16-Bit, 8-Channel, 500 kSPS PulSAR ADC AD7699 FEATURES APPLICATIONS Battery-powered equipment Medical instruments: ECG/EKG Mobile communications: GPS Personal digital assistants Power line monitoring Data acquisition Seismic data acquisition systems Instrumentation Process control FUNCTIONAL BLOCK DIAGRAM 0.5V TO 4.096V 0.1µF 5V 0.5V TO VDD 10µF REF REFIN BAND GAP REF VDD 1.8V VIO TO VDD AD7699 TEMP SENSOR IN0 IN1 IN2 IN3 IN4 IN5 IN6 IN7 CNV 16-BIT SAR ADC MUX SPI SERIAL INTERFACE ONE-POLE LPF SCK SDO DIN SEQUENCER COM 07354-001 16-bit resolution with no missing codes 8-channel multiplexer with choice of inputs Unipolar single-ended Differential (GND sense) Pseudobipolar Throughput: 500 kSPS INL: ±0.5 LSB typical, ±1.5 LSB maximum (±23 ppm or FSR) Dynamic range: 93.3 dB SINAD: 91.5 dB @ 20 kHz THD: −97 dB @ 20 kHz Analog input range: 0 V to VREF with VREF up to VDD Multiple reference types Internal 4.096 V External buffered (up to 4.096 V) External (up to VDD) Internal temperature sensor Channel sequencer, selectable 1-pole filter, busy indicator No pipeline delay, SAR architecture Single-supply 5 V operation with 1.8 V to 5 V logic interface Serial interface compatible with SPI, MICROWIRE, QSPI, and DSP Power dissipation 26 mW @ 500 kSPS 5.2 μW @ 100 SPS Standby current: 50 nA 20-lead 4 mm × 4 mm LFCSP package GND Figure 1. Table 1. Multichannel 14-/16-Bit PulSAR® ADC Type 14-Bit 16-Bit 16-Bit Channels 8 4 8 250 kSPS AD7949 AD7682 AD7689 500 kSPS AD7699 ADC Driver ADA4841-x ADA4841-x ADA4841-x GENERAL DESCRIPTION The AD7699 is an 8-channel, 16-bit, charge redistribution successive approximation register (SAR) analog-to-digital converter (ADC) that operates from a single power supply, VDD. The AD7699 contains all components for use in a multichannel, low power data acquisition system, including a true 16-bit SAR ADC with no missing codes; an 8-channel low crosstalk multiplexer useful for configuring the inputs as single-ended (with or without ground sense), differential, or bipolar; an internal 4.096 V low drift reference and buffer; a temperature sensor; a selectable one-pole filter; and a sequencer that is useful when channels are continuously scanned in order. The AD7699 uses a simple serial port interface (SPI) for writing to the configuration register and receiving conversion results. The SPI interface uses a separate supply, VIO, which is set to the host logic level. Power dissipation scales with throughput. The AD7699 is housed in a tiny 20-lead LFCSP with operation specified from −40°C to +85°C. Rev. 0 Information furnished by Analog Devices is believed to be accurate and reliable. However, no responsibility is assumed by Analog Devices for its use, nor for any infringements of patents or other rights of third parties that may result from its use. Specifications subject to change without notice. No license is granted by implication or otherwise under any patent or patent rights of Analog Devices. Trademarks and registered trademarks are the property of their respective owners. One Technology Way, P.O. Box 9106, Norwood, MA 02062-9106, U.S.A. www.analog.com Tel: 781.329.4700 Fax: 781.461.3113 ©2008 Analog Devices, Inc. All rights reserved. AD7699 TABLE OF CONTENTS Features .............................................................................................. 1 Voltage Reference Output/Input .............................................. 18 Applications ....................................................................................... 1 Power Supply............................................................................... 19 Functional Block Diagram .............................................................. 1 Supplying the ADC from the Reference.................................. 19 General Description ......................................................................... 1 Digital Interface .............................................................................. 20 Revision History ............................................................................... 2 Reading/Writing During Conversion, Fast Hosts.................. 20 Specifications..................................................................................... 3 Reading/Writing During Acquisition, Any Speed Hosts ...... 20 Timing Specifications....................................................................... 5 Reading/Writing Spanning Conversion, Any Speed Host .... 20 Absolute Maximum Ratings............................................................ 6 Configuration Register, CFG .................................................... 20 ESD Caution .................................................................................. 6 General Timing Without a Busy Indicator ............................. 22 Pin Configuration and Function Descriptions ............................. 7 Typical Performance Characteristics ............................................. 8 Read/Write Spanning Conversion Without a Busy Indicator ...................................................................................... 23 Terminology .................................................................................... 12 General Timing With a Busy Indicator ................................... 24 Theory of Operation ...................................................................... 13 Read/Write Spanning Conversion with a Busy Indicator ..... 25 Overview...................................................................................... 13 Application Hints ........................................................................... 26 Converter Operation .................................................................. 13 Layout .......................................................................................... 26 Transfer Functions...................................................................... 14 Evaluating AD7699 Performance............................................. 26 Typical Connection Diagrams .................................................. 15 Outline Dimensions ....................................................................... 27 Analog Inputs .............................................................................. 16 Ordering Guide .......................................................................... 27 Driver Amplifier Choice ............................................................ 18 REVISION HISTORY 10/08—Revision 0: Initial Version Rev. 0 | Page 2 of 28 AD7699 SPECIFICATIONS VDD = 4.5 V to 5.5 V, VREF = 4.096 to VDD, VIO = 1.8 V to VDD, all specifications TMIN to TMAX, unless otherwise noted. Table 2. Parameter RESOLUTION ANALOG INPUT Voltage Range Absolute Input Voltage Analog Input CMRR Leakage Current at 25°C Input Impedance 1 THROUGHPUT Conversion Rate Full Bandwidth2 ¼ Bandwidth2 Transient Response ACCURACY No Missing Codes Integral Linearity Error Differential Linearity Error Transition Noise Gain Error4 Gain Error Match Gain Error Temperature Drift Offset Error4 Offset Error Match Offset Error Temperature Drift Power Supply Sensitivity AC Accuracy Dynamic Range Signal-to-Noise SINAD Total Harmonic Distortion Spurious-Free Dynamic Range Channel-to-Channel Crosstalk SAMPLING DYNAMICS −3 dB Input Bandwidth Aperture Delay Conditions/Comments Min 16 Unipolar mode Bipolar mode Positive input, unipolar and bipolar modes Negative or COM input, unipolar mode Negative or COM input, bipolar mode fIN = 250 kHz Acquisition phase 0 −VREF/2 −0.1 −0.1 VREF/2 − 0.1 Typ VREF/2 68 1 0 0 Full-scale step, full bandwidth Full-scale step, ¼ bandwidth 16 −1.5 −1 Max Unit Bits +VREF +VREF/2 VREF + 0.1 +0.1 VREF/2 + 0.1 V V V V V dB nA 500 125 400 1600 kSPS kSPS ns ns VDD = 5 V ± 5% fIN = 20 kHz, VREF = 5 V fIN = 20 kHz, VREF = 4.096 V internal REF fIN = 20 kHz, VREF = 5 V fIN = 20 kHz, VREF = 5 V, −60 dB input fIN = 20 kHz, VREF = 4.096 V internal REF fIN = 20 kHz fIN = 20 kHz fIN = 100 kHz on adjacent channel(s) 93.3 92.5 91.5 91.5 33.5 90.5 −97 112 −125 dB5 dB dB dB dB dB dB dB dB 14 3.6 2.5 MHz MHz ns REF = VDD = 5 V All modes −10 −3 All modes −10 −3 Full bandwidth ¼ bandwidth VDD = 5 V Rev. 0 | Page 3 of 28 92 89.5 90 89 +1.5 +1.5 Bits LSB3 LSB LSB LSB LSB ppm/°C LSB LSB ppm/°C LSB ±0.5 ±0.25 0.5 ±1 ±1 ±0.3 ±1 ±1 ±0.3 ±1.5 +10 +3 +10 +3 AD7699 Parameter INTERNAL REFERENCE REF Output Voltage REFIN Output Voltage6 REF Output Current Temperature Drift Line Regulation Long-Term Drift Turn-On Settling Time EXTERNAL REFERENCE Voltage Range Current Drain TEMPERATURE SENSOR Output Voltage7 Temperature Sensitivity DIGITAL INPUTS Logic Levels VIL VIH IIL IIH DIGITAL OUTPUTS Data Format8 Pipeline Delay9 VOL VOH POWER SUPPLIES VDD VIO Standby Current10, 11 Power Dissipation Energy per Conversion TEMPERATURE RANGE12 Specified Performance Conditions/Comments Min Typ Max Unit @ 25°C @ 25°C 4.086 4.096 2.3 ±300 ±10 ±15 50 5 4.106 V V µA ppm/°C ppm/V ppm ms VDD + 0.3 VDD − 0.2 100 V V µA 283 1 mV mV/°C VDD = 5 V ± 5% 1000 hours CREF = 10 µF REF input REFIN input (buffered) 500 kSPS, REF = 5 V 0.5 0.5 @ 25°C −0.3 0.7 × VIO −1 −1 ISINK = +500 µA ISOURCE = −500 µA +0.3 × VIO VIO + 0.3 +1 +1 V V µA µA 0.4 V V 5.5 VDD + 0.3 V V nA µW mW mW nJ VIO − 0.3 Specified performance Specified performance VDD and VIO = 5 V, @ 25°C VDD = 5 V, 100 kSPS throughput VDD = 5 V, 500 kSPS throughput VDD = 5 V, 500 kSPS throughput with internal reference 4.5 1.8 TMIN to TMAX −40 1 50 5.2 26 28 52 29 32 +85 °C See the Analog Inputs section. The bandwidth is set with the configuration register. LSB means least significant bit. With the 5 V input range, one LSB = 76.3 µV. 4 See the Terminology section. These specifications include full temperature range variation but not the error contribution from the reference. 