AD811

High Performance Video Op Amp AD811 FEATURES High speed 140 MHz bandwidth (3 dB, G = +1) 120 MHz bandwidth (3 dB, G = +2) 35 MHz bandwidth (0.1 dB, G = +2) 2500 V/µs slew rate 25 ns settling time to 0.1% (for a 2 V step) 65 ns settling time to 0.01% (for a 10 V step) Excellent video performance (RL =150 Ω) 0.01% differential gain, 0.01° differential phase Voltage noise of 1.9 nV/√Hz Low distortion: THD = −74 dB @ 10 MHz Excellent dc precision: 3 mV max input offset voltage Flexible operation Specified for ±5 V and ±15 V operation ±2.3 V output swing into a 75 Ω load (VS = ±5 V) CONNECTION DIAGRAMS NC 1 –IN 2 +IN 3 –VS 4 8 NC 7 +VS 6 OUTPUT AD811 5 NC NC = NO CONNECT Figure 1. 8-Lead Plastic (N-8), CERDIP (Q-8), SOIC (R-8) NC 1 NC 2 –IN 3 16 15 14 13 12 11 NC NC +VS NC OUTPUT NC NC NC 00866-E-002 NC 4 +IN 5 NC 6 –VS 7 NC 8 AD811 10 9 APPLICATIONS Video crosspoint switchers, multimedia broadcast systems HDTV compatible systems Video line drivers, distribution amplifiers ADC/DAC buffers DC restoration circuits Medical Ultrasound PET Gamma Counter applications NC = NO CONNECT Figure 2. 16-Lead SOIC (R-16) NC NC NC NC NC 3 2 1 20 19 NC NC –IN NC +IN 4 5 6 7 8 9 10 11 12 13 AD811 17 NC NC 16 +VS 15 NC 14 OUTPUT 18 00866-E-003 NC = NO CONNECT GENERAL DESCRIPTION A wideband current feedback operational amplifier, the AD811 is optimized for broadcast-quality video systems. The −3 dB bandwidth of 120 MHz at a gain of +2 and the differential gain and phase of 0.01% and 0.01° (RL = 150 Ω) make the AD811 an excellent choice for all video systems. The AD811 is designed to meet a stringent 0.1 dB gain flatness specification to a bandwidth of 35 MHz (G = +2) in addition to low differential gain and phase errors. This performance is achieved whether driving one or two back-terminated 75 Ω cables, with a low power supply current of 16.5 mA. Furthermore, the AD811 is specified over a power supply range of ±4.5 V to ±18 V. (Continued on page 3) Figure 3. 20-Terminal LCC (E-20A) NC 1 NC 2 NC 3 –IN 4 NC 5 +IN 6 NC 7 –VS 8 20 NC 19 NC 18 NC 17 +VS 16 NC 15 OUTPUT 14 NC 13 NC NC 9 NC 10 –VS NC NC NC NC 11 NC NC = NO CONNECT Figure 4. 20-Lead SOIC (R-20) Rev. E 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 www.analog.com Fax: 781.326.8703 © 2004 Analog Devices, Inc. All rights reserved. 00866-E-004 AD811 12 NC 00866-E-001 AD811 TABLE OF CONTENTS Specifications..................................................................................... 4 Absolute Maximum Ratings............................................................ 6 Maximum Power Dissipation ..................................................... 6 Metalization Photograph ............................................................. 6 Typical Performance Characteristics ............................................. 7 Applications ..................................................................................... 12 General Design Considerations................................................ 12 Achieving the Flattest Gain Response at High Frequency .... 12 Operation as a Video Line Driver ............................................ 14 An 80 MHz Voltage-Controlled Amplifier Circuit ................ 15 A Video Keyer Circuit................................................................ 16 Outline Dimensions ....................................................................... 18 Ordering Guide .......................................................................... 20 REVISION HISTORY 7/04—Data Sheet Changed from Rev. D to Rev. E Updated Format............................................................. Universal Change to Maximum Power Dissipation Section .................... 7 Changes to Ordering Guide ...................................................... 20 Updated Outline Dimensions ................................................... 20 Rev. E | Page 2 of 20 AD811 GENERAL DESCRIPTION (continued) The AD811 is also excellent for pulsed applications where transient response is critical. It can achieve a maximum slew rate of greater than 2500 V/µs with a settling time of less than 25 ns to 0.1% on a 2 V step and 65 ns to 0.01% on a 10 V step. The AD811 is ideal as an ADC or DAC buffer in data acquisition systems due to its low distortion up to 10 MHz and its wide unity gain bandwidth. Because the AD811 is a current feedback amplifier, this bandwidth can be maintained over a wide range of gains. The AD811 also offers low voltage and current noise of 1.9 nV/√Hz and 20 pA/√Hz, respectively, and excellent dc accuracy for wide dynamic range applications. 