19-3432; Rev 0; 11/04 12-Bit, 125Msps ADC with CMOS Outputs for Wideband Applications Applications Base-Station Power Amplifier Linearization Cable Head-End Receivers Wireless and Wired Broadband Communication Communications Test Equipment Radar and Satellite Subsystems ♦ ♦ ♦ ♦ ♦ ♦ ♦ ♦ ♦ ♦ ♦ 125Msps Conversion Rate SNR = 65dB, fIN = 100MHz at 125Msps SFDR = 77dBc, fIN = 100MHz at 125Msps ±0.7 LSB INL, ±0.25 DNL (typ) 861mW Power Dissipation at 125Msps On-Chip Selectable Divide-by-2 Clock Input Parallel or Demux Parallel Digital CMOS Outputs Reset Option for Synchronizing Multiple ADCs Data Clock Output Offset Binary or Two’s-Complement Output Evaluation Kit Available (MAX19541EVKIT) Ordering Information PART TEMP RANGE PINPACKAGE PKG CODE MAX19541EGK -40°C to +85°C 68 QFN-EP* G6800-4 EP = Exposed paddle. 63 62 61 60 59 58 DA6 DA5 DA8 DA9 DA10 DA11 ORA OVCC OGND 67 66 65 64 AVCC 68 AVCC AGND AVCC ITL TOP VIEW AGND Pin Configuration T/B The MAX19541 monolithic 12-bit, 125Msps analog-todigital converter (ADC) is optimized for outstanding dynamic performance at high-IF frequencies of 300MHz and beyond. This device operates with conversion rates up to 125Msps while consuming only 861mW. At 125Msps and an input frequency of 240MHz, the MAX19541 achieves a spurious-free dynamic range (SFDR) of 71.5dBc. The MAX19541 features an excellent signal-to-noise ratio (SNR) of 65.4dB at 10MHz that remains flat (within 3dB) for input tones up to 250MHz. This makes the MAX19541 ideal for wideband applications such as power-amplifier predistortion in cellular base-station transceiver systems. The MAX19541 operates in either parallel mode where the data outputs appear on a single parallel port at the sampling rate, or in demux parallel mode, where the outputs appear on two separate parallel ports at one-half the sampling rate. See the Mode of Operation section. The MAX19541 operates on a single 1.8V supply. The analog input is differential and can be AC- or DC-coupled. The ADC also features a selectable on-chip divide-by-2 clock circuit that allows clock frequencies as high as 250MHz. This helps to reduce the phase noise of the input clock source, allowing for higher dynamic performance. For best performance, a differential LVPECL sampling clock is recommended. The digital outputs are CMOS compatible and the data format can be selected to be either two’s complement or offset binary. A pin-compatible, 12-bit, 170Msps version of the MAX19541 is also available. Refer to the MAX19542 data sheet for more information. The MAX19541 is available in a 68-pin QFN with exposed paddle (EP) and is specified over the extended (-40°C to +85°C) temperature range. Features 57 56 55 54 53 52 AVCC 1 51 DA4 AGND 2 50 DA3 REFIO 3 49 DA2 REFADJ 4 48 DA1 AGND 5 47 DA0 AVCC 6 46 ORB AGND 7 45 OGND INP 8 44 OVCC INN 9 43 DCLKP MAX19541 AGND 10 42 DCLKN AVCC 11 41 OVCC AVCC 12 40 DB11 AVCC 13 39 DB10 AVCC 14 38 DB9 RESET 15 37 DB8 DEMUX 16 36 DB7 CLKDIV 17 35 DB6 DB5 DB4 DB3 DB2 DB1 DB0 OVCC OVCC AVCC NOTE: EXPOSED PADDLE CONNECTED TO AGND. OGND AGND CLKP CLKN AGND AVCC AGND AGND 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 QFN ________________________________________________________________ Maxim Integrated Products For pricing, delivery, and ordering information, please contact Maxim/Dallas Direct! at 1-888-629-4642, or visit Maxim’s website at www.maxim-ic.com. 1 MAX19541 General Description DA7 KIT ATION EVALU E L B AVAILA MAX19541 12-Bit, 125Msps ADC with CMOS Outputs for Wideband Applications ABSOLUTE MAXIMUM RATINGS Maximum Current into Any Pin .........................................±50mA ESD on All Pins (Human Body Model).............................±2000V Continuous Power Dissipation (TA = +70°C) 68-Pin QFN (derate 41.7mW/°C above +70°C) .........3333mW Operating Temperature Range ...........................-40°C to +85°C Junction Temperature ......................................................+150°C Storage Temperature Range .............................-60°C to +150°C Lead Temperature (soldering, 10s) .................................+300°C AVCC to AGND ......................................................-0.3V to +2.1V OVCC to OGND .....................................................-0.3V to +2.1V AVCC to OVCC .......................................................-0.3V to +2.1V AGND to OGND ....................................................-0.3V to +0.3V Analog Inputs (INP, INN) to AGND ..........-0.3V to (AVCC + 0.3V) All Digital Inputs to AGND........................-0.3V to (AVCC + 0.3V) REFIO, REFADJ to AGND ........................-0.3V to (AVCC + 0.3V) All Digital Outputs to OGND ....................-0.3V to (OVCC + 0.3V) Stresses beyond those listed under “Absolute Maximum Ratings” may cause permanent damage to the device. These are stress ratings only, and functional operation of the device at these or any other conditions beyond those indicated in the operational sections of the specifications is not implied. Exposure to absolute maximum rating conditions for extended periods may affect device reliability. ELECTRICAL CHARACTERISTICS (AVCC = OVCC = 1.8V, AGND = OGND = 0, fSAMPLE = 125MHz, DEMUX = 0, differential LVPECL clock input drive, 0.1µF capacitor on REFIO, internal reference, TA = TMIN to TMAX, unless otherwise noted. TA ≥ +25°C guaranteed by production test, TA < +25°C guaranteed by design and characterization. Typical values are at TA = +25°C.) PARAMETER SYMBOL CONDITIONS MIN TYP MAX UNITS DC ACCURACY Resolution 12 Bits Integral Nonlinearity INL fIN = 10MHz (Note 1) -2.5 ±0.7 +2.5 LSB Differential Nonlinearity DNL fIN = 10MHz, no missing codes (Note 1) -0.75 ±0.25 +0.75 LSB Transfer Curve Offset VOS (Note 1) +3 mV -3 Offset Temperature Drift 40 mV/°C ANALOG INPUTS (INP, INN) Full-Scale Input Voltage Range VFS (Note 1) 1300 Full-Scale Range Temperature Drift Common-Mode Input Range VCM Input Capacitance CIN Differential Input Resistance RIN Full-Power Analog Bandwidth 1410 1510 130 ppm/°C 1.365 ±0.15 V 3 3.00 FPBW mVP-P 4.3 pF 6.25 900 kΩ MHz REFERENCE (REFIO, REFADJ) Reference Output Voltage VREFIO 1.22 Reference Temperature Drift REFADJ Input High Voltage 1.245 90 VREFADJ Used to disable the internal reference 1.27 V ppm/°C AVCC 0.3 V 125 MHz SAMPLING CHARACTERISTICS Maximum Sampling Rate fSAMPLE Minimum Sampling Rate fSAMPLE Clock Duty Cycle Aperture Delay tAD Aperture Jitter tAJ 2 