AB-185 [ETC]
AB-185 - AUTOMATIC GAIN CONTROL (AGC) USING THE DIAMOND TRANSISTOR OPA660 ; AB - 185 - 自动增益控制( AGC )使用金刚石晶体管OPA660\n![AB-185](http://pdffile.icpdf.com/pdf1/p00015/img/icpdf/AB-18_70527_icpdf.jpg)
型号: | AB-185 |
厂家: | ![]() |
描述: | AB-185 - AUTOMATIC GAIN CONTROL (AGC) USING THE DIAMOND TRANSISTOR OPA660
|
文件: | 总10页 (文件大小:164K) |
中文: | 中文翻译 | 下载: | 下载PDF数据表文档文件 |
®
AP P LICATION BULLETIN
Mailing Address: PO Box 11400 • Tucson, AZ 85734 • Street Address: 6730 S. Tucson Blvd. • Tucson, AZ 85706
Tel: (602) 746-1111 • Twx: 910-952-111 • Telex: 066-6491 • FAX (602) 889-1510 • Immediate Product Info: (800) 548-6132
AUTOMATIC GAIN CONTROL (AGC)
USING THE DIAMOND TRANSISTOR OPA660
By Christian Henn, Burr-Brown International GmbH
Multiplication of analog signals has long been one of the
quiescent current programmer, it can also be used for mul-
tiplicative applications.
most important nonlinear functions of analog circuit tech-
nology. Many signal sources, however, such as CCD sen-
sors, pin diodes, or antennas, deliver weak, oscillating, and
simultaneously wide-band signals. But now a new multipli-
cation method is available. Used as a wide-bandwidth Auto-
matic Gain Control (AGC) application circuit, the integrated
circuit OPA660 varies its own gain to change the signal
amplitude and keep the output signal constant over a wide
input voltage range. The OPA660 thus makes it possible to
control and amplify signals with no additional multiplier.
Important parameters include the differential gain (DG), the
thermally induced pulse distortion, and the signal-to-noise
ratio (S/N).
Figure 1 illustrates the dependance of the transconductance
(gm = d(IOUT)/d(VIN)) upon the resistance, RQC. The follow-
ing equation can be derived from the idealized OPA660
model circuits shown in Figure 2.
VT
l
n (n)
IQC
=
RQC
When the temperature voltage (VT) is 25.86mV, the quies-
cent current resistance (RQC) is 250Ω, and the scale factor (n)
of the transistor R122 is 10, the cross current IQC can be
calculated as follows:
25.86mV
IQC
=
ln(10) = 238µA
An analog multiplier delivers an output signal (voltage or
current) that is proportional to the product of two or more
inputs. The application circuit presented here is concerned
primarily with two inputs. In the simplest case, each of the
two inputs can function with both polarities. In this case, the
input voltage swing covers all four quadrants; that is, there
are four polarity combinations. In contrast to a four quadrant
multiplier, a two quadrant multiplier allows only one input
to be connected to a signal of any polarity. The second input
can only process unipolar signals.
250Ω
The quiescent current of the subsequent transistor stages can
be calculated with a scale factor (a) of 7.3 for transistors 31,
32, 81, and 82 to
IQC' = a • IQC = 7.3 • 238µA = 1.74mA
IOUT (mA)
RQC
1.5
Multipliers are nonlinear and thus can not be implemented
as simply and exactly as linear components. In developing
the circuit, various design methods were used depending
upon the accuracy, bandwidth, and justifiable complexity.
Multipliers do have several disadvantages, including linear-
ity errors, temperature dependence, less than ideal crosstalk,
and limited bandwidth, but the multiplication function pre-
sented here functions directly and has variable
transconductance, enabling it to achieve the largest possible
bandwidth.
250
1.0
500
0.5
–20
–15
–10
10
15
20
VIN (mV)
–0.5
–1.0
–1.5
AGC WITH THE DIAMOND TRANSISTOR
The voltage-controlled current source of the OPA660 from
Burr-Brown has acquired various nicknames according to its
applications:
M
RQC
VOUT = VIN
Operational Transconductance Amplifier (OTA)
Current Conveyor
VIN
VOUT
M
Diamond Transistor
Ideal Transistor
Macrotransistor
RQC
Applications for the OPA660 are usually amplifier
circuits. But although the OPA660’s connection pin,
IQ, adjusts functions primarily as a power supply switch or
FIGURE 1. Schematic Diagram of the Multiplication
Function.