0.10 0.09 0.08 0.20 RF = 649Ω 0.18 FC = 3.58MHz 100 IRE MODULATED RAMP 0.16 RL = 150Ω 0.14 0.12 0.10 PHASE 0.08 0.06 0.04 GAIN 0.01 0 5 6 7 8 9 10 11 12 13 14 15 SUPPLY VOLTAGE ( ±V) 0.02 00866-E-005 12 G = +2 RL = 150Ω RG = RFB VS = ±15V 9 6 GAIN (dB) 3 VS = ±5V 0 –3 1 10 FREQUENCY (MHz) 100 DIFFERENTIAL PHASE (DEGREES) Figure 6. Frequency Response DIFFERENTIAL GAIN (%) 0.07 0.06 0.05 0.04 0.03 0.02 0 Figure 5. Differential Gain and Phase Rev. E | Page 3 of 20 00866-E-006 –6 AD811 SPECIFICATIONS @ TA = +25°C, VS = ±15 V dc, RLOAD = 150 Ω, unless otherwise noted. Table 1. Parameter DYNAMIC PERFORMANCE Small Signal Bandwidth (No Peaking) −3 dB G = +1 G = +2 G = +2 G = +10 0.1 dB Flat G = +2 Full Power Bandwidth3 Slew Rate Settling Time to 0.1% Settling Time to 0.01% Settling Time to 0.1% Rise Time, Fall Time Differential Gain Differential Phase THD @ fC = 10 MHz Third-Order Intercept4 INPUT OFFSET VOLTAGE TMIN to TMAX Offset Voltage Drift INPUT BIAS CURRENT −Input TMIN to TMAX +Input TRANSRESISTANCE TMIN to TMAX TMIN to TMAX VOUT = ±10 V RL = ∞ RL = 200 Ω VOUT = ±2.5 V RL = 150 Ω ±5 V, ±1 5 V 2 5 ±5 V, ±15 V 2 5 15 10 20 Conditions VS AD811J/A1 Min Typ Max AD811S2 Min Typ Max Unit RFB = 562 Ω RFB = 649 Ω RFB = 562 Ω RFB = 511 Ω RFB = 562 Ω RFB = 649 Ω VOUT = 20 V p-p VOUT = 4 V p-p VOUT = 20 V p-p 10 V Step, AV = − 1 10 V Step, AV = − 1 2 V Step, AV = − 1 RFB = 649, AV = +2 f = 3.58 MHz f = 3.58 MHz VOUT = 2 V p-p, AV = +2 @ fC = 10 MHz ±15 V ±15 V ±15 V ±15 V ±15 ±15 ±15 ±15 ±15 ±15 ±15 ±15 ±15 ±15 ±15 ±15 ±15 ±15 ±5 V, ±15 V 140 120 80 100 25 35 40 400 2500 50 65 25 3.5 0.01 0.01 −74 36 43 0.5 140 120 80 100 25 35 40 400 2500 50 65 25 3.5 0.01 0.01 −74 36 43 0.5 5 2 2 5 30 10 25 MHz MHz MHz MHz MHz MHz MHz V/µs V/µs ns ns ns ns % Degree dBc dBm dBm mV mV µV/°C µA µA µA µA 3 5 3 5 ±15 V ±15 V ±5 V 0.75 0.5 0.25 1.5 0.75 0.4 0.75 0.5 0.125 1.5 0.75 0.4 MΩ MΩ MΩ 1 2 3 The AD811JR is specified with ±5 V power supplies only, with operation up to ±12 V. See the Analog Devices military data sheet for 883B tested specifications. FPBW = slew rate/(2 π VPEAK). 4 Output power level, tested at a closed-loop gain of two. Rev. E | Page 4 of 20 AD811 Parameter COMMON-MODE REJECTION VOS (vs. Common Mode) TMIN to TMAX TMIN to TMAX Input Current (vs. Common Mode) POWER SUPPLY REJECTION VOS +Input Current −Input Current INPUT VOLTAGE NOISE INPUT CURRENT NOISE OUTPUT CHARACTERISTICS Voltage Swing, Useful Operating Range3 Output Current Short-Circuit Current Output Resistance INPUT CHARACTERISTIC +Input Resistance −Input Resistance Input Capacitance Common-Mode Voltage Range POWER SUPPLY Operating Range Quiescent Current TRANSISTOR COUNT 1 2 3 Conditions Vs AD811J/A1 Min Typ Max AD811S2 Min Typ Max Unit VCM = ±2.5 V VCM = ±10 V TMIN to TMAX VS = ±4.5 V to ±18 V TMIN to TMAX TMIN to TMAX TMIN to TMAX f = 1 kHz f = 1 kHz ±5 V ±15 V 56 60 60 66 1 70 0.3 0.4 1.9 20 50 56 3 60 2 2 60 66 1 70 0.3 0.4 1.9 20 ±2.9 ±12 100 150 9 1.5 14 7.5 ±3 ±13 3 dB dB µA/V dB µA/V µA/V nV/√Hz pA/√Hz V V mA mA Ω MΩ Ω pF V V 60 2 2 ±5 V ±15 V TJ = 25°C (Open Loop @ 5 MHz) ±2.9 ±12 100 150 9 1.5 14 7.5 ±3 ±13 ±4.5 ±18 16.0 18.0 ±4.5 +Input ±5 V ±15 V ±5 V ±15 V Number of Transistors 14.5 16.5 40 14.5 16.5 40 ±18 16.0 18.0 V mA mA The AD811JR is specified with ±5 V power supplies only, with operation up to ±12 V. See the Analog Devices military data sheet for 883B tested specifications. Useful operating range is defined as the output voltage at which linearity begins to degrade. Rev. E | Page 5 of 20 AD811 ABSOLUTE MAXIMUM RATINGS Table 2. Parameter Supply Voltage AD811JR Grade Only Internal Power Dissipation 8-Lead PDIP Package 8-Lead CERDIP Package 8-Lead SOIC Package 16-Lead SOIC Package 20-Lead SOIC Package 20-Lead LCC Package Output Short-Circuit Duration Common-Mode Input Voltage Differential Input Voltage Storage Temperature Range (Q, E) Storage Temperature Range (N, R) Operating Temperature Range AD811J AD811A AD811S Lead Temperature Range (Soldering 60 sec) Rating ±18 V ±12 V Observe Derating Curves θJA = 90°C/ W θJA = 110°C/W θJA = 155°C/W θJA = 85°C/W θJA = 80°C/W θJA = 70°C/W Observe Derating Curves ±VS ±6 V −65°C to +150°C −65°C to +125°C 0°C to +70°C −40°C to +85°C −55°C to +125°C 300°C MAXIMUM POWER DISSIPATION The maximum power that can be safely dissipated by the AD811 is limited by the associated rise in junction temperature. For the plastic packages, the maximum safe junction temperature is 145°C. For the CERDIP and LCC packages, the maximum junction temperature is 175°C. If these maximums are exceeded momentarily, proper circuit operation is restored as soon as the die temperature is reduced. Leaving the device in the “overheated” condition for an extended period can result in device burnout. To ensure proper operation, it is important to obser ve the derating cur ves in Figure 22 and Figure 25. While the AD811 is internally short-circuit protected, this may not be sufficient to guarantee that the maximum junction temperature is not exceeded under all conditions. An important example is when the amplifier is driving a reverse-terminated 75 Ω cable and the cable’s far end is