ADC081500 High Performance, Low Power, 8-Bit, 1.5 GSPS A/D Converter General Description Features The ADC081500 is a low power, high performance CMOS analog-to-digital converter that digitizes signals to 8 bits resolution at sample rates up to 1.7 GSPS. Consuming a typical 1.2 W at 1.5 GSPS from a single 1.9 Volt supply, this device is guaranteed to have no missing codes over the full operating temperature range. The unique folding and interpolating architecture, the fully differential comparator design, the innovative design of the internal sample-and-hold amplifier and the self-calibration scheme enable a very flat response of all dynamic parameters beyond Nyquist, producing a high 7.3 ENOB with a 748 MHz input signal and a 1.5 GHz sample rate while providing a 10-18 B.E.R. Output formatting is offset binary and the LVDS digital outputs are compatible with IEEE 1596.3-1996, with the exception of an adjustable output offset voltage between 0.8V and 1.2V. The converter output has a 1:2 demultiplexer that feeds two LVDS buses and reduces the output data rate on each bus to one-half the sample rate. The converter typically consumes less than 3.5 mW in the Power Down Mode and is available in a 128-lead, thermally enhanced exposed pad LQFP and operates over the Industrial (-40°C ≤ TA ≤ +85°C) temperature range. ■ ■ ■ ■ ■ ■ ■ ■ Internal Sample-and-Hold Single +1.9V ±0.1V Operation Choice of SDR or DDR output clocking Multiple ADC Synchronization Capability Guaranteed No Missing Codes Serial Interface for Extended Control Fine Adjustment of Input Full-Scale Range and Offset Duty Cycle Corrected Sample Clock Key Specifications ■ ■ ■ ■ ■ ■ ■ Resolution Max Conversion Rate Bit Error Rate ENOB @ 748 MHz Input DNL Power Consumption — Operating — Power Down Mode 8 Bits 1.5 GSPS (min) 10-18 (typ) 7.3 Bits (typ) ±0.15 LSB (typ) 1.2 W (typ) 3.5 mW (typ) Applications ■ ■ ■ ■ ■ Direct RF Down Conversion Digital Oscilloscopes Satellite Set-top boxes Communications Systems Test Instrumentation Block Diagram 20153153 © 2009 National Semiconductor Corporation 201531 www.national.com ADC081500 High Performance, Low Power, 8-Bit, 1.5 GSPS A/D Converter April 21, 2009 ADC081500 Ordering Information Industrial Temperature Range (-40°C < TA < +85°C) NS Package ADC081500CIYB 128-Pin Exposed Pad LQFP ADC081500DEV Development Board Pin Configuration 20153101 * Exposed pad on back of package must be soldered to ground plane to ensure rated performance. www.national.com 2 ADC081500 Pin Descriptions and Equivalent Circuits Pin Functions Pin No. Symbol Equivalent Circuit Description OutV / SCLK Output Voltage Amplitude and Serial Interface Clock. Tie this pin high for normal differential DCLK and data amplitude. Ground this pin for a reduced differential output amplitude and reduced power consumption. See Section 1.1.6. When the extended control mode is enabled, this pin functions as the SCLK input which clocks in the serial data. See Section 1.3 4 OutEdge / DDR / SDATA DCLK Edge Select, Double Data Rate Enable and Serial Data Input. This input sets the output edge of DCLK+ at which the output data transitions. (See Section 1.1.5.2). When this pin is floating or connected to 1/2 the supply voltage, DDR clocking is enabled. When the extended control mode is enabled, this pin functions as the (SDATA) input. See Section 1.2 for details on the extended control mode. 15 DCLK_RST DCLK Reset. A positive pulse on this pin is used to reset and synchronize the DCLK outputs of multiple converters. See Section 1.5 for detailed description. 26 PD Power Down Pin. A logic high on the PD pin puts the device into the Power Down Mode. CAL Calibration Cycle Initiate. A minimum 80 input clock cycles logic low followed by a minimum of 80 input clock cycles high on this pin initiates the self calibration sequence. See Section 2.4.2. FSR/ECE Full Scale Range Select and Extended Control Enable. In non-extended control mode, a logic low on this pin sets the full-scale differential input range to a reduced VIN input level . A logic high on this pin sets the full-scale differential input range to a higher VIN input level. See Converter Electrical Characteristics. To enable the extended control mode, whereby the serial interface and control registers are employed, allow this pin to float or connect it to a voltage equal to VA/2. See Section 1.2 for information on the extended control mode. CalDly / SCS Calibration Delay and Serial Interface Chip Select. With a logic high or low on pin 14, this pin functions as Calibration Delay and sets the number of input clock cycles after power up before calibration begins (See Section 1.1.1). With pin 14 floating, this pin acts as the enable pin for the serial interface input and the CalDly value becomes "0" (short delay with no provision for a long power-up calibration delay). 3 30 14 127 3 www.national.com ADC081500 Pin Functions Pin No. Symbol 18 19 CLK+ CLK- LVDS Clock input pins for the ADC. The differential clock signal must be a.c. coupled to these pins. The input signal is sampled on the falling edge of CLK+. See Section 2.3. VIN+ VIN− Analog signal inputs to the ADC. The differential full-scale input range of this input is programmable using the FSR pin 14 in normal mode and the Input Full-Scale Voltage Adjust register in the extended control mode. Refer to the VIN specification in the Converter Electrical Characteristics for the full-scale input range in the normal mode. Refer to ?? 1.4 for the full-scale input range in the extended control mode. 7 VCMO Common Mode Voltage. The voltage output at this pin is required to be the common mode input voltage at VIN+ and VIN− when d.c. coupling is used. This pin should be grounded when a.c. coupling is used at the analog inputs. This pin is capable of sourcing or sinking 100 μA. See Section 2.2. 31 VBG Bandgap output voltage capable of 100 μA source/sink. 126 CalRun Calibration Running indication. This pin is at a logic high when calibration is running. 32 REXT External bias resistor connection. Nominal value is 3.3 kOhms (±0.1%) to ground. See Section 1.1.1. 34 35 Tdiode_P Tdiode_N Temperature Diode Positive (Anode) and Negative (Cathode). These pins may be used for die temperature measurements, however no specified accuracy is implied or guaranteed. See Section 2.6.2. 11 10 www.national.com Equivalent Circuit Description 4 ADC081500 Pin Functions Pin No. Symbol 83 84 85 86 89 90 91 92 93 94 95 96 100 101 102 103 D7− D7+ D6− D6+ D5− D5+ D4− D4+ D3− D3+ D2− D2+ D1− D1+ D0− D0+ The LVDS Data Outputs that are not delayed in the output demultiplexer. Compared with the Dd outputs, these outputs represent the later time samples. These outputs should always be terminated with a 100Ω differential resistor. 104 105 106 107 111 112 113 114 115 116 117 118 122 123 124 125 Dd7− Dd7+ Dd6− Dd6+ Dd5− Dd5+ Dd4− Dd4+ Dd3− Dd3+ Dd2− Dd2+ Dd1− Dd1+ Dd0 Dd0 The LVDS Data Outputs that are delayed by one CLK cycle in the output demultiplexer. Compared with the D outputs, these outputs represent the earlier time sample. These outputs should always be terminated with a 100Ω differential resistor. OR+ OR- Out Of Range output. A differential high at these pins indicates that the differential input is out of range (outside the range ±VIN/2 as programmed by the FSR pin in nonextended control mode or the Input Full-Scale Voltage Adjust register setting in the extended control mode). 82 81 DCLK+ DCLK- Differential Clock outputs used to latch the output data. Delayed and non-delayed data outputs are supplied synchronous to this signal. This signal is at 1/2 the input clock rate in SDR mode and at 1/4 the input clock rate in the DDR mode. The DCLK outputs are not active during a calibration cycle. The DCLK outputs are not active during a calibration cycle, therefore this is not recommended as a system clock. 2, 5, 8, 13, 16, 17, 20, 25, 28, 33, 128 VA Analog power supply pins. Bypass these pins to ground. 40, 51 ,62, 73, 88, 99, 110, 121 VDR Output Driver power supply pins. Bypass these pins to DR GND. 1, 6, 9, 12, 21, 24, 27, 41 GND Ground return for VA. 79 80 Equivalent Circuit Description 5 www.national.com ADC081500 Pin Functions Pin No. Symbol 42, 53, 64, 74, 87, 97, 108, 119 DR GND 22, 23, 29, 36-39, 43-50, 52, 54-61, 63, 65-72, 75-78, 98, 109, 120 NC www.national.com Equivalent Circuit Description Ground return for VDR. No Connection. Make no connection to these pins. 6 (Notes 1, 2) −40°C ≤ TA ≤ +85°C (Notes 1, 2) Ambient Temperature Range If Military/Aerospace specified devices are required, please contact the National Semiconductor Sales Office/ Distributors for availability and specifications. Supply Voltage (VA) Driver Supply Voltage (VDR) Analog Input Common Mode Voltage VIN+, VIN- Voltage Range (Maintaining Common Mode) Supply Voltage (VA, VDR) Supply Difference VDR - VA Voltage on Any Input Pin (Except VIN+, VIN-) Voltage on VIN+, VIN(Maintaining Common Mode) Ground Difference |GND - DR GND| Input Current at Any Pin (Note 3) Package Input Current (Note 3) 2.2V +1.8V to +2.0V +1.8V to VA VCMO ±50mV 0V to 2.15V (100% duty cycle) 0V to 2.5V (10% duty cycle) 0V to 100 mV Ground Difference (|GND - DR GND|) CLK Pins Voltage Range Differential CLK Amplitude −0.15V to (VA +0.15V) -0.15V to 2.5V 0V to 100 mV ±25 mA ±50 mA Power Dissipation at TA ≤ 85°C ESD Susceptibility (Note 4) Human Body Model Machine Model Soldering Temperature, Infrared, 10 seconds, (Note 5), (Applies to standard plated package only) Storage Temperature 0V 0V to VA 0.4VP-P to 2.0VP-P Package Thermal Resistance Package 2.0 W 128-Lead Exposed Pad LQFP 2500V 250V θJA 26°C / W θJC (Top of θJ-PAD Package) (Thermal Pad) 10°C / W 2.8°C / W Figure 13Soldering process must comply with National Semiconductor’s Reflow Temperature Profile specifications. Refer to www.national.com/packaging.Figure 13 (Note 5) 235°C −65°C to +150°C Converter Electrical Characteristics The following specifications apply after calibration for VA = VDR = +1.9VDC, OutV = 1.9V, VIN (a.c. coupled) Full Scale Range = differential 870mVP-P, CL = 10 pF, Differential (a.c. coupled) sinewave input clock, fCLK = 1.5 GHz at 0.5VP-P with 50% duty cycle, VBG = Floating, Normal Control Mode, Single Data Rate Mode, REXT = 3300Ω ±0.1%, Analog Signal Source Impedance = 100Ω Differential. Boldface limits apply for TA = TMIN to TMAX. All other limits TA = 25°C, unless otherwise noted. (Notes 6, 7) Symbol Parameter Conditions Typical (Note 8) Limits (Note 8) Units (Limits) STATIC CONVERTER CHARACTERISTICS INL Integral Non-Linearity (Best fit) DC Coupled, 1MHz Sine Wave Over ranged ±0.3 ±0.9 LSB (max) DNL Differential Non-Linearity DC Coupled, 1MHz Sine Wave Over ranged ±0.15 ±0.6 LSB (max) 8 Bits −1.5 1.0 LSB (min) LSB (max) Resolution with No Missing Codes VOFF Offset Error VOFF_ADJ Input Offset Adjustment Range Extended Control Mode ±45 PFSE Positive Full-Scale Error (Note 9) −0.6 ±25 mV (max) NFSE Negative Full-Scale Error (Note 9) −1.31 ±25 mV (max) FS_ADJ Full-Scale Adjustment Range Extended Control Mode ±20 ±15 %FS -0.45 mV DYNAMIC CONVERTER CHARACTERISTICS FPBW Full Power Bandwidth B.E.R. Bit Error Rate Gain Flatness ENOB Effective Number of Bits SINAD Signal-to-Noise Plus Distortion Ratio 1.7 GHz 10-18 Error/Sample d.c. to 500 MHz ±0.5 dBFS d.c. to 1 GHz ±1.0 fIN = 373 MHz, VIN = FSR − 0.5 dB 7.4 fIN = 748 MHz, VIN = FSR − 0.5 dB 7.3 fIN = 373 MHz, VIN = FSR − 0.5 dB 46.3 fIN = 748 MHz, VIN = FSR − 0.5 dB 45.4 7 dBFS 7.0 Bits (min) Bits (min) 43.9 dB (min) dB (min) www.national.com ADC081500 Operating Ratings Absolute Maximum Ratings ADC081500 Symbol SNR THD 2nd Harm Typical (Note 8) Limits (Note 8) Units (Limits) fIN = 373 MHz, VIN = FSR − 0.5 dB 47 44.5 dB (min) fIN = 748 