©1993 Burr-Brown Corporation
AB-185
Printed in U.S.A. October, 1993
BC
DB
DT
7
7
7
(13)
(13)
(13)
IQC'
7.3(x)
IQC'
IQC'
IQC
IQC
81
31
8
2
7.3(x)
7.3(x)
Rgm
IQC
IQC
6
5
3
82
32
10(x)
(14)
4
IQC'
IQC'
IQC
IQC
122
7.3(x)
(14)
4
(14)
4
IQC'
RQC
250Ω
FIGURE 2. Idealized Model Circuit.
Now it is easy to determine the transconductance using the
following equation:
gm, since it is dependent upon the modulation. This change
results in turn in signal distortion. The following equations
derive the relation between the signal amplitude and distor-
tion.
IQC
VT
a • ln(n)
RQC
gm =
=
= 67mA/V
∆V
∆V
The circuit diagram of the actual multiplier circuit as illus-
trated in Figure 3 makes it easier to determine the multipli-
cation constant, M. The signal current at Pin 8 produces the
i = I1 – I2 = IQC' Exp
+
– Exp –
( V ) ( )
VT
T
i = –IQC' [Exp (–ϕ) – Exp (+ϕ)] = –2IQC' • sinh (ϕ)
following output voltage at the resistor ROUT
:
M
VIN – Rgm • i
2VT
VIN + 2IQC' Rgm • sinh (ϕ)
∆V
VOUT = i • ROUT = VIN • gm • ROUT = VIN
•
ϕ =
=
=
RQC
VT
2VT
When the resistor (ROUT) has 2.08kΩ and the input voltage
is ±10mV, the output voltage reaches the following value:
VIN = 2VT • ϕ –2IQC' Rgm • sinh (ϕ)
a • ln(n) • ROUT
d(i)
VOUT
=
= –2IQC' • cosh (ϕ)
RQC
d(ϕ)
7.3 • ln(10) • 2.08kΩ
250Ω
= ±10mV
= ±1.4V
d(VIN)
= –2VT –2IQC' Rgm • cosh (ϕ)
d(ϕ)
The multiplication constant M can be derived directly from
the equation as follows:
2
i/IQC'
cosh (ϕ) = sinh2 (ϕ) + 1 =
+ 1
( )
2
M = a • ln(n) • ROUT = 7.3 • ln(10) • 2.08kΩ = 35kΩ
The gain G can be calculated using the equation:
d(i)
d(i)/d(ϕ)
1
d(VOUT
)
M
35kΩ
gm =
=
=
G =
=
=
= 140
d(VIN)
d(VIN)/d(ϕ)
VT
d(VIN)
RQC 250Ω
Rgm
–
IQC' cosh (ϕ)
DETERMINING THE
DIFFERENTIAL GAIN (DG)
1
=
VT
Figure 4 shows the circuit part important for the multiplica-
tion. When VIN = 0, i = 0, and I1 = I2 = IQC’, i increases with
rising VIN, resulting in variation of the currents I1 and I2. The
increase in both currents also changes the transconductance
Rgm
+
2
i/IQC'
IQC'
+ 1
( )
2
2
The following applies for low modulation:
VINMAX
a • ln(n) •VT
RQC
IQC' =
iMAX
≈
Rgm + VT/IQC'
In the extreme case in which Rgm = 0, the following results:
or for low modulation:
1
gm0 =
i = 0
Rgm + VT/IQC'
R
gm + VT/IQC'
VT/IQC'
gmMAX
Rgm + VT/IQC'
VT/IQC'
DG ≈
DG =
– 1 =
Rgm
+
gm0
Rgm
+
2
VIN/IQC'
2
+ 1
iMAX/IQC'
+ 1
2 (Rgm +VT/IQC')
( )
2
a • ln(n) Rgm/ (RQC + 1)
1
– 1
≈
a ln(n) Rgm/RQC
+
ROUT
2.08kΩ
2
VIN/VT
2 (a • ln(n) Rgm/ (RQC + 1)
+ 1
+5V
VOUT
±1.4Vp0
i
7
8
i
MAX/IQC'
( )
2
5
3
DG0 =
+1 – 1
+1 – 1
VIN
±10mVp0
DB
6
DT
OPA660
4
2
1
Rgm
1mΩ
2
VINMAX
( )
2VT
DG0 ≈
RQC
250Ω
Figures 5 through 8 show an analysis of the equation
DG = f (VIN; Rgm; RQC), which determines the differential
gain error dependent upon the input voltage. The figures
include the open-loop gain resistance (Rgm) and quiescent
current resistance (RQC).
