ADC12EU050 Ultra-Low Power, Octal, 12-bit, 40-50 MSPS Sigma-Delta Analog-to-Digital Converter General Description Features The ADC12EU050 is a 12-bit, ultra-low power, octal A/D converter for use in high performance analog to digital applications. The ADC12EU050 uses an innovative continuous time sigma delta architecture offering ultra low power consumption and an alias free sample bandwidth up to 25MHz. The input stage of each channel features a proprietary system to ensure instantaneous recovery from overdrive. Instant overload recovery (IOR) with no memory effect guarantees the elimination of phase errors resulting from out of range input signals. The ADC12EU050 reduces interconnection complexity by using programmable serialized outputs which offer the industry standard LVDS and SLVS modes. Power consumption of only 48mW per channel @ 50MSPS gives a total chip power consumption of 384mW. The ADC12EU050 can operate entirely from a 1.2V supply, although a separate output driver supply of up to 1.8V can be used. The device operates from -40 to +85 °C and is supplied in a 10 x 10 mm2, 68 pin package. ■ ■ ■ ■ ■ ■ ■ CT∑Δ ADC architecture with 40-50MSPS throughput Anti-alias filter free Nyquist sample range Unique Instant Overload Recovery (IOR) Wide 2.10 VPP input range 1.2V supply voltage Integrated precision LC PLL Serial control via SPI compatible interface Key Specifications ■ Resolution ■ Conversion Rate ■ SNR ■ THD ■ ■ ■ ■ Per Channel Power Total Active Power Inter-Channel Isolation Operating Temp. Range 12 Bits 40 to 50 MSPS 69.3 dBFS (typ) @ 50 MSPS fIN = 4.4MHz –76.6 dB (typ) @ 50 MSPS fIN = 4.4MHz 48 mW/ch (typ) @ 50MSPS 385 mW (typ) @ 50MSPS >110 dB @ fIN = 4.4MHz -40 to +85 °C Applications ■ ■ ■ ■ © 2008 National Semiconductor Corporation 300511 Medical imaging, ultrasound Industrial ultrasound, such as non-destructive testing Communications Battery powered portable systems www.national.com ADC12EU050 Ultra-Low Power, Octal, 12-bit, 40-50 MSPS Sigma-Delta Analog-to-Digital Converter November 6, 2008 ADC12EU050 Block Diagram 30051102 www.national.com 2 ADC12EU050 Connection Diagram 30051101 Ordering Information Industrial (−40°C ≤ TA ≤ +85°C) Package ADC12EU050CIPLQ 68 Pin LLP ADC12EU050EB Evaluation Board 3 www.national.com ADC12EU050 Pin Descriptions Pin No. Name Type Function and Connection Input Differential analog inputs to the ADC, for channels 0 to 7. The negative input pin may be connected via a capacitor to AGND or the inputs may be transformer coupled for single ended operation. Differential inputs are recommended for best performance. ANALOG I/O 2 3 67 68 64 65 61 62 58 59 55 56 52 53 49 50 VIN0+ VIN0VIN1+ VIN1VIN2+ VIN2VIN3+ VIN3VIN4+ VIN4VIN5+ VIN5VIN6+ VIN6VIN7+ VIN7- 4 VREFB Optional negative reference voltage to improve multi-channel ADC matching. This pin must be connected to AGND. VREFT Optional positive reference voltage to improve multi-channel ADC matching. If using the internal reference, this pin should be left tied to AGND through a 100nF capacitor. If using an external reference voltage, this pin should be connected to the positive reference voltage, which must lie in the range specified in the Electrical Characteristics table. DCAP This pin provides the capacitance for the low pass filter in the modulator’s DAC. It must be connected to AGND through a minimum 100nF capacitor. It is possible to decrease the noise close to the carrier by increasing this capacitor, up to a maximum of 10μF. See Applications Information for further information on the selection of this capacitor. 5 6 7 Input External bias reference resistor. This pin must always be connected to AGND through a resistor, whether the internal reference or an external reference voltage is used. The resistor value must be 10kΩ ±1%. RREF Input/Output RST Input This pin is an active low reset for the entire ADC, both analog and digital components. The pin must be held low for 500ns then returned to high in order to ensure that the chip is reset correctly. Input Sleep mode. Toggling this pin to high will cause the ADC to enter the low power sleep mode. When the pin is returned to low, the chip will, after the specified time to exit sleep mode, return to normal operation. DIGITAL I/O 9 10 www.national.com SLEEP 4 Name 15 16 18 19 20 21 23 24 25 26 28 29 31 32 33 34 DO0+ DO0DO1+ DO1DO2+ DO2DO3+ DO3DO4+ DO4DO5+ DO5DO6+ DO6DO7+ DO7- 36 37 BCLK+ BCLK- ADC12EU050 Pin No. Type Function and Connection Output Differential Serial Outputs for channels 0 to 7. Each pair of outputs provides the serial output for the specific channel. The default output is reduced common mode LVDS format, but by programming the appropriate control registers, the output format can be changed to SLVS or LVDS. By programming TX_term (bit 4) in the LVDS Control register, it is possible to internally terminate these outputs with 100 ohm resistors. Output Bit clock. Differential output clock used for sampling the serial outputs. Information on timing can be seen in the Electrical Specifications section of the datasheet. By programming TX_term (bit 4) in the LVDS Control register, it is possible to internally terminate these outputs with 100 ohm resistors. 38 39 WCLK+ WCLK- Output Word Clock. Differential output frame clock. Information on timing can be seen in the Electrical Specifications section of the datasheet. By programming TX_term (bit 4) in the LVDS Control register, it is possible to internally terminate these outputs with 100 ohm resistors. 44 SDATA Input/Output SPI data input and output. This pin is used to send and receive SPI address and data information. The direction of the pin is controlled internally by the ADC based on the SPI protocol. 45 SCLK Input SPI clock. In order to use the SPI interface, a clock must be provided on this pin. See Electrical Specifications for SPI clock and timing information. 46 SSEL Input SPI chip select. This active low pin is used to enable the serial interface. 47 48 CLK+ (SE) CLK- Input Differential Input Clock. The input clock must lie in the range of 40MHz to 50MHz. It is used by the PLL to generate the internal sampling clocks. A single ended clock can also be used, and should be connected to pin 47. 1, 8, 51, 54, 57, 60, 63, 66 VA Power Analog Power Supply. All pins should be connected to the same 1.2V supply, with voltage limits as in the Electrical Specification. 0 AGND Ground Analog Ground Return. 11, 12, 42, 43 VD Power Digital Power Supply. Connect to 1.2V, with voltage limits as in the Electrical Specification. 13, 14, 22, 30, 40, 41 DGND Ground Digital and Output Driver Ground Return. 17, 27, 35 VDR Power Output Driver Power Supply. Can be connected to 1.2V – 1.8V, depending on application requirements. Voltage limits are described in more detail in the Electrical Specification. POWER SUPPLY 5 www.national.com ADC12EU050 Storage Temperature Range −65°C to +125°C Soldering process must comply with National Semiconductor's Reflow Temperature Profile specifications. Refer to www.national.com/packaging. Absolute Maximum Ratings (Notes 1, 3) If Military/Aerospace specified devices are required, please contact the National Semiconductor Sales Office/ Distributors for availability and specifications. Supply Voltage (VA, VD) IO Supply Voltage (VDR) Voltage at Analog Inputs Voltage at SPI Inputs Input Current, VIN+, VINInput Current, other pins ESD Susceptibility Human Body Model Machine Model Charged Device Mode Soldering Temperature Infrared, 10 seconds Operating Ratings −0.3V to 1.4V -0.3 to 2.0V -0.3 to 2.0V -0.3 to 2.5V ±1mA ±10mA (Notes 2, 3) Operating Temperature Range Supply Voltage (VA=VD) IO Supply Voltage (VDR) Minimum rise time on VA, VD, VDR at power-up Analog Inputs (VINN+, VINN-) SPI Inputs (SDATA, SSEL, SCLK) VREFT (When using external reference) VREFB VCM Input Common Mode Range (Differential Input) CLK Input Frequency External Clock Duty Cycle Ground Difference |AGNDDGND| 2500V 250V 1,250V 235°C −40°C to +85°C +1.14 to +1.26V +1.14 to +1.89V 40µs -0.10 to VA +1.14 to +2.50V 475mV to 525mV AGND 0.4V to 1.2V 40 to 50 MHz 20% to 80% <50mV Electrical Characteristics Unless otherwise specified, the following conditions apply: VA = VD = 1.2V; VDR = 1.2V; VREF = internal; RREF = 10kohm ±1%; CL = 5pF; 100Ω terminated at the receiver; fCLK = 50MHz; fS = 50MSPS. Boldface limits apply for TA = TMIN to TMAX; All other limits apply for TA = +25°C. Symbol Parameter Conditions Typical (Note 4) Limits Units 12 Bits LSB Static Converter Characteristics Resolution (No missing codes guaranteed) INL Integral Non Linearity ±0.75 ±3.0 DNL Differential Non Linearity ±0.35 ±0.75 LSB PSE Positive Full Scale Error ±0.66 ±3 %FS NSE Negative Full Scale Error ±0.58 ±3 %FS GE Gain Error ±1.23 %FS Dynamic Converter Characteristics – Instant Overload Recovery (IOR) Off SNR SINAD ENOB THD Signal to Noise Ratio(Note 5) Signal to Noise and Distortion(Note 5) Effective Number of Bits Total Harmonic Distortion www.national.com fCLK = 50MHz, fIN = 4.4MHz, VIN = -0.5dBFS 69.3 fCLK = 50MHz, fIN = 9.5MHz, VIN = -0.5dBFS 69.0 dBFS fCLK = 40MHz, fIN = 4.4MHz, VIN = -0.5dBFS 69.9 dBFS fCLK = 40MHz, fIN = 9.5MHz, VIN = -0.5dBFS 69.6 fCLK = 50MHz, fIN = 4.4MHz, VIN = -0.5dBFS 68.5 fCLK = 50MHz, fIN = 9.5MHz, VIN = -0.5dBFS 68.5 dBFS