MHz, VIN = FSR − 0.5 dB 46.3 fIN = 373 MHz, VIN = FSR − 0.5 dB -54.5 -47 dB (max) Parameter Signal-to-Noise Ratio Total Harmonic Distortion Second Harmonic Distortion 3rd Harm Third Harmonic Distortion SFDR Spurious-Free dynamic Range IMD Intermodulation Distortion Out of Range Output Code (In addition to OR Output high) Conditions dB (min) fIN = 748 MHz, VIN = FSR − 0.5 dB -53 dB (max) fIN = 373 MHz, VIN = FSR − 0.5 dB −60 dB fIN = 748 MHz, VIN = FSR − 0.5 dB -57 dB fIN = 373 MHz, VIN = FSR − 0.5 dB −62 dB fIN = 748 MHz, VIN = FSR − 0.5 dB -65 dB fIN = 373 MHz, VIN = FSR − 0.5 dB 56 fIN = 748 MHz, VIN = FSR − 0.5 dB 53 dB (min) fIN1 = 321 MHz, VIN = FSR − 7 dB fIN2 = 326 MHz, VIN = FSR − 7 dB -50 dB 48.5 (VIN+) − (VIN−) > + Full Scale 255 (VIN+) − (VIN−) < − Full Scale 0 dB (min) ANALOG INPUT AND REFERENCE CHARACTERISTICS FSR pin 14 Low VIN 650 Full Scale Analog Differential Input Range FSR pin 14 High 870 570 mVP-P (min) 730 mVP-P (max) 790 mVP-P (min) 950 mVP-P (max) VCMO − 50 VCMO + 50 mV (min) mV (max) VCMI Analog Input Common Mode Voltage CIN Analog Input Capacitance (Notes Differential 10, 11) Each input pin to ground 0.02 pF 1.6 pF RIN Differential Input Resistance 100 VCMO 94 Ω (min) 106 Ω (max) 0.95 1.45 V (min) V (max) ANALOG OUTPUT CHARACTERISTICS VCMO Common Mode Output Voltage ICMO = ±100 µA 1.26 TC VCMO Common Mode Output Voltage Temperature Coefficient TA = −40°C to +85°C 118 ppm/°C VCMO_LVL VCMO input threshold to set DC Coupling mode VA = 1.8V 0.60 V VA = 2.0V 0.66 V CLOAD VCMO Maximum VCMO load Capacitance VBG Bandgap Reference Output Voltage IBG = ±100 µA TC VBG Bandgap Reference Voltage Temperature Coefficient TA = −40°C to +85°C, IBG = ±100 µA CLOAD VBG Maximum Bandgap Reference load Capacitance 1.26 80 pF 1.20 1.33 V (min) V (max) 28 ppm/°C 80 pF TEMPERATURE DIODE CHARACTERISTICS ΔVBE Temperature Diode Voltage www.national.com 192 µA vs. 12 µA, TJ = 25°C 71.23 mV 192 µA vs. 12 µA, TJ = 85°C 85.54 mV 8 Typical (Note 8) Limits (Note 8) Units (Limits) Sine Wave Clock 0.6 0.4 2.0 VP-P (min) VP-P (max) Square Wave Clock 0.6 0.4 2.0 VP-P (min) VP-P (max) Parameter Conditions CLOCK INPUT CHARACTERISTICS VID II CIN Differential Clock Input Level Input Current VIN = 0 or VIN = VA ±1 µA Input Capacitance (Notes 10, 11) Differential 0.02 pF Each input to ground 1.5 pF DIGITAL CONTROL PIN CHARACTERISTICS VIH Logic High Input Voltage (Note 12) 0.85 x VA V (min) VIL Logic Low Input Voltage (Note 12) 0.15 x VA V (max) CIN Input Capacitance (Notes 11, 13) Each input to ground 1.2 Measured differentially, OutV = VA, VBG = Floating (Note 15) 710 Measured differentially, OutV = GND, VBG = Floating (Note 15) 510 pF DIGITAL OUTPUT CHARACTERISTICS VOD LVDS Differential Output Voltage 400 mVP-P (min) 920 mVP-P (max) 280 mVP-P (min) 720 mVP-P (max) Δ VO DIFF Change in LVDS Output Swing Between Logic Levels VOS Output Offset Voltage VOS Output Offset Voltage Δ VOS Output Offset Voltage Change Between Logic Levels IOS Output Short Circuit Current ZO Differential Output Impedance VOH CalRun High level output IOH = -400uA (Note 12) 1.65 1.5 V CalRun Low level output IOH = 400uA (Note 12) 0.15 0.3 V VOL ±1 mV VBG = Floating 800 mV VBG = VA (Note 15) 1200 mV ±1 mV ±4 mA Output+ & Output- connected to 0.8V 100 Ohms POWER SUPPLY CHARACTERISTICS IA Analog Supply Current PD = Low PD = High 524 1.8 600 mA (max) mA IDR Output Driver Supply Current PD = Low PD = High 116 0.012 165 mA (max) mA PD Power Consumption PD = Low PD = High 1.2 3.5 1.45 W (max) mW PSRR1 D.C. Power Supply Rejection Ratio Change in Full Scale Error with change in VA from 1.8V to 2.0V 30 dB PSRR2 A.C. Power Supply Rejection Ratio 248 MHz, 50mVP-P riding on VA 51 dB AC ELECTRICAL CHARACTERISTICS fCLK1 Maximum Input Clock Frequency 1.7 fCLK2 Minimum Input Clock Frequency 200 1.5 GHz (min) MHz Input Clock Duty Cycle 200 MHz ≤ Input clock frequency ≤ 1.5 GHz (Note 12) 50 20 80 % (min) % (max) tCL Input Clock Low Time (Note 11) 333 133 ps (min) tCH Input Clock High Time (Note 11) 333 133 ps (min) 45 55 % (min) % (max) tRS DCLK Duty Cycle (Note 11) 50 Reset Setup Time (Note 11) 150 9 ps www.national.com ADC081500 Symbol ADC081500 Symbol Parameter Conditions Typical (Note 8) Limits (Note 8) Units (Limits) tRH Reset Hold Time tSD Synchronizing Edge to DCLK Output Delay tRPW Reset Pulse Width tLHT Differential Low to High Transition 10% to 90%, CL = 2.5 pF Time 250 ps tHLT Differential High to Low Transition 10% to 90%, CL = 2.5 pF Time 250 ps tOSK DCLK to Data Output Skew 50% of DCLK transition to 50% of Data transition, SDR Mode and DDR Mode, 0° DCLK (Note 11) ±50 ps (max) tSU Data to DCLK Set-Up Time DDR Mode, 90° DCLK (Note 11) 400 ns tH DCLK to Data Hold Time DDR Mode, 90° DCLK (Note 11) 560 ns tAD Sampling (Aperture) Delay Input CLK+ Fall to Acquisition of Data 1.3 ns tAJ Aperture Jitter 0.4 ps rms tOD Input Clock to Data Output Delay 50% of Input Clock transition to 50% of (in addition to Pipeline Delay) Data transition 3.1 ns Pipeline Delay (Latency) (Notes 11, 14) Over Range Recovery Time (Note 11) 250 ps tOD + tOSK (Note 11) 4 D Outputs 13 Dd Outputs 14 Differential VIN step from ±1.2V to 0V to get accurate conversion CLK± Cycles (min) Input CLK± Cycles 1 Input CLK± Cycle 500 ns tWU PD low to Rated Accuracy Conversion (Wake-Up Time) fSCLK Serial Clock Frequency (Note 11) 100 MHz tSSU Data to Serial Clock Setup Time (Note 11) 2.5 ns (min) tSH Data to Serial Clock Hold Time 1 ns (min) (Note 11) Serial Clock Low Time 4 Serial Clock High Time 4 ns (min) tCAL Calibration Cycle Time tCAL_L CAL Pin Low Time See Figure 9 (Note 11) 80 CLK± Cycles (min) tCAL_H CAL Pin High Time See Figure 9 (Note 11) 80 CLK± Cycles (min) CalDly = Low See Section 1.1.1, Figure 9, (Note 11) 225 CLK± Cycles (min) CalDly = High See Section 1.1.1, Figure 9, (Note 11) 231 CLK± Cycles (max) tCalDly Calibration delay determined by pin 127 1.4 x 105 ns (min) CLK± Cycles Note 1: Absolute Maximum Ratings indicate limits beyond which damage to the device may occur. There is no guarantee of operation at the Absolute Maximum Ratings. Operating Ratings indicate conditions for which the device is functional, but do not guarantee specific performance limits. For guaranteed specifications and test conditions, see the Electrical Characteristics. The guaranteed specifications apply only for the test conditions listed. Some performance characteristics may degrade when the device is not operated under the listed test conditions. Note 2: All voltages are measured with respect to GND = DR GND = 0V, unless otherwise specified. Note 3: When the input voltage at any pin exceeds the power supply limits (that is, less than GND or greater than VA), the current at that pin should be limited to 25 mA. The 50 mA maximum package input current rating limits the number of pins that can safely exceed the power supplies with an input current of 25 mA to two. This limit is not placed upon the power, ground and digital output pins. Note 4: Human body model is 100 pF capacitor discharged through a 1.5 kΩ resistor. Machine model is 220 pF discharged through ZERO Ohms. Note 5: See AN-450, “Surface Mounting Methods and Their Effect on Product Reliability”. Note 6: The analog inputs are protected as shown below. Input voltage magnitudes beyond the Absolute Maximum Ratings may damage this device. www.national.com 10 ADC081500 20153104 Note 7: To guarantee accuracy, it is required that VA and VDR be well bypassed. Each supply pin must be decoupled with separate bypass capacitors. Additionally, achieving rated performance requires that the backside exposed pad be well grounded. Note 8: Typical figures are at TA = 25°C, and represent most likely parametric norms. Test limits are guaranteed to National's AOQL (Average Outgoing Quality Level). Note 9: Calculation of Full-Scale Error for this device assumes that the actual reference voltage is exactly its nominal value. Full-Scale Error for this device, therefore, is a combination of Full-Scale Error and Reference Voltage Error. See Figure 2. For relationship between Gain Error and Full-Scale Error, see Specification Definitions for Gain Error. Note 10: The analog and clock input capacitances are die capacitances only. Additional package capacitances of 0.65 pF differential and 0.95 pF each pin to ground are isolated from the die capacitances by lead and bond wire inductances. Note 11: This parameter is guaranteed by design and is not tested in production. Note 12: This parameter is guaranteed by design and/or characterization and is not tested in production. Note 13: The digital control pin capacitances are die capacitances only. Additional package capacitance of 1.6 pF each pin to ground are isolated from the die capacitances by lead and bond wire inductances. Note 14: The ADC081500 converter has two LVDS output buses, which each clock data out at one half the sample rate. The second bus (D0 through D7) has a pipeline latency that is one Input Clock cycle less than the latency of the first bus (Dd0 through Dd7). Note 15: Tying VBG to the supply rail will increase the output offset voltage (VOS) by 400mv (typical), as shown in the VOS specification above. Tying VBG to the supply rail will also affect the differential LVDS output voltage (VOD), causing it to increase by 40mV (typical). 11 www.national.com ADC081500 Specification Definitions APERTURE (SAMPLING) DELAY is that time required after the fall of the clock input for the sampling switch to open. The Sample/Hold circuit effectively stops capturing the input signal and goes into the “hold” mode the aperture delay time (tAD) after the input clock goes low. APERTURE JITTER (tAJ) is the variation in aperture delay from sample to sample. Aperture jitter shows up as input noise. Bit Error Rate (B.E.R.) is the probability of error and is defined as the probable number of errors per unit of time divided by the number of bits seen in that amount of time. A B.E.R. of 10-18 corresponds to a statistical error in one bit about every four (4) years. CLOCK DUTY CYCLE is the ratio of the time that the clock wave form is at a logic high to the total time of one clock period. DIFFERENTIAL NON-LINEARITY (DNL) is the measure of the maximum deviation from the ideal step size of 1 LSB. Measured at sample rate = ?? MSPS with a 1MHz input sinewave. EFFECTIVE NUMBER OF BITS (ENOB, or EFFECTIVE BITS) is another method of specifying Signal-to-Noise and Distortion Ratio, or SINAD. ENOB is defined as (SINAD − 1.76) / 6.02 and says that the converter is equivalent to a perfect ADC of this (ENOB) number of bits. FULL POWER BANDWIDTH (FPBW) is a measure of the frequency at which the reconstructed output fundamental drops 3 dB below its low frequency value for a full scale input. GAIN ERROR is the deviation from the ideal slope of the transfer function. It can be calculated from Offset and FullScale Errors: Pos. Gain Error = Offset Error − Pos. Full-Scale Error Neg. Gain Error = −(Offset Error − Neg. Full-Scale Error) Gain Error = Neg. Full-Scale Error − Pos. Full-Scale Error = Pos. Gain Error + Neg. Gain Error INTEGRAL NON-LINEARITY (INL) is a measure of worst case deviation of the ADC transfer function from an ideal straight line drawn through the ADC transfer function. The deviation of any given code from this straight line is measured from the center of that code value step. The best fit method is used. INTERMODULATION DISTORTION (IMD) is the creation of additional spectral components as a result of two sinusoidal frequencies being applied to the ADC input at the same time. it is defined as the ratio of the power in the second and third order intermodulation products to the power in one of the original frequencies. IMD is usually expressed in dBFS. LSB (LEAST SIGNIFICANT BIT) is the bit that has the smallest value or weight of all bits. This value is 20153146 FIGURE 1. LVDS OUTPUT OFFSET VOLTAGE (VOS) is the midpoint between the D+ and D- pins output voltage; i.e., [(VD+) + ( VD-)]/2. MISSING CODES are those output codes that are skipped and will never appear at the ADC outputs. These codes cannot be reached with any input value. MSB (MOST SIGNIFICANT BIT) is the bit that has the largest value or weight. Its value is one half of full scale. NEGATIVE FULL-SCALE ERROR (NFSE) is a measure of how far the first code transition is from the ideal 1/2 LSB above a differential −VIN /2. For the ADC081500 the reference voltage is assumed to be ideal, so this error is a combination of full-scale error and reference voltage error. OFFSET ERROR (VOFF) is a measure of how far the midscale point is from the ideal zero voltage differential input. Offset Error = Actual Input causing average of 8k samples to result in an average code of 127.5. OUTPUT DELAY (tOD) is the time delay (in addition to Pipeline Delay) after the falling edge of CLK+ before the data update is present at the output pins. OVER-RANGE RECOVERY TIME is the time required after the differential input voltages goes from ±1.2V to 0V for the converter to recover and make a conversion with its rated accuracy. PIPELINE DELAY (LATENCY) is the number of input clock cycles between initiation of conversion and when that data is presented to the output driver stage. New data is available at every clock cycle, but the data lags the conversion by the Pipeline Delay plus the tOD. POSITIVE FULL-SCALE ERROR (PFSE) is a measure of how far the last code transition is from the ideal 1-1/2 LSB below a differential + VIN. For the ADC081500 the reference voltage is assumed to be ideal, so this error is a combination of full-scale error and reference voltage error. POWER SUPPLY REJECTION RATIO (PSRR) can be one of two specifications. PSRR1 (DC PSRR) is the ratio of the change in full-scale error that results from a power supply voltage change from 1.8V to 2.0V. PSRR2 (AC PSRR) is a measure of how well an a.c. signal riding upon the power supply is rejected from the output and is measured with a 248 MHz, 50 mVP-P signal riding upon the power supply. It is the ratio of the output amplitude of that signal at the output to its amplitude on the power supply pin. PSRR is expressed in dB. VFS / 2n where VFS is the differential full-scale amplitude VIN as set by the FSR input and "n" is the ADC resolution in bits, which is 8 for the ADC081500. LVDS DIFFERENTIAL OUTPUT VOLTAGE (VOD) is the absolute value of the difference between the VD+ & VD- outputs; each measured with respect to Ground. www.national.com 12 where Af1 is the RMS power of the fundamental (output) frequency and Af2 through Af10 are the RMS power of the first 9 harmonic frequencies in the output spectrum. – Second Harmonic Distortion (2nd Harm) is the difference, expressed in dB, between the RMS power in the input frequency seen at the output and the power in its 2nd harmonic level at the output. – Third Harmonic Distortion (3rd Harm) is the difference expressed in dB between the RMS power in the input frequency seen at the output and the power in its 3rd harmonic level at the output. Transfer Characteristic 20153122 FIGURE 2. Input / Output Transfer Characteristic 13 www.national.com ADC081500 SIGNAL TO NOISE RATIO (SNR) is the ratio, expressed in dB, of the rms value of the input signal at the output to the rms value of the sum of all other spectral components below onehalf the sampling frequency, not including harmonics or d.c. SIGNAL TO NOISE PLUS DISTORTION (S/(N+D) or SINAD) is the ratio, expressed in dB, of the rms value of the input signal at the output to the rms value of all of the other spectral components below half the input clock frequency, including harmonics but excluding d.c. SPURIOUS-FREE DYNAMIC RANGE (SFDR) is the difference, expressed in dB, between the rms values of the input signal at the output and the peak spurious signal, where a spurious signal is any signal present in the output spectrum that is not present at the input, excluding d.c. TOTAL HARMONIC DISTORTION (THD) is the ratio expressed in dB, of the rms total of the first nine harmonic levels at the output to the level of the fundamental at the output. THD is calculated as ADC081500 Timing Diagrams 20153114 FIGURE 3. ADC081500 Timing — SDR Clocking 20153159 FIGURE 4. ADC081500 Timing — DDR Clocking www.national.com 14 ADC081500 20153119 FIGURE 5. Serial Interface Timing 20153120 FIGURE 6. Clock Reset Timing in DDR Mode 20153123 FIGURE 7. Clock Reset Timing in SDR Mode with OUTEDGE Low 15 www.national.com ADC081500 20153124 FIGURE 8. Clock Reset Timing in SDR Mode with OUTEDGE High 20153125 FIGURE 9. Self Calibration and On-Command Calibration Timing www.national.com 16 VA = VDR = 1.9V, FCLK = 1500MHz, TA = 25°C unless otherwise stated. INL vs. CODE INL vs. TEMPERATURE 20153164 20153165 DNL vs. CODE DNL vs. TEMPERATURE 20153166 20153167 POWER DISSIPATION vs. SAMPLE RATE ENOB vs. TEMPERATURE 20153181 20153176 17 www.national.com ADC081500 Typical Performance Characteristics ADC081500 ENOB vs. SUPPLY VOLTAGE ENOB vs. SAMPLE RATE 20153177 20153178 ENOB vs. INPUT FREQUENCY SNR vs. TEMPERATURE 20153179 20153168 SNR vs. SUPPLY VOLTAGE SNR vs. SAMPLE RATE 20153169 www.national.com 20153170 18 ADC081500 SNR vs. INPUT FREQUENCY THD vs. TEMPERATURE 20153171 20153172 THD vs. SUPPLY VOLTAGE THD vs. SAMPLE RATE 20153173 20153174 THD vs. INPUT FREQUENCY SFDR vs. TEMPERATURE 20153175 20153185 19 www.national.com ADC081500 SFDR vs. SUPPLY VOLTAGE SFDR vs. SAMPLE RATE 20153184 20153182 SFDR vs. INPUT FREQUENCY Spectral Response at FIN = 373 MHz 20153183 20153187 Spectral Response at FIN = 745 MHz FULL POWER BANDWIDTH 20153188 www.national.com 20153186 20 The ADC081500 is a versatile A/D Converter with an innovative architecture permitting very high speed operation. The controls available ease the application of the device to circuit solutions. Optimum performance requires adherence to the provisions discussed here and in the Applications Information Section. While it is generally poor practice to allow an active pin to float, pins 4 and 14 of the ADC081500 are designed to be left floating without jeopardy. In all discussions throughout this data sheet, whenever a function is called by allowing a control pin to float, connecting that pin to a potential of one half the VA supply voltage will have the same effect as allowing it to float. 1.1 OVERVIEW The ADC081500 uses a calibrated folding and interpolating architecture that achieves 7.4 effective bits. The use of folding amplifiers greatly reduces the number of comparators and power consumption. Interpolation reduces the number of front-end amplifiers required, minimizing the load on the input signal and further reducing power requirements. In addition to other things, on-chip calibration reduces the INL bow often seen with folding architectures. The result is an extremely fast, high performance, low power converter. The analog input signal that is within the converter's input voltage range is digitized to eight bits at speeds of 200 MSPS to 1.7 GSPS, typical. Differential input voltages below negative full-scale will cause the output word to consist of all zeroes. Differential input voltages above positive full-scale will cause the output word to consist of all ones. Either of these conditions at the input will cause the OR (Out of Range) output to be activated. That is, the single OR output indicates the output code is below negative full scale or above positive full scale. The ADC081500 has a 1:2 demultiplexer that feeds two LVDS output buses. The data on these buses provide an output word rate on each bus at half the ADC sampling rate and must be interleaved by the user to provide output words at the full conversion rate. The output levels may be selected to be normal or reduced. Using reduced levels saves power but could result in erroneous data capture of some or all of the bits, especially at higher sample rates and in marginally designed systems. Calibration Operation Notes: • During the calibration cycle, the OR output may be active as a result of the calibration algorithm. All data on the output pins and the OR output are invalid during the calibration cycle. • During the power-up calibration and during the oncommand calibration, all clocks are halted on chip, including internal clocks and DCLK, while the input termination resistor is trimmed to a value that is equal to REXT / 33. This is to reduce noise during the input resistor calibration portion of the calibration cycle. See??2.4.2 for information on maintaining DCLK operation during oncommand calibration. This external resistor is located between pin 32 and ground. REXT must be 3300 Ω ±0.1%. With this value, the input termination resistor is trimmed to be 100 Ω. Because REXT is also used to set the proper current for the Track and Hold amplifier, for the preamplifiers and for the comparators, other values of REXT should not be used. • The CalRun output is high whenever the calibration procedure is running. This is true whether the calibration is done at power-up or on-command. 1.1.1 Self-Calibration A self-calibration is performed upon power-up and can also be invoked by the user upon command. Calibration trims the 100Ω analog input differential termination resistor and minimizes full-scale error, offset error, DNL and INL, resulting in maximizing SNR, THD, SINAD (SNDR) and ENOB. Internal bias currents are also set with the calibration process. All of this is true whether the calibration is performed upon power up or is performed upon command. Running the self calibration is an important part of this chip's functionality and is required in order to obtain adequate performance. In addition to the requirement to be run at power-up, self calibration must be re-run whenever the sense of the FSR pin is changed. For best performance, we recommend that self calibration be run 20 seconds or more after application of power and whenever the operating temperature changes significantly relative to the specific system performance requirements. See Section 2.4.2.2 for more information. Calibration can not be initiated or run while the device is in the power-down mode. See Section 1.1.7 for information on the interaction between Power Down and Calibration. 1.1.2 Acquiring the Input Data is acquired at the falling edge of CLK+ (pin 18) and the digital equivalent of that data is available at the digital outputs 13 input clock cycles later for the D output bus and 14 input clock cycles later for the Dd output bus. There is an additional internal delay called tOD before the data is available at the outputs. See the Timing Diagram. The ADC081500 will convert as long as the input clock signal is present. The fully differential comparator design and the innovative design of the sample-and-hold amplifier, together with self calibration, enables a very flat SINAD/ENOB response beyond 1.5 GHz. The ADC081500 output data signaling is LVDS and the output format is offset binary. 