5 All specifications expressed in decibels are referred to a full-scale input FSR and tested with an input signal at 0.5 dB below full scale, unless otherwise specified. 6 This is the output from the internal band gap. 7 The output voltage is internal and present on a dedicated multiplexer input. 8 Unipolar mode: serial 16-bit straight binary. Bipolar mode: serial 16-bit twos complement. 9 Conversion results available immediately after completed conversion. 10 With all digital inputs forced to VIO or GND as required. 11 During acquisition phase. 12 Contact an Analog Devices, Inc., sales representative for the extended temperature range. 2 3 Rev. 0 | Page 4 of 28 AD7699 TIMING SPECIFICATIONS VDD = 4.5 V to 5.5 V, VREF = 4.096 to VDD, VIO = 1.8 V to VDD, all specifications TMIN to TMAX, unless otherwise noted. Table 3. Parameter1 Conversion Time: CNV Rising Edge to Data Available Acquisition Time Time Between Conversions CNV Pulse Width Data Write/Read During Conversion SCK Period SCK Low Time SCK High Time SCK Falling Edge to Data Remains Valid SCK Falling Edge to Data Valid Delay VIO Above 4.5 V VIO Above 3 V VIO Above 2.7 V VIO Above 2.3 V VIO Above 1.8 V CNV Low to SDO D15 MSB Valid VIO Above 4.5 V VIO Above 3 V VIO Above 2.7 V VIO Above 2.3 V VIO Above 1.8 V CNV High or Last SCK Falling Edge to SDO High Impedance CNV Low to SCK Rising Edge DIN Valid Setup Time from SCK Falling Edge DIN Valid Hold Time from SCK Falling Edge Min Max 1.6 1.2 tDSDO + 2 11 11 4 Unit µs ns µs ns µs ns ns ns ns 16 17 18 21 28 ns ns ns ns ns 15 17 18 22 25 32 ns ns ns ns ns ns ns ns ns tEN tDIS tCLSCK tSDIN tHDIN 10 5 5 See Figure 2 and Figure 3 for load conditions. 500µA IOL 1.4V TO SDO CL 50pF 500µA IOH Figure 2. Load Circuit for Digital Interface Timing 70% VIO 30% VIO 2V OR VIO – 0.5V1 2V OR VIO – 0.5V1 0.8V OR 0.5V2 0.8V OR 0.5V2 2V IF VIO ABOVE 2.5V, VIO – 0.5V IF VIO BELOW 2.5V. 0.8V IF VIO ABOVE 2.5V, 0.5V IF VIO BELOW 2.5V. Figure 3. Voltage Levels for Timing Rev. 0 | Page 5 of 28 07354-003 tDELAY tDELAY 1 2 Typ 400 2 10 07354-002 1 Symbol tCONV tACQ tCYC tCNVH tDATA tSCK tSCKL tSCKH tHSDO tDSDO AD7699 ABSOLUTE MAXIMUM RATINGS Table 4. Parameter Analog Inputs INx, 1 COM1 REF, REFIN Supply Voltages VDD, VIO to GND VDD to VIO DIN, CNV, SCK to GND SDO to GND Storage Temperature Range Junction Temperature θJA Thermal Impedance (LFCSP) θJC Thermal Impedance (LFCSP) 1 Rating GND − 0.3 V to VDD + 0.3 V or VDD ± 130 mA GND − 0.3 V to VDD + 0.3 V −0.3 V to +7 V ±7 V −0.3 V to VIO + 0.3 V −0.3 V to VIO + 0.3 V −65°C to +150°C 150°C 47.6°C/W 4.4°C/W Stresses above those listed under Absolute Maximum Ratings may cause permanent damage to the device. This is a stress rating only; functional operation of the device at these or any other conditions above those indicated in the operational section of this specification is not implied. Exposure to absolute maximum rating conditions for extended periods may affect device reliability. ESD CAUTION See the Analog Inputs section. Rev. 0 | Page 6 of 28 AD7699 20 19 18 17 16 VDD IN3 IN2 IN1 IN0 PIN CONFIGURATION AND FUNCTION DESCRIPTIONS 1 2 3 4 5 PIN 1 INDICATOR AD7699 TOP VIEW (Not to Scale) 15 VIO 14 SDO 13 SCK 12 DIN 11 CNV NOTES 1. THE EXPOSED PADDLE IS NOT CONNECTED INTERNALLY. FOR INCREASED RELIABILITY OF THE SOLDER JOINTS, IT IS RECOMMENDED THAT THE PAD BE SOLDERED TO THE GND PLANE. 07354-004 IN4 6 IN5 7 IN6 8 IN7 9 COM 10 VDD REF REFIN GND GND Figure 4. Pin Configuration Table 5. Pin Function Descriptions Pin No. 1, 20 2 Mnemonic VDD REF Type1 P AI/O 3 REFIN AI/O 4, 5 6 to 9 10 GND IN4 to IN7 COM P AI AI 11 CNV DI 12 DIN DI 13 SCK DI 14 SDO DO 15 VIO P 16 to 19 21 (EPAD) IN0 to IN3 Exposed Paddle (EPAD) AI Description Power Supply. Nominally 4.5 to 5.5 V and should be decoupled with 10 μF and 100 nF capacitors. Reference Input/Output. See the Voltage Reference Output/Input section. When the internal reference is enabled, this pin produces 4.096 V. When the internal reference is disabled and the buffer is enabled, REF produces a buffered version of the voltage present on the REFIN pin (VDD – 0.5 V maximum) useful when using low cost, low power references. For improved drift performance, connect a precision reference to REF (0.5 V to VDD). For any reference method, this pin needs decoupling with an external 10 μF capacitor connected as close to REF as possible. See the Reference Decoupling section. Internal Reference Output/Reference Buffer Input. See the Voltage Reference Output/Input section. When using the internal reference, the internal unbuffered reference voltage is present and needs decoupling with a 0.1 μF capacitor. When using the internal reference buffer, apply a source between 0.5 V and 4.096 V that is buffered to the REF pin as previously described. Power Supply Ground. Analog Input Channel 4, Analog Input Channel 5, Analog Input Channel 6, and Analog Input Channel 7. Common Channel Input. All input channels, IN[7:0], can be referenced to a common-mode point of 0 V or VREF/2 V. Conversion Input. On the rising edge, CNV initiates the conversion. During conversion, if CNV is held high, the busy indictor is enabled. Data Input. This input is used for writing to the 14-bit configuration register. The configuration register can be written to during and after conversion. Serial Data Clock Input. This input is used to clock out the data on SDO and clock in data on DIN in an MSB first fashion. Serial Data Output. The conversion result is output on this pin and synchronized to SCK. In unipolar modes, conversion results are straight binary; in bipolar modes, conversion results are twos complement. Input/Output Interface Digital Power. Nominally at the same supply as the host interface (1.8 V, 2.5 V, 3 V, or 5 V). Analog Input Channel 0, Analog Input Channel 1, Analog Input Channel 2, and Analog Input Channel 3. The exposed paddle is not connected internally. For increased reliability of the solder joints, it is recommended that the pad be soldered to the GND plane. 1 AI = analog input, AI/O = analog input/output, DI = digital input, DO = digital output, and P = power. Rev. 0 | Page 7 of 28 AD7699 TYPICAL PERFORMANCE CHARACTERISTICS VDD = 5V, VREF = 5V, VIO = VDD, unless otherwise noted 1.5 1.5 1.0 1.0 0.5 (LSBS) (LSBS) 0.5 0 0 –0.5 –0.5 0 16,384 32,768 CODES 49,152 65,536 –1.0 07354-006 –1.5 0 16,384 Figure 5. Integral Nonlinearity vs. Code 49,152 65,536 Figure 8. Differential Nonlinearity vs. Code 250,000 250,000 σ = 0.51 LSB VREF = 5V 220,840 σ = 0.78 LSB VREF = 4.096V 200,000 200,000 150,000 150,000 COUNTS COUNTS 32,768 CODES 07354-009 –1.0 100,000 50,000 191,013 100,000 50,000 31,411 26,926 38,420 0 3 10 7FF9 7FFA 7FFB 7FFC 7FFD 7FFE 7FFF CODE IN HEX 0 0 8000 8001 0 Figure 6. Histogram of a DC Input at Code Center 0 119 157 0 0 8000 8001 Figure 9. Histogram of a DC Input at Code Center 0 0 VREF = 5V fS = 500kSPS fIN = 19.94kHz SNR = 92.3dB SINAD = 91.5dB THD = –98dB SFDR = 100dB SECOND HARMONIC = –111dB THIRD HARMONIC = –101dB –40 –60 –80 VREF = 4.096V fS = 500kSPS fIN = 19.94kHz SNR = 91.1dB SINAD = 90.4dB THD = –98dB SFDR = 100dB SECOND HARMONIC = –104dB THIRD HARMONIC = –101dB –20 AMPLITUDE (dB OF FULL SCALE) –20 –100 –120 –140 –160 –40 –60 –80 –100 –120 –140 –180 0 25 50 75 100 125 150 175 FREQUENCY (kHz) 200 225 250 Figure 7. 