19-2726; Rev 3; 12/03 3.3V, 16-Bit, 500Msps High Dynamic Performance DAC with Differential LVDS Inputs Single 3.3V Supply Operation Excellent SFDR and IMD Performance SFDR = 76dBc at fOUT = 40MHz (to Nyquist) IMD = -85dBc at fOUT = 10MHz ACLR = 73dB at fOUT = 61MHz 2mA to 20mA Full-Scale Output Current Differential, LVDS-Compatible Digital and Clock Inputs On-Chip 1.2V Bandgap Reference Low 130mW Power Dissipation 68-Lead QFN-EP Package Ordering Information PART MAX5888AEGK -40C to +85C 68 QFN-EP* MAX5888EGK -40C to +85C 68 QFN-EP* *EP = Exposed paddle. 68 Applications Communications: LMDS, MMDS, Point-to-Point Microwave Digital Signal Synthesis Automated Test Equipment (ATE) Instrumentation 63 62 61 60 59 58 B10N B10P B9P B9N B8P B8N B7P B7N DGND DVDD DGND B6N 67 66 65 64 B6P B5P B5N Pin Configuration TOP VIEW Base Stations: Single-/Multicarrier UMTS, CDMA, GSM PINPACKAGE TEMP RANGE B4P The digital and clock inputs of the MAX5888 are designed for differential low-voltage differential signal (LVDS)-compatible voltage levels. The MAX5888 is available in a 68-lead QFN package with an exposed paddle (EP) and is specified for the extended industrial temperature range (-40C to +85C). Refer to the MAX5887 and MAX5886 data sheets for pin-compatible 14- and 12-bit versions of the MAX5888. 500Msps Output Update Rate B4N The MAX5888 utilizes a current-steering architecture, which supports a full-scale output current range of 2mA to 20mA, and allows a differential output voltage swing between 0.1VP-P and 1VP-P. The MAX5888 features an integrated 1.2V bandgap reference and control amplifier to ensure high accuracy and low noise performance. Additionally, a separate reference input pin enables the user to apply an external reference source for optimum flexibility and to improve gain accuracy. Features 57 56 55 54 53 52 EP B3P 1 B3N 2 50 B11P B2P 3 49 B12N B2N 4 48 B12P B1P 5 47 B13N B1N 6 46 B13P B0P 7 45 B14N B0N 8 DGND 9 51 B11N 44 B14P MAX5888 43 B15N DVDD 10 42 B15P VCLK 11 41 DGND CLKGND 12 40 DVDD CLKP 13 39 SEL0 CLKN 14 38 N.C. CLKGND 15 37 N.C. VCLK 16 36 N.C. PD 17 35 N.C. N.C. AGND AVDD AGND AVDD AVDD AGND IOUTP IOUTN AVDD AGND N.C. DACREF FSADJ REFIO AVDD 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 MAX5888 General Description The MAX5888 is an advanced, 16-bit, 500Msps digitalto-analog converter (DAC) designed to meet the demanding performance requirements of signal synthesis applications found in wireless base stations and other communications applications. Operating from a single 3.3V supply, this DAC offers exceptional dynamic performance such as 76dBc spurious-free dynamic range (SFDR) at fOUT = 40MHz. The DAC supports update rates of 500Msps and a power dissipation of only 250mW. MAX5888 3.3V, 16-Bit, 500Msps High Dynamic Performance DAC with Differential LVDS Inputs ABSOLUTE MAXIMUM RATINGS Continuous Power Dissipation (TA = +70C) 68-Lead QFN-EP (derate 41.7mW/C above +70C) ...3333mW Thermal Resistance (JA) ..............................................+24C/W Operating Temperature Range ..........................-40C to +85C Junction Temperature .....................................................+150C Storage Temperature Range ............................-60C to +150C Lead Temperature (soldering, 10s) ................................+300C AVDD, DVDD, VCLK to AGND................................-0.3V to +3.9V AVDD, DVDD, VCLK to DGND ...............................-0.3V to +3.9V AVDD, DVDD, VCLK to CLKGND ...........................-0.3V to +3.9V AGND, CLKGND to DGND....................................-0.3V to +0.3V DACREF, REFIO, FSADJ to AGND.............-0.3V to AVDD + 0.3V IOUTP, IOUTN to AGND................................-1V to AVDD + 0.3V CLKP, CLKN to CLKGND...........................-0.3V to VCLK + 0.3V B0P/B0N-B15P/B15N, SEL0, PD to DGND ...........................................-0.3V to DVDD + 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 (AVDD = DVDD = VCLK = 3.3V, AGND = DGND = CLKGND = 0, external reference, VREFIO = 1.25V, differential transformer-coupled analog output, 50 double terminated (Figure 7), IOUT = 20mA, TA = TMIN to TMAX, unless otherwise noted. +25C guaranteed by production test, <+25C guaranteed by design and characterization. Typical values are at TA = +25C.) PARAMETER SYMBOL CONDITIONS MIN TYP MAX UNITS STATIC PERFORMANCE Resolution Integral Nonlinearity 16 INL MAX5888A___, measured differentially, TA +25C -0.008 Offset Error DNL MAX5888A___, measured differentially, TA +25C 0.006 -0.006 -0.025 0.003 +0.025 50 GEFS Gain Drift Full-Scale Output Current +0.006 0.003 Offset Drift Full-Scale Gain Error 0.002 % FS MAX5888___, measured differentially, TA +25C OS +0.008 % FS MAX5888___, measured differentially, TA +25C Differential Nonlinearity 0.004 Bits IOUT Min Output Voltage Max Output Voltage External reference, TA +25C -3.1 +1.1 Internal reference 100 External reference 50 (Note 1) 2 -0.5 Single ended %FS ppm/C 20 Single ended %FS ppm/C mA V 1.1 V Output Resistance ROUT 1 M Output Capacitance COUT 5 pF DYNAMIC PERFORMANCE Output Update Rate fCLK Noise Spectral Density Spurious-Free Dynamic Range to Nyquist 2 SFDR 1 500 fCLK = 300MHz fOUT = 16MHz, -12dB FS -165 fCLK = 500MHz fOUT = 16MHz, -12dB FS -164 fCLK = 100MHz fOUT = 1MHz, 0dB FS 88 fOUT = 1MHz, -6dB FS 89 fOUT = 1MHz, -12dB FS 85 _______________________________________________________________________________________ Msps dB FS/ Hz dBc 3.3V, 16-Bit, 500Msps High Dynamic Performance DAC with Differential LVDS Inputs (AVDD = DVDD = VCLK = 3.3V, AGND = DGND = CLKGND = 0, external reference, VREFIO = 1.25V, differential transformer-coupled analog output, 50 double terminated (Figure 7), IOUT = 20mA, TA = TMIN