGA in wide bandwidth gain control systems.
For example, the logic pin, MODE, allows the slope of the gain
function to be inverted, so that the basic gain starts at +50 dB
for a gain voltage, VDBS, of zero and runs down to 0 dB when
this voltage is at its maximum specified value of 1.5 V. The basic
forms of these two gain control modes are shown in Figure 48.
MODE PIN
HIGH, GAIN
INCREASES
WITH VDBS
MODE PIN
LOW, GAIN
DECREASES
WITH VDBS
40
(3)
Using this to calculate the output voltage,
Inversion of the Gain Slope
50
VMAG
0. 5 V
30
20
(5)
whereas for a load resistance of RL directly across OPHI and
OPLO, it is
VOUT _ PK =
±2 VMAG RL
(6)
(RL + 150 )
These capabilities are illustrated in Figure 49, where VS = 6 V,
RL = O/C, VDBS = 0 V, VIN is swept from −2.5 V dc to +2.5 V dc, and
VMAG is set to 0.25 V, 0.5 V, 1 V, and 2 V. Except for the last value
of VMAG, the peak output follows Equation 5. This exceeds the
supply-limited value when VMAG = 2 V and the peak output is
±5.65 V, that is, ±6 V − 0.35 V. Figure 50 demonstrates the high
speed multiplication capability. The signal input is a 100 MHz,
0.1 V sine wave, VDBS is set to 0.6 V, and VMAG is a square wave at
5 MHz alternating from 0.25 V to 1 V. The output is ideally a
sine wave switching in amplitude between 0.5 V and 2 V.
8
VMAG = 2V
6
1V
0.25
0.50
0.75
1.0
VDBS (V)
1.25
1.50
0.5V
2
VOUT (V)
0
03217-049
4
0
Figure 48. Two Gain Directions of the AD8330
Gain Magnitude Control (VMAG)
–2
In addition to the basic linear-in-dB control, two more gain
control features are provided. The voltage applied to Pin VMAG
provides accurate linear-in-magnitude gain control with a very
rapid response. The bandwidth of this interface is >100 MHz.
When this pin is unconnected, VMAG assumes its default value of
500 mV (see Figure 47) to set up the basic 0 dB to 50 dB range.
However, any voltage from ~15 mV to 5 V can be applied either
to lower the gain by up to 30 dB or to raise it by 20 dB. The
combined gain span is thus 100 dB, that is, the 50 dB basic gain
span provided by VDBS plus a 60 dB linear-in-magnitude span
–4
Rev. H | Page 17 of 32
0.25V
0
–6
–8
–3
–2
–1
0
VIN ( V)
1
2
Figure 49. Effect of VMAG on Gain and Peak Output
3
03217-050
10
AD8330
Data Sheet
VIN
0.10
Amplitude/Phase Response
0.05
The ac response of the AD8330 is remarkably consistent not
only over the full 50 dB of its basic gain range, but also with
changes of gain due to alteration of VMAG, as demonstrated in
Figure 51. This is an overlay of two sets of results: first, with a
very low VMAG of 16 mV that reduces the overall gain by 30 dB
[20 × log10(500 mV/16 mV)]; second, with VMAG = 5 V that
increases the gain by 20 dB = 20 × log10(5 V/0.5 V).
0
–0.05
–0.10
VMAG
1.2
1.0
0.8
0.6
0.4
0.2
0
90
VOUT
–200
–100
0
100
200
TIME (ns)
300
50
30
10
–10
–30
–50
100k
0
Another gain related feature allows both gain control ranges
to be accurately raised by 200 mV. To enable this offset, open
circuit CMGN (Pin 6, LFCSP; Pin 8, QSOP) and add a 0.1 µF
capacitor to ground. In use, the nominal range for VDBS extends
from 0.2 V to 1.7 V and VMAG from 0.2 V to 5.2 V. These
specifications apply for any supply voltage. This allows the use
of DACs whose output range does not include ground as sources
for the gain control function(s).
Note that the 200 mV that appears on this pin affects the
response to an externally applied VMAG, but when Pin VMAG is
unconnected, the internally set default value of 0.5 V still applies.
Furthermore, Pin CMGN can, if desired, be driven by a usersupplied voltage to reposition the baseline for VDBS (or for an
externally applied VMAG) to any other voltage up to 500 mV. In
all cases, the gain scaling, its law conformance, and temperature
stability are unaffected.
Two Classes of Variable Gain Amplifiers
Note that there are two broad classes of VGAs. The first type is
designed to cope with a very wide range of input amplitudes
and, by virtue of its gain control function, compress this range
down to an essentially constant output. This is the function
needed in an AGC system. Such a VGA is called an IVGA,
referring to a structure optimized to address a wide range of
input amplitudes. By contrast, an OVGA is optimized to deliver
a wide range of output values while operating with an essentially
constant input amplitude. This function might be needed, for
example, in providing a variable drive to a power amplifier.
It is apparent from the foregoing sections that the AD8330 is
both an IVGA and an OVGA in one package. This is an unusual
and possibly confusing degree of versatility for a VGA; therefore,
these two distinct control functions are described at separate
points throughout this data sheet to explain the operation and
applications of this product. It is, nevertheless, useful to briefly
describe the capabilities of these features when used together.
PHASE (Degrees)
Figure 50. Using VMAG in Modulation Mode
1M
10M
100M
300M
G = +70dB
–50
–100
–150
–200
–250
–300
–350
10k
G = –20dB
100k
1M
10M
FREQUENCY (Hz)
100M 300M
03217-052
–300
GAIN (dB)
70
03217-051
2.5
2.0
1.5
1.0
0.5
0
–0.5
–1.0
–1.5
–2.0
–400
Figure 51. AC Performance over a 100 dB Gain Range Obtained by
Using Two Values of VMAG
This 50 dB step change in gain produces two sets of gain curves,
having a total gain span of 100 dB. It is apparent that the amplitude and phase response are essentially independent of the gain
over this wide range, an aspect of the AD8330 performance
potential unprecedented in any prior VGA.
It is unusual for an application to require such a wide range of
gains; and, as a practical matter, the peak output voltage for
VMAG = 16 mV is reduced by the factor 16/500, compared to its
nominal value of ±2 V, to only ±64 mV. As previously noted,
most applications of VGAs require that they operate in a mode
that is predominantly of either an IVGA or OVGA style, rather
than mixed modes.
With this limitation in mind, and simply to illustrate the
unusual possibilities afforded by the AD8330, note that, with
appropriate drive to VDBS and VMAG in tandem, the gain span is a
remarkable 120 dB, extending from −50 dB to +70 dB, as shown
in Figure 52 for operation at 1 MHz and 100 MHz. In this case,
VDBS and VMAG are driven from a common control voltage,
VGAIN, that varies from 1.2 mV to 5 V, with 30% (1.5/5) of VGAIN
applied to VDBS, and 100% applied to VMAG.
