10-/12-Bit, Low Power, Broadband MxFE AD9961/AD9963 Data Sheet FEATURES FUNCTIONAL BLOCK DIAGRAM Dual 10-bit/12-bit, 100 MSPS ADC SNR = 67 dB, fIN = 30.1 MHz Dual 10-bit/12-bit, 170 MSPS DAC ACLR = 74 dBc 5 channels of analog auxiliary input/output Low power, <425 mW at maximum sample rates Supports full and half-duplex data interfaces Small 72-lead LFCSP lead-free package AUX ADC DLLFILT The high level of integrated functionality, small size, and low power dissipation of the AD9961/AD9963 make them wellsuited for portable and low power applications. AUXIN1 AUX DAC MUX AUX DAC CLKN AUXIO2 AUXIO3 INTERNAL TXCLK TXIQ/TXnRX LPF 1/2/4/8 TXD[11:0] DATA ASSEMBLER LPF 1/2/4/8 TRXCLK TRXIQ LPF 1/2 TRXD[11:0] The AD9961/AD9963 are pin-compatible, 10-/12-bit, low power MxFE(R) converters that provide two ADC channels with sample rates of 100 MSPS and two DAC channels with sample rates to 170 MSPS. These converters are optimized for transmit and receive signal paths of communication systems requiring low power and low cost. The digital interfaces provide flexible clocking options. The transmit is configurable for 1x, 2x, 4x, and 8x interpolation. The receive path has a bypassable 2x decimating low-pass filter. The AD9961 and AD9963 have five auxiliary analog channels. Three are inputs to a 12-bit ADC. Two of these inputs can be configured as outputs by enabling 10-bit DACs. The other two channels are dedicated outputs from two independent 12-bit DACs. DLL AND CLOCK DISTRIBUTION CLKP Wireless infrastructure Picocell, femtocell basestations Medical instrumentation Ultrasound AFE Portable instrumentation Signal generators, signal analyzers LPF 1/2 RESET SDIO SCLK SERIAL PORT LOGIC 12-BIT DAC TXIP 12-BIT DAC TXQP 12-BIT ADC RXIP 12-BIT ADC RXQP AUX DAC DAC12A AUX DAC DAC12B TXIN TXQN RXIN RXQN 08801-001 RXCML RXBIAS LDO VREGs AUXADCREF REFERENCES AND BIAS LDO_EN CS REFIO GENERAL DESCRIPTION TEMPERATURE SENSOR TXCML APPLICATIONS AD9961/AD9963 Figure 1. PRODUCT HIGHLIGHTS 1. High Performance with Low Power Consumption. The DACs operate on a single 1.8 V to 3.3 V supply. Transmit path power consumption is <100 mW at 170 MSPS. Receive path power consumption is <350 mW at 100 MSPS from 1.8 V supply. Sleep and power-down modes are provided for low power idle periods. 2. High Integration. The dual transmit and dual receive data converters, five channels of auxiliary data conversion and clock generation offer complete solutions for many modem designs. 3. Flexible Digital Interface. The interface mates seamlessly to most digital baseband processors. Rev. A 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 (c)2010-2012 Analog Devices, Inc. All rights reserved. AD9961/AD9963 Data Sheet TABLE OF CONTENTS Features .............................................................................................. 1 Transmit DAC Outputs ............................................................. 42 Applications ....................................................................................... 1 Device Clocking.............................................................................. 45 General Description ......................................................................... 1 Clock Distribution ..................................................................... 45 Functional Block Diagram .............................................................. 1 Driving the Clock Input ............................................................ 46 Product Highlights ........................................................................... 1 Clock Multiplication Using the DLL ....................................... 46 Revision History ............................................................................... 2 Configuring the Clock Doublers .............................................. 47 Specifications..................................................................................... 3 Digital Interfaces ............................................................................ 48 Absolute Maximum Ratings ............................................................ 8 TRx Port Operation (Full-Duplex Mode) ............................... 48 Thermal Resistance ...................................................................... 8 Single ADC Mode ...................................................................... 48 ESD Caution .................................................................................. 8 Tx Port Operation (Full-Duplex Mode) ................................. 49 Pin Configurations and Function Descriptions ........................... 9 Half-Duplex Mode ..................................................................... 50 Typical Performance Characteristics ........................................... 13 Auxiliary Converters ...................................................................... 52 Terminology .................................................................................... 18 Auxiliary ADC ............................................................................ 52 Theory of Operation ...................................................................... 19 Conversion Clock ....................................................................... 52 Serial Control Port .......................................................................... 20 Auxiliary DACs........................................................................... 53 General Operation of Serial Control Port ............................... 20 Power Supplies ................................................................................ 55 Sub Serial Interface Communications ..................................... 21 Power Supply Configuration Examples ................................... 55 Configuration Registers ................................................................. 23 Power Dissipation....................................................................... 55 Configuration Register Bit Descriptions ................................. 24 Example Start-Up Sequences ........................................................ 58 Receive Path..................................................................................... 35 Configuring the DLL ................................................................. 58 Receive ADC Operation ............................................................ 35 Configuring the Clock Doublers (DDLL)............................... 58 Decimation Filter and Digital Offset ....................................... 36 Sensing temperature with the AUXADC ................................ 58 Transmit Path .................................................................................. 38 Outline Dimensions ....................................................................... 59 Interpolation Filters.................................................................... 38 Ordering Guide .......................................................................... 59 Transmit DAC Operation .......................................................... 40 REVISION HISTORY 8/12--Rev. 0 to Rev. A Changes to Table 15 ........................................................................ 24 Changes to Figure 65 ...................................................................... 45 Added DLL Duty Cycle Caution Section .................................... 46 Changes to Table 22 ........................................................................ 47 Changes to Figure 93 and Power Supply Configuration Examples Section ............................................................................ 55 Added Example Start-Up Sequences Section ............................. 58 Updated Outline Dimensions ....................................................... 59 7/10--Revision 0: Initial Version Rev. A | Page 2 of 60 Data Sheet AD9961/AD9963 SPECIFICATIONS TMIN to TMAX, RX33V = TXVDD = CLK33V = DRVDD = AUX33V = 3.3 V. All LDOs enabled, IOUTFS = 2 mA, DAC sample rate = 125 MSPS. No interpolation, unless otherwise noted. Table 1. Tx Path Specifications Parameter TxDAC DC CHARACTERISTICS Resolution Differential Nonlinearity Gain Variation (Internal Reference) Gain Matching Offset Error Full-Scale Output Current (Default Setting) Output Compliance Range TXVDD = 3.3 V, VTXCML = 0 V TXVDD = 3.3 V, VTXCML = 0.5 V TXVDD = 1.8 V, VTXCML = 0 V Offset Temperature Drift Gain Temperature Drift (Internal Reference) Tx REFERENCE (DEFAULT REGISTER SETTINGS) Internal Reference Voltage (REFIO) Output Resistance Temperature Drift Adjustment Range (TXVDD = 3 V) Adjustment Range (TXVDD = 1.8 V) TxDAC AC CHARACTERISTICS Maximum Update Rate Spurious-Free Dynamic Range fOUT = 5 MHz fOUT = 20 MHz Two-Tone Intermodulation Distortion fOUT1 = 5 MHz, fOUT2 = 6 MHz fOUT1 = 20 MHz, fOUT2 = 21 MHz Noise Spectral Density fOUT = 5 MHz fOUT = 20 MHz W-CDMA Adjacent Channel Leakage Ratio, 1 Carrier fDAC = 122.88 MHz, fOUT = 11 MHz Tx PATH DIGITAL FILTER INPUT RATES SRRC (8x Interpolation Mode) INT0 (4x Interpolation Mode) INT1 (2x Interpolation Mode Transmit DAC (1x Interpolation Mode) Min AD9961 Typ Max -10 -2.4 -0.03 10 0.1 0.4 0.4 +10 +2.4 +0.03 Min -10 -2.4 -0.03 2.0 -0.5 +0.7 -0.5 -0.5 +0.7 -0.5 +10 +2.4 +0.03 +1.0 +1.7 +0.8 0 40 0 40 1.02 10 25 1.02 10 25 1.2 REFIO 175 Rev. A | Page 3 of 60 12 0.3 0.4 0.4 2.0 +1.0 +1.7 +0.8 0.8 0.8 21.875 43.75 87.5 175 AD9963 Typ Max 0.8 0.8 1.2 REFIO 175 Unit Bits LSB %FSR %FSR %FSR mA V V V ppm/C ppm/C V k ppm/C V V MSPS 78 68 81 70 dBc dBc 85 78 89 80 dBc dBc -140 -136 -145 -141 dBm/Hz dBm/Hz 70 74 dBc 21.875 43.75 87.5 175 MHz MHz MHz MHz AD9961/AD9963 Data Sheet TMIN to TMAX, RX33V = TXVDD = CLK33V = DRVDD = AUX33V = 3.3 V. All LDOs enabled, ADC sample rate = 100 MSPS. No decimation, unless otherwise noted. Table 2. Rx Path Specifications Parameter Rx ADC DC CHARACTERISTICS Resolution Differential Nonlinearity Gain Error Offset Error Input Voltage Range Input Capacitance Rx ADC AC SPECIFICATIONS Maximum Sample Rate Spurious Free Dynamic Range fIN = 10.1 MHz fIN = 70.1 MHz Two-Tone Intermodulation Distortion fIN1 = 10 MHz, fIN2 = 11 MHz fIN1 = 29 MHz, fIN2 = 32 MHz Signal-to-Noise Ratio fIN = 10.1 MHz fIN = 30.1 MHz fIN = 70.1 MHz RXCML OUTPUTS Output Voltage Output Current Rx DIGITAL FILTER CHARACTERISTICS 2x Decimation Latency (ADC Clock Cycles) Passband Ripple; fOUT/fDAC (0.4 x fDATA) Stop-Band Rejection (fDATA 0.4 x fDATA) Min AD9961 Typ Max Min 10 0.1 1 0.5 1.56 8 100 AD9963 Typ Max Unit 12 0.3 1 0.5 1.56 8 Bits LSB %FSR %FSR V p-p diff pF 100 MSPS 77 75 77 73 dBc dBc 78 76 82 80 dBc dBc 61 60 60 68 67 66 dBFS dBFS dBFS 1.4 1.4 0.1 49 0.2 70 Rev. A | Page 4 of 60 0.1 49 0.2 70 V mA Cycles fOUT/fDAC dB Data Sheet AD9961/AD9963 TMIN to TMAX, RX33V = TXVDD = CLK33V = DRVDD = AUX33V = 3.3 V. All LDOs enabled, unless otherwise noted. Table 3. Auxiliary Converter Specifications Parameter AUXILIARY DAC12A/AUXDAC12B Resolution Differential Nonlinearity Gain Error Settling Time (1%) AUXILIARY DAC10A/DAC10B (Range = 0.5 V to 1.5 V) Resolution Differential Nonlinearity Gain Error Settling Time (1%) AUXILIARY ADC Resolution Differential Nonlinearity Gain Error (Internal Reference) Input Voltage Range Maximum Sample Rate Min AD9961 Typ Max 12 Min AD9963 Typ Max 12 0.8 2.0 1 10 0.8 2.0 1 Bits LSB % s 1.0 2.0 10 Bits LSB % s 10 1.0 2.0 10 12 -1.0 -2.0 0 50 +1.0 +2.0 3.2 Rev. A | Page 5 of 60 12 -1.0 -2.0 0 50 Units +1.0 +2.0 3.2 Bits LSB % V kHz AD9961/AD9963 Data Sheet fCLK = 125 MHz, fDLL = 250 MHz, DAC sample rate = 125 MSPS, ADC sample rate = 62.5 MSPS, unless otherwise noted. Table 4. Power Consumption Specifications Parameter 1.8 V ONLY OPERATION (EXTERNAL 1.8 V) CLK33V TXVDD DRVDD DVDD18V CLK18V DLL18V RX18V RX18VF 3.3 V ONLY OPERATION (ON-CHIP REGULATORS) TXVDD CLK33V RX33V DRVDD AUX33V SUPPLY VOLTAGE RANGE CLK33V, TXVDD (These Supplies Must Be Tied Together) DRVDD DVDD18V CLK18V DLL18V RX18V RX18VF RX33V AUX33V (AUXADC Enabled) AUX33V (AUXADC Disabled) Min 1.72 1.72 1.72 1.72 1.72 1.72 1.72 2.50 3.14 1.72 Rev. A | Page 6 of 60 AD9961 Typ Max Min AD9963 Typ Max Unit 1.65 10.7 29.4 21.0 3.84 9.98 79.2 34.3 1.65 10.7 34.9 22.7 3.84 9.98 79.2 34.3 mA mA mA mA mA mA mA mA 12.1 17.0 113 93 0.55 12.1 17.0 113 108 0.55 mA mA mA mA mA 3.63 3.63 1.89 1.89 1.89 1.89 1.89 3.63 3.63 3.63 1.72 1.72 1.72 1.72 1.72 1.72 1.72 2.50 3.14 1.72 3.63 3.63 1.89 1.89 1.89 1.89 1.89 3.63 3.63 3.63 V V V V V V V V V V Data Sheet AD9961/AD9963 Table 5. Digital Logic Level Specifications Parameter CMOS INPUT LOGIC LEVEL VIN Logic High VIN Logic High VIN Logic High VIN Logic Low VIN Logic Low VIN Logic Low CMOS OUTPUT LOGIC LEVEL VOUT Logic High VOUT Logic High VOUT Logic High VOUT Logic Low VOUT Logic Low VOUT Logic Low DAC CLOCK INPUT Differential Peak-to-Peak Voltage Duty Cycle Slew Rate DIRECT CLOCKING Clock Rate DLL ENABLED Clock Rate SERIAL PERIPHERAL INTERFACE Maximum Clock Rate Minimum Pulse Width High (tHIGH) Minimum Pulse Width Low (tLOW) Setup Time, SDIO (Data In) to SCLK (tDS) Hold Time, SDI to SCLK (tDH) Data Valid, SDIO (Data Out) to SCLK (tDV) Setup Time, CS to SCLK (tS) Conditions Min DRVDD = 1.8 V DRVDD = 2.5 V DRVDD = 3.3 V DRVDD = 1.8 V DRVDD = 2.5 V DRVDD = 3.3 V 1.2 1.7 2.0 DRVDD = 1.8 V DRVDD = 2.5 V DRVDD = 3.3 V DRVDD = 1.8 V DRVDD = 2.5 V DRVDD = 3.3 V 1.35 2.05 2.4 200 45 0.1 Typ 400 Max Unit 0.5 0.7 0.8 V V V V V V 0.4 0.4 0.4 V V V V V V CLK33V 55 mV p-p diff % V/ns MHz % MHz CLKP/CLKN inputs 0.1 200 DLL delay line output 100 310 50 10 10 5.0 5.0 5.0 5.0 Rev. A | Page 7 of 60 MHz ns ns ns ns ns ns AD9961/AD9963 Data Sheet ABSOLUTE MAXIMUM RATINGS Table 6. Parameter RX33V, AUX33V TXVDD DRVDD CLK33V RX18V, RX18VF DVDD18V CLK18V, DLL18V RXGND, TXGND, DGND, TXIP, TXIN, TXQP, TXQN With Respect to RXGND TXGND DGND EPAD RXGND EPAD EPAD EPAD TXGND RXIP, RXIN, RXQP, RXQN RXGND CS, SCLK, SDIO, RESET, DGND LDO_EN TRXD[11:0], TXD[11:0], TXIQ, TRXIQ, TXCLK, TRXCLK CLKP, CLKN Junction Temperature Storage Temperature Range DGND EPAD Rating -0.3 V to +3.9 V -0.3 V to +3.9 V -0.3 V to +3.9 V -0.3 V to +3.9 V -0.3 to +2.1 V -0.3 to +2.1 V -0.3 to +2.1 V -0.3 V to +0.3 V -1.0 V to TXVDD + 0.3 V -0.3 V to RX18V + 0.3 V -0.3V to DRVDD + 0.3 V -0.3 V to DRVDD + 0.3 V -0.3 V to CLK33V + 0.3 V +125C -65C to +150C 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. THERMAL RESISTANCE The exposed paddle must be soldered to the ground plane for the LFCSP package. Soldering the exposed paddle to the customer board increases the reliability of the solder joints, maximizing the thermal capability of the package. Table 7. Thermal Resistance Airflow 1 m/sec 0 m/sec JA 17.1 20.3 JB 10.6 JC 1.0 Unit C/W C/W Typical JA, JB, and JC are specified for a JEDEC standard 51-7 High- thermal test board. Airflow increases heat dissipation, effectively reducing JA. In addition, metal in direct contact with the package leads from metal traces, through holes, ground, and power planes, reduces the JA. ESD CAUTION Rev. A | Page 8 of 60 Data Sheet AD9961/AD9963 72 71 70 69 68 67 66 65 64 63 62 61 60 59 58 57 56 55 AUXIN1 AUXIO2 AUXIO3 DAC12A DAC12B TXVDD TXIN TXIP TXGND REFIO TXCML TXVDD TXQP TXQN CLK33V CLKP CLKN CLK18V PIN CONFIGURATIONS AND FUNCTION DESCRIPTIONS 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 PIN 1 INDICATOR AD9961 (TOP VIEW) 54 53 52 51 50 49 48 47 46 45 44 43 42 41 40 39 38 37 DLLFILT DLL18V DVDD18 DRVDD NC NC TXD0 TXD1 TXD2 TXD3 TXD4 TXD5 TXD6 TXD7 TXD8 TXD9 TXIQ/TXnRX TXCLK NOTES 1. EXPOSED PAD MUST BE SOLDERED TO PCB. 2. NC = NO CONNECT. 