to TMAX, unless otherwise noted. +25C guaranteed by production test, <+25C guaranteed by design and characterization. Typical values are at TA = +25C.) PARAMETER SYMBOL CONDITIONS fCLK = 100MHz Spurious-Free Dynamic Range to Nyquist 2-Tone IMD fOUT = 30MHz, -12dB FS 79 fOUT = 10MHz, -12dB FS 73 SFDR 69 fOUT = 50MHz, -12dB FS 72 66 fOUT = 10MHz, -12dB FS 67 fOUT = 30MHz, -12dB FS 65 fOUT = 50MHz, -12dB FS 65 fOUT = 80MHz, -12dB FS 63 fCLK = 150MHz fOUT = 20MHz, -12dB FS 82 fCLK = 100MHz fOUT1 = 9MHz, -6dB FS, fOUT2 = 10MHz, -6dB FS -85 fCLK = 300MHz fOUT1 = 49MHz, -12dB FS, fOUT2 = 50MHz, -12dB FS -83 TTIMD MAX UNITS 77 fOUT = 80MHz, -12dB FS fCLK = 500MHz Spurious-Free Dynamic Range, 25MHz Bandwidth TYP 82 fOUT = 16MHz, -12dB FS, fCLK = 200MHz TA +25C SFDR MIN fOUT = 10MHz, -12dB FS dBc dBc dBc 4-Tone IMD, 1MHz Frequency Spacing, GSM Model FTIMD fCLK = 300MHz fOUT = 32MHz, -12dB FS -78 dBc Adjacent Channel Leakage Power Ratio, 4.1MHz Bandwidth, WCDMA Model ACLR fCLK = 184.32MHz 73 dB 450 MHz Output Bandwidth BW-1dB fOUT = 61.44MHz (Note 2) REFERENCE Internal Reference Voltage Range VREFIO Reference Voltage Drift TCOREF Reference Input Compliance Range VREFIOCR Reference Input Resistance RREFIO 1.13 1.22 1.3 50 0.125 V ppm/C 1.250 V 10 k ANALOG OUTPUT TIMING Output Fall Time tFALL 90% to 10% (Note 3) 375 ps Output Rise Time tRISE 10% to 90% (Note 3) 375 ps Output Voltage Settling Time Output Propagation Delay tSETTLE tPD Output settles to 0.025% FS (Note 3) 11 ns (Note 3) 1.8 ns 1 pV-s Glitch Energy Output Noise NOUT IOUT = 2mA 30 IOUT = 20mA 30 pA/Hz _______________________________________________________________________________________ 3 MAX5888 ELECTRICAL CHARACTERISTICS (continued) MAX5888 3.3V, 16-Bit, 500Msps High Dynamic Performance DAC with Differential LVDS Inputs ELECTRICAL CHARACTERISTICS (continued) (AVDD = DVDD = VCLK = 3.3V, AGND = DGND = CLKGND = 0, external reference, VREFIO = 1.25V, differential transformer-coupled analog output, 50 double terminated (Figure 7), IOUT = 20mA, TA = TMIN to TMAX, unless otherwise noted. +25C guaranteed by production test, <+25C guaranteed by design and characterization. Typical values are at TA = +25C.) PARAMETER SYMBOL CONDITIONS MIN TYP MAX UNITS TIMING CHARACTERISTICS Data to Clock Setup Time tSETUP Referenced to rising edge of clock (Note 4) -0.8 Data to Clock Hold Time tHOLD Referenced to rising edge of clock (Note 4) 1.8 ns ns Data Latency 3.5 Clock cycles Minimum Clock Pulse Width High tCH CLKP, CLKN Minimum Clock Pulse Width Low tCL CLKP, CLKN LVDS LOGIC INPUTS (B0N-B15N, B0P-B15P) 0.9 0.9 ns ns Differential Input Logic High VIH Differential Input Logic Low VIL Common-Mode Voltage Range 100 VCOM 1.125 Differential Input Resistance RIN 85 Input Capacitance CIN mV 100 -100 mV 1.375 V 125 5 pF CMOS LOGIC INPUTS (PD, SEL0) Input Logic High VIH Input Logic Low VIL Input Leakage Current IIN Input Capacitance CIN 0.7 DVDD V 0.3 DVDD -15 +15 5 V A pF CLOCK INPUTS (CLKP, CLKN) Differential Input Voltage Swing VCLK Sine wave 1.5 Square wave 0.5 (Note 5) >100 V/s VP-P Differential Input Slew Rate SRCLK Common-Mode Voltage Range VCOM 1.5 20% V Input Resistance RCLK 5 k Input Capacitance CCLK 5 pF POWER SUPPLIES Analog Supply Voltage Range AVDD 3.135 3.3 3.465 V Digital Supply Voltage Range DVDD 3.135 3.3 3.465 V Clock Supply Voltage Range VCLK 3.135 3.3 3.465 V Analog Supply Current IAVDD Digital Supply Current IDVDD Clock Supply Current 4 IVCLK fCLK = 100Msps, fOUT = 1MHz 27 Power-down 0.3 mA fCLK = 100Msps, fOUT = 1MHz 7 Power-down 10 A fCLK = 100Msps, fOUT = 1MHz 5.6 mA Power-down 10 A _______________________________________________________________________________________ mA 3.3V, 16-Bit, 500Msps High Dynamic Performance DAC with Differential LVDS Inputs (AVDD = DVDD = VCLK = 3.3V, AGND = DGND = CLKGND = 0, external reference, VREFIO = 1.25V, differential transformer-coupled analog output, 50 double terminated (Figure 7), IOUT = 20mA, TA = TMIN to TMAX, unless otherwise noted. +25C guaranteed by production test, <+25C guaranteed by design and characterization. Typical values are at TA = +25C.) PARAMETER SYMBOL Power Dissipation PDISS Power-Supply Rejection Ratio PSRR Note 1: Note 2: Note 3: Note 4: Note 5: Note 6: CONDITIONS MIN TYP fCLK = 100Msps, fOUT = 1MHz MAX UNITS 130 Power-down mW 1 AVDD = VCLK = DVDD = 3.3V 5% (Note 6) -1 +1 %FS/V Nominal full-scale current IOUT = 32 IREF. This parameter does not include update-rate depending effects of sin(x)/x filtering inherent in the MAX5888. Parameter measured single ended into a 50 termination resistor. Parameter guaranteed by design. A differential clock input slew rate of >100V/ms is required to achieve the specified dynamic performance. Parameter defined as the change in midscale output caused by a 5% variation in the nominal supply voltage. Typical Operating Characteristics (AVDD = DVDD = VCLK = 3.3V, external reference, VREFIO = 1.25V, RL = 50, IOUT = 20mA, TA = +25C, unless otherwise noted.) 