20 kHz FFT, VREF = 5 V –180 0 25 50 75 100 125 150 175 FREQUENCY (kHz) 200 Figure 10. 20 kHz FFT, VREF = 4.096 V Rev. 0 | Page 8 of 28 225 250 07354-010 –160 07354-007 AMPLITUDE (dB OF FULL SCALE) 0 7FF9 7FFA 7FFB 7FFC 7FFD 7FFE 7FFF CODE IN HEX 07354-008 0 07354-005 13,341 0 AD7699 100 100 VREF = 5V VREF = 5V 95 95 –10dB –10dB 90 90 85 SINAD (dB) –0.5dB 80 75 80 75 70 70 65 65 0 50 100 150 200 250 300 350 FREQUENCY (kHz) 400 450 500 60 07354-011 60 –0.5dB 0 100 50 Figure 11. SNR vs. Frequency 150 200 250 300 350 FREQUENCY (kHz) 400 450 500 07354-014 SNR (dB) 85 Figure 14. SINAD vs. Frequency 16 –60 VREF = 5V –65 VREF = 5V –10dB 15 –70 –75 –0.5dB 14 –80 ENOB (Bits) –10dB THD (dB) –85 –90 –95 13 –0.5dB 12 –100 –105 11 –110 50 100 150 200 250 300 350 FREQUENCY (kHz) 400 450 500 10 0 50 100 150 200 250 300 350 FREQUENCY (kHz) 450 115 –80 96 fIN = 20kHz fIN = 20kHz SNR, VREF = 5V SFDR, VREF = 5V SINAD, VREF = 5V 94 500 Figure 15. ENOB vs. Frequency Figure 12. THD vs. Frequency SFDR, VREF = 4.096V –85 THD (dB) 92 110 105 –90 90 SNR, VREF = 4.096V SINAD, VREF = 4.096V –95 THD, VREF = 5V THD, VREF = 4.096V 100 86 –55 –35 –15 5 25 45 65 TEMPERATURE (°C) 85 105 125 –100 –55 –35 –15 5 25 45 65 TEMPERATURE (°C) 85 Figure 16. THD, SFDR vs. Temperature Figure 13. SNR, SINAD vs.Temperature Rev. 0 | Page 9 of 28 105 95 125 07354-017 88 07354-013 SNR, SINAD (dB) 400 SFDR (dB) 0 07354-012 –120 07354-015 –115 AD7699 94 17 –80 110 –85 105 fIN = 20kHz SNR 16 15 ENOB 88 THD (dB) 90 ENOB (Bits) SINAD 100 95 –95 THD –100 90 –105 85 14 –110 4.0 80 5.5 4.5 5.0 REFERENCE VOLTAGE (V) Figure 17. SNR, SINAD, ENOB vs. Reference Voltage Figure 20. THD, SFDR vs. Reference Voltage 5500 180 fIN = 20kHz fs = 500kSPS VDD, INT REF 15.5 93 SNR 15.4 92 SINAD 15.3 160 89 15.0 88 14.9 87 14.8 86 14.7 VDD CURRENT (µA) 15.1 ENOB ENOB (Bits) 15.2 91 140 120 5000 100 80 VDD, EXT REF 4750 40 14.6 –8 –6 –4 INPUT LEVEL (dB) –2 60 VIO 4500 –55 07354-018 85 –10 0 Figure 18. SNR, SINAD, and ENOB vs. Input Level –35 –15 5 25 45 65 TEMPERATURE (°C) 85 105 20 125 Figure 21. Operating Currents vs. Temperature 3 5750 100 fS = 500kSPS 4.096V INTERNAL REF 5500 2 90 INTERNAL BUFFER, TEMP ON UNIPOLAR GAIN 5250 1 0 VDD CURRENT (µA) ZERO ERROR, GAIN ERROR (LSB) BIPOLAR GAIN UNIPOLAR OFFSET –1 BIPOLAR OFFSET 80 INTERNAL BUFFER, TEMP OFF 5000 70 4750 60 4500 50 EXTERNAL REF, TEMP ON 4250 40 EXTERNAL REF, TEMP OFF –2 4000 30 –3 –55 –35 –15 5 25 45 65 TEMPERATURE (°C) 85 105 125 3750 4.5 Figure 19. Offset and Gain Errors vs. Temperature, Not Normalized 5.0 VDD SUPPLY (V) Figure 22. Operating Currents vs. Supply Rev. 0 | Page 10 of 28 20 5.5 07354-040 VIO 07354-020 SNR (dB) 5250 07354-022 94 VREF = 5V VIO CURRENT (µA) 15.6 95 90 07354-019 13 5.5 4.5 5.0 REFERENCE VOLTAGE (V) VIO CURRENT (µA) 86 4.0 07354-016 SNR, SINAD (dB) SFDR –90 SFDR (dB) 92 AD7699 25 4.099 4.098 20 tDSDO DELAY (ns) VREF (V) 4.097 4.096 4.095 15 VDD = 5V, 85°C 10 4.094 VDD = 5V, 25°C 5 –35 –15 5 25 45 65 TEMPERATURE (°C) 85 105 125 Figure 23. Internal Reference Output Voltage vs. Temperature, Three Devices Rev. 0 | Page 11 of 28 0 0 20 40 60 80 SDO CAPACITIVE LOAD (pF) 100 120 Figure 24. tDSDO Delay vs. SDO Capacitance Load and Supply 07354-021 4.092 –55 07354-041 4.093 AD7699 TERMINOLOGY Least Significant Bit (LSB) The LSB is the smallest increment that can be represented by a converter. For an analog-to-digital converter with N bits of resolution, the LSB expressed in volts is LSB (V) = V REF 2N Integral Nonlinearity Error (INL) INL refers to the deviation of each individual code from a line drawn from negative full scale through positive full scale. The point used as negative full scale occurs ½ LSB before the first code transition. Positive full scale is defined as a level 1½ LSB beyond the last code transition. The deviation is measured from the middle of each code to the true straight line (see Figure 26). Differential Nonlinearity Error (DNL) In an ideal ADC, code transitions are 1 LSB apart. DNL is the maximum deviation from this ideal value. It is often specified in terms of resolution for which no missing codes are guaranteed. Offset Error For unipolar mode, the first transition should occur at a level ½ LSB above analog ground. The unipolar offset error is the deviation of the actual transition from that point. For bipolar mode, the first transition should occur at a level ½ LSB above VREF/2. The bipolar offset error is the deviation of the actual transition from that point. Gain Error The last transition (from 111 … 10 to 111 … 11) should occur for an analog voltage 1½ LSB below the nominal full scale. The gain error is the deviation in LSB (or percentage of full-scale range) of the actual level of the last transition from the ideal level after the offset error is adjusted out. Closely related is the full-scale error (also in LSB or percentage of full-scale range), which includes the effects of the offset error. Aperture Delay Aperture delay is the measure of the acquisition performance. It is the time between the rising edge of the CNV input and the point at which the input signal is held for a conversion. Transient Response Transient response is the time required for the ADC to accurately acquire its input after a full-scale step function is applied. Dynamic Range Dynamic range is the ratio of the rms value of the full scale to the total rms noise measured with the inputs shorted together. The value for dynamic range is expressed in decibels. Signal-to-Noise Ratio (SNR) SNR is the ratio of the rms value of the actual input signal to the rms sum of all other spectral components below the Nyquist frequency, excluding harmonics and dc. The value for SNR is expressed in decibels. Signal-to-(Noise + Distortion) Ratio (SINAD) SINAD is the ratio of the rms value of the actual input signal to the rms sum of all other spectral components below the Nyquist frequency, including harmonics but excluding dc. The value for SINAD is expressed in decibels. Total Harmonic Distortion (THD) THD is the ratio of the rms sum of the first five harmonic components to the rms value of a full-scale input signal and is expressed in decibels. Spurious-Free Dynamic Range (SFDR) SFDR is the difference, in decibels, between the rms amplitude of the input signal and the peak spurious signal. Effective Number of Bits (ENOB) ENOB is a measurement of the resolution with a sine wave input. It is related to SINAD by the formula ENOB = (SINADdB − 1.76)/6.02 and is expressed in bits. Channel-to-Channel Crosstalk Channel-to-channel crosstalk is a measure of the level of crosstalk between any two adjacent channels. It is measured by applying a dc to the channel under test and applying a full-scale, 100 kHz sine wave signal to the adjacent channel(s). The crosstalk is the amount of signal that leaks into the test channel and is expressed in decibels. Reference Voltage Temperature Coefficient Reference voltage temperature coefficient is derived from the typical shift of output voltage at 25°C on a sample of parts at the maximum and minimum reference output voltage (VREF) measured at TMIN, T (25°C), and TMAX. It is expressed in ppm/°C as TCVREF (ppm/°C) = VREF ( Max ) – VREF ( Min) VREF (25°C) × (TMAX – TMIN ) × 106 where: VREF (Max) = maximum VREF at TMIN, T (25°C), or TMAX. VREF (Min) = minimum VREF at TMIN, T (25°C), or TMAX. VREF (25°C) = VREF at 25°C. TMAX = +85°C. TMIN = –40°C. Rev. 0 | Page 12 of 28 AD7699 THEORY OF OPERATION INx+ SWITCHES CONTROL MSB 32,768C 16,384C LSB 4C 2C C SW+ C BUSY REF COMP GND 32,768C 16,384C 4C 2C C CONTROL LOGIC OUTPUT CODE C MSB LSB SW– 07354-023 CNV INx– OR COM Figure 25. ADC Simplified Schematic OVERVIEW CONVERTER OPERATION The AD7699 is an 8-channel, 16-bit, charge redistribution successive approximation register (SAR) analog-to-digital converter (ADC). It is capable of converting 500,000 samples per second (500 kSPS) and power down between conversions. For example, when operating with an external reference at 1 kSPS, it consumes 52 µW typically, ideal for battery-powered applications. The AD7699 is a successive approximation ADC based on a charge redistribution DAC. Figure 25 shows the simplified schematic of the ADC. The capacitive DAC consists of two identical arrays of 16 binary-weighted capacitors, which are connected to the two comparator inputs. The AD7699 contains all of the components for use in a multichannel, low power data acquisition system, including • • • • • • 16-bit SAR ADC with no missing codes 8-channel, low crosstalk multiplexer Internal