shorted to a power supply. With power supplies of ±12 V (or less) at an ambient temperature of +25°C or less, and the cable shorted to a supply rail, the amplifier is not destroyed, even if this condition persists for an extended period. Stresses above those listed under Absolute Maximum Ratings may cause permanent damage to the device. This is a stress rating only ; functional operation of the device at these or any other conditions above those indicated in the operational section of this specification is not implied. Exposure to absolute maximum rating conditions for extended periods may affect device reliability. METALIZATION PHOTOGRAPH Contact the factor y for the latest dimensions. V+ 7 VOUT 6 0.0618 (1.57) 0.098 (2.49) Figure 7. Metalization Photograph Dimensions Shown in Inches and (Millimeters) ESD CAUTION ESD (electrostatic discharge) sensitive device. Electrostatic charges as high as 4000 V readily accumulate on the human body and test equipment and can discharge without detection. Although this product features proprietary ESD protection circuitry, permanent damage may occur on devices subjected to high energy electrostatic discharges. Therefore, proper ESD precautions are recommended to avoid performance degradation or loss of functionality. Rev. E | Page 6 of 20 00866-E-007 –INPUT 2 3 AD811 +INPUT 4 V– AD811 TYPICAL PERFORMANCE CHARACTERISTICS 20 MAGNITUDE OF THE OUTPUT VOLTAGE (±V) COMMON-MODE VOLTAGE RANGE ( ±V) 20 TA = 25°C 15 TA = 25°C 15 RL = 150Ω 10 NO LOAD 5 10 5 0 5 10 SUPPLY VOLTAGE (±V) 15 20 0 5 10 SUPPLY VOLTAGE (±V) 15 20 Figure 8. Input Common-Mode Voltage Range vs. Supply Voltage 35 21 Figure 11. Output Voltage Swing vs. Supply Voltage QUIESCENT SUPPLY CURRENT (mA) 30 18 VS = ±15V 15 OUTPUT VOLTAGE (V p-p) 25 20 VS = ±15V 12 15 VS = ±5V 10 VS = ±5V 5 9 6 00866-E-009 0 10 100 1k 10k LOAD RESISTANCE (Ω) –40 –20 0 20 40 60 80 100 120 140 JUNCTION TEMPERATURE (°C) Figure 9. Output Voltage Swing vs. Resistive Load 10 5 0 –5 INVERTING INPUT –10 VS = ±15V –15 –20 –25 VS = ±5V NONINVERTING INPUT ±5 TO ±15V Figure 12. Quiescent Supply Current vs. Junction Temperature 10 8 MASTER CLOCK FREQUENCY (MHz) INPUT OFFSET VOLTAGE (mV) 6 4 2 0 VS = ±15V –2 –4 –6 –8 VS = ±5V 00866-E-010 –40 –20 0 20 40 60 80 100 120 140 –40 –20 0 20 40 60 80 100 120 140 JUNCTION TEMPERATURE (°C) JUNCTION TEMPERATURE (°C) Figure 10. Input Bias Current vs. Junction Temperature Figure 13. Input Offset Voltage vs. Junction Temperature Rev. E | Page 7 of 20 00866-E-013 –30 –60 –10 –60 00866-E-012 3 –60 00866-E-011 00866-E-008 0 0 AD811 250 2.0 VS = ±15V RL = 200Ω VOUT = ±10V SHORT-CIRCUIT CURRENT (mA) VS = ±15V 150 VS = ±5V TRANSRESISTANCE (MΩ) 200 1.5 1.0 VS = ±5V RL = 150Ω VOUT = ±2.5V 100 0.5 00866-E-014 –40 –20 0 20 40 60 80 100 120 140 –40 –20 0 20 40 60 80 100 120 140 JUNCTION TEMPERATURE (°C) JUNCTION TEMPERATURE (°C) Figure 14. Short-Circuit Current vs. Junction Temperature 10 100 Figure 17. Transresistance vs. Junction Temperature 100 CLOSED-LOOP OUTPUT RESISTANCE (Ω) NOISE VOLTAGE (nV/ Hz) 1 VS = ±15V NONINVERTING CURRENT VS = ±5V TO ±15V INVERTING CURRENT VS = ±5V TO ±15V 10 10 VS = ±5V 0.1 GAIN = –2 RFB = 649Ω 00866-E-015 VOLTAGE NOISE VS = ±15V VOLTAGE NOISE VS = ±5V 100M 10 100 1k FREQUENCY (Hz) 10k 00866-E-018 00866-E-019 0.01 10k 1 100k 1M FREQUENCY (Hz) 10M 1 100k Figure 15. Closed-Loop Output Resistance vs. Frequency 10 100 200 Figure 18. Input Noise vs. Frequency 10 8 RISE TIME 60 160 8 –3dB BANDWIDTH (MHz) RISE TIME (ns) 6 OVERSHOOT 4 40 120 6 20 80 4 PEAKING 40 2 2 0 0.6 0.8 1.0 1.2 1.4 00866-E-016 0 0.4 –20 1.6 0 0.4 0.6 0.8 1.0 1.2 1.4 0 1.6 VALUE OF FEEDBACK RESISTOR [RFB] (kΩ) VALUE OF FEEDBACK RESISTOR [RFB] (kΩ) Figure 16. Rise Time and Overshoot vs. Value of Feedback Resistor, RFB Figure 19. −3 dB Bandwidth and Peaking vs. Value of RFB Rev. E | Page 8 of 20 PEAKING (dB) VS = ±15V VO = 1V p-p RL = 150Ω GAIN = +2 OVERSHOOT (%) BANDWIDTH VS = ±15V VO = 1V p-p RL = 150Ω GAIN = +2 NOISE CURRENT (pA/ Hz) 00866-E-017 50 –60 0 –60 AD811 110 649Ω 100 VIN 90 150Ω 80 70 VS = ±15V 60 VS = ±5V 50 40 00866-E-020 00866-E-023 00866-E-025 00866-E-024 25 649Ω VS = ±15V VOUT 20 150Ω OUTPUT VOLTAGE (V p-p) CMRR (dB) 15 GAIN = +10 OUTPUT LEVEL FOR 3% THD 10 5 VS = ±5V 30 1k 10k 100k FREQUENCY (Hz) 1M 10M 0 100k 1M 10M FREQUENCY (Hz) 100M Figure 20. Common-Mode Rejection Ratio vs. Frequency 80 70 60 50 VS = ±15V 40 30 20 10 00866-E-021 Figure 23. Large Signal Frequency Response –50 RL = 100Ω VOUT = 2V p-p GAIN = +2 –70 SECOND HARMONIC ±5V SUPPLIES RF = 649Ω AV = +2 VS = ±5V HARMONIC DISTORTION (dBc) PSRR (dB) CURVES ARE FOR WORST CASE CONDITION WHERE ONE SUPPLY IS VARIED WHILE THE OTHER IS HELD CONSTANT. –90 THIRD HARMONIC –110 SECOND HARMONIC THIRD HARMONIC –130 1k 10k 100k FREQUENCY (Hz) ±15V SUPPLIES 5 1k 10k 100k FREQUENCY (Hz) 1M 10M 1M 10M Figure 21. Power Supply Rejection Ration