08801-002 DGND DRVDD TRXD9 TRXD8 TRXD7 TRXD6 TRXD5 TRXD4 TRXD3 TRXD2 TRXD1 TRXD0 NC NC DRVDD DGND TRXIQ TRXCLK 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 AUX33V AUXADCREF RXQP RXQN RXGND RXBIAS RX18V RX33V RX18VF RXCML RXGND RXIN RXIP LDO_EN RESET SCLK CS SDIO Figure 2. AD9961 Pin Configuration Table 8. AD9961 Pin Function Descriptions Pin No. 1 Mnemonic AUX33V 2 3, 4 5, 11 6 AUXADCREF RXQP, RXQN RXGND RXBIAS 7 8 RX18V RX33V 9 10 12, 13 14 RX18VF RXCML RXIN, RXIP LDO_EN 15 16 17 18 19, 34 20, 33, 51 21 to 30 31, 32, 49, 50 35 RESET SCLK CS SDIO DGND DRVDD TRXD9 to TRXD0 NC Description Analog Supply for the Auxiliary ADC and Auxiliary DACs (3.3 V 5%, 1.8 V 5% If Auxiliary ADC Is Powered Down). Reference Output (Or Input) for Auxiliary ADC. Differential ADC Q Inputs. The default full-scale input voltage range is 1.56 V p-p differential. Receive Path Ground. External Bias Resistor Connection. An optional 10 k resistor can be connected between this pin and the analog ground to improve the accuracy of the full-scale range of the Rx ADCs. Output of RX18V Voltage Regulator. Input to RX18V and RX18VF Voltage Regulators (2.5 V to 3.3 V). If LDOs are not being used, short Pin 8 to Pin 7. Output of RX18VF Voltage Regulator. ADC Common-Mode Voltage Output. Differential ADC I Inputs. The default full-scale input voltage range is 1.56 V p-p differential. Control Pin for LDOs (GND = Disable all LDOs, Float = Enable DVDD18 LDO Only, DRVDD = Enable All LDOs). Reset. Active low to reset the configuration registers to default values and reset device. Clock Input for Serial Port. Active Low Chip Select. Bidirectional Data Line for Serial Port. Digital Core Ground. Input/Output Pad Ring Supply Voltage (1.8 V to 3.3 V). ADC Output Data in Full Duplex Mode. ADC output data and DAC input data in half-duplex mode. Not Connected. TRXIQ Output Signal Indicating from Which ADC the Output Data Is Sourced. Rev. A | Page 9 of 60 AD9961/AD9963 Pin No. 36 37 38 Mnemonic TRXCLK TXCLK TXIQ/TXnRX 39 to 48 52 53 54 55 56, 57 58 TXD9 to TXD0 DVDD18 DLL18V DLLFILT CLK18V CLKN, CLKP CLK33V 59, 60 61, 67 62 63 64 65, 66 68 69 70 TXQN, TXQP TXVDD TXCML REFIO TXGND TXIP, TXIN DAC12B DAC12A AUXIO3 71 AUXIO2 72 AUXIN1 EPAD Data Sheet Description Qualifying Clock for the TRXD Bus. Qualifying Clock for the TXD Bus. It can be configured as either an input or output. Dual Function Pin. In half-duplex mode (TXnRX), this pin controls the direction of the TRX port. In fullduplex mode (TXIQ), this input signal indicates to which DAC, I or Q, the TxDAC input data is intended. TxDAC Input Data. Digital Core 1.8 V Supply. Output of DLL18V Voltage Regulator. DLL Filter Output. Output of CLK18V Voltage Regulator. Differential Input Clock. Input to CLK18V and DLL18V Voltage Regulators (1.8 V to 3.3 V). If LDOs are not being used, short Pin 58 to Pin 55. CLK33V must track TXVDD. Complementary DAC Q Current Outputs. Analog Supply Voltage for Tx Path (1.8 V to 3.3 V). TXVDD must track CLK33V. Common-Mode Input Voltage for the I and Q Tx DACs. Decoupling Point for Internal DAC 1.0 V Bandgap Reference. Use a 0.1 F capacitor to AGND. Transmit Path Ground. Complementary DAC I Current Outputs. Auxiliary DAC B Output. Auxiliary DAC A Output. Selectable Analog Pin. Programmable to either Input 3 of the auxiliary ADC or to the auxiliary DAC10B output. Selectable Analog Pin. Programmable to either Input 2 of the auxiliary ADC or to the auxiliary DAC10A output. Input 1 of Auxiliary ADC. Thermal Pad Under Chip. This must be connected to AGND for proper chip operation. It provides both a thermal and electrical connection to the PCB. Rev. A | Page 10 of 60 AD9961/AD9963 72 71 70 69 68 67 66 65 64 63 62 61 60 59 58 57 56 55 AUXIN1 AUXIO2 AUXIO3 DAC12A DAC12B TXVDD TXIN TXIP TXGND REFIO TXCML TXVDD TXQP TXQN CLK33V CLKP CLKN CLK18V Data Sheet 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 PIN 1 INDICATOR AD9963 (TOP VIEW) 54 53 52 51 50 49 48 47 46 45 44 43 42 41 40 39 38 37 DLLFILT DLL18V DVDD18 DRVDD TXD0 TXD1 TXD2 TXD3 TXD4 TXD5 TXD6 TXD7 TXD8 TXD9 TXD10 TXD11 TXIQ/TXnRX TXCLK NOTES 1. EXPOSED PAD MUST BE SOLDERED TO PCB. 08801-003 DGND DRVDD TRXD11 TRXD10 TRXD9 TRXD8 TRXD7 TRXD6 TRXD5 TRXD4 TRXD3 TRXD2 TRXD1 TRXD0 DRVDD DGND TRXIQ TRXCLK 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 AUX33V AUXADCREF RXQP RXQN RXGND RXBIAS RX18V RX33V RX18VF RXCML RXGND RXIN RXIP LDO_EN RESET SCLK CS SDIO Figure 3. AD9963 Pin Configuration Table 9. AD9963 Pin Function Descriptions Pin No. 1 Mnemonic AUX33V 2 3, 4 5, 11 6 AUXADCREF RXQP, RXQN RXGND RXBIAS 7 8 RX18V RX33V 9 10 12, 13 14 RX18VF RXCML RXIN, RXIP LDO_EN 15 16 17 18 19, 34 20, 33, 51 21 to 32 35 36 37 38 RESET SCLK CS SDIO DGND DRVDD TRXD11 to TRXD0 TRXIQ TRXCLK TXCLK TXIQ/TXnRX 39 to 50 52 53 TXD11 to TXD0 DVDD18 DLL18V Description Analog Supply for the Auxiliary ADC and Auxiliary DACs (3.3 V 10%, 1.8 V 10% If Auxiliary ADC Is Powered Down). Reference Output (or input) for Auxiliary ADC. Differential ADC Q Inputs. Full-scale input voltage range is 1.56 V p-p differential. Receive Path Ground. External Bias Resistor Connection. This voltage is nominally 0.5 V. A 10 k resistor can be connected between this pin and analog ground to improve the Rx ADC full-scale accuracy. Output of RX18V Voltage Regulator. Input to RX18V and RX18VF Voltage Regulators (2.5 V to 3.3 V). If LDOs are not being used, short Pin 8 to Pin 7. Output of RX18VF Voltage Regulator. ADC Common-Mode Voltage Output. Differential ADC I Inputs. Full-scale input voltage range is 1.56 V p-p differential. Control pin for LDOs (GND = Disable all LDOs, Float = Enable DVDD18 LDO Only, DRVDD = Enable All LDOs). Reset. Active low to reset the configuration registers to default values and reset device. Clock Input for Serial Port. Active Low Chip Select. Bidirectional Data Line for Serial Port. Digital Core Ground. Input/Output Pad Ring Supply Voltage (1.8 V to 3.3 V). ADC Output Data in Full Duplex Mode. ADC output data and DAC input data in half-duplex mode. Output Signal Indicating from Which ADC the Output Data Is Sourced. Qualifying Clock for the TRXD Bus. Qualifying Clock for the TXD Bus. It can be configured as either an input or output. Dual Function Pin. In half-duplex mode (TXnRX), this pin controls the direction of the TRX port. In fullduplex mode (TXIQ), this input signal indicates to which DAC, I or Q, the TxDAC Input Data is intended. TxDAC Input Data. Digital Core 1.8 V Supply. Output of DLL18V Voltage Regulator. Rev. A | Page 11 of 60 AD9961/AD9963 Pin No. 54 55 56,57 58 Mnemonic DLLFILT CLK18V CLKN, CLKP CLK33V 59, 60 61, 67 62 63 64 65, 66 68 69 70 TXQN, TXQP TXVDD TXCML REFIO TXGND TXIP, TXIN DAC12B DAC12A AUXIO3 71 AUXIO2 72 AUXIN1 EPAD Data Sheet Description DLL Filter Output. Output of CLK18V Voltage Regulator. Differential Input Clock. Input to CLK18V and DLL18V Voltage Regulators (1.8 V to 3.3 V). If LDOs are not being used, short Pin 58 to Pin 55. CLK33V must track TXVDD. Complementary DAC Q Current Outputs. Analog Supply Voltage for Tx Path (1.8 V to 3.3V). TXVDD must track CLK33V. Common-Mode Input Voltage for the I and Q Tx DACs. Decoupling Point for Internal DAC 1.0 V Bandgap Reference. Use a 0.1 F capacitor to AGND. Transmit Path Ground. Complementary DAC I Current Outputs. Auxiliary DAC B Output. Auxiliary DAC A Output. Selectable Analog Pin. Programmable to either Input 3 of the auxiliary ADC or to the auxiliary DAC10B output. Selectable Analog Pin. Programmable to either Input 2 of the auxiliary ADC or to the auxiliary DAC10A output. Input 1 of Auxiliary ADC. Thermal Pad Under Chip. This must be connected to AGND for proper chip operation. It provides both a thermal and electrical connection to the PCB. Rev. A | Page 12 of 60 Data Sheet AD9961/AD9963 TYPICAL PERFORMANCE CHARACTERISTICS 100 95 95 90 90 85 85 IFS = 4mA 80 SFDR (dBc) SFDR (dBc) IFS = 2mA 80 75 75 70 IFS = 1mA 70 IFS = 2mA 65 IFS = 1mA 60 65 50 0 10 20 30 40 50 60 fOUT (MHz) 08801-201 60 0 10 20 30 40 50 60 fOUT (MHz) Figure 4. Second Harmonic Distortion vs. fOUT Over Full-Scale Current, fDAC = 125 MHz, 1x, Digital Scale = 0 dBFS, TXVDD = 1.8 V 08801-204 55 Figure 7. Third Harmonic Distortion vs. fOUT Over Full-Scale Current, fDAC = 125 MHz, 1x, Digital Scale = 0 dBFS, TXVDD = 3.3 V 90 95 85 90 80 SFDR (dBc) SFDR (dBc) 85 75 IFS = 2mA 70 IFS = 1mA 65 0dBFS 80 75 -3dBFS 60 70 55 -6dBFS 10 20 30 40 50 60 fOUT (MHz) 65 0 10 20 30 40 50 60 fOUT (MHz) Figure 5. Third Harmonic Distortion vs. fOUT Over Full-Scale Current, fDAC = 125 MHz, 1x, Digital Scale = 0 dBFS, TXVDD = 1.8 V 08801-205 0 08801-202 50 Figure 8. Second Harmonic Distortion vs. fOUT Over Digital Scale, fDAC = 125 MHz, 1x, Full-Scale Current = 2 mA, TXVDD = 1.8 V 100 90 95 85 90 80 0dBFS IFS = 4mA 80 SFDR (dBc) SFDR (dBc) 85 IFS = 2mA 75 70 IFS = 1mA 75 70 65 -6dBFS 65 -3dBFS 60 60 55 55 50 20 30 fOUT (MHz) 40 50 60 Figure 6. Second Harmonic Distortion vs. fOUT Over Full-Scale Current, fDAC = 125 MHz, 1x, Digital Scale = 0 dBFS, TXVDD = 3.3 V Rev. A | Page 13 of 60 0 10 20 30 40 50 60 fOUT (MHz) Figure 9. Third Harmonic Distortion vs. fOUT Over Digital Scale, fDAC = 125 MHz, 1x, Full-Scale Current = 2 mA, TXVDD = 1.8 V 08801-206 10 08801-203 50 0 AD9961/AD9963 Data Sheet 100 -10 95 -20 DIRECT CLOCK DLL x 25 -30 90 POWER (dBm) SFDR (dBc) -40 85 80 0dBFS 75 -3dBFS -50 -60 -70 70 -80 65 -90 0 10 20 30 40 50 -100 08801-207 60 60 fOUT (MHz) 0 Figure 10. Second Harmonic Distortion vs. fOUT Over Digital Scale, fDAC = 125 MHz, 1x, Full-Scale Current = 2 mA, TXVDD = 3.3 V 50 100 150 FREQUENCY (MHz) 200 250 08801-210 -6dBFS Figure 13. Transmit DAC Output Spectrum, Full-Scale Current = 2 mA, TXVDD = 3.3 V, fOUT = 10 MHz, fDAC = 125 MHz 100 1.0 95 -6 dBFS 90 0.5 80 0 dBFS DNL (LSB) SFDR (dBc) 85 75 -3 dBFS 70 0 65 -0.5 60 0 10 20 30 40 50 60 fOUT (MHz) -1.0 08801-208 50 0 512 1024 1536 2048 2560 3072 3584 4096 3584 4096 SAMPLES Figure 11. Third Harmonic Distortion vs. fOUT Over Digital Scale, fDAC = 125 MHz, 1x, Full-Scale Current = 2 mA, TXVDD = 3.3 V 08801-211 55 Figure 14. Auxiliary ADC DNL -10 1.0 DIRECT CLOCK DLL x25 -20 -30 0.5 -50 INL (LSB) POWER (dBm) -40 -60 0 -70 -80 -0.5 0 50 100 150 FREQUENCY (MHz) 200 250 Figure 12. Transmit DAC Output Spectrum, Full-Scale Current = 2 mA, TXVDD = 3.3 V, fOUT = 50 MHz, fDAC = 125 MHz Rev. A | Page 14 of 60 -1.0 0 512 1024 1536 2048 2560 3072 SAMPLES Figure 15. Auxiliary ADC INL 08801-212 -100 08801-209 -90 AD9961/AD9963 100 8 95 6 90 4 85 2 80 SFDR (dBc) 10 0 -2 70 -4 65 -6 60 -8 55 50 5 15 25 35 45 55 65 75 85 95 TEMPERATURE (C) 0 10 20 30 40 50 60 fOUT (MHz) Figure 16. Typical Die Temperature Readback Error vs. Ambient Temperature Figure 19. AD9961, Second and Third Harmonic Distortion vs. fOUT, fDAC = 125 MHz, 1x, Digital Scale = 0 dBFS, TXVDD = 3.3 V 100 ATTEN 2dB REF -38.23dBm #AVG LOG 10dB/ IDAC 3.3V CMOS THIRD HARMONIC (dBc) 08801-216 -10 -40 -35 -25 -15 -5 IDAC 3.3V CMOS SECOND HARMONIC (dBc) 75 08801-213 ERROR (C) Data Sheet SFDR (dBFS) 90 EXT REF DC COUPLED SNR OR SFDR (dBc, dBFS) 80 SNR (dBFS) 70 60 50 40 SFDR (dBc) 30 20 SNR (dBc) PAVG 10 W1 S2 CENTER 21.00MHz #RES BW 30kHz Figure 17. One-Carrier W-CDMA ACLR Performance, IF = ~21 MHz 0 -80 -60 -50 -40 -30 -20 -10 0 fIN (dBm) Figure 20. SNR/SFDR vs. Analog Input Level, fIN = 10 MHz, fADC = 100 MSPS 100 100 95 90 90 SFDR (dBFS) SNR OR SFDR (dBc, dBFS) 80 IDAC 1.8V CMOS SECOND HARMONIC (dBc) 85 80 75 70 65 IDAC 1.8V CMOS THIRD HARMONIC (dBc) 60 SNR (dBFS) 70 60 50 40 SFDR (dBc) 30 20 55 SNR (dBc) 10 50 10 20 30 40 50 60 fOUT (MHz) Figure 18. AD9961, Second and Third Harmonic Distortion vs. fOUT, fDAC = 125 MHz, 1x, Digital Scale = 0 dBFS, TXVDD = 1.8 V 0 -80 08801-215 0 -70 -60 -50 -40 fIN (dBm) -30 -20 -10 0 08801-218 SFDR (dBc) -70 08801-214 RMS RESULTS CARRIER POWER -25.07dBm/ 3.84000MHz SPAN 33.84MHz VBW 300kHz SWEEP 109.8ms (601pts) UPPER LOWER dBc dBm FREQ OFFSET REF BW dBc dBm 5.000MHz 3.840MHz -73.49 -98.57 -73.85 -98.92 10.00MHz 3.840MHz -72.90 -97.97 -73.11 -98.19 15.00MHz 3.840MHz -73.44 -98.51 -73.56 -98.63 08801-217 10 Figure 21. SNR/SFDR vs. Analog Input Level, fIN = 70 MHz, fADC = 100 MSPS Rev. A | Page 15 of 60 AD9961/AD9963 Data Sheet 1.2 100 1.0 90 0.8 IDAC 125MHz 80 0.4 IMD (dB) 0.2 LSB IDAC 35MHz INL DNL 0.6 0 -0.2 70 IDAC 70MHz 60 -0.4 -0.6 50 -0.8 -1.0 512 1024 1536 2048 2560 3072 3584 4096 CODE 08801-219 0 Figure 22. Rx Path ADC, INL and DNL 0 10 20 30 40 50 60 fOUT (MHz) 08801-222 40 -1.2 Figure 25. Intermodulation Distortion vs. fOUT Over fDAC, TXVDD = 3.3 V, Full-Scale Current = 2 mA 155 100 IDAC, 125MHz, 4mA, 0dB 153 90 151 QDAC, BOARD 4 IDAC, 125MHz, 2mA, 0dB 80 147 IMD (dB) NSD (-dBm/Hz) 149 145 QDAC, 125MHz, 1mA, 0dB 143 QDAC, BOARD 3 70 QDAC, BOARD 1 60 141 139 50 137 10 20 30 40 50 60 70 fOUT (MHz) 40 Figure 23. Transmit DAC Noise Spectral Density vs. fOUT Over Full-Scale Current 0 10 20 30 40 50 60 fOUT (MHz) 08801-223 0 08801-220 135 Figure 26. Intermodulation Distortion vs. fOUT , TXVDD = 3.3 V, Full-Scale Current = 2 mA, Board-to-Board Variation 155 100 153 90 151 IDAC, 125MHz, 2mA, 0dB 80 IDAC, 125MHz, 2mA, -3dB QDAC -6dB 147 IMD (dB) NSD (-dBm/Hz) 149 145 143 QDAC -3dB 70 QDAC 0dB IDAC, 125MHz, 2mA, -6dB 60 141 139 50 0 10 20 30 fOUT (MHz) 40 50 60 Figure 24. Transmit DAC Noise Spectral Density vs. fOUT Over Digital Scale Rev. A | Page 16 of 60 40 0 10 20 30 40 50 60 fOUT (MHz) Figure 27. Intermodulation Distortion vs. fOUT Over Digital Scale, TXVDD = 3.3 V, Full-Scale Current = 2 mA 08801-224 135 08801-221 137 Data Sheet AD9961/AD9963 100 -60 95 85 80 75 -65 MIN PIPE SFDR (dBFS) MID PIPE SFDR (dBFS) MAX PIPE SFDR (dBFS) MIN PIPE SNR (dBFS) MID PIPE SNR (dBFS) MAX PIPE SNR (dBFS) THD (dBc) SNR OR SFDR (dBFS) 90 -70 -75 70 65 60 50 40 30 20 10 0 fIN (dBm) 0 60 80 100 120 140 Figure 31. AD9963 1.8 V CMOS IADC, 100 MSPS Single Tone AC 80 -70 78 -72 SECOND HARMONIC (dBc) 76 74 -74 -76 -78 72 20 40 60 80 100 120 140 fIN (MHz) 0 08801-226 0 20 40 60 80 100 120 140 fIN (MHz) 08801-229 -80 70 Figure 32. AD9963 1.8 V CMOS IADC, 100 MSPS Single Tone AC Figure 29. AD9963 100 MSPS Single Tone AC -65 68 -70 THIRD HARMONIC (dBc) 70 66 64 -75 -80 -85 62 -90 0 20 40 60 80 100 120 140 fIN (MHz) 08801-227 60 0 20 40 60 80 100 120 140 fIN (MHz) Figure 33. AD9963 1.8 V CMOS IADC, 100 MSPS Single Tone AC Figure 30. AD9963 1.8 V CMOS IADC, 100 MSPS Single Tone AC Rev. A | Page 17 of 60 08801-230 SFDR (dB) 40 fIN (MHz) Figure 28. SNR/SFDR vs. Analog Input Level Over Full-Scale Input Range, fIN = 70 MHz, fADC = 100 MSPS SNR (dB) 20 08801-228 -80 70 08801-225 60 80 AD9961/AD9963 Data Sheet TERMINOLOGY Linearity Error (Integral Nonlinearity or INL) Linearity error is defined as the maximum deviation of the actual analog output from the ideal output, determined by a straight line drawn from zero scale to full scale. Spurious Free Dynamic Range (SFDR) The difference, in decibels, between the peak amplitude of the output signal and the peak spurious signal between dc and the frequency equal to half the input data rate. Differential Nonlinearity (DNL) DNL is the measure of the variation in analog value, normalized to full scale, associated with a 1 LSB change in digital input code. Total Harmonic Distortion (THD) THD is the ratio of the rms sum of the first six harmonic components to the rms value of the measured fundamental. It is expressed as a percentage or in decibels. Monotonicity A DAC is monotonic if the output either increases or remains constant as the digital input increases. Offset Error The deviation of the output current from the ideal of zero is called offset error. For TXIN, 0 mA output is expected when the inputs are all 0s. For TXIP, 0 mA output is expected when all inputs are set to 1. Gain Error The difference between the actual and ideal output span. The actual span is determined by the difference between the output when all inputs are set to 1 and the output when all inputs are set to 0. Output Compliance Range The range of allowable voltage at the output of a current-output DAC. Operation beyond the maximum compliance limits can cause either output stage saturation or breakdown, resulting in nonlinear performance. Signal-to-Noise Ratio (SNR) SNR is the ratio of the rms value of the measured output signal to the rms sum of all other spectral components below the Nyquist frequency, excluding the first six harmonics and dc. The value for SNR is expressed in decibels. Adjacent Channel Leakage Ratio (ACLR) The ratio in dBc between the measured power within a channel relative to its adjacent channel. Complex Image Rejection In a traditional two-part upconversion, two images are created around the second IF frequency. These images have the effect of wasting transmitter power and system bandwidth. By placing the real part of a second complex modulator in series with the first complex modulator, either the upper or lower frequency image near the second IF can be rejected. Temperature Drift Temperature drift is specified as the maximum change from the ambient (25C) value to the value at either TMIN or TMAX. For offset and gain drift, the drift is reported in parts per million of full-scale range (FSR) per degree Celsius (C). For reference drift, the drift is reported in parts per ppm/C. Power Supply Rejection The maximum change in the full-scale output as the supplies are varied from minimum to maximum specified voltages. Settling Time The time required for the output to reach and remain within a specified error band around its final value, measured from the start of the output transition. Rev. A | Page 18 of 60 Data Sheet AD9961/AD9963 THEORY OF OPERATION The AD9961/AD9963 are targeted to cover the mixed-signal front-end needs of multiple wireless communications systems. They feature a receive path that consists of dual 10-/12-bit receive ADCs and a transmit path that consists of dual 10-/12-bit transmit DACs (TxDAC). The AD9961/AD9963 integrate additional functionality typically required in most systems, such as power scalability, Tx gain control, and clock multiplication circuitry. In full duplex mode, the AD9961/AD9963 use two 12-bit buses, along with qualifying clock signals, to transfer Rx path data and Tx path data. These two buses support either single data rate or double data rate data transfers. The data bus, along with many other device options, is configurable through the serial port by writing internal registers. The device can also be used in a single-port, half-duplex configuration. The AD9961/AD9963 minimize both size and power consumption to address the needs of a range of applications from the low power portable market to the high performance femto base station market. The part is provided in a 72-lead lead frame chip scale package (LFCSP) that has a footprint of only 10 mm x 10 mm. Power consumption can be optimized to suit the particular application by incorporating power-down controls, low power ADC modes, and TxDAC power scaling. Rev. A | Page 19 of 60 AD9961/AD9963 Data Sheet SERIAL CONTROL PORT Table 10. Byte Transfer Count The AD9961/AD9963 serial control ports are a flexible, synchronous, serial communications port that allows an easy interface with many industry-standard microcontrollers and microprocessors. The AD9961/AD9963 serial control ports are compatible with most synchronous transfer formats, including both the Motorola SPI and Intel(R) SSR(R) protocols. The serial control port allows read/write access to all registers that configure the AD9961/AD9963. Single or multiple byte transfers are supported, as well as MSB first or