30 MHz Dual Programmable Filters and Variable Gain Amplifiers ADRF6510 Matched pair of programmable filters and VGAs Continuous gain control range: −5 dB to +45 dB 6-pole filter 1 MHz to 30 MHz in 1 MHz steps, 0.5 dB corner frequency SPI programmable 6 dB front-end gain step IMD3: >55 dBc for 1.5 V p-p composite output HD2, HD3: >60 dBc for 1.5 V p-p output Differential input and output Adjustable output common-mode voltage Optional dc output offset correction Power-down feature Single 5 V supply operation APPLICATIONS FUNCTIONAL BLOCK DIAGRAM ENBL INP1 INM1 VPS COM GNSW OFS1 VPS VPSD OPP1 COMD OPM1 COM LE CLK ADRF6510 SPI GAIN VOCM DATA SDO COM COM OPM2 VPS OPP2 COM INP2 INM2 VPS COM OFDS OFS2 VPS 09002-001 FEATURES Figure 1. Baseband I/Q receivers Diversity receivers ADC drivers GENERAL DESCRIPTION The ADRF6510 is a matched pair of fully differential low noise and low distortion programmable filters and variable gain amplifiers (VGAs). Each channel is capable of rejecting large out-ofband interferers while reliably boosting the wanted signal, thus reducing the bandwidth and resolution requirements on the analog-to-digital converters (ADCs). The excellent matching between channels and their high spurious-free dynamic range over all gain and bandwidth settings make the ADRF6510 ideal for quadrature-based (IQ) communication systems with dense constellations, multiple carriers, and nearby interferers. The filters provide a six-pole Butterworth response with 0.5 dB corner frequencies programmable through the SPI port from 1 MHz to 30 MHz in 1 MHz steps. The preamplifier that precedes the filters offers a pin-programmable option of either 6 dB or 12 dB of gain. The preamplifier sets a differential input impedance of 400 Ω and has a common-mode voltage that defaults to 2.1 V but can be driven from 1.5 V to 2.5 V. The variable gain amplifiers that follow the filters provide 50 dB of continuous gain control with a slope of 30 mV/dB. The output buffers provide a differential output impedance of 20 Ω that is capable of driving 1.5 V p-p into 1 kΩ loads. The output commonmode voltage defaults to VPS/2, but it can be programmed via the VOCM pin. The built-in dc offset correction loop can be disabled if dc-coupled operation is desired. The high-pass corner frequency is defined by external capacitors on the OFS1 and OFS2 pins. The ADRF6510 operates from a 4.75 V to 5.25 V supply and consumes a maximum supply current of 258 mA when programmed to the highest bandwidth setting. When disabled, it consumes 2 mA. The ADRF6510 is fabricated in an advanced silicon-germanium BiCMOS process and is available in a 32-lead, exposed paddle LFCSP. Performance is specified over the −40°C to +85°C temperature range. Rev. 0 Information furnished by Analog Devices is believed to be accurate and reliable. However, no responsibility is assumed by Analog Devices for its use, nor for any infringements of patents or other rights of third parties that may result from its use. Specifications subject to change without notice. No license is granted by implication or otherwise under any patent or patent rights of Analog Devices. Trademarks and registered trademarks are the property of their respective owners. One Technology Way, P.O. Box 9106, Norwood, MA 02062-9106, U.S.A. Tel: 781.329.4700 www.analog.com Fax: 781.461.3113 ©2010 Analog Devices, Inc. All rights reserved. ADRF6510 TABLE OF CONTENTS Features .............................................................................................. 1 Programming the Filters ........................................................... 16 Applications ....................................................................................... 1 Noise Characteristics ................................................................. 16 Functional Block Diagram .............................................................. 1 Distortion Characteristics ......................................................... 17 General Description ......................................................................... 1 Maximizing the Dynamic Range.............................................. 17 Revision History ............................................................................... 2 Key Parameters for Quadrature-Based Receivers .................. 18 Specifications..................................................................................... 3 Applications Information .............................................................. 19 Timing Diagrams.......................................................................... 5 Basic Connections ...................................................................... 19 Absolute Maximum Ratings............................................................ 6 Error Vector Magnitude (EVM) Performance ........................... 19 ESD Caution .................................................................................. 6 Low IF Image Rejection............................................................. 20 Pin Configuration and Function Descriptions ............................. 7 Example Baseband Interface ..................................................... 21 Typical Performance Characteristics ............................................. 8 Evaluation Board ............................................................................ 23 Theory of Operation ...................................................................... 14 Evaluation Board Control Software ......................................... 23 Input Buffers ............................................................................... 14 Schematics and Artwork ........................................................... 23 Programmable Filters ................................................................. 14 Evaluation Board Configuration Options ............................... 25 Variable Gain Amplifiers (VGAs) ............................................ 15 Outline Dimensions ....................................................................... 27 Output Buffers/ADC Drivers ................................................... 15 Ordering Guide .......................................................................... 27 DC Offset Compensation Loop................................................ 15 REVISION HISTORY 4/10—Revision 0: Initial Version Rev. 0 | Page 2 of 28 ADRF6510 SPECIFICATIONS VPS = 5 V, TA = 25°C, ZSOURCE = 400 Ω, ZLOAD = 1 kΩ, VOUT = 1.5 V p-p, bandwidth setting = 30 MHz, GNSW = 0 V, unless otherwise noted. Table 1. Parameter FREQUENCY RESPONSE Low-Pass Corner Frequency, fC Step Size Corner Frequency Absolute Accuracy Corner Frequency Matching Pass-Band Ripple Gain Matching Group Delay Variation Corner Frequency = 1 MHz Corner Frequency = 30 MHz Group Delay Matching Corner Frequency = 1 MHz Corner Frequency = 30 MHz Stop-Band Rejection Relative to Pass Band INPUT STAGE Maximum Input Swing Differential Input Impedance Input Common-Mode Range GAIN CONTROL Voltage Gain Range Gain Slope Gain Error Gain Step OUTPUT STAGE Maximum Output Swing Differential Output Impedance Output DC Offset Output Common-Mode Range NOISE/DISTORTION 1 MHz Corner Frequency Output Noise Density Second Harmonic, HD2 Third Harmonic, HD3 Test Conditions/Comments Min 6-pole Butterworth filter, 0.5 dB bandwidth 1 Over operating temperature range Channel A and Channel B at same gain and bandwidth settings Channel A and Channel B at same gain and bandwidth settings From midband to peak Typ Max Unit 30 1 ±15 MHz MHz % fC ±0.5 % fC 0.5 ±0.1 dB p-p dB 135 11 ns ns 5 0.2 ns ns 30 75 