1.1.3 Control Modes Much of the user control can be accomplished with several control pins that are provided. Examples include initiation of the calibration cycle, power down mode and full scale range setting. However, the ADC081500 also provides an Extended Control mode whereby a serial interface is used to access 21 www.national.com ADC081500 In normal operation, calibration is performed just after application of power and whenever a valid calibration command is given, which is holding the CAL pin low for at least tCAL_L clock cycles, then hold it high for at least another tCAL_H clock cycles as defined in the Converter Electrical Characteristics. The time taken by the calibration procedure is specified as tCALin Converter Electrical Characteristics. Holding the CAL pin high upon power up will prevent the calibration process from running until the CAL pin experiences the above-mentioned tCAL_L clock cycles followed by tCAL_H clock cycles. CalDly (pin 127) is used to select one of two delay times after the application of power to the start of calibration. This calibration delay time is depedent on the setting of the CalDly pin and is specified as tCalDly in the Converter Electrical Characteristics. These delay values allow the power supply to come up and stabilize before calibration takes place. If the PD pin is high upon power-up, the calibration delay counter will be disabled until the PD pin is brought low. Therefore, holding the PD pin high during power up will further delay the start of the power-up calibration cycle. The best setting of the CalDly pin depends upon the power-on settling time of the power supply. 1.0 Functional Description ADC081500 register-based control of several advanced features. The Extended Control mode is not intended to be enabled and disabled dynamically. Rather, the user is expected to employ either the Normal Control mode or the Extended Control mode at all times. When the device is in the Extended Control mode, pin-based control of several features is replaced with register-based control and those pin-based controls are disabled. These pins are OutV (pin 3), OutEdge/DDR (pin 4), FSR (pin 14) and CalDly (pin 127). See Section 1.2 for details on the Extended Control mode. es. With double data rate the DCLK frequency is half the data rate and data is sent to the outputs on both DCLK edges. DDR clocking is enabled in Normal Control mode by allowing pin 4 to float. 1.1.6 The LVDS Outputs The data outputs, the Out Of Range (OR) and DCLK, are LVDS. Output current sources provide 3 mA of output current to a differential 100 Ohm load when the OutV input (pin 14) is high or 2.2 mA when the OutV input is low. For short LVDS lines and low noise systems, satisfactory performance may be realized with the OutV input low, which results in lower power consumption. If the LVDS lines are long and/or the system in which the ADC081500 is used is noisy, it may be necessary to tie the OutV pin high. The LVDS data output have a typical common mode voltage of 800mV when the VBG pin is unconnected and floating. This common mode voltage can be increased to 1.2V by tying the VBG pin to VA if a higher common mode is required. IMPORTANT NOTE: Tying the VBG pin to VA will also increase the differential LVDS output voltage by up to 40mV. 1.1.4 The Analog Inputs The ADC081500 must be driven with a differential input signal. Operation with a single-ended signal is not recommended. It is important that the inputs either be a.c. coupled to the inputs with the VCMO pin grounded, or d.c. coupled with the VCMO pin left floating. An input common mode voltage equal to the VCMO output must be provided when d.c. coupling is used. The input full-scale range is programmable in the normal mode by setting a level on pin 14 (FSR) as defined in by the specification VIN in the Converter Electrical Characteristics. The full-scale range setting operates equally on both ADCs. In the Extended Control mode, programming the Input FullScale Voltage Adjust register allows the input full-scale range to be adjusted as described in Section 2.2 and Section 1.1.7 Power Down The ADC081500 is in the active state when the Power Down pin (PD) is low. When the PD pin is high, the device is in the power down mode. In this power down mode the data output pins (positive and negative) are put into a high impedance state and the devices power consumption is reduced to a minimal level. The DCLK+/- and OR +/- are not tri-stated, they are weakly pulled down to ground internally. Therefore when the device is powered down the DCLK +/- and OR +/- should not be terminated to a DC voltage. Also note, that upon return to normal operation after power down mode, the pipeline will contain meaningless information. If the PD input is brought high while a calibration is running, the device will not go into power down until the calibration sequence is complete. However, if power is applied and PD is already high, the device will not begin the calibration sequence until the PD input goes low. If a manual calibration is requested while the device is powered down, the calibration will not begin at all. That is, the manual calibration input is completely ignored in the power down state. 1.1.5 Clocking The ADC081500 must be driven with an a.c. coupled, differential clock signal. Section 2.3 describes the use of the clock input pins. A differential LVDS output clock is available for use in latching the ADC output data into whatever device is used to receive the data. The ADC081500 offers options for output clocking. These options include a choice of which DCLK edge the output data transitions on, and a choice of Single Data Rate (SDR) or Double Data Rate (DDR) outputs. The ADC081500 also has the option to use a duty cycle corrected clock receiver as part of the input clock circuit. This feature is enabled by default and provides improved ADC clocking. This circuitry allows the ADC to be clocked with a signal source having a duty cycle ratio of 80 / 20 % (worst case). 1.2 NORMAL/EXTENDED CONTROL MODES The ADC081500 may be operated in one of two modes. In the simpler Normal Control mode, the user affects available configuration and control of the device through several control pins. The Extended Control mode provides additional configuration and control options through a serial interface and a set of 3 registers. The two control modes are selected with pin 14 (FSR/ECE: Extended Control Enable). The choice of control modes is required to be a fixed selection and is not intended to be switched dynamically while the device is operational. Table 1 shows how several of the device features are affected by the control mode chosen. 1.1.5.1 OutEdge Setting To help ease data capture in the SDR mode, the output data may be caused to transition on either the positive or the negative edge of the output data clock (DCLK). This is chosen with the OutEdge input (pin 4). A high on the OutEdge input pin causes the output data to transition on the rising edge of DCLK, while grounding this input causes the output to transition on the falling edge of DCLK. See Section 2.4.3. 1.1.5.2 Double Data Rate A choice of single data rate (SDR) or double data rate (DDR) output is offered. With single data rate the output clock DCLK frequency is the same as the data rate of the two output bus- www.national.com 22 Feature Normal Control Mode Extended Control Mode SDR or DDR Clocking Selected with nDE in the Configuration DDR Clocking selected with pin 4 Register (1h; bit-10). When the device is floating. SDR clocking selected when pin in DDR mode, address 1h, bit-8 must be 4 not floating. set to 0b. DDR Clock Phase Not Selectable (0° Phase Only) Selected with DCP in the Configuration Register (1h; bit-11). SDR Data transitions with rising edge of SDR Data transitions with rising or falling Selected with OE in the Configuration DCLK+ when pin 4 is high and on falling DCLK edge Register (1h; bit-8). edge when low. LVDS output level Normal differential data and DCLK amplitude selected when pin 3 is high and reduced amplitude selected when low. Power-On Calibration Delay Short delay selected when pin 127 is low Short delay only. and longer delay selected when high. Full-Scale Range Normal input full-scale range selected when pin 14 is high and reduced range when low. Selected range applies to both channels. Up to 512 step adjustments over a nominal range specified in ??1.4. Selected using the Input Full-Scale Adjust register (3h; bits-7 thru 15). Input Offset Adjust Not possible 512 steps of adjustment using the Input Offset register (2h; bits-7 thru 15) as specified in ??1.4. SDATA: Each register access requires a specific 32-bit pattern at this input. This pattern consists of a header, register address and register value. The data is shifted in MSB first. Setup and hold times with respect to the SCLK must be observed. See the Timing Diagram. Each Register access consists of 32 bits, as shown in Figure 5 of the Timing Diagrams. The fixed header pattern is 0000 0000 0001 (eleven zeros followed by a 1). The loading sequence is such that a "0" is loaded first. These 12 bits form the header. The next 4 bits are the address of the register that is to be written to and the last 16 bits are the data written to the addressed register. The addresses of the various registers are indicated in Table 3. Refer to the Register Description (Section 1.4) for information on the data to be written to the registers. Subsequent register accesses may be performed immediately, starting with the 33rd SCLK. This means that the SCS input does not have to be de-asserted and asserted again between register addresses. It is possible, although not recommended, to keep the SCS input permanently enabled (at a logic low) when using extended control. IMPORTANT NOTE: The Serial Interface should not be used when calibrating the ADC. Doing so will impair the performance of the device until it is re-calibrated correctly. Programming the serial registers will also reduce dynamic performance of the ADC for the duration of the register access time. The default state of the Extended Control Mode is set upon power-on reset (internally performed by the device) and is shown in Table 2. TABLE 2. Extended Control Mode Operation (Pin 14 Floating) Feature Extended Control Mode Default State SDR or DDR Clocking DDR Clocking DDR Clock Phase Data changes with DCLK edge (0° phase) LVDS Output Amplitude Normal amplitude (710 mVP-P) Calibration Delay Short Delay Full-Scale Range 700 mV nominal Input Offset Adjust No adjustment Selected with the OV in the Configuration Register (1h; bit-9). 