20 Set by clock-management circuit Figure 4 MHz 40 to 60 % 620 ps 0.2 psRMS _______________________________________________________________________________________ 12-Bit, 125Msps ADC with CMOS Outputs for Wideband Applications (AVCC = OVCC = 1.8V, AGND = OGND = 0, fSAMPLE = 125MHz, DEMUX = 0, differential LVPECL clock input drive, 0.1µF capacitor on REFIO, internal reference, TA = TMIN to TMAX, unless otherwise noted. TA ≥ +25°C guaranteed by production test, TA < +25°C guaranteed by design and characterization. Typical values are at TA = +25°C.) PARAMETER SYMBOL CONDITIONS MIN TYP MAX UNITS 200 500 mVP-P 1.15 ±0.25 V CLOCK INPUTS (CLKP, CLKN) Differential Clock Input Amplitude (Note 2) Clock Input Common-Mode Voltage Range Clock Differential Input Resistance RCLK 11 ±25% kΩ Clock Differential Input Capacitance CCLK 5 pF DYNAMIC CHARACTERISTICS (at -2dBFS) Signal-to-Noise Ratio SNR fIN = 10MHz 63.7 65.4 fIN = 100MHz 63.3 65 fIN = 180MHz fIN = 240MHz Signal-to-Noise and Distortion SINAD 63.4 fIN = 10MHz 63.1 65.2 fIN = 100MHz 62.5 64.2 fIN = 180MHz SFDR Worst Harmonics (HD2 or HD3) Two-Tone Intermodulation Distortion IMD100 dB 63.4 fIN = 240MHz Spurious-Free Dynamic Range dB 64.1 62.7 fIN = 10MHz 72 82 fIN = 100MHz 70.5 77 fIN = 180MHz dBc 75 fIN = 240MHz 71.5 fIN = 10MHz -88.7 -72 fIN = 100MHz -73.1 -70.5 fIN = 180MHz -72.8 fIN = 240MHz -71.5 fIN1 = 150MHz at -7dBFS, fIN2 = 153MHz at -7dBFS, fSAMPLE = 125MHz -75 dBc dBc CMOS DIGITAL OUTPUTS (DA0–DA11, DB0–DB11, ORA, ORB) Logic-High Output Voltage VOH Logic-Low Output Voltage VOL OVCC 0.1 V 0.1 V 0.2 x AVCC V LVCMOS DIGITAL INPUTS (CLKDIV, T/B, DEMUX, ITL) Digital Input-Voltage Low VIL Digital Input-Voltage High VIH 0.8 x AVCC V _______________________________________________________________________________________ 3 MAX19541 ELECTRICAL CHARACTERISTICS (continued) MAX19541 12-Bit, 125Msps ADC with CMOS Outputs for Wideband Applications ELECTRICAL CHARACTERISTICS (continued) (AVCC = OVCC = 1.8V, AGND = OGND = 0, fSAMPLE = 125MHz, DEMUX = 0, differential LVPECL clock input drive, 0.1µF capacitor on REFIO, internal reference, TA = TMIN to TMAX, unless otherwise noted. TA ≥ +25°C guaranteed by production test, TA < +25°C guaranteed by design and characterization. Typical values are at TA = +25°C.) PARAMETER SYMBOL CONDITIONS MIN TYP MAX UNITS Input Resistance RIN 46.5 kΩ Input Capacitance CIN 5 pF TIMING CHARACTERISTICS CLKP-to-DA0–DA11 Propagation Delay tPDL Figures 5, 6, 7 2.5 ns CLK-to-DCLKP Propagation Delay tCPDL Figures 5, 6, 7 2.1 ns DCLKP Rising Edge to DA0–DA11 tPDL tCPDL Figures 5, 6, 7 (Note 2) CMOS Output Rise Time tRISE 20% to 80%, CL = 5pF 1 ns CMOS Output Fall Time tFALL 20% to 80%, CL = 5pF 1 ns 180 400 710 ns RESET Hold tHR Figure 4 100 ps RESET Setup tSR Figure 4 500 ps tLATENCY Figure 4 11 Clock cycles Output Data Pipeline Delay POWER REQUIREMENTS Analog Supply Voltage Range AVCC 1.7 1.7 1.8 1.9 V Digital Supply Voltage Range OVCC Analog Supply Current IAVCC fIN = 100MHz 1.8 1.9 V 460 500 mA Digital Supply Current IOVCC fIN = 100MHz 18 25 mA Analog Power Dissipation PDISS fIN = 100MHz 861 945 mW Power-Supply Rejection Ratio PSRR Offset (Note 3) 1.8 mV/V Gain (Note 3) 1.5 %FS/V Note 1: Static linearity and offset parameters are computed from a straight line drawn between the end points of the code transition transfer function. The full-scale range (FSR) is defined as 4096 x slope of the line. Note 2: Parameter guaranteed by design and characterization; TA = TMIN to TMAX. Note 3: PSRR is measured with both analog and digital supplies connected to the same potential. 4 _______________________________________________________________________________________ 12-Bit, 125Msps ADC with CMOS Outputs for Wideband Applications -40 -50 -60 3 -70 -80 -30 5 -50 -60 3 -70 2 4 6 5 -50 -60 -100 -100 -100 -110 -110 20 30 40 50 60 3 -70 -90 2 4 5 6 7 -110 0 10 20 30 40 50 0 60 10 20 30 40 50 60 ANALOG INPUT FREQUENCY (MHz) ANALOG INPUT FREQUENCY (MHz) ANALOG INPUT FREQUENCY (MHz) FFT PLOT (16,384-POINT DATA RECORD) SNR/SINAD vs. ANALOG INPUT FREQUENCY (fSAMPLE = 125.0043MHz, AIN = -1dBFS) SFDR vs. ANALOG INPUT FREQUENCY (fSAMPLE = 125.0043MHz, AIN = -1dBFS) -30 -40 -50 2 -60 3 -70 4 -80 5 7 6 MAX19541 toc06 SNR 67 90 85 80 75 SFDR (dBc) -20 70 MAX19541 toc05 fIN = 240.0977201MHz fSAMPLE = 125.0043232MHz AIN = -1.001dBFS SNR = 63.96dB SINAD = 62.798dB SFDR = 70.011dBc HD2 = -70.011dBc HD3 = -78.367dBc SNR/SINAD (dB) 0 -10 64 SINAD 61 70 65 60 55 58 -90 50 45 -100 40 55 -110 0 10 20 30 40 50 60 0 50 100 150 200 250 0 50 100 150 200 250 ANALOG INPUT FREQUENCY (MHz) ANALOG INPUT FREQUENCY (MHz) ANALOG INPUT FREQUENCY (MHz) HD2/HD3 vs. ANALOG INPUT FREQUENCY (fSAMPLE = 125.0043MHz, AIN = -1dBFS) THD vs. ANALOG INPUT FREQUENCY (fSAMPLE = 125.0043MHz, AIN = -1dBFS) SNR/SINAD vs. ANALOG INPUT AMPLITUDE (fSAMPLE = 125.0043MHz, fIN = 60.0225MHz) 68 -65 MAX19541 toc09 HD3 -70 -60 MAX19541 toc08 -65 MAX19541 toc07 -60 SNR 62 THD (dBc) HD2 -80 -85 -90 SNR/SINAD (dB) -70 -75 HD2/HD3 (dBc) -40 -80 7 MAX19541 toc03 -20 -90 10 fIN = 183.48565448MHz fSAMPLE = 125.0043232MHz AIN = -0.982dBFS SNR = 64.812dB SINAD = 63.424dB SFDR = 70.001dBc HD2 = -79.233dBc HD3 = -70.001dBc -30 -80 7 2 -40 0 -10 -90 0 AMPLITUDE (dB) 6 4 fIN = 60.0225226MHz fSAMPLE = 125.0043232MHz AIN = -0.985dBFS SNR = 66.359dB SINAD = 65.581dB SFDR = 74.285dBc HD2 = -87.016dBc HD3 = -74.285dBc -20 MAX19541 toc04 AMPLITUDE (dB) -30 FFT PLOT (16,384-POINT DATA RECORD) AMPLITUDE (dB) -20 0 -10 AMPLITUDE (dB) fIN = 11.5284138MHz fSAMPLE = 125.0043232MHz AIN = -0.959dBFS SNR = 66.416dB SINAD = 66.211dB SFDR = 83.199dBc HD2 = -95.203dBc HD3 = -83.199dBc -10 MAX19541 toc01 0 FFT PLOT (16,384-POINT DATA RECORD) MAX19541 toc02 FFT PLOT (16,384-POINT DATA RECORD) -75 -80 -85 -95 56 SINAD 50 44 -90 -100 38 -95 -105 -110 -100 0 50 100 150 200 ANALOG INPUT FREQUENCY (MHz) 250 32 0 50 100 150 200 ANALOG INPUT FREQUENCY (MHz) 250 -30 -25 -20 -15 -10 -5 0 ANALOG INPUT AMPLITUDE (dBFS) _______________________________________________________________________________________ 5 MAX19541 Typical Operating Characteristics (AVCC = OVCC = 1.8V, AGND = OGND = 0, fSAMPLE = 125MHz, AIN = -1dBFS; see TOCs for detailed information on test conditions, differential input drive, differential LVPECL clock input drive, 0.1µF capacitor on REFIO, internal reference, digital outputs differential RL = 100Ω, TA = +25°C.) Typical