dB dB 1 400 V p-p Ω V V Channel A and Channel B at same gain 2 × fC 5 × fC INP1, INM1, INP2, INM2 At minimum gain, VGAIN = 0 V 1 V p-p input voltage Input pins left floating GAIN, GNSW GNSW = 0 V, VGAIN from 0 V to 2 V GNSW = 5 V 1.5 −5 1 Gain = 0 dB at fC/2 Gain = 20 dB at fC/2 Gain = 40 dB at fC/2 250 kHz fundamental, 1.5 V p-p output voltage Gain = 0 dB Gain = 40 dB 250 kHz fundamental, 1.5 V p-p output voltage Gain = 0 dB Gain = 40 dB Rev. 0 | Page 3 of 28 +45 51 30 0.2 6 VGAIN from 500 mV to 1.7 V GNSW = 0 V to 5 V OPP1, OPM1, OPP2, OPM2, VOCM At maximum gain, RLOAD = 1 kΩ HD2 > 60 dBc, HD3 > 60 dBc Inputs shorted, offset loop disabled 1.5 V p-p output voltage VOCM left floating VPS/2 VPS/2 2 1.5 20 35 dB dB mV/dB dB dB VPS/2 V p-p V p-p Ω mV V V −129 −127 −111 dBV/√Hz dBV/√Hz dBV/√Hz 46.2 43.2 dBc dBc 52.2 51.2 dBc dBc 1.5 3.0 ADRF6510 Parameter IMD3 IMD3 with Input CW Blocker 30 MHz Corner Frequency Output Noise Density Second Harmonic, HD2 Third Harmonic, HD3 IMD3 IMD3 with Input CW Blocker DIGITAL LOGIC Input High Voltage, VINH Input Low Voltage, VINL Input Current, IINH/IINL Input Capacitance, CIN SPI TIMING fSCLK tDH tDS tLH tLS tPW tD POWER AND ENABLE Supply Voltage Range Total Supply Current Disable Current Disable Threshold Enable Response Time Disable Response Time Test Conditions/Comments f1 = 500 kHz, f2 = 550 kHz, 1.5 V p-p composite output voltage Gain = 5 dB Gain = 35 dB f1 = 500 kHz, f2 = 550 kHz, 1.5 V p-p composite output, gain = 5 dB; blocker at 5 MHz, 10 dBc relative to two-tone composite output voltage Min Midband, gain = 0 dB Midband, gain = 20 dB Midband, gain = 40 dB 8 MHz fundamental, 1.5 V p-p output voltage Gain = 0 dB Gain = 40 dB 8 MHz fundamental, 1.5 V p-p output voltage Gain = 0 dB Gain = 40 dB f1 = 15 MHz, f2 = 16 MHz, 1.5 V p-p composite output voltage Gain = 5 dB Gain = 35 dB f1 = 15 MHz, f2 = 16 MHz, 1.5 V p-p composite output, gain = 5 dB; blocker at 150 MHz, 10 dBc relative to two-tone composite output voltage LE, CLK, DATA, SDO, OFDS, GNSW LE, CLK, DATA, SDO 1/tSCLK DATA hold time DATA setup time LE hold time LE setup time CLK high pulse width CLK to SDO delay VPS, VPSD, COM, COMD, ENBL 4.75 ENBL = 5 V Maximum bandwidth setting Minimum bandwidth setting ENBL = 0 V Delay following ENBL low-to-high transition Delay following ENBL high-to-low transition Rev. 0 | Page 4 of 28 Typ Max Unit 61 57 40 dBc dBc dBc −130 −130 −123 dBV/√Hz dBV/√Hz dBV/√Hz 63 84 dBc dBc 54 87 dBc dBc 59 77.5 55 dBc dBc dBc >2 <0.8 <1 2 V V μA pF 20 5 5 5 5 5 5 MHz ns ns ns ns ns ns 5.0 258 131 2 2.5 20 300 5.25 V mA mA mA V μs ns ADRF6510 TIMING DIAGRAMS tCLK tPW CLK tLH tLS LE DATA tDH WRITE BIT LSB LSB + 1 MSB – 2 MSB – 1 MSB- 2 MSB NOTES 1. THE FIRST DATA BIT DETERMINES WHETHER THE PART IS WRITING TO OR READING FROM THE INTERNAL CORNER FREQUENCY WORD REGISTER. FOR A WRITE OPERATION, THE FIRST BIT SHOULD BE A LOGIC 1. THE CORNER FREQUENCY WORD BIT IS THEN REGISTERED INTO THE DATA PIN ON CONSECUTIVE RISING EDGES OF THE CLOCK. 09002-003 tDS Figure 2. Write Mode Timing Diagram tPW tCLK tD CLK tLH tLS LE DATA SDO tDH READ BIT DC LSB DC DC LSB + 1 MSB – 2 DC DC MSB – 1 DC DC MSB NOTES 1. THE FIRST DATA BIT DETERMINES WHETHER THE PART IS WRITING TO OR READING FROM THE INTERNAL CORNER FREQUENCY WORD REGISTER. FOR A READ OPERATION, THE FIRST BIT SHOULD BE A LOGIC 0. THE CORNER FREQUENCY WORD BIT IS THEN UPDATED AT THE SDO PIN ON CONSECUTIVE FALLING EDGES OF THE CLOCK. Figure 3. Read Mode Timing Diagram Rev. 0 | Page 5 of 28 09002-004 tDS ADRF6510 ABSOLUTE MAXIMUM RATINGS Table 2. Parameter Supply Voltages, VPS, VPSD ENBL, GNSW, OFDS, LE, CLK, DATA, SDO INP1, INM1, INP2, INM2 OPP1, OPM1, OPP2, OPM2 OFS1, OFS2 GAIN Internal Power Dissipation θJA (Exposed Pad Soldered to Board) Maximum Junction Temperature Operating Temperature Range Storage Temperature Range Lead Temperature (Soldering 60 sec) Rating 5.25 V VPS + 0.6 V VPS + 0.6 V, GND − 0.6 V VPS + 0.6 V VPS + 0.6 V VPS + 0.6 V 1.4 W 37.4°C/W 150°C −40°C to +85°C −65°C to +150°C 300°C 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. ESD CAUTION Rev. 0 | Page 6 of 28 ADRF6510 32 31 30 29 28 27 26 25 ENBL INP1 INM1 VPS COM GNSW OFS1 VPS PIN CONFIGURATION AND FUNCTION DESCRIPTIONS 1 2 3 4 5 6 7 8 PIN 1 INDICATOR ADRF6510 TOP VIEW (Not to Scale) 24 23 22 21 20 19 18 17 OPP1 OPM1 COM GAIN VOCM COM OPM2 OPP2 NOTES 1. CONNECT THE EXPOSED PADDLE TO A LOW IMPEDANCE GROUND PAD. 09002-002 COM INP2 INM2 VPS COM OFDS OFS2 VPS 9 10 11 12 13 14 15 16 VPSD COMD LE CLK DATA SDO COM VPS Figure 4. Pin Configuration Table 3. Pin Function Descriptions Pin No. 1 2 3 4 5 6 7, 9, 13, 19, 22, 28 8, 12, 16, 25, 29 10, 11, 30, 31 14 15, 26 17, 18, 23, 24 20 21 27 32 Mnemonic VPSD COMD LE CLK DATA SDO COM VPS INP2, INM2, INM1, INP1 OFDS OFS2, OFS1 OPP2, OPM2, OPM1, OPP1 VOCM GAIN GNSW ENBL EP Description Digital Positive Supply Voltage: 4.75 V to 5.25 V. Digital Common. Connect to external circuit common using the lowest possible impedance. Latch Enable. SPI programming pin. CMOS levels: VLOW < 0.8 V, VHIGH > 2 V. SPI Port Clock. CMOS levels: VLOW < 0.8 V, VHIGH > 2 V. SPI Data Input. CMOS levels: VLOW < 0.8 V, VHIGH > 2 V. SPI Data Output. CMOS levels: VLOW < 0.8 V, VHIGH > 2 V. Analog Common. Connect to external circuit common via a 1 kΩ resistor. Analog Positive Supply Voltage: 4.75 V to 5.25 V. Differential Inputs. 400 Ω input impedance. Common-mode range is 1.5 V to 2.5 V; default is 2.1 V. Offset Correction Loop Disable. Pull high to disable the offset correction loop. Offset Correction Loop Compensation Capacitors. Connect capacitors to circuit common. Differential Outputs. 20 Ω output impedance. Common-mode range is 1.5 V to 3 V; default is VPS/2. Output Common-Mode Setpoint. Defaults to VPS/2 if left open. Analog Gain Control. 0 V to 2 V, 30 mV/dB gain scaling. Front-End Gain Switch, 6 dB or 12 dB. Pull low for 6 dB; pull high for 12 dB. Chip Enable. Pull high to enable. Exposed Paddle. Connect the exposed paddle to a low impedance ground pad. Rev. 0 | Page 7 of 28 ADRF6510 TYPICAL PERFORMANCE CHARACTERISTICS VPS = 5 V, TA = 25°C, ZSOURCE = 400 Ω, ZLOAD = 1 kΩ, VOUT = 1.5 V p-p, GNSW = 0 V, unless otherwise noted. 3.0 50 BANDWIDTH = 30MHz 30 1.0 GAIN ERROR (dB) 1.5 25 –40°C VPS = 4.75V, 5V, 5.25V 15 +85°C VPS = 4.75V, 5V, 5.25V 10 –1.0 –1.5 0 –2.0 –5 –2.5 0 200 400 600 800 1000 1200 1400 1600 1800 2000 2200 VGAIN (mV) 50 45 40 35 30 25 20 15 10 5 0 –5 –10 –15 –20 –25 –30 –35 –40 –45 –50 +25°C VPS = 4.25V, 5V, 5.25V –40°C VPS = 4.25V, 5V, 5.25V –3.0 0 200 400 600 800 1000 1200 1400 1600 1800 2000 2200 2400 VGAIN (mV) Figure 5. In-Band Gain vs. VGAIN over Supply and Temperature (Bandwidth Setting = 30 MHz) Figure 8. Gain Conformance vs. VGAIN over Supply and Temperature (Bandwidth Setting = 30 MHz) 9 5 BANDWIDTH = 30MHz BANDWIDTH = 30MHz 8 3 1 10 100 FREQUENCY (MHz) 7 –1 6 –3 5 PREAMP GAIN = 6dB –5 4 –7 3 2 –9 0 5 10 15 20 25 30 35 40 FREQUENCY (MHz) Figure 6. Gain vs. Frequency by VGAIN (Bandwidth Setting = 30 MHz) Figure 9. 