1.3 THE SERIAL INTERFACE IMPORTANT NOTE: During the initial write using the serial interface, all 3 user registers must be written with desired or default values. Once all registers have been written once, other desired settings can be loaded. The 3-pin serial interface is enabled only when the device is in the Extended Control mode. The pins of this interface are Serial Clock (SCLK), Serial Data (SDATA) and Serial Interface Chip Select (SCS) Three write only registers are accessible through this serial interface. SCS: This signal should be asserted low while accessing a register through the serial interface. Setup and hold times with respect to the SCLK must be observed. SCLK: Serial data input is accepted with the rising edge of this signal. There is no minimum frequency requirement for SCLK. 23 www.national.com ADC081500 TABLE 1. Features and modes ADC081500 TABLE 3. Register Addresses Bit 10 4-Bit Address Loading Sequence: A3 loaded after Fixed Header Pattern, A0 loaded last A3 A2 A1 A0 Hex Register Addressed 0 0 0 0 0h Reserved 0 0 0 1 1h Configuration 0 0 1 0 2h Input Offset 0 0 1 1 3h Input Full-Scale Voltage Adjust 0 1 0 0 4h Reserved 0 1 0 1 5h Reserved 0 1 1 0 6h Reserved 0 1 1 1 7h Reserved 1 0 0 0 8h Reserved 1 0 0 1 9h Reserved 1 0 1 0 Ah Reserved 1 0 1 1 Bh Reserved 1 1 0 0 Ch Reserved 1 1 0 1 Dh Reserved 1 1 1 0 Eh Reserved 1 1 1 1 Fh Reserved Bit 9 Bit 8 1.4 REGISTER DESCRIPTION Three write-only registers provide several control and configuration options in the Extended Control Mode. These registers have no effect when the device is in the Normal Control Mode. Each register description below also shows the PowerOn Reset (POR) state of each control bit. Bits 7:0 Input Offset Addr: 2h (0010b) Configuration Register Addr: 1h (0001b) D15 D14 D13 1 0 1 W only (0xB2FF) D12 D15 D12 D11 D10 D9 Offset Value D8 (LSB) D11 D10 D9 D8 nDE OV OE D7 D6 D5 D4 D3 D2 D1 D0 1 1 1 1 1 1 1 D6 D5 D4 D3 D2 D1 D0 1 1 1 1 1 1 1 1 Bits 15:8 www.national.com D13 W only (0x007F) DCS DCP D7 Bit 11 D14 (MSB) Sign Bit 15 Bit 14 Bit 13 Bit 12 nDE: DDR Enable. When this bit is set to 0b, data bus clocking follows the DDR (Double Data Rate) mode whereby a data word is output with each rising and falling edge of DCLK. When this bit is set to a 1b, data bus clocking follows the SDR (single data rate) mode whereby each data word is output with either the rising or falling edge of DCLK, as determined by the OutEdge bit. POR State: 0b OV: Output Voltage. This bit determines the LVDS outputs' voltage amplitude and has the same function as the OutV pin that is used in the normal control mode. When this bit is set to 1b, the standard output amplitude of 710 mVP-P is used. When this bit is set to 0b, the reduced output amplitude of 510 mVP-P is used. POR State: 1b OE: Output Edge. This bit selects the DCLK edge with which the data words transition in the SDR mode and has the same effect as the OutEdge pin in the normal control mode. When this bit is 1, the data outputs change with the rising edge of DCLK+. When this bit is 0, the data output change with the falling edge of DCLK+. POR State: 0b Must be set to 1b. Must be set to 1b Must be set to 0b Must be set to 1b DCS: Duty Cycle Stabilizer. When this bit is set to 1b, a duty cycle stabilization circuit is applied to the clock input. When this bit is set to 0b the stabilization circuit is disabled. POR State: 1b DCP: DDR Clock Phase. This bit only has an effect in the DDR mode. When this bit is set to 0b, the DCLK edges are time-aligned with the data bus edges ("0° Phase"). When this bit is set to a 1b, the DCLK edges are placed in the middle of the data bit-cells ("90° Phase"). POR State: 0b Bit 7 Bit 6:0 24 Input Offset Value. The input offset of the ADC is adjusted linearly and monotonically by the value in this field. 00h provides a nominal zero offset, while FFh provides a nominal 45 mV of offset. Thus, each code step provides 0.176 mV of offset. POR State: 0000 0000 b Sign bit. 0b gives positive offset, 1b gives negative offset. POR State: 0b Must be set to 1b D15 D14 D13 W only (0x807F) D12 (MSB) D11 D10 D9 D8 Adjust Value D7 D6 D5 D4 D3 D2 D1 D0 (LSB) 1 1 1 1 1 1 1 Bit 15:7 Bits 6:0 Input Full Scale Voltage Adjust Value. The input full-scale voltage or gain of the ADC is adjusted linearly and monotonically with a 9 bit data value. The adjustment range is ±20% of the nominal 700 mVP-P differential value. 0000 0000 0 560mVP-P 1000 0000 0 Default Value 1111 1111 1 700mVP-P 840mVP-P For best performance, it is recommended that the value in this field be limited to the range of 0110 0000 0b to 1110 0000 0b. i.e., limit the amount of adjustment to ±15%. The remaining ±5% headroom allows for the ADC's own full scale variation. A gain adjustment does not require ADC re-calibration. POR State: 1000 0000 0b (no adjustment) Must be set to 1b 2.0 Applications Information 1.4.1 Note Regarding Extended Mode Offset Correction When using the Input Offset Adjust register, the following information should be noted. For offset values of +0000 0000 and -0000 0000, the actual offset is not the same. By changing only the sign bit in this case, an offset step in the digital output code of about 1/10th of an LSB is experienced. This is shown more clearly in the Figure below. Table 52.1 THE REFERENCE VOLTAGE The voltage reference for the ADC081500 is derived from a 1.254V bandgap reference which is made available at pin 31, VBG for user convenience. This output has an output current capability of ±100 μA and should be buffered if more current than this is required. The internal bandgap-derived reference voltage has a nominal value of VIN, as determined by the FSR pin and described in Section 1.1.4. There is no provision for the use of an external reference voltage, but the full-scale input voltage can be adjusted through a Configuration Register in the Extended Control mode, as explained in Section 1.2. Differential input signals up to the chosen full-scale level will be digitized to 8 bits. Signal excursions beyond the full-scale range will be clipped at the output. These large signal excursions will also activate the OR output for the time that the signal is out of range. See Section 2.2.2. One extra feature of the VBG pin is that it can be used to raise the common mode voltage level of the LVDS outputs. The output offset voltage (V OS) is typically 800 mV when the VBG pin is used as an output or left unconnected. To raise the LVDS offset voltage to a typical value of 1200 mV the VBG pin can be connected directly to the supply rails. 20153130 2.2 THE ANALOG INPUT The analog input is a differential one to which the signal source may be a.c. coupled or d.c. coupled. In the normal mode, the full-scale input range is selected using the FSR pin as specified in the Converter Electrical Characteristics. In the Extended Control mode, the full-scale input range is selected by programming the Full-Scale Voltage Adjust register FIGURE 10. Extended Mode Offset Behavior 1.5 MULTIPLE ADC SYNCHRONIZATION The ADC081500 has the capability to precisely reset its sampling clock input to DCLK output relationship as determined by the user-supplied DCLK_RST pulse. This allows multiple 25 www.national.com ADC081500 ADCs in a system to have their DCLK (and data) outputs transition at the same time with respect to the shared CLK input that all the ADCs use for sampling. The DCLK_RST signal must observe some timing requirements that are shown in Figure 6, Figure 7 and Figure 8 of the Timing Diagrams. The DCLK_RST pulse must be of a minimum width and its deassertion edge must observe setup and hold times with respect to the CLK input rising edge. These timing specifications are listed as tRH, tRS, and tRPW in the Converter Electrical Characteristics. The DCLK_RST signal can be asserted asynchronous to the input clock. If DCLK_RST is asserted, the DCLK output is held in a designated state. The state in which DCLK is held during the reset period is determined by the mode of operation (SDR/ DDR) and the setting of the Output Edge configuration pin or bit. (Refer to Figure 6, Figure 7 and Figure 8 for the DCLK reset state conditions). Therefore, depending upon when the DCLK_RST signal is asserted, there may be a narrow pulse on the DCLK line during this reset event. When the DCLK_RST signal is de-asserted in synchronization with the CLK rising edge, the next CLK falling edge synchronizes the DCLK output with those of other ADC081500s in the system. The DCLK output is enabled again after a constant delay (relative to the input clock frequency) which is equal to the CLK input to DCLK output delay (tSD). The device always exhibits this delay characteristic in normal operation. The DCLK_RST pin should NOT be brought high while the calibration process is running (while CalRun is high). Doing so could cause a digital glitch in the digital circuitry, resulting in corruption and invalidation of the calibration. Input Full-Scale Voltage Adjust Addr: 3h (0011b) ADC081500 single-ended signals is to convert them to differential signals before presenting them to the ADC. through the Serial Interface. For best performance when adjusting the input full-scale range in the Extended Control, refer to ?? 1.4 for guidelines on limiting the amount of adjustment. Table 4 gives the input to output relationship with the FSR pin high when the normal (non-extended) mode is used. With the FSR pin grounded, the millivolt values inTable 4 are reduced to 75% of the values indicated. In the Enhanced Control Mode, these values will be determined by the full scale range and offset settings in the Control Registers. 2.2.1.1 A.C. Coupled Input The easiest way to accomplish single-ended a.c. input to differential a.c. signal is by using an appropriate balun, as shown in Figure 12. TABLE 4. DIFFERENTIAL INPUT TO OUTPUT RELATIONSHIP (Normal Control Mode, FSR High) VIN+ VIN− Output Code VCM − 217.5mV VCM + 217.5mV 0000 0000 VCM − 109 mV VCM + 109 mV 0100 0000 VCM 0111 1111 / 1000 0000 VCM + 109 mV VCM −109 mV 1100 0000 VCM + 217.5mV VCM − 217.5mV 1111 1111 VCM 20153143 FIGURE 12. Single-Ended To Differential Signal Conversion Using a Balun Figure 12 is a generic depiction of a single-ended to differential signal conversion using a balun. The circuitry specific to the balun will depend on the type of balun selected and the overall board layout. It is recommended that the system designer contact the manufacturer of the balun they have selected to aid in designing the best performing single-ended to differential conversion circuit using that particular balun. When selecting a balun, it is important to understand the input architecture of the ADC. There are specific balun parameters of which the system designer should be mindful. They should match the impedance of their analog source to the ADC081500's on-chip 100 differential input termination resistor. The range of this termination resistor is described in the electrical table as the specification RIN. Also, as a result of the ADC architecture, the phase and amplitude balance are important. The lowest possible phase and amplitude imbalance is desired when selecting a balun. The phase imbalance should be no more than ±2.5° and the amplitude imbalance should be limited to less than 1dB at the desired input frequency range. Finally, when selecting a balun, the VSWR (Voltage Standing Wave Ratio), bandwidth and insertion loss of the balun should also be considered. The VSWR aids in determining the overall transmission line termination capability of the balun when interfacing to the ADC input. The insertion loss should be considered so that the signal at the balun output is within the specified input range of the ADC as described in the Converter Electrical Characteristics as the specification VIN. The buffered analog inputs simplify the task of driving these inputs and the RC pole that is generally used at sampling ADC inputs is not required. If it is desired to use an amplifier circuit before the ADC, use care in choosing an amplifier with adequate noise and distortion performance and adequate gain at the frequencies used for the application. Note that a precise d.c. common mode voltage must be present at the ADC inputs. This common mode voltage, VCMO, is provided on-chip when a.c. input coupling is used and the input signal is a.c. coupled to the ADC. When the inputs are a.c. coupled, the VCMO output must be grounded, as shown in Figure 11. This causes the on-chip VCMO voltage to be connected to the inputs through on-chip 50k-Ohm resistors. 