APPLICATION 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 most important nonlinear functions of analog circuit technology. Many signal sources, however, such as CCD sensors, pin diodes, or antennas, deliver weak, oscillating, and simultaneously wide-band signals. But now a new multiplication method is available. Used as a wide-bandwidth Automatic 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). 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. 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 linearity errors, temperature dependence, less than ideal crosstalk, and limited bandwidth, but the multiplication function presented here functions directly and has variable transconductance, enabling it to achieve the largest possible bandwidth. AGC WITH THE DIAMOND TRANSISTOR The voltage-controlled current source of the OPA660 from Burr-Brown has acquired various nicknames according to its applications: Operational Transconductance Amplifier (OTA) Current Conveyor Diamond Transistor Ideal Transistor Macrotransistor Applications for the OPA660 are usually amplifier circuits. But although the OPA660s connection pin, IQ, adjusts functions primarily as a power supply switch or

quiescent current programmer, it can also be used for multiplicative applications. Figure 1 illustrates the dependance of the transconductance (gm = d(IOUT)/d(VIN)) upon the resistance, RQC. The following equation can be derived from the idealized OPA660 model circuits shown in Figure 2. V IQC = T ln (n) RQC When the temperature voltage (VT) is 25.86mV, the quiescent current resistance (RQC) is 250, and the scale factor (n) of the transistor R122 is 10, the cross current IQC can be calculated as follows: IQC = 25.86mV ln(10) = 238A 250

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 238A = 1.74mA
IOUT (mA) 1.5 RQC 250 1.0 500 0.5

20

15

10 0.5

10

15

20 VIN (mV)

1.0

1.5

VOUT = VIN

M RQC VOUT

VIN

RQC

FIGURE 1. Schematic Diagram of the Multiplication Function.

AB-185 Printed in U.S.A. October, 1993

1993 Burr-Brown Corporation

SBOA072

BC

7 (13)

7 (13) IQC'

DB

DT

7 (13)

7.3(x) IQC 81 IQC IQC 5 6 Rgm 8 2

IQC' 31 7.3(x)

IQC'

IQC

3 7.3(x)

10(x) 122 (14) (14) 4 4 RQC 250 IQC

82 IQC' 7.3(x) IQC'

32 IQC' IQC (14) 4

FIGURE 2. Idealized Model Circuit. Now it is easy to determine the transconductance using the following equation: gm = IQC VT = a ln(n) = 67mA/V RQC 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 distortion. i = I1 I2 = IQC' Exp +

The circuit diagram of the actual multiplier circuit as illustrated in Figure 3 makes it easier to determine the multiplication constant, M. The signal current at Pin 8 produces the following output voltage at the resistor ROUT: VOUT = i ROUT = VIN gm ROUT = VIN M RQC

( )

V VT

Exp

( )

V VT

i = IQC' [Exp () Exp (+)] = 2IQC' sinh () = VIN Rgm i VIN + 2IQC' Rgm sinh () V = = 2VT 2VT VT VIN = 2VT 2IQC' Rgm sinh () d(i) = 2IQC' cosh () d()

When the resistor (ROUT) has 2.08k and the input voltage is 10mV, the output voltage reaches the following value: VOUT = a ln(n) ROUT RQC 7.3 ln(10) 2.08k 250 = 1.4V

= 10mV

d(VIN) d()

= 2VT 2IQC' Rgm cosh ()

The multiplication constant M can be derived directly from the equation as follows: M = a ln(n) ROUT = 7.3 ln(10) 2.08k = 35k The gain G can be calculated using the equation: G= d(VOUT) d(VIN) = M 35k = = 140 RQC 250

cosh () = sinh2 () + 1 =

( )
i/IQC' 2 1 Rgm

+1

gm =

d(i) d(VIN)

d(i)/d() d(VIN)/d()

VT IQC' cosh ()

DETERMINING THE DIFFERENTIAL GAIN (DG) Figure 4 shows the circuit part important for the multiplication. 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 2

= Rgm + IQC'

1 VT

( )
i/IQC' 2

+1

IQC' =

a ln(n) VT RQC 1 Rgm + VT/IQC'

The following applies for low modulation: iMAX VINMAX Rgm + VT/IQC'

gm0 =

i=0

In the extreme case in which Rgm = 0, the following results: or for low modulation: Rgm + VT/IQC' Rgm +
2

DG =

gmMAX gm0

1= Rgm +

Rgm + VT/IQC' VT/IQC'

DG

VT/IQC' VIN/IQC'
2

iMAX/IQC' 2

+1

2 (Rgm +VT/IQC')

+1

ROUT 2.08k +5V i 7 5 8 3 VOUT 1.4Vp0

a ln(n) Rgm/ (RQC + 1) a ln(n) Rgm/RQC + 1 VIN/VT 2 (a ln(n) Rgm/ (RQC + 1)

2

+1

VIN 10mVp0

DB OPA660 DT

DG0 =

( (

iMAX/IQC' 2

)
2

+1 1

Rgm 1m

DG0
RQC 250

VINMAX 2VT

+1 1

5V

FIGURE 3. Multiplier Circuit.