LSB first transfer formats. Serial Control Port Pin Descriptions The serial control port has three pins, SCLK, SDIO, and CS: * SCLK (serial clock) is the input clock used to register serial control port reads and writes. Write data bits are registered on the rising edge of this clock, and read data bits are registered on the falling edge. This pin is internally pulled down by a 30 k resistor to ground. * SDIO (serial data input/output) functions as both the input and output data pin. * CS (chip select bar) is an active low control that gates the read and write cycles. When CS is high, SDIO is in a high impedance state and SCLK is disabled. This pin is internally pulled up by a 30 k resistor to DRVDD. N1 0 0 1 1 N0 0 1 0 1 Bytes to Transfer 1 2 3 Streaming mode A12 to A0 select the address within the register map that is written to or read from during the data transfer portion of the communications cycle. For multibyte transfers, the address is the starting byte address. Only Address Bits[A7:A0] are needed to cover the range of the 0xFF registers used by the AD9961/AD9963. Address Bits[A12:A8] must always be 0. Write Transfer If the instruction header indicates a write operation, the bytes of data written onto the SDIO line are loaded into the serial control port buffer of the AD9961/AD9963. Data bits are registered on the rising edge of SCLK. The length of the transfer (1 byte, 2 byte, 3 bytes, or streaming mode) is indicated by two bits (N1:N0) in the instruction byte. During a write, streaming mode does not skip over unused or reserved registers; therefore, the user must know what bit pattern to write to the reserved registers to preserve proper operation of the part. It does not matter what data is written to unused registers. GENERAL OPERATION OF SERIAL CONTROL PORT Read Transfer The falling edge of CS, in conjunction with the rising edge of SCLK, determines the start of a communication cycle. There are two parts to a communication cycle with the AD9961/ AD9963. The first part writes a 16-bit instruction word into the AD9961/AD9963, coincident with the first 16 SCLK rising edges. The instruction word provides the AD9961/AD9963 serial control ports with information regarding the data transfer, which is the second part of the communication cycle. The instruction word defines whether the upcoming data transfer is a read or a write, the number of bytes in the data transfer, and the starting register address for the first byte of the data transfer. If the instruction word is for a read operation, the next N x 8 SCLK cycles clock out the data from the address specified in the instruction word, where N is 1 to 3 as determined by N1:N0. If N = 4, the read operation is in streaming mode, and continues until CS is raised. Streaming mode does not skip over reserved or unused registers. The readback data is valid on the falling edge of SCLK. Instruction Header The MSB of the instruction word is R/W, which indicates whether the serial port transfer is a read or a write. The next two bits, N1:N0, indicate the length of the transfer in bytes. The final 13 bits are the address (A12 to A0) at which to begin the read or write operation. For a write, the instruction word is followed by the number of bytes of data indicated by Bit N1 to Bit N0 (see Table 10). MSB/LSB First Transfers The AD9961/AD9963 instruction word and byte data formats can be selected to be MSB first or LSB first. The default for the AD9961/AD9963 is MSB first. When MSB first mode is active, the instruction and data bytes must be written from MSB to LSB. Multibyte data transfers in MSB first format start with an instruction byte that includes the register address of the most significant data byte. Subsequent data bytes must follow in order from the high address to the low address. In MSB first mode, the serial control port internal address generator decrements for each data byte of the multibyte transfer cycle. When LSB first is active, the instruction and data bytes must be written from LSB to MSB. Multibyte data transfers in LSB first format start with an instruction byte that includes the register address of the least significant data byte followed by multiple data bytes. The internal byte address generator of the serial control port increments for each byte of the multibyte transfer cycle. Rev. A | Page 20 of 60 Data Sheet AD9961/AD9963 When LSB first is set by Register 0x00, Bit 2 and Register 0x00, Bit 6, it takes effect immediately. In multibyte transfers, subsequent bytes reflect any changes in the serial port configuration. To avoid problems reconfiguring the serial port operation, any data written to 0x00 must be mirrored (the eight bits should read the same, forward or backward). Mirroring the data makes it irrelevant whether LSB first or MSB first is in effect. As an example of this mirroring, the default setting for Register 0x00 is 00011000. Table 11. Streaming Mode (No Addresses Are Skipped) Write Mode LSB First MSB First Address Direction Increment Decrement Stop Sequence 0xFD, 0xFE, 0xFF, stop 0x01, 0x00, 0xFF, stop SUB SERIAL INTERFACE COMMUNICATIONS The AD9963/AD9961 have two registers that require a different communication sequence. These registers are 0x0F and 0x10. The write sequence for these two registers requires a write to Register 0x05, a write to the Register (0x0F or 0x10), and then a write to Register 0xFF. The write takes effect when the write to Register 0xFF is completed. Ending Transfers When the transfer is 1, 2, or 3 bytes, the data transfer ends after the required number of clock cycles have been received. CS can be raised after each sequence of eight bits to stall the bus (except after the last byte, where it ends the cycle). When the bus is stalled, the serial transfer resumes when CS is lowered. Raising CS on a non byte boundary resets the serial control port. For example, to enable the RXCML pin output buffer, the register write sequence is: The AD9961/AD9963 serial control port register addresses decrement from the register address just written toward 0x00 for multibyte I/O operations if the MSB first mode is active (default). If the LSB first mode is active, the register address of the serial control port increments from the address just written toward 0xFF for multibyte I/O operations. Streaming mode transfers always terminate when CS is raised. Streaming mode transfers also terminate whenever the address reaches 0xFF. Note that unused addresses are not skipped during multibyte I/O operations. To avoid unpredictable device behavior, do not write to reserved registers. 1. Write 0x03 into Register 0x05. This addresses both of the Rx ADCs. 2. Write 0x02 into Register 0x0F. This sets the RXCML enable bit. 3. Write 0x01 into Register 0xFF. This updates the internal register, which activates the RXCML buffer. 4. Write 0x00 into Register 0x05. This returns the SPI to the normal addressing mode. An example of updating Register 0x10 is given in the ADC Digital Offset Adjustment section. Table 12. Serial Control Port, 16-Bit Instruction Word, MSB First MSB I15 I14 I13 I12 I11 I10 I9 I8 I7 I6 I5 I4 I3 I2 I1 LSB I0 R/W N1 N0 0 0 0 0 0 A7 A6 A5 A4 A3 A2 A1 A0 CS SCLK DON'T CARE DON'T CARE A6 A5 A4 A3 A2 A1 A0 D7 D6 16-BIT INSTRUCTION HEADER D5 D4 D3 D2 D1 D0 REGISTER (N) DATA D7 D6 D5 D4 D3 D2 D1 D0 DON'T CARE 08801-038 SDIO DON'T CARE R/W N1 N0 A12 A11 A10 A9 A8 A7 REGISTER (N - 1) DATA Figure 34. Serial Control Port Access--MSB First, 16-Bit Instruction, 2-Byte Data tDS tS tHIGH tDH CS DON'T CARE SDIO DON'T CARE DON'T CARE R/W N1 N0 A12 A11 A10 A9 A8 A7 A6 A5 D4 D3 D2 D1 Figure 35. Serial Control Port Write--MSB First, 16-Bit Instruction, Timing Measurements Rev. A | Page 21 of 60 D0 DON'T CARE 08801-040 SCLK tC tCLK tLOW AD9961/AD9963 Data Sheet CS SCLK DATA BIT N 08801-041 tDV SDIO SDO DATA BIT N - 1 Figure 36. Timing Diagram for Serial Control Port Register Read CS SCLK DON'T CARE A0 A1 A2 A3 A4 A5 A6 A7 A8 A9 A10 A11 A12 N0 N1 R/W D0 16-BIT INSTRUCTION HEADER D1 D2 D3 D4 D5 REGISTER (N) DATA D6 D7 D0 D1 D2 D3 D4 D5 D6 D7 DON'T CARE 08801-042 SDIO DON'T CARE DON'T CARE REGISTER (N + 1) DATA Figure 37. Serial Control Port Access--LSB First, 16-Bit Instruction, Two Bytes Data tS tC CS tCLK tHIGH tLOW tDS SCLK SDIO BIT N BIT N + 1 08801-043 tDH Figure 38. Serial Control Port Timing--Write Table 13. Serial Control Port Timing Parameter tDS tDH tCLK tS Timing (Min, ns) 5.0 5.0 20.0 5.0 tC 2 tHIGH tLOW tDV 10 10 5.0 Description Setup time between data and rising edge of SCLK. Hold time between data and rising edge of SCLK. Period of the clock. Setup time between CS falling edge and SCLK rising edge (start of communication cycle). Setup time between SCLK rising edge and CS rising edge (end of communication cycle). Minimum period that SCLK should be in a logic high state. Minimum period that SCLK should be in a logic low state. SCLK to valid SDIO and SDO (see Figure 36). Rev. A | Page 22 of 60 Data Sheet AD9961/AD9963 CONFIGURATION REGISTERS Table 14. Configuration Register Map Addr 0x00 0x05 0x0F 0x10 0x30 0x31 0x32 0x33 0x34 0x35 0x36 0x37 0x38 0x39 0x3A 0x3B 0x3C 0x3D 0x3E 0x3F 0x40 Default 0x18 0x00 0x00 0x00 0x3F 0xA7 0xA7 Varies Varies 0x10 0x08 0x10 0x06 0x00 0x51 0x51 0xF0 0x00 0x09 0x07 0x01 0x41 0x42 0x43 0x44 0x45 0x46 0x47 0x48 0x49 0x4A 0x50 0x51 0x52 0x53 0x54 0x55 0x5C 0x60 0x61 0x62 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x93 0x34 0x5F 0x36 0x08 0x00 0x00 0xF8 0x63 0x66 0x68 0x69 0x6A 0x6B 0x6C 0x6D 0x6E 0x00 0x28 0x00 0x00 0x00 0x00 0x00 0x00 0x40 Bit 7 SDIO Bit 6 LSB First Unused Unused TX_SDR TXCKO_INV RX_SDR Unused Unused FIFO_INIT Bit 5 Reset Bit 4 1 Unused Bit 3 1 Bit 2 Reset Bit 1 LSB First ADDRQ RXCML Bit 0 SDIO ADDRI ADC_OFFSET[5:0] INT1_BP INT0_BP SRRC_BP TXCLK_EN RXCLK_EN TXCLK_MD[1:0] TXCKI_INV TXIQ_HILO TX_IFIRST TX_BNRY RXCLK_MD[1:0] RXCLK_INV RXIQ_HILO RX_IFIRST RX_BNRY Aligned ALIGN_ACK ALIGN_REQ FIFO_OFFSET[2:0] FIFO_LVL[7:0] Unused SRRC_SCALE[4:0] Unused INT0_SCALE[4:0] Unused INT1_SCALE[4:0] Unused DEC_SCALE[4:0] RXDLLRST TXDLLRST Unused RXDLL_LKD TXDLL_LKD RXDBL_SEL TXDBL_SEL TX_UNLOCK[1:0] TX_LOCK[1:0] TX_DLYOFS[1:0] TX_HYST[1:0] RX_UNLOCK[1:0] RX_LOCK[1:0] RX_DLYOFS[1:0] RX_HYST[1:0] DBL_TAPDLY[7:0] Unused RX_INVQ RX_INVI TX_INVQ TX_INVI Unused TX_DBLPW[2:0] RX_DBLPW[2:0] Unused RX_CLK RX_BUS SINGLERX TXCLK_MD HD_BUSCTL HD_CLKMD FULL_DUPLEX AUXDAC_ DAC_ DAC12B_EN DAC12A_EN DAC12B_TOP DAC12A_TOP Unused REF UPDATE DAC12A[11:4] Unused DAC12A[3:0] DAC12B[11:4] Unused DAC12B[3:0] DAC10B_EN Unused DAC10B_TOP[2:0] DAC10B_RNG[1:0] DAC10B[9:2] Unused DAC10B[1:0] DAC10A_EN Unused DAC10A_TOP[2:0] DAC10A_RNG[1:0] DAC10A[9:2] Unused DAC10A[1:0] Unused TX_PTTRN TX_INSEL TX_CONT TX_START TX_BISTEN Unused RX_PTTRN RX_INSEL RX_CONT RX_START RX_BISTEN TXI_CHK[15:8] TXI_CHK[7:0] TXQ_CHK[15:8] TXQ_CHK[7:0] Chip ID[7:0] DLL_EN TXDAC_PD TXI_SLEEP TXQ_SLEEP CLK_PD RXADC_PD RXQ_SLEEP RXI_SLEEP Unused DLL_LDO_PD DLLBIAS_PD CLK_LDO_PD RX_LDO_PD RXF_LDO_PD AUXADC_PD AUX_REF_PD DLL_LDO_ RXF_LDO_ DIG_LDO_ CLK_LDO_STAT RX_LDO_ Unused Unused RSET_SEL STAT STAT STAT STAT TRXD_DRV TRXIQ_DRV TRXCLK_DRV TXCLK_DRV TXI_DCLK TXQ_DCLK Unused RXI_DCLK RXQ_DCLK DCS_BP ADCDIV[1:0] Unused IGAIN1[5:0] Unused IGAIN2[5:0] Unused IRSET[5:0] Unused QGAIN1[5:0] Unused QGAIN2[5:0] Unused QRSET[5:0] Unused REFIO_ADJ[5:0] DEC_BP Rev. A | Page 23 of 60 AD9961/AD9963 Addr 0x71 0x72 0x75 0x77 0x78 0x79 0x7A 0x7B 0x7D 0x7E 0x7F 0x80 Default 0x00 0x01 0x00 0x00 Varies Varies 0x00 0x00 0x00 0x00 0x00 0x00 0x81 0x82 0x00 0x00 0xFF 0x00 Bit 7 ADCCLKSEL DLL_Locked Data Sheet Bit 6 DACCLKSEL Bit 5 Unused DLLDIV 0 Bit 4 DLL_REF_EN Unused Bit 2 Bit 1 N[3:0] Bit 0 M[4:0] DLL_RESB CONV_TIME[1:0] AUXADC_EN TMPSNS_EN Bit 3 AUXADC[3:0] AUXADC_RESB Unused Unused RXTrim_EN RXTrim_Fine 0 AUXADC_CH[2:0] Unused AUXADC[11:4] CONV_COMPL Unused AUXREF_ADJ[2:0] AUXCML_EN RXI_Trim[9:2] Unused CHAN_SEL[2:0] AUXDIV[2:0] Unused RX_FSADJ[4:0] 0 RX_DC RXI_Trim[1:0] GAINCAL_ ENI RXQ_Trim[1:0] GAINCAL_ ENQ Update RXQ_Trim[9:2] Unused Unused CONFIGURATION REGISTER BIT DESCRIPTIONS Table 15. Register Name Serial Port Config Register Address 0x00 Bit(s) 7, 0 Parameter SDIO 6, 1 LSB_First 5, 2 RESET ADC Address 0x05 1:0 ADDRQ, ADDRI CM Buffer Enable 0x0F 1 RXCML ADC Offset 0x10 5:0 ADC_OFFSET[5:0] Digital Filters 0x30 7:6 5 4 3 Unused DEC_BP INT1_BP INT0_BP Function 0: use SDIO as both input and output data 1: use SDIO pin as input data only 0: first bit of serial data is MSB of data byte. 1: first bit of serial data is LSB of data byte. A transition from 0 to 1 on this bit resets the device. All registers but Register 0x00 revert to their default values. Bits are set to determine which device on chip receives ADC specific write commands. ADC specific write commends include writes to Registers 0x0F and Register 0x10. These writes also require a rising end on the Update bit (Register 0xFF, Bit 0). 00: no ADCs are addressed. 01: I ADC is addressed. 10: Q ADC is addressed 11: both I and Q ADCs are addressed. Enable control for the RXCML output buffer. Note that updating this bit also requires writing to Register 0x05 and Register 0xFF as described in the Sub Serial Interface Communications section. 0: RXCML pin is high impedance. 1: RXCML pin is a low impedance 1.4 V output. Adds a dc offset to the ADC output of whichever ADC is addressed by Register 0x05. The offset applied is as follows: 011111: offset = +31 LSBs ... 000001: offset = +1 LSB 000000: offset = 0 LSB 111111: offset = -1 LSB ... 100000: offset = -32 LSBs 1: bypass 2x decimator in Rx path (D0). 1: bypass 2x Half-Band Interpolation Filter 1 (INT1). 1: bypass 2x Half-Band Interpolation Filter 0 (INT0). Rev. A | Page 24 of 60 Data Sheet Register Name Tx Data Interface AD9961/AD9963 Register Address 0x31 Bit(s) 2 Parameter SRRC_BP 1 0 TXCLK_EN RxNTx 7 TX_SDR 6 TXCKO_INV 5:4 TXCLK_MD[1:0] 3 TXCKI_INV 2 TXIQ_HILO 1 TX_IFIRST 0 TX_BNRY 7 RX_SDR 6 5:4 Unused RXCLK_MD[1:0] Function 1: bypass 2x SRRC interpolation filter (SRRC). The filter chain is SRRCINT0INT1. If SRRC filter is enabled, the other two filters are enabled too. 1: enables data clocks for transmit path. 0: in HD SPI pin mode, TRx port operates in Tx mode. 1: in HD SPI pin mode, TRx port operates in Rx mode. 0: chooses DDR clocking mode. Tx data is driven out on both edges of the TXCLK signal. 1: chooses bus rate clocking mode. Tx data is driven out on one edge of the TXCLK signal. This signal inverts the phase of the transmit path output clock signal. 0: does not invert TxCLK output. 1: inverts TxCLK output. Controls the mode of the TXCLK pin. The TXCLK pin can be configured as an input or an output. When configured as an output, it can have two possible sources, the internal TXCLK signal or the DLL output signal. 00: disabled. 01: the TXCLK pin is configured as an input. 10: the TXCLK pin is configured as an output. The source signal is the transmit path clock signal. 11: the TXCLK pin is configured as an output. The source signal is the DLL output signal. Note that the TXCLK signal may appear on either the TXCLK pin or the TRXCLK pin, depending on the mode of the device. In Half-Duplex 1Clock mode, this signal is present on the TRXCLK pin when TX is active. In Half-Duplex 2-Clock mode and Full-Duplex mode, this signal is present on the TXCLK pin. Selects which edge of the TXCLK signal samples the transmit path data. 0: TXPCLK negative edge latches transmit path data. 1: TXPCLK positive edge latches transmit path data. Data appears on the TXD bus sequentially but is loaded into the transmit path in pairs. TXIQ_HILO selects how the TXIQ signal marks each data pair. 0: each data pair is marked by TXIQ being low then high. 1: each data pair is marked by TXIQ being high then low. This bit sets the data pairing order of the I and Q samples on transmit path. 0: selects that Q is first, followed by I. 1: selects that I is first, followed by Q. Rx Data Interface 0x32 This bit selects the data format of the transmit path data. 0: Tx binary. 1: Tx twos complement. 0: chooses DDR clocking mode. Rx data is driven out on both edges of the TRXCLK signal. 1: chooses bus rate clocking mode. Rx data is driven out on one edge of the TRXCLK signal. This sets the way the internal RXCLK signal in the chip is driven. 00: disabled. 01: disabled. 10: RXCLK is driven by internal Rx path clock. Rev. A | Page 25 of 60 AD9961/AD9963 Register Name Register Address Data Sheet Bit(s) Parameter 3 RXCLK_INV 2 RXIQ_HILO Function 11: RXCLK is driven by the DLL output. Note that the RXCLK signal is present on the TRXCLK pin with one exception. In Half-Duplex 1-Clock mode, the RXCLK signal is present on the TRXCLK pin when Rx is active, but the TXCLK signal appears on the TRXCLK pin when TX is active. 0: uses TRxCLK negative edge to drive out Rxdata. 1: uses TRxCLK positive edge to drive out Rxdata. Data appears on the RXD bus sequentially but is sampled in the Rx path in pairs. RXIQ_HILO selects how the RXIQ signal marks each data pair. 0: each data pair is marked by RXIQ being low then high. 1: each data pair is marked by RXIQ being high then low. FIFO Alignment 0x33 1 RX_IFIRST 0 RX_BNRY 7 6 5 4 3 2:0 Unused Unused Unused Unused ALIGN_REQ FIFO_OFFSET[2:0] FIFO Status 0x34 7:0 FIFO_LVL[7:0] Tx Scale P 0x35 7:5 4:0 Unused SRRC_SCALE[4:0] 7:5 4:0 Unused INT0_SCALE[4:0] 7:5 4:0 Unused INT1_SCALE[4:0] Tx Scale 0 Tx Scale 1 0x36 0x37 The Rx path I and Q ADCs sample simultaneously producing a pair of samples. Because the RXD bus is shared, the sampled I and Q data appears on the TRXD bus sequentially. This bit determines the order of the paired samples. 