fCLK = 40MHz, fIN = 4.4MHz, VIN = -0.5dBFS 69.3 dBFS fCLK = 40MHz, fIN = 9.5MHz, VIN = -0.5dBFS 69.2 fCLK = 50MHz, fIN = 4.4MHz, VIN = -0.5dBFS 11.1 fCLK = 50MHz, fIN = 9.5MHz, VIN = -0.5dBFS 11.1 Bits fCLK = 40MHz, fIN = 4.4MHz, VIN = -0.5dBFS 11.2 Bits fCLK = 40MHz, fIN = 9.5MHz, VIN = -0.5dBFS 11.2 fCLK = 50MHz, fIN = 4.4MHz, VIN = -0.5dBFS -76 fCLK = 50MHz, fIN = 9.5MHz, VIN = -0.5dBFS -78 dBc fCLK = 40MHz, fIN = 4.4MHz, VIN = -0.5dBFS -77 dBc fCLK = 40MHz, fIN = 9.5MHz, VIN = -0.5dBFS -79 dBc 6 67.0 dBFS (min) dBFS 62.5 dBFS (min) dBFS 10.1 Bits (min) Bits -65 dBc (max) H2 H3 SFDR IMD Parameter Second Harmonic Distortion Third Harmonic Distortion Spurious Free Dynamic Range Intermodulation Distortion Typical (Note 4) Limits Units fCLK = 50MHz, fIN = 4.4MHz, VIN = -0.5dBFS -81 -66 dBc (max) fCLK = 50MHz, fIN = 9.5MHz, VIN = -0.5dBFS -79 dBc fCLK = 40MHz, fIN = 4.4MHz, VIN = -0.5dBFS -82 dBc fCLK = 40MHz, fIN = 9.5MHz, VIN = -0.5dBFS -80 dBc fCLK = 50MHz, fIN = 4.4MHz, VIN = -0.5dBFS -83 fCLK = 50MHz, fIN = 9.5MHz, VIN = -0.5dBFS -97 dBc fCLK = 40MHz, fIN = 4.4MHz, VIN = -0.5dBFS -84 dBc fCLK = 40MHz, fIN = 9.5MHz, VIN = -0.5dBFS -108 dBc fCLK = 50MHz, fIN = 4.4MHz, VIN = -0.5dBFS 77 fCLK = 50MHz, fIN = 9.5MHz, VIN = -0.5dBFS 78 dBc fCLK = 40MHz, fIN = 4.4MHz, VIN = -0.5dBFS 78 dBc fCLK = 40MHz, fIN = 9.5MHz, VIN = -0.5dBFS 79 dBc -70 dBFS Conditions f1 = 9.6MHz, VIN = -6dBFS f2 = 10.1MHz, VIN = -6dBFS -67 66 dBc (max) dBc (min) Dynamic Converter Characteristics – Instant Overload Recovery (IOR) On SNR SINAD ENOB THD H2 H3 SFDR IMD Signal-to-Noise Ratio(Note 5) Signal-to-Noise and Distortion(Note 5) Effective Number of Bits Total Harmonic Disortion Second Harmonic Distortion Third Harmonic Distortion Spurious Free Dynamic Range Intermodulation Distortion fCLK = 50MHz, fIN = 4.4MHz, VIN = -0.5dBFS 67.6 fCLK = 50MHz, fIN = 9.5MHz, VIN = -0.5dBFS 67.4 dBFS fCLK = 40MHz, fIN = 4.4MHz, VIN = -0.5dBFS 68.4 dBFS fCLK = 40MHz, fIN = 9.5MHz, VIN = -0.5dBFS 68.2 dBFS fCLK = 50MHz, fIN = 4.4MHz, VIN = -0.5dBFS 67.0 fCLK = 50MHz, fIN = 9.5MHz, VIN = -0.5dBFS 67.0 dBFS fCLK = 40MHz, fIN = 4.4MHz, VIN = -0.5dBFS 67.9 dBFS fCLK = 40MHz, fIN = 9.5MHz, VIN = -0.5dBFS 67.8 dBFS fCLK = 50MHz, fIN = 4.4MHz, VIN = -0.5dBFS 10.8 fCLK = 50MHz, fIN = 9.5MHz, VIN = -0.5dBFS 10.8 Bits fCLK = 40MHz, fIN = 4.4MHz, VIN = -0.5dBFS 11.0 Bits fCLK = 40MHz, fIN = 9.5MHz, VIN = -0.5dBFS 11.0 Bits fCLK = 50MHz, fIN = 4.4MHz, VIN = -0.5dBFS -76 fCLK = 50MHz, fIN = 9.5MHz, VIN = -0.5dBFS -77 dBc fCLK = 40MHz, fIN = 4.4MHz, VIN = -0.5dBFS -77 dBc fCLK = 40MHz, fIN = 9.5MHz, VIN = -0.5dBFS -78 dBc fCLK = 50MHz, fIN = 4.4MHz, VIN = -0.5dBFS -80 fCLK = 50MHz, fIN = 9.5MHz, VIN = -0.5dBFS -78 dBc fCLK = 40MHz, fIN = 4.4MHz, VIN = -0.5dBFS -81 dBc fCLK = 40MHz, fIN = 9.5MHz, VIN = -0.5dBFS -79 dBc fCLK = 50MHz, fIN = 4.4MHz, VIN = -0.5dBFS -83 fCLK = 50MHz, fIN = 9.5MHz, VIN = -0.5dBFS -96 dBc fCLK = 40MHz, fIN = 4.4MHz, VIN = -0.5dBFS -85 dBc fCLK = 40MHz, fIN = 9.5MHz, VIN = -0.5dBFS -107 dBc fCLK = 50MHz, fIN = 4.4MHz, VIN = -0.5dBFS 76 fCLK = 50MHz, fIN = 9.5MHz, VIN = -0.5dBFS 77 dBc fCLK = 40MHz, fIN = 4.4MHz, VIN = -0.5dBFS 78 dBc fCLK = 40MHz, fIN = 9.5MHz, VIN = -0.5dBFS 78 dBc -70 dBFS f1 = 9.6MHz, VIN = -6dBFS f2 = 10.1MHz, VIN = -6dBFS 65.5 61.5 9.9 -64 -65 -67 65 dBFS (min) dBFS (min) Bits (min) dBc (max) dBc (max) dBc (max) dBc (min) Inter-channel Characteristics Channel to channel gain match ±0.1 7 ±0.5 dB www.national.com ADC12EU050 Symbol ADC12EU050 Symbol Parameter Inter-channel Isolation Conditions fIN = 4.4MHz @ -0.1dBFS Typical (Note 4) Limits Units 110 dB IOR mode off 2.10 VPP IOR mode on 1.56 VPP Adjacent channel terminated Reference And Analog Input Characteristics VIN Full Scale Analog Input Voltage Maximum Input for Instantaneous Recovery from Overload RIN Differential Input Impedance VCM Internal Input Common Mode IOR mode on, fIN < 12MHz 2.61 Generated internally Input Impedance of VReft VREFOUT Internal Reference Voltage 2.75 605 2.4 kΩ (min) 2.8 kΩ (max) 574 637 mV (min) mv (max) 20 Generated internally VPP (max) kΩ 502 480 520 mV (min) mV (max) Power Characteristics IA Analog Supply Current fCLK = 50 MHz 152 163 mA (max) ID Digital Supply Current fCLK = 50 MHz 130 147 mA (max) IDR Output Driver Supply Current LVDS, VDR = 1.8V, fCLK = 50 MHz 38 45 mA (max) fCLK = 50 MHz, Equalizer off 385 412 mW (max) fCLK = 50 MHz, Equalizer on 435 470 mW (max) fCLK = 40 MHz, Equalizer off 343 mW fCLK = 40 MHz, Equalizer on 383 mW Sleep 40 50 mW (max) Power Down 5 15 mW (max) fCLK = 50 MHz, Equalizer off 48 mW Power consumption Per channel power consumption fCLK = 40 MHz, Equalizer off 43 mW PSRR Power supply rejection ratio 100mV, 100kHz to 1MHz sinusoid on VA 65 dB CMRR Common mode rejection ratio 100mV, 1MHz sinusoid on VIN+ and VIN- 60 dB Recovery time from sleep 12 µs (max) Recovery time from power down 18 ms (max) Recovery time from single channel power down 6 µs (max) www.national.com 8 Unless otherwise specified, the following conditions apply: VA = VD = 1.2V; VDR = 1.2V; VREF = internal; RREF = 10kohm ±1%; CL = 5pF; 100Ω terminated at the receiver; fCLK = 50MHz; fS = 50MSPS. Boldface limits apply for TA = TMIN to TMAX; All other limits apply for TA = +25°C. Symbol Parameter Pass band Pass band transition Conditions Typical (Note 4) Units fCLK = 50MHz 22 MHz fCLK = 40MHz 17.6 MHz fCLK = 50MHz, -3dB attenuation 25 MHz fCLK = 40MHz, -3dB attenuation 20 MHz Pass band ripple fIN < 22MHz Stop band fCLK = 50MHz 34.5 fCLK = 40MHz 27.6 ±0.01 Stop band attenuation Group delay ripple (peak to peak) Limits fIN < 22MHz, Equalizer on 9 dB MHz MHz 72 dB (min) 0.05 Samples (max) www.national.com ADC12EU050 Digital Decimation Filter Characteristics ADC12EU050 External Input Clock and PLL Characteristics Unless otherwise specified, the following conditions apply: VA = VD = 1.2V; VDR = 1.2V; VREF = internal; RREF = 10kohm ±1%; CL = 5pF; 100Ω terminated at the receiver; fCLK = 50MHz; fS = 50MSPS. Boldface limits apply for TA = TMIN to TMAX; All other limits apply for TA = +25°C. Symbol Parameter Conditions Typical (Note 4) Limits Units 40 50 MHz (min) MHz (max) 20 80 % (min) % (max) External Input Clock fCLK External input clock frequency tCLK Input clock period 1/fCLK ns fCLK DC External clock duty cycle tCH Input clock high time 0.45* fCLK tCL Input clock low time 0.45* fCLK tCR Input clock rise time 0.05* fCLK ns tCF Input clock fall time 0.05* fCLK ns 300 fs tJIN RMS Clock Jitter 50 Generated by PLL. Integrated from 0MHz to BWloop 0.2* tCLK ns (min) 0.8* tCLK ns (max) 0.8* tCLK ns (min) 0.2* tCLK ns (max) PLL f∑Δ Over-sampling frequency BWloop PLL Loop filter bandwidth tJ RMS Bit Clock Jitter www.national.com 640 800 Low Bandwidth 400 High Bandwidth 1.4 kHz MHz 2 10 MHz (min) MHz (max) ps peak Unless otherwise specified, the following conditions apply: VA = VD = 1.2V; VDR = 1.2V; VREF = internal; RREF = 10kohm ±1%; CL = 5pF; 100Ω terminated at the receiver; fCLK = 50MHz; fS = 50MSPS. Boldface limits apply for TA = TMIN to TMAX; All other limits apply for TA = +25°C. Symbol Parameter Conditions Typical (Note 4) Limits Units Digital Inputs (SDATA, SSEL, SCLK, SLEEP, RST) VIH Logical input “1” voltage Test run at 2MHz 900 mV (min) VIL Logical input “0” voltage Test run at 2MHz 300 mV (max) IIN1 Logical "1" Input current 1 µA (max) IIN0 Logical "0" Input current -1 µA (min) CIN Input Capacitance Guaranteed by design 5 pF Digital Outputs (SDATA) VOH Logical output “1” voltage Test run at 2MHz, VDR = 1.2V VDR 850 mV (min) VOL Logical output “0” voltage Test run at 2MHz, VDR = 1.2V DRGND IOH Logical "1" Output Current 250 mV (max) -0.75 mA (min) IOL Logical "0" Output Current 1 mA (max) Output Drive Capability (SDATA) CLOAD Load capacitance R = 4.7 kohm, VDR > 1.8V 50 pF R = 4.7 kohm, VDR = 1.2V 50 pF Open Drain Mode VEXT Maximum allowed external voltage on Open Drain mode activated SDATA RSDATA Recommended SDATA external pullup resistor Open Drain mode activated 11 2.5 4.7 V kΩ www.national.com ADC12EU050 Digital Input and Output Characteristics ADC12EU050 AC and Timing Characteristics Unless otherwise specified, the following conditions apply: VA = VD = 1.2V; VDR = 1.2V; VREF = internal; RREF = 10kohm ±1%; CL = 5pF; 100Ω terminated at the receiver; fCLK = 50MHz; fS = 50MSPS. Boldface limits apply for TA = TMIN to TMAX; All other limits apply for TA = +25°C. Symbol Parameter Conditions Typical (Note 4) Limits Units 40 50 MSPS (min) MSPS (max) General ADC Output Timing Parameters fs Sample Rate 19 Samples tBCLK Conversion Latency Bit clock period fCLK = 50MHz 3.33 ns tWCLK Word clock period fCLK = 50MHz 20 ns tS Data valid setup time 830 325 ps (min) tH Data valid hold time 900 470 ps (min) tR, tF Output Rise/Fall time 320 tDFS Data Edge to Word Edge Skew -295 -720 220 ps (min) ps (max) ps LVDS Output Parameters, OCM = 0 (VDR = 1.2V) LVDS mode, I_drive[1:0] = 00 (2.5mA), RL = 270 mV 100Ω VOD Differential Output Voltage LVDS mode, I_drive[1:0] = 01 (3.5mA), RL = 370 100Ω LVDS mode, I_drive[1:0] = 11 (5.0mA), RL = 318 428 520 mV (min) mV (max) mV 100Ω VOCM Output Common Mode Voltage LVDS mode, OCM = 0 (for VDR = 1.2V) 945 895 1000 mV (min) mV (max) LVDS Output Parameters, OCM = 1 (VDR = 1.8V) LVDS mode, I_drive[1:0] = 00 (2.5mA), RL = 265 mV 100Ω