The gain varies in a linear-in-dB manner with VDBS, although
the response from VMAG is linear-in-magnitude. Consequently, the
overall numerical gain as the product of these two functions is
GAIN = VGAIN / 0.5 V × 0.3 ×10
VGAIN
0. 6 V
In rare cases where such a wide gain range is of value, the
calibration is still accurate and the temperature is stable.
Rev. H | Page 18 of 32
(7)
Data Sheet
AD8330
80
perfect, the noise figure cannot be better than 3 dB. The 1 kΩ
internal termination resistance results in a minimum noise
figure of 3 dB for an RS of 1 kΩ if the amplifier were noise-free.
However, this is not the case, and the minimum noise figure
occurs at a slightly different value of RS (for an example, see
Figure 53 and the Using the AD8330 section).
40
20
0
–20
–40
15
10k
14
1k
13
100
12
10
0.1
VGAIN (V)
1
10
Figure 52. Gain Control Function and Input Referred Noise Spectral Density
over a 120 dB Range
10
9
8
7
Noise, Input Capacity, and Dynamic Range
Below midgain (25 dB, VDBS = 0.75 V), noise in the output
section dominates, and the total input noise is 11 nV/√Hz, or
4.9 µV rms in a 200 kHz bandwidth, and the peak input is
78 mV rms. Thus, the dynamic range increases to 84 dB.
At minimum gain, the input noise is up to 120 nV/√Hz, or
53.7 mV rms in the assumed 200 kHz bandwidth, while the
input capacity is ±2 V, or +1.414 V rms (sine), a dynamic range
of 88.4 dB. In calculating the dynamic range for other channel
bandwidths, ∆f, subtract 10 log10(∆f/200 kHz) from these
illustrative values. A system operating with a 2 MHz bandwidth,
for example, exhibits dynamic range values that are uniformly
10 dB lower; used in an audio application with a 20 kHz bandwidth, they are 10 dB higher.
Noise figure is a misleading metric for amplifiers that are not
impedance matched at their input, which is the special condition resulting only when both the voltage and current components
of a signal, that is, the signal power, are used at the input port.
When a source of impedance (RS) is terminated using a resistor
of RS (a condition that is not to be confused with matching),
only one of these components is used, either the current (as in
the AD8330) or the voltage. Then, even if the amplifier is
6
5
10
100
RS (Ω)
1k
10k
Figure 53. Noise Figure for Source Resistance of 50 Ω to 5 kΩ, at f = 10 MHz
(Lower) and 100 MHz (Simulation)
144
140
DYNAMIC RANGE (dB/√Hz)
The design of variable gain amplifiers invariably incurs some
compromises in noise performance. However, the structure of
the AD8330 is such that this penalty is minimal. Examination
of the simplified schematic (Figure 47) shows that the input
voltage is converted to current-mode form by the two 500 Ω
resistors at Pin INHI and Pin INLO, whose combined Johnson
noise contributes 4.08 nV/√Hz. The total input noise at full
gain, when driven from a low impedance source, is typically
5 nV/√Hz after accounting for the voltage and current noise
contributions of the loop amplifier. For a 200 kHz channel
bandwidth, this amounts to 2.24 μV rms. The peak input at full
gain is ±6.4 mV, or +4.5 mV rms for a sine wave signal. The
signal-to-noise ratio at full input, that is, the dynamic range, for
these conditions is, thus, 20 log10(4.5 mV/2.24 μV), or 66 dB.
The value of VMAG has essentially no effect on the input referred
noise, but it is assumed to be 0.5 V.
03217-054
0.01
11
CONSTANT 1V rms
OUTPUT, BOTH CASES
136
132
128
X-AMP WITH 40dB
OF GAIN AND AN
INPUT NSD
OF nV/√Hz
124
120
116
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4 1.5
VDBS (V)
03217-055
1
0.001
03217-053
NOISE (nV/√Hz)
–60
100k
NOISE FIGURE
GAIN (dB)
60
Figure 54. Dynamic Range in dB/√Hz vs. VDBS (VMAG = 0.5 V, 1 V rms Output)
Compared with a Representative X-AMP (Simulation)
Dynamic Range
The ratio of peak output swing, expressed in rms terms, to the
output-referred noise spectral density provides a measure of
dynamic range, in dB/√Hz. For a certain class of variable gain
amplifiers, exemplified by the Analog Devices X-AMP® family,
the dynamic range is essentially independent of the gain setting
because the peak output swing and noise are both constant. The
AD8330 provides a different dynamic range profile because
there is no longer a constant relationship between these two
parameters. Figure 54 compares the dynamic range of the
AD8330 to a representative X-AMP.
Input Common-Mode Range and Rejection Ratio
AC-couple the input pins, INHI and INLO, in most applications
to achieve the stated noise performance. In general, when direct
coupling is used, care must be taken in setting the dc voltage
level at these inputs, and particularly when minimizing noise is
Rev. H | Page 19 of 32
AD8330
Data Sheet
critical. This objective is complicated by the fact that the
common-mode level varies with the basic gain voltage, VDBS.
Figure 55 shows this relationship for a supply voltage of 5 V, for
temperatures of −40°C, +25°C, and +85°C. Figure 56 shows the
input noise spectral density (RS = 0) vs. the input commonmode voltage, for VDBS = 0.5 V, 0.6 V, 0.75 V, and 1.5 V. It is
apparent that there is a broad range over which the noise is
unaffected by this dc level. The input CMRR is excellent (see
Figure 16).
T = +85°C
VNOISE_OUT = (0.1 + 0.32 VMAG) µV/√Hz
for RS = 0 and VDBS = 1.5 V, assuming an input noise of 5 nV/√Hz.
The output noise for very small values of VMAG (at or below 15 mV)
is not precise, partly because the small input offset associated
with this interface has a large effect on the gain.
The AD8330 includes an offset compensation feature that is
operational in the default condition (no connection to Pin OFST).
This loop introduces a high-pass filter function into the signal
path, whose −3 dB corner frequency is at
3.0
T = –40°C
2.9
f HPF =
2.8
2.7
2.6
0
0.2
0.4
0.6
0.8
VDBS (V)
1.0
1.2
1.4
1.6
VDBS = 0.6V
16
14
VDBS = 0.75V
12
SIMULATION
6
4
0
0.4
VDBS = 1.5V
0.8 1.2 1.6 2.0 2.4 2.8 3.2 3.6 4.0
COMMON-MODE VOLTAGE AT INHI, INLO (V)
4.4
C HP (C in μF)
HP
(12)
The offset compensation feature can be disabled simply by
grounding the OFST pin. This provides a dc-coupled signal
path, with no other effects on the overall ac response. Input
offsets must be externally nulled in this mode of operation, as
shown in Figure 58.
10
8
330 µ
A small amount of peaking at this corner when using small
capacitor values can be avoided by adding a series resistor.
Useful combinations are CHP = 3 nF, RHP = 180 Ω, f = 100 kHz;
CHP = 33 nF, RHP = 10 Ω, f = 10 kHz; CHP = 0.33 μF, RHP = 0 Ω,
f = 1 kHz; CHP = 3.3 μF, RHP = 0 Ω, f = 100 Hz.