0: Q appears first on Rx path. 1: I appears first on Rx path. 0: straight binaryon Rx path. 1: twos compliment on Rx path. 1: request FIFO read and write pointers alignment Sets the FIFO read and write pointer phase offset following FIFO reset. Normally this should be set to 000 to set the FIFO to half full. For valid transmit data path operation, the FIFO should be running half full, that is, it should always contain 4 valid DAC input samples for each DAC. FIFO_LVL values of 00011110, 00011111, 000001110, and 00001111 all indicate that the FIFO is half full. This phenomenon is due to ambiguities in reading back the FIFO_LVL level from this register using the SPI port versus the actual FIFO pointer values. Value of 1.4 multiplier applied to both I and Q channels just after the SRRC filter. 00000: multiply by 0.0. 00001: multiply by 0.0625. ... 11111: multiply by 1.9375. Value of 1.4 multiplier applied to both I and Q channels just after Interpolation Filter 0. 00000: multiply by 0.0. 00001: multiply by 0.0625. ... 11111: multiply by 1.9375. Value of 1.4 multiplier applied to both I and Q channels just after Interpolation Filter 1. 00000: multiply by 0.0. 00001: multiply by 0.0625. 11111: multiply by 1.9375. Rev. A | Page 26 of 60 Data Sheet Register Name Rx Scale Clock Doubler Config TX Clock Doubler Config RX Clock Doubler Config Clock Doubler Config Data Spectral Inversion Clock Doubler Pulse Width Rx Data Interface AD9961/AD9963 Register Address 0x38 Bit(s) 7:5 4:0 Parameter Unused DEC_SCALE[4:0] 7 RXDDLLRST 6 5:4 3 2 1 TXDDLLRST Unused Unused Unused RXDBL_SEL 0 TXDBL_SEL 0x3A 7:4 TX_UNLOCK[1:0] 0: selects fixed pulse width clock doubler. 1: selects fixed duty cycle clock doubler. See Table 22 for configuration recommendations. 0: selects fixed pulse width clock doubler. 1: selects fixed duty cycle clock doubler. See Table 22 for configuration recommendations. Sets the number of clock cycles for the unlock indicator. Set to 01. 0x3B 3 2 1 7:4 TX_LOCK[1:0] TX_DLYOFS[1:0] TX_HYST[1:0] RX_UNLOCK[1:0] Sets the number of clock cycles for the lock indicator. Set to 01. Sets delay line offset of clock doubler. Set to 01. Sets delay line hysteresis of clock doubler. Set to 01. Sets the number of clock cycles for the unlock indicator. Set to 01. 0x3C 3 2 1 7:0 RX_LOCK[1:0] RX_DLYOFS[1:0] RX_HYST[1:0] DBL_TAPDLY[7:0] Sets the number of clock cycles for the lock indicator. Set to 01. Sets delay line offset of clock doubler. Set to 01. Sets delay line hysteresis of clock doubler. Set to 01. Sets the initial tap delay of the Rx and Tx clock doublers. Set to 0x00. 0x3D 7:4 Unused RX_INVQ RX_INVI TX_INVQ TX_INVI Unused 1: multiply Rxdata from QADC by -1. 1: multiply Rxdata from IADC by -1. 1: multiply Txdata for QDAC by -1. 1: multiply Txdata for IDAC by -1. 0x3E 3 2 1 0 7:6 5:3 2:0 7 6 TX_DBLPW[2:0] RX_DBLPW[2:0] Unused RX_CLK Sets the pulse width of the Tx clock doubler. See Table 22 for details. Sets the pulse width of the Rx clock doubler. See Table 22 for details. 5 RX_BUS 4 SINGLERX 3 TXCLK_MD 0x39 0x3F Function Value of 3.2 multiplier applied to both I and Q channels just after the decimation filter. The value of the gain applied is equal to DEC_SCALE/4. 00000: multiply by 0.0. 00001: multiply by 0.25. 11111: multiply by 7.75. 1: resets the Rx signal path clock doubler. 1: resets the Tx signal path clock doubler. . 0: when SINGLERX is active, use Q side clock. 1: when SINGLERX is active, use I side clock. 0: when SINGLERX is active, use the Q ADC. 1: when SINGLERX is active, use the I ADC. 0: use both Rx paths. 1: use only one Rx path. This bit controls the operation of the TXCLK pin when the chip is configured in half-duplex 1-clock mode. This bit is otherwise ignored. 0: the TXCLK pin is set to a high impedance output. 1: the DLL clock output is driven onto the TXCLK pin. Rev. A | Page 27 of 60 AD9961/AD9963 Register Name DAC12 Config Register Address 0x40 DAC12A MSBs DAC12A LSBs 0x41 0x42 DAC12B MSBs DAC12B LSBs 0x43 0x44 DAC10B Config DAC10BMSBs DAC10BLSBs 0x45 0x46 0x47 Data Sheet Bit(s) 2 Parameter HD_BUSCTL 1 HD_CLKMD 0 FULL_DUPLEX 7 DAC12B_EN 6 DAC12A_EN 5 DAC12B_TOP 4 DAC12A_TOP 3:2 1 Unused AUXDAC_REF 0 DAC_UPDATE 7:0 7:4 3:0 7:0 7:4 3:0 7 DAC12A[11:4] Unused DAC12A[3:0] DAC12B[11:4] Unused DAC12B[3:0] DAC10B_EN 6:5 4:2 Unused DAC10B_TOP[2:0] 1:0 DAC10B_RNG[1:0] 7:0 7:2 1:0 DAC10B[9:2] Unused DAC10B[1:0] Function 0: selects SPI mode to control bus direction in half-duplex mode. 1: selects Pin mode to control bus direction in half-duplex mode. SPI bit to set Tx or Rx is Register 0x30, Bit 0. Register 0x30, Bit 1 is ignored in this case. 0: selects 1-clock submode if in half-duplex mode. 1: selects 2-clock submode if in half-duplex mode. 0: configures the digital interface for half-duplex mode (covers both 1clock and 2-clock submodes). 1: configures the digital interface for full-duplex mode. 0: powers down DAC12B. 1: enables DAC12B. 0: powers down DAC12A. 1: enables DAC12A. 0: sets DAC12B range to 3.3 x VAUXDACREF. 1: sets DAC12B range to 1.8 x VAUXDACREF. 0: sets DAC12A range to 3.3 x VAUXACREF. 1: sets DAC12A range to 1.8 x VAUXDACREF. Selects where the voltage reference for all of the auxiliary DACs is derived. 0: resistive divider from AUX33V. VAUXDACREF = VAUX33V /3.3. 1: selects the 1.0 V bandgap voltage. VAUXDACREF = 1.0 V. This bit determines which of the two data words updates all four of the auxiliary DACs. 0: update DACs after LSB write. 1: update DACs after MSB write. DAC12A voltage control word (upper eight bits). DAC12A voltage control word (lower four bits). DAC12B voltage control word (upper eight bits). DAC12B voltage control word (lower four bits). 0: powers down DAC10B. 1: enables DAC10B. Sets the DAC output voltage at the top range as follows: 000: 1.0 V. 001: 1.5 V. 010: 2.0 V. 011: 2.5 V. 100: 3.0 V. The total range of the DAC extends from top-of-range, to top-of-range minus the span. The span is set as: 00: 2.0 V. 01: 1.5 V. 10: 1.0 V. 11: 0.5 V. DAC10B voltage control word (eight most significant bits). DAC10Bvoltage control word (two least significant bits). Rev. A | Page 28 of 60 Data Sheet Register Name DAC10A Config DAC10A MSBs DAC10A LSBs TX BIST Control RX BIST Control TXI Check MSB TXI Check LSB TXQ Check MSB TXQ Check LSB Version Power Down 0 AD9961/AD9963 Register Address 0x48 0x49 0x4A 0x50 0x51 0x52 0x53 0x54 0x55 0x5C 0x60 Bit(s) 7 Parameter DAC10A_EN 6:5 4:2 Unused DAC10A_TOP[2:0] 1:0 DAC10A_RNG[1:0] 7:0 7:2 1:0 7:5 4 DAC10A[9:2] Unused DAC10A[1:0] Unused TX_PTTRN 3 TX_INSEL 2 TX_CONT 1 TX_START 0 TX_BISTEN 7:5 4 Unused RX_PTTRN 3 RX_INSEL 2 RX_CONT 1 RX_START 0 RX_BISTEN 7:0 7:0 7:0 7:0 7:0 7 TXI_CHK[15:8] TXI_CHK[7:0] TXQ_CHK[15:8] TXQ_CHK[7:0] Chip ID[7:0] DLL_EN 6 TXDAC_PD Function 0: powers down DAC10A. 1: enables DAC10A. Sets the DAC output voltage at the top range as follows: 000: 1.0 V. 001: 1.5 V. 010: 2.0 V. 011: 2.5 V. 100: 3.0 V. The total range of the DAC extends from top-of-range to top-of-range minus the span. The span is set as: 00: 2.0 V. 01: 1.5 V. 10: 1.0 V. 11: 0.5 V. DAC10A voltage control word (eight most significant bits). DAC10A voltage control word (two least significant bits). Unused Chooses the pattern type for the BIST sequence. 0: selects PRN output. 1: selects checker board pattern (0xA5A, 0x5A5, 0xA5A, ...). 0: selects pattern input from internal pattern generator. 1: selects pattern from the external pins of the Tx port. 0: runs the BIST for 512 cycles. 1: runs the BIST continuously. 0: keep the BIST engine in an idle state. 1: start the BIST sequence. 0: disable the BIST engine. 1: enable the BIST engine. Chooses the pattern type for the BIST sequence. 0: selects PRN output. 1: selects checker board pattern (0xA5A, 0x5A5, 0xA5A, ...). 0: selects pattern input from internal pattern generator. 1: selects pattern from the external pins of the Rx path. 0: runs the BIST for 512 cycles. 1: runs the BIST continuously. 0: keep the BIST engine in an idle state. 1: start the BIST sequence. 0: disable the BIST engine. 1: enable the BIST engine. MSB of the BIST signature value for the I side transmit path. LSB of the BIST signature value for the I side transmit path. MSB of the BIST signature value for the Q side transmit path. LSB of the BIST signature value for the Q side transmit path. Indicates device hardware revision number. Should read back as 0x08. 0: powers down DLL block. 1: enables DLL block. 1: powers down the bandgap reference voltage common to both transmit DACs and all of the auxiliary DACs. Rev. A | Page 29 of 60 AD9961/AD9963 Register Name Power Down 1 LDO Status Output Drive Register Address 0x61 0x62 0x63 Data Sheet Bit(s) 5 4 3 Parameter TXI_SLEEP TXQ_SLEEP CLK_PD 2 RXADC_PD 1 0 7 6 5 4 3 2 1 0 RXQ_SLEEP RXI_SLEEP Unused DLL_LDO_PD DLLBIAS_PD CLK_LDO_PD RX_LDO_PD RXF_LDO_PD AUXADC_PD AUX_REF_PD 7 6 5 4 3 2 1 0 DLL_LDO_STAT CLK_LDO_STAT RX_LDO_STAT RXF_LDO_STAT DIG_LDO_STAT Unused Unused RSET_SEL 7:6 TRXD_DRV 5:4 TRXIQ_DRV 3:2 TRXCLK_DRV 1:0 TXCLK_DRV Function 1: turns off IDAC output current. 1: turns off QDAC output current. 1: turns off clock receiver. This disables all clocks on the chip except for the serial port clock. 1: powers down main ADC clock and the bandgap reference voltage common to both receive ADCs. 1: powers down the Q ADC core. 1: powers down the I ADC core. 1: powers down LDO that supplies the DLL18V voltage rail. 1: powers down bias sub-block inside DLL block. 1: powers down LDO that supplies the CLK18V voltage rail. 1: powers down LDO that supplies the RX18V voltage rail. 1: powers down LDO that supplies the RX18VF voltage rail. 1: powers down AUXADC block. 1: powers down the auxiliary ADC voltage reference, allowing an external reference to be used. 1: LDO to DLL block is on (read only). 1: LDO to CLOCK block is on (read only). 1: LDO to ADC blocks is on (read only). 1: LDO to FLASH section of ADC is on (read only). 1: LDO to digital core is on (read only). 0: selects internal 10 k to generate 1 V reference. 1: selects external RSET to generate voltage reference. Controls the drive strength of the TRXD[11:0] pins. 00: 4 mA output drive. 01: 8 mA output drive. 10: 12 mA output drive. 11: not valid. Controls the drive strength of the TRXIQ pin. 00: 4 mA output drive. 01: 8 mA output drive. 10: 12 mA output drive. 11: not valid. Controls the drive strength of the TRXCLK pin. 00: 4 mA output drive. 01: 8 mA output drive. 10: 12 mA output drive. 11: not valid. Controls the drive strength of the TXCLK pin. 00: 4 mA output drive. 01: 8 mA output drive. 10: 12 mA output drive. 11: not valid. Rev. A | Page 30 of 60 Data Sheet Register Name Clock Mode I DAC Gain Ctrl 0 I DAC Gain Ctrl 1 I DAC Gain Ctrl 2 Q DAC Gain Ctrl 0 Q DAC Gain Ctrl 1 Q DAC Gain Ctrl 2 REFIO Adjust DLL Control 0 AD9961/AD9963 Register Address 0x66 0x68 0x69 0x6A 0x6B 0x6C 0x6D 0x6E 0x71 Bit(s) 7 6 5 4 3 2 1:0 Parameter TXI_DCLK TXQ_DCLK Unused RXI_DCLK RXQ_DCLK DCS_BP ADCDIV[1:0] 7:6 5:0 Unused IGAIN1[5:0] 7:6 5:0 Unused IGAIN2[5:0] 7:6 5:0 Unused IRSET[5:0] 7:6 5:0 Unused QGAIN1[5:0] 7:6 5:0 Unused QGAIN2[5:0] 7:6 5:0 Unused QRSET[5:0] 7:6 5:0 Unused REFIO_ADJ[5:0] 7 ADCCLKSEL 6 DACCLKSEL 5 4 Unused DLL_REF_EN Function 1: disables internal clock to I DAC. 1: disables internal clock to Q DAC. 1: disables internal clock to I ADC. 1: disables internal clock to Q ADC. 1: disables duty cycle stabilizer block. 00: selects divide by 1. Bypasses internal divider block for RXCLK. 01: selects divide by 1. Bypasses internal divider block for RXCLK. 10: selects divide by 2. 11: selects divide by 4. Linear in dB adjustment of the full-scale current of I DAC. Provides an adjustment range of approximately 6 dB in 0.25 dB steps. See Figure 57 for details. Linear adjustment of the full-scale current of I DAC. Provides an adjustment range of approximately 2.5% in 0.08% steps. See Figure 55 for details. Linear adjustment of the full-scale current of I DAC. Provides an adjustment range of approximately 20% in 0.625% steps. See Figure 55 for details. Linear in dB adjustment of the full-scale current of Q DAC. Provides an adjustment range of approximately 6 dB in 0.25 dB steps. See Figure 56 for details. Linear adjustment of the full-scale current of Q DAC. Provides an adjustment range of approximately 2.5% in 0.08% steps. See Figure 57 for details. Linear adjustment of the full-scale current of Q DAC. Provides an adjustment range of approximately 20% in 0.625% steps. See Figure 55 for details. Adjusts the on-chip reference voltage and output at REFIO. The transmit DAC full-scale currents and the auxiliary DAC full-scale voltages are proportional to the REFIO voltage. The approximate REFIO output voltage by code is: 000000: VREF = 1.0 V. 000001: VREF = 1.00625 V. ... 011111: VREF = 1.19375 V. 100000: VREF = 0.8 V. 100001: VREF = 0.80625 V. ... 111111 : VREF = 0.99375 V. 1: selects DLL output as the ADC sampling clock. 0: selects external clock as the ADC sampling clock. 1: selects DLL output as the DAC sampling clock. 0: selects external clock as the DAC sampling clock. 1: enables the input reference clock to the DLL. Rev. A | Page 31 of 60 AD9961/AD9963 Register Name Register Address DLL Control 1 0x72 DLL Control 2 0x75 Aux ADC Config and Conversion Start 0x77 Data Sheet Bit(s) 3:0 Parameter N[3:0] 7 6:5 DLL_Locked DLLDIV[1:0] 4:0 M[4:0] 7:4 3 0 DLL_RESB 2:0 7:6 0 CONV_TIME[1:0] Function Sets DLL divide ratio (1 to 8) at the output of the DLL. 0000: not valid. 0001: 1. 0010: 2. ... 0110: 6. 0111: not valid. 1000: 8. 1001: not valid. ... 1111: not valid. 1: DLL has locked to reference clock (read only). 00: DLL output is directly driven out. Divider is bypassed. 01: DLL output is directly driven out. Divider is bypassed. 10: DLL output is divided by 2. 11: DLL output is divided by 4. Sets DLL multiplication factor (1 to 32). 00000: 1. 00001: 2. ... 11111: 32. Set these bits to 0. Reset DLL. The DLL must be reset by a low to high transition on this bit each time the DLL configuration is changed or the reference frequency is changed. Set these bits to 0. Sets the number of AUXADCCLK cycles required to perform a conversion. 00: 20 AUXADCCLK cycles. 01: 22 AUXADCCLK cycles. 10: 26 AUXADCCLK cycles. 11: 34 AUXADCCLK cycles. Aux ADC MSBs Aux ADC LSBs 0x78 0x79 5:3 2:0 Unused AUXADC_CH[2:0] 7:0 7:4 3 AUXADC[11:4] AUXADC[3:0] CONV_COMPL 2:0 CHAN_SEL[2:0] Selects analog input channel to the auxiliary ADC. 000: AUXIN1, Pin 72. 001: AUXIO2, Pin 71. 010: AUXIO3, Pin 70. 011: internal VPTAT voltage. 100: internal VCMLI voltage. 101: internal VCMLQ voltage. 110: RXCML voltage. 111: not connected. Any write to this register initiates an ADC conversion cycle. This is the 8 MSBs of the most recent AUXADC conversion result. This is the 4 LSBs of the most recent AUXADC conversion result. 0: indicates that the request auxiliary ADC conversion is in progress. 1: indicates that the auxiliary ADC conversion result is valid. Indicates the actual auxiliary ADC input channel selected for the conversion. This should match the channel that was selected in the write to Register 0x77 that initiated the conversion. Rev. A | Page 32 of 60 Data Sheet Register Name Aux ADC CTRL 0 Aux ADC CTRL 1 ADC Full-Scale Adj Rx ADC Trim Ctrl AD9961/AD9963 Register Address 0x7A 0x7B 0x7D 0x7E Bit(s) 7 Parameter AUXADC_EN 6 RES 5:3 2:0 Unused AUXDIV[2:0] 7 6:5 4:2 TEMPSNS_EN Unused AUXREF_ADJ[2:0] 1:0 7:5 4:0 Unused Unused RX_FSADJ[4:0] 7 6 5 Unused RXTrim_EN RXTrim_Fine 4 AUXCML_EN 3:1 0 Function 0: powers down the auxiliary ADC clock. 1: enables the auxiliary ADC clock. 1: resets the AUXADC. A transition from 0 to 1 triggers the reset. The bit should be returned to 0 after issuing the reset. Sets the frequency division ratio of the input clock driving the CLKP, CLKN pins over the AUXADCCLK. 000: 256. 001: 128. ... 110: 4. 111: 2. The frequency of the AUXADCCLK should be less than 10 MHz. The sample conversion rate of the AUXADC is determined by the AUXCLK rate and CONV_TIME. 1: enables the on-chip temperature sensor. Adjustment for tuning the internal auxiliary ADC reference voltage. 011: +18 mV. 010: +12 mV. 001: +6 mV. 000: default. 111: -6 mV. 110: -12 mV. 101: -18 mV. 100: -24 mV. This parameter adjusts the full-scale input voltage range of the Rx path ADCs. The peak-to-peak input voltage range can be set as follows: 10000: 1.25 V. 10001:1.27 V. 10010: 1.29 V. 10011: 1.31 V. ... 11111: 1.54 V. 00000: 1.56 V. 00001: 1.58 V. ... 01110: 1.873 V. 01111: 1.875 V. 1: enables ADC gain calibration. 1: decreases the step size (increases resolution) of the gain calibration adjustment. Controls the buffers of internal bias points within each of the Rx ADCs to allow for checking of this voltage. These voltages should read back about 0.9 V. 0: disables the buffers. 1: enables the buffers. Set to 000. Rev. A | Page 33 of 60 AD9961/AD9963 Register Name Register Address IGAIN CAL MSBs IGAIN CAL LSBS 0x7F 0x80 IGAIN CAL MSBs 0x81 IGAIN CAL LSBs 0x82 DDLL Lock Bits IGAIN CAL LSBS 0x84 0xFF Data Sheet Bit(s) 0 Parameter RX_DC Function 0: the ADC common-mode buffer is active. This sets the ADC inputs to the desired common-mode voltage through 10 k resistors to each single sided input. 1: disables the common-mode buffer. The buffer should be disabled whenever the user DC couples to the ADC inputs. 7:0 7:3 2:1 0 7:0 RXI_Trim[9:2] Unused RXI_Trim[1:0] GAINCAL_ENI RXQ_Trim[9:2] The RXI_Trim[9:0] word is used to adjust the gain of the receive path I ADC. These bits have no effect unless the RXTrim_EN bit is set. The RXTrim_Fine bit reduces the LSB size of the calibration word by 1/2. 