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). As is evident, Rgm produces transfer linearization, but it also reduces the gain, GRgm.

IQC'

IQC'

GRgm =

d(VOUT) d(VIN) ROUT

ROUT Rgm +VT/IQC'

VIN

I1 I2

Rgm

I2 I1

= Rgm +

RQC a ln(n)

i=0

IQC'

IQC'

FIGURE 4. Multiplier Section.

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.

10

DG max

DG max

0 RQC = 250

10 RQC = 500 3

3 10

10 20

Rgm ()

1 20 0.3 30 40 0.1 20 10 0 10 VIN (mVpo) 20

30 40

0.3

50

0.1 20 10 0 10 VIN (mVpo) 20

FIGURE 5. Differential Gain Error (RQC = 250).

FIGURE 6. Differential Gain Error (RQC = 500).

10

DG max

RQC = 1k 3

DG max

0 10 20 30 40 50

10 RQC = 2k 3

0 10 20 30 40 50

Rgm ()

0.3

0.3

0.1 20 10 0 10 VIN (mVpo) 20

0.1 20 10 0 10 VIN (mVpo) 20

FIGURE 7. Differential Gain Error (RQC = 1k). When Rgm is inserted, the relation between the gain, GRgm, and the control value, 1/RQC, becomes disproportionate. AUTOMATIC GAIN CONTROL (AGC) Circuit tolerances and insufficient temperature compensation 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. 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 (4k/56) to about 10mV. The 4k 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 4

FIGURE 8. Differential Gain Error (RQC = 2k).

Reference for VOUT +

Automatic Gain Control VIN Multiplier Amplifier VOUT

Level Control

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.

Rgm ()

(%)

(%)

Rgm ()

(%)

(%)

+5V 3 2 7 OPA621 4 ROUT 2.08k +5V 10mVp-p i 7 OPA660 4k 22k 8 +5V 5V 6

VOUT 1.4Vp-p

10k Offset 5V

5 56

DB

DT

VIN 0.7Vp-p

4 5V

Rgm 1m

56

RQC 250

Peak Detector and Comparator 1.4V 5V 5V +5V Reference for VOUT

FIGURE 10. AGC Amplifier for Various Signals.

+5V To Multiplier Pin 1 2.7k +1.6V Reference for VOUT +0.4V 0.47F 100 1k 330 2811 2.2M RQC 2N5460 2* BC577 100 2811 2.2M 0.47F 1M 5V VOUT 1.4Vp-p

+5V 100k 0.47F

FIGURE 11. Peak Level. 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. DETERMINING THE MAXIMUM DIFFERENTIAL GAIN (DGMAX) OF AGC AMPLIFIERS 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 R QCMAX. The largest signal distortions measured as 5
+5V VOUT 3 2 4 10k 5V Hk 4Vp-p 5V 47k 0.1F 2N3904 1N4148 7 CA3080 5 C hold 1F 6 To Multiplier Pin 1

FIGURE 12. Clamp Circuit for TV Signals.

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: q= VINMAX VINMIN

RQCMAX = q RQCMIN + a ln(n) Rgm (q 1) B = a ln(n) Rgm/RQCMAX

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 results. The measurements, for example, also include distortion from the subsequent amplifier OPA621. Figures 13 to 15 give an overview of the achievable distortion. For maximum input voltages (VINMAX) from 10mV to 20mV and open-loop resistances from 0 to 50, the differential gain shown in simulations is a function of the ratio VINMAX/VINMIN and equals 9. Figure 16 presents measured achievable distortions in the AGC structure, as already shown in Figure 10. THERMALLY INDUCED DISTORTION 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.

From these equations, it is possible to derive the maximum distortion, DGMAX, as a function of B and the maximum input voltage.

B+1 DGMAX = B+ 1 VINMAX/VT 2 (B + 1)

2

+1

As Figure 17 shows, Rgm can reduce this thermally induced pulse distortion.

10 DGMAX 3.0 0 10 20 30 40 50 VINMAX/VINMIN 0.1 1 2 3 4 5 6 7 8

Rgm ()

10 VINMAX = 10mVp0 3.0 DGMAX VINMAX = 15mVp0 0 10 20 30 40 50

1.0

1.0

0.3

0.3 VINMAX/VINMIN 0.1 1 2 3 4 5 6 7 8

FIGURE 13. DGMAX of the AGC Amplifier (Simulation) (VINMAX = 10mV).

FIGURE 14. DGMAX of the AGC Amplifier (Simulation) (VINMAX = 15mV).