6 dB Gain Step and Gain Error vs. Frequency (Bandwidth Setting = 30 MHz, VGAIN = 0 V) 0.25 40 PREAMP GAIN = 6dB PREAMP GAIN = 12dB BANDWIDTH = 30MHz 35 0.15 30 OP1dB (dBV rms) 0.20 0.10 0.05 25 20 15 –0.05 0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 VGAIN (V) Figure 7. Gain Matching vs. VGAIN (Bandwidth Setting = 30 MHz) 10 0 5 10 15 20 GAIN (dB) 25 30 35 40 09002-010 0 09002-007 GAIN MISMATCH (dB) GAIN STEP (dB) 1 09002-009 GAIN ERROR (dB) PREAMP GAIN = 12dB 09002-006 GAIN (dB) 0 –0.5 5 –10 +85°C VPS = 4.25V, 5V, 5.25V 0.5 09002-005 GAIN (dB) 2.0 35 20 BANDWIDTH = 30MHz 2.5 +25°C VPS = 4.75V, 5V, 5.25V 40 09002-008 45 Figure 10. Output P1dB vs. GAIN at 15 MHz (Bandwidth Setting = 30 MHz) Rev. 0 | Page 8 of 28 ADRF6510 40 1000 35 900 30 800 25 700 GAIN = 20dB 20 15 10 5 600 500 400 BW = 5MHz 300 0 200 –5 100 10M 100M FREQUENCY (Hz) BW = 30MHz 0 0.5 09002-011 –10 1M BW = 10MHz BW = 20MHz 5 50 FREQUENCY (MHz) Figure 11. Frequency Response vs. Bandwidth Setting (Gain = 30 dB), Log Scale 09002-014 GROUP DELAY (ns) GAIN (dB) BW = 1MHz Figure 14. Group Delay vs. Frequency (Gain = 20 dB) 32 1.0 BANDWIDTH = 30MHz 29 27 1 6 11 16 21 26 31 36 40 FREQUENCY (MHz) 0 GAIN = 40dB –0.5 –1.0 0.5 09002-012 28 GAIN = 20dB 1.4 Figure 15. Group Delay Mismatch vs. Frequency (Bandwidth Setting = 30 MHz) 5 –40°C 4 IQ GROUP DELAY MATCHING (ns) 28 +25°C 27 26 25 24 +85°C 23 22 21 20 19 18 17 BANDWIDTH = 1MHz 3 GAIN = 20dB 2 1 0 –1 GAIN = 0dB –2 –3 –4 16 15 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 FREQUENCY (MHz) 09002-013 GAIN (dB) 30 FREQUENCY (MHz) Figure 12. Frequency Response vs. Bandwidth Setting (Gain = 30 dB), Linear Scale 30 29 5 09002-015 30 0.5 09002-016 GROUP DELAY MATCHING (ns) GAIN (dB) 31 Figure 13. Frequency Response over Temperature (Gain = 26 dB, Bandwidth Setting = 30 MHz) –5 0.2 0.4 0.6 0.8 1.0 1.2 FREQUENCY (MHz) Figure 16. IQ Group Delay Mismatch vs. Frequency (Bandwidth Setting = 1 MHz) Rev. 0 | Page 9 of 28 ADRF6510 0 FREQUENCY (MHz) 1.0 1.5 2.0 0.5 100 2.5 3.0 0.75 90 BANDWIDTH = 1MHz 0.25 BANDWIDTH = 30MHz 0 80 70 +25°C, VPS = 4.75V +25°C, VPS = 5V +25°C, VPS = 5.25V +85°C, VPS = 4.75V +85°C, VPS = 5V +85°C, VPS = 5.25V –40°C, VPS = 4.75V –40°C, VPS = 5V –40°C, VPS = 5.25V 60 –0.25 50 –0.50 40 0 0 5 10 15 20 30 25 FREQUENCY (MHz) 15 20 25 30 35 40 Figure 20. HD3 vs. Gain over Supply and Temperature (Bandwidth Setting = 30 MHz) 100 1.5V p-p OUTPUT @ 8MHz BANDWIDTH = 30MHz 90 80 HD3 @ 24MHz (dBc) 75 70 65 +25°C, VPS = 4.75V +25°C, VPS = 5V +25°C, VPS = 5.25V +85°C, VPS = 4.75V +85°C, VPS = 5V +85°C, VPS = 5.25V –40°C, VPS = 4.75V –40°C, VPS = 5V –40°C, VPS = 5.25V 55 50 45 40 0 5 10 15 20 25 30 35 70 60 VOCM = 1.5V VOCM = 1.75V VOCM = 2V VOCM = 2.5V 50 40 GAIN (dB) 40 09002-018 60 80 0 5 10 15 20 25 30 35 40 GAIN (dB) Figure 18. HD2 vs. Gain over Supply and Temperature (Bandwidth Setting = 30 MHz) 09002-021 90 85 Figure 21. HD3 vs. Gain over Output Common-Mode Voltage (Bandwidth Setting = 30 MHz) 100 30 1.5V p-p OUTPUT @ 8MHz BANDWIDTH = 30MHz BANDWIDTH = 30MHz f1 = 14MHz, f2 = 15MHz 90 25 PREAMP GAIN = 6dB 80 OIP3 (dBV) 20 70 60 15 PREAMP GAIN = 12dB 10 50 VOCM = 1.5V VOCM = 1.75V VOCM = 2V VOCM = 2.5V 5 30 0 5 10 15 20 25 30 35 40 GAIN (dB) 09002-019 40 Figure 19. HD2 vs. Gain over Output Common-Mode Voltage (Bandwidth Setting = 30 MHz) 0 0 5 10 15 20 25 30 35 40 45 50 GAIN (dB) Figure 22. In-Band OIP3 vs. Gain (Bandwidth Setting = 30 MHz) Rev. 0 | Page 10 of 28 09002-022 HD2 @ 16MHz (dBc) 10 GAIN (dB) Figure 17. IQ Amplitude Mismatch vs. Frequency HD2 @ 16MHz (dBc) 5 09002-017 –0.75 09002-020 HD3 @ 24MHz (dBc) IQ AMPLITUDE MISMATCH (dB) 0.50 30 BANDWIDTH = 30MHz f1 = 14MHz, f2 = 15MHz 25 –40°C OIP3 (dBV) IMD2 @ 15MHz (dBV) +25°C 20 +85°C 15 10 0 0 5 10 15 20 25 30 35 40 GAIN (dB) 09002-023 5 65 55 BANDWIDTH = 30MHz 45 35 25 15 5 –5 2:1 SLOPE –15 –25 –35 –45 PREAMP –55 GAIN = 12dB –65 OUT-OF-BAND IIP2 –75 –85 –95 PREAMP –105 GAIN = 6dB –115 –125 –135 –145 –45 –35 –25 –15 –5 55 5 15 25 35 45 Figure 23. In-Band OIP3 vs. Gain over Temperature (Preamp Gain = 6 dB, Bandwidth Setting = 30 MHz) Figure 26. Out-of-Band IIP2: IMD2 Tone at Midband (Bandwidth Setting = 30 MHz) 120 10 GAIN = 0dB GAIN = 10dB GAIN = 20dB GAIN = 30dB GAIN = 40dB BANDWIDTH = 30MHz 0 –10 –20 IMD3 @ 15MHz (dBV) 100 80 IMD3 (dBc) 65 INPUT LEVEL @ 115MHz AND 130MHz (dBV/Tone) 09002-060 ADRF6510 60 40 –30 –40 3:1 SLOPE –50 –60 PREAMP GAIN = 12dB –70 OUT-OF-BAND IIP3 –80 –90 –100 20 PREAMP GAIN = 6dB –110 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 COMPOSITE OUTPUT VOLTAGE (V p-p) –130 –45 09002-024 0 –35 –30 –25 –20 –15 –10 –5 0 5 INPUT LEVEL @ 115MHz AND 215MHz (dBV/Tone) Figure 24. In-Band Third-Order Intermodulation Distortion (Preamp Gain = 6 dB, Bandwidth Setting = 30 MHz) Figure 27. Out-of-Band IIP3: IMD3 Tone at Midband (Bandwidth Setting = 30 MHz) 120 70 GAIN = 6dB GAIN = 16dB GAIN = 26dB GAIN = 36dB GAIN = 46dB 1MHz BW 2MHz BW 4MHz BW 6MHz BW 8MHz BW 30MHz BW 65 60 NOISE FIGURE (dB re 50Ω) 100 80 60 40 55 50 45 40 35 30 20 0 0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 COMPOSITE OUTPUT VOLTAGE (V p-p) Figure 25. In-Band Third-Order Intermodulation Distortion (Preamp Gain = 12 dB, Bandwidth Setting = 30 MHz) 20 –5 0 5 10 15 20 25 30 35 40 GAIN (dB) Figure 28. Noise Figure vs. Gain over Bandwidth Setting, Preamp Gain = 6 dB (Noise Figure at 1/2 Bandwidth) Rev. 0 | Page 11 of 28 45 09002-028 25 09002-025 IMD3 (dBc) –40 09002-027 –120 0 ADRF6510 –90 60 1MHz BW 2MHz BW 4MHz BW 6MHz BW 8MHz BW 30MHz BW 45 40 35 30 25 20 –110 –115 –120 –125 –130 5 10 15 20 25 30 35 40 45 50 GAIN (dB) –140 0.5 Figure 29. Noise Figure vs. Gain over Bandwidth Setting, Preamp Gain = 12 dB (Noise Figure at 1/2 Bandwidth) 1.0 1.5 2.0 2.5 3.0 FREQUENCY (MHz) 09002-032 0 09002-029 10 Figure 32. Output Noise Density vs. Frequency (Bandwidth Setting = 1 MHz) –110 –100 1MHz BW 2MHz BW 4MHz BW 6MHz BW 8MHz BW 30MHz BW –110 GAIN = 0dB GAIN = 20dB GAIN = 40dB –115 OUTPUT NOISE (dBV rms/√Hz) –105 –115 –120 –125 –120 –125 –130 –135 –130 0 5 10 15 20 25 30 35 40 45 GAIN (dB) –140 10 09002-030 –135 –5 Figure 30. Output Noise Density vs. Gain by Bandwidth Setting, Preamp Gain = 6 dB (Noise at 1/2 Bandwidth) 12 14 16 18 20 22 24 26 28 30 FREQUENCY (MHz) 09002-033 OUTPUT NOISE (dBV rms/√Hz) –105 –135 15 Figure 33. Output Noise Density vs. Frequency (Bandwidth Setting = 20 MHz) –90 –100 1MHz BW 2MHz BW 4MHz BW 6MHz BW 8MHz BW 30MHz BW –110 –95 OUTPUT NOISE @ 15MHz (dBV/√Hz) –105 –115 –120 –125 –130 –100 –105 GAIN = 40dB –110 –115 –120 GAIN = 20dB –125 –130 GAIN = 0dB –135 –135 0 5 10 15 20 25 30 35 40 45 50 GAIN (dB) 09002-031 OUTPUT NOISE (dBV rms/√Hz) –100 Figure 31. Output Noise Density vs. Gain by Bandwidth Setting, Preamp Gain = 12 dB (Noise at 1/2 Bandwidth) –140 –35 –30 –25 –20 –15 BLOCKER LEVEL @ 150MHz (dBV rms) Figure 34. Output Noise Density vs. Blocker Level (Bandwidth Setting = 30 MHz, Blocker at 150 MHz) Rev. 0 | Page 12 of 28 –10 09002-034 NOISE FIGURE (dB re 50Ω) 50 GAIN = 0dB GAIN = 20dB GAIN = 40dB –95 OUTPUT NOISE (dBV rms/√Hz) 55 450 10 280 440 5 260 430 0 240 420 –5 410 –10 400 –15 390 –20 380 –25 370 –30 360 –35 120 –40 100 5 10 15 20 25 ISUPPLY (mA) CIN (pF) 160 140 30 FREQUENCY (MHz) 0 200 266 27 100 264 26 0 25 –100 24 –200 23 –300 22 –400 21 –500 250 –600 248 –50 5 10 15 20 25 20 25 30 ISUPPLY (mA) LSERIES_OUT (nH) 260 258 256 254 252 30 FREQUENCY (MHz) VPS = 4.75V VPS = 5V VPS = 5.25V –30 –10 10 30 50 70 90 TEMPERATURE (°C) Figure 36. Output Impedance vs. Frequency (Bandwidth Setting = 30 MHz) Figure 39. Current Consumption vs. Temperature over Supply (Bandwidth Setting = 30 MHz) 90 70 BANDWIDTH = 30MHz BANDWIDTH = 30MHz 80 60 70 GAIN = 40dB GAIN = 40dB 50 GAIN = 20dB CMRR (dB) 60 50 GAIN = 20dB 40 40 30 GAIN = 0dB 30 20 20 0 0.1 1 10 100 FREQUENCY (MHz) Figure 37. Channel Isolation, Output to Output, vs. Frequency (Bandwidth Setting = 30 MHz) 0 0 5 10 15 20 25 FREQUENCY (MHz) Figure 40. Common-Mode Rejection Ratio vs. Frequency (Bandwidth Setting = 30 MHz) Rev. 0 | Page 13 of 28 30 09002-040 10 10 09002-037 ISOLATION (dB) 15 262 09002-036 0 10 Figure 38. Current Consumption vs. Bandwidth Setting (Gain = 20 dB) 28 20 5 BANDWIDTH SETTING (MHz) Figure 35. Input Impedance vs. Frequency (Bandwidth