7:3 2:1 0 1 Unused RXQ_Trim[1:0] GAINCAL_ENQ TXDDLL lock bit 0 RXDDLL lock bit 7:1 0 Unused Update 1: enables the gain calibration DAC for the I Rx ADC. The RXQ_Trim[9:0] word is used to adjust the gain of the receive path Q ADC. These bits have no effect unless the RXTrim_EN bit is set. The RXTrim_Fine bit reduces the LSB size of the calibration word by 1/2. Bottom two LSBs of RXQ_Trim described in Register 0x81 above. 1: enables the gain calibration DAC for the Q Rx ADC. 0: TXDDLL is unlocked. 1: TXDDLL is locked. 0: RXDDLL is unlocked. 1: RXDDLL is locked. Synchronously transfers ADC configuration data from the global register set to the local ADC register set and activates the changes. A 0-to-1 transition is required to initiate the transfer. 1: transfer ADC parameters to ADC to make changes active. Rev. A | Page 34 of 60 Data Sheet AD9961/AD9963 RXBIAS RECEIVE PATH Rx Path General Description The AD9961/AD9963 Rx paths consist of dual, differential input, 100 MSPS ADCs followed by an optional 2x decimation filter. The Rx path also has digital offset and gain adjustments. I OFFSET DECIMATION SCALE RXQN RXQP TRXD[11:0] RXCML TRXIQ The RXCML pin of the AD9961/AD9963 provides the user with a buffered version of the expected ADC common-mode bias voltage. The RXCML output nominally is at 1.4 V. Bypassing the RXCML output to analog ground maintains the stability of the output buffer and lowers the noise. To maintain the accuracy of the RXCML bias voltage, the current draw from the pin should be kept below 1 mA. LPF 1/2 DATA ASSEMBLER Q ADC LPF 1/2 TRXCLK Q OFFSET Figure 39. Receive Path Block Diagram RXIP The dual ADC paths share the same clocking and reference circuitry to provide optimal matching characteristics. The ADCs have a multistage differential pipelined switched capacitor architecture with output error correction logic. The ADCs support IF sampling frequencies up to 140 MHz, making them suitable for undersampling receivers. Also, one of the ADCs can be powered down and the digital interface can be placed into single ADC mode. This flexibility makes the part well-suited for sampling real signals as well. 2k IADC 2k RXIN REG 0x7E[0] PD ~1.4V RXQP 2k QADC CMBIAS 2k RECEIVE ADC OPERATION RXQN The Rx path analog inputs look into a nominal differential impedance of 4 k. The Rx inputs are self-biasing, so they can be either ac-coupled or direct coupled. The nominal dc bias level of the inputs is 1.4 volts. A buffered version of the bias voltage is available at the RXCML pin. This voltage can be used for biasing external buffer circuits when dc coupling is required. For optimal dynamic performance, the analog inputs should be driven differentially. The source impedances driving the Rx inputs should be matched so that common-mode settling errors are symmetrical. The Rx inputs can be driven with a singleended source, but SNR and SINAD performance is degraded. ADC Reference Voltage An internal differential voltage reference creates positive and negative reference voltages that define the full-scale input voltage of the ADCs. This full-scale input voltage range can be adjusted by means of the RX_FSADJ[4:0] parameter in configuration Register 0x7D. See the Configuration Registers section for more details on setting the voltage. REG 0x0F[1] RXCML ~1.4V EN AD9961/AD9963 08801-012 I ADC 08801-112 RXIP RXIN The AD9961/AD9963 provide the user with the option to place a 10 k resistor between the RXBIAS pin and ground. This resistor is used to set the master current reference of the ADC core. The RXBIAS resistor should have a tolerance of 1% or better to preserve the accuracy of the ADC full-scale range. Care should be taken in the layout to avoid any noise from coupling into the RXBIAS pin. Figure 40. Simplified Schematic of Rx Path Inputs Differential Input Configurations Optimum performance is achieved by driving the analog inputs in a differential input configuration. For baseband applications, the ADA4937 differential driver provides excellent performance and a flexible interface to the ADC. Figure 41 shows an ac-coupled input configuration. The VOCM pin should be connected to a voltage that provides sufficient headroom for the output driver of the differential amp. Usually, setting VOCM to 1/2 of the amplifier supply voltage is the optimal setting. Placing source resistance in series with the amplifiers outputs isolates the amplifier from on-board parasitic capacitances and leads to more stable operation. The nominal input voltage range is 1.56 V. In general, a tradeoff can be made between linearity and SNR. Increasing the input voltage range leads to higher SNR. Decreasing the input voltage range leads to better linearity. Rev. A | Page 35 of 60 AD9961/AD9963 Data Sheet 200 Single-Ended Input Configuration ADA4937 1k 1k -VIN 33 0.1F 33 0.1F Driving the Rx inputs with a single-ended signal typically limits the achievable ADC performance. When using this configuration, best performance is achieved by maintaining a balanced impedance off each of the Rx inputs as shown in Figure 44. RXIP VOCM RXIN 0.1F 200 49.9 200 AD9961/ AD9963 1.25V p-p 08801-013 VCC The output common-mode voltage of the ADA4937 is set to match the common-mode voltage required by the ADC by connecting the RXCML output to the VOCM input of the amplifier. The RXCML output nominally is at 1.4 V. Bypassing the RXCML output to analog ground maintains the stability of the output buffer and lowers the noise. 200 ADA4937 200 33 RXIP VOCM RXIN -VIN 33 200 C RXIP 0.1F ADC *CDIFF 25 Figure 41. Differential Input Configuration, AC-Coupled +VIN 49.9 33 AD9963 33 0.1F RXIN C 08801-016 200 +VIN *CDIFF IS OPTIONAL. Figure 44. Single-Ended Input Configuration Interfacing to the ADF4602 Rx Baseband Outputs The ADF4602 is an RF transceiver suitable for femtocell and other wireless communications applications. The ADF4602 Rx baseband outputs have a nominal output common-mode voltage that can be set to 1.4 V. The ADF4602 can be dccoupled to the AD9963. It is recommended that a first-order low-pass filter be placed between the two devices to reject unwanted high frequency signals that may alias into the desired baseband signal. 200 68pF RXBBI ADC At higher input frequencies, the amplifiers required to maintain the full dynamic power of the AD9963 requires considerable supply current. For higher frequency power sensitive applications, differential transformer coupling is the recommended input configuration. The signal characteristics must be considered when selecting a transformer. Most RF transformers saturate at frequencies below a few megahertz, and excessive signal power can also cause core saturation, which leads to distortion. In any configuration, the value of the shunt capacitor, C, is dependent on the input frequency and may need to be reduced or removed. 50 C RXIP ADC 0.1F AD9963 *CDIFF 33 0.1F RXIN C *CDIFF IS OPTIONAL. 08801-015 1.25V p-p 33 RXIP AD9963 Figure 42. Differential Input Configuration, DC-Coupled ADT1-1WT 1:1 Z RATIO 100 RXBBIB ADF4602 100 RXIN 68pF 08801-118 AD9961/ AD9963 08801-014 RXCML 0.1F Figure 45. ADF4602 to AD9963 Receive Interface Circuit In this application, the ADF4602 is setting the common-mode input voltage of the AD9963 ADCs. The input common-mode buffer of the AD9963 should be disabled (set Register 0x7E, Bit 1 = 1) to avoid contention with the ADF4602 output driver. DECIMATION FILTER AND DIGITAL OFFSET Decimation Filter The I and Q receive paths each have a bypassable 2x decimating low-pass filter. The half-band digital filter reduces the output sample rate by a factor of 2 while rejecting aliases that fall into the band of interest. These low-pass filters provide >7 dB of stop-band rejection for 40% of the output data rate. When used with quadrature signals, the complex output band is 80% of the quadrature output data rate. A graph of the pass-band response of the decimation filter is shown in Figure 46. Figure 43. Differential Transformer--Coupled Configuration Rev. A | Page 36 of 60 Data Sheet AD9961/AD9963 ADC Digital Offset Adjustment 0 -60 The Rx paths also have individual digital offsets that can be applied to the data captured by the ADCs. The offset is a 6-bit digital value that is added directly to the LSBs of the ADC output data. The offset values are configured by first addressing the ADC by setting the appropriate address in Register 0x05, then writing the desired offset (in LSBs) into Register 0x10. For example, to set offsets of +6 and -2 to the I and Q channels respectively, the register write sequence is: -70 1. Write 0x01 into Register 0x05. This addresses the I channel ADC. 2. Write 0x06 into Register 0x10. This sets the IADC_Offset value to +6 LSBs. 3. The filter coefficients of the 2x decimation low-pass are shown in Table 16. Write 0x02 into Register 0x05. This addresses the Q channel ADC. 4. Table 16. Write 0xFE into Register 0x10. This sets the QADC_Offset value to -2 LSBs. 5. Write 0x01 into Register 0xFF. This updates the data path registers and applies the offset to the data. 6. Write 0x00 into Register 0x05. This returns the SPI to the normal addressing mode. -10 MAGNITUDE (dBc) -20 -30 -40 -50 1.00 NORMALIZED FREQUENCY (Relative to fDAC ) 08801-119 0.95 0.90 0.85 0.80 0.75 0.70 0.65 0.60 0.55 0.50 0.45 0.40 0.35 0.30 0.25 0.20 0.15 0.10 0 0.05 -80 Figure 46. Pass-Band Response of the Rx Path Decimation Filter Lower Coefficient H(1) H(3) H(5) H(7) H(9) H(11) H(13) H(15) H(17) H(19) H(21) H(22) Upper Coefficient H(43) H(41) H(39) H(37) H(35) H(33) H(31) H(29) H(27) H(25) H(23) Value 12 -32 72 -140 252 -422 682 -1086 1778 -3284 10364 16384 Rev. A | Page 37 of 60 AD9961/AD9963 Data Sheet Tx Path General Description The transmit section consists of two complete paths of interpolation filters stages, each followed by a high speed current output DAC. A data assembler receives interleaved data from one of two digital interface ports, and de-interleaves and buffers the data before supplying the data samples into the two datapaths. The interpolation filter bank consists of three stages that can be completely bypassed or cascaded to provide 2x, 4x, or 8x interpolation. The supported clock rates for each of the interpolation filters and the transmit DACs are listed in Table 1. TXVDD 0 -10 -20 -30 -40 -50 I GAIN I SCALE -60 TXIP I DAC TX PORT -70 TXIN TXQN 1.00 08801-122 0.95 0.90 0.85 0.80 0.75 0.70 0.65 0.60 0.55 0.50 0.45 0.40 0.35 0.30 0.25 NORMALIZED FREQUENCY (Relative to fDAC ) Q DAC LPF 1/2/4/8x 0.20 0 TXQP TRX PORT 0.15 -80 TXCML Q SCALE 0.10 LPF 1/2/4/8x 0.05 DATA ASSEMBLER AND FIFO than 70 dB. In 8x interpolation mode, the image rejection is greater than 65 dB. The usable bandwidth of the filters is typically limited by the stop-band attenuation they provide, rather than the passband flatness. The transfer functions of the interpolation filters configured for 2x, 4x, and 8x interpolation ratios are shown in Figure 49 through Figure 51. MAGNITUDE (dBc) TRANSMIT PATH Figure 49. Digital Filter Transfer Function for 2x Interpolation 0 Q GAIN REFIO Rev. A | Page 38 of 60 08801-123 1.00 0.95 0.90 0.85 0.80 0.75 0.70 0.65 0.60 0.55 0.50 0.45 0.40 0.35 0.30 0.25 0.20 0.15 0.10 -50 -60 -70 NORMALIZED FREQUENCY (Relative to fDAC ) Figure 51. Digital Filter Transfer Function for 8x Interpolation 08801-124 1.00 0.95 0.90 0.85 0.80 0.75 0.70 0.65 0.60 0.55 -80 0.50 The INT0 and INT1 filters have bandwidths of 40% of the input data rate. Over their usable bandwidth, the filters have a passband ripple of less than 0.1 dB. The SRRC has a roll-off factor of 0.22 with a 60 dB stop-band attenuation. In 2x and 4x interpolation modes, the interpolation filters have an image rejection of greater -40 0.45 The digital filters should be cascaded such that INT0 is enabled for an interpolation factor of 2x, INT0 and INT1 should be enabled for an interpolation factor of 4x, and INT0, INT1, and the SRRC should be enabled for an interpolation factor of 8x. -30 0.40 Figure 48. Block Diagram of Transmit Datapath -20 0.35 INT1_SCALE 0x37[4:0] -10 0.30 0 0.25 INT0_SCALE 0x36[4:0] INT1 0 TO DAC 0.20 0 1 Figure 50. Digital Filter Transfer Function for 4x Interpolation 0.15 SRRC_SCALE 0x35[4:0] 1 INT0 NORMALIZED FREQUENCY (Relative to fDAC ) 0.10 0 -80 0 SRRC -70 MAGNITUDE (dBc) FROM FIFO -60 INT1_BP 0x30[4] 08801-018 1 -50 0.05 INT0_BP 0x30[3] -40 0 The I and Q transmit paths contain three interpolation filters designated as INT0, INT1, and SRRC. Each of the interpolation filters provides a 2x increase in output data rate. The filters can be completely bypassed or cascaded to provide 2x, 4x, or 8x upsampling ratios. The interpolation filters effectively increase the DAC update rate while suppressing the images at the input date rate. This reduces the requirements on the analog output reconstruction filter. -30 0.05 INTERPOLATION FILTERS -20 MAGNITUDE (dBc) Figure 47. Transmit Path Block Diagram SRRC_BP 0x30[2] -10 08801-017 RFSADJ Data Sheet AD9961/AD9963 Interpolation Filter Coefficients The interpolation filters, INT0 and INT1, are half-band filters implemented with a symmetric set of coefficients. Every other coefficient (even coefficients) except the center coefficient is zero. The coefficient values for the three interpolation filters are listed in Table 17 to Table 19. Table 17. Coefficient Values for INT0 Lower Coefficient H(1) H(3) H(5) H(7) H(9) H(11) H(13) H(15) H(17) H(19) H(21) H(22) Upper Coefficient H(43) H(41) H(39) H(37) H(35) H(33) H(31) H(29) H(27) H(25) H(23) Value 12 -32 72 -140 252 -422 682 -1086 1778 -3284 10364 16384 Table 18. Coefficient Values for INT1 Lower Coefficient H(1) H(3) H(5) H(7) H(9) H(10) Upper Coefficient H(19) H(17) H(15) H(13) H(11) Value 26 -138 466 -1314 5058 8191 Table 19. Coefficient Values for SRRC Filter Lower Coefficient H(1) H(2) H(3) H(4) H(5) H(6) H(7) H(8) H(9) H(10) H(11) H(12) H(13) H(14) H(15) H(16) H(17) H(18) H(19) H(20) H(21) H(22) H(23) H(24) H(25) H(26) H(27) Upper Coefficient H(53) H(52) H(51) H(50) H(49) H(48) H(47) H(46) H(45) H(44) H(43) H(42) H(41) H(40) H(39) H(38) H(37) H(36) H(35) H(34) H(33) H(32) H(31) H(30) H(29) H(28) Value -2 -2 8 -4 -21 10 44 -29 -79 66 123 -127 -183 232 251 -394 -326 642 401 -1034 -469 1704 523 -3160 -560 9996 16383 Data Flow and Clock Generation The transmit port TXD[11:0] and TXIQ signals are captured from by the device with an input latch. The data is then formatted and buffered in an 8-word deep FIFO. The data exits the FIFO and is processed by whichever interpolation filters are enabled. The data is then sampled by the transmit DACs. The FIFO absorbs any phase drift between the two clock domains that drive the transmit data. The data is read from the FIFO by the RDCLK signal. The RDCLK signal is always the DACCLK divided by the interpolation ratio, I. Data is written to the FIFO by the WRCLK signal at the quadrature data input rate, fDATA. fDATA is equal to one-half the bus speed because the I and Q samples are interleaved. Figure 52 shows the block diagram of the transmit path data flow in full-duplex mode. Also shown in the diagram are the input data clocking options and the clock doubler selections. Rev. A | Page 39 of 60 AD9961/AD9963 Data Sheet 24 BITS WRITE POINTER READ POINTER REG 0 REG 1 12 I DATA PATH 12 12 Q DATA PATH 12 REG 2 INPUT LATCH 26 24 DATA FORMAT REG 3 REG 4 REG 5 REG 6 TX_BNRY Reg 0x31[0] WRCLK TXIQ_HILO Reg 0x31[2] 1 0 FIFO_OFFSET Reg 0x33[2:0] FIFO_PTR Reg 0x34[7:0] EN TXCLK 0 1 EN FIFO RDPTR = 0 TXCKI_INV Reg 0x31[3] TX_IFIRST Reg 0x31[1] 1 0 WRPTR TXCLK_MD Reg 0x31[0] Q DAC REG 7 RDCLK /I /2 0 1 FIFO RESET AND MONITOR 0 1 DOUBLER TX_SDR Reg 0x31[7] TXCLK_MD Reg 0x31[1] TXSMPCLK I=1 * I DENOTES INTERPOLATION RATIO TXCKO_INV Reg 0x31[6] DACCLK TXDBL_SEL TX_DBLPW[2:0] Reg 0x39[0] Reg 0x3E[5:3] 08801-150 13 TXD[11:0] TXIQ I DAC Figure 52. Transmit Path Data Flow and Clock Generation In Full Duplex Mode TXVDD The signal on the TXCLK pin can be configured as either an input or an output. This is configured by the TXCLK_MD variable (Register 0x31, Bits[5:4]). Whether configured as an input or an output, the TXCLK signal has the option of being inverted by configuring the TXCKI_INV or TXCKO_INV variables. 