–5V
FIGURE 3. Multiplier Circuit.
As is evident, Rgm produces transfer linearization, but it also
reduces the gain, GRgm
.
d(VOUT
)
ROUT
IQC'
IQC'
GRgm
=
=
Rgm +VT/IQC'
d(VIN)
Rgm
I1
I2
I2
I1
i
i
∆V
∆V
ROUT
RQC
a • ln(n)
VIN
=
i = 0
Rgm
+
IQC'
IQC'
As will be shown later, the gain reduction results in a poorer
signal-to-noise ratio (S/N). Designers can determine the best
performance compromise for DG and S/N by choosing
appropriate values for VINMAX and Rgm. However, the larger
the control range —that is, the greater the variation of RQC
—the poorer the quality of the compromise that can be
attained.
FIGURE 4. Multiplier Section.
3
10
3
10
3
0
0
RQC = 500Ω
RQC = 250Ω
10
20
10
20
1
1
30
40
50
0.3
0.1
0.3
0.1
30
40
–20
–10
0
10
20
–20
–10
0
10
20
VIN (mVpo)
VIN (mVpo)
FIGURE 5. Differential Gain Error (RQC = 250Ω).
FIGURE 6. Differential Gain Error (RQC = 500Ω).
10
10
0
0
10
20
30
40
50
10
20
RQC = 1kΩ
RQC = 2kΩ
3
3
30
40
50
1
1
0.3
0.1
0.3
0.1
–20
–10
0
10
20
–20
–10
0
10
20
VIN (mVpo)
VIN (mVpo)
FIGURE 8. Differential Gain Error (RQC = 2kΩ).
FIGURE 7. Differential Gain Error (RQC = 1kΩ).
When Rgm is inserted, the relation between the gain, GRgm
and the control value, 1/RQC, becomes disproportionate.
,
Reference for VOUT
–
+
AUTOMATIC GAIN CONTROL (AGC)
Automatic Gain Control
Amplifier
Circuit tolerances and insufficient temperature compensa-
tion result in undefined gains (GRgm = f(RQC)) of about ±25%.
If RQC is implemented by a FET, this undefined gain range
increases even more. These problems can be avoided by
using an AGC circuit as shown in Figure 9.
VIN
VOUT
Multiplier
Level Control
In the detailed circuit in Figure 10, the ±0.7V input signal
(VIN), which is assumed for now as a constant, is divided by
the input divider (4kΩ/56Ω) to about ±10mV. The 4kΩ
resistor in front of the circuit can, of course, be removed if
the input amplitude is only in the mV range, as is the case
in fiber optic transmission receivers. The amplifier (OPA621)
placed after the circuit converts the output current i of the
multiplier (OPA660) into voltage. The peak detector and
comparator compare the ±1.4V output signal (VOUT) with the
given reference value +1.4V and connect the control voltage
to the FET. This control ensures that the peak value of VOUT
is identical to the adjustable reference DC voltage and is
FIGURE 9. AGC Circuit (Schematic).
unaffected by circuit tolerances. It is also possible to control
the output voltage against the black level or synchronization
level by acquiring the output voltage for comparison only
during the horizontal sync time. While the luminance signal
changes over time, the sync level is always transmitted with
constant amplitude. Such regulation enables the video signal
to be transmitted at a constant amplitude despite changes in
the luminance signal.
4
VOUT
±1.4Vp-p
+5V
3
2
7
OPA621
4
6
–5V
ROUT
2.08kΩ
+5V
7
i
±10mVp-p
8
+5V
–5V
OPA660
10kΩ
Offset
4kΩ
22kΩ
5
3
DB
6
DT
2
VIN
56Ω
56Ω
4
1
±0.7Vp-p
Rgm
1mΩ
–5V
RQC
250Ω
Peak Detector
and Comparator
1.4V
–5V
+5V
–5V
Reference for VOUT
FIGURE 10. AGC Amplifier for Various Signals.
+5V
To Multiplier
Pin 1
+5V
100kΩ
2.7kΩ
0.47µF
+1.6V
+0.4V
0.47µF
Reference
for
100Ω
100Ω
VOUT
2*
BC577
VOUT
±1.4Vp-p
1kΩ
2811
2811
RQC
2N5460
330Ω
2.2MΩ
2.2MΩ
+
0.47µF
1MΩ
–5V
FIGURE 11. Peak Level.