Operating Characteristics (continued) (AVCC = OVCC = 1.8V, AGND = OGND = 0, fSAMPLE = 125MHz, AIN = -1dBFS; see TOCs for detailed information on test conditions, differential input drive, differential LVPECL clock input drive, 0.1µF capacitor on REFIO, internal reference, digital outputs differential RL = 100Ω, TA = +25°C.) -60 70 65 60 55 50 -25 -20 -15 -10 -5 -65 -70 -75 HD2 -80 -75 -80 -85 -85 -90 -90 -95 -95 0 -100 -30 -25 -20 -15 -10 -5 0 -30 -20 -15 -10 -5 SNR/SINAD vs. fSAMPLE (fIN = 60.0225MHz. AIN = -1dBFS) SFDR vs. fSAMPLE (fIN = 60.0225MHz, AIN = -1dBFS) HD2/HD3 vs. fSAMPLE (fIN = 65.0225MHz, AIN = -1dBFS) 85 66 -50 -55 -60 -65 MAX19541 toc14 MAX19541 toc13 90 SFDR (dBc) 65 SINAD 64 63 HD2/HD3 (dBc) 80 75 70 65 62 60 61 55 60 60 80 100 120 140 160 180 200 HD3 -70 -75 -80 -85 -90 -95 -100 -105 -110 50 40 20 40 60 80 100 120 140 160 180 200 HD2 20 40 60 80 100 120 140 160 180 200 fSAMPLE (MHz) fSAMPLE (MHz) fSAMPLE (MHz) THD vs. fSAMPLE (fIN = 60.0225MHz, AIN = -1dBFS) TWO-TONE IMD PLOT (16,384-POINT DATA RECORD) INL vs. DIGITAL OUTPUT CODE (512k-POINT DATA RECORD) -20 fIN2 fIN1 = 150.0067138MHz fIN2 = 152.9822805MHz fSAMPLE = 125.0043232MHz AIN1 = AIN2 = -7dBFS IMD = - 75dBc AMPLITUDE (dB) -30 -70 -75 -80 -40 -50 -60 2fIN1 - fIN2 -70 2fIN2 - fIN1 0.6 0.4 0.2 0 -0.2 -80 -0.4 -85 -90 -0.6 -100 -0.8 -90 -110 40 60 80 100 120 140 160 180 200 fSAMPLE (MHz) fIN = 13.0390862MHz 0.8 INL (LSB) -65 -10 fIN1 1.0 MAX19541toc17 0 MAX19541 toc16 -60 0 MAX19541 toc15 ANALOG INPUT AMPLITUDE (dBFS) SNR 20 -25 ANALOG INPUT AMPLITUDE (dBFS) 67 20 -70 ANALOG INPUT AMPLITUDE (dBFS) 68 SNR/SINAD (dB) -60 -100 -30 6 -55 HD3 -65 HD2/HD3 (dBc) SFDR (dBc) 75 -50 MAX19541 toc12 -55 THD (dBc) 80 THD vs. ANALOG INPUT AMPLITUDE (fSAMPLE = 125.0043MHz, fIN = 60.0225MHz) MAX19541 toc11 -50 MAX19541 toc10 85 HD2/HD3 vs. ANALOG INPUT AMPLITUDE (fSAMPLE = 125.0043MHz, fIN = 60.0225MHz) MAX19541 toc18 SFDR vs. ANALOG INPUT AMPLITUDE (fSAMPLE = 125.0043MHz, fIN = 60.0225MHz) THD (dBc) MAX19541 12-Bit, 125Msps ADC with CMOS Outputs for Wideband Applications -1.0 0 10 20 30 40 50 ANALOG INPUT FREQUENCY (MHz) 60 0 512 1024 1536 2048 2560 3072 3584 4096 DIGITAL OUTPUT CODE _______________________________________________________________________________________ 12-Bit, 125Msps ADC with CMOS Outputs for Wideband Applications GAIN BANDWIDTH PLOT (fSAMPLE = 125.0043232MHz, AIN = -1dBFS) 0.6 1 0 SNR/SINAD (dB) -2 0.2 GAIN (dB) 0 -0.2 -0.4 -3 -4 SNR 65 64 SINAD 63 -5 62 -6 -0.8 -7 -1.0 61 1 512 1024 1536 2048 2560 3072 3584 4096 10 1000 100 -40 -15 10 35 60 DIGITAL OUTPUT CODE ANALOG INPUT FREQUENCY (MHz) TEMPERATURE (°C) SFDR vs. TEMPERATURE (fSAMPLE = 125MHz, AIN = -2dBFS) TOTAL POWER DISSIPATION vs. fSAMPLE (fIN = 60MHz, AIN = -1dBFS) FULL-SCALE ADJUSTMENT RANGE vs. ADJUSTMENT RESISTANCE 76 0.92 0.90 0.88 75 PDISS (W) 74 73 72 0.86 0.84 0.82 71 0.80 70 1.34 1.32 IFULL-SCALE VOLTAGE (V) fIN = 100MHz 77 MAX19541 toc23 78 MAX19541 toc22 0 1.30 RADJ BETWEEN REFADJ AND REFIO 1.28 85 MAX19541 toc24 -0.6 1.26 1.24 1.22 1.20 RADJ BETWEEN REFADJ AND GND 1.18 1.16 69 0.78 68 0.76 -15 10 35 60 1.12 20 85 TEMPERATURE (°C) 40 60 RADJ (kΩ) fSAMPLE (MHz) AVCC = OVCC SNR 66 INTERNAL REFERENCE vs. SUPPLY VOLTAGE 1.246 MAX19541 toc26 68 1.245 1.244 VREFIO (V) SINAD 64 0 100 200 300 400 500 600 700 800 900 1000 80 100 120 140 160 180 200 SNR/SINAD vs. SUPPLY VOLTAGE (fIN = 60.0225MHz, AIN = -1dBFS) SNR/SINAD (dB) -40 1.14 MAX19541 toc25 DNL (LSB) fIN = 100MHz 66 -1 0.4 SFDR (dBc) 67 MAX19541 toc20 fIN = 13.0390862MHz 0.8 MAX19541 toc19 1.0 SNR/SINAD vs. TEMPERATURE (fSAMPLE = 125MHz, AIN = -2dBFS) MAX19541 toc21 DNL vs. DIGITAL OUTPUT CODE (524k-POINT DATA RECORD) 62 1.243 1.242 60 1.241 58 1.240 1.6 1.7 1.8 1.9 SUPPLY VOLTAGE (V) 2.0 2.1 1.6 1.7 1.8 1.9 2.0 2.1 SUPPLY VOLTAGE (V) _______________________________________________________________________________________ 7 MAX19541 Typical Operating Characteristics (continued) (AVCC = OVCC = 1.8V, AGND = OGND = 0, fSAMPLE = 125MHz, AIN = -1dBFS; see TOCs for detailed information on test conditions, differential input drive, differential LVPECL clock input drive, 0.1µF capacitor on REFIO, internal reference, digital outputs differential RL = 100Ω, TA = +25°C.) MAX19541 12-Bit, 125Msps ADC with CMOS Outputs for Wideband Applications Pin Description PIN NAME 1, 6, 11–14, 20, 25, 62, 63, 65 AVCC Analog Supply Voltage. Bypass each AVCC pin with a 0.1µF capacitor for best decoupling results. Additional board decoupling might be required. See the Grounding, Bypassing, and Layout Considerations section. 2, 5, 7, 10, 18, 19, 21, 24, 64, 66 AGND Analog Converter Ground. Connect the converter’s exposed paddle (EP) to AGND. 3 REFIO Reference Input/Output. Drive REFADJ high to allow an external reference source to be connected to the MAX19541. Drive REFADJ low to activate the internal 1.23V bandgap reference. Connect a 0.1µF capacitor from REFIO to AGND. 8 INP Reference Adjust Input. REFADJ allows for full-scale range adjustments by placing a resistor or trim potentiometer between REFADJ and AGND (decreases FS range) or REFADJ and REFIO (increases FS range). If REFADJ is connected to AVCC, the internal reference can be overdriven with an external source connected to REFIO. If REFADJ is connected to AGND, the internal reference is used to determine the full-scale range of the data converter. Positive Analog Input Terminal 9 INN Negative Analog Input Terminal 15 RESET Active-High RESET Input. RESET controls the latency of the MAX19541. RESET has an internal pulldown resistor. See the Reset Operation section. 16 DEMUX Output-Mode-Select Input. Drive DEMUX low for the parallel output mode (full-rate CMOS outputs on A ports only). Drive DEMUX high for the demux parallel or demux interleaved modes (half-rate outputs on both ports A and B) depending on the state of the ITL input. See the Modes of Operation section. 17 CLKDIV Clock-Divider Input. CLKDIV is an LVCMOS-compatible input that controls the sampling frequency relative to the input clock frequency. CLKDIV has an internal pulldown resistor: CLKDIV = 0: sampling frequency is 1/2 the input clock frequency. CLKDIV = 1: sampling frequency is equal to the input clock frequency. 22 CLKN Complementary Clock Input. CLKN ideally requires an LVPECL-compatible input level to maintain the converter’s excellent performance. 