80 80 0dB FS 50 40 60 -12dB FS 50 0dB FS 40 50 40 30 30 20 20 20 10 10 10 0 20 30 40 50 fOUT (MHz) fT1 = 9.0252MHz fT2 = 10.0417MHz -20 fT2 -40 -50 -60 -70 -90 -12dB FS -80 -70 -60 -6dB FS 2 x fT2 - fT1 2 x fT1 - fT2 -80 -100 TWO-TONE IMD (dBc) fT1 -30 0 100 150 200 250 2-TONE INTERMODULATION DISTORTION (fCLK = 450MHz) 0 AOUT = -6dB FS BW = 5MHz -10 fT1 = 79.2114MHz fT2 = 80.0903MHz -20 fT1 -30 fT2 -40 -50 -60 2 x fT2 - fT1 2 x fT1 - fT2 -70 -50 -80 -40 -100 -90 50 fOUT (MHz) MAX5888 toc05 AOUT = -6dB FS BW = 5MHz MAX5888 toc04 0 -10 10 20 30 40 50 60 70 80 90 100 fOUT (MHz) 2-TONE IMD vs. OUTPUT FREQUENCY (1MHz CARRIER SPACING, fCLK = 300MHz) 2-TONE INTERMODULATION DISTORTION (fCLK = 100MHz) 0dB FS 0 0 2-TONE IMD (dBm) 10 -12dB FS 60 30 0 -6dB FS 70 SFDR (dBc) -6dB FS 60 0 2-TONE IMD (dBm) 90 70 SFDR (dBc) SFDR (dBc) 70 100 MAX5888 toc03 80 -6dB FS 90 SPURIOUS-FREE DYNAMIC RANGE vs. OUTPUT FREQUENCY (fCLK = 500MHz) MAX5888 toc02 -12dB FS 90 100 MAX5888 toc01 100 SPURIOUS-FREE DYNAMIC RANGE vs. OUTPUT FREQUENCY (fCLK = 200MHz) MAX5888 toc06 SPURIOUS-FREE DYNAMIC RANGE vs. OUTPUT FREQUENCY (fCLK = 100MHz) -90 -100 7 8 9 10 fOUT (MHz) 11 12 0 25 50 75 fOUT (MHz) 100 77 78 79 80 81 82 fOUT (MHz) _______________________________________________________________________________________ 5 MAX5888 ELECTRICAL CHARACTERISTICS (continued) Typical Operating Characteristics (continued) (AVDD = DVDD = VCLK = 3.3V, external reference, VREFIO = 1.25V, RL = 50, IOUT = 20mA, TA = +25C, unless otherwise noted.) SFDR vs. OUTPUT FREQUENCY (fCLK = 200MHz, AOUT = -6dB FS) fOUT = 10MHz fOUT = 40MHz 82 4.5 MAX5888toc9 90 MAX5888toc08 IOUT = 20mA 80 INTEGRAL NONLINEARITY vs. DIGITAL INPUT CODE SFDR vs. TEMPERATURE (fCLK = 300MHz, AOUT = -6dB FS, IOUT = 20mA) MAX5888toc07 100 3.0 60 IOUT = 10mA 40 20 66 fOUT = 120MHz -1.5 58 -3.0 fOUT = 80MHz 0 -4.5 50 0 10 20 30 40 50 60 70 80 90 100 -40 fOUT (MHz) -15 10 35 60 0 10000 20000 30000 40000 50000 60000 70000 85 DIGITAL INPUT CODE TEMPERATURE (C) 8-TONE MULTITONE POWER RATIO PLOT (fCLK = 300MHz, fCENTER = 31.9702MHz) DIFFERENTIAL NONLINEARTIY vs. DIGITAL INPUT CODE 0 MAX5888toc10 3.0 0 AOUT = -18dB FS BW = 12MHz -10 -20 8-TONE MTPR (dBm) 1.5 DNL (LSB) 0 MAX5888 toc11 IOUT = 5mA 74 INL (LSB) SFDR (dBc) 1.5 SFDR (dBc) -1.5 fT5 fT1 -30 fT2 -40 fT6 fT3 -50 fT7 fT4 -60 fT8 -70 -80 -90 -100 -3.0 26 0 10000 20000 30000 40000 50000 60000 70000 28 30 DIGITAL INPUT CODE fT1 = 28.0151MHz fT2 = 29.0405MHz fT3 = 30.0659MHz fT4 = 31.0181MHz 200 160 38 fT5 = 33.06881MHz fT6 = 34.0209MHz fT7 = 35.0464MHz fT8 = 36.0718MHz 132 EXTERNAL REFERENCE 128 INTERNAL REFERENCE 124 120 120 116 80 100 200 300 fCLK (MHz) 6 36 MAX5888toc13 240 34 136 POWER DISSIPATION (mW) MAX5888toc12 280 32 fOUT (MHz) POWER DISSIPATION vs. SUPPLY VOLTAGE (fCLK = 100MHz, fOUT = 10MHz, IFS = 20mA) POWER DISSIPATION vs. CLOCK FREQUENCY (fOUT = 10MHz, AOUT = 0dB FS, IOUT = 20mA) POWER DISSIPATION (mW) MAX5888 3.3V, 16-Bit, 500Msps High Dynamic Performance DAC with Differential LVDS Inputs 400 500 3.135 3.190 3.245 3.300 3.355 3.410 SUPPLY VOLTAGE (V) _______________________________________________________________________________________ 3.465 3.3V, 16-Bit, 500Msps High Dynamic Performance DAC with Differential LVDS Inputs PIN NAME 1 B3P Data Bit 3 FUNCTION 2 B3N Complementary Data Bit 3 3 B2P Data Bit 2 4 B2N Complementary Data Bit 2 5 B1P Data Bit 1 6 B1N Complementary Data Bit 1 7 B0P Data Bit 0 (LSB) 8 B0N Complementary Data Bit 0 (LSB) 9, 41, 60, 62 DGND Digital Ground 10, 40, 61 DVDD Digital Supply Voltage. Accepts a supply voltage range of 3.135V to 3.465V. Bypass each pin with a 0.1F capacitor to the nearest DGND. 11, 16 VCLK Clock Supply Voltage. Accepts a supply voltage range of 3.135V to 3.465V. Bypass each pin with a 0.1F capacitor to the nearest CLKGND. 12, 15 CLKGND 13 CLKP 14 CLKN Complementary Converter Clock Input. Negative input terminal for the differential converter clock. 17 PD Power-Down Input. PD pulled high enables the DAC's power-down mode. PD pulled low allows for normal operation of the DAC. This pin features an internal pulldown resistor. 18, 24, 29, 30, 32 AVDD Analog Supply Voltage. Accepts a supply voltage range of 3.135V to 3.465V. Bypass each pin with a 0.1F capacitor to the nearest AGND. 19, 25, 28, 31, 33, EP AGND Analog Ground. Exposed paddle (EP) must be connected to AGND. 20 REFIO Reference I/O. Output of the internal 1.2V precision bandgap reference. Bypass with a 1F capacitor to AGND. Can be driven with an external reference source. 21 FSADJ Full-Scale Adjust Input. This input sets the full-scale output current of the DAC. For 20mA full-scale output current, connect a 2k resistor between FSADJ and DACREF. 22 DACREF Return Path for the Current Set Resistor. For 20mA full-scale output current, connect a 2k resistor between FSADJ and DACREF. 23, 34-38 N.C. 26 IOUTN Complementary DAC Output. Negative terminal for differential current output. The full-scale output current range can be set from 2mA to 20mA. 27 IOUTP DAC Output. Positive terminal for differential current output. The full-scale output current range can be set from 2mA to 20mA. 39 SEL0 Mode Select Input SEL0. Set high to activate the segment shuffling function. Since this pin features an internal pulldown resistor, it can be left open or pulled low to disable the segment-shuffling function. See Segment Shuffling in the Detailed Description section for more information. 42 B15P Data Bit 15 (MSB) 43 B15N Complementary Data Bit 15 (MSB) 44 B14P Data Bit 14 Clock Ground Converter Clock Input. Positive input terminal for the differential converter clock. Not Connected. Do not connect to these pins. Do not tie these pins together. _______________________________________________________________________________________ 7 MAX5888 Pin Description 3.3V, 16-Bit, 500Msps High Dynamic Performance DAC with Differential LVDS Inputs MAX5888 Pin Description (continued) PIN NAME 45 B14N Complementary Data Bit 14 FUNCTION 46 B13P Data Bit 13 47 B13N Complementary Data Bit 13 48 B12P Data Bit 12 49 B12N Complementary Data Bit 12 50 B11P Data Bit 11 51 B11N Complementary Data Bit 11 52 B10P Data Bit 10 53 B10N 54 B9P 55 B9N Complementary Data Bit 9 56 B8P Data Bit 8 57 B8N Complementary Data Bit 8 58 B7P Data Bit 7 59 B7N Complementary Data Bit 7 63 B6P Data Bit 6 64 B6N Complementary Data Bit 6 65 B5P Data Bit 5 66 B5N Complementary Data Bit 5 67 B4P Data Bit 4 68 B4N Complementary Data Bit 