To Multiplier
Pin 1
+5V
7
Variations in the input signal amplitude cause the control
system to produce constant output signal amplitudes
corresponding to the reference value. Simultaneous changes
in VIN and the reference value are also possible.
CA3080
3
2
VOUT
6
∞
5
C hold
1µF
4
10kΩ
DETERMINING THE MAXIMUM DIFFERENTIAL
GAIN (DGMAX) OF AGC AMPLIFIERS
–5V
47kΩ 0.1µF
Hk
2N3904
The input voltage of AGC amplifiers varies from VINMIN to
VINMAX. To maintain a constant output voltage (VOUT) over
this range, the control voltage from the peak level control
varies the resistance RQC correspondingly from RQCMIN to
4Vp-p
1N4148
–5V
R
QCMAX. The largest signal distortions measured as
FIGURE 12. Clamp Circuit for TV Signals.
5
differential gain (DGMAX) happen at VINMAX or RQCMAX, thus
during operation of the OPA660 with the smallest quiescent
current IQ. For the control range q of the AGC amplifier, the
following conditions apply:
It should be kept in mind, however, that this equation is
based upon the simplified model shown in Figure 2 and
sometimes deviates from measurements and simulation re-
sults. The measurements, for example, also include distor-
tion from the subsequent amplifier OPA621. Figures 13 to
15 give an overview of the achievable distortion. For maxi-
mum input voltages (VINMAX) from ±10mV to ±20mV and
open-loop resistances from 0Ω to 50Ω, the differential gain
shown in simulations is a function of the ratio VINMAX/VINMIN
and equals 9. Figure 16 presents measured achievable distor-
tions in the AGC structure, as already shown in Figure 10.
VINMAX
q =
VINMIN
RQCMAX = q • RQCMIN + a • ln(n) • Rgm • (q – 1)
B = a • ln(n) • Rgm/RQCMAX
THERMALLY INDUCED DISTORTION
From these equations, it is possible to derive the maximum
distortion, DGMAX, as a function of B and the maximum
input voltage.
As shown in Figure 2, the power consumption of transistors
31, 32, 81, and 82 varies according to the signal curve. This
variation leads to temperature oscillation and finally to
change in the transconductance gm.
At first glance, it looks as if the pulse distortion is caused by
RC parts. The visible thermal time constant, however, is in
the microsecond range and is negatively affected by unequal
temperature distribution on the chip.
B + 1
DGMAX
=
1
B +
2
VINMAX/VT
+ 1
As Figure 17 shows, Rgm can reduce this thermally induced
pulse distortion.
2 (B + 1)
10
3.0
1.0
0.3
10
DGMAX
VINMAX = 10mVp0
DGMAX
VINMAX = 15mVp0
0
3.0
1.0
0.3
10
20
30
40
50
0
10
20
30
40
50
VINMAX/VINMIN
VINMAX/VINMIN
0.1
0.1
1
2
3
4
5
6
7 8
1
2
3
4
5
6
7 8
FIGURE 13. DGMAX of the AGC Amplifier (Simulation)
FIGURE 14. DGMAX of the AGC Amplifier (Simulation)
(VINMAX = ±10mV).
(VINMAX = ±15mV).
10
10
0
DGMAX
DGMAX
VINMAX = 20mVp0
VINMAX = 20mVp0
10
20
30
40
50
0
3.0
1.0
0.3
3.0
10
20
30
1.0
40
50
0.3
VINMAX/VINMIN
VINMAX/VINMIN
0.1
0.1
1
2
3
4
5
6
7
8
1
2
3
4
5
6
7 8
FIGURE 15. DGMAX of the AGC Amplifier (Simulation)
FIGURE 16. DGMAX of the AGC Amplifier (Measurement)
(VINMAX = ±20mV).
(VINMAX = ±20mV).
6
In contrast, periodic RF signals less than 1MHz are barely
affected by the pulse distortion. The temperature change can
no longer follow the signal change, resulting in more bal-
anced temperature distribution on the chip.
10
8
VINMAX = 20mVp0
0
DGMAX
6
4
10
20
DEMO BOARD
30
40
50
All available measurements were conducted using the com-
pletely dimensioned circuit shown in Figure 19. The demo
board designed for this application contains four circuit
blocks. As a differential amplifier with current output, the
OPA660 allows users to control the transconductance by
varying the total quiescent current. Functioning mainly as a
multiplier, it also enables a shift in DC position of the output
voltage by varying the noninverting OPA660 input. The
OPA621 functions as a current-to-voltage converter and
amplifies the signal. The switch, S1, in the shift block lets
the user choose between manual, and automatic offset com-
pensation, and clamped DC restoration. At active LOW, the
clamp pulse triggers the OTA module CA3080, checks the
output voltage (VOUT) against the reference value for the
black level voltage, and stores the correction voltage up to
the next clamp pulse (HK) in the capacitor CHOLD. The fourth
block is the already mentioned peak level control circuit.