20
4.8
03217-057
INPUT REFERRED NOISE (nV/√Hz)
f HPF =
VDBS = 0.5V
18
(11)
This evaluates to
26
22
1
(2π RINT CHP )
where:
CHP is the external capacitance added from OFST to CNTR.
RINT is an internal resistance of approximately 480 Ω, having a
maximum uncertainty of about ±20%.
Figure 55. Common-Mode Voltage at Input Pins vs. VDBS, for VS = 5 V,
T = −40°C, + 25°C, and + 85°C
24
(10)
Offset Compensation
T = +25°C
3.1
03217-056
DC VOLTAGE AT INHI, INLO (V)
3.2
For example, using a reduced value of VMAG = 0.25 V that lowers all
gain values by 6 dB, the peak output swing is ±1 V (differentially)
and the output noise spectral density evaluates to 102.5 nV/√Hz.
The peak output swing is no different at full gain, but the noise
becomes
Effects of Loading on Gain and AC Response
Figure 56. Input Noise vs. Common-Mode Input Voltage for
VDBS = 0.5 V, 0.6 V, 0.75 V, and 1.5 V
Output Noise and Peak Swing
The output noise of the AD8330 is the input noise multiplied by
the overall gain, including any optional change to the voltage,
VMAG, applied to Pin VMAG. The peak output swing is also
proportional to this voltage, which, at low gains and high values
of VMAG, affects the output noise.
The scaling for VDBS = 0 V is as follows:
VOUT_PK = ±4 VMAG
(8)
VNOISE_OUT = (85 + 70 VMAG) nV/√Hz
(9)
The differential output impedance (RO) is 150 Ω, and the frequency response of the output stage is optimized for operation
with a certain load capacitance on each output pin (OPHI and
OPLO) to ground, in combination with a load resistance (RL)
directly across these pins. In the absence of these capacitances,
there is a small amount of peaking at the top extremity of the ac
response. Suitable combinations are: RL = ∞, CL = 12 pF; RL =
150 Ω, CL = 25 pF; RL = 75 Ω, CL = 40 pF; or RL = 50 Ω, CL = 50 pF.
The gain calibration is specified for an open-circuited load,
such as the high input resistance of an ADC. When resistively
loaded, all gain values are nominally lowered as follows:
G LOADED =
Rev. H | Page 20 of 32
GUNLOADED R L
(150 Ω + R L )
(13)
Data Sheet
AD8330
Gain Errors Due to On-Chip Resistor Tolerances
In all cases where external resistors are used, keep in mind that
all on-chip resistances, including the RO and the input resistance
(RI), are subject to variances of up to ±20%.
the output stage where decoupling can be useful in maintaining
a glitch-free output. Figure 57 shows the general case, where VPSI
and VPSO are each provided with their own decoupling network,
but this is not needed in all cases.
VS 2.7V TO 6V
RD1
These variances need to be accounted for when calculating the
gain with input and output loading. This sensitivity can be avoided
by adjusting the source and load resistances to bear an inverse
relationship as follows:
CHPF
ENBL
CD1
VPSI
If RS = αRI, then make RL = RO/α; or,
if RL = αRO, then make RS = RI/α
The simplest case is when RS = 1 kΩ and RL = 150 Ω, therefore,
the gain is 12 dB lower than the basic value. The reduction of
peak swing at the load can be corrected by using VMAG = 1 V,
thereby restoring 6 dB of gain; using VMAG = 2 V restores the full
basic gain and doubles the peak available output swing.
The input common-mode voltage, VCMI, at Pin INHI and
Pin INLO is slaved to the output. It bears a y = mx + b linear
and offset relationship to VCNTR as shown in Equation 14 where
y = VCMI, m = 0.757, x = VCNTR, and b = 1.12 V for VDBS = 0.75 V
and T = 25°C.
VCMI = 0.757 VCNTR + 1.12 V
(14)
The effects of VDBS and ambient temperature on VCMI are shown
in Figure 55. Thus, the default value for VCMI for VDBS = 0.75V,
T = 25°C and VS = 5 V is 3.01 V.
USING THE AD8330
This section describes a few general aspects of using the AD8330.
Applying the AD8330 to a wide variety of circumstances requires
very few precautions.
As in all high frequency circuits, careful observation of the ground
nodes associated with each function is important. Three positive
supply pins are provided: VPSI supports the input circuitry that
often operates at a relatively high sensitivity; VPOS supports
general bias sources and needs no decoupling; and VPSO biases
BIAS AND
V-REF
VPOS
INPUT,
0V TO ±2V MAX
VGA CORE
MODE
VDBS
BASIC GAIN BIAS
VDBS: 0V TO 1.5V
CD3
OPHI
OUTPUT,
±2V MAX
OUTPUT
STAGES
OPLO
INLO
NC
CNTR
VPSO
CM MODE AND
OFFSET CONTROL
INHI
Output (Input) Common-Mode Control
The output voltages are nominally positioned at the midpoint of
the supply, VS/2, over the range 2.7 V < VS < 6 V, and this voltage
appears at Pin CNTR, which is not normally expected to be loaded
(the source resistance is ~4 kΩ). However, some circumstances
require a small change in this voltage, and a resistor from CNTR to
ground can lower this voltage, whereas a resistor to the supply
raises it. On the other hand, this pin can be driven by an external
voltage source to set the common-mode level to satisfy, for
example, the needs of a following ADC. Any value from 0.5 V
above ground to 0.5 V below the supply is permissible. Of course,
when using an extreme common-mode level, the available
output swing is limited, and it is recommended that a value equal
or close to the default of VCNTR = VS/2 be used. There may be a
few millivolts of offset between the applied voltage and the
actual common-mode level at the output pins.
OFST
RD2
CD2
GAIN INTERFACE
CMGN
OUTPUT
CONTROL CMOP
COMM
VMAG
NC
GROUND
03217-058
Thus, when RL = 150 Ω, the gain is reduced by 6 dB; for RL = 75 Ω,
the reduction is 9.5 dB; and for RL = 50 Ω, it is 12 dB.
Figure 57. Power Supply Decoupling and Basic Connections
Because of the differential nature of the signal path, power
supply decoupling is, in general, much less critical than in a
single-sided amplifier; and where the minimization of boardlevel components is especially crucial, it is possible that these
pins need no decoupling at all. On the other hand, when the
signal source is single-sided, giving extra attention to the
decoupling on Pin VPSI is sometimes required. Likewise, care is
required in decoupling the VPSO pin if the output is loaded on
only one of its two output pins. The general common (COMM)
and the output stage common (CMOP) are usually grounded as
shown in the Figure 57; however, the Applications Information
section shows how a negative supply can optionally be used.