4 Complementary Data Bit 10 Data Bit 9 Detailed Description Architecture The MAX5888 is a high-performance, 16-bit, currentsteering DAC (Figure 1) capable of operating with clock speeds up to 500MHz. The converter consists of separate input and DAC registers, followed by a currentsteering circuit. This circuit is capable of generating differential full-scale currents in the range of 2mA to 20mA. An internal current-switching network in combination with external 50 termination resistors convert the differential output currents into a differential output voltage with a peak-to-peak output voltage range of 0.1V to 1V. An integrated 1.2V bandgap reference, control amplifier, and user-selectable external resistor determine the data converter's full-scale output range. Reference Architecture and Operation The MAX5888 supports operation with the on-chip 1.2V bandgap reference or an external reference voltage source. REFIO serves as the input for an external, lowimpedance reference source, and as the output if the DAC is operating with the internal reference. For stable 8 operation with the internal reference, REFIO should be decoupled to AGND with a 0.1F capacitor. Due to its limited output drive capability REFIO must be buffered with an external amplifier, if heavier loading is required. The MAX5888's reference circuit (Figure 2) employs a control amplifier, designed to regulate the full-scale current IOUT for the differential current outputs of the DAC. Configured as a voltage-to-current amplifier, the output current can be calculated as follows: IOUT = 32 IREFIO - 1LSB IOUT = 32 IREFIO - (IOUT / 216) where IREFIO is the reference output current (IREFIO = VREFIO/RSET) and IOUT is the full-scale output current of the DAC. Located between FSADJ and DACREF, RSET is the reference resistor, which determines the amplifier's output current for the DAC. See Table 1 for a matrix of different IOUT and RSET selections. _______________________________________________________________________________________ 3.3V, 16-Bit, 500Msps High Dynamic Performance DAC with Differential LVDS Inputs DGND SEL0 FUNCTION SELECTION BLOCK 1.2V REFERENCE MAX5888 DVDD PD AGND AVDD MAX5888 REFIO IOUTP IOUTN CURRENT-STEERING DAC REFADJ CLKN CLKP SEGMENT SHUFFLING/LATCH DECODER LVDS RECEIVER INPUT/LATCH 16 DIFFERENTIAL DIGITAL INPUT B0 THROUGH B15 Figure 1. Simplified MAX5888 Block Diagram Table 1. IOUT and RSET Selection Matrix Based on a Typical 1.200V Reference Voltage FULL-SCALE CURRENT IOUT (mA) REFERENCE CURRENT IREF (A) RSET (k) CALCULATED 1% EIA STD OUTPUT VOLTAGE VIOUTP/N* (mVP-P) 2 62.5 19.2 19.1 100 5 156.25 7.68 7.5 250 10 312.5 3.84 3.83 500 15 468.75 2.56 2.55 750 20 625 1.92 1.91 1000 *Terminated into a 50 load. Analog Outputs (IOUTP, IOUTN) The MAX5888 outputs two complementary currents (IOUTP, IOUTN) that can be operated in a singleended or differential configuration. A load resistor can convert these two output currents into complementary single-ended output voltages. The differential voltage existing between IOUTP and IOUTN can also be con- verted to a single-ended voltage using a transformer or a differential amplifier configuration. If no transformer is used, the output should have a 50 termination to the analog ground and a 50 resistor between the outputs. Although not recommended, because of additional noise pickup from the ground plane, for single-ended _______________________________________________________________________________________ 9 MAX5888 3.3V, 16-Bit, 500Msps High Dynamic Performance DAC with Differential LVDS Inputs AVDD 1.2V REFERENCE AVDD CURRENT SOURCES 10k REFIO CURRENT SWITCHES 0.1F IOUTP FSADJ IREF RSET CURRENT-STEERING DAC IOUT IOUTN DACREF IOUT IOUTN IOUTP IREF = VREFIO/RSET Figure 2. Reference Architecture, Internal Reference Configuration operation IOUTP should be selected as the output, with IOUTN connected to AGND. Note that a single-ended output configuration has a higher 2nd-order harmonic distortion at high output frequencies than a differential output configuration. Figure 3 displays a simplified diagram of the internal output structure of the MAX5888. Figure 3. Simplified Analog Output Structure WIDEBAND RF TRANSFORMER PERFORMS SINGLE-ENDED TO DIFFERENTIAL CONVERSION. 10 CLKP 25 TO DAC 1:1 SINGLE-ENDED CLOCK SOURCE (e.g., HP 8662A) 25 Clock Inputs (CLKP, CLKN) The MAX5888 features a flexible differential clock input (CLKP, CLKN) operating from separate supplies (VCLK, CLKGND) to achieve the lowest possible jitter performance. The two clock inputs can be driven from a single-ended or a differential clock source. For single-ended operation, CLKP should be driven by a logic source, while CLKN should be bypassed to AGND with a 0.1F capacitor. The CLKP and CLKN pins are internally biased to 1.5V. This allows the user to AC-couple clock sources directly to the device without external resistors to define the DC level. The input resistance of CLKP and CLKN is >5k. See Figure 4 for a convenient and quick way to apply a differential signal created from a single-ended source (e.g., HP 8662A signal generator) and a wideband transformer. These inputs can also be driven from an LVDS-compatible clock source; however, it is recommended to use sinewave or AC-coupled ECL drive for best performance. 