23 CLKP True Clock Input. CLKP ideally requires an LVPECL-compatible input level to maintain the converter’s excellent performance. 26, 45, 61 OGND Digital Converter Ground. Ground connection for digital circuitry and output drivers. 27, 28, 41, 44, 60 OVCC Digital Supply Voltage. Bypass OVCC with a 0.1µF capacitor for best decoupling results. Additional board decoupling might be required. See the Grounding, Bypassing, and Layout Considerations section. 29 DB0 Port B CMOS Digital Output Bit 0 (LSB) 30 DB1 Port B CMOS Digital Output Bit 1 31 DB2 Port B CMOS Digital Output Bit 2 32 DB3 Port B CMOS Digital Output Bit 3 33 DB4 Port B CMOS Digital Output Bit 4 34 DB5 Port B CMOS Digital Output Bit 5 35 DB6 Port B CMOS Digital Output Bit 6 36 DB7 Port B CMOS Digital Output Bit 7 37 DB8 Port B CMOS Digital Output Bit 8 4 8 FUNCTION REFADJ _______________________________________________________________________________________ 12-Bit, 125Msps ADC with CMOS Outputs for Wideband Applications PIN NAME FUNCTION 38 DB9 Port B CMOS Digital Output Bit 9 39 DB10 Port B CMOS Digital Output Bit 10 40 DB11 Port B CMOS Digital Output Bit 11 (MSB) 42 DCLKN Inverted CMOS Digital Clock Output. DCLKN provides a CMOS-compatible output level and can be used to synchronize external devices to the converter clock. When DEMUX is high, the frequency at DCLKN is half the sampling clock’s frequency. 43 DCLKP True CMOS Digital Clock Output. DCLKP provides a CMOS-compatible output level and can be used to synchronize external devices to the converter clock. When DEMUX is high, the frequency at DCLKP is half the sampling clock’s frequency. 46 ORB Port B CMOS Digital Output Overrange 47 DA0 Port A CMOS Digital Output Bit 0 (LSB) 48 DA1 Port A CMOS Digital Output Bit 1 49 DA2 Port A CMOS Digital Output Bit 2 50 DA3 Port A CMOS Digital Output Bit 3 51 DA4 Port A CMOS Digital Output Bit 4 52 DA5 Port A CMOS Digital Output Bit 5 53 DA6 Port A CMOS Digital Output Bit 6 54 DA7 Port A CMOS Digital Output Bit 7 55 DA8 Port A CMOS Digital Output Bit 8 56 DA9 Port A CMOS Digital Output Bit 9 57 DA10 Port A CMOS Digital Output Bit 10 58 DA11 Port A CMOS Digital Output Bit 11 (MSB) 59 ORA Port A CMOS Digital Output Overrange 67 ITL Interleaved/Parallel-Select Input. Drive ITL low for the demux parallel mode. Drive ITL high for the demux interleaved mode. 68 T/B Output-Format-Select Input. T/B is an LVCMOS-compatible input that controls the digital output format of the MAX19541. T/B has an internal pulldown resistor: T/B = 1: binary output format. T/B = 0: two’s-complement output format. EP AGND Exposed Paddle. Connect EP to the analog ground (AGND) for optimum performance. The exposed paddle is located on the backside of the chip. EP is internally connected to the die substrate. _______________________________________________________________________________________ 9 MAX19541 Pin Description (continued) MAX19541 12-Bit, 125Msps ADC with CMOS Outputs for Wideband Applications Detailed Description— Theory of Operation The MAX19541 uses a fully differential, pipelined architecture that allows for high-speed conversion, optimized accuracy and linearity, while minimizing power consumption. Both positive (INP) and negative/complementary analog input terminals (INN) are centered around a 1.365V common-mode voltage, and accept a ±350mV differential analog input voltage swing each, resulting in a typical 1.41VP-P differential full-scale signal swing. Inputs INP and INN are buffered prior to entering each track-and-hold (T/H) stage and are sampled when the differential sampling clock signal transitions high. The CLKDIV CLKP RESET CLOCKDIVIDER CONTROL CLKN CLOCK MANAGEMENT T/H INN 2.15kΩ CM BUFFER DEMUX ITL 12 BITS BUFFER INP 2.15kΩ ADC following the first T/H stage then digitizes the signal, and controls a digital-to-analog converter (DAC). Digitized and reference signals are then subtracted, resulting in a fractional residue signal that is amplified before it is passed on to the next stage through another T/H amplifier. This process is repeated until the applied input signal has successfully passed through all stages of the 12-bit quantizer. Finally, the digital outputs of all stages are combined and corrected for in the digital correction logic to generate the final output code. The result is a 12-bit parallel digital output word in userselectable two’s complement or binary output formats with CMOS-compatible output levels. See the functional diagram (Figure 1) for a more detailed view of the MAX19541’s architecture. REFERENCE REFIO 12-BIT PIPELINE QUANTIZER CORE CMOS DATA PORTS CLK GENERATOR MAX19541 DA0–DA11, ORA DB0–DB11, ORB 12 BITS DCLKP DCLKN REFADJ Figure 1. MAX19541 Functional Diagram 10 ______________________________________________________________________________________ 12-Bit, 125Msps ADC with CMOS Outputs for Wideband Applications REFT G MAX19541 ADC FULL SCALE = REFT - REFB REFERENCESCALING AMPLIFIER REFB AVCC REFERENCE BUFFER INN INP REFIO 1V 0.1µF MAX19541 2.15kΩ REFADJ* 2.15kΩ CONTROL LINE TO DISABLE REFERENCE BUFFER TO COMMON-MODE INPUT TO COMMON-MODE INPUT AGND AVCC AVCC / 2 *REFADJ CAN BE SHORTED TO AGND THROUGH A 1kΩ RESISTOR OR POTENTIOMETER. Figure 2. Simplified Analog Input Architecture Figure 3. Simplified Reference Architecture Analog Inputs (INP, INN) Clock Inputs (CLKP, CLKN) INP and INN are the fully differential inputs of the MAX19541. Differential inputs usually feature good rejection of even-order harmonics, which allows for enhanced AC performance as the signals are progressing through the analog stages. The MAX19541 analog inputs are self-biased at a 1.365V commonmode voltage and allow a 1.41VP-P differential input voltage swing. Both inputs are self-biased through 2.15kΩ resistors, resulting in a typical differential input resistance of 4.3kΩ (Figure 2). It is recommended driving the analog inputs of the MAX19541 in an AC-coupled configuration to achieve the best dynamic performance. See the Transformer-Coupled, Differential Analog Input Drive section for a detailed discussion of this configuration. Drive the clock inputs of the MAX19541 differentially with an LVPECL-compatible clock to achieve the best dynamic performance. The clock signal source must be high-quality, low phase noise to avoid any degradation in the noise performance of the ADC. The clock inputs (CLKP, CLKN) are internally biased to typically 1.15V, accept a typical 0.5VP-P