The AD8330 is enabled by taking the ENBL pin to a logical high
(or, in all cases, the supply). The UP gain mode is enabled either
by leaving the MODE pin unconnected or taking it to a logical
high. When the opposite gain direction is needed, ground the
MODE pin or drive it to a logical low. The CHPF capacitor
determines the low-pass corner of the offset loop; this is
preferably tied to the CNTR pin that in turn, must be decoupled
to ground. The gain interface common pin (CMGN) is
grounded, and the output magnitude control pin (VMAG) is
left unconnected, or can optionally be connected to a 500 mV
source for basic gain calibration.
Connections to the input and output pins are not shown in Figure 57
because of the many options that are available. When the AD8330 is
used to drive an ADC, connect the OPHI and OPLO pins directly
to the differential inputs of a suitable converter, such as an AD9214.
Rev. H | Page 21 of 32
AD8330
Data Sheet
When the loads to be driven introduce a dc resistive path to
ground, coupling capacitors must be used. These must be of
sufficient value to pass the lowest frequency components of the
signal without excessive attenuation. Keep in mind that the voltage
swing on such loads alternates both above and below ground,
requiring that the subsequent component must be able to cope
with negative signal excursions.
Gain and Swing Adjustments When Loaded
The output can also be coupled to a load via a transformer to
achieve a higher load power by impedance transformation. For
example, using a 2:1 turns ratio, a 50 Ω final load presents a 200 Ω
load on the output. The gain loss (relative to the basic value with no
termination) is 20 log10{(200+150)/200} or 4.86 dB, which can
be restored by raising the voltage on the VMAG pin by a factor
of 104.86/20 or × 1.75, from its basic value of 0.5 V to 0.875 V. This
also restores the peak swing at the 200 Ω level to ±2 V, or ±1 V
into the 50 Ω final load.
Whenever a stable supply voltage is available, additional voltage
swing can be provided by adding a resistor from the VMAG pin
to the supply. The calculation is based on knowing that the internal
bias is delivered via a 5 kΩ source; because an additional 0.375 V is
needed, the current in this external resistor must be 0.375 V/
5 kΩ = 75 μA. Thus, using a 5 V supply, a resistor of 5 V − 0.875 V/
75 μA = 55 kΩ is used. Based on this example, the corrections
for other load conditions are easy to calculate. If the effects on
gain and peak output swing due to supply variations cannot be
tolerated, VMAG must be driven by an accurate voltage.
DC-Coupled Signal Path
In many cases, where the VGA is not required to provide its
lowest noise, the full common-mode input range of zero to VS
can be used without problems, avoiding the need for any ac
coupling means. However, such direct coupling at both the input
and output does not automatically result in a fully dc-coupled
signal path. The internal offset compensation loop must also be
disengaged by connecting the OFST pin to ground. Note that at
the maximum basic gain of 50 dB (×316), every millivolt of offset at
the input, arising from whatever source, causes an output offset of
316 mV, which is an appreciable fraction of the peak output swing.
Because the offset correction loop is placed after the front-end
variable gain sections of the AD8330, the most effective way of
dealing with such offsets is at the input pins, as shown in Figure 58.
For example, assume, for illustrative purposes, that the resistances
associated with each side of the source in a certain application
are 50 Ω. If this source has a very low (op amp) output impedance,
insert the extra resistors, with a negligible noise penalty and an
attenuation of only 0.83 dB. The resistor values shown provide a
trim range of about ±2 mV.
Using Single-Sided Sources and Loads
Where the source provides a single-sided output, either INHI or
INLO can be used for the input, with a polarity change when using
INLO. The unused pin must be connected either through a capacitor
to ground, or through a dc bias point that corresponds closely to
the dc level on the active signal pin. The input CMRR over the
full frequency range is illustrated in Figure 59. In some cases, an
additional element such as a SAW filter (having a single-sided
balanced configuration) or a flux-coupled transformer can be
interposed. Where this element must be terminated in the correct
impedance, other than 1 kΩ, it is necessary to add either shunt
or series resistors at this interface.
The dc common-mode voltage at the input pins varies with the
supply, the basic gain bias, and temperature (see Figure 55); for
this reason, many applications need to use coupling capacitors
from the source that are large enough to support the lowest
frequencies to be transmitted. Using one capacitor at each input
pin, their minimum values can be readily found from the expression
CIN_CPL= (320 µF/fHPF)
VS 2.7V TO 6V
RD1
Input Coupling
(15)
where fHPF is the –3dB frequency expressed in hertz. Thus, for
an fHPF of 10 kHz, 33 nF capacitors are used.
CD2
ENBL
CD1
VPSI
RS ASSUMED
50kΩ
TO BE 50Ω
ON EACH
SIDE
Occasionally, it is possible to avoid the use of coupling capacitors
when the dc level of the driving source is within a certain range,
as shown in Figure 56. This range extends from 3.5 V to 4.5 V when
using a 5 V supply, and at high basic gains, where the effect of an
incorrect dc level degrades the noise level due to internal aspects of
the input stage. For example, suppose the driver, IC, is an LNA
having an output topology in which its load resistors are taken
to the supply, and the output is buffered by emitter followers.
This presents a source for the AD8330 that can be directly coupled.
Rev. H | Page 22 of 32
75kΩ
OFST
BIAS AND
V-REF
VPOS
CM MODE AND
OFFSET CONTROL
INHI
BASIC GAIN BIAS
VDBS: 0V TO 1.5V
OUTPUT,
±2V MAX
OUTPUT
STAGES
INLO
VDBS
CD3
VPSO
OPHI
VGA CORE
MODE
CNTR
RD2
OPLO
GAIN INTERFACE
CMGN
OUTPUT
CONTROL CMOP
COMM
VMAG
NC
GROUND
Figure 58. Input Offset Nulling in a DC-Coupled System
03217-059
If an adjustment is needed to this common-mode level, it can be
introduced by applying that voltage to the CNTR pin, or, more
simply, by using a resistor from this pin to either ground or the
supply (see the Applications Information section). The CNTR pin
can also supply the common-mode voltage to an ADC that
supports such a feature.
Data Sheet
80
VDBS = 1.5V
VDBS = 0.75V
In Case 3 and Case 4, a further factor of ×1.33 is needed to make
up the 2.5 dB loss, that is, raise VMAG to 2 V. With the restoration
of gain, the peak output swing at the load is, likewise restored to
±2 V.
OFST: ENABLED
DISABLED
70
CMRR (dB)
60
50
40
Pulse Operation
VDBS = 0V
30
20
10
1M
10M
FREQUENCY (Hz)
1 00 M
03217-060
0
–10
50k 100k
Because of the considerable variation between applications, only
general recommendations can be made with regard to minimizing
pulse overshoot and droop. The former can be optimized by
adding small load capacitances, if necessary; the latter requires
the use of sufficiently large capacitors (C1).