0.1F 0.1F CLKN CLKGND Figure 4. Differential Clock Signal Generation Data Timing Relationship Figure 5 shows the timing relationship between differential, digital LVDS data, clock, and output signals. The MAX5888 features a 1.4ns hold, a -1ns setup, and a 1.8ns propagation delay time. There is a 3.5 clockcycle latency between CLKP/CLKN transitioning high/low and IOUTP/IOUTN. LVDS-Compatible Digital Inputs (B0P-B15P, B0N-B15N) The MAX5888 features LVDS receivers on the bus input interface. These LVDS inputs (B0P/N through B15P/N) allow for a low-differential voltage swing with low constant power consumption across a large range of ______________________________________________________________________________________ 3.3V, 16-Bit, 500Msps High Dynamic Performance DAC with Differential LVDS Inputs B0 TO B15 OUTPUT DATA IS UPDATED ON THE FALLING EDGE OF CLKP N N-1 tSETUP MAX5888 DIGITAL DATA IS LATCHED ON THE RISING EDGE OF CLKP N+1 tHOLD N+2 tCH tCL CLKP CLKN tPD IOUT N-5 N-4 N-3 N-2 N-1 Figure 5. Detailed Timing Relationship frequencies. Their differential characteristic supports the transmission of high-speed data patterns without the negative effects of electromagnetic interference (EMI). All MAX5888 LVDS inputs feature on-chip termination with differential 100 resistors. See Figure 6 for a simplified block diagram of the LVDS inputs. A common-mode level of 1.25V and an 800mV differential input swing can be applied to these inputs. Segment Shuffling (SEL0) Segment shuffling can improve the SFDR of the MAX5888. The improvement is most pronounced at higher output frequencies and amplitudes. Note that an improvement in SFDR can only be achieved at the cost of a slight increase in the DAC's noise floor. Pin SEL0 controls the segment-shuffling function. If SEL0 is pulled low, the segment-shuffling function of the DAC is disabled. SEL0 can also be left open, because an internal pulldown resistor helps to deactivate the segment-shuffling feature. To activate the MAX5888 segment-shuffling function, SEL0 must be pulled high. Power-Down Operation (PD) The MAX5888 also features an active-high power-down mode, which allows the user to cut the DAC's digital current consumption to less than 6A and the analog current consumption to less than 0.3mA. A single pin (PD) is used to control the power-down mode (PD = 1) or reactivate the DAC (PD = 0) after power-down. B0P-B15P D Q TO DECODE LOGIC 100 D Q B0N-B15N CLOCK Figure 6. Simplified LVDS-Compatible Input Structure Enabling the power-down mode of the MAX5888 allows the overall power consumption to be reduced to less than 1mW. The MAX5888 requires 10ms to wake up from power-down and enter a fully operational state. Applications Information Differential Coupling Using a Wideband RF Transformer The differential voltage existing between IOUTP and IOUTN can also be converted to a single-ended voltage using a transformer (Figure 7) or a differential amplifier configuration. Using a differential transformer coupled output, in which the output power is limited to 0dBm, can optimize the dynamic performance. However, make sure to pay close attention to the transformer core saturation characteristics when selecting a transformer for the MAX5888. Transformer core saturation can introduce strong 2nd-harmonic distortion, especially at low output frequencies and high signal ______________________________________________________________________________________ 11 MAX5888 3.3V, 16-Bit, 500Msps High Dynamic Performance DAC with Differential LVDS Inputs AVDD DVDD VCLK 50 T2, 1:1 VOUT, SINGLE ENDED IOUTP B0-B15 100 MAX5888 IOUTN 16 T1, 1:1 AGND DGND WIDEBAND RF TRANSFORMER T2 PERFORMS THE DIFFERENTIAL TO SINGLE-ENDED CONVERSION. 50 CLKGND Figure 7. Differential to Single-Ended Conversion Using a Wideband RF Transformer AVDD DVDD VCLK 50 OUTP IOUTP B0-B15 100 MAX5888 IOUTN 16 AGND DGND CLKGND OUTN 50 Figure 8. MAX5888 Differential Output Configuration amplitudes. It is also recommended to center tap the transformer to ground. If no transformer is used, each DAC output should be terminated to ground with a 50 resistor. Additionally, a 100 resistor should be placed between the outputs (Figure 8). If a single-ended unipolar output is desirable, IOUTP should be selected as the output, with IOUTN grounded. However, driving the MAX5888 single ended is not recommended since additional noise is added (from the ground plane) in such configurations. The distortion performance of the DAC depends on the load impedance. The MAX5888 is optimized for a 50 double termination. It can be used with a transformer output as shown in Figure 7 or just one 50 resistor from each output to ground and one 50 resistor between the outputs. This produces a full-scale output power of up to 0dBm depending on the output current setting. Higher termination impedance can be used at the cost of degraded distortion performance and increased output noise voltage. 12 Adjacent Channel Leakage Power Ratio (ACLR) Testing for CDMA- and WCDMA-Based Base Station Transceiver Systems (BTS) The transmitter sections of BTS applications serving CDMA and WCDMA architectures must generate carriers with minimal coupling of carrier energy into the adjacent channels. Similar to the GSM/EDGE model (see the Multitone Testing for GSM/EDGE Applications section in the Applications section), a transmit mask (Tx mask) exists for this application. The spread-spectrum modulation function applied to the carrier frequency generates a spectral response, which is uniform over a given bandwidth (up to 4MHz) for a WCDMA-modulated carrier. A dominant specification is ACLR, a parameter which reflects the ratio of the power in the desired carrier band to the power in an adjacent carrier band. The specification covers the first two adjacent bands, and is measured on both sides of the desired carrier. According to the transmit mask for CDMA and WCDMA architectures, the power ratio of the integrated carrier channel energy to the integrated adjacent channel energy must be >45dB for the first adjacent carrier slot (ACLR 1) and >50dB for the second adjacent carrier slot (ACLR 2). This specification