VOD Differential Output Voltage LVDS mode, I_drive[1:0] = 01 (3.5mA), RL = 350 100Ω LVDS mode, I_drive[1:0] = 11 (5.0mA), RL = 280 417 485 mV (min) mV (max) mV 100Ω VOCM Output Common Mode Voltage LVDS mode, OCM = 1 1265 1200 mV (min) 1340 mv (max) SLVS Output Parameters SLVS mode, I_drive[1:0] = 00 (2.5mA), RL = 245 mV 100Ω VOD Differential Output Voltage SLVS mode, I_drive[1:0] = 01 (3.5mA), RL = 330 100Ω SLVS mode, I_drive[1:0] = 11 (5.0mA), RL = 262 393 475 mV (min) mV (max) mV 100Ω VOCM Output Common Mode Voltage www.national.com SLVS mode 225 12 185 270 mV (min) mV (max) Unless otherwise specified, the following conditions apply: VA = VD = 1.2V; VDR = 1.2V; VREF = internal; RREF = 10kohm ±1%; CL = 5pF; 100Ω terminated at the receiver; fCLK = 50MHz; fS = 50MSPS. Boldface limits apply for TA = TMIN to TMAX; All other limits apply for TA = +25°C. Symb ol Parameter Conditions Typical (Note 4) Limits Units Serial Interface tSSELS SSEL setup time 250 tSSELH SSEL hold time 250 tWS SDATA setup time, write transaction 250 tWH SDATA hold time, write transaction tSCLK SCLK period tSCLKL SCLK low time 450 ns (min) tSCLKH SCLK high time 450 ns (min) tSCLKR SCLK rise time 50 ns tSCLKF SCLK fall time 50 ns 500 ns tSSELHI SSEL high time Applies to read and write transactions ns ns 15 ns (max) 250 10 ns (max) 1 0.2 µs (min) tRS SDATA valid setup time, read transaction 100 -5 ns (min) tRH SDATA valid hold time, read transaction 250 10 ns (min) Note 1: Absolute maximum ratings are those values beyond which the safety of the device cannot be guaranteed. They are not meant to imply that the device should be operated at these limits. Note 2: Operating Ratings indicate conditions for which the device is guaranteed to be functional, but do not guarantee specific performance limits. Guaranteed specifications and test conditions are specified in the Electrical Characterisitcs section. Operation of the device beyond the Operating Ratings is not recommended as it may degrade the device lifetime. Note 3: All voltages are measured with respect to GND = AGND = DGND = 0V, unless otherwise specified. Note 4: Typical figures are at TA = 25°C, and represent most likely parametric norms at the time of product characterization. The typical specifications are not guaranteed. Note 5: This parameter is specified in dBFS. This indicates the value which would be obtained with a full-scale input. 13 www.national.com ADC12EU050 AC and Timing Characteristics (Serial Interface) ADC12EU050 Timing Diagrams 30051103 FIGURE 1. LVDS/SLVS Output Timing 30051129 FIGURE 2. Output Level Definitions www.national.com 14 ADC12EU050 30051104 FIGURE 3. SPI Write Timing 30051105 FIGURE 4. SPI Read Timing 15 www.national.com ADC12EU050 Typical Performance Characteristics Unless otherwise specified, the following conditions apply: VA = VD = 1.2V; VDR = 1.2V; VREF = internal; CL = 5pF; fCLK = 50MHz; fS = 50MSPS; fIN = 10MHz. Units for SNR and SINAD are dBFS. Units for SFDR and Distortion are dBc. DNL INL 30051130 30051131 Spectral Response @ fIN=10MHz, fCLK= 40MHz, IOR off Spectral Response @ fIN=10MHz, fCLK= 40MHz, IOR on 30051132 30051133 Spectral Response @ fIN=10MHz, fCLK= 50MHz, IOR on Spectral Response @ fIN=10MHz, fCLK= 50MHz, IOR off 30051134 www.national.com 30051135 16 Unless otherwise specified, the following conditions apply: VA = VD = 1.2V; VDR = 1.2V; VREF = internal; CL = 5pF; fCLK = 50MHz; fS = 50MSPS; fIN = 10MHz. Units for SNR and SINAD are dBFS. Units for SFDR and Distortion are dBc. SNR, SINAD, SFDR vs fCLK, IOR off Distortion vs fCLK, IOR off 30051139 30051140 SNR, SINAD, SFDR vs fCLK, IOR on Distortion vs fCLK, IOR on 30051141 30051142 SNR, SINAD, SFDR vs VA, fCLK = 40MHz, IOR off Distortion vs VA, fCLK = 40MHz, IOR off 30051143 30051153 17 www.national.com ADC12EU050 Typical Performance Characteristics ADC12EU050 Typical Performance Characteristics Unless otherwise specified, the following conditions apply: VA = VD = 1.2V; VDR = 1.2V; VREF = internal; CL = 5pF; fCLK = 50MHz; fS = 50MSPS; fIN = 10MHz. Units for SNR and SINAD are dBFS. Units for SFDR and Distortion are dBc. SNR, SINAD, SFDR vs VA, fCLK = 50MHz, IOR off Distortion vs VA, fCLK = 50MHz, IOR off 30051144 30051154 SNR, SINAD, SFDR vs Temperature, fCLK = 40MHz, IOR off Distortion vs Temperature, fCLK = 40MHz, IOR off 30051156 30051157 SNR, SINAD, SFDR vs Temperature, fCLK = 50MHz, IOR off Distortion vs Temperature, fCLK = 50MHz, IOR off 30051160 www.national.com 30051161 18 Unless otherwise specified, the following conditions apply: VA = VD = 1.2V; VDR = 1.2V; VREF = internal; CL = 5pF; fCLK = 50MHz; fS = 50MSPS; fIN = 10MHz. Units for SNR and SINAD are dBFS. Units for SFDR and Distortion are dBc. SNR, SINAD, SFDR vs fIN, fCLK = 40MHz, IOR off Distortion vs fIN, fCLK = 40MHz, IOR off 30051145 30051146 SNR, SINAD, SFDR vs fIN, fCLK = 40MHz, IOR on Distortion vs fIN, fCLK = 40MHz, IOR on 30051147 30051148 SNR, SINAD, SFDR vs fIN, fCLK = 50MHz, IOR off Distortion vs fIN, fCLK = 50MHz, IOR off 30051149 30051150 19 www.national.com ADC12EU050 Typical Performance Characteristics ADC12EU050 Typical Performance Characteristics Unless otherwise specified, the following conditions apply: VA = VD = 1.2V; VDR = 1.2V; VREF = internal; CL = 5pF; fCLK = 50MHz; fS = 50MSPS; fIN = 10MHz. Units for SNR and SINAD are dBFS. Units for SFDR and Distortion are dBc. SNR, SINAD, SFDR vs fIN, fCLK = 50MHz, IOR on Distortion vs fIN, fCLK = 50MHz, IOR on 30051151 30051152 Spectral Response @ fIN1= 9.6MHz, fIN2= 10.1MHz, IOR off Spectral Response @ fIN1= 9.6MHz, fIN2= 10.1MHz, IOR on 30051136 30051137 Histogram of output code for zero input Current vs fCLK, Equalizer off, LVDS output 30051138 www.national.com 30051155 20 ADC12EU050 Functional Description The ADC12EU050 employs a number of unique strategies to provide a high performance multi-channel ADC that offers a significant power consumption reduction when compared to compteting architectures, as well as easing system level design. The ultra-low power performance of the ADC12EU050 is derived from the implementation of a fast continuous time sigma delta (CT∑Δ) modulator. Other features of this technology are: • Intrinsic anti-alias filter – the digital decimating filter provides an intrinsic anti-alias filter, eliminating external analog filter components, and simplifying multi-channel designs. • Instant overload recovery (IOR) system guarantees extremely fast recovery from overload (<1ps), and no settling errors on return from overload. • Ultra-low inter-channel crosstalk. • Digital Equalizer provides low group delay and hence minimizes signal path delay variation. The major signal path blocks are: clipping control; CT∑Δ modulator; digital decimation filter; 12 bit serializer; and finally the LVDS/SLVS outputs. The PLL is critical to the operation of the ADC12EU050, and the PLL also provides the bit and word clock outputs. The SPI Control Interface gives uncomplicated user access to the ADC registers. 30051123 FIGURE 5. SHA Input Stage 1.0 12-BIT SIGMA DELTA ADC CORE The ADC12EU050 comprises eight analog ADC channels using a CT∑Δ architecture, which provides very high dynamic performance with ultra-low power, while operating from a minimal 1.2V supply. The CT∑Δ ADC architecture uses a third order sigma delta modulator operating at a nominal 16 times over-sampling rate in combination with a 3-bit quantizer. The modulator output is coupled to a power efficient digital decimation filter that decimates the high rate modulator output (640 to 800MHz) to provide output data at a sample rate between 40 MSPS and 50 MSPS. A benefit of the CT∑Δ design is that the ADC requires no external anti-alias filters for most applications. This benefit is derived from a combination of the design of the analog sigma delta modulator and digital decimation filter. The digital filter achieves a steep transition band, and provides 72 dB of attenuation in the stop band. Using the digital equalizer, the signal transfer characteristics including phase performance can be optimized so as to minimise group delay variation. In applications where it is not required, the digital equalizer can be disabled to further save power. 