vs. Frequency 2.5 TJ MAX = –145°C 16-LEAD SOIC 3.4 3.2 3.0 Figure 24. Harmonic Distortion vs. Frequency TJ MAX = –175°C TOTAL POWER DISSIPATION (W) TOTAL POWER DISSIPATION (W) 2.8 2.6 2.4 2.2 2.0 1.8 1.6 1.4 1.2 1.0 0.8 0.6 8-LEAD CERDIP 20-LEAD LCC 2.0 8-LEAD PDIP 1.5 20-LEAD SOIC 1.0 8-LEAD SOIC 0 10 20 30 40 50 60 70 80 90 00866-E-022 0.5 –50 –40 –30 –20 –10 0.4 –60 –40 –20 0 20 40 60 80 100 120 140 AMBIENT TEMPERATURE (°C) AMBIENT TEMPERATURE (°C) Figure 22. Maximum Power Dissipation vs. Temperature for Plastic Packages Figure 25. Maximum Power Dissipation vs. Temperature for Hermetic Packages Rev. E | Page 9 of 20 AD811 9 RFB 6 +VS 0.1µF RG 2 G = +1 RL = 150Ω RG = ∞ – GAIN (dB) 7 6 VOUT TO TEKTRONIX P6201 FET PROBE RL 3 VS = ±15V RFB = 750Ω 0 –3 VIN 3 AD811 + 5 HP8130 50Ω PULSE GENERATOR 0.1µF –VS –6 VS = ±5V RFB = 619Ω –9 00866-E-026 1 10 FREQUENCY (MHz) 100 Figure 26. Noninverting Amplifier Connection 26 Figure 29. Closed-Loop Gain vs. Frequency, Gain = +1 1V 100 10ns 23 G = +1 RL = 150Ω VS = ±15V RFB = 511Ω VIN 90 20 GAIN (dB) 17 VS = ±5V RFB = 442Ω 14 VOUT 10 0% 11 1V 00866-E-027 1 10 FREQUENCY (MHz) 100 Figure 27. Small Signal Pulse Response, Gain = +1 Figure 30. Closed-Loop Gain vs. Frequency, Gain = +10 100mV 100 10ns 100 1V VIN 90 20ns VIN 90 VOUT 10 0% VOUT 10 0% 1V 00866-E-028 10V 00866-E-031 Figure 28. Small Signal Pulse Response, Gain = +10 Figure 31. Large Signal Pulse Response, Gain = +10 Rev. E | Page 10 of 20 00866-E-030 8 00866-E-029 –12 AD811 RFB +VS 0.1µF VIN HP8130 PULSE GENERATOR 3 6 G = –1 RL = 150Ω VS = ±15V RFB = 590Ω 3 2 – GAIN (dB) RG 7 VOUT TO TEKTRONIX P6201 FET PROBE 0 AD811 + 4 –3 VS = ±5V RFB = 562Ω 6 RL –6 –9 0.1µF 00866-E-032 00866-E-035 00866-E-037 00866-E-036 –12 1 10 FREQUENCY (MHz) 100 –VS Figure 32. Inverting Amplifier Connection 26 Figure 35. Closed-Loop Gain vs. Frequency, Gain = −1 1V 100 10ns 23 G = –1 RL = 150Ω VIN 90 20 VS = ±15V RFB = 511Ω GAIN (dB) 17 VS = ±5V RFB = 442Ω VOUT 10 0% 14 11 1V 00866-E-033 8 1 10 FREQUENCY (MHz) 100 Figure 33. Small Signal Pulse Response, Gain = −1 Figure 36. Closed-Loop Gain vs. Frequency, Gain = −10 100mV 100 10ns 100 1V VIN 90 20ns VIN 90 VOUT 10 0% VOUT 10 0% 1V 00866-E-034 10V Figure 34. Small Signal Pulse Response, Gain = −10 Figure 37. Large Signal Pulse Response, Gain = −10 Rev. E | Page 11 of 20 AD811 APPLICATIONS GENERAL DESIGN CONSIDERATIONS The AD811 is a current feedback amplifier optimized for use in high performance video and data acquisition applications. Because it uses a current feedback architecture, its closed-loop −3 dB bandwidth is dependent on the magnitude of the feedback resistor. The desired closed-loop gain and bandwidth are obtained by varying the feedback resistor (RFB) to tune the bandwidth and by varying the gain resistor (RG) to obtain the correct gain. Table 3 contains recommended resistor values for a variety of useful closed-loop gains and supply voltages. Table 3. −3 dB Bandwidth vs. Closed-Loop Gain and Resistance Values VS = ±15 V Closed-Loop Gain +1 +2 +10 −1 −10 VS = ±5 V Closed-Loop Gain +1 +2 +10 −1 −10 VS = ±10 V Closed-Loop Gain +1 +2 +10 −1 −10 RFB 750 Ω 649 Ω 511 Ω 590 Ω 511 Ω RFB 619 Ω 562 Ω 442 Ω 562 Ω 442 Ω RFB 649 Ω 590 Ω 499 Ω 590 Ω 499 Ω RG 649 Ω 56.2 Ω 590 Ω 51.1 Ω RG 562 Ω 48.7 Ω 562 Ω 44.2 Ω RG 590 Ω 49.9 Ω 590 Ω 49.9 Ω −3 dB BW (MHz) 140 120 100 115 95 −3 dB BW (MHz) 80 80 65 75 65 −3 dB BW (MHz) 105 105 80 105 80 ACHIEVING THE FLATTEST GAIN RESPONSE AT HIGH FREQUENCY Achieving and maintaining gain flatness of better than 0.1 dB at frequencies above 10 MHz requires careful consideration of several issues. Choice of Feedback and Gain Resistors Because of the previously mentioned relationship between the 3 dB bandwidth and the feedback resistor, the fine scale gain flatness varies, to some extent, with feedback resistor tolerance. Therefore, it is recommended that resistors with a 1% tolerance be used if it is desired to maintain flatness over a wide range of production lots. In addition, resistors of different construction have different associated parasitic capacitance and inductance. Metal film resistors were used for the bulk of the characterization for this data sheet. It is possible that values other than those indicated are optimal for other resistor types. Printed Circuit Board Layout Considerations As is expected for a wideband amplifier, PC board parasitics can affect the overall closed-loop performance. Of concern are stray capacitances at the output and the inverting input nodes. If a ground plane is used on the same side of the board as the signal traces, a space (3/16" is plenty) should be left around the signal lines to minimize coupling. Additionally, signal lines connecting the feedback and gain resistors should be short enough so that their associated inductance does not cause high frequency gain errors. Line lengths less than 1/4" are