low drift reference and buffer Temperature sensor Selectable one-pole filter Channel sequencer These components are configured through an SPI-compatible, 14-bit register. Conversion results, also SPI compatible, can be read after or during conversions with the option for reading back the configuration. The AD7699 provides the user with an on-chip track-and-hold and does not exhibit pipeline delay or latency. The AD7699 is specified from 4.5 V to 5.5 V and can be interfaced to any 1.8 V to 5 V digital logic family. It is housed in a 20-lead, 4 mm × 4 mm LFCSP that combines space savings and allows flexible configurations and is also pin-for-pin compatible with the 16-bit AD7682 and AD7689, and the 14-bit AD7949. During the acquisition phase, terminals of the array tied to the comparator input are connected to GND via SW+ and SW−. All independent switches are connected to the analog inputs. Thus, the capacitor arrays are used as sampling capacitors and acquire the analog signal on the INx+ and INx− (or COM) inputs. When the acquisition phase is complete and the CNV input goes high, a conversion phase is initiated. When the conversion phase begins, SW+ and SW− are opened first. The two capacitor arrays are then disconnected from the inputs and connected to the GND input. Therefore, the differential voltage between the INx+ and INx− (or COM) inputs captured at the end of the acquisition phase is applied to the comparator inputs, causing the comparator to become unbalanced. By switching each element of the capacitor array between GND and REF, the comparator input varies by binary-weighted voltage steps (VREF/2, VREF/4, ... VREF/32,768). The control logic toggles these switches, starting with the MSB, to bring the comparator back into a balanced condition. After the completion of this process, the part returns to the acquisition phase, and the control logic generates the ADC output code and a busy signal indicator. Because the AD7699 has an on-board conversion clock, the serial clock, SCK, is not required for the conversion process. Rev. 0 | Page 13 of 28 AD7699 TRANSFER FUNCTIONS 011...111 111...111 011...110 011...101 111...110 111...101 The ideal transfer characteristic for the AD7699 is shown in Figure 26 and for both unipolar and bipolar ranges with the internal 4.096 V reference. 100...010 000...010 100...001 000...001 100...000 000...000 –FSR –FSR + 1LSB –FSR + 0.5LSB +FSR – 1LSB +FSR – 1.5LSB ANALOG INPUT 07354-024 With the inputs configured for bipolar range (COM = VREF/2 or paired differentially with INx− = VREF/2), the data outputs are twos complement. TWOS STRAIGHT COMPLEMENT BINARY ADC CODE With the inputs configured for unipolar range (single ended, COM with ground sense, or paired differentially with INx− as ground sense), the data output is straight binary. Figure 26. ADC Ideal Transfer Function Table 6. Output Codes and Ideal Input Voltages Description FSR − 1 LSB Midscale + 1 LSB Midscale Midscale − 1 LSB −FSR + 1 LSB −FSR Unipolar Analog Input1 VREF = 4.096 V 4.095938 V 2.048063 V 2.048 V 2.047938 V 62.5 μV 0V Digital Output Code (Straight Binary Hex) 0xFFFF3 0x8001 0x8000 0x7FFF 0x0001 0x00003 1 With COM or INx− = 0 V or all INx referenced to GND. With COM or INx− = VREF/2. 3 This is also the code for an overranged analog input ((INx+) − (INx−), or COM, above VREF − VGND). 4 This is also the code for an underranged analog input ((INx+) − (INx−), or COM, below VGND). 2 Rev. 0 | Page 14 of 28 Bipolar Analog Input2 VREF = 4.096 V 2.047938 V 62.5 μV 0V −62.5 μV −2.047938 V −2.048 V Digital Output Code (Twos Complement Hex) 0x7FFF3 0x0001 0x0000 0xFFFF4 0x8001 0x8000 AD7699 TYPICAL CONNECTION DIAGRAMS 5V 100nF 100nF 10µF2 V+ 100nF REFIN VDD REF 1.8V TO VDD VIO 0V TO VREF ADA4841-x 3 IN0 V– V+ IN[7:1] AD7699 0V TO VREF ADA4841-x 3 DIN MOSI SCK SCK SDO MISO CNV SS V– 0V OR VREF /2 COM NOTES 1. INTERNAL REFERENCE SHOWN. SEE THE VOLTAGE REFERENCE OUTPUT/INPUT SECTION FOR REFERENCE SELECTION. 2. CREF IS USUALLY A 10µF CERAMIC CAPACITOR (X5R). 3. SEE THE DRIVER AMPLIFIER CHOICE SECTION FOR ADDITIONAL RECOMMENDED AMPLIFIERS. 4. SEE THE DIGITAL INTERFACE SECTION FOR CONFIGURING AND READING CONVERSION DATA. 07354-025 GND Figure 27. Typical Application Diagram with Multiple Supplies 5V V+ 100nF 100nF 10µF2 REF REFIN VDD 1.8V TO VDD 100nF VIO ADA4841-x 3 IN0 V+ IN[7:1] AD7699 ADA4841-x 3 VREF /2 MOSI SCK SDO MISO CNV SS COM GND NOTES 1. INTERNAL REFERENCE SHOWN. SEE THE VOLTAGE REFERENCE OUTPUT/INPUT SECTION FOR REFERENCE SELECTION. 2. CREF IS USUALLY A 10µF CERAMIC CAPACITOR (X5R). 3. SEE THE DRIVER AMPLIFIER CHOICE SECTION FOR ADDITIONAL RECOMMENDED AMPLIFIERS. 4. SEE THE DIGITAL INTERFACE SECTION FOR CONFIGURING AND READING CONVERSION DATA. Figure 28. Typical Application Diagram Using Bipolar Input Rev. 0 | Page 15 of 28 07354-026 VREF p-p DIN SCK AD7699 Unipolar or Bipolar 70 Figure 27 shows an example of the recommended connection diagram for the AD7699 when multiple supplies are available. 65 60 Bipolar Single Supply 55 ANALOG INPUTS Input Structure Figure 29 shows an equivalent circuit of the input structure of the AD7699. The two diodes, D1 and D2, provide ESD protection for the analog inputs, IN[7:0] and COM. Care must be taken to ensure that the analog input signal does not exceed the supply rails by more than 0.3 V because this causes the diodes to become forward-biased and to start conducting current. These diodes can handle a forward-biased current of 130 mA maximum. For instance, these conditions may eventually occur when the input buffer supplies are different from VDD. In such a case, for example, an input buffer with a short circuit, the current limitation can be used to protect the part. VDD D1 INx+ OR INx– OR COM CIN 35 30 1 10 100 FREQUENCY (kHz) 1k 10k Figure 30. Analog Input CMRR vs. Frequency During the acquisition phase, the impedance of the analog inputs can be modeled as a parallel combination of the capacitor, CPIN, and the network formed by the series connection of RIN and CIN. CPIN is primarily the pin capacitance. RIN is typically 400 Ω (8.8 kΩ when the one-pole filter is active) and is a lumped component made up of serial resistors and the on resistance of the switches. CIN is typically 27 pF and is mainly the ADC sampling capacitor. Selectable Low-Pass Filter During the conversion phase, where the switches are opened, the input impedance is limited to CPIN. While the AD7699 is acquiring, RIN and CIN make a one-pole, low-pass filter that reduces undesirable aliasing effects and limits the noise from the driving circuitry. The low-pass filter can be programmed for the full bandwidth or ¼ of the bandwidth with CFG[6], as shown in Table 8. Note that the converter throughput must also be reduced by ¼ when using the filter. If the maximum throughput is used with the BW set to ¼, the acquisition time of the converter, tACQ, is violated, resulting in poor THD. Input Configurations Figure 31 shows the different methods for configuring the analog inputs with the configuration register (CFG[12:10]). Refer to the Configuration Register, CFG section for more details. D2 GND 45 40 07354-027 CPIN RIN 50 07354-028 For this circuit, a rail-to-rail input/output amplifier can be used; however, the offset voltage vs. input common-mode range should be noted and taken into consideration (1 LSB = 62.5 μV with VREF = 4.096 V). Note that the conversion results are in twos complement format when using the bipolar input configuration. Refer to the AN-581 Application Note, Biasing and Decoupling Op Amps in Single Supply Applications, at www.analog.com for additional details about using single-supply amplifiers. CMRR (dB) Figure 28 shows an example of a system with a bipolar input using single supplies with the internal reference (optional different VIO supply). This circuit is also useful when the amplifier/signal conditioning circuit is remotely located with some common mode present. Note that for any input configuration, the inputs, INx, are unipolar and always referenced to GND (no negative voltages even in bipolar range). Figure 29. Equivalent Analog Input Circuit This analog input structure allows the sampling of the true differential signal between INx+ and COM or INx+ and INx−. (COM or INx− = GND ± 0.1 V or VREF ± 0.1 V). By using these differential inputs, signals common to both inputs are rejected, as shown in Figure 30. Rev. 0 | Page 16 of 28 AD7699 CH0+ IN0 CH0+ IN0 CH1+ IN1 CH1+ IN1 CH2+ IN2 CH2+ IN2 CH3+ IN3 CH3+ IN3 CH4+ IN4 CH4+ IN4 CH5+ IN5 CH5+ IN5 CH6+ IN6 