The discrete differential amplifier checks the peak value of
the output voltage (VOUT) against the reference voltage set by
2
1
VINMAX/VINMIN
1
2
3
4
5
6
7 8
FIGURE 17. Effect of Rgm on Thermal Pulse Distortion.
64
VINMAX = 20mVp0
60
56
0
52
48
44
10
20
30
40
50
PREF. The transistor 2N5460 changes the quiescent current
VINMAX/VINMIN
according to the difference, thus varying the transconductance
gm.
40
1
2
3
4
5
6
7 8
For applications requiring frequencies of more than 80MHz
and a controlled output voltage (VOUT) of more than ±1V, we
recommend two-stage gain using two OPA621s. With the
amplifiers OPA622 and OPA623, it will be possible to
increase the bandwidth even more.
FIGURE 18. S/N of AGC Amplifiers.
7
+5V
Amplfier
R6
100Ω
10nF
2.2µF
VOUT
2.8Vp-p
R9
51Ω
7
3
2
9
OPA621
R7
VOUT
20kΩ
4
Level-Shifting
+5V
POFFSET
–5V
2.2µF
10nF
–5V
R4
R5
56Ω
22kΩ
0.2 ... 0.8Vp-p
+5V
2.2µF
10nF
Manual
±13mV
Automatic
–
C5
470µF
i
2
1
+
VIN
3
S1
R8
10kΩ
Clamped
+5V
4
7
8
R3
56Ω
10nF
2.2µF
RIN
2kΩ
OPA660
5
3
DB
DT
R11
1kΩ
7
CA3080
3
2
R1
6
56Ω
R22
1kΩ
4
6
2
1
R2
51Ω
CHOLD
C4
1µF
5
10nF
R20
10Ω
Rgm
2.2µF
10nF
2.2µF
C6
R21
47kΩ
–5V
0.1µF
2N3904
HK
Multiplier
–5V
R23
560kΩ
0.1µF
4Vp-p
–5V
–5V
+5V
Peak Level-Control
+5V
220µF
R10
R18
100kΩ
R16
2.7kΩ
+1.6V
–
+
–
C2
0.47µF
C3
0.47µF
R15
100Ω
R13
100Ω
PREF
D2
D1
2811
+
2*
BC577
VREF
1kΩ
2811
RQC
2N5460
+0.4V
R17
R14
R12
+
–
330Ω
2.2MΩ
2.2MΩ
C1
4.7µF
R19
1MΩ
220µF
–5V
FIGURE 19. Circuit Diagram of the AGC Amplifier Demo Board.
8
FIGURE 20. Layout of the AGC Amplifier Circuit Board — Back.
FIGURE 21. Layout of the AGC Amplifier Circuit Board — Front.
9
PARTS LIST
NUMBER
NO. DESIGNATION
PART NAME/VALUE
OF PARTS
1
IC1
IC2
OPA621KP
CA3080
OPA660AP
BC577
2N5460
2N3904
2N2811
IN4148
56Ω
1
1
1
2
1
1
2
1
3
1
1
3
1
1
2
2
1
1
1
2
1
1
1
1
6
6
2
2
1
1
1
1
1
3
3
2
3
IC3
4
T1, T2
T3
5
6
T4
7
D1, D2
D3
8
9
R1, R3, R5
RIN
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
2kΩ
Rgm
51Ω
R6, R13, R15
R7
100Ω
20kΩ
R4
22kΩ
R8, R20
R11, R22
R23
10kΩ
1kΩ
560kΩ
47kΩ
R21
R10
R12, R14
R18
2.2MΩ
100kΩ
R17
330Ω
R16
2.7kΩ
R19
1MΩ
Capacitor 2.2µF
Capacitor 10nF
Capacitor 220µF
Capacitor 0.47µF
Capacitor 470µF
Capacitor 0.1µF
Capacitor 1µF
PREF POT 1kΩ
POFFSET POT 10kΩ
VIN, VOUT, HK SMA
POS, GND, NEG Mini-Banana
C2, C3
C5
C6
C4
10
相关型号:
©2020 ICPDF网 联系我们和版权申明