differential signal swing, and are usually driven in an AC-coupled configuration. See the Differential, ACCoupled Clock Input section for more circuit details on how to drive CLKP and CLKN appropriately. The MAX19541 features an internal clock-management circuit (duty-cycle equalizer). The clock-management circuit ensures that the clock signal applied to inputs CLKP and CLKN is processed to provide a near 50% duty-cycle clock signal. This desensitizes the performance of the converter to variations in the duty cycle of the input clock source. Note that the clock duty-cycle equalizer cannot be turned off externally. On-Chip Reference Circuit The MAX19541 features an internal 1.24V bandgap reference circuit (Figure 3), which, in combination with an internal reference-scaling amplifier, determine the fullscale range of the MAX19541. Bypass REFIO with a 0.1µF capacitor to AGND. To compensate for gain errors or increase the ADC’s full-scale range, the voltage of this bandgap reference can be indirectly adjusted by adding an external resistor (e.g., 100kΩ trim potentiometer) between REFADJ and AGND or REFADJ and REFIO. See Figure 8 and the Applications Information section for a detailed description of this process. Clock Outputs (DCLKP, DCLKN) The MAX19541 features CMOS-complementary clock outputs (DCLKP, DCLKN) to latch the digital output data with an external latch or receiver. Additionally, the clock outputs can be used to synchronize external devices (e.g., FPGAs) to the ADC. There is a 2.1ns delay time between the rising (falling) edge of CLKP (CLKN) and the rising (falling) edge of DCLKP (DCLKN). See Figure 4 for timing details. ______________________________________________________________________________________ 11 MAX19541 12-Bit, 125Msps ADC with CMOS Outputs for Wideband Applications Divide-by-Two Clock Control (CLKDIV) RESET Operation The MAX19541 offers a clock control line (CLKDIV) that allows the reduction of clock jitter and phase noise in a system as higher frequency oscillators usually exhibit better phase noise and jitter characteristics. Connect CLKDIV to OGND to enable the ADC’s internal divideby-2 clock divider, which allows the user to use an oscillator of twice the maximum sampling frequency. The sampling frequency now becomes 1/2 of the input clock frequency. CLKDIV has an internal pulldown resistor and can be left open for applications that require this divide-by-2 mode. Connecting CLKDIV to OVCC disables the divide-by-2 mode. The RESET input defines the pipeline latency of the MAX19541. Drive RESET high to place the MAX19541 in reset mode with the CMOS outputs tri-stated. During the time when RESET is high, no sample information is available at the outputs. For pipeline latency, the first sample is defined at the first rising edge of CLKP after RESET goes low. The conversion information is available at the outputs after 11 clock cycles. Synchronize RESET with the input clock of the device by observing the minimum RESET hold (tHR) and RESET setup (tSR) times (Figure 4). RESET is only used to control the latency of the device and, in applications where this is not critical, drive RESET low or leave unconnected. RESET has an internal pulldown resistor. SAMPLING EVENT SAMPLING EVENT SAMPLING EVENT SAMPLING EVENT INN INP tAD CLKN N N+1 N + 11 CLKP tSR RESET tCH N + 12 tCL tHR Figure 4. RESET Timing Diagram 12 ______________________________________________________________________________________ 12-Bit, 125Msps ADC with CMOS Outputs for Wideband Applications Modes of Operation The MAX19541 features three modes of operation. In each mode of operation, the conversion data is output in a different format. SAMPLING EVENT SAMPLING EVENT Parallel Mode Drive DEMUX low to place the MAX19541 in the parallel mode. In this mode, the output clock has the same frequency as the sampling frequency and conversion data is output at full rate on parallel ports DA0–DA11. Note that the sampling frequency may not be the same as the input clock frequency. See the Divide-by-Two Clock Control (CLKDIV) section. In parallel mode, samples are taken on the rising edge of CLKP. Conversion data appears at the outputs on the rising edge of DCLKP after the latency period of 11 clock cycles and is stable for one clock period (Figure 5). If an overrange condition occurs, it is reflected on the ORA port. SAMPLING EVENT SAMPLING EVENT SAMPLING EVENT INN INP tAD CLKN N N+1 N + 11 N + 12 CLKP tCL tCH RESET tCPDL DCLKP N - 11 N - 10 DCLKN N+1 tLATENCY tPDL DA0–DA11, ORA N N - 11 N - 10 N -1 N N+1 Figure 5. Parallel Mode Timing Diagram ______________________________________________________________________________________ 13 MAX19541 System Timing Requirements Figures 5, 6, and 7 depict the relationship between the clock input and output, analog input, sampling event, and data output. The MAX19541 samples on the rising (falling) edge of CLKP (CLKN). In all these figures, CLKDIV is assumed to be high; otherwise, the sampling events would occur at every other rising edge of CLKP. Output data is latched on the next rising (falling) edge of the DCLKP (DCLKN) clock, but has an internal latency of 11 input clock cycles. MAX19541 12-Bit, 125Msps ADC with CMOS Outputs for Wideband Applications Demux Parallel Mode Drive DEMUX high and ITL low to place the MAX19541 in the demux parallel mode. In this mode, the output clock’s frequency is 1/2 the sampling frequency. The sampling frequency may not be the same as the input clock frequency. See the Divide-by-Two Clock Control (CLKDIV) section. Each conversion starts with a sampling event on the rising edge of CLKP. Conversion data now appears on both DA0–DA11 and DB0–DB11. The first conversion result is output on the A ports on the rising edge of DCLKP after 12 input clock cycles from the initial sampling event. The second conversion SAMPLING EVENT SAMPLING EVENT result is output on the B ports on the rising edge of DCLKP after 11 input clock cycles from the initial sampling event. Both conversion results are output simultaneously (Figure 6). The conversion results on ports A and B remain stable for one period of DCLKP after they become valid. Thus, the overall throughput rate is the same as in parallel mode; however, now each data line is allowed to be valid for a longer time (two sampling periods, one digital clock period). Overrange conditions are reflected on the appropriate