CH6+ IN6 CH7+ IN7 CH7+ IN7 COM COM– COM GND To enable the sequencer, CFG[2:1] are written to for initializing the sequencer. After CFG[13:0] are updated, DIN must be held low while reading data out (at least for Bit 13), or the CFG register begins updating again. B—8 CHANNELS, COMMON REFERENCE IN0 CH0+ (–) CH0– (+) IN1 CH0– (+) IN1 CH1+ (–) IN2 CH1+ (–) IN2 CH1– (+) IN3 CH1– (+) IN3 CH2+ (–) IN4 CH2+ IN4 CH2– (+) IN5 CH3+ IN5 CH3+ (–) IN6 CH4+ IN6 CH3– (+) IN7 CH5+ IN7 IN0 COM– While operating in a sequence, the CFG register can be changed by writing 012 to CFG[2:1]. However, if changing CFG11 (paired or single channel) or CFG[9:7] (last channel in sequence), the sequence reinitializes and converts IN0 (or IN1) after CFG is updated. Examples COM GND C—4 CHANNELS, DIFFERENTIAL The sequencer starts with IN0 and finishes with IN[7:0] set in CFG[9:7]. For paired channels, the channels are paired depending on the last channel set in CFG[9:7]. Note that the channel pairs are always paired as IN (even) = INx+ and IN (odd) = INx− regardless of CFG[7]. GND CH0+ (–) COM The AD7699 includes a channel sequencer useful for scanning channels in a IN0 to IN[7:0] fashion. Channels are scanned as singles or pairs, with or without the temperature sensor, after the last channel is sequenced. Bit[13], Bits[6:3], and Bit 0 are configured for the input and sequencer. GND D—COMBINATION 07354-029 A—8 CHANNELS, SINGLE ENDED Sequencer As a first example, scan all IN[7:0] referenced to COM = GND with the temperature sensor. Figure 31. Multiplexed Analog Input Configurations The analog inputs can be configured as • • • • Figure 31A, single-ended referenced to system ground; CFG[12:10] = 1112. Figure 31B, bipolar differential with a common reference point; COM = VREF/2; CFG[12:10] = 0102. Unipolar differential with COM connected to a ground sense; CFG[12:10] = 1102. Figure 31C, bipolar differential pairs with INx− referenced to VREF/2; CFG[12:10] = 00X2. Unipolar differential pairs with INx− referenced to a ground sense; CFG[12:10] = 10X2. In this configuration, the INx+ is identified by the channel in CFG[9:7]. For example, for IN0 = IN1+ and IN1 = IN1−, CFG[9:7] = 0002; for IN1 = IN1+ and IN0 = IN1−, CFG[9:7] = 0012. Figure 31D, inputs configured in any of the above combinations (showing that the AD7699 can be configured dynamically). 13 CFG 12 1 11 10 INCC 1 0 9 8 7 INx 1 1 1 6 BW 5 4 3 REF 2 1 SEQ 1 0 0 RB As a second example, scan three paired channels without the temperature sensor and referenced to VREF/2. 13 CFG 12 0 1 11 10 INCC 0 X1 9 1 8 7 INx 0 X1 6 BW 5 4 3 REF 2 1 SEQ 1 1 0 RB X = don’t care. Source Resistance When the source impedance of the driving circuit is low, the AD7699 can be driven directly. Large source impedances significantly affect the ac performance, especially total harmonic distortion (THD). The dc performances are less sensitive to the input impedance. The maximum source impedance depends on the amount of THD that can be tolerated. The THD degrades as a function of the source impedance and the maximum input frequency. Rev. 0 | Page 17 of 28 AD7699 DRIVER AMPLIFIER CHOICE Although the AD7699 is easy to drive, the driver amplifier must meet the following requirements: • The noise generated by the driver amplifier must be kept as low as possible to preserve the SNR and transition noise performance of the AD7699. Note that the AD7699 has a noise much lower than most of the other 16-bit ADCs and, therefore, can be driven by a noisier amplifier to meet a given system noise specification. The noise from the amplifier is filtered by the AD7699 analog input circuit low-pass filter made by RIN and CIN or by an external filter, if one is used. Because the typical noise of the AD7699 is 35 µV rms (with VREF = 5 V), the SNR degradation due to the amplifier is SNRLOSS 35 = 20 log π 2 2 35 + f − 3dB (NeN ) 2 • For ac applications, the driver should have a THD performance commensurate with the AD7699. Figure 12 shows THD vs. frequency for the AD7699. For multichannel, multiplexed applications on each input or input pair, the driver amplifier and the AD7699 analog input circuit must settle a full-scale step onto the capacitor array at a 16-bit level (0.0015%). In amplifier data sheets, settling at 0.1% to 0.01% is more commonly specified. This may differ significantly from the settling time at a 16-bit level and should be verified prior to driver selection. Table 7. Recommended Driver Amplifiers Amplifier ADA4841-x AD8655 AD8021 AD8022 OP184 AD8605, AD8615 Internal Reference/Temperature Sensor The internal reference can be set for a 4.096 V output as detailed in Table 8. With the internal reference enabled, the band gap voltage is also present on the REFIN pin, which requires an external 0.1 μF capacitor. Because the current output of REFIN is limited, it can be used as a source if followed by a suitable buffer, such as the AD8605. Enabling the reference also enables the internal temperature sensor, which measures the internal temperature of the AD7699 and is thus useful for performing a system calibration. Note that, when using the temperature sensor, the output is straight binary referenced from the AD7699 GND pin. The internal reference is temperature-compensated to within 15 mV. The reference is trimmed to provide a typical drift of 3 ppm/°C. where: f−3dB is the input bandwidth in megahertz of the AD7699 (14.7 MHz in full BW or 670 kHz in ¼ BW) or the cutoff frequency of an input filter, if one is used. N is the noise gain of the amplifier (for example, 1 in buffer configuration). eN is the equivalent input noise voltage of the op amp, in nV/√Hz. • described in Table 8 with more details in each of the following sections. Typical Application Very low noise, small, and low power 5 V single supply, low noise Very low noise and high frequency Low noise and high frequency Low power, low noise, and low frequency 5 V single supply, low power External Reference and Internal Buffer For improved drift performance, an external reference can be used with the internal buffer. The external reference is connected to REFIN, and the output is produced on the REF pin. An external reference can be used with the internal buffer with or without the temperature sensor enabled. Refer to Table 8 for register details. With the buffer enabled, the gain is unity and is limited to an input/output of 4.096 V. The internal reference buffer is useful in multiconverter applications because a buffer is typically required in these applications. In addition, a low power reference can be used because the internal buffer provides the necessary performance to drive the SAR architecture of the AD7699. External Reference In any of the five voltage reference schemes, an external reference can be connected directly on the REF pin because the output impedance of REF is >5 kΩ. To reduce power consumption, the reference and buffer can be powered down independently or together for the lowest power consumption. However, for applications requiring the use of the temperature sensor, the reference must be active. Refer to Table 8 for register details. For improved drift performance, an external reference such as the ADR43x or ADR44x is recommended. Reference Decoupling VOLTAGE REFERENCE OUTPUT/INPUT The AD7699 allows the choice of a very low temperature drift internal voltage reference, an external reference, or an external buffered reference. The internal reference of the AD7699 provides excellent performance and can be used in almost all applications. There are five possible choices of voltage reference schemes briefly Whether using an internal or external reference, the AD7699 voltage reference output/input, REF, has a dynamic input impedance and should therefore be driven by a low impedance source with efficient decoupling between the REF and GND pins. This decoupling depends on the choice of the voltage reference but usually consists of a low ESR capacitor connected to REF and GND with minimum parasitic inductance. A 10 µF (X5R, 1206 size) ceramic chip capacitor is appropriate when using the internal reference, the ADR43x/ADR44x external Rev. 0 | Page 18 of 28 AD7699 10,000 If desired, smaller reference decoupling capacitor values down to 2.2 µF can be used with a minimal impact on performance, especially on DNL. Regardless, there is no need for an additional lower value ceramic decoupling capacitor (for example, 100 nF) between the REF and GND pins. For applications that use multiple AD7699s or other PulSAR devices, it is more effective