30051124 FIGURE 6. Continuous Time Sigma Delta Input Stage 1.2 INSTANT OVERLOAD RECOVERY The ADC12EU050 features an overload handling system which provides instantaneous recovery from signals driving the ADC inputs beyond the full-scale input range. The ADC can operate in two different modes. In the default ADC mode (IOR mode off) a full-scale input range of 2.10 VPP is supported, here the ADC operates with some inherent overload recovery time, similar to a conventional ADC. In the IOR mode, the ADC has a reduced 1.56 VPP full scale input range, but provides a significant benefit in that the ADC can now be driven by input voltages as high as 5 dB beyond the nominal full-scale (fIN < 12MHz), that is 2.75 VPP, and will recover instantaneously. In a number of applications this feature can help simplify input stage design and manufacturing set-up and calibration. The ADC12EU050 recovers immediately from overload with no missing codes and no settling time. The proprietary strategy used within the ADC12EU050 uses high speed patented clamp techniques to limit the input signal and keep it within the stable input range of the ADC. This process happens at a speed equivalent to the on-chip oversampling rate of 640 to 800 MHz. The advantage of this system is that it responds immediately to out of range signals. While the inputs are over-range the ADC outputs a full scale result. As the over-range input is removed the ADC adjusts to the input signal level and is able to provide sampled data instantaneously. The ADC’s behaviour on emerging from overload is repeatable and independent of whether the input signal was positive or negative going at the point of overload. The diagram below shows a 5dB overloaded input (2.75 VPP versus 1.56 VPP Full scale), with 240,000 sample periods 1.1 DIFFERENTIAL INPUT STAGE The ADC can capture high speed analog signals without resorting to a complex fast sample-and-hold amplifier (SHA) as used in pipeline ADCs. This is where CT∑Δ technology derives much of its power and performance benefits. This feature also assists external circuit design. In the case of the SHA inputs of pipeline ADCs, the effective input capacitance is time variant, requiring a powerful input buffer to drive to the resolution limits of the system. The input stage of the ADC is purely resistive (1.3kΩ single ended) driving into virtual earth. As a result the ADC12EU050 is extremely easy to drive as its input impedance is not complex. It also means that external lower power input buffering circuitry can used, and can be completely eliminated in some cases. 21 www.national.com ADC12EU050 overlaid. There is no ringing and recovery from overload is instantaneous. 30051106 FIGURE 7. Instant Overload Recovery 1.3 INTEGRATED PRECISION LC PLL The ADC12EU050 family includes an integrated high performance “clean up” phase locked loop (PLL), simplifying the need for a low jitter external clock. The PLL serves three important functions; it generates a highly accurate internal sampling clock source of up to 800 MHz; a clock for the LVDS serializers at 600 MHz; and it provides a low jitter clock for other internal components. With its jitter clean-up capability this PLL allows lower performance system clocks to be used. 1.4 DIGITAL DECIMATION FILTER AND EQUALIZER The digital decimation filter is an integral part of the sigma delta architecture. It decimates the over-sampled data from the modulator down to the sample rate, and its extremely sharp low pass characteristic combined with the modulator’s broad band response provides the intrinsic anti-alias filter. The digital low pass filter exhibits 72dB of attenuation in the stop band. The following diagram shows the digital filter transfer function at 40MSPS, compared to a third order Butterworth transfer function. Due to the digital implementation of the filter, the filter parameters automatically scale with the ADC sampling frequency. www.national.com 30051107 FIGURE 8. Digital Filter Transfer Function 22 1.6 POWER MANAGEMENT MODES The ADC12EU050 operates normally at ultra-low power levels. In addition, several power management modes are provided: • Power Down (accessible through PD bit of Top Control Register) • Sleep (SLEEP pin, or SLEEP bit of Top Control Register) • Single channel power down (PD0-7 of ADC/LVDS Channel Power Down Register) Power Down is the lowest power consumption mode, but with a longer wake-up time than Sleep mode. In power down mode, all circuits in the chip are turned off, including the PLL, reference and bias circuits. Power consumption in Sleep mode is higher than in Power Down mode, but pin access (SLEEP pin) and fast wake-up enables duty cycle powering of the ADC. The device also allows channel by channel power-down through the ADC/LVDS Channel Power Down register. When a single channel is powered down, the sigma delta modulator, digital decimating filter and LVDS outputs for that channel will be shut off, with the corresponding single channel reduction in power consumption. 1.7 SPI CONTROL INTERFACE The ADC12EU050 provides configurability via the serial control interface. This provides IOR mode control power management control, output configuration control, data output test patterns to provide LVDS/SLVS training sequences, as well as many other user configurable options. Full details of the SPI registers can be found in the Programming Guide section of this datasheet. The SPI pins (SDATA, SCLK, SSEL), as well as the pins RST and SLEEP, have been designed to operate with voltage levels up to 2.5V, despite the low 1.2V core voltage. As a result, no external level shift components are required for this control interface. 30051108 FIGURE 9. Group Delay with Equalizer Off 1.8 UNCORRELATED NOISE REFERENCE FOR EACH CHANNEL In many early multi-channel ADC designs, a single voltage reference was used to provide the reference level for each channel. Unfortunately, this ensures that the noise at each ADC’s reference terminal is cross correlated. Multi-channel systems often make use of a 3 dB processing gain increase that results from each channel doubling. Without a specific technique to prevent the reference terminals seeing correlated noise the expected 3 dB gain is compromised. In the case of the ADC12EU050, a unique system has been implemented to de-correlate the noise at each ADC channel. 30051109 FIGURE 10. Group Delay with Equalizer On 1.5 SERIAL DATA OUTPUTS Sampled data is transformed into high speed serial LVDS/ SLVS output data streams. The low amplitude differential signal swings of LVDS/SLVS help to reduce digital system noise. It is possible to select between LVDS and SLVS modes by simple programming through the SPI control interface. The output common mode can also be programmed through the SPI control interface, allowing it to be adjusted based on the value of VDR. 23 www.national.com ADC12EU050 Such steep digital filters introduce group delay problems, but the ADC12EU050 includes a digital equalizer, which reduces group delay ripple variation to less than 0.05 samples. In applications where group delay is not of concern, the equalizer can be turned off through the SPI interface in order to save power. The following two diagrams show the group delay ripple of the digital decimation filter at 50MSPS, firstly with the equalizer disabled, and secondly with the equalizer enabled. ADC12EU050 differential clock, the clock traces should be routed as 100Ω differential pairs, and terminated with a 100Ω resistor close to the chip. A single ended clock input should be connected to pin 47 (CLK+/SE), and pin 48 (CLK-) should be grounded. On-chip PLL The benefit of having an on chip PLL is that in most applications a high precision clock source is not required. The impact of aperture jitter on the ADC’s performance is reduced dramatically by the jitter clean-up properties of the PLL, which ensures that any RMS jitter outside of the PLL bandwidth is attenuated. The PLL also significantly relaxes the input clock duty cycle requirements, accepting input clock duty cycles of 20% to 80%. The PLL offers two choices of bandwidth. For the majority of systems, the default bandwidth of 400kHz is suitable. If the system already contains a high performance clock, with excellent RMS jitter performance up to a 1.4MHz bandwidth, then the PLL’s high bandwidth mode may be used. Application Information 2.0 POWER-UP SEQUENCE The ADC12EU050 has three separate power supplies: Analog (VA), Digital (VD) and the output drive voltage ( VDR). The ADC contains a power on reset circuit, connected to VA, and so to ensure correct reset of both analog and digital logic of the ADC, the power supplies should be provided in the following order: 1. VDR 2. VD 3. VA If this order is not followed, then the user should issue a reset via the reset pin (RST) immediately after power up. Additionally, it is required that the rise time for each voltage supply is longer than the minimum rise time stated in the Electrical Specifications section of this data sheet. There is no required sequence for powering down the ADC. 2.1 ADC START-UP SEQUENCE After any reset, either power-on reset, software reset via SPI or hardware reset via the RST pin, the chip undergoes a series of internal calibrations and the PLL/VCO will lock to the external clock. After reset, the ADC12EU050’s registers have the default values shown in register tables. The registers can be programmed via the SPI after reset, even during the period while the chip is performing the internal calibrations mentioned in the previous paragraph. During reset and until the PLL is locked, the LVDS outputs will not provide valid data. Furthermore, the ADC has an inherent data conversion latency, which is related to the pipeline stages of the digital decimating filter. Until the data conversion latency has passed, the data outputs will be invalid. Thus the maximum time until valid sampled data is received at the outputs is: PLL lock time + ADC Latency Specific values for these times can be found in the Electrical Specifications section of this datasheet. 