100A REFIO REFIO_ADJ[5:0] 0x6E[5:0] IRSET[5:0] 0x6A[5:0] The transmit path clock doubler is only used when all of the interpolation filters are bypassed (I = 1) and the transmit path is configured in bus rate mode (TX_SDR = 1). For more information about configuring the clock doubler, see Table 22. IGAIN1[5:0] 0x68[5:0] IGAIN2 0x69[5:0] RREF RSET TX1P TXDATA IDAC TXCML DACCLK TRANSMIT DAC OPERATION 08801-019 TX1N Figure 53. Simplified Block Diagram of I DAC Core Figure 53 shows a simplified block diagram of one of the transmit path DACs. Each DAC consists of a current source array, switch core, digital control logic, and full-scale output current control. The DAC contains a current source array capable of providing a nominal full-scale current (IOUTFS) of 2 mA. The output currents from the TXIP and TXIN pins are complementary, meaning that the sum of the two currents always equals the full-scale current of the DAC. The digital input code to the DAC determines the effective differential current delivered to the load. The DACs are powered through the TXVDD pin and can operate over a 1.8 V to 3.3 V supply range. To facilitate interfacing the output of the AD9961/AD9963 directly to a range of commonmode levels, an internal bias voltage is made available through the TXCML pin. The DAC full-scale output current is regulated by the reference control amplifier and is determined by the product of a reference current, a programmable reference resistor, RREF, an internal programmable resistor, RSET, and a pair of programmable gain scaling parameters. Transmit DAC Transfer Function The output currents from the TXIP and TXIN pins are complementary, meaning that the sum of the two currents always equals the full-scale current of the DAC. The digital input code to the DAC determines the effective differential current delivered to the load. TXIP provides maximum output current when all bits are high. The output currents vs. DACCODE for the DAC outputs are expressed as: DACCODE I TXIP I OUTFS 2N (1) I TXIN I OUTFS I TXIP (2) where DACCODE = 0 to 2N - 1. There are a number of adjustments that can be made to scale IOUTFS to provide programmability in the output signal level. Rev. A | Page 40 of 60 Data Sheet AD9961/AD9963 Transmit Path Gain Adjustment Table 20. Reference Operation Adjusting the output signal level is implemented by scaling the full-scale output current of the transmit DAC. There are four separate programmable parameters that can be used to adjust the full-scale output of the DACs; the REFIO voltage, the RSET resistance, and the fine and coarse gain control parameters. Reference Mode Internal REFIO Pin Connect 0.1 F capacitor External Apply external reference Register Setting Register 0x60, Bit 6 = 0 (default) Register 0x60, Bit 6 = 1 (disables internal reference) Adjusting the REFIO Voltage There is a single reference voltage that is used by both the I and Q channel DACs. The REFIO reference voltage is generated by an internal 100 A current source terminated into a programmable resistor, RREF. The nominal RREF resistance is 10 k resulting in a 1.0 V reference voltage. The resistance can be varied by adjusting the REFIO_ADJ[5:0] bits in Register 0x6E. This adds or subtracts up to 20% from the RREF resistance and hence the REFIO voltage and the DAC full-scale current. A secondary effect to changing the REFIO voltage is that the full-scale voltage in the auxiliary DACs also changes by the same magnitude. Adjusting the Current Scaling Resistor Each transmit DAC has a resistor that is used to adjust the fullscale current. The nominal resistance is 16 k, which results in a full-scale current of 2 mA (when VREFIO equals 1.0 V). The 6-bit programmable values, IRSET[5:0] and QRSET[5:0] (Register 0x6A and Register 0x6D), provide an output current adjustment range of 20% as shown in Figure 55. 2.6 2.4 FSC (mA) The register uses twos complement format in which 011111 maximizes the voltage on the REFIO node and 100000 minimizes the voltage. A curve illustrating the variation of REFIO voltage vs. REFIO_ADJ value is shown in Figure 54. 1.3 2.2 2.0 1.2 1.8 1.6 0 1.0 8 16 24 32 40 48 56 RSET () 08801-021 VREF (V) 1.1 Figure 55. Output Current Scaling vs. IRSET and QRSET Values 0.9 Adjusting the GAIN Parameters 0.7 0 8 16 24 32 40 48 56 REFIO ADJ 08801-020 0.8 Figure 54. Typical VREFIO Voltage vs. REFIO_ADJ Value The REFIO pin should be decoupled to AGND with a 0.1 F capacitor. If the voltage at REFIO is to be used for external purposes, an external buffer amplifier with an input bias current of less than 100 nA should be used. Each transmit DAC has coarse and fine gain control parameters for scaling the full-scale output currents. These adjustments change only the full-scale current of the DAC and have no impact on the REFIO voltage. The coarse scale adjust (GAIN1) allows the nominal output current to be changed by 6 dB in approximately 0.25 dB steps. The adjustment range of the fine scale adjust (GAIN2) is about 2.5%. Figure 56 and Figure 57 show the resulting gain scaling vs. the GAIN1 and GAIN2 parameters. An external reference can be used in applications requiring tighter gain tolerances or lower temperature drift. Also, a variable external voltage reference can be used to implement a method for gain control of the DAC output. The external reference is applied to the REFIO pin. Note that the 0.1 F compensation capacitor is not required. The internal reference can be directly overdriven by the external reference, or the internal reference can be powered down. The input impedance of REFIO is 10 k when powered up and 1 M when powered down. Rev. A | Page 41 of 60 AD9961/AD9963 Data Sheet Figure 58 shows the most basic DAC output circuitry. A pair of resistors, RO, are used to convert each of the complementary output currents to a differential voltage output, VOUTX. Because the current outputs of the DAC are very high impedance, the differential driving point impedance of the DAC outputs, ROUT, is equal to 2 x RO. 8 6 4 dBFS 2 0 Figure 59 illustrates the output voltage waveforms. -2 VPEAK -4 VCM -6 9 17 25 33 41 49 57 GAIN1 VN VP Figure 56. Typical DAC Full-Scale Current vs. GAIN1 Code VOUT 2.06 -VPEAK Figure 59. Voltage Output Waveforms 2.04 FULL-SCALE CURRENT (mA) 08801-025 1 08801-022 0 -8 The common-mode signal voltage, VCM, is calculated as: 2.02 VCM = 2.00 I FS x RO 2 The peak output voltage, VPEAK, is calculated as: VPEAK = I FS x RO 1.98 With this circuit configuration, the single-ended peak voltage is the same as the peak differential output voltage. 1.96 Setting the TXCML Pin Voltage 0 8 16 24 32 40 48 56 GAIN2 08801-023 1.94 Figure 57. Typical DAC Full-Scale Current vs. GAIN2 Code TRANSMIT DAC OUTPUTS The optimum noise and distortion performances of the AD9961/ AD9963 are realized when they are configured for differential operation. The common-mode error sources of the DAC outputs are significantly reduced by the common-mode rejection of a transformer or differential amplifier. These common-mode error sources include even-order distortion products and noise. The enhancement in distortion performance becomes more significant as the frequency content of the reconstructed waveform increases and/or its amplitude increases. This is due to the firstorder cancellation of various dynamic common-mode distortion mechanisms, digital feedthrough, and noise. + VIP TXIP RO RO TXIN VOUTI VIN - VQP + TXQP RO TXQN RO VOUTQ VQN - When the TXVDD supply is 1.8 V, the output signal commonmode voltage should be kept close to 0 V and the TXCML pin should be connected directly to ground. When the TXVDD supply is 3.3 V, the output signal common mode can be operated as high as 1.25 V. The circuit shown in Figure 60 shows a typical output circuit configuration that provides a non zero bias voltage at the TXCML pin. Resistance values of 499 for RL and 249 for RCML produces a 2 V p-p differential output voltage swing with a 1.0 V output common-mode voltage and a voltage of 0.5 V supplied to the TXCML pin. The 2 mA full-scale current flows through the 249 RCML creating the 0.5 V TXCML voltage. The decoupling capacitor, assures a low ac driving impedance for the TXCML pin. 08801-024 TXCML The TXCML pin serves to change the DAC bias voltages in the part, allowing it to operate with higher output signal commonmode voltages. When the output signal common mode is below 0.8 V, the TXCML pin should be tied directly to AGND. When the output signal common mode is greater then 0.8 V, then the TXCML pin should be set to 0.5 V. The TXCML pin should be a low ac impedance source (capacitive decoupling is recommended). Figure 58. Basic Transmit DAC Output Circuit Rev. A | Page 42 of 60 Data Sheet AD9961/AD9963 center tap of the transformer should provide a path for this dc current. In most applications, AGND provides the most convenient voltage for the transformer center tap. The complementary voltages appearing at TXIP and TXIN (that is, VIOUTP and VIOUTN) swing symmetrically around AGND and should be maintained with the specified output compliance range of the AD9961/AD9963. AD9961/AD9963 + TXIP 65 RL VOUT RL R TXIN 66 - TXCML 62 RCML 08801-030 C Figure 60. Circuit for Setting TXCML Level Using RCML Transmit DAC Output Circuit Configurations The following section illustrates some typical output configurations for the AD9961/AD9963 transmit DACs. Unless otherwise noted, it is assumed that IOUTFS is set to a nominal 2.0 mA. For applications requiring the optimum dynamic performance, a differential output configuration is suggested. A differential output configuration can consist of either an RF transformer or a differential op amp configuration. The transformer configuration provides the optimum high frequency performance and is recommended for any application that allows ac coupling. The differential op amp configuration is suitable for applications requiring dc coupling, signal gain, and/or a low output impedance. A single-ended output is suitable for applications where low cost and low power consumption are primary concerns. A differential resistor, RDIFF, can be inserted in applications where the output of the transformer is connected to the load, RLOAD, via a passive reconstruction filter or cable. RDIFF, as reflected by the transformer, is chosen to provide a source termination that results in a low voltage standing wave ratio (VSWR). Note that approximately half the signal power is dissipated across RDIFF. Differential Buffered Output Using an Op Amp A dual op amp (see the circuit shown in Figure 62) can be used in a differential version of the single-ended buffer shown in Figure 63. The same R-C network is used to form a one-pole, differential, low-pass filter to isolate the op amp inputs from the high frequency images produced by the DAC outputs. The feedback resistor, RFB, determines the differential peakto-peak signal swing by the formula VOUT = 2 x RFB x IFS The minimum single-ended voltages out of the amplifier are, respectively, Differential Coupling Using a Transformer An RF transformer can be used to perform a differential-tosingle-ended signal conversion, as shown in Figure 61. The distortion performance of a transformer typically exceeds that available from standard op amps, particularly at higher frequencies. Transformer coupling provides excellent rejection of common-mode distortion (that is, even-order harmonics) over a wide frequency range. It also provides electrical isolation and can deliver voltage gain without adding noise. Transformers with different impedance ratios can also be used for impedance matching purposes. The main disadvantages of transformer coupling are low frequency roll-off, lack-of-power gain, and high output impedance. VMIN = VMAX - RFB x IFS The common-mode voltage of the differential output is determined by the formula VCM = VMAX - RFB x IFS CF RFB RB AD9961/AD9963 RS - TXIP 65 ADA4841-2 + REFIO 63 VOUT C TXGND 64 AD9961/AD9963 + RS TXIN 66 ADA4841-2 TXIP 65 - RLOAD RB OPTIONAL RDIFF 08801-031 TXIN 66 RFB Figure 62. Single-Supply Differential Buffer Figure 61. Differential Output Using a Transformer The center tap on the primary side of the transformer must be connected to a voltage that keeps the voltages on TXIP and TXIN within the output common-mode voltage range of the device. Note that the dc component of the DAC output current is equal to IOUTFS and flows out of both TXIP and TXIN. The Rev. A | Page 43 of 60 08801-032 CF AD9961/AD9963 Data Sheet Single-Ended Buffered Output Using an Op Amp Interfacing to the ADF4602 An op amp such as the ADA4899-1 can be used to perform a single-ended current-to-voltage conversion, as shown in Figure 63. The AD9961/AD9963 are configured with a pair of series resistors, RS, off each output. For best distortion performance, RS should be set to 0 . The feedback resistor, RFB, determines the peak-to-peak signal swing by the formula The ADF4602 is an RF transceiver suitable for Femtocell and other wireless communications applications. The ADF4602 Tx baseband inputs have a nominal input common-mode voltage requirement of 1.2 V. The AD9963 can be dc coupled to the ADF4602 as shown in Figure 64. When configured for a 2 mA full-scale current, the output swing of the circuit is 1 V ppd centered at 1.2 V. The TXMCL pin is biased at 0.5 V to increase the headroom of the DAC outputs. The TXVDD and CLK33V supplies must be supplied with 3.3 V to support this output compliance range from the DACs. VOUT = RFB x IFS The maximum and minimum voltages out of the amplifier are, respectively, VMAX = VREFIO TXBBI VMIN = VMAX - IFS x RFB TXIP 249 CF 249 +5V TXBBIB ADA4899-1 TXBBQ TXQP 249 C -5V TXGND 64 ADF4602 08801-033 RS AUXIO2 VOUT + REFIO 63 TXIN 66 TXIN 100k - TXIP 65 TXCML 0.1uF 249 226 AD9963 249 TXQN TXBBQB Figure 63. Single-Supply Single-Ended Buffer 100k AUXIO3 Figure 64. AD9963 to ADF4602 Tx Interface Circuitry The optional 100 k resistors connected between the AUXIO pins and the TXIN (and TXQN) pins allow a dc offset to be provided to null out LO feedthrough at the ADF4602 outputs. Rev. A | Page 44 of 60 08801-142 RS 249 RFB RB AD9961/AD9963 226 Data Sheet AD9961/AD9963 DEVICE CLOCKING clock duty cycle. The DCS can be bypassed. Recommendations for using the DCS can be found in the Clock Duty Cycle Considerations section. CLOCK DISTRIBUTION The clock distribution diagram shown in Figure 65 gives an overview of the clocking options for each of the data converters. The receive path ADCs and the transmit path DACs can be clocked directly from the CLKP/CLKN inputs or from the output of the on-chip DLL. The auxiliary ADC sampling clock is always a divided down version of the input clock. The auxiliary DACs are updated synchronously with the serial port clock and have no relationship with the CLKP/CLKN inputs. The ADC clock divider and the DLL clock multiplication supports a variety of ratios between the receive path ADC sampling clock and the transmit path DAC sampling clock. Table 21 details the specific values the device supports and which register bits are require configuration. Table 21. Clock Tree Configuration Variables Variable DCS_BP ADCDIV ADCCLKSEL DACCLKSEL N M DLLDIV AUXDIV The best data converter performance is realized when a low jitter clock source drives the CLKP/CLKN inputs, and this signal is used directly (or through the on-chip divider) as the data converter sampling clocks. The ADC and DAC sampling clocks are independently selected to be derived from either the CLKP/CLKN input or from the DLL output, DLLCLK. Using DLLCLK as the data converter sampling clock signal may degrade the noise and SFDR performance of the converters. More information is given in the Clock Multiplication Using the DLL section. Values 0 or1 1, 2, 4 0 or 1 0 or 1 1 to 6, 8 1, 2, 3,..., 32 1, 2, 4 2J, J = 1 to 8 The receive path ADC has a duty cycle stabilizer (DCS) to help make the ADC performance insensitive to changes in the input ADCCLKSEL DOUBLER AND /ADCDIV DCS_BP ADC 0 1 CLK_PD 1 DCS CLKP 0 DLL /DLLDIV M N EXTDLLCLK DAC DLLCLK 1 DACCLK 0 DACCLKSEL AUXADC AUXCLK /AUXDIV Figure 65. Clock Distribution Diagram Rev. A | Page 45 of 60 08801-300 CLKN ADCCLK Address Register Bit(s) 0x66 2 0x66 [1:0] 0x71 7 0x71 6 0x71 [3:0] 0x72 [4:0] 0x72 [6:5] 0x7A [2:0] AD9961/AD9963 Data Sheet DRIVING THE CLOCK INPUT AD9510/AD9511/ AD9512/AD9513/ AD9514/AD9515/ AD9516/AD9518 0.1F CLK+ CLK_P 0.1F CLK- 100 LVDS DRIVER 0.1F 50* 08801-035 50* ADC AD9963 CLK_N CLK *50 RESISTORS ARE OPTIONAL. Figure 66. Differential LVDS Sample Clock In applications where the receive analog input signals and the transmit analog output signals are at low frequencies, it is acceptable to drive the sample clock inputs with a single-ended CMOS signal. In such applications, CLKP should be driven directly from a CMOS gate, and the CLKN pin should be bypassed to ground with a 0.1 F capacitor in parallel with a 39 k resistor (see Figure 67). A series termination resistor off the clock driver output may improve the dynamic response of the driver. CLK+ CLK 50 CMOS DRIVER OPTIONAL 100 CLK_P 0.1F CLK_N 08801-036 39k Figure 67. Single-Ended 1.8 V CMOS Sample Clock AD9510/AD9511/ AD9512/AD9513/ AD9514/AD9515/ AD9516/AD9518 CLK_P 100 PECL DRIVER 0.1F CLK- 50* 0.1F CLK 0.1F 50* CLK_N 240 A 2.5 V CMOS driver may also be used. In this case, the minimum CLK33V supply voltage should be 2.5 V. The 39 k resistor should be removed in this case. Connecting CLKN to ground with just a 0.1 F capacitor results in the CLKN voltage being biased to about 1.2 V. Clock Duty Cycle Considerations The duty cycle of the input clock should be maintained between 45% and 55%. Duty cycles outside of this range affects the dynamic performance of the ADC. This is especially true at sample rates greater than 75 MHz. It is recommended that the duty cycle stabilizer (DCS) be used at clock rates above 75 MHz to ensure the sampling clock maintains the proper duty cycle inside the device. Below 75 MHz, the DCS should be bypassed. The DCS is bypassed by setting Register 0x66, Bit 2 high. ADC AD9963 The AD9961/AD9963 contain a recirculating DLL, as shown in Figure 70. This circuit allows the incoming CLK signal (REFCLK) to be multiplied by a programmable M/N factor. This provides a means of generating a wide range of DLL output clock (DLLCLK) frequencies. The DLLCLK signal can be used for either the receive ADC sampling clock, the transmit DAC sampling clock, or both. The EXTDLLCLK signal can be programmed to appear on the TXCLK pin or TRXCLK if desired. 