Setting = 30 MHz) RSERIES_OUT (Ω) 180 09002-038 0 200 09002-039 350 220 09002-035 RIN (Ω) ADRF6510 ADRF6510 THEORY OF OPERATION band rejection, ripple, and group delay. The 0.5 dB bandwidth is programmed from 1 MHz to 30 MHz in 1 MHz steps via the serial programming interface (SPI) as described in the Programming the Filters section. The ADRF6510 consists of a matched pair of buffered, programmable filters followed by variable gain amplifiers and output ADC drivers. The block diagram of a single channel is shown in Figure 41. The programmability of the bandwidth and of the pre- and post-filtering gain offers great flexibility when coping with signals of varying levels in the presence of noise and large, undesired signals nearby. The entire differential signal chain is dc-coupled with flexible interfaces at the input and output. The bandwidth and gain setting controls for the two channels are shared, ensuring close matching of their magnitude and phase responses. The ADRF6510 can be fully disabled through the ENBL pin. 50dB VGA where: φ is the phase in radians. ω = 2πf is the frequency in radians/second. OUTPUT ADC DRIVER BASEBAND INPUTS BASEBAND OUTPUTS Note that for a frequency scaled filter prototype, the absolute magnitude of the group delay scales inversely with the bandwidth; however, the shape is retained. For example, the peak group delay for a 28 MHz bandwidth setting is 14× less than for a 2 MHz setting. PREAMP GAIN SWITCH FILTER PROGRAMMING SPI BUS ANALOG GAIN CONTROL 30mV/dB OUTPUT COMMON-MODE CONTROL 09002-042 SPI INTERFACE 0 –20 Filtering and amplification are fundamental operations in any signal processing system. Filtering is necessary to select the intended signal while rejecting out-of-band noise and interferers. Amplification increases the level of the desired signal to overcome noise added by the system. When used together, filtering and amplification can extract a low level signal of interest in the presence of noise and out-of-band interferers. Such analog signal processing alleviates the requirements on the analog, mixed signal, and digital components that follow. RELATIVE MAGNITUDE (Hz) Figure 41. Signal Path Block Diagram for a Single Channel of the ADRF6510 –40 –60 –80 –100 –120 –140 –160 –180 1M INPUT BUFFERS 10M 100M 1G FREQUENCY (Hz) (VCM – (VPS/2))/200 Ω The input buffers in both channels can be configured simultaneously to a gain of 6 dB or 12 dB through the GNSW pin. When configured for a 6 dB gain, the buffers support up to a 1 V p-p differential input level with >50 dBc harmonic distortion. For a 12 dB gain setting, the buffers support 0.5 V p-p inputs. Figure 42. Sixth-Order Butterworth Magnitude Response for 0.5 dB Bandwidths; Programmed from 2 MHz to 29 MHz in 1 MHz Steps 500 400 2MHz GROUP DELAY (ns) The input buffers provide a convenient interface to the sensitive filter sections that follow. They set a differential input impedance of 400 Ω and sit at a nominal common-mode voltage of VPS/2. The inputs can be dc-coupled or ac-coupled. If using direct dc-coupling, the common-mode voltage, VCM, can range from 1.5 V to 3 V. A current flows into or out of the input pins to accommodate the difference in common-mode voltages. The current into each pin is given by 09002-043 1MHz TO 30MHz PROG. FILTERS τg = −∂φ/∂ω PROGRAMMABLE FILTERS 28MHz 300 200 14x 100 0 –100 100k 1M 10M 100M FREQUENCY (Hz) The integrated programmable filter is the key signal processing function in the ADRF6510. The filters follow a six-pole Butterworth prototype response that provides a compromise between Rev. 0 | Page 14 of 28 Figure 43. Sixth-Order Butterworth Group Delay Response for 0.5 dB Bandwidths; Programmed to 2 MHz and 28 MHz 09002-044 6dB/12dB PREAMP The filters are designed so that the Butterworth prototype filter shape and group delay responses vs. frequency are retained for any bandwidth setting. Figure 42 and Figure 43 illustrate the ideal six-pole Butterworth gain and group delay responses, respectively. The group delay, τg, is defined as ADRF6510 The corner frequency of the filters is defined by RC products, which can vary by ±30% in a typical process. Therefore, all the parts are factory calibrated for corner frequency, resulting in a residual ±10% corner frequency variation over the −40°C to +85°C temperature range. Although absolute accuracy requires calibration, the matching of RC products between the pair of channels is better than 1% by observing careful design and layout practices. Calibration and excellent matching ensure that the magnitude and group delay responses of both channels track together, a critical requirement for digital IQ-based communication systems. VARIABLE GAIN AMPLIFIERS (VGAs) The VGAs are implemented using the Analog Devices, Inc., patented X-AMP® architecture, consisting of a tapped 50 dB attenuator followed by a fixed-gain amplifier. The X-AMP architecture generates a linear-in-dB monotonic gain response with low ripple. The gain is controlled through the high impedance GAIN pin with an accurate slope of 30 mV/dB. The gain response shown in Figure 44 shows the GAIN pin voltage range and the absence of gain foldback at high VGAIN. 50 DC OFFSET COMPENSATION LOOP In many signal processing applications, no information is carried in the dc level. In fact, dc voltages and other low frequency disturbances can often dominate the intended signal and consume precious dynamic range in the analog path and bits in the data converters. These dc voltages can be present with the desired input signal or can be generated inside the signal path by inherent dc offsets or other unintended signaldependent processes such as self-mixing or rectification. Because the ADRF6510 is fully dc-coupled, it may be necessary to remove these offsets to realize the maximum signal-to-noise ratio (SNR). This can be achieved with ac-coupling capacitors at the input and output pins, but that would require large values because the impedances are fairly low, and high-pass corners may need to be <10 Hz in some cases. To address the issue of dc offsets, the ADRF6510 provides an offset correction loop that nulls the output differential dc level as shown in Figure 45. If the correction loop is not required, it can be disabled through the OFDS pin. COFS OFDS 0.3 OFSx 0.2 30 0.1 0 20 –0.1 10 BASEBAND OUTPUT FROM FILTERS 50dB VGA OUTPUT ADC DRIVER GAIN Figure 45. Offset Compensation Loop Operates Around the VGA and Output Buffer –10 0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 –0.3 4.0 VGAIN (V) 09002-045 –0.2 0 09002-066 40 GAIN ERROR (dB) GAIN (dB) 30mV/dB Figure 44. Linear-in-dB Gain Control Response of the X-Amp VGA Showing Consistent Slope and Low Error OUTPUT BUFFERS/ADC DRIVERS The low impedance (20 Ω) output buffers of the ADRF6510 are designed to drive either ADC inputs or subsequent amplifier stages. They are capable of delivering up to 4 V p-p composite two-tone signals into 500 Ω differential loads with >60 dBc IM3. The output common-mode voltage defaults to VPS/2, but it can be adjusted from 1.5 V to 3.0 V without loss of drive capability by presenting the VOCM pin with the desired common-mode voltage. The high input impedance of VOCM allows the ADC reference output to be connected directly. Even though the signal path is fully dc-coupled and the dc offset compensation loop can remove undesired dc offsets (see the DC Offset Compensation Loop section), the output buffers can be accoupled to the next stage by properly selecting the coupling capacitors according to the load impedance. The offset control loop creates a high-pass corner, fHP, that is superimposed on