recommended. Quality of Coaxial Cable Optimum flatness when driving a coax cable is possible only when the driven cable is terminated at each end with a resistor matching its characteristic impedance. If the coax is ideal, then the resulting flatness is not affected by the length of the cable. While outstanding results can be achieved using inexpensive cables, note that some variation in flatness due to varying cable lengths may occur. Figure 18 and Figure 19 illustrate the relationship between the feedback resistor and the frequency and time domain response characteristics for a closed-loop gain of +2. (The response at other gains is similar.) The 3 dB bandwidth is somewhat dependent on the power supply voltage. As the supply voltage is decreased, for example, the magnitude of the internal junction capacitances is increased, causing a reduction in closed-loop bandwidth. To compensate for this, smaller values of feedback resistor are used at lower supply voltages. Power Supply Bypassing Adequate power supply bypassing can be critical when optimizing the performance of a high frequency circuit. Inductance in the power supply leads can form resonant circuits that produce peaking in the amplifier’s response. In addition, if large current transients must be delivered to the load, then bypass capacitors (typically greater than 1 µF) are required to provide the best settling time and lowest distortion. Although the recommended 0.1 µF power supply bypass capacitors are sufficient in many applications, more elaborate bypassing (such as using two paralleled capacitors) may be required in some cases. Rev. E | Page 12 of 20 AD811 Driving Capacitive Loads The feedback and gain resistor values in Table 3 result in ver y flat closed-loop responses in applications where the load capacitances are below 10 pF. Capacitances greater than this result in increased peaking and overshoot, although not necessarily in a sustained oscillation. There are at least two ver y effective ways to compensate for this effect. One way is to increase the magnitude of the feedback resistor, which lowers the 3 dB frequency. The other method is to include a small resistor in series with the output of the amplifier to isolate it from the load capacitance. The results of these two techniques are illustrated in Figure 39. Using a 1.5 kΩ feedback resistor, the output ripple is less than 0.5 dB when driving 100 pF. The main disadvantage of this method is that it sacrifices a little bit of gain flatness for increased capacitive load drive capability. With the second method, using a series resistor, the loss of flatness does not occur. RFB +VS 0.1µF RG 2 7 100 90 80 70 GAIN = +2 VS = ±15V RS VALUE SPECIFIED IS FOR FLATTEST FREQUENCY RESPONSE VALUE OF RS (Ω) 60 50 40 30 20 10 00866-E-040 00866-E-041 0 10 100 LOAD CAPACITANCE (pF) 1000 Figure 40. Recommended Value of Series Resistor vs. the Amount of Capacitive Load Figure 40 shows recommended resistor values for different load capacitances. Refer again to Figure 39 for an example of the results of this method. Note that it may be necessar y to adjust the gain setting resistor, RG, to correct for the attenuation which results due to the divider formed by the series resistor, RS, and the load resistance. Applications that require driving a large load capacitance at a high slew rate are often limited by the output current available from the driving amplifier. For example, an amplifier limited to 25 mA output current cannot drive a 500 pF load at a slew rate greater than 50 V/µs. However, because of the AD811’s 100 mA output current, a slew rate of 200 V/µs is achievable when driving the same 500 pF capacitor, as shown in Figure 41. 2V 100 – RS (OPTIONAL) AD811 VIN 3 6 VOUT CL RL + 4 RT 0.1µF 00866-E-038 –VS 100ns Figure 38. Recommended Connection for Driving a Large Capacitive Load 12 VIN 90 9 RFB = 1.5kΩ RS = 0 6 GAIN (dB) 3 VS = ±15V CL = 100pF RL = 10kΩ GAIN = +2 RFB = 649Ω RS = 30Ω VOUT 10 0% 0 5V –3 1 10 FREQUENCY (MHz) 100 Figure 39. Performance Comparison of Two Methods for Driving a Capacitive Load Rev. E | Page 13 of 20 00866-E-039 –6 Figure 41. Output Waveform of an AD811 Driving a 500 pF Load. Gain = +2, RFB = 649 Ω, RS = 15 Ω, RS = 10 kΩ AD811 OPERATION AS A VIDEO LINE DRIVER The AD811 has been designed to offer outstanding performance at closed-loop gains of +1 or greater, while driving multiple reverse-terminated video loads. The lowest differential gain and phase errors are obtained when using ±15 V power supplies. With ±12 V supplies, there is an insignificant increase in these errors and a slight improvement in gain flatness. Due to power dissipation considerations, ±12 V supplies are recommended for optimum video performance. Excellent performance can be achieved at much lower supplies as well. The closed-loop gain versus the frequency at different supply voltages is shown in Figure 43. Figure 44 is an oscilloscope photograph of an AD811 line driver’s pulse response with ±15 V supplies. The differential gain and phase error versus the supply are plotted in Figure 45 and Figure 46, respectively. Another important consideration when driving multiple cables is the high frequency isolation between the outputs of the cables. Due to its low output impedance, the AD811 achieves better than 40 dB of output-to-output isolation at 5 MHz driving back-terminated 75 Ω cables. 