20153144 2.2.1.2 D.C. Coupled Input When d.c. coupling to the ADC081500 analog inputs is required, single-ended to differential conversion may be easily accomplished with the LMH6555. An example of this type of circuit is shown in Figure 13. In such applications, the LMH6555 performs the task of single-ended to differential conversion while delivering low distortion and noise, as well as output balance, that supports the operation of the ADC081500. Connecting the ADC081500 VCMO pin to the VCM_REF pin of the LMH6555, through the appropriate buffer, will ensure that the common mode input voltage is as needed for optimum performance of the ADC081500. The LMV321 was chosen to buffer VCMO for its low voltage operation and reasonable offset voltage. Be sure that the current drawn from the VCMO output does not exceed 100 μA. FIGURE 11. Differential Input Drive When the d.c. coupled mode is used, a common mode voltage must be provided at the differential inputs that should track the VCMO output voltage. The VCMO output potential will change with temperature and the common mode output of the driving device should track this change. Full-scale distortion performance falls off rapidly as the input common mode voltage deviates from VCMO. This is a direct result of using a very low supply voltage to minimize power. Keep the input common voltage within 50 mV of VCMO. Performance is as good in the d.c. coupled mode as in the a.c. coupled mode, provided the input common mode voltage at both analog inputs remain within 50 mV of VCMO. 2.2.1 Handling Single-Ended Input Signals There is no provision for the ADC081500 to adequately process single-ended input signals. The best way to handle www.national.com 26 20153155 FIGURE 13. Example of Servoing the Analog Input with VCMO In Table 5, RADJ- and RADJ+ are used to adjust the differential offset that can be measured at the ADC inputs VIN+ / VIN-. An unadjusted positive offset with reference to VIN- greater than |15mV| should be reduced with a resistor in the RADJ- position. Likewise, an unadjusted negative offset with reference to VIN- greater than |15mV| should be reduced with a resistor in the RADJ+ position. Table 5 gives suggested RADJ- and RADJ+ values for various unadjusted differential offsets to bring the VIN+ / VIN- offset back to within |15mV|. 20153147 FIGURE 14. Differential (LVDS) Input Clock Connection The differential Clock line pair should have a characteristic impedance of 100Ω and be terminated at the clock source in that (100Ω) characteristic impedance. The clock line should be as short and as direct as possible. The ADC081500 clock input is internally terminated with an untrimmed 100Ω resistor. Insufficient clock levels will result in poor dynamic performance. Excessively high clock levels could cause a change in the analog input offset voltage. To avoid these problems, keep the clock level within the range specified as VID in the Converter Electrical Characteristics. The low and high times of the input clock signal can affect the performance of any A/D Converter. The ADC081500 features a duty cycle clock correction circuit which can maintain performance over temperature. The ADC will meet its performance specification if the input clock high and low times are maintained within the duty cycle range as specified in the Converter Electrical Characteristics. High speed, high performance ADCs such as the ADC081500 require a very stable input clock signal with minimum phase noise or jitter. ADC jitter requirements are defined by the ADC resolution (number of bits), maximum ADC input frequency and the input signal amplitude relative to the ADC input full scale range. The maximum jitter (the sum of the jitter from all sources) allowed to prevent a jitter-induced reduction in SNR is found to be TABLE 5. D.C. Coupled Offset Adjustment Unadjusted Offset Reading Resistor Value 0mV to 10mV no resistor needed 11mV to 30mV 20.0kΩ 31mV to 50mV 10.0kΩ 51mV to 70mV 6.81kΩ 71mV to 90mV 4.75kΩ 91mV to 110mV 3.92kΩ 2.2.2 Out Of Range (OR) Indication When the conversion result is clipped the Out of Range output is activated such that OR+ goes high and OR- goes low. This output is active as long as accurate data on the output bus would be outside the range of 00h to FFh. 2.2.3 Full-Scale Input Range As with all A/D Converters, the input range is determined by the value of the ADC's reference voltage. The reference voltage of the ADC081500 is derived from an internal band-gap reference. The FSR pin controls the effective reference voltage of the ADC081500 such that the differential full-scale input range at the analog inputs is a normal amplitude with the FSR pin high, or a reduced amplitude with FSR pin low as defined by the specification VIN in the Converter Electrical Characteristics. Best SNR is obtained with FSR high, but better distortion and SFDR are obtained with the FSR pin low. tJ(MAX) = (VINFSR / VIN(P-P)) x (1/(2(N+1) x π x fIN)) where tJ(MAX) is the rms total of all jitter sources in seconds, VIN(P-P) is the peak-to-peak analog input signal, VINFSR is the full-scale range of the ADC, "n" is the ADC resolution in bits and fIN is the maximum input frequency, in Hertz, to the ADC analog input. Note that the maximum jitter described above is the arithmetic sum of the jitter from all sources, including that in the ADC input clock, that added by the system to the ADC input clock and input signals and that added by the ADC itself. Since the effective jitter added by the ADC is beyond user control, the best the user can do is to keep the sum of the externally added input clock jitter and the jitter added by the analog circuitry to the analog signal to a minimum. 2.3 THE CLOCK INPUTS The ADC081500 has differential LVDS clock inputs, CLK+ and CLK-, which must be driven with an a.c. coupled, differential clock signal. Although the ADC081500 is tested and its performance is guaranteed with a differential 1.5 GHz clock, it typically will function well with input clock frequencies indicated in the Converter Electrical Characteristics. The clock inputs are internally terminated and biased. The input clock 27 www.national.com ADC081500 signal must be capacitively coupled to the clock pins, as indicated in Figure 14. Operation up to the sample rates indicated in the Converter Electrical Characteristics is typically possible if the maximum ambient temperatures indicated are not exceeded. Operating at higher sample rates than indicated for the given ambient temperature may result in reduced device reliability and product lifetime. This is because of the higher power consumption and die temperatures at high sample rates. Important also for reliability is proper thermal management . See Section 2.6.2. ADC081500 Input clock amplitudes above those specified in the Converter Electrical Characteristics may result in increased input offset voltage. This would cause the converter to produce an output code other than the expected 127/128 when both input pins are at the same potential. 2.4.2.2 for best performance, a self calibration should be performed 20 seconds or more after power up and repeated when the operating temperature changes significantly according to the particular system performance requirements. ENOB drops slightly as junction temperature increases and executing a new self calibration cycle will essentially eliminate the change. During a Power-On calibration cycle, both the ADC and the input termination resistor are calibrated. As ENOB changes slightly with junction temperature, an On-Command calibration can be executed to bring the performance of the ADC in line. 2.4 CONTROL PINS Six control pins (without the use of the serial interface) provide a wide range of possibilities in the operation of the ADC081500 and facilitate its use. These control pins provide Full-Scale Input Range setting, Calibration, Calibration Delay, Output Edge Synchronization choice, LVDS Output Level choice and a Power Down feature. 2.4.2.3 Calibration Delay The CalDly input (pin 127) is used to select one of two delay times after the application of power to the start of calibration, as described in Section 1.1.1. The calibration delay values allow the power supply to come up and stabilize before calibration takes place. With no delay or insufficient delay, calibration would begin before the power supply is stabilized at its operating value and result in non-optimal calibration coefficients. If the PD pin is high upon power-up, the calibration delay counter will be disabled until the PD pin is brought low. Therefore, holding the PD pin high during power up will further delay the start of the power-up calibration cycle. The best setting of the CalDly pin depends upon the power-on settling time of the power supply. Note that the calibration delay selection is not possible in the Extended Control mode and the short delay time is used. 2.4.1 Full-Scale Input Range Setting The input full-scale range can be selected with the FSR control input (pin 14) in the normal mode of operation. The is specified as VIN in the Converter Electrical Characteristics. In the extended control mode, the input full-scale range may be programmed using the Full-Scale Adjust Voltage register. See Section 2.2 for more information. 2.4.2 Self Calibration The ADC081500 self-calibration must be run to achieve specified performance. The calibration procedure is run upon power-up and can be run any time on command. The calibration procedure is exactly the same whether there is an input clock present upon power up or if the clock begins some time after application of power. The CalRun output indicator is high while a calibration is in progress. Note that DCLK outputs are not active during a calibration cycle, therefore it is not recommended as a system clock. 2.4.3 Output Edge Synchronization DCLK signals are available to help latch the converter output data into external circuitry. The output data can be synchronized with either edge of these DCLK signals. That is, the output data transition can be set to occur with either the rising edge or the falling edge of the DCLK signal, so that either edge of that DCLK signal can be used to latch the output data into the receiving circuit. When OutEdge (pin 4) is high, the output data is synchronized with (changes with) the rising edge of the DCLK+ (pin 82). When OutEdge is low, the output data is synchronized with the falling edge of DCLK+. At the very high speeds of which the ADC081500 is capable, slight differences in the lengths of the DCLK and data lines can mean the difference between successful and erroneous data capture. The OutEdge pin is used to capture data on the DCLK edge that best suits the application circuit and layout. 2.4.2.1 Power-On Calibration Power-on calibration begins after a time delay following the application of power. This time delay is determined by the setting of CalDly, as described in the Calibration Delay Section, below. The calibration process will be not be performed if the CAL pin is high at power up. In this case, the calibration cycle will not begin until the on-command calibration conditions are met. The ADC081500 will function with the CAL pin held high at power up, but no calibration will be done and performance will be impaired. A manual calibration, however, may be performed after powering up with the CAL pin high. See OnCommand Calibration Section 2.4.2.2. The internal power-on calibration circuitry comes up in an unknown logic state. If the input clock is not running at power up and the power on calibration circuitry is active, it will hold the analog circuitry in power down and the power consumption will typically be less than 200 mW. The power consumption will be normal after the clock starts. 