the normal Butterworth filter response. Typically, fHP is many orders of magnitude lower than the lower programmed filter bandwidth so that there is no interaction between them. Setting fHP is accomplished with capacitors, COFS, from the OFS1 and OFS2 pins to ground. Because the correction loop works around the VGA section, fHP is also dependent on the gain of the VGA. In general, the expression for fHP is given by fHP (Hz) = 1.2 × (Gain/COFS) where: Gain is expressed in linear terms, not in decibels (dB). COFS is expressed in microfarads (μF). Note that fHP increases in proportion to the gain. For this reason, COFS should be chosen at the highest operating gain to guarantee that fHP is always below the maximum limit required by the system. Rev. 0 | Page 15 of 28 ADRF6510 PROGRAMMING THE FILTERS NOISE CHARACTERISTICS The 0.5 dB corner frequencies for both filters are programmed simultaneously through the SPI port. A 5-bit register stores the codes for corner frequencies of 1 MHz through 30 MHz (see Table 4). The SPI protocol not only allows frequency codes to be written to the DATA pin but also allows the stored code to be read back from the SDO pin. The output noise behavior of the ADRF6510 depends on the gain and bandwidth settings. Both the filter sections and the VGAs contribute to the total noise at the output. The filter contributes a noise spectral density profile that is flat at low frequencies, peaks near the corner frequency, and then rolls off as the filter poles roll off the gain. The magnitude of the noise spectral density, expressed in nV/√Hz, varies inversely with the square root of the bandwidth setting, resulting in a total integrated noise in nV that is nearly constant with bandwidth setting. The latch enable (LE) pin must first go to a Logic 0 for a read or write cycle to begin. On the next rising edge of the clock (CLK), a Logic 1 on the DATA pin initiates a write cycle, whereas a Logic 0 on the DATA pin initiates a read cycle. In a write cycle, the next five CLK rising edges latch the frequency code, LSB first. When LE goes high, the write cycle is completed and the frequency code is presented to the filter. In a read cycle, the next five CLK falling edges present the stored frequency code, LSB first. When LE goes high, the read cycle is completed. Detailed timing diagrams are shown in Figure 2 and Figure 3. Table 4. Frequency Code vs. Corner Frequency Lookup Table 1 –115 BANDWIDTH = 20MHz GAIN = 40dB –120 –125 GAIN = 20dB –130 GAIN = 0dB –135 10 15 20 25 30 35 40 45 50 55 60 FREQUENCY (MHz) Figure 46. Total Output Noise with a 20 MHz Corner Frequency for Three Different Gain Settings MSB first. Rev. 0 | Page 16 of 28 09002-046 Corner Frequency (MHz) 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 30 30 Figure 46 and Figure 47 show the total output noise spectral density vs. frequency for different bandwidth settings. At low values of VGA gain, the noise at the output is the flat spectral density contributed by the VGA because the filter noise is suppressed by the VGA attenuation. As the gain increases, more of the filter noise appears at the output. Because the filter noise increases at lower bandwidth settings, it overwhelms the VGA noise floor. In either case, the noise density asymptotically approaches the −130 dBV/√Hz limit set by the VGA at the highest frequencies. For other values of VGA gain and bandwidth setting, the detailed shape of the noise spectral density changes. OUTPUT NOISE (dBV/√Hz) 5-Bit Binary Frequency Code1 00000 00001 00010 00011 00100 00101 00110 00111 01000 01001 01010 01011 01100 01101 01110 01111 10000 10001 10010 10011 10100 10101 10110 10111 11000 11001 11010 11011 11100 11101 11110 11111 The X-AMP type VGAs used in the ADRF6510 contribute a fixed noise spectral density to the output, independent of the gain setting, of −130 dBV/√Hz, which is equivalent to 316 nV/√Hz. Although the VGA noise contribution to the output is fixed, the gain of the VGA controls the relative contribution of the filter noise. ADRF6510 –100 To distinguish and quantify the distortion performance of the input section, two different IP3 specifications are presented. The first is called in-band IP3 and refers to a two-tone test where the signals are inside the filter bandwidth. This is exactly the same figure of merit familiar to communications engineers in which the third-order intermodulation level, IM3, is measured. BANDWIDTH = 1MHz OUTPUT NOISE (dBV/√Hz) –105 GAIN = 40dB –110 –115 –120 GAIN = 20dB –125 –130 1.0 1.5 2.0 2.5 3.0 FREQUENCY (MHz) 09002-047 GAIN = 0dB –135 0.5 Figure 47. Total Output Noise with a 1 MHz Corner Frequency for Three Different Gain Settings Note that the noise spectral density outside the filter bandwidth is limited by the fixed VGA output noise. It may be necessary to use an external, fixed-frequency, passive filter prior to an analogto-digital conversion to prevent noise aliasing from degrading the signal-to-noise ratio. The higher the sampling rate relative to the maximum ADRF6510 corner frequency setting to be used, the lower the order of the external filter. DISTORTION CHARACTERISTICS The distortion performance of the ADRF6510 is similar to its noise performance. The filters and the VGAs contribute to the overall distortion and signal handling capabilities. Furthermore, the front end must also cope with out-of-band signals that can be larger than the in-band signals. These out-of-band signals are filtered before reaching the VGA. It is important to understand the signals presented to the ADRF6510 and to match these signals with the input and output characteristics of the part. When the gain is low, the distortion is typically limited by the input section because the output is not driven to its maximum capacity. When the gain is high, the distortion is likely limited by the output section because the input is not driven to its maximum capacity. An exception to this is when the input is driven with a small desired signal in combination with a large out-of-band signal. In this case, the out-of-band signal may drive the input to distort. As long as the input is not overdriven, the out-of-band signal is removed by the filter. A high VGA gain is still needed to raise the small desired signal to a higher level at the output. The overall distortion introduced by the part depends on the input drive level, including the out-of-band signals, and the desired output signal level. As noted in the Input Buffers section, the input section can handle a total signal level of 1 V p-p for a 6 dB preamplifier and 500 mV p-p for a 12 dB preamplifier with >50 dBc harmonic distortion. This includes both in-band and out-of-band signals. To quantify the effect of out-of-band signals, a new out-of-band (OOB) IIP3 figure of merit is introduced. This test also involves a two-tone stimulus; however, the two tones are placed out-ofband so that the lower IM3 product lands in the middle of the filter pass band. At the output, only the IM3 product is visible because the original two tones are filtered out. To calculate the OOB IP3 at the input, the IM3 level is referred to the input by the overall gain. The OOB IIP3 allows the user to predict the impact of out-of-band blockers or interferers at an arbitrary signal level on the in-band performance. The ratio of the desired input signal level to the input-referred IM3 at a given blocker level represents a signal-to-distortion limit imposed by the out-of-band signals. MAXIMIZING THE DYNAMIC RANGE The role of the ADRF6510 is to increase the level of a variable in-band signal while minimizing out-of-band