output port, ORA or ORB, depending on which conversion they occur. SAMPLING EVENT SAMPLING EVENT SAMPLING EVENT INN INP tAD CLKN N N + 12 CLKP tCPDL tCL tCH RESET DCLKP N N+2 DCLKN tLATENCY tPDL DA0–DA11, ORA N N+2 DB0–DB11, ORB N+1 N+3 DEMUX PARALLEL MODE Figure 6. Demux Parallel Mode Timing Diagram 14 ______________________________________________________________________________________ SAMPLING EVENT 12-Bit, 125Msps ADC with CMOS Outputs for Wideband Applications SAMPLING EVENT SAMPLING EVENT ports on the rising edge of DCLKN after 12 input clock cycles from the initial sampling event. In this way, the two conversion results are interleaved with respect to each other (Figure 7). The conversion results on ports A and B remain stable for one period of DCLKP and DCLKN, respectively, after they become valid. Overrange conditions are reflected on the appropriate output port, ORA or ORB, depending on which conversion they occur. The demux interleaved mode is the recommended demux mode of operation due to the fact that output bus switching is more evenly distributed over sample clock edges. SAMPLING EVENT SAMPLING EVENT SAMPLING EVENT SAMPLING EVENT INN INP tAD CLKN N N + 12 CLKP tCPDL tCL tCH RESET DCLKP N N+2 DCLKN tLATENCY tPDL DA0–DA11, ORA N DB0–DB11, ORB N+2 N+1 N+3 DEMUX INTERLEAVED MODE Figure 7. Demux Interleaved Mode Timing Diagram ______________________________________________________________________________________ 15 MAX19541 Demux Interleaved Mode Drive DEMUX high and ITL high to place the MAX19541 in the demux interleaved mode of operation. In this mode, the output clock’s frequency is 1/2 the sampling frequency. The sampling frequency may not be the same as the input clock frequency. See the Divide-byTwo Clock Control (CLKDIV) section. Each conversion starts with a sampling event on the rising edge of CLKP. Conversion data now appears on both DA0–DA11 and DB0–DB11. The first conversion result is output on the A ports on the rising edge of DCLKP after 12 input clock cycles from the initial sampling event. The second conversion result is output on the B MAX19541 12-Bit, 125Msps ADC with CMOS Outputs for Wideband Applications Digital Outputs (DA0–DA11, DCLKP, DCLKN, ORA, DB0–DB11, ORB) and Control Input T/B Digital outputs DA0/DB0–DA11/DB11, DCLKP, DCLKN, ORA/ORB are CMOS compatible, and data on DA0/DB DA11/DB11 are presented in either binary or two’scomplement format (Table 1). The T/B control line is an LVCMOS-compatible input that allows the user to select the desired output format. Drive T/B high to select data to be output in offset binary format and drive it low to select data to be output in two’s complement format on the 12-bit parallel bus. T/B has an internal pulldown resistor and can be left unconnected in applications using only two’s-complement output format. The CMOS outputs are powered from a separate power supply that can be operated between 1.7V and 1.9V. The MAX19541 offers an additional differential output pair (ORA, ORB) to flag overrange conditions, where overrange is above positive or below negative full scale. An overrange condition is identified with ORA/ORB transitioning high. Note: Keep the capacitive load on the digital outputs as low as possible. Use digital buffers on the digital outputs of the ADC when driving larger loads to improve overall performance and reduce system timing constraints. Further improvements in dynamic performance can be achieved by adding small series resistors (100Ω) to the digital output paths, close to the ADC. Table 1. MAX19541 Digital Output Coding BINARY DIGITAL OUTPUT CODE (D_11–D_0) TWO’S-COMPLEMENT DIGITAL OUTPUT CODE (D_11–D_0) INP ANALOG INPUT VOLTAGE LEVEL INN ANALOG INPUT VOLTAGE LEVEL OVERRANGE ORA/ORB > VREF + 0.35V < VREF - 0.35V 1 1111 1111 1111 (exceeds +FS, OR set) 0111 1111 1111 (exceeds +FS, OR set) VREF + 0.35V VREF - 0.35V 0 1111 1111 1111 (+FS) 0111 1111 1111 (+FS) VREF VREF 0 1000 0000 0000 or 0111 1111 1111 (FS/2) 0000 0000 0000 or 1111 1111 1111 (FS/2) VREF - 0.35V VREF + 0.35V 0 0000 0000 0000 (-FS) 1000 0000 0000 (-FS) < VREF + 0.35V > VREF - 0.35V 1 00 0000 0000 (exceeds -FS, OR set) 10 0000 0000 (exceeds -FS, OR set) 16 ______________________________________________________________________________________ 12-Bit, 125Msps ADC with CMOS Outputs for Wideband Applications Full-Scale Range Adjustments Using the Internal Bandgap Reference The MAX19541 supports a full-scale adjustment range of ±10%. To decrease the full-scale range, an external resistor value ranging from 13kΩ to 1MΩ can be added between REFADJ and AGND. A similar approach can be taken to increase the ADCs full-scale range. Add a variable resistor, potentiometer, or predetermined resistor value between REFADJ and REFIO to increase the full-scale range of the data converter. Figure 8 shows the two possible configurations and their impact on the ADC FULL SCALE = REFT - REFB REFT G VREF (%) = 1.25% x 100kΩ RADJ The percentage change is positive when R ADJ is added between REFADJ and REFIO, and is negative when RADJ is added between REFADJ and GND. ADC FULL SCALE = REFT - REFB REFERENCESCALING AMPLIFIER REFT G REFERENCESCALING AMPLIFIER REFB REFB REFERENCE BUFFER REFERENCE BUFFER 1V REFIO MAX19541 0.1µF 1V 13kΩ TO 100kΩ REFIO MAX19541 REFADJ CONTROL LINE TO DISABLE REFERENCE BUFFER AVCC 0.1µF REFADJ CONTROL LINE TO DISABLE REFERENCE BUFFER AVCC / 2 AVCC 13kΩ TO 100kΩ AVCC / 2 Figure 8. Circuit Suggestions to Adjust the ADC’s Full-Scale Range (Simplified Schematic) ______________________________________________________________________________________ 17 MAX19541 overall full-scale range adjustment of the MAX19541. Do not use resistor values of less than 13kΩ to avoid instability of the internal gain regulation loop for the bandgap reference. Use the following formula to calculate the percentage change of the reference voltage: Applications Information MAX19541 12-Bit, 125Msps ADC with CMOS Outputs for Wideband Applications Differential, AC-Coupled, LVPECLCompatible Clock Input The MAX19541 dynamic performance depends on a very clean clock source. The phase noise floor of the clock source has a negative impact on the SNR performance. Spurious signals on the clock signal source also affect the ADC’s dynamic range. The preferred method of clocking the MAX19541 is differentially with LVPECLcompatible input levels. The fast data transition rates