649Ω 649Ω 75Ω CABLE 75Ω +VS 0.1µF 2– 7 100 1V VIN 90 10ns VOUT 10 0% 1V 00866-E-044 Figure 44. Small Signal Pulse Response, Gain = +2, VS = ±15 V 0.10 0.09 0.08 RF = 649Ω FC = 3.58MHz 100 IRE MODULATED RAMP DIFFERENTIAL GAIN (%) 0.07 0.06 0.05 0.04 0.03 0.02 0.01 0 a 00866-E-045 00866-E-046 VOUT No. 1 75Ω a. DRIVING A SINGLE, BACKTERMINATED, 75Ω COAX CABLE b. DRIVING TWO PARALLEL, BACKTERMINATED, COAX CABLES b 75Ω CABLE 5 VOUT No. 2 75Ω 6 7 8 9 10 11 12 13 14 15 75Ω CABLE VIN 75Ω 3+ AD811 4 6 75Ω SUPPLY VOLTAGE (V) Figure 45. Differential Gain Error vs. Supply Voltage for the Video Line Driver of Figure 42 0.20 RF = 649Ω FC = 3.58MHz 100 IRE MODULATED RAMP 0.1µF 00866-E-042 0.18 DIFFERENTIAL PHASE (DEGREES) 0.16 0.14 b 0.12 0.10 0.08 a 0.06 0.04 0.02 –VS Figure 42. A Video Line Driver Operating at a Gain of +2 12 G = +2 RL = 150Ω RG = RFB 9 a. DRIVING A SINGLE, BACKTERMINATED, 75Ω COAX CABLE b. DRIVING TWO PARALLEL, BACKTERMINATED, COAX CABLES 6 VS = ±15V RFB = 649Ω GAIN (dB) 3 VS = ±5V RFB = 562Ω 0 5 6 7 8 9 10 11 12 13 14 15 SUPPLY VOLTAGE (V) 0 –3 Figure 46. Differential Phase Error vs. Supply Voltage for the Video Line Driver of Figure 42 1 10 FREQUENCY (MHz) 100 Figure 43. Closed-Loop Gain vs. Frequency, Gain = +2 Rev. E | Page 14 of 20 00866-E-043 –6 AD811 AN 80 MHZ VOLTAGE-CONTROLLED AMPLIFIER CIRCUIT The voltage-controlled amplifier (VCA) circuit of Figure 48 shows the AD811 being used with the AD834, a 500 MHz, 4-quadrant multiplier. The AD834 multiplies the signal input by the dc control voltage, VG. The AD834 outputs are in the form of differential currents from a pair of open collectors, ensuring that the full bandwidth of the multiplier (which exceeds 500 MHz) is available for certain applications. Here, the AD811 op amp provides a buffered, single-ended, groundreferenced output. Using feedback resistors R8 and R9 of 511 Ω, the overall gain ranges from −70 dB for VG = 0 dB to +12 dB (a numerical gain of +4) when VG = 1 V. The overall transfer function of the VCA is VOUT = 4 (X1 − X2)(Y1 − Y2), which reduces to VOUT = 4 VG VIN using the labeling conventions shown in Figure 47. The circuit’s −3 dB bandwidth of 80 MHz is maintained essentially constant—that is, independent of gain. The response can be maintained flat to within ±0.1 dB from dc to 40 MHz at full gain with the addition of an optional capacitor of about 0.3 pF across the feedback resistor R8. The circuit produces a full-scale output of ±4 V for a ±1 V input and can drive a reverse-terminated load of 50 Ω or 75 Ω to ±2 V. The gain can be increased to 20 dB (×10) by raising R8 and R9 to 1.27 kΩ, with a corresponding decrease in −3 dB bandwidth to approximately 25 MHz. The maximum output voltage under these conditions is increased to ±9 V using ±12 V supplies. The gain-control input voltage, VG, may be a positive or negative ground-referenced voltage, or fully differential, depending on the choice of connections at Pins 7 and 8. A positive value of VG results in an overall noninverting response. Reversing the sign of VG simply causes the sign of the overall response to invert. In fact, although this circuit has been classified as a voltagecontrolled amplifier, it is also quite useful as a general-purpose, four-quadrant multiplier, with good load driving capabilities and fully symmetrical responses from the X and Y inputs. The AD811 and AD834 can both be operated from power supply voltages of ±5 V. While it is not necessar y to power them from the same supplies, the common-mode voltage at W1 and W2 must be biased within the common-mode range of the AD811’s input stage. To achieve the lowest differential gain and phase errors, it is recommended that the AD811 be operated from power supply voltages of ±10 V or greater. This VCA circuit operates from a ±12 V dual power supply. FB +12V C1 0.1µF + VG – R1 100Ω R8* R2 100Ω 8 7 6 5 X2 X1 +VS W1 R4 182Ω R6 294Ω 7 2– U1 AD834 3+ U3 AD811 4 6 VOUT Y1 1 Y2 2 –VS 3 W2 4 R5 182Ω R7 294Ω RL VIN R9* R3 249Ω C2 0.1µF FB *R8 = R9 = 511Ω FOR ×4 GAIN R8 = R9 = 1.27kΩ FOR ×10 GAIN –12V 00866-E-047 Figure 47. An 80 MHz Voltage-Controlled Amplifier Rev. E | Page 15 of 20 AD811 A VIDEO KEYER CIRCUIT By using two AD834 multipliers, an AD811, and a 1 V dc source, a special form of a two-input VCA circuit called a video keyer can be assembled. Keying is the term used in reference to