10 DGMAX 3.0 VINMAX = 20mVp0 0 10 20 30 40 50

10 DGMAX 3.0
Rgm ()

VINMAX = 20mVp0 0 10 20 30 40 50

1.0

1.0

0.3 VINMAX/VINMIN 0.1 1 2 3 4 5 6 7 8

0.3 VINMAX/VINMIN 0.1 1 2 3 4 5 6 7 8

FIGURE 15. DGMAX of the AGC Amplifier (Simulation) (VINMAX = 20mV). 6

FIGURE 16. DGMAX of the AGC Amplifier (Measurement) (VINMAX = 20mV).

Rgm ()

(%)

(%)

Rgm ()

(%)

(%)

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 balanced temperature distribution on the chip. DEMO BOARD All available measurements were conducted using the completely 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 compensation, 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 PREF. The transistor 2N5460 changes the quiescent current according to the difference, thus varying the transconductance gm. 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.

10 8 6 10 4 20 30 40 50 DGMAX VINMAX = 20mVp0 0

2 VINMAX/VINMIN 1 1 2 3 4 5 6 7 8

FIGURE 17. Effect of Rgm on Thermal Pulse Distortion.

64 60 56 VINMAX = 20mVp0

S/N (dB)

52 48 44

VINMAX/VINMIN 1 2 3 4 5 6 7 8

0 10 20 30 40 50

40

FIGURE 18. S/N of AGC Amplifiers.

Rgm ()

Rgm ()

(%)

+5V Amplfier R6 100 10nF 3 2 7 OPA621 4 2.2F 10nF 5V 9 R7 VOUT 20k 2.2F VOUT 2.8Vp-p

R9 51

Level-Shifting POFFSET +5V 5V R4 22k R5 56 C5 +470F R8 10k 10nF 2.2F

0.2 ... 0.8Vp-p

+5V

2.2F i 10nF 1 S1 2 3

Manual

13mV Automatic

VIN 7 RIN 2k 5 R1 56 10nF OPA660 DB DT 3 8

4 Clamped R3 56

+5V

7 CA3080 3 2 1 CHOLD C4 R 20 1F 10 6 5

R11 1k R22 1k

R2 51 Rgm

2.2F

10nF 5V 0.1F

2.2F

R21 47k

HK

Multiplier

5V

2N3904 R23 560k 5V 5V C6 0.1F

4Vp-p

+5V +5V 220F R16 2.7k +1.6V PREF VREF 1k +0.4V R17 330 D1 2811 R14 2.2M C2 + 0.47F R15 100 2N5460 2* BC577 R12 2.2M 220F R18 100k R13 100 C 3 + 0.47F Peak Level-Control R10

D2 2811

RQC

C1 4.7F

R19 1M

5V

FIGURE 19. Circuit Diagram of the AGC Amplifier Demo Board.

FIGURE 20. Layout of the AGC Amplifier Circuit Board Back.

FIGURE 21. Layout of the AGC Amplifier Circuit Board Front. 9

PARTS LIST
NO. 1 2 3 4 5 6 7 8 9 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 C2 , C3 C5 C6 C4 DESIGNATION IC1 IC2 IC3 T1, T2 T3 T4 D1 , D2 D3 R1, R3, R5 RIN Rgm R6, R13, R15 R7 R4 R8, R20 R11, R22 R23 R21 R10 R12, R14 R18 R17 R16 R19 2.2M 100k 330 2.7k 1M Capacitor 2.2F Capacitor 10nF Capacitor 220F Capacitor 0.47F Capacitor 470F Capacitor 0.1F Capacitor 1F PREF POT 1k POFFSET POT 10k VIN, VOUT, HK SMA POS, GND, NEG Mini-Banana PART NAME/VALUE OPA621KP CA3080 OPA660AP BC577 2N5460 2N3904 2N2811 IN4148 56 2k 51 100 20k 22k 10k 1k 560k 47k NUMBER OF PARTS 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

10

IMPORTANT NOTICE Texas Instruments and its subsidiaries (TI) reserve the right to make changes to their products or to discontinue any product or service without notice, and advise customers to obtain the latest version of relevant information to verify, before placing orders, that information being relied on is current and complete. All products are sold subject to the terms and conditions of sale supplied at the time of order acknowledgment, including those pertaining to warranty, patent infringement, and limitation of liability. TI warrants performance of its semiconductor products to the specifications applicable at the time of sale in accordance with TIs standard warranty. Testing and other quality control techniques are utilized to the extent TI deems necessary to support this warranty. Specific testing of all parameters of each device is not necessarily performed, except those mandated by government requirements. Customers are responsible for their applications using TI components. In order to minimize risks associated with the customers applications, adequate design and operating safeguards must be provided by the customer to minimize inherent or procedural hazards. TI assumes no liability for applications assistance or customer product design. TI does not warrant or represent that any license, either express or implied, is granted under any patent right, copyright, mask work right, or other intellectual property right of TI covering or relating to any combination, machine, or process in which such semiconductor products or services might be or are used. TIs publication of information regarding any third partys products or services does not constitute TIs approval, warranty or endorsement thereof.

Copyright 2000, Texas Instruments Incorporated