signals. Ideally, this is achieved without degrading the SNR of the incoming signal or introducing distortion to the incoming signal. The first goal is to maximize the output signal swing, which can be defined by the ADC input range or the input signal capacity of the next analog stage. For the complex waveforms often encountered in communication systems, the peak-to-average ratio, or crest factor, must be considered when choosing the peak-to-peak output. From the chosen output signal and the maximum gain of the ADRF6510, the minimum input level can be defined. Lower signal levels do not yield the maximum output and suffer a greater degradation in SNR. As the input signal level increases, the VGA gain is reduced from its maximum gain point to maintain the desired fixed output level. The output noise, initially dominated by the filter, follows the gain reduction, yielding a progressively better SNR. At some point, the VGA gain drops sufficiently that the constant VGA noise becomes dominant, resulting in a constant SNR from that point. From the perspective of SNR alone, the maximum input level is reached when the VGA reaches its minimum gain. Distortion must also be considered when maximizing the dynamic range. At low and moderate signal levels, the output distortion is constant and assumed to be adequate for the selected output level. At some point, the input signal becomes large enough that distortion at the input limits the system. The maximum tolerable input signal depends on whether the input distortion becomes unacceptably large or the minimum gain is reached. Rev. 0 | Page 17 of 28 ADRF6510 The most challenging scenario in terms of dynamic range is the presence of a large out-of-band blocker accompanying a weaker in-band wanted signal. In this case, the maximum input level is dictated by the blocker and its inclination to cause distortion. After filtering, the weak wanted signal must be amplified to the desired output level, possibly requiring maximum gain. Both the distortion limits associated with the blocker at the input and the SNR limits created by the weaker signal and higher gains are present simultaneously. Furthermore, not only does the blocker scenario degrade the dynamic range, it also reduces the range of input signals that can be handled because a larger part of the gain range is used to simply extract the weak desired signal from the stronger blocker. KEY PARAMETERS FOR QUADRATURE-BASED RECEIVERS The majority of digital communication receivers makes use of quadrature signaling, in which bits of information are encoded onto pairs of baseband signals that then modulate in-phase (I) and quadrature (Q) sinusoidal carriers. Both the baseband and modulated signals appear quite complex in the time domain with dramatic peaks and valleys. In a typical receiver, the goal is to recover the pair of quadrature baseband signals in the presence of noise and interfering signals after quadrature demodulation. In the process of filtering out-of-band noise and unwanted interferers and restoring the levels of the wanted I and Q baseband signals, it is critical to retain their gain and phase integrity over the bandwidth. The ADRF6510 delivers flat in-band gain and group delay, consistent with a six-pole Butterworth prototype filter as described in the Programmable Filters section. Furthermore, careful design ensures excellent matching of these parameters between the I and Q channels. Although absolute gain flatness and group delay can be corrected with digital equalization, mismatch introduces quadrature errors and intersymbol interference that degrade bit error rates in digital communication systems. Rev. 0 | Page 18 of 28 ADRF6510 APPLICATIONS INFORMATION BASIC CONNECTIONS ERROR VECTOR MAGNITUDE (EVM) PERFORMANCE Figure 48 shows the basic connections for operating the ADRF6510. A voltage from 4.75 V to 5.25 V should be applied to the supply pins. Each supply pin should be decoupled with at least one low inductance, surface-mount ceramic capacitor of 0.1 μF placed as close as possible to the device. Error vector magnitude (EVM) is a measure used to quantify the performance of a digital radio transmitter or receiver. A signal received by a receiver has all constellation points at their ideal locations; however, various imperfections in the implementation (such as magnitude imbalance, noise floor, and phase imbalance) cause the actual constellation points to deviate from their ideal locations. The input buffers provide an interface to the sensitive filter sections that follow. They set a differential input impedance of 400 Ω and sit at a nominal common-mode voltage of VPS/2. The inputs can be dc-coupled or ac-coupled. If using direct dc-coupling, the common-mode voltage, VCM, can range from 1.5 V to 3 V. In general, a receiver exhibits three distinct EVM limitations vs. received input signal power. • The output buffers of the ADRF6510 are low impedance (~20 Ω) designed to drive either ADC inputs or subsequent amplifier stages. The output common-mode voltage defaults to VPS/2 but can be adjusted from 1.5 V to 3.0 V without loss of drive capability by presenting the VOCM pin with the desired common-mode voltage. The high input impedance of VOCM allows the ADC reference output to be connected directly. At strong signal levels, the distortion components falling in-band due to nonlinearities in the device components cause strong degradation to EVM as signal levels increase. At medium signal levels, where the signal chain behaves in a linear manner and the signal is well above any notable noise contributions, EVM has a tendency to reach an optimum level determined dominantly by the quadrature accuracy and the precision of the test equipment. As signal levels decrease such that noise is a major contribution, the EVM performance vs. the signal level exhibits a decibelfor-decibel degradation with decreasing signal level. At lower signal levels, where noise proves to be the dominant limitation, the decibel EVM proves to be directly proportional to the SNR. • To enable the ADRF6510, the ENBL pin must be pulled high. Taking ENBL low disables the device, reducing current consumption to approximately 2 mA at ambient temperature. • INPUT1(–) 0.1µF VPS VPS INPUT1(+) SERIAL CONTROL INTERFACE VPS INPUT2(+) VPSD 0.1µF VPS 0.1µF OUTPUT1(+) ENBL INM1 COM OFS1 INP1 VPS GNSW VPS OPP1 VPSD OPM1 COMD OUTPUT1(–) COM LE CLK ADRF6510 GAIN DATA VOCM SDO COM 0.1µF COM OPM2 VPS VPS 0.1µF 0.1µF OUTPUT2(–) OPP2 VPS OFDS VPS INP2 OFS2 COM INM2 COM VPS 0.1µF R2 OUTPUT2(+) VPS 0.1µF VPS 09002-050 INPUT2(–) Figure 48. Basic Connections Rev. 0 | Page 19 of 28 ADRF6510 32 • To the right of the 5 MHz center frequency, the filter is programmed to be 5 MHz greater than the centered baseband frequency. This ensures that the signal edge is well within the pass band of the filter. In such cases, where the filter bandwidth is set to be greater than the signal bandwidth, the image rejection of the ADL5387 tends to be the limiting factor, and the ADRF6510 has minimal effects. To the left of 5 MHz—specifically at a center frequency of 3.5 MHz—the filter corner is lowered to the baseband signal edge, degrading the image rejection. When the centered baseband frequency is 3.5 MHz and the filter is set at 7 MHz (instead of a safer 8 MHz), the filter corner conflicts with the edge of the modulated signal. Channel mismatch in group delay characteristics and variation in absolute group delay (from the normal flat response) tend to degrade image rejection. 