2.4.4 LVDS Output Level Control The output level can be set to one of two levels with OutV (pin3). The strength of the output drivers is greater with OutV high. With OutV low there is less power consumption in the output drivers, but the lower output level means decreased noise immunity. For short LVDS lines and low noise systems, satisfactory performance may be realized with the OutV input low. If the LVDS lines are long and/or the system in which the ADC081500 is used is noisy, it may be necessary to tie the OutV pin high. 2.4.2.2 On-Command Calibration To initiate an on-command calibration, bring the CAL pin high for a minimum of tCAL_H input clock cycles after it has been low for a minimum of tCAL_L input clock cycles. Holding the CAL pin high upon power up will prevent execution of poweron calibration until the CAL pin is low for a minimum of tCAL_L input clock cycles, then brought high for a minimum of another tCAL_H input clock cycles. The calibration cycle will begin tCAL_H input clock cycles after the CAL pin is thus brought high. The CalRun signal should be monitored to determine when the calibration cycle has completed. The minimum tCAL_H and tCAL_L input clock cycle sequences are required to ensure that random noise does not cause a calibration to begin when it is not desired. As mentioned in ?? www.national.com 2.4.5 Power Down Feature The Power Down pin (PD) suspends device operation and puts the ADC081500 into a minimum power dissipation state. See Section 1.1.7 for details on the power down feature. The digital data (+/-) output pins are put into a high impedance state when the PD pin is high. Upon return to normal opera- 28 2.6.1 Supply Voltage The ADC081500 is specified to operate with a supply voltage of 1.9V ±0.1V. It is very important to note that, while this device will function with slightly higher supply voltages, these higher supply voltages may reduce product lifetime. No pin should ever have a voltage on it that is in excess of the supply voltage or below ground by more than 150 mV, not even on a transient basis. This can be a problem upon application of power and power shut-down. Be sure that the supplies to circuits driving any of the input pins, analog or digital, do not come up any faster than does the voltage at the ADC081500 power pins. The Absolute Maximum Ratings should be strictly observed, even during power up and power down. A power supply that produces a voltage spike at turn-on and/or turn-off of power can destroy the ADC081500. The circuit of Figure 15 will provide supply overshoot protection. Many linear regulators will produce output spiking at poweron unless there is a minimum load provided. Active devices draw very little current until their supply voltages reach a few hundred millivolts. The result can be a turn-on spike that can destroy the ADC081500, unless a minimum load is provided for the supply. The 100Ω resistor at the regulator output provides a minimum output current during power-up to ensure there is no turn-on spiking. In the circuit of Figure 15, an LM317 linear regulator is satisfactory if its input supply voltage is 4V to 5V . If a 3.3V supply is used, an LM1086 linear regulator is recommended. 2.5 THE DIGITAL OUTPUTS The ADC081500 demultiplexes the converter output data into two LVDS output buses. The results of successive conversions started on the odd falling edges of the CLK+ pin are available on one of the two LVDS buses, while the results of conversions started on the even falling edges of the CLK+ pin are available on the other LVDS bus. This means that, the word rate at each LVDS bus is 1/2 the ADC081500 input clock rate and the two buses must be multiplexed to obtain the entire 1.5 GSPS conversion result. Since the minimum recommended input clock rate for this device is 200 MHz, the effective data rate can be reduced to as low as 100 MSPS by using the results available on just one of the output buses with a 200 MHz input clock, decimating the 200 MSPS data by two. There is one LVDS output clock pair (DCLK+/-) available for use to latch the LVDS outputs on all buses. Whether the data is sent at the rising or falling edge of DCLK is determined by the sense of the OutEdge pin, as described in Section 2.4.3. DDR (Double Data Rate) clocking can also be used. In this mode a word of data is presented with each edge of DCLK, reducing the DCLK frequency to 1/4 the input clock frequency. See the Timing Diagram section for details. The OutV pin is used to set the LVDS differential output levels. See Section 2.4.4. The output format is Offset Binary. Accordingly, a full-scale input level with VIN+ positive with respect to VIN− will produce an output code of all ones, a full-scale input level with VIN− positive with respect to VIN+ will produce an output code of all zeros and when VIN+ and VIN− are equal, the output code will vary between codes 127 and 128. 20153154 FIGURE 15. Non-Spiking Power Supply The output drivers should have a supply voltage, VDR, that is within the range specified in the Operating Ratings table. This voltage should not exceed the VA supply voltage. If the power is applied to the device without an input clock signal present, the current drawn by the device might be below 200 mA. This is because the ADC081500 gets reset through clocked logic and its initial state is unknown. If the reset logic comes up in the "on" state, it will cause most of the analog circuitry to be powered down, resulting in less than 100 mA of current draw. This current is greater than the power down current because not all of the ADC is powered down. The device current will be normal after the input clock is established. 2.6 POWER CONSIDERATIONS A/D converters draw sufficient transient current to corrupt their own power supplies if not adequately bypassed. A 33 µF capacitor should be placed within an inch (2.5 cm) of the A/D converter power pins. A 0.1 µF capacitor should be placed as close as possible to each VA pin, preferably within one-half centimeter. Leadless chip capacitors are preferred because they have low lead inductance. The VA and VDR supply pins should be isolated from each other to prevent any digital noise from being coupled into the analog portions of the ADC. A ferrite choke, such as the JW Miller FB20009-3B, is recommended between these supply lines when a common source is used for them. As is the case with all high speed converters, the ADC081500 should be assumed to have little power supply noise rejection. Any power supply used for digital circuitry in a system where a lot of digital power is being consumed should not be used to supply power to the ADC081500. The ADC supplies should be the same supply used for other analog circuitry, if not a dedicated supply. 2.6.2 Thermal Management The ADC081500 is capable of impressive speeds and performance at very low power levels for its speed. However, the power consumption is still high enough to require attention to thermal management. For reliability reasons, the die temperature should be kept to a maximum of 130°C. That is, TA (ambient temperature) plus ADC power consumption times θJA (junction to ambient thermal resistance) should not exceed 130°C. This is not a problem if the ambient temperature is kept to a maximum of +85°C as specified in the Operating Ratings section. 29 www.national.com ADC081500 tion, the pipeline will contain meaningless information and must be flushed. If the PD input is brought high while a calibration is running, the device will not go into power down until the calibration sequence is complete. However, if power is applied and PD is already high, the device will not begin the calibration sequence until the PD input goes low. If a manual calibration is requested while the device is powered down, the calibration will not begin at all. That is, the manual calibration input is completely ignored in the power down state. ADC081500 be barrel plated to avoid solder wicking into the vias during the soldering process as this wicking could cause voids in the solder between the package exposed pad and the thermal land on the PCB. Such voids could increase the thermal resistance between the device and the thermal land on the board, which would cause the device to run hotter. If it is desired to monitor die temperature, a temperature sensor may be mounted on the heat sink area of the board near the thermal vias. Allow for a thermal gradient between the temperature sensor and the ADC081500 die of θJ-PAD times typical power consumption = 2.8 x 1.2 = 3.4°C. Allowing for a 5°C temperature drop (including an extra 1.6°C margin) from the die to the temperature sensor, then, would mean that maintaining a maximum pad temperature reading of 125°C will ensure that the die temperature does not exceed 130°C, assuming that the exposed pad of the ADC081500 is properly soldered down and the thermal vias are adequate. (The inaccuracy of the temperature sensor is in addition to the above calculation). As a convenience to the user, the ADC081500 incorporates a thermal diode to aid in temperature measurement. However, this diode has not been characterized and National Semiconductor has no information to provide regarding its characteristics. Hence, no information is available as to the temperature accuracy attainable when using this diode. Please note that the following are general recommendations for mounting exposed pad devices onto a PCB. This should be considered the starting point in PCB and assembly process development. It is recommended that the process be developed based upon past experience in package mounting. The package of the ADC081500 has an exposed pad on its back that provides the primary heat removal path as well as excellent electrical grounding to the printed circuit board. The land pattern design for lead attachment to the PCB should be the same as for a conventional LQFP, but the exposed pad must be attached to the board to remove the maximum amount of heat from the package, as well as to ensure best product parametric performance. To maximize the removal of heat from the package, a thermal land pattern must be incorporated on the PC board within the footprint of the package. The exposed pad of the device must be soldered down to ensure adequate heat conduction out of the package. The land pattern for this exposed pad should be at least as large as the 5 x 5 mm of the exposed pad of the package and be located such that the exposed pad of the device is entirely over that thermal land pattern. This thermal land pattern should be electrically connected to ground. A clearance of at least 0.5 mm should separate this land pattern from the mounting pads for the package pins. 2.7 LAYOUT AND GROUNDING Proper grounding and proper routing of all signals are essential to ensure accurate conversion. A single ground plane should be used, instead of splitting the ground plane into analog and digital areas. Since digital switching transients are composed largely of high frequency components, the skin effect tells us that total ground plane copper weight will have little effect upon the logic-generated noise. Total surface area is more important than is total ground plane volume. Coupling between the typically noisy digital circuitry and the sensitive analog circuitry can lead to poor performance that may seem impossible to isolate and remedy. The solution is to keep the analog circuitry well separated from the digital circuitry. High power digital components should not be located on or near any linear component or power supply trace or plane that services analog or mixed signal components as the resulting common return current path could cause fluctuation in the analog input “ground” return of the ADC, causing excessive noise in the conversion result. Generally, we assume that analog and digital lines should cross each other at 90° to avoid getting digital noise into the analog path. In high frequency systems, however, avoid crossing analog and digital lines altogether. The input clock lines should be isolated from ALL other lines, analog AND digital. The generally accepted 90° crossing should be avoided as even a little coupling can cause problems at high frequencies. Best performance at high frequencies is obtained with a straight signal path. The analog input should be isolated from noisy signal traces to avoid coupling of spurious signals into the input. This is especially important with the low level drive required of the ADC081500. Any external component (e.g., a filter capacitor) connected between the converter's input and ground should be connected to a very clean point in the analog ground plane. All analog circuitry (input amplifiers, filters, etc.) should be separated from any digital components. 