of these logic families minimize the clock-input circuitry’s transition uncertainty, thereby improving the SNR performance. Apply a 50Ω reverse-terminated clock signal source with low phase noise AC-coupled into a fast differential receiver such as the MC100LVEL16 (Figure 9). The receiver produces the necessary LVPECL output levels to drive the clock inputs of the data converter. Transformer-Coupled, Differential Analog Input Drive The MAX19541 provides the best SFDR and THD with fully differential input signals and it is not recommended driving the ADC inputs in single-ended configuration. In differential input mode, even-order harmonics are usually lower since INP and INN are balanced, and each of the ADC inputs requires only half the signal swing compared to a single-ended configuration. Wideband RF transformers provide an excellent solution to convert a single-ended source signal to a fully differential signal, required by the MAX19541 for optimum dynamic performance. A secondary-side termination of a 1:1 transformer (e.g., Mini-Circuit’s ADT1-1WT) into two separate 24.9Ω ±0.1% resistors (use tight resistor tolerances to minimize effects of imbalance; 0.1% would be an ideal choice) placed between top/bottom and center tap of the transformer is recommended to maximize the ADC’s dynamic range. This configuration optimizes THD and SFDR performance of the ADC by reducing the effects of transformer parasitics. However, the source impedance combined with the shunt capacitance provided by a PC board and the ADC’s parasitic capacitance limit the ADC’s full-power input bandwidth to approximately 600MHz. To further enhance THD and SFDR performance at high input frequencies (>100MHz), a second transformer (Figure 10) should be placed in series with the singleended-to-differential conversion transformer. This transformer reduces the increase of even-order harmonics at high frequencies. For more detailed information on transformer termination methods, refer to the Application Note: SecondarySide Transformer Termination Improves Gain Flatness in High-Speed ADCs from the Maxim website: www.maxim-ic.com. VCLK 0.1µF SINGLE-ENDED INPUT TERMINAL 0.1µF 8 0.1µF 2 7 150Ω MC100LVEL16 0.1µF 50Ω AVCC OVCC 6 3 510Ω 150Ω 510Ω 4 0.01µF 5 INP CLKN CLKP D_0–D_11, OR_ MAX19541 VGND 12 INN AGND OGND Figure 9. Differential, AC-Coupled, LVPECL-Compatible Clock Input Configuration 18 ______________________________________________________________________________________ 12-Bit, 125Msps ADC with CMOS Outputs for Wideband Applications SINGLE-ENDED INPUT TERMINAL 10Ω 0.1µF OVCC INP ADT1-1WT ADT1-1WT MAX19541 AVCC D_0–D_11, OR_ 25Ω MAX19541 25Ω 0.1µF 12 INN 10Ω AGND OGND Figure 10. Analog Input Configuration with Back-to-Back Transformers and Secondary-Side Termination AVCC SINGLE-ENDED INPUT TERMINAL 0.1µF OVCC INP D_0–D_11, OR_ 50Ω 0.1µF MAX19541 INN 12 25Ω AGND OGND Figure 11. Single-Ended AC-Coupled Analog Input Configuration Single-Ended, AC-Coupled Analog Input Although not recommended, the MAX19541 can be used in single-ended mode (Figure 11). Analog signals can be AC-coupled to the positive input INP through a 0.1µF capacitor and terminated with a 49.9Ω resistor to AGND. Terminate the negative input with a 24.9Ω resistor and AC ground it with a 0.1µF capacitor. Grounding, Bypassing, and Board Layout Considerations The MAX19541 requires board layout design techniques suitable for high-speed data converters. This ADC provides separate analog and digital power supplies. The analog and digital supply voltage inputs AV CC and OV CC accept 1.7V to 1.9V input voltage ranges. Although both supply types can be combined and supplied from one source, it is recommended using separate sources to cut down on performance degradation caused by digital switching currents that can couple into the analog supply network. Isolate analog and digital supplies (AVCC and OVCC) where they enter the PC board with separate networks of ferrite beads and capacitors to their corresponding grounds (AGND, OGND). To achieve optimum performance, provide each supply with a separate network of a 47µF tantalum capacitor in parallel with 10µF and 1µF ceramic capacitors. Additionally, the ADC requires each supply pin to be bypassed with separate 0.1µF ceramic capacitors (Figure 12). Locate these capacitors directly at the ADC supply pins or as close as possible to the MAX19541. Choose surface-mount capacitors, whose preferred location should be on the same side as the converter, to save space and minimize the inductance. If close placement on the same side is not possible, these bypassing capacitors may be routed through vias to the bottom side of the PC board. Multilayer boards with separated ground and power planes produce the highest level of signal integrity. Consider the use of a split ground plane arranged to match the physical location of analog and digital ground on the ADC’s package. The two ground planes should be joined at a single point so the noisy digital ground currents do not interfere with the analog ground plane. A major concern with this approach are the dynamic currents that may need to travel long distances before they are recombined at a common source ground, resulting in large and undesirable ______________________________________________________________________________________ 19 MAX19541 12-Bit, 125Msps ADC with CMOS Outputs for Wideband Applications ground loops. Ground loops can add to digital noise by coupling back to the analog front end of the converter, resulting in increased spur activity and a decreased noise performance. Alternatively, all ground pins could share the same ground plane if the ground plane is sufficiently isolated from any noisy, digital systems ground. To minimize the effects of digital noise coupling, ground return vias can be positioned throughout the layout to divert digital switching currents away from the sensitive analog sections of the ADC. This does not require additional ground splitting, but can be accomplished by placing substantial ground connections between the analog front end and the digital outputs. The MAX19541 is packaged in a 68-pin QFN-EP package (package code: G6800-4), providing greater design flexibility, increased thermal dissipation, and optimized AC performance of the ADC. The EP must be soldered down to AGND. In this package, the data converter die is attached to an EP lead