2.5 2.0 28 1.5 26 1.0 24 0.5 22 ANALOG GAIN VOLTAGE (V) • 30 EVM (dB) Figure 50 shows the image rejection of the ADL5387 and the ADRF6510 for various baseband frequencies. The modulation is 64 QAM with a 7 MHz symbol rate. Note the following: 45 –55 –45 –35 –25 –15 RF INPUT POWER (dBm) –5 AT CENTER FREQUENCY = 3.5MHz, FILTER CORNER IS SET TO 7MHz 40 35 IMAGE REJECTION (dB) 0 09002-058 20 –65 Figure 49. EVM vs. RF Input Power Level LOW IF IMAGE REJECTION The image rejection ratio is the ratio of the intermediate frequency (IF) signal level produced by the desired input frequency to that produced by the image frequency. The image rejection ratio is expressed in decibels (dB). Appropriate image rejection is critical because the image power can be much higher than that of the desired signal, thereby plaguing the downconversion process. 30 25 AT CENTER FREQUENCIES GREATER THAN 5MHz, FILTER CORNER IS SET TO 5MHz HIGHER THAN SIGNAL EDGE 20 15 10 5 0 3 Figure 51 illustrates the image problem. If the upper sideband is the desired band, a 90° shift to the Q channel cancels the image at the lower sideband. In the same way, if the lower sideband is the desired band, a 90° shift to the I channel cancels the image at the upper sideband. Phase and gain balance between the I and Q channels are critical for high levels of image rejection. 5 7 9 11 13 BASEBAND FREQUENCY (MHz) Figure 50. Image Rejection of the ADL5387 and ADRF6510 COSωLOt 0° ωIF ωIF –ωIF 0 +ωIF –90° 0 +ωIF 0 +ωIF +90° ωLO ωUSB 0° –ωIF 0 +ωIF 09002-062 ωLSB SINωLOt Figure 51. Illustration of the Image Problem Rev. 0 | Page 20 of 28 15 09002-059 An overall RF-to-baseband EVM performance was obtained with the ADL5387 IQ demodulator preceding the ADRF6510. An 840 MHz RF signal with a modulation setting of 64 QAM and a 7 MHz symbol rate was used. The local oscillator (LO) is set at 1680 MHz to obtain 840 MHz at the quadrature core after the divide-by-2 stage. The analog gain of the ADRF6510 was adjusted to maintain 1.5 V p-p into a 1 kΩ load impedance. Figure 49 shows EVM vs. input power and the corresponding analog gain voltage. ADRF6510 ETC1-1-13 RFC VPS 120nH VPS VPOS VPS 100pF 2 COM VPOS 0.1µF VPB 18 VPB 17 100pF 3 BIAS 4 VPL 5 VPL CML 7 8 9 10 11 12 COM CML ILO 13 LOIN 6 VPL IHI 14 LOIP 0.1µF 100pF QLO 15 CML VPOS QHI 16 ADL5387 VPS 0.1µF 19 SERIAL CONTROL INTERFACE 20 VPX 21 CMRF RFIP 22 RFIN 23 CMRF 1 VPA 0.1µF CMRF 24 VPSD 0.1µF ENBL INM1 COM OFS1 INP1 VPS GNSW VPS OPP1 OPM1 COMD VPSD COM LE CLK ADRF6510 DATA GAIN 0.1µF VPS COM 0.1µF OPM2 OPP2 VPS INP2 VPS OFDS VPS COM INM2 COM OFS2 COM VPS 0.1µF R2 1000pF 0.1µF VOCM SDO 1000pF VECTOR SIGNAL ANALYZER/ADC 1000pF 1000pF 120nH LO VPS 0.1µF 09002-063 VPS Figure 52. ADL5387 and ADRF6510 Interfacing Example—Block Diagram The noise spectral density of the ADRF6510 outside the filter bandwidth is limited by the fixed VGA output noise. It may be necessary to use an external, fixed-frequency, passive filter prior to an analog-to-digital conversion to prevent noise aliasing from degrading the signal-to-noise ratio. As shown in Figure 46 and Figure 47, the noise density at higher frequencies tends to be flat, and any higher IF noise aliasing into the Nyquist zone has minimal effects. Using the AD9639, a 12-bit ADC with a 210 MSPS sampling rate, the effects of an antialiasing filter present between the ADRF6510 and the ADC showed a minimal 1.5 dB improvement. When designing an antialiasing filter, it is necessary to consider the overall source and load impedance presented by the ADRF6510 and the ADC input to design the filter network. The differential baseband output impedance of the ADRF6510 is 20 Ω and is designed to drive a high impedance ADC input. It may be desirable to terminate the ADC input to a lower impedance by using a terminating resistor, such as 500 Ω. The terminating resistor helps to better define the input impedance at the ADC input at the cost of a slightly reduced gain. The order and type of filter network depend on the desired high frequency rejection required, the pass-band ripple, and the group delay. Filter design tables provide outlines for various filter types and orders, illustrating the normalized inductor and capacitor values for a 1 Hz cutoff frequency and 1 Ω load. After scaling the normalized prototype element values by the actual desired cutoff frequency and load impedance, the series reactance elements are halved to realize the final balanced filter network component values. As an example, a second-order Butterworth, low-pass filter design is shown in Figure 53 where the differential load impedance is 500 Ω and the source impedance is 50 Ω. The normalized series inductor value for the 10-to-1, load-to-source impedance ratio is 0.074 H, and the normalized shunt capacitor is 14.814 F. For a 10.9 MHz cutoff frequency, the single-ended equivalent circuit consists of a 0.54 μH series inductor followed by a 433 pF shunt capacitor. The balanced configuration is realized as the 0.54 μH inductor is split in half to achieve the network that is shown in Figure 53. RS = 50Ω LN = 0.074H NORMALIZED SINGLE-ENDED CONFIGURATION VS CN 14.814F RS = 0.1Ω RL RS = 50Ω fC = 1Hz 0.54µH DENORMALIZED SINGLE-ENDED EQUIVALENT VS RL = 500Ω 433pF RL = 500Ω fC = 10.9MHz RS = 25Ω 2 0.27µH BALANCED CONFIGURATION VS RS = 25Ω 2 433pF RL 2 = 250Ω RL = 250Ω 2 0.27µH Figure 53. Second-Order Butterworth, Low-Pass Filter Design Example Rev. 0 | Page 21 of 28 09002-061 EXAMPLE BASEBAND INTERFACE ADRF6510 10 5 0 –5 –10 GAIN (dB) A complete design example is shown in Figure 54. A third-order Chebyshev differential filter with a 31 MHz corner frequency interfaces the output of the ADRF6510 to that of an ADC input. The 20 Ω source impedance reflects the impedance of the output buffer stage. The 500 Ω load resistor defines the input impedance of the ADC. The filter adheres to a 0.1 dB in-band flatness and offers sufficient out-of-band rejection to act as an antialiasing filter. –15 –20 –25 –30 1.8µH 1µH –40 500Ω 1.8µH 1µH 09002-064 –50 0 20 40 60 80 100 120 140 160 180 200 BASEBAND FREQUENCY (MHz) 09002-048 –45 Figure 55. Third-Order Baseband Filter Response Figure 54. Third-Order Chebyshev Differential Filter Design Example 20 Figure 55 and Figure 56 show the measured frequency response and group delay of the third-order Chebyshev differential filter. 18 GROUP DELAY (ns) 16 14 12 10 8 6 4 2 0 0 20 40 60 80 100 120 140 160 180 200 220 240 260 280 300 BASEBAND FREQUENCY (MHz) Figure 56. Third-Order Baseband Filter Group Delay Response Rev. 0 | Page 22 of 28 09002-049 100pF 20Ω –35 ADRF6510 EVALUATION BOARD The ADRF6510 evaluation board is available with software control to program the filter bandwidth. It is a 4-layer board with split ground plane for analog and digital sections. Special care is taken to place the power decoupling capacitors close to the device pins. The board is designed for easy single-ended (through a Mini-Circuits ADT8-1T+ 8:1 balun) or differential configuration for each channel. EVALUATION BOARD CONTROL SOFTWARE 09002-051 The ADRF6510 evaluation board is configured with a USBfriendly interface to program the filter bandwidth of the ADRF6510. The software GUI (see Figure 57) allows users to select a particular frequency to write to the device and also to read back data from the SDO pin that shows the currently programmed filter setting. The software setup files can be downloaded from the ADRF6510 product page at www.analog.com. Figure 57. Evaluation Control Software SCHEMATICS AND ARTWORK VPS INM1_SE_P R4 R45 R17 P3 R47 R31 R55 T1 C6 R57 C12 C9 VPS R58 C10 R56 R48 INP1 C14 VPS R43 R18 VPSD C4 C28 R30 SDO R1 INP2_SE_P VPS ADRF6510 C20 GAIN DATA VOCM SDO COM COM OPM2 VPS