20153121 FIGURE 16. Recommended Package Land Pattern Since a large aperture opening may result in poor release, the aperture opening should be subdivided into an array of smaller openings, similar to the land pattern of Figure 16. To minimize junction temperature, it is recommended that a simple heat sink be built into the PCB. This is done by including a copper area of about 2 square inches (6.5 square cm) on the opposite side of the PCB. This copper area may be plated or solder coated to prevent corrosion, but should not have a conformal coating, which could provide some thermal insulation. Thermal vias should be used to connect these top and bottom copper areas. These thermal vias act as "heat pipes" to carry the thermal energy from the device side of the board to the opposite side of the board where it can be more effectively dissipated. The use of 9 to 16 thermal vias is recommended. The thermal vias should be placed on a 1.2 mm grid spacing and have a diameter of 0.30 to 0.33 mm. These vias should www.national.com 2.8 DYNAMIC PERFORMANCE The ADC081500 is a.c. tested and its dynamic performance is guaranteed. To meet the published specifications and avoid jitter-induced noise, the clock source driving the CLK input must exhibit low rms jitter. The allowable jitter is a function of the input frequency and the input signal level, as described in Section 2.3. 30 2.10 COMMON APPLICATION PITFALLS Failure to write all register locations when using extended control mode. When using the serial interface, all 3 user registers must be written at least once with the default or desired values before calibration and subsequent use of the ADC. Once all registers have been written once, other desired settings can be loaded. Driving the inputs (analog or digital) beyond the power supply rails.For device reliability, no input should go more than 150 mV below the ground pins or 150 mV above the supply pins. Exceeding these limits on even a transient basis may not only cause faulty or erratic operation, but may impair device reliability. It is not uncommon for high speed digital circuits to exhibit undershoot that goes more than a volt below ground. Controlling the impedance of high speed lines and terminating these lines in their characteristic impedance should control overshoot. Care should be taken not to overdrive the inputs of the ADC081500. Such practice may lead to conversion inaccuracies and even to device damage. Incorrect analog input common mode voltage in the d.c. coupled mode. As discussed in section 1.3 and 3.0, the Input common mode voltage must remain within 50 mV of the VCMO output, which has a variability with temperature that must also be tracked. Distortion performance will be degraded if the input common mode voltage is more than 50 mV from VCMO . Using an inadequate amplifier to drive the analog input. Use care when choosing a high frequency amplifier to drive the ADC081500 as many high speed amplifiers will have higher distortion than will the ADC081500, resulting in overall system performance degradation. Driving the VBG pin to change the reference voltage. As mentioned in Section 2.1, the reference voltage is intended to be fixed by FSR pin or Full-Scale Voltage Adjust register settings. Over driving this pin will not change the full scale value, but can otherwise upset operation. Driving the clock input with an excessively high level signal. The ADC input clock level should not exceed the level described in the Operating Ratings Table or the input offset could change. Inadequate input clock levels. As described in Section 2.3, insufficient input clock levels can result in poor performance. Excessive input clock levels could result in the introduction of an input offset. Using a clock source with excessive jitter, using an excessively long input clock signal trace, or having other signals coupled to the input clock signal trace. This will cause the sampling interval to vary, causing excessive output noise and a reduction in SNR performance. Failure to provide adequate heat removal. As described in Section 2.6.2, it is important to provide adequate heat removal to ensure device reliability. This can be done either with adequate air flow or the use of a simple heat sink built into the board. The backside pad should be grounded for best performance. 2.9 USING THE SERIAL INTERFACE The ADC081500 may be operated in the Normal control mode (using control pins) or in the Extended control mode (using a serial interface and register set). Table 6 and Table 7 describe the functions of pins 3, 4, 14 and 127 in the Normal control mode and the Extended control mode, respectively. 2.9.1 Normal Control Mode Operation Normal control mode operation means that the Serial Interface is not active and all controllable functions are controlled with various pin settings. That is, the full-scale range, the power on calibration delay, the output voltage and the input coupling (a.c. or d.c.). The Normal control mode is used by setting pin 14 high or low, as opposed to letting it float. Table 6 indicates the pin functions of the ADC081500 in the Normal control mode. TABLE 6. Normal Control Mode Operation (Pin 14 High or Low) Pin Low High Floating 3 Reduced VOD Normal VOD n/a 4 OutEdge = Neg OutEdge = Pos DDR 127 14 CalDly Short CalDly Long Reduced VIN Normal VIN n/a Extended Control Mode Pin 3 can be either high or low in the Normal control mode. Pin 14 must not be left floating to select this mode. See Section 1.2 for more information. Pin 4 can be high or low or can be left floating in the Normal control mode. In the Normal control mode, pin 4 high or low defines the edge at which the output data transitions. See Section 2.4.3 for more information. If this pin is floating, the output clock (DCLK) is a DDR (Double Data Rate) clock (see Section 1.1.5.3) and the output edge synchronization is irrelevant since data is clocked out on both DCLK edges. Pin 127 in the non-extended control mode sets the calibration delay. Pin 127 is not designed to remain floating. TABLE 7. Extended Control Mode Operation (Pin 14 Floating) Pin Function 3 SCLK (Serial Clock) 4 SDATA (Serial Data) 127 SCS (Serial Interface Chip Select) 31 www.national.com ADC081500 It is good practice to keep the ADC input clock line as short as possible, to keep it well away from any other signals and to treat it as a transmission line. Other signals can introduce jitter into the input clock signal. The clock signal can also introduce noise into the analog path if not isolated from that path. Best dynamic performance is obtained when the exposed pad at the back of the package has a good connection to ground. This is because this path from the die to ground is a lower impedance than offered by the package pins. ADC081500 Physical Dimensions inches (millimeters) unless otherwise noted NOTES: UNLESS OTHERWISE SPECIFIED REFERENCE JEDEC REGISTRATION MS-026, VARIATION BFB. 128-Lead Exposed Pad LQFP Order Number ADC081500CIYB NS Package Number VNX128A www.national.com 32 ADC081500 Notes 33 www.national.com ADC081500 High Performance, Low Power, 8-Bit, 1.5 GSPS A/D Converter Notes For more National Semiconductor product information and proven design tools, visit the following Web sites at: Products Design Support Amplifiers www.national.com/amplifiers WEBENCH® Tools www.national.com/webench Audio www.national.com/audio App Notes www.national.com/appnotes Clock and Timing www.national.com/timing Reference Designs www.national.com/refdesigns Data Converters www.national.com/adc Samples www.national.com/samples Interface www.national.com/interface Eval Boards www.national.com/evalboards LVDS www.national.com/lvds Packaging www.national.com/packaging Power Management www.national.com/power Green Compliance www.national.com/quality/green Switching Regulators www.national.com/switchers Distributors www.national.com/contacts LDOs www.national.com/ldo Quality and Reliability www.national.com/quality LED Lighting www.national.com/led Feedback/Support www.national.com/feedback Voltage Reference www.national.com/vref Design Made Easy www.national.com/easy PowerWise® Solutions www.national.com/powerwise Solutions www.national.com/solutions Serial Digital Interface (SDI) www.national.com/sdi Mil/Aero www.national.com/milaero Temperature Sensors www.national.com/tempsensors SolarMagic™ www.national.com/solarmagic Wireless (PLL/VCO) www.national.com/wireless Analog University® www.national.com/AU THE CONTENTS OF THIS DOCUMENT ARE PROVIDED IN CONNECTION WITH NATIONAL SEMICONDUCTOR CORPORATION (“NATIONAL”) PRODUCTS. NATIONAL MAKES NO REPRESENTATIONS OR WARRANTIES WITH RESPECT TO THE ACCURACY OR COMPLETENESS OF THE CONTENTS OF THIS PUBLICATION AND RESERVES THE RIGHT TO MAKE CHANGES TO SPECIFICATIONS AND PRODUCT DESCRIPTIONS AT ANY TIME WITHOUT NOTICE. NO LICENSE, WHETHER EXPRESS, IMPLIED, ARISING BY ESTOPPEL OR OTHERWISE, TO ANY INTELLECTUAL PROPERTY RIGHTS IS GRANTED BY THIS DOCUMENT. TESTING AND OTHER QUALITY CONTROLS ARE USED TO THE EXTENT NATIONAL DEEMS NECESSARY TO SUPPORT NATIONAL’S PRODUCT WARRANTY. EXCEPT WHERE MANDATED BY GOVERNMENT REQUIREMENTS, TESTING OF ALL PARAMETERS OF EACH PRODUCT IS NOT NECESSARILY PERFORMED. NATIONAL ASSUMES NO LIABILITY FOR APPLICATIONS ASSISTANCE OR BUYER PRODUCT DESIGN. BUYERS ARE RESPONSIBLE FOR THEIR PRODUCTS AND APPLICATIONS USING NATIONAL COMPONENTS. PRIOR TO USING OR DISTRIBUTING ANY PRODUCTS THAT INCLUDE NATIONAL COMPONENTS, BUYERS SHOULD PROVIDE ADEQUATE DESIGN, TESTING AND OPERATING SAFEGUARDS. EXCEPT AS PROVIDED IN NATIONAL’S TERMS AND CONDITIONS OF SALE FOR SUCH PRODUCTS, NATIONAL ASSUMES NO LIABILITY WHATSOEVER, AND NATIONAL DISCLAIMS ANY EXPRESS OR IMPLIED WARRANTY RELATING TO THE SALE AND/OR USE OF NATIONAL PRODUCTS INCLUDING LIABILITY OR WARRANTIES RELATING TO FITNESS FOR A PARTICULAR PURPOSE, MERCHANTABILITY, OR INFRINGEMENT OF ANY PATENT, COPYRIGHT OR OTHER INTELLECTUAL PROPERTY RIGHT. LIFE SUPPORT POLICY NATIONAL’S PRODUCTS ARE NOT AUTHORIZED FOR USE AS CRITICAL COMPONENTS IN LIFE SUPPORT DEVICES OR SYSTEMS WITHOUT THE EXPRESS PRIOR WRITTEN APPROVAL OF THE CHIEF EXECUTIVE OFFICER AND GENERAL COUNSEL OF NATIONAL SEMICONDUCTOR CORPORATION. As used herein: Life support devices or systems are devices which (a) are intended for surgical implant into the body, or (b) support or sustain life and whose failure to perform when properly used in accordance with instructions for use provided in the labeling can be reasonably expected to result in a significant injury to the user. A critical component is any component in a life support device or system whose failure to perform can be reasonably expected to cause the failure of the life support device or system or to affect its safety or effectiveness. National Semiconductor and the National Semiconductor logo are registered trademarks of National Semiconductor Corporation. All other brand or product names may be trademarks or registered trademarks of their respective holders. Copyright© 2009 National Semiconductor Corporation For the most current product information visit us at www.national.com National Semiconductor Americas Technical Support Center Email: [email protected] Tel: 1-800-272-9959 www.national.com National Semiconductor Europe Technical Support Center Email: [email protected] National Semiconductor Asia Pacific Technical Support Center Email: [email protected] National Semiconductor Japan Technical Support Center Email: [email protected]