applies to the output of the entire transmitter signal chain. The requirement for only the DAC block of the transmitter must be tighter, with a typical margin of >15dB, requiring the DAC's ACLR 1 to be better than 60dB. Adjacent channel leakage is caused by a single-spread spectrum carrier, which generates intermodulation (IM) products between the frequency components located within the carrier band. The energy at one end of the carrier band generates IM products with the energy from the opposite end of the carrier band. For single-carrier WCDMA modulation, these IMD products are spread 3.84MHz over the adjacent sideband. Four contiguous WCDMA ______________________________________________________________________________________ 3.3V, 16-Bit, 500Msps High Dynamic Performance DAC with Differential LVDS Inputs -25 -30 -40 The transmitter sections of multicarrier base station transceiver systems for GSM/EDGE usually present communication DAC manufacturers with the difficult task of providing devices with higher resolution, while simultaneously reducing noise and spurious emissions over a desired bandwidth. To specify noise and spurious emissions from base stations, a GSM/EDGE Tx mask is used to identify the DAC requirements for these parameters. This mask shows that the allowable levels for noise and spurious emissions are dependent on the offset frequency from the transmitted carrier frequency. The GSM/EDGE mask and its specifications are based on a single active carrier with any other carriers in the transmitter being disabled. Specifications displayed in Figure 11 support per-carrier output power levels of 20W or greater. Lower output power levels yield less stringent emission requirements. For GSM/EDGE applications, the DAC demands spurious emission levels of less than -80dBc for offset frequencies 6MHz. Spurious products from the DAC can combine with both random noise and spurious products from other circuit elements. The spurious products from the DAC should therefore be backed off by 6dB more to allow for these other sources and still avoid signal clipping. -30 fCENTER = 61.44MHz fCLK = 184.32Mbps ACLR = 73dB -40 -50 OUTPUT POWER (dBm) -50 OUTPUT POWER (dBm) Multitone Testing for GSM/EDGE Applications -60 -70 -80 -90 -100 -60 -70 -80 -90 -100 -110 -110 -120 -120 -125 fCENTER = 61.44MHz fCLK = 184.32Mbps ACLR = 65dB -130 3.5MHz/div Figure 9. ACLR for WCDMA Modulation, Single Carrier 3.5MHz/div Figure 10. ACLR for WCDMA Modulation, Four Carriers *Note that due to their own IM effects and noise limitations, spectrum analyzers introduce ACLR errors, which can falsify the measurement. For a single-carrier ACLR measurement greater than 70dB, these measurement limitations are significant, becoming even more restricting for multicarrier measurement. Before attempting an ACLR measurement, it is recommended consulting application notes provided by major spectrum analyzer manufacturers that provide useful tips on how to use their instruments for such tests. ______________________________________________________________________________________ 13 MAX5888 carriers spread their IM products over a bandwidth of 20MHz on either side of the 20MHz total carrier bandwidth. In this four-carrier scenario, only the energy in the first adjacent 3.84MHz side band is considered for ACLR 1. To measure ACLR, drive the converter with a WCDMA pattern. Make sure that the signal is backed off by the peak-to-average ratio, such that the DAC is not clipping the signal. ACLR can then be measured with the ACLR measurement function built into your spectrum analyzer. Figure 9 shows the ACLR performance for a single WCDMA carrier (fCLK = 184.32MHz, fOUT = 61.44MHz) applied to the MAX5888 (including measurement system limitations*). Figure 10 illustrates the ACLR test results for the MAX5888 with a four-carrier WCDMA signal at an output frequency of 61.44MHz and sampling frequency of 184.32MHz. Again, the noise floor of the instrument restricts the signal's real dynamic range of the signal, and the measured ACLR 1 understates the actual by more than 2.5dB. Considerable care must be taken to ensure accurate measurement of this parameter. MAX5888 3.3V, 16-Bit, 500Msps High Dynamic Performance DAC with Differential LVDS Inputs The number of carriers and their signal levels with respect to the full scale of the DAC are important as well. Unlike a full-scale sine wave, the inherent nature of a multitone signal contains higher peak-to-RMS ratios, raising the prospect for potential clipping, if the signal level is not backed off appropriately. If a transmitter operates with four/eight in-band carriers, each individual carrier must be operated at less than -12dB FS/-18dB FS to avoid waveform clipping. The noise density requirements (Table 2) for a GSM/EDGE-based system can again be derived from the system's Tx mask. With a worst-case noise level of -80dBc at frequency offsets of 6MHz and a measurement bandwidth of 100kHz, the minimum noise density per hertz is calculated as follows: SNRMIN = -80dBc - 10 log10(100 103Hz) SNRMIN = -130dBc/Hz Since random DAC noise adds to both the spurious tones and to random noise from other circuit elements, it is recommended reducing the specification limits by about 10dB to allow for these additional noise contributions while maintaining compliance with the Tx mask values. Other key factors in selecting the appropriate DAC for the Tx path of a multicarrier GSM/EDGE system is the converter's ability to offer superior IMD and MTPR performance. Multiple carriers in a designated band generate unwanted intermodulation distortion between the individual carrier frequencies. A multitone test vector usually consists of several