to use the internal reference buffer to buffer the external reference voltage, thus reducing SAR conversion crosstalk. The voltage reference temperature coefficient (TC) directly impacts full scale; therefore, in applications where full-scale accuracy matters, care must be taken with the TC. For instance, a ±15 ppm/°C TC of the reference changes full scale by ±1 LSB/°C. POWER SUPPLY The AD7699 uses two power supply pins: an analog and digital core supply (VDD) and a digital input/output interface supply (VIO). VIO allows direct interface with any logic between 1.8 V and VDD. To reduce the supplies needed, the VIO and VDD pins can be tied together. The AD7699 is independent of power supply sequencing between VIO and VDD. The only restriction is that CNV must be low when powering up the AD7699. Additionally, it is very insensitive to power supply variations over a wide frequency range, as shown in Figure 32. VDD = 5V, INTERNAL REF 1000 100 VDD = 5V, EXTERNAL REF 10 1 VIO 0.1 0.010 0.001 100 10 1k 10k SAMPLING RATE (sps) 100k Figure 33. Operating Currents vs. Sampling Rate SUPPLYING THE ADC FROM THE REFERENCE For simplified applications, the AD7699, with its low operating current, can be supplied directly using the reference circuit, as shown in Figure 34. The reference line can be driven by • • The system power supply directly A reference voltage with enough current output capability, such as the ADR43x/ADR44x A reference buffer, such as the AD8605, which can also filter the system power supply, as shown in Figure 34 • 5V 5V 10Ω 5V 10kΩ 1µF AD8605 1µF 10µF 0.1µF REF 70 VDD VIO 60 1OPTIONAL 55 REFERENCE BUFFER AND FILTER. Figure 34. Example of an Application Circuit 50 45 40 1 10 100 FREQUENCY (kHz) 1k 10k 07354-030 35 Figure 32. PSRR vs. Frequency Rev. 0 | Page 19 of 28 07354-032 AD7699 65 PSSR (dB) 0.1µF 1 75 30 1M 07354-031 The placement of the reference decoupling capacitor is also important to the performance of the AD7699, as explained in the Layout section. Mount the decoupling capacitor on the same side as the ADC at the REF pin with a thick PCB trace. The GND should also connect to the reference decoupling capacitor with the shortest distance and to the analog ground plane with several vias. The AD7699 powers down automatically at the end of each conversion phase; therefore, the operating currents and power scale linearly with the sampling rate. This makes the part ideal for low sampling rates (even of a few hertz) and low batterypowered applications. OPERATING CURRENT (µA) reference, or a low impedance buffer such as the AD8031 or the AD8605. AD7699 DIGITAL INTERFACE The AD7699 uses a simple 4-wire interface and is compatible with SPI, MICROWIRE™, QSPI™, digital hosts, and DSPs, for example, Blackfin® ADSP-BF53x, SHARC®, ADSP-219x, and ADSP-218x. The SCK frequency required is calculated by The interface uses the CNV, DIN, SCK, and SDO signals and allows CNV, which initiates the conversion, to be independent of the readback timing. This is useful in low jitter sampling or simultaneous sampling applications. The time between tDATA and tCONV is a safe time when digital activity should not occur, or sensitive bit decisions may be corrupt. A 14-bit register, CFG[13:0], is used to configure the ADC for the channel to be converted, the reference selection, and other components, which are detailed in the Configuration Register, CFG section. f SCK Number _ SCK _ Edges t DATA READING/WRITING DURING ACQUISITION, ANY SPEED HOSTS When reading/writing after conversion, or during acquisition (n), conversion results are for the previous (n − 1) conversion, and writing is for the (n + 1) acquisition. When CNV is low, reading/writing can occur during conversion, acquisition, and spanning conversion (acquisition plus conversion), as detailed in the following sections. The CFG word is updated on the first 14 SCK rising edges, and conversion results are output on the first 15 (or 16 if busy mode is selected) SCK falling edges. If the CFG readback is enabled, an additional 14 SCK falling edges are required to output the CFG word associated with the conversion results, with the CFG MSB following the LSB of the conversion result. For the maximum throughput, the only time restriction is that the reading/writing take place during the tACQ (min) time. For slow throughputs, the time restriction is dictated by throughput required by the user, and the host is free to run at any speed. Thus for slow hosts, data access must take place during the acquisition phase. A discontinuous SCK is recommended because the part is selected with CNV low, and SCK activity begins to write a new configuration word and clock out data. When reading/writing spanning conversion, the data access starts at the current acquisition (n) and spans into the conversion (n). Conversion results are for the previous (n − 1) conversion, and writing the CFG register is for the next (n + 1) acquisition and conversion. Note that in the following sections, the timing diagrams indicate digital activity (SCK, CNV, DIN, SDO) during the conversion. However, due to the possibility of performance degradation, digital activity should occur only prior to the safe data reading/ writing time, tDATA, because the AD7699 provides error correction circuitry that can correct for an incorrect bit during this time. From tDATA to tCONV, there is no error correction and conversion results may be corrupted. The user should configure the AD7699 and initiate the busy indicator (if desired) prior to tDATA. It is also possible to corrupt the sample by having SCK or DIN transitions near the sampling instant. Therefore, it is recommended to keep the digital pins quiet for approximately 30 ns before and 10 ns after the rising edge of CNV, using a discontinuous SCK whenever possible to avoid any potential performance degradation. READING/WRITING DURING CONVERSION, FAST HOSTS When reading/writing during conversion (n), conversion results are for the previous (n − 1) conversion, and writing the CFG is for the next (n + 1) acquisition and conversion. After the CNV is brought high to initiate conversion, it must be brought low again to allow reading/writing during conversion. Reading/writing should only occur up to tDATA and, because this time is limited, the host must use a fast SCK. READING/WRITING SPANNING CONVERSION, ANY SPEED HOST Similar to reading/writing during conversion, reading/writing should only occur up to tDATA. For the maximum throughput, the only time restriction is that reading/writing take place during the tACQ (min) + tDATA time. For slow throughputs, the time restriction is dictated by the user’s required throughput, and the host is free to run at any speed. Similar to reading/writing during acquisition, for slow hosts, the data access must take place during the acquisition phase with additional time into the conversion. Note that data access spanning conversion requires the CNV to be driven high to initiate a new conversion, and data access is not allowed when CNV is high. Thus, the host must perform two bursts of data access when using this method. CONFIGURATION REGISTER, CFG The AD7699 uses a 14-bit configuration register (CFG[13:0]) as detailed in Table 8 for configuring the inputs, the channel to be converted, one-pole filter bandwidth, the reference, and the channel sequencer. The CFG register is latched (MSB first) on DIN with 14 SCK rising edges. CFG update is edge dependent, allowing for asynchronous or synchronous hosts. Rev. 0 | Page 20 of 28 AD7699 • • • The register can be written to during conversion, during acquisition, or spanning acquisition/conversion and is updated at the end of conversion, tCONV (maximum). There is always a one deep delay when writing the CFG register. Note that at power-up, the CFG register is undefined and two dummy conversions are required to update the register. To preload the CFG register with a factory setting, hold DIN high for two conversions. Thus CFG[13:0] = 0x3FFF. This sets the AD7699 for the following: 13 CFG 12 INCC 11 INCC 10 INCC 9 INx 8 INx • • IN[7:0] unipolar referenced to GND, sequenced in order Full bandwidth for a one-pole filter Internal reference/temperature sensor disabled, buffer enabled Enables the sequencer No readback of the CFG register Table 8 summarizes the configuration register bit details. See the Theory of Operation section for more details. 7 INx 6 BW 5 REF 4 REF 3 REF 2 SEQ 1 SEQ 0 RB Table 8. Configuration Register Description Bit(s) [13] Name CFG [12:10] INCC [9:7] INx [6] BW [5:3] REF [2:1] SEQ 0 RB 1 Description Configuration update. 