Figure 59. Input CMRR vs. Frequency for Various Values of VDBS
30
LINE 1
LINE 3
10
0
LINE 4
–10
LINE 2
–20
PHASE (Degrees)
–30
0
LINE 2
–100
LINE 3
–200
LINE 4
–300
–400
LINE 1
–500
–600
1M
10M
FREQUENCY (Hz)
100M
500M
03217-061
GAIN (dB)
20
When using the AD8330 in applications where its transient
response is of greater interest and the outputs are conveyed to
their loads via coaxial cables, the added capacitances can slightly
differ in value, and can be placed either at the sending or load
end of the cables, or divided between these nodes. Figure 61
shows an illustrative example where dual, 1 meter, 75 Ω cables
are driven through dc-blocking capacitors and are independently
terminated at ground level.
Figure 60. AC Gain and Phase for Various Loading Conditions
When driving a single-sided load, either OPHI or OPLO can be
used. These outputs are very symmetric, so the only effect of
this choice is to select the desired polarity. However, when the
frequency range of interest extends to the upper limits of the
AD8330, attach a dummy resistor of the same value to the
unused output. Figure 60 illustrates the ac gain and phase
response for various loads and VDBS = 0.75 V. Line 1 shows the
unloaded (CL = 12 pF) case for reference; the gain is 6 dB lower
(20 dB) using only the single-sided output. Adding a 75 Ω load
from OPHI to an ac ground results in Line 2. The gain becomes
a factor of ×1.5 V or 3.54 dB lower, but artifacts of the output
common-mode control loop appear in both the magnitude and
phase response.
Figure 62 shows typical results for VDBS = 0.24 V, a square wave
input amplitude of 450 mV (the actual combination is not
important), a rise time of 2 ns, and VMAG raised to 2.0 V. In the
upper waveforms, the load capacitors are both zero, and a small
amount of overshoot is visible; with 40 pF, the response is cleaner.
A shunt capacitance of 20 pF from OPHI to OPLO has a similar
effect. Coupling capacitors for this demonstration are sufficiently
large to prevent any visible droop over this time scale. The outputs
at the load side eventually assume a mean value of zero, with
negative and positive excursions depending on the duty cycle.
CD2
ENBL
VPSI
OFST
BIAS AND
V-REF
VPOS
CNTR
CM MODE AND
OFFSET CONTROL
RD2
VS 2.7V–6V
CD3
VPSO
RL1
C1
INHI
OPHI
VGA CORE
INLO
MODE
Adding a dummy 75 Ω to OPLO results in Line 3: the gain is a
further 2.5 dB lower, at about 14 dB. The CM artifacts are no longer
present but a small amount of peaking occurs. If objectionable, this
can be eliminated by raising both of the capacitors on the
output pins to 25 pF, as shown in Line 4 of Figure 60.
The gain reduction incurred both by using only one output and
by the additional effect of loading can be overcome by taking
advantage of the VMAG feature, provided primarily for just such
circumstances. Thus, to restore the basic gain in the first case
(Line 1), apply a 1 V source to this pin; to restore the gain in the
second case, this voltage must be raised by a factor of ×1.5 to 1.5 V.
Rev. H | Page 23 of 32
VDBS
CL1
OUTPUT
STAGES
C1
OPLO
GAIN INTERFACE
CMGN
CL2
RL2
OUTPUT
CONTROL CMOP
COMM
VMAG
NC
03217-062
90
AD8330
Figure 61. Driving Dual Cables with Grounded Loads
AD8330
Table 5. Preserving Absolute Gain
0
5ns
10ns
15ns
20ns
25ns
03217-063
1.2
1.0
0.8
0.6
0.4
0.2
0
–0.2
0.2
0
–0.2
–0.4
–0.6
–0.8
–1.0
–1.2
1.2
1.0
0.8
0.6
0.4
0.2
0
–0.2
0.2
0
–0.2
–0.4
–0.6
–0.8
–1.0
–1.2
Data Sheet
Figure 62. Typical Pulse Response for Figure 61
The bandwidth from Pin VMAG to these outputs is somewhat
higher than from the normal input pins. Thus, when this pin is
used to rapidly modulate the primary signal, some further
experimentation with response optimization may be required.
In general, the AD8330 is very tolerant of a wide range of
loading conditions.
Preserving Absolute Gain
Although the AD8330 is not laser trimmed, its absolute gain
calibration, based mainly on ratios, is very good. Full details are
found in the Specifications section and in the typical performance
curves (see the Typical Performance Characteristics section).
Nevertheless, having finite input and output impedances, the
gain is necessarily dependent on the source and load conditions.
The loss that is incurred when either of these is finite causes an
error in the absolute gain. The absolute gain can also be
uncertain due to the approximately ±20% tolerance in the
absolute value of the input and output impedances.
Often, such losses and uncertainties can be tolerated and
accommodated by a correction to the gain control bias. On the
other hand, the error in the loss can be essentially nulled by
using appropriate modifications to either the source impedance
(RS) or the load impedance (RL), or both (in some cases by
padding them with series or shunt components).
The formulation for this correction technique was previously
described. However, to simplify its use, Table 5 shows spot
values for combinations of RS and RL resulting in an overall loss
that is not dependent on sample-to-sample variations in on chip
resistances. Furthermore, this fixed and predictable loss can be
corrected by an adjustment to VMAG, as indicated in Table 5.
RS (Ω)
10
15
20
30
50
75
100
150
200
300
500
750
1k
1.5 k
2k
Uncorrected Loss
Factor
dB
0.980
0.17
0.971
0.26
0.961
0.34
0.943
0.51
0.907
0.85
0.865
1.26
0.826
1.66
0.756
2.43
0.694
3.17
0.592
4.56
0.444
7.04
0.327
9.72
0.250
12.0
0.160
15.9
0.111
19.1
RL (Ω)
15 k
10 k
7.5 k
5.0 k
3.0 k
2.0 k
1.5 k
1.0 k
750
500
300
200
150
100
75
VMAG Required to
Correct Loss
0.510
0.515
0.520
0.530
0.551
0.578
0.605
0.661
0.720
0.845
1.125
1.531
2.000
3.125
4.500
Calculation of Noise Figure
The AD8330 noise is a consequence of its intrinsic voltage noise
spectral density (ENSD) and the current noise spectral density (INSD).
Their combined effect generates a net input noise, VNOISE_IN, that
is a function of the input resistance of the device (RI), nominally
1 kΩ, and the differential source resistance (RS) as follows:
VNOISE _ IN =
{ENSD2 + I NSD2 (RI + RS )2 }
(16)
Note that purely resistive source and input impedances as a concession to simplicity is assumed. A more thorough treatment of
noise mechanisms, for the case where the source is reactive, is
beyond the scope of these brief notes. Also note that VNOISE_IN is
the voltage noise spectral density appearing across INHI and
INLO, the differential input pins. In preparing for the calculation
of the noise figure, VSIG is defined as the open-circuit signal
voltage across the source, and VIN is defined as the differential
input to the AD8330. The relationship is simply
VIN =
VSIG RI
(17)
(RI + RS )
At maximum gain, ENSD is 4.1 nV/√Hz, and INSD is 3 pA/√Hz.