30051115 FIGURE 11. PLL Phase Noise Transfer Function: fs = 40MHz 2.2 USING ADC LOW POWER MODES As explained previously in the Functional Description, the ADC12EU050 offers several power management modes. Sleep mode offers the fastest wake-up time, and should be used in applications where duty cycle powering of the ADC is required. In this case it is recommended to toggle sleep mode via the SLEEP pin, which will give a faster cycle time than programming the SLEEP bit through the SPI, due to the extra time required to send a command through the SPI port. The Power Down mode is accessible via the SPI port. Due to the power-up time of the ADC coupled with the programming time of the SPI port, this mode should be used to power the chip down for longer time periods. Channel power down allows one or more channels to be turned off independently, with the corresponding power saving. 30051116 2.3 CLOCK SELECTION CONSIDERATIONS The ADC12EU050 has an on-chip PLL, which simplifies the task of clock source selection and clock network design. Clock Input Connection The ADC is designed to accept either single ended or differential clock inputs. In order to guarantee the specified performance, a differential clock should be used. When using a www.national.com FIGURE 12. PLL Phase Noise Transfer Function: fs = 50MHz On the input clock, excessive RMS jitter within the PLL bandwidth will be seen in the output spectrum as sidebands, or close in phase noise, around the fundamental signal. 24 Differential Input Configurations The ADC12EU050 can be driven either actively or passively. The LMH6512 offers a combined LNA/VGA with differential output, which performs well with the ADC12EU050 in systems requiring input signal amplification, such as Medical Ultrasound. Transformer coupling provides another possibility for converting a single ended signal into a differential signal. The diagram below shows a transformer coupled input configuration. 2.4 ADC INPUT CONSIDERATIONS The ADC12EU050’s sigma delta architecture offers many flexible options for connecting input signals. In order to obtain maximum performance from the device, it is recommended to use a differential input connection. The device, however, also supports single ended analog input. 30051121 FIGURE 13. Transformer Coupled Input mum input voltage allowed is 3dB less than the 2.10V full scale input. The diagram below shows a single ended input configuration. Single Ended Input Configurations In cost sensitive applications, a single ended input may provide adequate performance, however ADC performance will degrade slightly. When using single ended inputs, the maxi- 30051120 FIGURE 14. Single Ended Input 25 www.national.com ADC12EU050 Input Clock Selection For systems which do not have a requirement for a high performance clock, any standard product 40MHz – 50MHz crystal oscillator will allow the ADC12EU050 to perform to specifications. If the system requires high performance clocks for other system components, then National Semiconductor's LMK family of clock conditioners are recommended. ADC12EU050 Due to the purely resistive input circuit of the sigma delta architecture, the ADC12EU050 allows the user to scale down large input signals by adding external series resistors. The gain achieved by adding external resistors can be calculated as a simple voltage divider, as follows: VFS/VIN = 20 * log (RADC /( RADC + REXT)) dB The diagram below shows this configuration, and defines the values in the equation above. Input Coupling and Common Mode The ADC12EU050 internally generates a common mode of 0.62V. It is possible to provide input signals with other common modes however, the full scale input range of the ADC must be kept in mind. For this reason, it is recommended that the device inputs are AC coupled. The recommended capacitor value is 100nF. External Series Resistance 30051118 FIGURE 15. External Series Resistance RADC, the input resistance of the ADC, is nominally 1.3kΩ. Due to manufacturing the value of this resistance can vary by up to 15%. This is not important for the operation of the ADC, since the ADC depends only on internal resistors being matched, but it should be taken into account when performing calculations. 2.5 ADC OUTPUT CONSIDERATIONS The ADC12EU050 offers a variety of output settings in order to cater for different system design and integration needs. Output Driver Voltage, VDR The ADC output driver voltage, VDR, can be set between 1.2V and 1.8V. A VDR of 1.2V will offer the lowest power consumption. Because VDR can be varied, the ADC12EU050 provides, via the SPI registers, the ability to adjust the output common mode voltage. Output Modes And Output Common Mode Three different output modes are also supported: SLVS, LVDS and reduced common mode LVDS. SLVS and LVDS modes output data according to their respective specifications. Reduced common mode LVDS must be used when the output driver voltage, VDR, is 1.2V. The standard LVDS common mode voltage is 1.2V, which is obviously not feasible if VDR is 1.2V. Therefore, the output common mode voltage must be set to 1.0V by setting the bit OCM in the LVDS Control Register to 0. www.national.com 30051126 FIGURE 16. Output Driver Circuit: Reduced Common Mode LVDS When VDR is 1.8V, the standard LVDS common mode voltage of 1.2V must be used, by setting OCM equal to 1. 26 30051125 FIGURE 17. Output Driver Circuit: LVDS SLVS mode offers the lowest power consumption, followed by reduced common mode LVDS then standard LVDS. 30051127 FIGURE 18. Output Driver Circuit: SLVS 27 www.national.com ADC12EU050 As well as the different output modes, the output drive current can also be controlled via the LVDS Control Register. The default output drive current is 2.5mA, but this can be increased to 3.5mA or 5mA, depending on output trace routing and receiver requirements. Power consumption of the ADC12EU050 will increase slightly as the output driver current is increased. Termination The final control feature available in the LVDS Control Register is the choice between internal and external 100Ω termination. Although the termination is recommended to be as close to the receiver as possible, in some cases it may be necessary or desirable to perform this termination at the transmitter. Internal 100Ω termination at the transmitter (the ADC12EU050) is enabled by setting the bit TX_term to 1. LVDS Output Training Sequences Often it is necessary to calibrate the LVDS receiver, for example an FPGA or DSP, so that skew between the eight ADC output channels is minimized. In order to simplify this process, the ADC12EU050 provides three LVDS training modes, where a pre-defined or custom pattern is output on all eight channels simultaneously. While a training mode is active, the word and bit clocks are output as usual. In order to select a training mode, the TSEL bits of the Decimator Control Register (16h) must be programmed via the SPI interface. There are two pre-defined training patterns, or a custom pattern can be loaded via the SPI into the Serializer Custom Pattern 0 and 1 Registers (10h and 12h). In order to return to normal ADC operation after skew calibration, the TSEL bits should be returned to their default value of 00. ADC12EU050 30051164 FIGURE 19. LVDS Training Select operation The implementation of IOR mode in the ADC consists of analog and digital parts working in tandem. The analog clipping circuitry, controlled by OL, is designed to protect the sigma delta modulator from large signal inputs. Using an analog clamp, signals are soft-limited to the less than the 2.10Vpp full scale range of the modulator. OL gives the value at which the circuit will begin to clamp. The digital filter of the ADC12EU050 is where the full scale input range is selected and the hard limiting of the signal takes place. DGF selects the gain of the digital filter, and hence the new full scale input range of the ADC. In order to set a custom value for DGF, CGS, bit 7 of the Decimator Clipping Control register, must be set. The DGF can then be set, based on the application requirements. OL should then be set to a value approximately half-way between the new full scale input range (which was just selected by DGF) and the default full scale input range of 2.10Vpp. OL must be set to a value higher than DGF, otherwise the signal will be limited by the analog clipping circuitry, rather than the digtal circuitry, and overload recovery will be impacted. 2.6 USING IOR MODE As discussed in the Functional Description, IOR mode provides instantaneous recovery from overload conditions, with no ringing and correct data output as soon as the input returns in range. Standard Use of IOR Mode The recommended way to enable IOR mode is by setting bit 4 (IOR) of the Modulator Overload Control register (04h). Setting this bit will enable IOR mode with the default settings for DGF in the Decimator Clipping Control register (14h) and OL in the Modulator Overload Control register (04h). Setting the IOR mode bit to 0 will restore DGF and OL to their default values, hence putting the chip back into ADC mode. As can be seen in the Electrical Specifications, using IOR mode gives a slight reduction in SNR performance, and also a reduction of the full scale input range to 1.56Vpp differential. Advanced Use of IOR Mode The registers described above allow the user to customize IOR mode. In order to correctly set the DGF and OL values, it is necessary to understand how the IOR mode functions. www.national.com 28 ADC12EU050 30051128 FIGURE 20. IOR Mode Signal Modification When using the internal reference, VREFT should be connected to AGND through a 100nF capacitor, while VREFB must be connected to AGND. Chip-to-chip gain matching between several ADC12EU050 ADCs can be improved by connecting the VREFT pins of the ADCs. This is show in the figure below. 