0 = Keep current configuration settings. 1 = Overwrite contents of register. Input channel configuration. Selection of pseudobipolar, pseudodifferential, pairs, single-ended, or temperature sensor. Refer to the Input Configurations section. Bit 12 Bit 11 Bit 10 Function 0 0 X1 Bipolar differential pairs; INx− referenced to VREF/2 ± 0.1 V. 0 1 0 Bipolar; INx referenced to COM = VREF/2 ± 0.1 V. 0 1 1 Temperature sensor. 1 0 X1 Unipolar differential pairs; INx− referenced to GND ± 0.1 V. 1 1 0 Unipolar, IN0 to IN7 referenced to COM = GND ± 0.1 V (GND sense). 1 1 1 Unipolar, IN0 to IN7 referenced to GND. Input channel selection in binary fashion. Bit 9 Bit 8 Bit 7 Channel 0 0 0 IN0 0 0 1 IN1 … … … … 1 1 1 IN7 Select bandwidth for low-pass filter. Refer to the Selectable Low-Pass Filter section. 0 = ¼ of BW, uses an additional series resistor to further bandwidth limit the noise. Maximum throughput must also be reduced to ¼. 1 = Full BW. Reference/buffer selection. Selection of internal, external, and external buffered references, and enabling of the on-chip temperature sensor. Refer to the Voltage Reference Output/Input section. Bit 5 Bit 4 Bit 3 Function 0 0 0 Not used 0 0 1 Internal reference, REF = 4.096 V output. 0 1 0 External reference, temperature enabled. 0 1 1 External reference, internal buffer, temperature enabled. 1 1 0 External reference, temperature disabled. 1 1 1 External reference, internal buffer, temperature disabled. Channel sequencer. Allows for scanning channels in an IN0 to IN[7:0] fashion. Refer to the Sequencer section. Bit 2 Bit 1 Function 0 0 Disable sequencer. 0 1 Update configuration during sequence. 1 0 Scan IN0 to IN[7:0] (set in CFG[9:7]), then temperature. 1 1 Scan IN0 to IN[7:0] (set in CFG[9:7]). Read back the CFG register. 0 = Read back current configuration at end of data. 1 = Do not read back contents of configuration. X = don’t care. Rev. 0 | Page 21 of 28 AD7699 MSB of the current conversion. For detailed timing, refer to Figure 36 and Figure 37, which depict reading/writing spanning conversion with all timing details, including setup, hold, and SCK. GENERAL TIMING WITHOUT A BUSY INDICATOR Figure 35 details the timing for all three modes: reading/writing during conversion, after conversion, and spanning conversion. Note that the gating item for both CFG and data readback is at the end of conversion (EOC). At the end of conversions (EOC), if CNV is high, the busy indicator is disabled. When CNV is brought low after EOC, SDO is driven from high impedance to the MSB. Falling SCK edges clock out bits starting with MSB − 1. The SCK can idle high or low depending on the clock polarity (CPOL) and clock phase (CPHA) settings if SPI is used. A simple solution is to use CPOL = CPHA = 0 as shown in Figure 35 with SCK idling low. As detailed previously, the data access should occur up to safe data reading/writing time, tDATA. If the full CFG word was not written to prior to EOC, it is discarded and the current configuration remains. If the conversion result is not read out fully prior to EOC, it is lost as the ADC updates SDO with the START OF CONVERSION tCYC tCONV PHASE END OF CONVERSION (EOC) EOC EOC tDATA POWER UP CONVERSION (n – 2) ACQUISITION (n – 1) CONVERSION (n – 1) ACQUISITION (n) CONVERSION (n) ACQUISITION (n + 1) CONVERSION (n + 1) ACQUISITION (n + 2) READ/WRITE DURING CONVERT CNV DIN XXX SDO XXX SCK 1 CFG (n) MSB (n – 2) 16/30 CFG (n + 1) MSB (n – 1) DATA (n – 2) 1 16/30 CFG (n + 2) MSB (n) DATA (n – 1) 1 16/30 MSB (n + 1) DATA (n) 1 16/30 DIN CFG (n) CFG (n + 1) CFG (n + 2) CFG (n + 3) SDO DATA (n – 2) DATA (n – 1) DATA (n) DATA (n + 1) READ/WRITE SPANNING CONVERT SCK 1 16/30 1 16/30 1 16/30 1 CNV CFG (n) DIN DATA (n – 2) SDO SCK 1 CFG (n + 1) DATA (n – 2) 16/30 DATA (n – 1) CFG (n + 2) DATA (n – 1) 1 16/30 DATA (n) 1 NOTES 1. CNV MUST BE HIGH PRIOR TO THE END OF CONVERSION (EOC) TO AVOID THE BUSY INDICATOR. A TOTAL OF 16 SCK FALLING EDGES IS REQUIRED TO RETURN SDO TO HIGH-Z. IF CFG READBACK IS ENABLED, A TOTAL OF 30 SCK FALLING EDGES IS REQUIRED TO RETURN SDO TO HIGH-Z. Figure 35. General Interface Timing for the AD7699 Without a Busy Indicator Rev. 0 | Page 22 of 28 CFG (n + 3) DATA (n + 1) DATA (n) 16/30 1 07354-033 READ/WRITE AFTER CONVERT CNV AD7699 READ/WRITE SPANNING CONVERSION WITHOUT A BUSY INDICATOR the CFG update. While CNV is low, both a CFG update and a data readback take place. The first 14 SCK rising edges are used to update the CFG, and the first 15 SCK falling edges clock out the conversion results starting with MSB − 1. The restriction for both configuring and reading is that they both must occur before the tDATA time of the next conversion elapses. All 14 bits of CFG[13:0] must be written, or they are ignored. In addition, if the 16-bit conversion result is not read back before tDATA elapses, it is lost. This mode is used when the AD7699 is connected to any host using an SPI, serial port, or FPGA. The connection diagram is shown in Figure 36, and the corresponding timing is given in Figure 37. For SPI, the host should use CPHA = CPOL = 0. Reading/writing spanning conversion is shown, which covers all three modes detailed in the Digital Interface section. For this mode, the host must generate the data transfer based on the conversion time. For an interrupt driven transfer, refer to the next section, which uses a busy indicator. The SDO data is valid on both SCK edges. Although the rising edge can be used to capture the data, a digital host using the SCK falling edge allows a faster reading rate, provided it has an acceptable hold time. After the 16th (or 30th) SCK falling edge, or when CNV goes high (whichever occurs first), SDO returns to high impedance. A rising edge on CNV initiates a conversion, forces SDO to high impedance, and ignores data present on DIN. After a conversion is initiated, it continues until completion irrespective of the state of CNV. CNV must be returned high before the safe data transfer time, tDATA, and then held high beyond the conversion time, tCONV, to avoid generation of the busy signal indicator. If CFG readback is enabled, the CFG associated with the conversion result is read back MSB first following the LSB of the conversion result. A total of 30 SCK falling edges is required to return SDO to high impedance if this is enabled. After the conversion is complete, the AD7699 enters the acquisition phase and powers down. When the host brings CNV low after tCONV (max), the MSB is enabled on SDO. The host also must enable the MSB of CFG at this time (if necessary) to begin CNV SS SDO MISO DIN MOSI SCK SCK FOR SPI USE CPHA = 0, CPOL = 0. 07354-034 DIGITAL HOST AD7699 Figure 36. Connection Diagram for the AD7699 Without a Busy Indicator tCYC > tCONV tCONV tCONV tDATA tDATA tCNVH RETURN CNV HIGH FOR NO BUSY RETURN CNV HIGH FOR NO BUSY CNV tACQ (QUIET TIME) CONVERSION (n – 1) tSCK tSCKH SCK 14 UPDATE (n) CFG/SDO 15 16/ 30 1 CFG LSB DIN X tEN END CFG (n) SDO LSB + 1 tDIS END DATA (n – 2) MSB LSB tDIS CFG LSB X X tHDIN CFG MSB – 1 BEGIN CFG (n + 1) tHSDO tDSDO tEN 15 2 tSDIN CFG MSB X 14 16/ 30 ACQUISITION (n + 1) UPDATE (n + 1) CFG/SDO SEE NOTE tCLSCK tSCKL (QUIET TIME) CONVERSION (n) ACQUISITION (n) tEN END CFG (n + 1) SEE NOTE MSB – 1 BEGIN DATA (n – 1) LSB + 1 tDIS END DATA (n – 1) NOTES 1. THE LSB IS FOR CONVERSION RESULTS OR THE CONFIGURATION REGISTER CFG (n – 1) IF. 15 SCK FALLING EDGES = LSB OF CONVERSION RESULTS. 