240 *50 RESISTORS ARE OPTIONAL. SCHOTTKY DIODES: HSM2812 Figure 69. Transformer Coupled Clock CLK_N CLK AD9963 0.1F 0.1F 08801-037 0.1F CLK+ CLK_P ADC 50 CLOCK MULTIPLICATION USING THE DLL AD9963 0.1F 0.1F Stability of the DLL output requires the main clock input to have a duty cycle of 50% or less. In systems where the duty cycle is greater than 50%, care should be taken to swap the CLKP and CLKN pins to reverse this effect. ADC CLK XFMR DLL Duty Cycle Caution AD9510/AD9511/ AD9512/AD9513/ AD9514/AD9515/ AD9516/AD9518 0.1F 0.1F CLK+ Note that the 39 k resistor shown in the CMOS clock driver example shifts the CLK_N input to about 0.9 V. This is optimal when the CMOS driver is supplied from a 1.8 V supply. 0.1F CLK Mini-Circuits(R) ADT1-1WT, 1:1Z 08801-138 For optimum performance, the AD9961/AD9963 clock inputs (CLKP and CLKN) should be clocked with a low jitter, fast rise time differential signal. This signal should be ac-coupled to the CLKP and CLKN pins via a transformer or capacitors. The CLKP/CLKN inputs are internally biased and require no external bias circuitry. Figure 66 through Figure 69 show preferred methods for clocking the AD9961/AD9963. Figure 68. Differential PECL Sample Clock Rev. A | Page 46 of 60 Data Sheet AD9961/AD9963 DLLBIASPD REG 0x61[5] PHASE DETECTOR RZ 22.5 DLLFILT PIN 54 CHARGE PUMP DLLFILT CP CZ 820pF M[4:0] REG 0x72[4:0] 68nF DLLDIV[1:0] REG 0x72[6:5] 08801-039 DLLLOCKED REG 0x72[7] /M /DLLDIV EXTDLLCLK Figure 71. Recommended DLL Loop Filter DLL_REF_EN REG 0x71[4] 1 /N REFCLK 0 MCLK DLL Start-Up Routine DLLCLK DELAY LINE SELECT LOGIC M[3:0] REG 0x71[3:0] 08801-148 DLL_RESB REG 0x75[3] DLL_EN REG 0x60[7] Figure 70. Functional Block Diagram of Clock Multiplier DLL The DLL is composed of a ring oscillator made from a programmable delay line. The ring oscillator output signal is labeled as MCLK. The MCLK signal is set to oscillate at a frequency M times greater than the REFCLK signal. The DLL output clock, DLLCLK, is the MCLK signal divided by a programmable factor, N. M can be set to values from 1 to 32 and N can be set to values from 1 to 6 and 8. DLL Frequency Locking Range The DLL frequency lock range is determined by the output frequency of the ring oscillator, MCLK. The DLL locks over an MCLK frequency range of 100 MHz to 310 MHz. Verifying that the DLL is locked can be done by polling the DLL_Locked bit (Register 0x72, Bit 7). DLL Filter Considerations The DLL requires an external loop filter between the DLLFILT pin (Pin 54) and ground for stable operation. The circuit diagram in Figure 71 shows the recommended DLL filter configuration. The external components should be placed as close as possible to the device pins. It is important that no noise be coupled into the filter circuit or DLL output clock jitter performance is degraded. To enable the DLL, three bits should be set. The DLL_EN bit (Register 0x60, Bit 7) and the DLL_REF_EN bit (Register 0x71, Bit 4) should be set to 1 and the DLLBIAS_PD bit (Register 0x61, Bit 5) should be set to 0. The CLK input signal should be stable. The DLL_RESB bit should be asserted low for a minimum of 25 s, and then brought inactive (high) to start the frequency acquisition. The DLL takes several REFCLK cycles to acquire lock. The DLL_Locked bit can be queried to verify the DLL is locked. CONFIGURING THE CLOCK DOUBLERS The receive and transmit data paths each have a clock doubler used for clocking data through the device. These clock doublers are only used in single data rate clocking mode, when there is no interpolation or decimation being used. These doublers should be configured according to the following guidelines. Register 0x3A, Register 0x3B, and Register 0x3C configure the operating points of the doublers and should be initialized with the following values: 0x3A = 0x55, 0x3B = 0x55, 0x3C = 0x00 The clock doubler mode and pulse widths should be configured based on the DAC and ADC sample rates. These should be configured according to Table 22. Table 22. Clock Doubler Configuration Guidelines DACCLK/ADCCLK Freq (MHz) 0 to 15 15 to 30 30 to 45 45 to 55 55 to 65 65 to 70 70 to 70 1 2 TXDBLSEL Register 0x39, Bit 0 0 1 1 1 1 1 1 TX_DBLPW[2:0] Register 0x3E, Bits[5:3] 111 X2 X2 X2 X2 X2 X2 RXDBLSEL Register 0x39, Bit 1 0 0 0 0 0 0 1 The DCS_BP bit should be set based on the AUXADCCLK frequency. X = don't care. Rev. A | Page 47 of 60 RX_DBLPW[2:0] Register 0x3E, Bits[2:0] 111 111 110 101 100 011 X2 DCS_BP 1 Register 0x66, Bit 2 1 1 1 1 1 1 0 AD9961/AD9963 Data Sheet options produce the four timing diagrams shown in Figure 73. In full-duplex mode, the TRx and Tx port operate independently. The TRx port outputs samples from the receive path and the Tx port accepts incoming samples for the transmit port. In half-duplex mode, the TRx port outputs samples from the receive path and accepts incoming samples for the transmit path. The Tx port is disabled. The operation of the digital interface is detailed in the sections that follow. TRX PORT OPERATION (FULL-DUPLEX MODE) In full-duplex mode, the TRX port sources the data from the AD9961/AD9963 I and Q receive channels. The interface consists of an output data bus (TRXD[11:0]) that carries the interleaved I and Q data. The data is accompanied by a qualifying output clock (TRXCLK) and an output signal (TRXIQ) that identifies the data as from either the I or Q channel. The maximum guaranteed data rate is 200 MSPS. TRXIQ I0 Q0 I1 Q1 I2 Q2 TRXD[11 :0] Q0 I1 Q1 I2 Q2 RX_IFIRST = 1 I3 RXIQ_HILO = 0 TRXD[11 :0] Q0 I0 Q1 I1 Q2 RX_IFIRST = 0 I2 RXIQ_HILO = 1 TRXD[11 :0] I0 Q1 I1 Q2 I2 Q3 RX_IFIRST = 0 RXIQ_HILO = 0 Figure 73. Receive Path Data Pairing Options The output clock on TRXCLK can also be configured as a double data rate (DDR) clock. In this mode the output clock is divided by 2 and samples are placed on the TRXD[11:0] bus on both the rising and falling edges of the TRXCLK. Figure 74 shows the timing. tOD2 TRXCLK The basic timing diagram for the Rx path is shown in Figure 72. By default, the time-aligned TRXD[11:0] and TRXIQ output signals are driven on the rising edge of the TRXCLK signal. The tOD parameters are specified in Table 23. TRXIQ TRXD[11:0] tOD1 RX_IFIRST = 1 RXIQ_HILO = 1 TRXD[11 :0] 08801-045 The AD9961/AD9963 have two parallel interface ports, the Tx port and the TRx port. The operation of the ports depends on whether the device is configured for full-duplex or halfduplex mode. I0 Q0 I1 Q1 TRXCLK 08801-156 DIGITAL INTERFACES Figure 74. Receive Path Timing Diagram (DDR Clock Mode) Table 23. Maximum Output Delay Between TRXCLK/ TRXD[11:0] and TRXIQ Signals from -40C to +85C TRXD[11:0] I0 Q0 I1 Q1 08801-154 TRXIQ Figure 72. Receive Path Timing Diagram (Bus Rate Clock Mode) An additional configuration bit, RXCLKPH, is available to invert the TRXCLK. In this case, the TRX data and the TRXIQ signals are driven out on the falling edge of TRXCLK and tOD is measured with respect to the falling edge of TRXCLK. The analog signals are sampled simultaneously, creating a quadrature pair of data. This creates two possible data pairing orders on the output bus, I data followed by Q data, or Q data followed by I data. There are also two possible ways to align the bus data with the TRXIQ signal, I data aligned with TRXIQ being high or I data aligned with TRXIQ being low. The IQ pairing and data to TRXIQ alignment relationships create four possible timing modes. The AD9961/AD9963 enable any of these four modes to be sourced from the device. The data pairing order is controlled by the RX_IFIRST bit. The phase relationship between the Rx data and the RXIQ signal is controlled by the RXIQ_HILO bit. The two programming Parameter Drive Strength tOD1 tOD2 Min Max Register 0x63 = 0x00 0.55 0.93 0.42 0.67 Min Max Register 0x63 = 0xAA 0.36 0.57 0.20 0.35 Units ns ns SINGLE ADC MODE The receive port can be operated with only one of the ADCs operational. In this mode the TRXCLK signal can operate in either bus rate clock mode or double data rate clock mode. The TRXIQ pin indicates which ADC is active. Figure 75 to Figure 78 show the timing options available. Rev. A | Page 48 of 60 Data Sheet AD9961/AD9963 tOD2 TX PORT OPERATION (FULL-DUPLEX MODE) The Tx port operates with a qualifying clock that can be configured as either an input or an output. The input data (TXD[11:0]) must be accompanied by the TXIQ signal which identifies to which transmit channel (I or Q) the data is intended. By default, the data and TXIQ signals are latched by the device on the rising edge of TXCLK. The timing diagram is shown in Figure 79 TRXCLK TRXD[11:0] I0 I1 08801-157 TRXIQ TXCLK Figure 75. Rx Timing, I ADC Only, Bus Rate Clock Mode tOD2 TXIQ TRXCLK tSU tHD TRXIQ TRXD[11:0] Q1 Q0 08801-158 Figure 76. Rx Timing, Q ADC Only, Bus Rate Clock Mode tOD2 TRXCLK TRXD[11:0] I1 I0 08801-159 TRXIQ 08801-051 TXD[11:0] Figure 79. Tx Port Timing Diagram (Data Rate Clock Mode) The setup and hold time requirements for the Tx port in data rate clock mode are given in Table 24. The input samples to the device are assembled to create a quadrature pair of data. The data can be arranged in two possible data pairing orders and with two possible data to TXIQ signal phase relationships. This creates four possible timing modes. The AD9961/AD9963 can be configured to accept data in any of these four modes. The data pairing order is controlled by the TX_IFIRST bit. The data to TXIQ phase relationship is controlled by the TXIQ_HILO bit. The two programming options produce the four timing diagrams shown in Figure 80. TXIQ Figure 77. Rx Timing, I ADC Only, DDR Clock Mode TX_IFIRST = 1 TXIQ_HILO = 1 TXD[11 :0] I0 Q0 I1 Q1 I2 Q2 TXD[11 :0] Q0 I1 Q1 I2 Q2 TX_IFIRST = 1 I3 TXIQ_HILO = 0 TXD[11 :0] Q0 I0 Q1 I1 Q2 TX_IFIRST = 0 I2 TXIQ_HILO = 1 TXD[11 :0] I0 Q1 I1 Q2 I2 Q3 TRXCLK TRXD[11:0] Q0 Q1 08801-160 TRXIQ Figure 78. Rx Timing, Q ADC Only, DDR Clock Mode In addition to the different timing modes listed in Figure 75 to Figure 78, the input data can also be delivered from the device in either unsigned binary or twos complement format. The format type is chosen via the RX_BNRY configuration bit. TX_IFIRST = 0 TXIQ_HILO = 0 08801-052 tOD2 Figure 80. Transmit Path Data Pairing Options In addition to the different timing modes listed above, the input data can also be accepted by the device in either unsigned binary or twos complement format. The format type is chosen via the TX_BNRY configuration bit. Rev. A | Page 49 of 60 AD9961/AD9963 Data Sheet The interleaved digital data for the I and Q DACs is accepted by the Tx bus (TXD([11:0]). The data must be presented to the device such that it is stable throughout the setup and hold times, tS and tH, around both the rising and falling edges of the TXCLK signal. A detailed timing diagram is shown in Figure 81. TXCLK TXIFIRST = 1 TXIQPH = 1 TXD[11 :0] I0 Q0 I1 Q1 I2 Q2 TXD[11 :0] Q0 I1 Q1 I2 Q2 TXIFIRST = 1 I3 TXIQPH = 0 TXD[11 :0] Q0 I0 Q1 I1 Q2 TXIFIRST = 0 I2 TXIQPH = 1 TXD[11 :0] I0 Q1 I1 Q2 I2 Q3 TXIFIRST = 0 TXIQPH = 0 08801-054 The Tx port has an optional double data rate (DDR) clock mode. In DDR mode, the transmit data is latched on both the rising and falling edges of TXCLK. The polarity of the edge identifies to which channel the input data is intended. In this mode, the TXIQ signal is not required. Figure 82. Transmit Path Timing Modes (DDR Mode) TXCLK HALF-DUPLEX MODE tSU tHD tSU tHD 08801-053 TXD[11:0] Figure 81. Tx Port Timing Diagram (DDR Clock Mode) In DDR mode, the TXCLK signal is always an input and must be supplied along with the data. The setup and hold time requirements for the Tx port in DDR mode are given Table 24 Table 24. Tx Port Setup and Hold Times From -40C to +85C1 Tx Port Operating Mode TXCLK_MD = 01 TXCLK_MD = 10, TXDBLSEL = 1 TXCLK_MD = 10, TXDBLSEL = 0 1 DRVDD = 1.8 V tSU tHD (Min) (Min) -0.02 +2.60 -1.04 +4.24 DRVDD = 3.3 V tSU tHD (Min) (Min) +0.29 +1.99 -0.28 +3.92 Unit ns ns -0.61 -0.14 ns +4.76 +4.82 Specifications are preliminary and subject to change. The input samples to the device are assembled to create a quadrature pair of data. The two possible data pairing orders and two possible data to TXIQ signal phase relationships create four possible timing modes. The AD9961/AD9963 can be configured to accept data in any of these four modes. The data pairing order is controlled by the TX_IFIRST bit. The data to TXIQ phase relationship is controlled by the TXIQ_HILO bit. The two programming options produce the four timing diagrams shown in Figure 82. The AD9961/AD9963 offer a half-duplex mode enabling a reduced width digital interface. In half-duplex mode, the transmit and receive ports are multiplexed onto the TRXD, TRXIQ, and TRXCLK lines. The direction of the bus can be controlled by either the TXIQ/TXnRX pin (for the rest of this section referred to as simply the TXnRX pin) or the serial port configuration registers. The operation of the transmit and receive ports in half-duplex mode is very similar to the way they operate in full-duplex mode. In half-duplex mode, the interface can be configured to operate with a single clock pin, or with two clock pins. When in Rx mode (sourcing data) the TRX port operates the same in half-duplex mode as it does in full duplex. When in Tx mode, the TXIQ and TXD[11:0] signals are mapped onto the TRXIQ and TRXD[11:0] pins respectively. The TXCLK pin is mapped to the TRXCLK pin in one-clock mode and remains on the TXCLK pin in two-clock mode. Therefore, in one-clock mode, the TRXCLK pin carries the RXCLK signal when set in the Rx direction and the TXCLK signal when set in the Tx direction. In two-clock mode, the TRX pin carries the RXCLK signal and the TXCLK pin carries the TXCLK signal regardless of the bus direction. By default, the clocks sourced by the device are only present when the corresponding direction of the bus is active. Setup and hold times for the TRx port are shown in Table 25. Table 25. TRx Port Setup and Hold Times From -40C to +85C TRx Port Operating Mode TXCLK_MD = 01 TXCLK_MD = 10, TXDBLSEL = 1 TXCLK_MD = 10, TXDBLSEL = 0 Rev. A | Page 50 of 60 DRVDD = 1.8 V tSU tHD (Min) (Min) +0.73 +1.61 -1.66 +5.84 DRVDD = 3.3 V tSU tHD (Min) (Min) +0.44 +1.90 -0.96 +4.55 Units ns ns -1.40 -1.15 ns +6.62 +5.11 Data Sheet AD9961/AD9963 Table 26 shows the operating modes vs. serial port configuration bits. tTXRDY TXnRX Table 26. TRx Bus Operation via Serial Port RXEN 0 1 0 1 TRXD Bus Direction High-Z Rx Tx Rx Tx Bus Function High-Z High-Z High-Z High-Z TRXIQ HIGH-Z TRXD[11:0] HIGH-Z Table 27 shows the operating modes of the TRXD bus as a function of the TXnRX signal. The Tx bus is high impedance in half-duplex mode. Table 27. Rx Bus Operation via TXnRX Pin TXnRX State 0 1 TRXD Bus Direction Rx Tx Tx Bus Function High-Z High-Z The timing of the bus turnaround is shown in the Figure 83 and Figure 84. TXnRX tTXRDY TRXD[11:0] HIGH-Z 08801-055 HIGH-Z TRXIQ Figure 83. Half-Duplex Bus Turnaround, Rx to Tx Rev. A | Page 51 of 60 08801-056 TXEN 0 0 1 1 Figure 84. Half-Duplex Bus Turnaround, Tx to Rx AD9961/AD9963 Data Sheet AUXILIARY CONVERTERS The AD9961/AD9963 have two fast settling servo DACs, along with an analog input and two analog I/O pins. All of the auxiliary converters run off a dedicated supply pin. The input and output compliance ranges depend on the voltage supplied. AUXADCREF pin. The input voltage range for external voltage references is from 1.0 V to 2.5 V. The input impedance of the AUXADCREF pin is 100 k. The full-scale input voltage of the ADC is a function of the voltage reference as: VAUXFS = 3. 2 x VAUXREF 2. 5 AUXILIARY ADC Analog Inputs The auxiliary ADC is a 12-bit SAR converter that is accessed and controlled through the serial port registers (Register 0x77 through Register 0x7B). The ADC voltage reference and clock signals are generated on chip. The auxiliary ADC is preceded by a seven-input multiplexer. The ADC inputs can be connected to either the AUXIN1, AUXIO2, AUXIO3 input pins, or one of four internal signals as shown in Figure 85. The ADC can be configured to sample one of eight analog inputs. The input is selected through the channels select bits (Register 0x77, Bits[2:0]). These eight signals are described in Table 28. REG 0x77[2:0] REG 0x7A[2:0] CLK /R AUXADCCLK SEL 110 100 101 011 111 AUX DAC 000 AUXREF 2.5V VRxCML VCMLI VCMLQ VPTAT VINT AUXIN1 001 AUXIO2 010 AUXIO3 Table 28. Auxiliary ADC Channel Selections Channel Select 000 001 Signal AUXIN1 AUXIO2 010 AUXIO3 011 VPTAT AUX DAC10A T (C o ) = 08801-057 AUX DAC10B Figure 85. Block Diagram of Auxiliary ADC Circuitry 100 101 VCMLI VCMLQ 110 RXCML 111 GND CONVERSION CLOCK The auxiliary ADC conversion clock is generated through a programmable binary division of the CLK input signal. The frequency of the ADC conversion clock is programmable and can be calculated from the following equation: f AUXCLK = Description Pin 72. Pin 71. The auxiliary DAC10A should be disabled when using this pin as an input. Pin 70. The auxiliary DAC10B should be disabled when using this pin as an input. Voltage proportional to absolute temperature scaled to 0.2 K per LSB. Therefore, the temperature in degrees C is: f CLK R where R is programmed through Register 0x7A, Bits[2:0]. For best performance and lowest power consumption, the conversion clock speed should be set to the lowest speed that meets the system conversion time requirements. The maximum allowable auxiliary ADC clock speed is 10 MHz. Voltage Reference The auxiliary ADC has an internal, temperature stable, 2.5 V reference. This results in an input voltage range of 0 V to 3.2 V. When using the internal voltage reference, the AUXADCREF pin should be decoupled to AGND through a 0.22 F capacitor. The AUXADCREF pin can be used as a reference output to external devices, but the current load on the pin should be limited to sourcing less than 5 mA and sinking less than 100 A. ADC _ CODE - 273.2 5 Common mode level of the I and Q Rx ADC buffers. Should measure approximately 0.9 V. The buffer must be enabled (see Configuration Register 0x7E). The RXCML output voltage on Pin 10. This should measure approximately 1.4 V. Should measure 0 V. When selected, Input Pin 70, Pin 71, and Pin 72 are connected to the sampling cap of the auxiliary ADC. Therefore, the circuits driving these inputs need to recover to the desired accuracy from having a discharged 10 pF capacitor connected to it at the initiation of the conversion, within the sampling window. A programmable delay (Register 0x7B, Bits[1:0]) can be added to the conversion cycle time to allow additional settling time of the input. If the ADC input is driven from a low source impedance, like the output of an op amp, a 20-cycle conversion time should yield good results. Higher impedance sources may require the 34-cycle conversion time to fully settle. Where the conversion cycle time is not an issue, it is recommended that the full 34-cycle conversion time be used. Conversions where the input multiplexer is switched between inputs require a longer conversion cycle time than consecutive conversions from the same multiplexer input. For systems with tight accuracy requirements, a higher accuracy external reference can be used to source a voltage into the Rev. A | Page 52 of 60 Data Sheet AD9961/AD9963 Digital Output Coding It should be noted that after initial power-up or recovery from power-down, the ADC needs about 100 S to stabilize. In many cases, the results of the first conversion should be discarded in order for the auxiliary ADC to reach an optimum operating condition. The digital output coding is straight binary. The ideal transfer characteristic for the auxiliary ADC is shown in Figure 86. 