Thus, the short-circuit voltage noise is
VNOISE _ IN =
{(4.1 n V/ Hz ) + (3pA / Hz ) (1 kΩ + 0) } =
2
2
2
5.08 nV/√Hz
(18)
Next, examine the net noise when RS = RI = 1 kΩ, often incorrectly called the matching condition, rather than source impedance
termination, which is the actual situation in this case.
Repeating the procedure,
V NOISE _ IN =
= 7.3 nV/√Hz
Rev. H | Page 24 of 32
(4.1 nV /
Hz
) + (3 pA /
2
Hz
) (1 kΩ + 1 kΩ)
2
2
(19)
Data Sheet
AD8330
The noise figure is the decibel representation of the noise factor,
NFAC, commonly defined as follows:
N FAC =
SNR at Input
SNR at Output
(20)
However, this is equivalent to
N FAC =
SNR at the Source
SNR at the Input Pins
(21)
Let VNSD be the voltage noise spectral density √kTRS due to the
source resistance. Using Equation 17 gives
N FAC =
=
VSIG {RI /(RI + RS )}/ VNSD
VIN /{VNOISE _ IN RS /(RI + RS )}
P1dB and V1dB
RI VNOISE _ IN
RS VNSD
(22)
Then, using the result from Equation 19 for a source resistance
of 1 kΩ, having a noise-spectral density of 4.08 nV/√Hz produces
N FAC =
(1 kΩ)(7.3 nV /
(1 kΩ)(4.08 nV/
Hz
) = 1.79
)
Hz
(23)
Finally, converting this to decibels using
NFIG = 10 log10(NFAC)
(24)
Thus, the resultant noise figure in this example is 5.06 dB,
which is somewhat lower than the value shown in Figure 53 for
this operating condition.
Noise as a Function of VDBS
The chief consequence of lowering the basic gain using VDBS is
that the current noise spectral density INSD increases with the
square root of the basic gain magnitude, GBN such that
INSD = (3 pA/√Hz)(√GBN)
In practice, however, the effect of device mismatches and junction resistances in the core cell, and other mechanisms in its
supporting circuitry inevitably cause distortion, further aggravated
by other effects in the later output stages. Some of these effects
are very consistent from one sample to the next, while those due
to mismatches (causing predominantly even-order distortion
components) are quite variable. Where the highest linearity
(and lowest noise) is demanded, consider using one of the XAMP products such as the AD603 (single-channel), AD604
(dual-channel), or AD8332 (wideband dual-channel with
ultralow noise LNAs).
(25)
Therefore, at the minimum basic gain of ×0, INSD rises to
53.3 pA/√Hz. However, the noise figure rises to 17.2 db if it
is recalculated using the procedures in Equation 16 through
Equation 24.
Distortion Considerations
Continuously variable gain amplifiers invariably employ
nonlinear circuit elements; consequently, it is common for their
distortion to be higher than well-designed fixed gain amplifiers.
The translinear multiplier principles used in the AD8330, in
theory, yield extremely low distortion, a result of the fundamental linearization technique that is an inherent aspect of
these circuits.
In addition to the nonlinearities that arise within the core of the
AD8330, at moderate output levels, another metric that is more
commonly stated for RF components that deliver appreciable
power to a load is the 1 dB compression point. This is defined
in a very specific manner: it is that point at which, with increasing
output level, the power delivered to the load eventually falls to a
value that is 1 dB lower than it would be for a perfectly linear
system. (Although this metric is sometimes called the 1 dB gain
compression point, it is important to note that this is not the
output level at which the incremental gain has fallen by 1 dB).
As shown in Figure 49, the output of the AD8330 limits quite
abruptly, and the gain drops sharply above the clipping level.
The output power, on the other hand, using an external resistive
load, RL, continues to increase. In the most extreme case, the
waveform changes from the sinusoidal form of the test signal,
with an amplitude just below the clipping level, VCLIP, to a
square wave of precisely the same amplitude. The change in
power over this range is from (VCLIP/√2)2/RL to (VCLIP)2/RL, that
is, a factor of 2, or 3 dB in power terms. It can be shown that for
an ideal limiting amplifier, the 1 dB compression point occurs
for an overdrive factor of 2 dB.
For example, if the AD8330 is driving a 150 Ω load and VMAG is
set to 2 V, the peak output is nominally ±4 V (as noted previously,
the actual value, when loaded. can differ because of a mismatch
between on-chip and external resistors), or 2.83 V rms for a sine
wave output that corresponds to a power of 53.3 mW, that is,
17.3 dBm in 150 Ω. Thus, the P1dB level, at 2 dB above
clipping, is 19.3 dBm.
Though not involving power transfer, it is sometimes useful
to state the V1dB, which is the output voltage (unloaded or
loaded) that is 2 dB above clipping for a sine waveform. In the
above example, this voltage is still 2.83 V rms, which can be
expressed as 9.04 dBV (0 dBV corresponds to a 1 V sine wave).
Thus, the V1dB is at 11.04 dBV.
Rev. H | Page 25 of 32
AD8330
Data Sheet
APPLICATIONS INFORMATION
from VMAG to ground. An overrange condition is signaled by a
high state on Pin OR of the AD9214. DFS/GAIN is unconnected
in this example producing an offset-binary output. To provide a
twos complement output, connect it to the REF pin.
The versatility of the AD8330, its very constant ac response over
a wide range of gains, the large signal dynamic range, output
swing, single supply operation, and low power consumption
commend this VGA to a diverse variety of applications. Only a
few can be described here, including the most basic uses and some
unusual ones.
For ADCs running at sampling rates substantially below the
bandwidth of the AD8330, an intervening noise filter is
recommended to limit the noise bandwidth. A one-pole filter
can easily be created with a single differential capacitor between
the OPHI and OPLO outputs. For a corner frequency of fC, the
capacitor must have a value of
ADC DRIVING
The AD8330 is well-suited to drive a high speed converter.
There are many high speed converters available, but to illustrate
the general features, the example in this data sheet uses one of
the least expensive, the AD9214. This is available in three
grades for operation at 65 MHz, 80 MHz, and 105 MHz; the
AD9214BRSZ-80 is a good complement to the general
capabilities of this VGA.
CFILT = 1/942 fC
(26)
For example, a 10 MHz corner requires about 100 pF.
SIMPLE AGC AMPLIFIER
Figure 64 illustrates the use of the inverted gain mode and the
offset gain range (0.2 V < VDBS < 1.7 V) in supporting a low cost
AGC loop. Q1 is used as a detector. When OPHI is sufficiently
higher than CNTR, due to the signal swing, it conducts and
charges C1. This raises VDBS and rapidly lowers the gain. Note
that MODE is grounded (see Figure 48). The minimum voltage
needed across R1 to set up the full gain is 0.2 V because CMGN
is dc open-circuited (this does not alter VMAG) and the maximum
voltage is 1.7 V.