29 SCK FALLING EDGES = LSB OF CONFIGURATION REGISTER. ON THE 16TH OR 30TH SCK FALLING EDGE, SDO IS DRIVEN TO HIGH IMPENDANCE. Figure 37. Serial Interface Timing for the AD7699 Without a Busy Indicator Rev. 0 | Page 23 of 28 LSB tDIS 07354-035 ACQUISITION (n – 1) AD7699 GENERAL TIMING WITH A BUSY INDICATOR fully prior to EOC, the last bit clocked out remains. If this bit is low, the busy signal indicator cannot be generated because the digital output requires a high impedance, or a bit remaining high, to low transition for the interrupt input of the host. A good example of this occurs when an SPI host sends 16 SCKs because these are usually limited to 8-bit or 16-bit bursts, thus the LSB remains. Because the transition noise of the AD7699 is 4 LSBs peak to peak (or greater), the LSB is low 50% of the time. For this interface, the SPI host needs to burst 24 SCKs, or a QSPI interface can be used and programmed for 17 SCKs. Figure 38 details the timing for all three modes: reading/writing during conversion, after conversion, and spanning conversion. Note that the gating item for both CFG and data readback is at the end of conversion (EOC). As detailed previously, the data access should occur up to safe data reading/writing time, tDATA. If the full CFG word is not written to prior to EOC, it is discarded and the current configuration remains. At the EOC, if CNV is low, the busy indicator is enabled. In addition, to generate the busy indicator properly, the host must assert a minimum of 17 SCK falling edges to return SDO to high impedance because the last bit of data on SDO remains active. Unlike the case detailed in the General Timing Without a Busy Indicator section, if the conversion result is not read out The SCK can idle high or low depending on the CPOL and CPHA settings if SPI is used. A simple solution is to use CPOL = CPHA = 1 (not shown) with SCK idling high. START OF CONVERSION tCYC tCONV POWER UP READ/WRITE AFTER CONVERT EOC CONVERSION (n – 2) ACQUISITION (n –1) CONVERSION (n – 1) ACQUISITION (n) CONVERSION (n) ACQUISITION (n + 1) CONVERSION (n + 1) ACQUISITION (n + 2) CNV DIN XXX CFG (n) SDO XXX DATA (n – 2) SCK 1 17/31 1 CFG (n + 2) CFG (n + 1) DATA (n – 1) 17/31 1 DATA (n) 17/31 1 17/31 CNV DIN CFG (n) CFG (n + 1) SDO DATA (n – 2) DATA (n – 1) SCK READ/WRITE SPANNING CONVERT EOC 1 17/31 1 CFG (n + 2) CFG (n + 3) DATA (n) 17/31 DATA (n + 1) 17/31 1 1 CNV SDO SCK DATA (n – 2) 1 CFG (n + 2) CFG (n + 1) CFG (n) DIN DATA (n – 2) 17/31 DATA (n – 1) DATA (n – 1) 1 17/31 DATA (n) 1 NOTES 1. CNV MUST BE HIGH PRIOR TO THE END OF CONVERSION (EOC) TO AVOID THE BUSY INDICATOR. A TOTAL OF 17 SCK FALLING EDGES IS REQUIRED TO RETURN SDO TO HIGH-Z. IF CFG READBACK IS ENABLED, A TOTAL OF 31 SCK FALLING EDGES IS REQUIRED TO RETURN SDO TO HIGH-Z. Figure 38. General Interface Timing for the AD7699 With a Busy Indicator Rev. 0 | Page 24 of 28 CFG (n + 3) DATA (n) 17/31 DATA (n + 1) 1 07354-036 READ/WRITE DURING CONVERT PHASE END OF CONVERSION (EOC) tDATA AD7699 READ/WRITE SPANNING CONVERSION WITH A BUSY INDICATOR update. While CNV is low, both a CFG update and a data readback take place. The first 14 SCK rising edges are used to update the CFG register, and the first 16 SCK falling edges clock out the conversion results starting with the MSB. The restriction for both configuring and reading is that they both occur before the tDATA time elapses for the next conversion. All 14 bits of CFG[13:0] must be written or they are ignored. Also, if the 16-bit conversion result is not read back before tDATA elapses, it is lost. This mode is used when the AD7699 is connected to any host using an SPI, serial port, or FPGA with an interrupt input. The connection diagram is shown in Figure 39, and the corresponding timing is given in Figure 40. For SPI, the host should use CPHA = CPOL = 1. Reading/writing spanning conversion is shown, which covers all three modes detailed in the Digital Interface section. The SDO data is valid on both SCK edges. Although the rising edge can be used to capture the data, a digital host using the SCK falling edge allows a faster reading rate, provided it has an acceptable hold time. After the optional 17th SCK falling edge, SDO returns to high impedance. Note that, if the optional SCK falling edge is not used, the busy feature cannot be detected if the LSB for the conversion is low. A rising edge on CNV initiates a conversion, forces SDO to high impedance, and ignores data present on DIN. After a conversion is initiated, it continues until completion irrespective of the state of CNV. CNV must be returned low before the safe data transfer time, tDATA, and then held low beyond the conversion time, tCONV, to generate the busy signal indicator. When the conversion is complete, SDO transitions from high impedance to low with a pull-up to VIO, which can be used to interrupt the host to begin data transfer. If CFG readback is enabled, the CFG register associated with the conversion result (n − 1) is read back MSB first following the LSB of the conversion result. A total of 31 SCK falling edges is required to return SDO to high impedance if this is enabled. After the conversion is complete, the AD7699 enters the acquisition phase and power-down. The host must enable the MSB of CFG at this time (if necessary) to begin the CFG VIO AD7699 DIGITAL HOST SDO MISO IRQ DIN MOSI SCK SCK 07354-037 SS CNV FOR SPI USE CPHA = 1, CPOL = 1. Figure 39. Connection Diagram for the AD7699 with a Busy Indicator tCYC tCONV tACQ tDATA tDATA tCNVH CNV (QUIET TIME) CONVERSION (n – 1) tSCK tSCKH 15 SCK 16 17/ 31 2 15 16 17/ 31 X X X ACQUISITION (n + 1) UPDATE (n + 1) CFG/SDO tHDIN tSDIN X X LSB +1 END DATA (n – 2) CFG CFG MSB MSB –1 X tDIS END CFG (n) SDO (QUIET TIME) CONVERSION (n) SEE NOTE 1 tSCKL DIN ACQUISITION (n) UPDATE (n) CFG/SDO BEIGN CFG (n + 1) MSB LSB tEN tHSDO tDSDO tEN MSB –1 BEGIN DATA (n – 1) LSB +1 tDIS LSB END DATA (n – 1) SEE NOTE NOTES 1. THE LSB IS FOR CONVERSION RESULTS OR THE CONFIGURATION REGISTER CFG (n – 1) IF. 16 SCK FALLING EDGES = LSB OF CONVERSION RESULTS. 30 SCK FALLING EDGES = LSB OF CONFIGURATION REGISTER. ON THE 17TH OR 31st SCK FALLING EDGE, SDO IS DRIVEN TO HIGH IMPENDANCE. OTHERWISE, THE LSB REMAINS ACTIVE UNTIL THE BUSY INDICATOR IS DRIVEN LOW. Figure 40. Serial Interface Timing for the AD7699 with a Busy Indicator Rev. 0 | Page 25 of 28 tDIS END CFG (n + 1) tEN 07354-038 CONVERSION (n – 1) AD7699 APPLICATION HINTS LAYOUT The printed circuit board that houses the AD7699 should be designed so that the analog and digital sections are separated and confined to certain areas of the board. The pinout of the AD7699, with all its analog signals on the left side and all its digital signals on the right side, eases this task. Avoid running digital lines under the device because these couple noise onto the die unless a ground plane under the AD7699 is used as a shield. Fast switching signals, such as CNV or clocks, should not run near analog signal paths. Crossover of digital and analog signals should be avoided. At least one ground plane should be used. It can be common or split between the digital and analog sections. In the latter case, the planes should be joined underneath the AD7699. The AD7699 voltage reference input, REF, has a dynamic input impedance and should be decoupled with minimal parasitic inductances. This is done by placing the reference decoupling ceramic capacitor close to, ideally right up against, the REF and GND pins and connecting them with wide, low impedance traces. Finally, the power supplies, VDD and VIO, of the AD7699 should be decoupled with ceramic capacitors, typically 100 nF, placed close to the AD7699 and connected using short, wide traces to provide low impedance paths and to reduce the effect of glitches on the power supply lines. EVALUATING AD7699 PERFORMANCE Other recommended layouts for the AD7699 are outlined in the documentation of the evaluation board for the AD7699 (EVALAD76MUXCBZ). The evaluation board package includes a fully assembled and tested evaluation board, documentation, and software for controlling the board from a PC via the evaluation controller board, EVAL-CONTROL BRD3. Rev. 0 | Page 26 of 28 AD7699 OUTLINE DIMENSIONS 0.60 MAX 4.00 BSC SQ 0.60 MAX 15 PIN 1 INDICATOR 20 16 1 PIN 1 INDICATOR 3.75 BSC SQ 0.50 BSC 2.65 2.50 SQ 2.35 EXPOSED PAD (BOTTOM VIEW) 5 10 1.00 0.85 0.80 12° MAX SEATING PLANE 0.50 0.40 0.30 0.80 MAX 0.65 TYP 0.30 0.23 0.18 0.05 MAX 0.02 NOM COPLANARITY 0.08 0.20 REF 0.25 MIN FOR PROPER CONNECTION OF THE EXPOSED PAD, REFER TO THE PIN CONFIGURATION AND FUNCTION DESCRIPTIONS SECTION OF THIS DATA SHEET. COMPLIANT TO JEDEC STANDARDS MO-220-VGGD-1 090408-B TOP VIEW 6 11 Figure 41. 20-Lead Lead Frame Chip Scale Package (LFCSP_VQ) 4 mm × 4 mm Body, Very Thin Quad (CP-20-4) Dimensions shown in millimeters ORDERING GUIDE Model AD7699BCPZ1 AD7699BCPZRL71 EVAL-AD7699CBZ1 EVAL-CONTROL BRD3Z1, 2 1 2 Integral Nonlinearity ±1.5 LSB max ±1.5 LSB max No Missing Code 16 bits 16 bits Temperature Range −40°C to +85°C −40°C to +85°C Package Description 20-Lead LFCSP_VQ 20-Lead LFCSP_VQ Evaluation Board Controller Board Package Option CP-20-4 CP-20-4 RoHS Compliant Part. This controller board allows a PC to control and communicate with all Analog Devices evaluation boards whose model numbers end in CB. Rev. 0 | Page 27 of 28 Ordering Quantity Tray, 490 Reel, 1500 AD7699 NOTES ©2008 Analog Devices, Inc. All rights reserved. Trademarks and registered trademarks are the property of their respective owners. D07354-0-10/08(0) Rev. 0 | Page 28 of 28