equally spaced carriers, usually four, with identical amplitudes. Each of these carriers is representative of a channel within the defined bandwidth of interest. To verify MTPR, one or more tones are removed such that the intermodulation distortion perfor- Table 2. GSM/EDGE Noise Requirements for Multicarrier Systems NUMBER OF CARRIERS CARRIER POWER LEVEL (dB FS) DAC NOISE DENSITY REQUIREMENT (dB FS/Hz) 2 -6 -146 4 -12 -152 8 -18 -158 mance of the DAC can be evaluated. Nonlinearities associated with the DAC create spurious tones, some of which may fall back into the area of the removed tone, limiting a channel's carrier-to-noise ratio. Other spurious components falling outside the band of interest can also be important, depending on the system's spectral mask and filtering requirements. Going back to the GSM/EDGE Tx mask, the IMD specification for adjacent carriers varies somewhat among the different GSM standards. For the PCS1800 and GSM850 standards, the DAC must meet an average IMD of -70dBc. Table 3 summarizes the dynamic performance requirements for the entire Tx signal chain in a four-carrier GSM/EDGE-based system and compares the previously established converter requirements with a new-generation high dynamic performance DAC. The four-tone MTPR plot in Figure 12 demonstrates the MAX5888's excellent dynamic performance. The center frequency (fCENTER = 31.97MHz) has been removed to allow detection and analysis of intermodulation or spurious components falling back into this empty spot from adjacent channels. The four carriers are observed over a 12MHz bandwidth and are equally spaced at 1MHz. Each individual output amplitude is backed off to -12dB FS. Under these conditions, the DAC yields an MTPR performance of -78dBc. Table 3. Summary of Important AC Performance Parameters for Multicarrier GSM/EDGE Systems SPECIFICATION SYSTEM TRANSMITTER OUTPUT LEVELS DAC REQUIREMENTS WITH MARGINS MAX5888 SPECIFICATIONS SFDR 80dBc 86dBc 82dBc* SNR -130dBc/Hz -152dB FS/Hz -165dB/Hz IMD -70dBc -75dBc -78dBc N/S -12dB FS -12dB FS Carrier Amplitude *Measured within a 25MHz window. 14 ______________________________________________________________________________________ 3.3V, 16-Bit, 500Msps High Dynamic Performance DAC with Differential LVDS Inputs MAX5888 4-TONE MULTITONE POWER RATIO PLOT (fCLK = 300MHz, fCENTER = 31.9702MHz) 0 AOUT = -12dB FS BW = 12MHz -10 INBAND 30kHz 100kHz -60 -20 4-TONE MTPR (dBm) MEASUREMENT BANDWIDTH -30 OUTBAND TRANSMITTER EDGE AMPLITUDE (dBc) O fT3 fT1 -30 fT4 fT2 -40 -50 -60 -70 IMD REQUIREMENT: < -70dBc -70 -73 -75 -80 -80 WORST-CASE NOISE LEVEL -90 -90 -100 26 0.2 0.4 0.6 1.2 1.8 28 30 6.0 FREQUENCY OFFSET FROM CARRIER (MHz) Figure 11. GSM/EDGE Tx Mask Grounding, Bypassing, and Power-Supply Considerations Grounding and power-supply decoupling can strongly influence the performance of the MAX5888. Unwanted digital crosstalk may couple through the input, reference, power supply, and ground connections, affecting dynamic performance. Proper grounding and powersupply decoupling guidelines for high-speed, high-frequency applications should be closely followed. This reduces EMI and internal crosstalk that can significantly affect the dynamic performance of the MAX5888. Use of a multilayer printed circuit (PC) board with separate ground and power-supply planes is recommended. High-speed signals should run on lines directly above the ground plane. Since the MAX5888 has separate analog and digital ground buses (AGND, CLKGND, and DGND, respectively), the PC board should also have separate analog and digital ground sections with only one point connecting the two planes. Digital signals should be run above the digital ground plane and analog/clock signals above the analog/clock ground plane. Digital signals should be kept as far away from sensitive analog inputs, reference inputs sense lines, common-mode input, and clock inputs as fT1 = 30.0659MHz fT2 = 31.0181MHz fT3 = 33.0688MHz fT4 = 34.0209MHz 32 34 36 38 fOUT (MHz) Figure 12. 4-Tone MTPR Test Results, fCENTER = 31.97MHz, fCLK = 300MHz practical. A symmetric design of clock input and analog output lines is recommended to minimize 2nd-order harmonic distortion components and optimize the DAC's dynamic performance. Digital signal paths should be kept short and run lengths matched to avoid propagation delay and data skew mismatches. The MAX5888 supports three separate power-supply inputs for analog (AVDD), digital (DVDD), and clock (VCLK) circuitry. Each AVDD, DVDD, and VCLK input should at least be decoupled with a separate 0.1F capacitor as close to the pin as possible and their opposite ends with the shortest possible connection to the corresponding ground plane (Figure 13). Try to minimize the analog and digital load capacitances for optimized operation. All three power-supply voltages should also be decoupled at the point they enter the PC board with tantalum or electrolytic capacitors. Ferrite beads with additional decoupling capacitors forming a pi network could also improve performance. The analog and digital power-supply inputs AV DD , VCLK, and DVDD of the MAX5888 allow a supply voltage range of 3.3V 5%. ______________________________________________________________________________________ 15 MAX5888 3.3V, 16-Bit, 500Msps High Dynamic Performance DAC with Differential LVDS Inputs DAC. An array of at least 4 4 vias (0.3mm diameter per via hole and 1.2mm pitch between via holes) is recommended for this 68-lead QFN-EP package. The MAX5888 is packaged in a 68-lead QFN-EP package (package code: G6800-4), providing greater design flexibility, increased thermal efficiency**, and optimized AC performance of the DAC. The exposed pad (EP) enables the user to implement grounding techniques, which are necessary to ensure highest performance operation. 