Figure 63 shows the connections to drive an ADC. A 3.3 V
supply is used for both parts. The ADC requires that its input
pins be positioned at one third of the supply, or 1.1 V. Given
that the default output level of the VGA is one-half the supply
or 1.65 V, a small correction is introduced by the 8 kΩ resistor
from CNTR to ground. The ADC specifications require that the
common-mode input be within ±0.2 V of the nominal 1.1 V;
variations of up to ±20% in the AD8330 on-chip resistors change
this voltage by only ±70 mV. With the connections shown in
Figure 63, the AD9214 is able to receive an input of 2 V p-p;
the peak output of the AD8330 can be reduced if desired by
adding a resistor
VS, 3.3V
0.1µF
3.3Ω
0.1µF
OVERRANGE
3.3Ω
CHPF
0.1µF
OFST
VPOS
CNTR
VPSO
CM MODE AND
OFFSET CONTROL
BIAS AND
V-REF
AVDD
OR
DrVDD
PWRDN
D8
DFS/GAIN
INHI
INPUT,
±2V MAX
OPHI
VGA CORE
D7
AIN
D6
OUTPUT
STAGES
D5
AD9214BRS-80
OPLO
INLO
D4
D3
AIN
D2
REFSENSE
NC
MODE
VDBS
GAIN BIAS,
VDBS , 0V–1.5V
GAIN INTERFACE
CMGN
OUTPUT
CONTROL CMOP
COMM
NC
D1
REF
D0
AGND
VMAG
D9
DATA OUTPUTS
ENBL
VPSI
0.1µF
CLK
DGND
0.1µF
ANALOG GROUND
CLOCK
Figure 63. Driving an Analog-to-Digital Converter (Preliminary)
Rev. H | Page 26 of 32
DIGITAL
GROUND
03217-064
10Ω
8kΩ
Data Sheet
AD8330
VS, 2.7V–6V
This simple detector exhibits a temperature variation in the
differential output amplitude of about 4 mV/°C. It provides a
fast attack time (an increase in the input is quickly leveled to the
nominal output, due to the high peak currents in Q1) and a
slow release time (a decrease in the input is not restored as
quickly). The voltage at the VDBS pin can be used as an RSSI
output, scaled 30 mV/dB. Note that the attack time can be
halved by adding a second transistor, labeled Q2 in Figure 64.
For operation at lower frequencies, the AGC hold capacitor
must be increased.
33nF
10Ω
ENBL
VPSI
0.1µF
OFST
VPOS
CM MODE AND
OFFSET CONTROL
BIAS AND
V-REF
OUTPUT
STAGES
VGA CORE
INLO
MODE
VPSO
0.1µF
OPHI
INHI
INPUT,
5mV TO 1V rms
4.7Ω
CNTR
SEE
TEXT
OUTPUT,
~1V rms
OPLO
GAIN INTERFACE
VDBS
CMGN
Q2
OUTPUT
CONTROL CMOP
COMM
WIDE RANGE TRUE RMS VOLTMETER
Q1
VMAG
0.1µF
NC
C1
0.1µF
0.1µF
03217-065
R1
10kΩ
Figure 64. Simple AGC Amplifier (Preliminary)
When the loop is settled, the average current in Q1 is VDBS/R1,
which varies from 20 µA at maximum gain (VDBS = 0.2 V) to
170 µA at minimum gain (VDBS = 1.7 V). This change in the Q1
current causes an increase of ~0.25 dB over the full gain range
in the differential output of nominally 0.75 dBV at midrange
(3.08 V p-p), corresponding to a 200:1 compression ratio. This
is plotted in Figure 65 for a representative 100 kHz input.
1.0
The AD8362 is an rms responding detector providing a dynamic
range of 60 dB from low frequencies to 2.7 GHz. This can increase
to 110 dB using an AD8330 as a preconditioner, provided the
noise bandwidth is limited by an interstage low-pass or bandpass filter.
The VGA also provides an input port that is easier to drive than
the 200 Ω input of the AD8362. Figure 67 shows the general
scheme.
Both the AD8330 and AD8362 provide linear-in-decibel control
interfaces. Thus, when the output of the AD8362 is used to control
the gain of the AD8330, the functional form is unaffected. The
overall scaling is 33 mV/dB. Figure 68 shows the time domain
response using a loop filter capacitor of 10 nF, for inputs ranging
from 10 μV to 1 V rms, that is, a 100 dB measurement range.
1.75
1.50
VDBS
1.25
1.00
0.75
0.8
GAIN ERROR (dB)
LEVELED OUTPUT (dBV)
0.9
0.7
0.6
0.50
0.25
0
3
2
1
0
–1
–30
–20
INPUT TO AD8330 (dBV)
–10
0
–2
OUTPUT
–3
–4
Figure 65. AGC Output vs. Input Amplitude (Simulation)
The upper panel in Figure 66 shows the time-domain output for
fourteen 3 dB steps in input amplitude from 5.4 mV to 1.7 V.
The waveforms in Figure 65 show the AGC voltage (VDBS).
Rev. H | Page 27 of 32
0
10
20
30
40
50
60
70
80
TIME (µs)
90
100 110 120 130 140 150
Figure 66. Time Domain Waveforms (Simulation)
03217-067
–40
03217-066
0.5
–50
AD8330
Data Sheet
5V
3.3Ω
3.3Ω
0.1µF
AD8362
0.1µF
VPOS
VPS1
CNTR
VPSO
INHI
OPHI
CFLT
18nF
AD8330
INPUT
INLO
OPLO
MODE
CMOP
CMGN
COMM
VMAG
2
CHPF
VREF 15
3
DECL
VTGT 14
4
INHI
VPOS 13
5
INLO
VOUT 12
6
DECL
VSET 11
7
PWDN
ACOM 10
8
COMM
CLPF 9
0.1µF
3.6V
VOUT
0.1µF
10µF
6.04kΩ
4.02kΩ
Figure 67. Wide Range True RMS Voltmeter (Preliminary)
4
3
OUTPUT (V)
VDBS
ACOM 16
2
1
0
0
0.4
0.8
1.2
1.6
2.0
2.4 2.8
TIME (ms)
3.2
3.6
4.0
4.4
4.8
Figure 68. Time Domain Response of RMS Voltmeter (Simulation)
Rev. H | Page 28 of 32
03217-068
OFST
COMM
10µF
3.6V
ENBL
1
03217-069
0.1µF
3.3Ω
Data Sheet
AD8330
EVALUATION BOARD
GENERAL DESCRIPTION
BASIC OPERATION
The AD8330-EVALZ is an easy-to-use accessory that enables a
hands-on evaluation of the AD8330 variable gain amplifier
(VGA). It includes test pins for connections to all of the functional
device inputs. Figure 69 is a full size photograph of the board.
The input SMA connector IN is terminated with a 49.9 Ω
resistor (see Figure 70). For convenience, the board includes an
AD8131 high speed differential amplifier to convert a singleended signal source to the differential input of the AD8330. If
desired, the AD8131 can be removed and the AD8330 can be
driven at one of its inputs from a single-ended source.
The AD8330 output is observed at the SMA connectors OUT_HI
and OUT_LO or by using the 2-pin header OUT_HI/OUT_LO
adjacent to the device.