blending two or more video sources under the control of a third signal or signals to create such special effects as dissolves and overlays. The circuit shown in Figure 48 is a two-input keyer, with video inputs VA and VB, and a control input VG. The transfer function (with VOUT at the load) is given by VOUT = GVA + (1−G)VB where G is a dimensionless variable (actually, just the gain of the A signal path) that ranges from 0 when VG = 0 to 1 when VG = 1 V. Thus, VOUT varies continuously between VA and VB as G varies from 0 to 1. Circuit operation is straightfor ward. Consider first the signal path through U1, which handles video input VA. Its gain is clearly 0 when VG = 0, and the scaling chosen ensures that it has a unity value when VG = 1 V; this takes care of the first term of the transfer function. On the other hand, the VG input to U2 is taken to the inverting input X2 while X1 is biased at an accurate 1 V. Thus, when VG = 0, the response to video input VB is already at its full-scale value of unity, whereas when VG = 1 V, the differential input X1−X2 is 0. This generates the second term. The bias currents required at the output of the multipliers are provided by R8 and R9. A dc level-shifting network comprising R10/R12 and R11/R13 ensures that the input nodes of the AD811 are positioned at a voltage within its common-mode range. At high frequencies, C1 and C2 bypass R10 and R11, respectively. R14 is included to lower the HF loop gain and is needed because the voltage-to-current conversion in the AD834s, via the Y2 inputs, results in an effective value of the feedback resistance of 250 Ω; this is only about half the value required for optimum flatness in the AD811’s response. (Note that this resistance is unaffected by G: when G = +1, all the feedback is via U1, while when G = 0 it is all via U2). R14 reduces the fractional amount of output current from the multipliers into the current-summing inverting input of the AD811 by sharing it with R8. This resistor can be used to adjust the bandwidth and damping factor to best suit the application. +5V R7 45.3Ω R6 226Ω 8 7 6 5 C1 0.1µF R14 SEE TEXT SETUP FOR DRIVING REVERSE-TERMINATED LOAD TO PIN 6 AD811 ZO 200Ω TO Y2 VOUT ZO R5 113Ω VG (0 TO +1V dc) R10 2.49kΩ X2 +5V R1 1.87kΩ X1 +VS W1 200Ω U4 AD589 U1 AD834 R8 29.4Ω R12 6.98kΩ +5V INSET R2 174Ω VA (±1V FS) Y1 1 Y2 2 –VS 3 W2 4 FB –5V R3 100Ω +5V –5V 7 2 8 7 6 5 C3 0.1µF LOAD GND – R4 1.02kΩ R9 29.4Ω R13 6.98kΩ 3 X2 X1 +VS W1 + C2 0.1µF U3 AD811 4 6 VOUT U1 AD834 C4 0.1µF Y1 1 Y2 2 –VS 3 W2 4 FB R11 2.49kΩ LOAD GND 00866-E-048 VB (±1V FS) –5V –5V Figure 48. A Practical Video Keyer Circuit Rev. E | Page 16 of 20 AD811 To generate the 1 V dc needed for the 1−G term, an AD589 reference supplies 1.225 V ± 25 mV to a voltage divider consisting of resistors R2 through R4. Potentiometer R3 should be adjusted to provide exactly 1 V at the X1 input. In this case, an arrangement is shown using dual supplies of ±5 V for both the AD834 and the AD811. Also, the overall gain is arranged to be unity at the load when it is driven from a reverse-terminated 75 Ω line. This means that the dual VCA has to operate at a maximum gain of +2, rather than +4 as in the VCA circuit of Figure 47. However, this cannot be achieved by lowering the feedback resistor because below a critical value (not much less than 500 Ω) the AD811’s peaking may be unacceptable. This is because the dominant pole in the openloop ac response of a current feedback amplifier is controlled by this feedback resistor. It would be possible to operate at a gain of ×4 and then attenuate the signal at the output. Instead, the signals have been attenuated by 6 dB at the input to the AD811; this is the function of R8 through R11. Figure 49 is a plot of the ac response of the feedback keyer when driving a reverse-terminated 50 Ω cable. Output noise and adjacent channel feedthrough, with either channel fully off and the other fully on, is about −50 dB to 10 MHz. The feedthrough at 100 MHz is limited primarily by board layout. For VG = 1 V, the −3 dB bandwidth is 15 MHz when using a 137 Ω resistor for R14 and 70 MHz with R14 = 49.9 Ω. For more information on the design and operation of the VCA and video keyer circuits, refer to the application note “Video VCAs and Keyers: Using the AD834 and AD811” by Brunner, Clarke, and Gilbert, available on the Analog Devices, Inc. website at www.analog.com. 10 0 –10 CLOSED-LOOP GAIN (dB) R14 = 49.9Ω GAIN R14 = 137Ω –20 –30 –40 –50 –60 –70 –80 00866-E-049 ADJACENT CHANNEL FEEDTHROUGH –90 10k 100k 1M FREQUENCY (Hz) 10M 100M Figure 49. A Plot of the AC Response of the Video Keyer Rev. E | Page 17 of 20 AD811 OUTLINE DIMENSIONS 0.375 (9.53) 0.365 (9.27) 0.355 (9.02) 8 5 5.00 (0.1968) 4.80 (0.1890) 8 5 4 1 4 0.295 (7.49) 0.285 (7.24) 0.275 (6.98) 0.325 (8.26) 0.310 (7.87) 0.300 (7.62) 0.015 (0.38) MIN SEATING PLANE 