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 PC board with standard infrared (IR) flow soldering techniques. A specially created land pattern on the PC board, matching the size of the EP (6mm 6mm), ensures the proper attachment and grounding of the DAC. Designing vias*** into the land area and implementing large ground planes in the PC board design allow for highest performance operation of the Static Performance Parameter Definitions Integral Nonlinearity (INL) Integral nonlinearity is the deviation of the values on an actual transfer function from either a best straight line fit (closest approximation to the actual transfer curve) or a line drawn between the end points of the transfer function, once offset and gain errors have been nullified. For a DAC, the deviations are measured at every individual step. Differential Nonlinearity (DNL) Differential nonlinearity is the difference between an actual step height 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. **Thermal efficiency is not the key factor, since the MAX5888 features low-power operation. The exposed pad is the key element to ensure a solid ground connection between the DAC and the PC board's analog ground layer. ***Vias connect the land pattern to internal or external copper planes. It is important to connect as many vias as possible to the analog ground plane to minimize inductance. BYPASSING--DAC LEVEL BYPASSING--BOARD LEVEL AVDD AVDD VCLK FERRITE BEAD 1F 0.1F 0.1F AGND 10F 47F ANALOG POWERSUPPLY SOURCE 47F DIGITAL POWERSUPPLY SOURCE 47F CLOCK POWERSUPPLY SOURCE CLKGND DVDD OUTP B0-B15 FERRITE BEAD MAX5888 1F 16 10F OUTN 0.1F VCLK FERRITE BEAD DGND 1F DVDD 10F Figure 13. Recommended Power-Supply Decoupling and Bypassing Circuitry 16 ______________________________________________________________________________________ 3.3V, 16-Bit, 500Msps High Dynamic Performance DAC with Differential LVDS Inputs Gain Error A gain error is the difference between the ideal and the actual full-scale output voltage on the transfer curve, after nullifying the offset error. This error alters the slope of the transfer function and corresponds to the same percentage error in each step. Settling Time The settling time is the amount of time required from the start of a transition until the DAC output settles its new output value to within the converter's specified accuracy. Glitch Energy Glitch impulses are caused by asymmetrical switching times in the DAC architecture, which generates undesired output transients. The amount of energy that appears at the DAC's output is measured over time and usually specified in the pV-s range. Dynamic Performance Parameter Definitions Signal-to-Noise Ratio (SNR) For a waveform perfectly reconstructed from digital samples, the theoretical maximum SNR is the ratio of the fullscale analog output (RMS value) to the RMS quantization error (residual error). The ideal, theoretical minimum can be derived from the DAC's resolution (N bits): SNRdB = 6.02dB N + 1.76dB Spurious-Free Dynamic Range (SFDR) SFDR is the ratio of RMS amplitude of the carrier frequency (maximum signal components) to the RMS value of their next-largest distortion component. SFDR is usually measured in dBc and with respect to the carrier frequency amplitude or in dB FS with respect to the DAC's full-scale range. Depending on its test condition, SFDR is observed within a predefined window or to Nyquist. Two-/Four-Tone Intermodulation Distortion (IMD) The two-tone IMD is the ratio expressed in dBc (or dB FS) of either input tone to the worst 3rd-order (or higher) IMD products. Note that 2nd-order IMD products usually fall at frequencies that can be easily removed by digital filtering; therefore, they are not as critical as 3rd-order IMDs. The two-tone IMD performance of the MAX5888 was tested with the two individual input tone levels set to at least -6dB FS and the four-tone performance was tested according to the GSM model at an output frequency of 32MHz and amplitude of -12dB FS. Adjacent Channel Leakage Power Ratio (ACLR) Commonly used in combination with WCDMA, ACLR reflects the leakage power ratio in dB between the measured power within a channel relative to its adjacent channel. ACLR provides a quantifiable method of determining out-of-band spectral energy and its influence on an adjacent channel when a bandwidth-limited RF signal passes through a nonlinear device. Chip Information TRANSISTOR COUNT: 10,629 PROCESS: CMOS However, noise sources such as thermal noise, reference noise, clock jitter, etc., affect the ideal reading; therefore, SNR is computed by taking the ratio of the RMS signal to the RMS noise, which includes all spectral components minus the fundamental, the first four harmonics, and the DC offset. ______________________________________________________________________________________ 17 MAX5888 Offset Error The offset error is the difference between the ideal and the actual offset current. For a DAC, the offset point is the value at the output for the two midscale digital input codes with respect to the full scale of the DAC. This error affects all codes by the same amount. 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 MAX5888 3.3V, 16-Bit, 500Msps High Dynamic Performance DAC with Differential LVDS Inputs PACKAGE OUTLINE, 68L QFN, 10x10x0.9 MM 1 C 21-0122 2 * *MAX5888 Package Code 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. 18 ____________________Maxim Integrated Products, 120 San Gabriel Drive, Sunnyvale, CA 94086 408-737-7600 (c) 2003 Maxim Integrated Products Printed USA is a registered trademark of Maxim Integrated Products.