R8 R6 C18 OPP2 VPS OFDS VPS INP2 OFS2 COM INM2 COM R2 R46 R15 T3 R7 R11 COM CLK C5 C19 OPM1 COMD LE R33 R34 ENBL INM1 COM OFS1 INP1 VPS GNSW VPS OPP1 VPSD R51 C3 INM2 R50 R23 C17 R9 R44 R16 R54 C8 VPS C11 R38 R42 R35 R14 C22 R10 T4 R13 OPM2_SE_P C24 OPP2 R36 C13 R53 C7 R22 R40 VPSD VPS L2 R3 R52 OPM1 R21 R39 0.1µF 0.1µF C21 VPS R49 T2 R20 R5 C15 R32 C23 VPS C1 P4 L1 Figure 58. Evaluation Board Schematic Rev. 0 | Page 23 of 28 C2 09002-065 CLK DATA R12 R29 C27 R41 R19 C16 P2 LE OPP1_SE_P R24 R37 VPS ADRF6510 XC11 3.3V 10PF XC6 3.3V 22PF DGND 3.3V XC13 0.1UF 3.3V XR1 2K 8 VCC XC14 0.1UF XC15 0.1UF XC16 0.1UF XC17 0.1UF XC18 0.1UF XC19 0.1UF XU2 1 A0 2 A1 3 A2 6 SCLSDA 7 WC_N GND 4 DGND XR2 2K 5 24LC64-I-SN XY1 DGND 1 4 XC1 2 3 XC10 XU3 ADP3303ARZ-3.3 10PF XC7 XC2 22PF 22PF XC9 24MHZ 21 22 23 XR4 100K XC3 0.1UF DGND 0.1UF DGND XC4 0.1UF DGND 50 51 52 PAD USB_POWER 1 2 43 55 11 27 32 XTALOUT PAD AGND OUT OUT1 ERR_N SD* GND 5 4 +5V 3 1 2 6 0.1UF XC12 3.3V 0.1UF DGND DGND LNJ308G8TRA(GREEN) XJ1 XU1 XTALIN 5 4 VCC 10PF XC8 8 DPLUS 9 DMINUS CLKOUT/PE1/T1OUT 54 13 IFCLK/T0OUT 29 CTL0/FLAGA 30 CTL1/FLAGB 31 CTL2/FLAGC GND DGND 3 4 5 6 7 8 SHIELD PINS 9 CY7C68013A-56LFXC DGND UX60A-MB-5ST240-0003-4 DGND USB_POWER CY7C68013A-56LFXC 09002-026 XR6 0 +5V Figure 59. Schematic for the USB Section of the Evaluation Board 09002-053 09002-052 PD7/FD15 17 AVCC 15 SCL 16 SDA 42 RESET* 44 WAKEUP 1 RDY0/SLRD 2 RDY1/SLWR 14 RESERVED XR5 100K 24 25 PD0/FD8 45 PD1/FD9 46 PD2/FD10 47 PD3/FD11 48 PD4/FD12 49 PD5/FD13 PD6/FD14 XD1 53 56 PB4/FD4 PB5/FD5 PB6/FD6 PB7/FD7 3.3V 12 26 28 41 PB3/TXD1/FD3 3.3V 7 IN 8 IN1 1K SDO DATA CLK LE 7 3 36 37 PA5/FIFOADR1 38 PA6/PKTEND 39 40 PA7/FLAGD/SLCS* PB0/FD0 18 PB1/FD1 19 PB2/FD2 20 6 33 34 35 XR3 DGND AC PA1/INT1* PA2/SLOE PA3/WU2 PA4/FIFOADR0 10 PA0/INT0* DGND 0.1UF DGND XU1 XC5 NR 3.3V Figure 60. Top Layer Silkscreen Figure 61. Component Side Layout Rev. 0 | Page 24 of 28 ADRF6510 EVALUATION BOARD CONFIGURATION OPTIONS Table 5 lists the components of the main section of the ADRF6510 evaluation board. Table 5. Components C1, C2, C4, C5, C11, C12, C15, C16, L1, L2, R2 Function Power supply and ground decoupling. Nominal supply decoupling consists of a 0.1 μF capacitor to ground. T1, T2, C3, C6, C7 to C10, R15 to R18, R31, R32, R43 to R58 Input interface. Input SMAs INM1_SE_P and INP2_SE_P are used to drive the baluns in a single-ended fashion. The default configuration of the evaluation board is for single-ended operation. T1 and T2 are 8:1 impedance ratio baluns to transform a 50 Ω, single-ended input into a 400 Ω balanced differential signal. R31, R32, and R47 to R50 are populated for appropriate balun interface. R51 to R58 are provided for generic placement of matching components. C3 and C6 are balun decoupling capacitors. R15 to R18 and R43 to R46 can be populated with 0 Ω, and the balun interfacing resistors can be removed to bypass T1 and T2 for differential interfacing. C7 to C10 can be used for ac coupling with differential interfacing. Output interface. Output SMAs OPP1_SE_P and OPM2_SE_P are used to drive the baluns in a single-ended fashion. The default configuration of the evaluation board is for single-ended operation. T3 and T4 are 8:1 impedance ratio baluns to transform a 50 Ω, single-ended output into a 400 Ω balanced differential load. R19, R20, R35, R36, R41, and R42 are populated for appropriate balun interface. R7 to R14 are provided for generic placement of matching components. R7 to R10 are set to 300 Ω to present a 1 kΩ load (with the balun used) at the DUT output. C19 to C22 are used for ac coupling when differential outputs are used. C23 and C24 are balun decoupling capacitors. R21 to R24 and R37 to R40 can be populated with 0 Ω, and the balun interfacing resistors can be removed to bypass T3 and T4 for differential interfacing. Enable interface. The ADRF6510 is powered up by applying a logic high voltage to the ENBL pin (Jumper P2 is connected to VPS). Serial interface control. The digital interface sets the corner frequency of the device using the serial interface via the LE, CLK, DATA, and SDO pins. T3, T4, C19 to C24, R7 to R14, R19 to R24, R35 to R42 P2 C27, C28, R1, R29, R30, R33, R34 P4, C13, C14, R3 DC offset correction loop compensation. The dc offset correction loop is enabled (low) with Jumper P4. When enabled, the capacitors are connected to circuit common. The high-pass corner frequency is expressed as follows: fHP (Hz) = 1.2 × ((Linear Gain)/COFS (μF)). C18, R6 C17, R5 Output common-mode setpoint. The output common-mode voltage can be set externally when applied to the VOCM pin. If the VOCM pin is left open, the output common-mode voltage defaults to VPS/2. Analog gain control. 0 V to 2 V, 30 mV/dB gain scaling. P3, R4 Front-end 6 dB or 12 dB gain switch. Pull low for 6 dB; pull high for 12 dB. Rev. 0 | Page 25 of 28 Default Conditions C1, C2 = 10 μF (Size 0603) C4, C5, C11, C12, C15, C16 = 0.1 μF (Size 0603) L1, L2 = 33 μH (Size 0805) R2 = 1 kΩ (Size 0402) T1, T2 = ADT8-1T+ (Mini-Circuits) C3, C6 = 0.1 μF (Size 0402) C7 to C10 = 100 nF (Size 0402) R15 to R18, R43 to R46, R51, R52, R55, R56 = open (Size 0402) R31, R32, R47 to R50, R53, R54, R57, R58 = 0 Ω (Size 0402) T3, T4 = ADT8-1T+ (Mini-Circuits) C19 to C22 = 100 nF (Size 0402) C23, C24 = 0.1 μF (Size 0402) R7 to R10 = 300 Ω (Size 0402) R11 to R14 = open R19, R20, R35, R36, R41, R42 = 0 Ω (Size 0402) R21 to R24, R37 to R40 = open (Size 0402) P2 = installed for enable R1 = 10 kΩ (Size 0402) C27, C28 = 330 pF (Size 0402) R29, R30 = 100 Ω (Size 0402) R33, R34 = 0 Ω (Size 0402) P4 = installed C13, C14 = 1000 pF (Size 0402) R3 = 10 kΩ (Size 0402) C18 = 0.1 μF (Size 0402) R6 = 0 Ω (Size 0402) C17 = 0.1 μF (Size 0402) R5 = 0 Ω (Size 0402) P3 = installed R4 = 10 kΩ (Size 0402) ADRF6510 USB Section Configuration Options Table 6 lists the components of the USB section of the ADRF6510 evaluation board. Table 6. Components XC1, XC2, XC6 XC3 to XC5, XC7, XC8, XC12 to XC19 XC9 to XC11 XD1 XJ1 XR1, XR2 XR3 XR4, XR5 XR6 XU1 XU2 XU3 XY1 Default Conditions 22 pF (Size 0603) 0.1 μF (Size 0402) 10 pF (Size 0402) Green LED ( Panasonic LNJ308G8TRA) USB SMT connector (Hirose Electric UX60A-MB-5ST 240-0003-4) 2 kΩ (Size 0603) 1 kΩ (Size 0603) 100 kΩ (Size 0603) 0 Ω (Size 0603) USB microcontroller (Cypress CY7C68013A-56LFXC) 64 kb EEPROM (Microchip 24LC64-I/SN) Low dropout regulator (Analog Devices ADP3303ARZ-3.3) 24 MHz crystal oscillator (AEL Crystals X24M000000S244) Rev. 0 | Page 26 of 28 ADRF6510 OUTLINE DIMENSIONS 0.60 MAX 5.00 BSC SQ 0.60 MAX PIN 1 INDICATOR 0.50 BSC 4.75 BSC SQ 0.50 0.40 0.30 12° MAX 17 16 0.80 MAX 0.65 TYP 0.30 0.23 0.18 3.25 3.10 SQ 2.95 EXPOSED PAD (BOTTOM VIEW) 9 8 0.25 MIN 3.50 REF 0.05 MAX 0.02 NOM SEATING PLANE 1 0.20 REF COPLANARITY 0.08 FOR PROPER CONNECTION OF THE EXPOSED PAD, REFER TO THE PIN CONFIGURATION AND FUNCTION DESCRIPTIONS SECTION OF THIS DATA SHEET. COMPLIANT TO JEDEC STANDARDS MO-220-VHHD-2 011708-A TOP VIEW 1.00 0.85 0.80 PIN 1 INDICATOR 32 25 24 Figure 62. 32-Lead Lead Frame Chip Scale Package [LFCSP_VQ] 5 mm × 5 mm Body, Very Thin Quad (CP-32-2) Dimensions shown in millimeters ORDERING GUIDE Model 1 ADRF6510ACPZ-R7 ADRF6510ACPZ-WP ADRF6510-EVALZ 1 Temperature Range −40°C to +85°C −40°C to +85°C Package Description 32-Lead LFCSP_VQ, 7” Tape and Reel 32-Lead LFCSP_VQ, Waffle Pack Evaluation Board Z = RoHS Compliant Part. Rev. 0 | Page 27 of 28 Package Option CP-32-2 CP-32-2 ADRF6510 NOTES ©2010 Analog Devices, Inc. All rights reserved. Trademarks and registered trademarks are the property of their respective owners. D09002-0-4/10(0) Rev. 0 | Page 28 of 28