2.7 THE VOLTAGE REFERENCE The ADC provides an on chip, ±5% tolerance voltage reference, together with all necessary biasing circuits and current sources. A 10kΩ (±1%) resistor must be connected between RREF and AGND in order to establish the biasing current of the ADC. The internal reference voltage, VREF, is available at the RREF pin. 30051122 FIGURE 21. Reference Sharing If a tighter tolerance reference is required for improved thermal stability, an external voltage reference can be connected between the VREFT and VREFB pins. The RREF resistor must be connected even when using an external reference. carrier. The filter circuit, which is a first order RC filter, is shown in the diagram below. 2.8 DCAP CAPACITOR SELECTION The DCAP pin provides the capacitance for the low pass filter between the DAC bias block and the DAC in the sigma-delta modulator. The filter blocks noise from the DAC Bias block from entering the DAC. Any noise which passes through this filter will be seen in the spectrum as side skirts around the 29 www.national.com ADC12EU050 This includes using a split ground plane, since the significant digital portion of the chip can produce noise on the digital/IO ground (DGND). When designing the ADC12EU050 into a system, It is critical that the exposed pad is connected to analog ground (AGND). The exposed pad provides the analog ground connection for the ADC12EU050, and so this connection is required for electrical rather than thermal reasons. It is recommended to decouple the power supplies using a large capacitor (e.g. 47µF) for low frequency noise, and small capacitors (e.g. 100nF) placed close to each supply pin. Analog and digital supplies (VA and VD) may be provided from the same supply, however in this case it is recommended that the supplies are isolated from each other with a ferrite bead or inductor. If the IO driver supply (VDR ) is 1.2V, then it may also be taken from the same supply, with isolation as described above. The clock and data output traces, as well as the clock input trace (when using a differential input clock), should be routed as 100Ω impedance differential pairs. If not using the option for 100Ω internal termination, then the clock and data output trances should be terminated with a 100Ω resistor close to the receiver. If the system requires regulators to provide the ADC12EU050 1.2V operating voltage, National Semiconductor recommends the LP3878SD-ADJ Low Noise “Ceramic Stable” Adjustable Regulator or the LP3879 Low Noise “Ceramic Stable” Regulator. Datasheets for both parts are available from the National Semiconductor website. 30051117 FIGURE 22. DCAP RC Filter The DCAP pin must be connected to AGND through a low leakage, minimum 100nF capacitor. If the application is especially sensitive to close to the carrier phase noise, then it is recommended to increase DCAP, up to a maximum of 10µF. For other applications where close to the carrier phase noise is not important, the capacitor can be kept small in order to reduce costs and minimise board space. The corner frequency of this filter is determined by the equation: f = 1/(2πRDCAPCDCAP) 2.9 BOARD LAYOUT CONSIDERATIONS Proper grounding, layout and routing are essential to ensure accurate conversion in any high speed ADC. Maintaining separate analog and digital areas of the board is recommended in order to achieve the specified performance. www.national.com 30 ADC12EU050 30051110 FIGURE 23. ADC12EU050 Application Diagram 31 www.national.com ADC12EU050 Programming Guide 3.0 THE SERIAL CONTROL INTERFACE The ADC12EU050 provides several user controlled functions which are accessed through a standard SPI compatible, 3 wire Serial Interface, as shown in the diagram below. 30051111 FIGURE 24. Three Wire Control Interface Wired OR mode is supported in order to connect multiple ADC12EU050 devices to one SPI Master. The clock and data buses are common to all ADC devices, and the chip select SSEL is used to control which SPI is currently active. The SPI master must have a unique pin available for each ADC’s SSEL. The diagram below illustrates the connection. 30051112 FIGURE 25. Multi-Wire Control Interface When connecting multiple devices, the SDATA pin must be set in Open Drain mode. Open Drain mode is enabled by setting the SPIOD bit in the Top Control Register of all connected ADC12EU050 devices. When SDATA is in open drain mode, the user must ensure that a pull-up resistor is connected to the SDATA bus. Further details on Open Drain mode are given in . www.national.com 3.1 SERIAL CONTROL INTERFACE PROTOCOL Both read and write transactions are made up of eight address bits and eight data bits. The final address bit of the address phase determines whether the transaction will be a read transaction or a write transaction – logic level low for write, logic level high for read. The following diagram shows the protocol. 32 ADC12EU050 30051113 FIGURE 26. Serial Control Interface Protocol The eight address bits, A[7:1] + R/W, are sent first. The data, D[7:0], is then sent for a write transaction, or D[7:0] is received for a read transaction. Address and data are sent and received with the most-significant-bit (MSB) first. The SPI is enabled using the active low input SSEL. If SSEL is high the SPI cannot be accessed, although SSEL is not a reset signal and registers will maintain their value when SSEL is toggled. SSEL must be held low during the entire transaction. Timing requirements for the Serial Interface are described in the Electrical Characteristics section of this document. be used. In Open Drain mode, the ADC’s SDATA will pull the output low, and SDATA will be pulled up to the external level by a pull-up resistor connected to the board’s positive voltage rail, VEXT. The intended use of Open Drain mode is when the ADC, including VDR, is running at 1.2V, and a VEXT of 1.8V is required. Open Drain mode is enabled by setting the SPIOD bit in the Top Control Register via the Serial Interface. When in Open Drain mode, a pull-up resistor (RSDATA) must be connected between SDATA and VEXT. The table of Electrical Specifications shows the required settings for VEXT and RSDATA. 3.2 SERIAL INTERFACE TRANSACTION CANCELLATION A transaction may be cancelled before the address and data stages are completed by toggling SSEL to high at any stage during an SPI access. This action is not recommended, as transaction cancellation during a write transaction may corrupt register contents and during a read transaction will result in incorrect data. After canceling a transaction with SSEL the ADC may be in an unknown state due to an incomplete and hence corrupted write to a register. It is therefore recommended to reset the chip via Software Reset (SRES) after a cancelled transaction. 3.4 SERIAL CONTROL INTERFACE READ AND WRITE SPEED SCLK (pin 45) controls the speed of interaction with the ADC. The SPI interface supports write to and read from speeds as defined in the Electrical Specifications section of this document. 3.5 SERIAL CONTROL INTERFACE REGISTER DESCRIPTIONS The following tables show the complete set of user accessible SPI registers, with descriptions of the functionality of each bit. Reset values of all registers are also described in the tables below. 3.3 SDATA PAD OPEN DRAIN MODE If the SDATA voltage at the board level is required to be higher than the ADC12EU050’s VDR, the Open Drain mode should 33 www.national.com ADC12EU050 Register Index Address b[7] b[6] b[5] b[4] b[3] b[2] b[1] b[0] Default Reserved CBR 40/50 SRES SPIOD SLEEP PD 00h PD5 PD4 PD3 PD2 PD1 PD0 00h Reserved Reserved IOR OL[3] OL[2] OL[1] OL[0] 00h Reserved Reserved Reserved Reserved Reserved SHBW STCAL 00h 100HYS 50HYS 20HYS Custom Pattern [4] Custom Pattern [3] Custom Pattern [2] Custom Pattern [1] Custom Pattern [0] 00h Reserved Reserved Custom Pattern [11] Custom Pattern [10] Custom Pattern [9] Custom Pattern [8] 00h a[2] a[1] a[0] b[2] b[1] b[0] 00h Reserved Reserved EQON DFS MSB TSEL[1] TSEL[0] 00h Reserved Reserved TX_term I_drive[1] I_drive[0] OCM SLVS 00h ID [6] ID [5] ID [4] ID [3] ID [2] ID [1] ID [0] Top Control Register 00h Reserved ADC / LVDS Channel Power Down Register 02h PD7 PD6 Modulator Overload Control Register 04h Reserved PLL Control Register 08h Reserved LVDS Input Clock Hysteresis 0Ah Reserved Reserved INVCLK 10HYSOFF HYSOFF 00h Serializer Custom Pattern 0 Register 10h Custom Pattern [7] Custom Pattern [6] Custom Pattern [5] Serializer Custom Pattern 1 Register 12h Reserved Reserved Decimator Clipping Control Register 14h Reserved Reserved Decimator Control Register 16h Reserved LVDS Control Register 18h Reserved Chip ID Register 1Eh www.national.com ID [7] 34 ADC12EU050 Top Control Register Address: Attributes 00h Write Only. Register 01h reads back contents of register 00h, if CBR is set. The Top Control Register is the basic initialization and control register for the device. b[7] Description b[6] Reserved Default 0 0 b[5] b[4] b[3] b[2] b[1] b[0] CBR 40/50 SRES SPIOD SLEEP PD 0 0 0 0 0 0 Bit 7:6 5 Description CBR: Control Bus Read. When asserted register 00h (this register) can be read, but no other registers. When de-asserted all other registers can be read, but not register 00h. Register 00h cannot be read from address 01h. All other registers can be read back. Register 00h can be read from address 01h. All other registers cannot be read back. 40/50: Selects the ADC sample rate. This bit should be set according to the applied input clock to obtain optimal performance. 