0.060 (1.52) 0.050 (1.27) 0.045 (1.14) 4.00 (0.1574) 3.80 (0.1497) 1 6.20 (0.2440) 5.80 (0.2284) 0.100 (2.54) BSC 0.180 (4.57) MAX 0.150 (3.81) 0.130 (3.30) 0.110 (2.79) 0.022 (0.56) 0.018 (0.46) 0.014 (0.36) 0.150 (3.81) 0.135 (3.43) 0.120 (3.05) 1.27 (0.0500) BSC 0.25 (0.0098) 0.10 (0.0040) 1.75 (0.0688) 1.35 (0.0532) 0.50 (0.0196) × 45° 0.25 (0.0099) 0.015 (0.38) 0.010 (0.25) 0.008 (0.20) 0.51 (0.0201) COPLANARITY SEATING 0.31 (0.0122) 0.10 PLANE 8° 0.25 (0.0098) 0° 1.27 (0.0500) 0.40 (0.0157) 0.17 (0.0067) COMPLIANT TO JEDEC STANDARDS MS-012AA CONTROLLING DIMENSIONS ARE IN MILLIMETERS; INCH DIMENSIONS (IN PARENTHESES) ARE ROUNDED-OFF MILLIMETER EQUIVALENTS FOR REFERENCE ONLY AND ARE NOT APPROPRIATE FOR USE IN DESIGN COMPLIANT TO JEDEC STANDARDS MO-095AA 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 Figure 52. 8-Lead Standard Small Outline Package [SOIC] Narrow Body (R-8) Dimensions shown in millimeters and (inches) Figure 50. 8-Lead Plastic Dual In-Line Package [PDIP] (N-8) Dimensions shown in inches and (millimeters) 0.005 (0.13) MIN 8 0.055 (1.40) MAX 5 0.100 (2.54) 0.064 (1.63) 0.075 (1.91) REF 0.095 (2.41) 0.075 (1.90) 19 18 20 1 4 PIN 1 1 4 0.310 (7.87) 0.220 (5.59) 3 0.200 (5.08) REF 0.100 (2.54) REF 0.015 (0.38) MIN 0.028 (0.71) 0.022 (0.56) 0.050 (1.27) BSC 0.100 (2.54) BSC 0.405 (10.29) MAX 0.200 (5.08) MAX 0.200 (5.08) 0.125 (3.18) 0.023 (0.58) 0.014 (0.36) 0.060 (1.52) 0.015 (0.38) 0.150 (3.81) MIN SEATING 0.070 (1.78) PLANE 0.030 (0.76) 15° 0° 0.015 (0.38) 0.008 (0.20) 0.320 (8.13) 0.290 (7.37) 0.358 (9.09) 0.342 (8.69) SQ 0.358 (9.09) MAX SQ 0.011 (0.28) 0.007 (0.18) R TYP 0.075 (1.91) REF 0.055 (1.40) 0.045 (1.14) BOTTOM VIEW 14 13 8 9 0.088 (2.24) 0.054 (1.37) 45° TYP 0.150 (3.81) BSC 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 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 Figure 53. 20-Terminal Ceramic Leadless Chip Carrier [LCC] (E-20A) Dimensions shown in inches and (millimeters) Figure 51. 8-Lead Ceramic Dual In-Line Package [CERDIP] (Q-8) Dimensions shown in inches and (millimeters) Rev. E | Page 18 of 20 AD811 10.50 (0.4134) 10.10 (0.3976) 20 13.00 (0.5118) 12.60 (0.4961) 16 9 11 7.60 (0.2992) 7.40 (0.2913) 1 8 7.60 (0.2992) 7.40 (0.2913) 1 10 10.65 (0.4193) 10.00 (0.3937) 10.65 (0.4193) 10.00 (0.3937) 1.27 (0.0500) BSC 0.30 (0.0118) 0.10 (0.0039) COPLANARITY 0.10 0.51 (0.0201) 0.31 (0.0122) 2.65 (0.1043) 2.35 (0.0925) 0.75 (0.0295) × 45° 0.25 (0.0098) 0.30 (0.0118) 0.10 (0.0039) 1.27 (0.0500) BSC 2.65 (0.1043) 2.35 (0.0925) 0.75 (0.0295) × 45° 0.25 (0.0098) SEATING PLANE 8° 0.33 (0.0130) 0° 0.20 (0.0079) 1.27 (0.0500) 0.40 (0.0157) COPLANARITY 0.10 8° 0.51 (0.0201) SEATING 0° 0.33 (0.0130) PLANE 0.31 (0.0122) 0.20 (0.0079) 1.27 (0.0500) 0.40 (0.0157) COMPLIANT TO JEDEC STANDARDS MS-013AA CONTROLLING DIMENSIONS ARE IN MILLIMETERS; INCH DIMENSIONS (IN PARENTHESES) ARE ROUNDED-OFF MILLIMETER EQUIVALENTS FOR REFERENCE ONLY AND ARE NOT APPROPRIATE FOR USE IN DESIGN COMPLIANT TO JEDEC STANDARDS MS-013AC CONTROLLING DIMENSIONS ARE IN MILLIMETERS; INCH DIMENSIONS (IN PARENTHESES) ARE ROUNDED-OFF MILLIMETER EQUIVALENTS FOR REFERENCE ONLY AND ARE NOT APPROPRIATE FOR USE IN DESIGN Figure 54. 16-Lead Standard Small Outline Package [SOIC] Wide Body (R-16) Dimensions shown in millimeters and (inches) Figure 55. 20-LeadStandard Small Outline Package [SOIC] Wide Body (R-20) Dimensions shown in millimeters and (inches) Rev. E | Page 19 of 20 AD811 ORDERING GUIDE Model AD811AN AD811ANZ1 AD811AR-16 AD811AR-16-REEL AD811AR-16-REEL7 AD811AR-20 AD811AR-20-REEL AD811JR AD811JR-REEL AD811JR-REEL7 AD811JRZ1 AD811SQ/883B 5962-9313101MPA AD811SE/883B 5962-9313101M2A AD811ACHIPS AD811SCHIPS 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 −40°C to +85°C 0°C to +70°C 0°C to +70°C 0°C to +70°C 0°C to +70°C −55°C to +125°C −55°C to +125°C −55°C to +125°C −55°C to +125°C −40°C to +85°C −55°C to +125°C Package Description 8-Lead Plastic Dual In-Line Package (PDIP) 8-Lead Plastic Dual In-Line Package (PDIP) 16-LeadStandard Small Outline Package (SOIC) 16-LeadStandard Small Outline Package (SOIC) 16-LeadStandard Small Outline Package (SOIC) 20-LeadStandard Small Outline Package (SOIC) 20-LeadStandard Small Outline Package (SOIC) 8-LeadStandard Small Outline Package (SOIC) 8-LeadStandard Small Outline Package (SOIC) 8-LeadStandard Small Outline Package (SOIC) 8-LeadStandard Small Outline Package (SOIC) 8-Lead Ceramic Dual In-Line Package (CERDIP) 8-Lead Ceramic Dual In-Line Package (CERDIP) 20-Terminal Ceramic Leadless Chip Carrier (LCC) 20-Terminal Ceramic Leadless Chip Carrier (LCC) Package Option N-8 N-8 R-16 R-16 R-16 R-20 R-20 R-8 R-8 R-8 R-8 Q-8 Q-8 E-20A E-20A DIE DIE Z = Pb-free part. © 2004 Analog Devices, Inc. All rights reserved. Trademarks and registered trademarks are the property of their respective owners. C00866–0–7/04(E) Rev. E | Page 20 of 20