03217-070
The AD8330 requires only a +5 V power supply; however, because
of the AD8131 buffer bipolar power supply requirements, ±5 V
supplies are required to power the board. The current required
for the board is approximately 40 mA from the +5 V supply and
10 mA from the −5 V supply.
Figure 69. Photograph of the AD8330 Evaluation Board
C13
1nF
FILTER_OFFSET
FLTR
OFST
1
+5V
R4
0Ω
C1 +
10µF
10V
C18
0.1µF
2
1
R1
24.9Ω
A1
6
CNTR
VPSI
VPSO
OPHI
INHI
+5V
GAIN_SLOPE
16
15
14
13
AD8330
INLO
OPLO
+5V
C12
0.1µF
C19
0.1µF
+5V
C11
0.1µF
C9
12pF
DUT1
5
C5
0.1µF
12
C8
12pF
UP
6
MODE
CMOP
VDBS
VMAG
CMGN
COMM
CNTR
C4
0.1µF
OUT_HI
OUT_TEST
OUT_ LO
C10
0.1µF
11
DOWN
VDBS
–5V
C17
10µF
10V
ENBL
IN_TEST
5
C16
10nF
+
VPOS
4
C3
0.1µF
AD8131 8 3
4
OFST
C20
0.1µF
7
C6
0.1µF
8
CMGN
Figure 70. Schematic Diagram
Rev. H | Page 29 of 32
10
9
C7
0.1µF
VMAG
03217-071
IN
2
C2
0.1µF 3
C14
10nF
C15
0.1µF
R2
49.9Ω
ENBL
GND1 GND2 GND3 GND4
R3
1kΩ
AD8330
Data Sheet
OPTIONS
MEASUREMENT SETUP
Table 6 lists the jumpers on the board and their functions.
The basic board connections for a typical measurement are
shown in Figure 71. To minimize circuit-loading effects, a low
capacitance FET probe is recommended for observing input or
output waveforms. Two-pin headers, IN_TEST and OUT_TEST,
are provided for this purpose. The SMA connectors OUT_HI
and OUT_LO can also be used, but the user may need to
account for load capacitance effects.
Table 6. Functions of Jumpers
OFST
UP
DOWN
Function
Connects a high-pass filter to the offset control loop
pin. This jumper is normally not installed.
Disables the offset correction loop. This jumper is
installed for dc or low frequency operation.
Mode up. Install for ascending gain with increasing
VDBS gain control voltage.
Mode down. Install for descending gain with
increasing VDBS gain control voltage.
AD8330-EVALZ BOARD DESIGN
The AD8330-EVALZ is a 4-layer design for maximum groundplane area. The evaluation board side silkscreen and wiring
patterns are shown in Figure 72 through Figure 77.
NETWORK
ANALYZER
PROBE
POWER SUPPLY
SIGNAL
INPUT
POWER SUPPLY
DIFFERENTIAL
PROBE
+5V
GND
PRECISION VOLTAGE REFERENCES
(FOR VDBS, VMAG)
Figure 71. Typical Connections
Rev. H | Page 30 of 32
03217-072
Name
FLTR
AD8330
Figure 72. Component Side Silkscreen
03217-076
03217-073
Data Sheet
Figure 73. Component Side Wiring
03217-077
03217-074
Figure 75. Wiring Side Silkscreen
03217-075
03217-078
Figure 76. Wiring Side Pattern
Figure 77. Inner Layer 2
Figure 74. Ground Plane
Rev. H | Page 31 of 32
AD8330
3.10
3.00 SQ
2.90
PIN 1
INDICATOR
0.30
0.25
0.20
0.50
BSC
12
13
16
PIN 1
INDICATOR
1
1.65
1.50 SQ
1.45
EXPOSED
PAD
9
0.50
0.40
0.30
TOP VIEW
0.80
0.75
0.70
5
8
BOTTOM VIEW
0.05 MAX
0.02 NOM
COPLANARITY
0.08
0.20 REF
SEATING
PLANE
4
0.20 MIN
FOR PROPER CONNECTION OF
THE EXPOSED PAD, REFER TO
THE PIN CONFIGURATION AND
FUNCTION DESCRIPTIONS
SECTION OF THIS DATA SHEET.
01-26-2012-A
OUTLINE DIMENSIONS
COMPLIANT TO JEDEC STANDARDS MO-220-WEED-6.
Figure 78. 16-Lead Lead Frame Chip Scale Package [LFCSP]
3 mm × 3 mm Body and 0.75 mm Package Height
(CP-16-27)
Dimensions shown in millimeters
0.197 (5.00)
0.193 (4.90)
0.189 (4.80)
9
1
8
0.244 (6.20)
0.236 (5.99)
0.228 (5.79)
0.010 (0.25)
0.006 (0.15)
0.069 (1.75)
0.053 (1.35)
0.065 (1.65)
0.049 (1.25)
0.010 (0.25)
0.004 (0.10)
COPLANARITY
0.004 (0.10)
0.158 (4.01)
0.154 (3.91)
0.150 (3.81)
0.025 (0.64)
BSC
0.012 (0.30)
0.008 (0.20)
SEATING
PLANE
8°
0°
0.050 (1.27)
0.016 (0.41)
COMPLIANT TO JEDEC STANDARDS MO-137-AB
CONTROLLING DIMENSIONS ARE IN INCHES; MILLIMETER DIMENSIONS
(IN PARENTHESES) ARE ROUNDED-OFF INCH EQUIVALENTS FOR
REFERENCE ONLY AND ARE NOT APPROPRIATE FOR USE IN DESIGN.
0.020 (0.51)
0.010 (0.25)
0.041 (1.04)
REF
09-12-2014-A
16
Figure 79. 16-Lead Shrink Small Outline Package [QSOP]
(RQ-16)
Dimensions shown in inches and (millimeters)
ORDERING GUIDE
Model 1
AD8330ACPZ-R2
AD8330ACPZ-RL
AD8330ACPZ-R7
AD8330ARQZ
AD8330ARQZ-RL
AD8330ARQZ-R7
AD8330-EVALZ
1
Temperature Range
−40°C to +85°C
−40°C to +85°C
−40°C to +85°C
−40°C to +85°C
−40°C to +85°C
−40°C to +85°C
Package Description
16-Lead Lead Frame Chip Scale Package [LFCSP]
16-Lead Lead Frame Chip Scale Package [LFCSP]
16-Lead Lead Frame Chip Scale Package [LFCSP]
16-Lead Shrink Small Outline Package [QSOP]
16-Lead Shrink Small Outline Package [QSOP]
16-Lead Shrink Small Outline Package [QSOP]
Evaluation Board
Z = RoHS Compliant Part.
©2002–2016 Analog Devices, Inc. All rights reserved. Trademarks and
registered trademarks are the property of their respective owners.
D03217-0-5/16(H)
Rev. H | Page 32 of 32
Package Option
CP-16-27
CP-16-27
CP-16-27
RQ-16
RQ-16
RQ-16
Branding
JFZ
JFZ
JFZ