0 1 3 00 h Reserved. Write as zero for future compatibility. 0 1 4 HEX 45-50MSPS 40-45MSPS SRES: Software Reset. When asserted the software reset will reset the whole device. SRES performs the same function as the hardware reset (RST pin). The SRES is self clearing in approximately 2µs. 0 1 2 SPIOD: SPI Open Drain mode. 0 1 1 Digital Logic Output Open Drain Mode. Enables SPI Driver to operate above VDR SLEEP: Sleep Mode. Powers down the device with the exception of the PLL and the reference blocks. The time to wake-up from sleep mode is < 10µs. 0 1 0 Software Reset Inactive Software Reset Active Sleep Mode Inactive Sleep Mode Active PD: Power Down Mode. Completely powers down the device. The power up time is approximately 20ms. 0 1 PD Mode Inactive, device operates normally PD Mode Active, device powered down 35 www.national.com ADC12EU050 ADC / LVDS Channel Power Down Register Address: Attributes 02h Write Only. Register 03h reads back contents of register 02h. The ADC/LVDS Channel Power Down Register provides the capability to independently power down each ADC channel. Description Default b[7] b[6] b[5] b[4] b[3] b[2] b[1] b[0] PD7 PD6 PD5 PD4 PD3 PD2 PD1 PD0 0 0 0 0 0 0 0 0 Bit 7 Description PD7: Power Down Channel 7 0 1 6 PD6: Power Down Channel 6 0 1 5 Channel Active Channel Power Down PD0: Power Down Channel 0 0 1 www.national.com Channel Active Channel Power Down PD1: Power Down Channel 1 0 1 0 Channel Active Channel Power Down PD2: Power Down Channel 2 0 1 1 Channel Active Channel Power Down PD3: Power Down Channel 3 0 1 2 Channel Active Channel Power Down PD4: Power Down Channel 4 0 1 3 Channel Active Channel Power Down PD5: Power Down Channel 0 1 4 Channel Active Channel Power Down Channel Active Channel Power Down 36 HEX 00 h ADC12EU050 Modulator Overload Control Register Address: Attributes 04h Write Only. Register 05h reads back contents of register 04h. b[7] b[6] Description b[5] Reserved Default 0 0 b[4] 4 b[2] 0 0 IOR 0 b[1] b[0] HEX 0 0 00 h OL[3:0] 0 Bit 7:5 b[3] Description Reserved. Write as zero for future compatibility. IOR: Enable IOR Mode (Instant Overload Recovery) This bit can be used to quickly enable IOR mode with the default IOR settings for DGF (see register 14h) and OL. 0 1 3:0 IOR Mode Disabled IOR Mode Enabled OL[3:0]: The bits define the differential peak voltage (in VPP) at which the analog input signal is clipped when in IOR mode. In IOR mode the analog clipping is set to 1.746 VPP. In the default ADC mode clipping of the analog input signal is disabled. Should it be decided to over-ride the default setting, it is important to follow the guidelines for setting OL, as described in the Applications Information section. OL[3:0] Clipping Voltage VPP 0 (IOR Mode default) 1.746 0001 1.694 0010 1.64 0011 1.586 0100 1.534 0101 1.480 0110 1.426 0111 1.374 1000 2.172 1001 2.120 1010 2.066 1011 2.012 1100 1.960 1101 1.906 1110 1.852 1111 1.800 37 www.national.com ADC12EU050 PLL Control Register Address: Attributes 08h Write Only. Register 09h reads back contents of register 08h. b[7] b[6] b[5] 0 0 0 Description Default b[4] 1 b[2] 0 0 Reserved 0 Bit 7:2 b[3] b[1] b[0] SHBW STCAL 0 0 HEX 00 h Description Reserved. Write as zero for future compatibility. SHBW: Set PLL to High Bandwidth. The selection of the PLL bandwidth permits to set the sensitivity of the PLL to input clock jitter. Less bandwidth decreases the sensitivity to input clock jitter. The PLL Bandwidth is related to the sampling frequency, the exact values of which can be found in the electrical specifications table. The PLL will pass any input clock jitter up to the PLL bandwidth, while jitter above the PLL bandwidth will be attenuated. Low bandwidth mode should be used for high jitter input clocks, while high bandwidth mode can be used for high-quality, low jitter input clocks. 0 1 0 STCAL: Start VCO calibration. The calibration can be manually started in order to assure that the frequency tuning margin is maximum, for example, in case of large temperature change during operation it can be useful to restart the calibration. 0 1 www.national.com PLL bandwidth is set to Low Bandwidth (400kHz). PLL bandwidth is set to High Bandwidth (1.4MHz). The VCO calibration starts automatically if a Loss of Lock is detected The VCO calibration is restarted. 38 ADC12EU050 LVDS Input Clock – Hysteresis Address: Attributes 0Ah Write Only. Register 0Bh reads back contents of register 0Ah. b[7] Description Default b[6] Reserved 0 0 b[5] b[4] b[3] b[2] b[1] b[0] INVCLK 100HYS 50HYS 20HYS 10HYS OFF HYSOFF 0 0 0 0 0 0 Bit 7:6 5 Reserved. Write as zero for future compatibility. INVCLK: Invert Input Reference Clock. This bit is used to invert the input clock. Normal operation (10mV hysteresis) 20mV hysteresis. (LVDS input clock only) 10HYSOFF: Disable 10mV hysteresis. 10mV hysteresis is the default setting. This bit is used to disable 10mV hysteresis, in the case where another hysteresis setting is desired, for example when using a CMOS input clock. 0 1 0 Normal operation (10mV hysteresis) 50mV hysteresis. (CMOS input clock only) 20HYS: Enable 20mV hysteresis. This bit enables 20mV hysteresis. It should be used for an LVDS input clock only. 0 1 1 Normal operation (10mV hysteresis) 100mV hysteresis (CMOS input clock only) 50HYS: Enable 50mV hysteresis. This bit enables 50mV hysteresis. It should be used for a CMOS input clock only. 0 1 2 Reference input clock not inverted. Reference input clock inverted. 100HYS: Enable 100mV hysteresis. This bit enables 100mV hysteresis. It should be used for a CMOS input clock only. 0 1 3 00 h Description 0 1 4 HEX 10mV hysteresis. (LVDS input clock only) 10mV hysteresis disabled. HYSOFF: Disable all hysteresis settings. This bit is used to disable all hysteresis settings. 0 1 Normal operation (10mV hysteresis) All hysteresis settings disabled. 39 www.national.com ADC12EU050 Serializer Custom Pattern 0 Register Address: Attributes 10h Write Only. Register 11h reads back contents of Register 10h. This register in conjunction with User Register 12 provides storage for the custom de-skew pattern. See User Register 16 for a description of how this training sequence is used. b[7] b[6] b[5] 0 0 0 Description Default b[4] b[3] b[2] b[1] b[0] HEX 0 0 0 00 h Custom Pattern [7:0] 0 0 Bit Description 7:0 Custom Pattern [7:0]. This pattern forms the lower byte of Custom Pattern [11:0] which is output by the serializer when the Training Sequence Select bits (bits 1:0) of the Decimator Control Register are set to select Training sequence 3. Serializer Custom Pattern 1 Register Address: Attributes 12h Write Only. Register 13h reads back contents of Register 12h. This register in conjunction with User Register 10 provides storage for the custom de-skew pattern. See User Register 16 for a description of how this training sequence is used. b[7] b[6] 0 0 Description Default b[5] b[4] b[3] 0 0 Reserved b[2] b[1] b[0] HEX 0 00 h Custom Pattern [11:8] 0 Bit 0 0 Description 7:4 Reserved. Write as zero for future compatibility. 3:0 Custom Pattern [11:8]. This pattern forms the upper 4 bits of Custom Pattern [11:0] which is output by the serializer when the Training Sequence Select bits (bits 1:0) of the Decimator Control Register are set to select Training sequence 3. www.national.com 40 ADC12EU050 Decimator Clipping Control Register Address: Attributes 14h Write Only. Register 15h reads back contents of Register 14h. Description Default b[7] b[6] CGS Reserved 0 0 b[5] b[4] b[3] b[2] 0 0 a[2:0] 0 b[1] b[0] HEX 0 00 h b[2:0] 0 0 Bit Description 7 CGS: Custom Gain Setting. This bit is used to override the automatic gain settings for ADC and IOR modes. If the user wishes to write a custom digital gain coefficient using a[2:0] and b[2:0] of this register, then the CGS bit must be set. 0 1 6 5:3 Normal operation Automatic gain settings used Custom Gain Setting Gain setting from a[2:0] and b[2:0] used. Reserved. Write as zero for future compatibility. a[2:0]: Digital Gain Coefficient. In clipping mode the input range of an ADC channel is limited to 1.56Vpp. In ADC mode the input range is 2.10Vpp. The output of the digital filter has to be scaled according to the selected mode (the filter data has to be mapped in to the 12bit output data), the difference between 1.6Vpp and 2.2Vpp is -2.6dB, hence the digital filter gain has to be set to 2.6dB when in IOR mode and to 0dB when in clipping mode (default mode) . This is performed by setting a Digital Gain Factor which is calculated using the following formula: The mapping of the coefficient values for a[2:0] is as follows: 011 = Not used. Defaults to 2 010 = 2 001 = 1 000 = 0 111 = -1 110 = -2 101 = Not used. Defaults to -2 100 = Not used. Defaults to -2 The mapping of the coefficient values for b[2:0] is shown below. The table on the following page shows the available Digital Gain Coefficient settings. 