frame with the back of this frame exposed at the package bottom surface, facing the PC board side of the package. This allows a solid attachment of the package to the board with standard infrared (IR) flow-soldering techniques. Thermal efficiency is one of the factors for the selection of a package with an exposed pad for the MAX19541. The exposed pad improves thermal dissipation and ensures a solid ground connection between the ADC and the PC board’s analog ground layer. Take considerable care when routing the digital output traces for a high-speed, high-resolution data converter. It is essential to keep trace lengths at a minimum and place minimal capacitive loading—less than 5pF—on any digital trace to prevent coupling to sensitive analog sections of the ADC. Route high-speed digital signal traces away from sensitive analog traces, and remove digital ground and power planes from underneath digital outputs. Keep all signal lines short and free of 90° turns. Static Parameter Definitions Integral Nonlinearity (INL) Integral nonlinearity is the deviation of the values on an actual transfer function from a straight line. This straight line can be either a best straight-line fit or a line drawn between the end points of the transfer function, once offset and gain errors have been nullified. However, the static linearity parameters for the MAX19541 are measured using the histogram method with a 10MHz input frequency. Differential Nonlinearity (DNL) Differential nonlinearity is the difference between an actual step width and the ideal value of 1 LSB. A DNL error specification of less than 1 LSB guarantees no missing codes and a monotonic transfer function. The MAX19541’s DNL specification is measured with the histogram method based on a 10MHz input tone. BYPASSING-ADC LEVEL BYPASSING-BOARD LEVEL OVCC AVCC 0.1µF AVCC 0.1µF 1µF D_0–D_11, OR_ 10µF 47µF ANALOG POWERSUPPLY SOURCE 10µF 47µF DIGITAL/OUTPUT DRIVER POWERSUPPLY SOURCE OVCC MAX19541 12 1µF AGND OGND NOTE: EACH POWER-SUPPLY PIN (ANALOG AND DIGITAL) SHOULD BE DECOUPLED WITH AN INDIVIDUAL 0.1µF CAPACITOR AS CLOSE AS POSSIBLE TO THE ADC. Figure 12. Grounding, Bypassing, and Decoupling Recommendations for the MAX19541 20 ______________________________________________________________________________________ 12-Bit, 125Msps ADC with CMOS Outputs for Wideband Applications Aperture Jitter Figure 13 depicts the aperture jitter (tAJ), which defines the sample-to-sample variation in the aperture delay. Aperture jitter is measured in psRMS. Aperture Delay Aperture delay (tAD) is the time defined between the 620ps rising edge of the sampling clock and the instant when an actual sample is taken (Figure 13). Signal-to-Noise Ratio (SNR) For a waveform perfectly reconstructed from digital samples, the theoretical maximum SNR is the ratio of the full-scale analog input (RMS value) to the RMS quantization error (residual error). The ideal, theoretical minimum analog-to-digital noise is caused by quantization error only and results directly from the ADC’s resolution (N bits): SNRdB[max] = 6.02dB x N + 1.76dB In reality, other noise sources such as thermal noise, clock jitter, signal phase noise, and transfer function nonlinearities are also contributing to the SNR calculation and should be considered when determining the SNR of an ADC. Signal-to-Noise Plus Distortion (SINAD) SINAD is computed by taking the ratio of the RMS signal to all spectral components excluding the fundamental and the DC offset. In the case of the MAX19541, SINAD is computed from a curve fit. Spurious-Free Dynamic Range (SFDR) SFDR is the ratio of RMS amplitude of the carrier frequency (maximum signal component) to the RMS value of the next-largest noise or harmonic distortion component. SFDR is usually measured in dBc with respect to the carrier frequency amplitude or in dBFS with respect to the ADC’s full-scale range. Two-Tone Intermodulation Distortion (IMD) The two-tone IMD is the ratio expressed in decibels of either input tone to the worst 2nd-order (or higher) intermodulation products. The individual input tone levels are usually set to 7dB below full scale and intermodulation products IM2 through IM5 are considered for the IMD calculation. The various intermodulation products are defined as follows: • 2nd-order intermodulation distortion (IM2): fIN1 + fIN2, fIN2 - fIN1 • 3rd-order intermodulation distortion (IM3): 2fIN1 + fIN2, 2fIN1 - fIN2, 2fIN2 + fIN1, 2fIN2 - fIN1 • 4th-order intermodulation distortion (IM4): 3fIN1 + fIN2, 3fIN1 - fIN2, 3fIN2 + fIN1, 3fIN2 - fIN1 • 5th-order intermodulation distortion (IM5): 4fIN1 + fIN2, 4fIN1 - fIN2, 4fIN2 + fIN1, 4fIN2 - fIN1 Full-Power Bandwidth A large -1dBFS analog input signal is applied to an ADC and the input frequency is swept up to the point where the amplitude of the digitized conversion result has decreased by 3dB. The -3dB point is defined as the full-power input bandwidth frequency of the ADC. CLKN CLKP ANALOG INPUT tAD tAJ SAMPLED DATA (T/H) T/H TRACK HOLD TRACK Figure 13. Aperture Jitter/Delay Specifications ______________________________________________________________________________________ 21 MAX19541 Dynamic Parameter Definitions Package Information (The package drawing(s) in this data sheet may not reflect the most current specifications. For the latest package outline information, go to www.maxim-ic.com/packages.) 68L QFN.EPS MAX19541 12-Bit, 125Msps ADC with CMOS Outputs for Wideband Applications PACKAGE OUTLINE, 68L QFN, 10x10x0.9 MM 1 C 21-0122 2 22 ______________________________________________________________________________________ 12-Bit, 125Msps ADC with CMOS Outputs for Wideband Applications PACKAGE OUTLINE, 68L QFN, 10x10x0.9 MM 1 C 21-0122 2 Maxim cannot assume responsibility for use of any circuitry other than circuitry entirely embodied in a Maxim product. No circuit patent licenses are implied. Maxim reserves the right to change the circuitry and specifications without notice at any time. Maxim Integrated Products, 120 San Gabriel Drive, Sunnyvale, CA 94086 408-737-7600 ____________________ 23 © 2004 Maxim Integrated Products Printed USA is a registered trademark of Maxim Integrated Products. MAX19541 Package Information (continued) (The package drawing(s) in this data sheet may not reflect the most current specifications. For the latest package outline information, go to www.maxim-ic.com/packages.)