111 ... 111 111 ... 110 111 ... 101 AUXILIARY DACs ADC CODE The AD9963 has two 10-bit auxiliary DACs and two 12-bit auxiliary DACs suitable for calibration and control functions. The DACs have voltage outputs with selectable full-scale voltages and output ranges. The auxiliary DACs are configured and updated through the serial port interface. 10-Bit Auxiliary DACs 000 ... 010 000 ... 001 The two 10-bit DACs have identical transfer functions and are output on the AUXIO2 and AUXIO3 pins. The two DACs can be independently enabled and configured. The DACs have five selectable top-of-scale voltages and four selectable output ranges, which result in 20 possible transfer functions. 000 ... 000 +VFS - 1 LSB 1 LSB +VFS - 1.5 LSB 08801-058 +0.5 LSB ANALOG INPUT Figure 86. Auxiliary ADC Transfer Function AVDD Auxiliary ADC Conversion Cycle Figure 87 shows a typical timing scenario for an auxiliary ADC conversion period. The scenario shows the write that initiates the conversion, followed by the read that retrieves the conversion result. In some cases, it may be required to add a wait time between the write and read to ensure that the conversion is complete. The wait time depends on the ADC conversion cycle time and the speed of the serial port clock. The minimum wait time is calculated as: DAC10_RNG DACCODE[9:0] SERIAL WRITE PORT INSTR. DATA REG 0x77 WAIT READ INSTR. ADC CONVERSION AUX ADC CYCLE 2 DATA REG 0x78 DATA REG 0x79 WRITE INSTR. DATA REG 0x77 WAIT ADC CONVERSION 08801-059 AUX ADC CYCLE 1 16k DAC10_TOP: 000 = 1.0V = 16k 001 = 1.5V = 8.0k 010 = 2.0V = 5.3k 011 = 2.5V = 4.0k 100 = 3.0V = 3.2k AUXIO - RTOP + 0.5V Figure 88. Simplified Circuit Diagram of the 10-Bit Auxiliary DAC The circuit is most easily analyzed using superposition of two inputs to the op amp, the 0.5 V reference voltage, and the programmable current source. The following equation describes the no-load output voltage: t wait (N 1) t AUXADCCLK 7 t SCLK where N is the number of auxiliary ADC clock cycles that result from the conversion time setting in Register 0x7B. tSCLK is the serial port clock period. A negative wait time indicates no wait time is required. ISPAN DAC10_RNG: 00 = 2.0V = 124A Ifs 01 = 1.5V = 93A Ifs 10 = 1.0V = 62A Ifs 11 = 0.5V = 31A Ifs 08801-060 A conversion is initiated by writing to SPI Register 0x77. The conversion starts on the first rising edge of the AUXADCCLK following a write to Register 0x77 (serial port register writes are completed on the eighth rising edge of SCLK during the data word write cycle). The conversion takes from 20 to 34 AUXADCCLK cycles to complete depending on the conversion time setting programmed in Register 0x77. In most cases, the ADC throughput is a function of both the serial port clock rate and the ADC conversion time. 0.5V DACCODE VOUT 0.5 16 k ISPAN RTOP 1024 The DACCODE (see Register 0x49 and Register 0x4A for DAC10A and Register 0x46 and Register 0x47 for DAC10B) is interpreted such that ISPAN is full scale at 0x000 and zero at 0x3FF. This leads to an increasing output voltage with increasing code as shown in Figure 89 and Figure 90. The five selectable gain setting resistors of 3.2 k, 4.0 k, 5.3 k, 8.0 k, and 16 k result in full-scale output voltage levels of 3.0 V, 2.5 V, 2.0 V, 1.5 V and 1.0 V respectively. The four selectable full-scale currents of 31 A, 62 A, 93 A and 124 A result in voltage output spans of 0.5 V, 1.0 V, 1.5 V, and 2.0 V, respectively. Figure 87. Timing Scenario for Auxiliary ADC Conversion Cycle Rev. A | Page 53 of 60 AD9961/AD9963 Data Sheet The curves in Figure 89 represent four of the possible DAC transfer functions with the full-scale voltage of 3.0 V and spans of 0.5 V, 1.0 V, 1.5 V, and 2.0 V. The curves in Figure 90 represent four of the possible DAC transfer functions with the full-scale voltage of 1.5 V and spans of 0.5 V, 1.0 V, 1.5 V, and 2.0 V. Note that the 2.0 V span results in clamping at the lower end of the scale at 0 V where the equation resultsin negative output voltages. AUX33V AUXDACREF 2.3 1 R REFIO 0 VREF DACCODE 0 TO VREF DAC12 R DAC12TOP: 0 = RTOP = 0.8R 1 = RTOP = 2.3R Figure 91. Simplified Schematic of the 12-Bit Auxiliary DAC 2.5 2.0 Note that VREF can be derived from a 1.0 V bandgap reference or be ratiometric with the AUX33V supply. An additional gain stage follows the DAC that sets the final full-scale output voltage . The following equation describes the no load output voltage: RNG00 RNG01 RNG10 RNG11 1.5 0.5 0 128 256 384 512 CODE 640 768 896 1024 08801-061 1.0 Figure 89. AUXDAC10 Voltage Output vs. Digital Code, VTOP = 3.0 V (RTOP = 3.2 k) where VFS is set with the combination of bits shown in Table 29. Table 29. 12-Bit Auxiliary DAC Full-Scale Voltage Selection DAC10x_RNG1 0 1 0 1 AUXDAC_REF 0 0 1 1 2.00 1.75 1.50 OUTPUT VOLTAGE (V) DACCODE VOUT = VFS x 1024 1.25 1 1.00 0.75 x = A or B. The curves in Figure 92 show the two transfer functions when using the internal 1.0 V bandgap reference. RNG00 RNG01 RNG10 RNG11 0.50 VFS AUX33V 0.54 x AUX33V 3.3 V 1.8 V 3.5 0.25 0 128 256 384 512 CODE 640 768 896 1024 Figure 90. AUXDAC10 Voltage Output vs. Digital Code, VTOP = 1.5 V (RTOP = 8.0 k) 12-Bit Auxiliary DACs The two 12-bit DACs have similar transfer functions and are output on the DAC12A and DAC12B pins. The two DACs can be independently enabled and configured. Figure 91 shows a simplified schematic of the 12-bit auxiliary DAC. OUTPUT VOLTAGE (V) 0 08801-062 3.0 2.5 VFS = 3.3V 2.0 1.5 VFS = 1.8V 1.0 0.5 0 0 128 256 384 512 CODE 640 768 896 Figure 92. AUXDAC12 Voltage Output vs. Digital Code Rev. A | Page 54 of 60 1024 08801-064 OUTPUT VOLTAGE (V) 3.0 RTOP 08801-063 3.5 Data Sheet AD9961/AD9963 POWER SUPPLY CONFIGURATION EXAMPLES POWER SUPPLIES The AD9961/AD9963 power distributions are shown in Figure 93. The functional blocks labeled Rx ANLG, Rx ADCs, SPI and digital core, clocking, and DLL operate from 1.8 V supplies. The functional blocks labeled Tx DACs, AUX DACs and digital I/O operate over a supply voltage range from 1.8 V to 3.3 V. The auxiliary ADC operates from a 3.3 V supply. TXVDD(2) Tx DACs AD9961/AD9963 AUX DACs TXVDD CLK33V DLL18V DVDD18 DRVDD 08801-066 3.3V Figure 93. AD9961/AD9963 Power Distribution Block Diagram Figure 94. 3.3 V Only Supply Configuration Figure 95 shows a power supply configuration where all 1.8 V voltage rails are powered by external supplies. The LDO_EN pin is grounded, and all of the internal LDOs are disabled. The transmit DAC, auxiliary converters and I/O pads run from a 3.3 V supply. The three DRVDD pins are internally connected together, therefore, these pins must be connected to the same voltage. The voltage applied to these pins affects the timing of the device as noted in the Digital Interfaces section. The TXVDD and AUX33V supplies can operate over a range from 1.8 V to 3.3 V. It should be noted that the auxiliary ADC requires AUX33V to be 3.3 V for operation. The performance of the Tx DACs vary with the TXVDD supply as indicated in the Table 1 and Figure 4 to Figure 11. 1.8V AUX33V 3.3V 08801-180 The LDO_EN pin (Pin 14) is a three-state input pin that controls the operation of the LDOs. When LDO_EN is high, all of the LDOs are enabled. When LDO_EN is low, all of the LDOs are disabled. When LDO_EN is floating or approximately DRVDD/2, only the DVDD18V LDO is enabled. All of the LDOs except the DVDD18V LDO can be independently disabled through serial port control as well by writing to Register 0x61. TXVDD When the LDO regulators are used, the RX18V, RX18VF, DLL18V, CLK18V, and DVDD18V pins should be decoupled to ground with a 0.1 F or larger capacitor. The LDO inputs can operate over a range from 2.5 V to 3.3 V. RX33V AD9961/AD9963 LDO_EN The 1.8 V only blocks can be supplied directly with 1.8 V by using the RX18V, RX18VF, DLL18V, CLK18V, and DVDD18V supply pins. In this mode, the on-chip voltage regulators must be disabled. To provide optimal ESD protection for the device, the inputs of the LDO regulators should not be left floating. When unused, the LDO regulator inputs should be tied to one of the LDO outputs (for example, if RX33V is unused, tie RX33V to either RX18V or RX18VF). DRVDD DRVDD(3) DVDD18 DIGITAL I/O LDO CLK18V AD9961/AD9963 CLK18V DVDD18V DLL18V RX18VF SPI AND DIGITAL CORE RX18VF Rx ADCs RX18V LDO LDO_EN LDO 08801-301 LDO RX33V RX18VF CLOCKING CLK33V RX18V LDO Rx ANLG RX33V DLL AUX33V RX18V REG 0x61 = 0x00 DLL18V CLK18V AUX ADCs Figure 94 shows a 3.3 V only power supply configuration. In this case, all of the internal circuits that require 1.8 V supplies are powered from the on-chip regulators. The LDO_EN pin is set high, and all of the internal LDOs are enabled. The transmit DAC, auxiliary converters, and I/O pads run from a 3.3 V supply. CLK33V AUX33V There are numerous ways of configuring the power supplies powering the AD9961/AD9963. Two power supply configuration examples are shown in Figure 94 and Figure 95. Figure 95. 3.3 V and 1.8 V Supply Configuration POWER DISSIPATION The AD9961/AD9963 power dissipation is highly dependent on operating conditions. Table 30 and Figure 96 to Figure 103 show the typical current consumption by power supply domain under different operating conditions. The current draw from the 1.8 V supplies are independent of whether they are supplied by the on-chip regulators or by an external 1.8 V supply. The quiescent current of the LDO regulators are about 100 A. The current drawn from the AUX33V supply by the auxiliary ADC is typically 350 A. The 10-bit auxiliary DACs each typically draw 275 A from the AUX33V supply. The 12-bit auxiliary DACs typically draw 550 A each from the AUX33V supply. Rev. A | Page 55 of 60 AD9961/AD9963 Data Sheet 7.0 80 70 6.5 60 6.0 ICLK18V (mA) IRX (mA) RX18V 50 5.5 5.0 40 4.5 30 4.0 20 20 40 60 80 100 fADC (MHz) 0 08801-181 0 25 50 75 100 125 175 150 fCLK (MHz) 08801-184 RX18VF Figure 99. ICLKVDD18 vs. fCLK Figure 96. IRX18V and IRX18VF vs. fADC, Both ADCs Enabled 12 26 fCLK = 20MHz, N = 1 22 IFS = 4mA 18 IDLL18V (mA) ITXVDD (mA) 10 IFS = 2mA 14 fCLK = 50MHz, N = 1 8 fCLK = 20MHz, N = 5 IFS = 1mA 6 4 6 25 50 75 100 125 150 175 fDAC (MHz) 80 08801-182 0 18 16 16 IDVDD18 (mA) IFS = 4mA IFS = 2mA 320 12 2x 8 1x 10 IFS = 1mA 4 0 6 0 25 50 75 100 125 150 175 fDAC (MHz) 0 25 50 75 100 fRXDATA (MHz) Figure 101. IDVDD18 vs. fRXDATA, 1x, 2x (Rx Only) Figure 98. ITXVDD vs. fDAC, FSC = 1 mA, 2 mA, 4 mA, TXVDD = 1.8 V Rev. A | Page 56 of 60 125 08801-186 8 08801-183 ITXVDD (mA) 260 Figure 100. IDLL18V vs. fDLL, fCLKIN= 19.2 MHz, 30.72 MHz 20 12 200 fDLL (MHz) Figure 97. ITXVDD vs. fDAC, FSC = 1 mA, 2 mA, 4 mA, TXVDD = 3.3 V 14 140 08801-185 10 Data Sheet AD9961/AD9963 Power Calculation Example The following example shows how to estimate the device power consumption under a typical operating condition. 100 80 IDVDD18 (mA) Operating conditions: fCLK = 60 MHz 60 4x 8x fDLL = 120 MHz 40 fDAC = 120 MHz fADC = 60 MHz 2x 1x 20 4x interpolation 2x decimation 0 25 50 75 100 125 150 175 fDAC (MHz) 08801-187 0 Figure 102. IDVDD18 vs. fDAC, 1x, 2x, 4x, 8x (Tx only) TXVDD = CLK33V = AUX33V = 3.3 V Auxiliary ADC enabled 35 All other supplies powered from external 1.8 V supplies. 30 Table 30. Example Power Supply Currents 25 20 3.3V 2.5V 15 1.8V 10 5 0 0 10 20 30 40 50 60 70 80 90 fDATA (MHz) Figure 103. IDRVDD vs. fDATA, (Tx Enable and Disabled) 100 08801-188 IDRVDD (mA) DAC full-scale current = 2 mA Supply RX18V RX18VF TXVDD CLKVDD18V DLL18V DVDD18V (Rx) DVDD18V (Tx) DRVDD AUX33V Total (1.8 V) Total (3.3 V) Rev. A | Page 57 of 60 Typical Current (mA) 74 30 16 5.2 7.5 9 35 5 0.5 169 16 Typical Power (mW) 133 54 53 9.5 13.5 16.2 63 9 1.7 298 55 AD9961/AD9963 Data Sheet EXAMPLE START-UP SEQUENCES CONFIGURING THE DLL The AD9963 DLL is shown in Figure 65, the clock distribution diagram. The register writes in Table 31 configures the DLL to drive the DACs with a multiplication in frequency of 10 and a division of 3 from the main CLKP/CLKN input. From the default register settings at reset, this would take a 20 MHz CLKP/CLKN clock, multiply it up to 200 MHz, then divide the clock down by 3 to produce 66.67 MHz. The write to Register 0x71 configures the DAC clock to be sourced from the DLL. By default, the Rx and Tx data buses operate in SDR mode. Each DAC is clocked at 66.67 MHz and the TxCLK pin outputs 133.33 MHz. Table 31. Register (hex) 0x60 0x71 Data (hex) 0x80 0x53 0x72 0x09 Delay 100 pS 0x75 0x08 0xDelay 100 pS 0x75 0x00 0x72 Read Comments % enable DLL % set DAC clock to DLL/enable DLL reference/N = 3 % M = 9, effective multiplication is M + 1 = 10 % hold DLL reset high % hold DLL reset low % check Bit 7 to verify the DLL has locked When operating below 75 MHz, bypass the duty cycle stabilizer in the ADCCLK generator circuit and take care to ensure a duty cycle 45% to 55% of the CLKP/CLKN clock input. The series of writes in Table 32 configures the Rx clock doubler to clock the ADCs from reset. These writes are for an ADC clock of < 75 MHz. This same sequence could be used for setting up a clock >75 MHz by removing the write to Register 0x66. Table 32. Register (hex) 0x3C 0x39 0x66 Data (hex) 0x00 0x02 0x04 0x3B 0x55 Delay 100 pS 0x39 0x82 Delay 100 pS 0x39 0x02 0x63 0x08 Comments % the recommended tap delay is 0 % configure RxCLK as DDLL % bypass duty cycle correction (for CLKP/CLKN < 75 MHz) % the recommended offset is 1 (changing Bit 3 from default) % reset Rx DDLL % pull Rx DDLL out of reset % set drive strength to 3 for the RxClk SENSING TEMPERATURE WITH THE AUXADC This sequence of register writes and reads configures the AUXADC to sense temperature. CONFIGURING THE CLOCK DOUBLERS (DDLL) The AD9963 includes two clock doublers. The Rx clock doubler, if enabled, doubles the frequency of the CLKP/CLKN signal on its way into the circuit that generates ADCCLK (Figure 65). The Tx clock doubler doubles the DACCLK signal and can be selected to be included in the TxCLK generator circuit (Figure 52). Use of both clock doublers is recommended when the ADCs and DACs are operated above 15 MHz. Register (hex) 0x77 0x7A 0x7B 0x77 Read Read Rev. A | Page 58 of 60 Data (hex) 0x03 0x80 0x80 0x83 0x78 0x79 Comments Channel temperature sensor Aux ADC enable Temperature sensor enable Choose channel to sample with AUX ADC MSB 7:0 = AUXADC[11:4] LSB bit 7:4 = AUXADC[3:0] Data Sheet AD9961/AD9963 OUTLINE DIMENSIONS 10.10 10.00 SQ 9.90 0.60 0.42 0.24 0.60 0.42 0.24 0.30 0.23 0.18 72 55 54 PIN 1 INDICATOR 1 PIN 1 INDICATOR 9.85 9.75 SQ 9.65 0.50 BSC 0.50 0.40 0.30 0.80 MAX 0.65 TYP 0.25 MIN 8.50 REF 0.05 MAX 0.02 NOM COPLANARITY 0.08 0.20 REF SEATING PLANE 19 BOTTOM VIEW 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-VNND-4 06-25-2012-A 1.00 0.85 0.80 18 37 36 TOP VIEW 12 MAX 7.25 7.10 SQ 6.95 EXPOSED PAD Figure 104. 72-Lead Lead Frame Chip Scale Package [LFCSP_VQ] 10 mm x 10 mm Body, Very Thin Quad (CP-72-4) Dimensions shown in millimeters ORDERING GUIDE Model1 Temperature Range Package Description Package Option AD9961BCPZ AD9961BCPZRL AD9963BCPZ AD9963BCPZRL AD9961-EBZ AD9963-EBZ AD-DPGIOZ -40C to +85C -40C to +85C -40C to +85C -40C to +85C -40C to +85C -40C to +85C -40C to +85C 72-Lead Lead Frame Chip Scale Package [LFCSP_VQ] 72-Lead Lead Frame Chip Scale Package [LFCSP_VQ] 72-Lead Lead Frame Chip Scale Package [LFCSP_VQ] 72-Lead Lead Frame Chip Scale Package [LFCSP_VQ] Evaluation Board Evaluation Board Pattern Generation and Capture Card CP-72-4 CP-72-4 CP-72-4 CP-72-4 1 Z = RoHS Compliant Part. Rev. A | Page 59 of 60 AD9961/AD9963 Data Sheet NOTES (c)2010-2012 Analog Devices, Inc. All rights reserved. Trademarks and registered trademarks are the property of their respective owners. D08801-0-8/12(A) Rev. A | Page 60 of 60