2:0 b[2:0]: Digital Gain Coefficient. The mapping of the coefficient values for b[2:0] is as follows: 011 = Not used. Defaults to 2 010 = 2 001 = 1 000 = 0 111 = -1 110 = -2 101 = Not used. Defaults to -2 100 = Not used. Defaults to -2 41 www.national.com ADC12EU050 Coefficent a[2:0] Coefficent b[2:0] Digital Gain (dB) Equivalent full scale input range (VPP) 010 010 4.16 1.30 010 001 3.95 1.33 010 000 3.74 1.37 010 111 3.52 1.40 010 110 3.29 1.44 001 001 3.06 1.48 001 000 2.82 1.52 001 111 2.58 1.56 001 110 2.33 1.61 000 001 2.07 1.65 000 000 1.80 1.71 000 111 1.53 1.76 000 110 1.24 1.82 111 001 0.95 1.88 111 000 0.64 1.95 111 111 0.33 2.02 111 110 000 2.10 110 001 -0.34 2.18 110 000 -0.70 2.28 110 111 -1.07 2.38 110 110 -1.45 2.48 www.national.com 42 IOR Mode default setting ADC mode default setting ADC12EU050 Decimator Control Register Address: Attributes 16h Write Only. Register 17h reads back contents of register 16h. b[7] Description Default b[6] b[5] Reserved 0 0 b[4] b[3] b[2] EQON DFS MSB 0 0 0 0 Bit 7:5 4 TSEL[1:0] 0 0 00 h EQON: Equalizer Enable. This bit is used to enable or disable the digital equalizer. The equalizer can be switched on in order to reduce the group delay of the output data, at the cost of increased power. Equalizer disabled Equalizer enabled DFS: Data Format Select. Selects the format, either Offset Binary or Twos Complement of the output data 2s Complement Offset Binary MSB: Select the bit order of the LVDS output data stream 0 1 1:0 HEX Reserved. Write as zero for future compatibility. 0 1 2 b[0] Description 0 1 3 b[1] LSB first MSB first TSEL[1:0]: Training Sequence Select. These bits select the LVDS output data. The default mode of operation is where the filter output data is serialized. In the remaining modes the selected training sequence is repeatedly output from the serializer this allows the receiving data capture circuitry to perform the de-skewing process. One of three known words can be selected, the first two words are hard-coded in the block, the third one, the custom pattern, is written into User Registers 10h and 12h the Serializer Custom Pattern Registers. Note. The outputs bit-clock and word-clock are not affected by the value of the Training Sequence Select bits. 00 01 10 11 ADC data[11:0] Training sequence 1: 000000111111 Training sequence 2: 101010101010 Training sequence 3: custom pattern 43 www.national.com ADC12EU050 LVDS Control Register Address: Attributes 18h Write Only. Register 19h reads back contents of register 18h. b[7] Description b[6] b[5] Reserved Default 0 0 b[4] b[3] TX_term 0 0 I_drive[1:0] 0 Bit 7:5 4 b[0] OCM SLVS 0 0 0 HEX 00 h Reserved. Write as zero for future compatibility. TX_term: Enable Internal 100 Ohm termination for data outputs. Internal 100 ohm termination disabled Internal 100 ohm termination enabled I_drive[1:0]: Controls the current drive of the data outputs. 00 01 10 11 1 b[1] Description 0 1 3:2 b[2] 2.5 mA 3.5 mA Reserved 5 mA OCM: Output Common mode. Allows the output common mode to be shifted depending on the setting of VDR. If bit 0 of this register, SLVS, is set to 1 then changing OCM will have no impact on the output common mode. The output common mode in SLVS mode is fixed, as described in the Electrical Specifications section of this datasheet. For VDR = 1.2V, OCM must be set to 0. For VDR = 1.8V, OCM must be set to 1. 0 0 Output Common Mode, VOCM = 1.0V 1 Output Common Mode, VOCM = 1.25V SLVS: Select the format for output data, either LVDS or SLVS. The differences in timing and electrical specifications between the two modes can be seen in the Electrical Specifications section of the datasheet. If this bit is set to 1 (SLVS mode), OCM has no effect and the output common mode will be set for SLVS as described in the Electrical Specifications section of this datasheet. When LVDS mode is selected, the output common mode must be selected using the OCM bit of this register. 0 1 www.national.com LVDS Mode SLVS Mode 44 Address: Attributes 1Eh Read Only. b[7] b[6] b[5] b[4] Description Default b[3] 0 0 0 0 0 Bit 7:0 b[2] b[1] b[0] HEX 0 0 0 00 h ID [7:0] Description ID[7:0]: Chip ID Register. Reading from this register will provide the chip version. The expected Chip ID for the ADC12EU050 is 12.4. X = ID[7]*8 + ID[6]*4 + ID[5]*2 + ID[4] Y = ID[3]*8 + ID[2]*4 + ID[1]*2 + ID[0] Chip ID = Version X.Y 45 www.national.com ADC12EU050 Chip ID Register ADC12EU050 Physical Dimensions inches (millimeters) unless otherwise noted 68-Lead LLP Package 10x10x1.0mm, 0.5mm Pitch Ordering Numbers ADC12EU050CIPLQ NS Package Number LQA68A www.national.com 46 ADC12EU050 Notes 47 www.national.com ADC12EU050 Ultra-Low Power, Octal, 12-bit, 40-50 MSPS Sigma-Delta Analog-to-Digital Converter Notes For more National Semiconductor product information and proven design tools, visit the following Web sites at: Products Design Support Amplifiers www.national.com/amplifiers WEBENCH® Tools www.national.com/webench Audio www.national.com/audio App Notes www.national.com/appnotes Clock and Timing www.national.com/timing Reference Designs www.national.com/refdesigns Data Converters www.national.com/adc Samples www.national.com/samples Interface www.national.com/interface Eval Boards www.national.com/evalboards LVDS www.national.com/lvds Packaging www.national.com/packaging Power Management www.national.com/power Green Compliance www.national.com/quality/green Switching Regulators www.national.com/switchers Distributors www.national.com/contacts LDOs www.national.com/ldo Quality and Reliability www.national.com/quality LED Lighting www.national.com/led Feedback/Support www.national.com/feedback Voltage Reference www.national.com/vref Design Made Easy www.national.com/easy PowerWise® Solutions www.national.com/powerwise Solutions www.national.com/solutions Serial Digital Interface (SDI) www.national.com/sdi Mil/Aero www.national.com/milaero Temperature Sensors www.national.com/tempsensors Solar Magic® www.national.com/solarmagic Wireless (PLL/VCO) www.national.com/wireless Analog University® www.national.com/AU THE CONTENTS OF THIS DOCUMENT ARE PROVIDED IN CONNECTION WITH NATIONAL SEMICONDUCTOR CORPORATION (“NATIONAL”) PRODUCTS. NATIONAL MAKES NO REPRESENTATIONS OR WARRANTIES WITH RESPECT TO THE ACCURACY OR COMPLETENESS OF THE CONTENTS OF THIS PUBLICATION AND RESERVES THE RIGHT TO MAKE CHANGES TO SPECIFICATIONS AND PRODUCT DESCRIPTIONS AT ANY TIME WITHOUT NOTICE. NO LICENSE, WHETHER EXPRESS, IMPLIED, ARISING BY ESTOPPEL OR OTHERWISE, TO ANY INTELLECTUAL PROPERTY RIGHTS IS GRANTED BY THIS DOCUMENT. TESTING AND OTHER QUALITY CONTROLS ARE USED TO THE EXTENT NATIONAL DEEMS NECESSARY TO SUPPORT NATIONAL’S PRODUCT WARRANTY. EXCEPT WHERE MANDATED BY GOVERNMENT REQUIREMENTS, TESTING OF ALL PARAMETERS OF EACH PRODUCT IS NOT NECESSARILY PERFORMED. NATIONAL ASSUMES NO LIABILITY FOR APPLICATIONS ASSISTANCE OR BUYER PRODUCT DESIGN. BUYERS ARE RESPONSIBLE FOR THEIR PRODUCTS AND APPLICATIONS USING NATIONAL COMPONENTS. PRIOR TO USING OR DISTRIBUTING ANY PRODUCTS THAT INCLUDE NATIONAL COMPONENTS, BUYERS SHOULD PROVIDE ADEQUATE DESIGN, TESTING AND OPERATING SAFEGUARDS. EXCEPT AS PROVIDED IN NATIONAL’S TERMS AND CONDITIONS OF SALE FOR SUCH PRODUCTS, NATIONAL ASSUMES NO LIABILITY WHATSOEVER, AND NATIONAL DISCLAIMS ANY EXPRESS OR IMPLIED WARRANTY RELATING TO THE SALE AND/OR USE OF NATIONAL PRODUCTS INCLUDING LIABILITY OR WARRANTIES RELATING TO FITNESS FOR A PARTICULAR PURPOSE, MERCHANTABILITY, OR INFRINGEMENT OF ANY PATENT, COPYRIGHT OR OTHER INTELLECTUAL PROPERTY RIGHT. LIFE SUPPORT POLICY NATIONAL’S PRODUCTS ARE NOT AUTHORIZED FOR USE AS CRITICAL COMPONENTS IN LIFE SUPPORT DEVICES OR SYSTEMS WITHOUT THE EXPRESS PRIOR WRITTEN APPROVAL OF THE CHIEF EXECUTIVE OFFICER AND GENERAL COUNSEL OF NATIONAL SEMICONDUCTOR CORPORATION. As used herein: Life support devices or systems are devices which (a) are intended for surgical implant into the body, or (b) support or sustain life and whose failure to perform when properly used in accordance with instructions for use provided in the labeling can be reasonably expected to result in a significant injury to the user. A critical component is any component in a life support device or system whose failure to perform can be reasonably expected to cause the failure of the life support device or system or to affect its safety or effectiveness. National Semiconductor and the National Semiconductor logo are registered trademarks of National Semiconductor Corporation. All other brand or product names may be trademarks or registered trademarks of their respective holders. Copyright© 2008 National Semiconductor Corporation For the most current product information visit us at www.national.com National Semiconductor Americas Technical Support Center Email: [email protected] Tel: 1-800-272-9959 www.national.com National Semiconductor Europe Technical Support Center Email: [email protected] German Tel: +49 (0) 180 5010 771 English Tel: +44 (0) 870 850 4288 National Semiconductor Asia Pacific Technical Support Center Email: [email protected] National Semiconductor Japan Technical Support Center Email: [email protected]