a LC2MOS Complete, 12-Bit, 100 kHz , Sampling ADC AD7870A FEATURES Complete Monolithic 12-Bit ADC with: 2 ms Track/Hold Amplifier 8 ms A/D Converter On-Chip Reference Laser-Trimmed Clock Parallel, Byte and Serial Digital Interface 70 dB SNR at 10 kHz Input Frequency 57 ns Data Access Time Low Power—60 mW typ FUNCTIONAL BLOCK DIAGRAM APPLICATIONS Digital Signal Processing Speech Recognition and Synthesis Spectrum Analysis High Speed Modems DSP Servo Control GENERAL DESCRIPTION The AD7870A is a fast, complete, 12-bit A/D converter. It consists of a track/hold amplifier, 8 µs successive approximation ADC, 3 V buried Zener reference and versatile interface logic. The ADC features a self-contained internal clock that is laser trimmed to guarantee accurate control of conversion time. No external clock timing components are required; the on-chip clock may be overridden by an external clock if required. AD7870A offers a choice of three data output formats: a single, parallel, 12-bit word, two 8-bit bytes or serial data. Fast bus access times and standard control inputs ensure easy interfacing to modern microprocessors and digital signal processors. The AD7870A operates from ± 5 V power supplies, accepts bipolar input signals of ± 3 V and can convert full power signals up to 50 kHz. In addition to the traditional dc accuracy specifications such as linearity, full-scale and offset errors, the AD7870A is also fully specified for dynamic performance parameters including harmonic distortion and signal-to-noise ratio. The AD7870A is fabricated in Analog Devices’ linear compatible CMOS (LC2MOS) process, a mixed technology process that combines precision bipolar circuits with low power CMOS logic. The part is available in a 24-pin, 0.3-inch wide, plastic dual inline package (DIP). PRODUCT HIGHLIGHTS 1. Complete 12-bit ADC on a chip. The AD7870A is the most complete monolithic ADC available and combines a 12-bit ADC with internal clock, track/ hold amplifier and reference on a single chip. 2. Dynamic specifications for DSP users. The AD7870A is fully specified and tested for ac parameters, including signal-to-noise ratio, harmonic distortion and intermodulation distortion. Key digital timing parameters are also tested and guaranteed over the full operating temperature range. 3. Fast microprocessor interface. Data access times of 57 ns make the AD7870A compatible with modern 8- and 16-bit microprocessors and digital signal processors. 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 which may result from its use. No license is granted by implication or otherwise under any patent or patent rights of Analog Devices. One Technology Way, P.O. Box 9106, Norwood, MA 02062-9106, U.S.A. Tel: 617/329-4700 World Wide Web Site: http://www.analog.com Fax: 617/326-8703 © Analog Devices, Inc., 1997 (VDD = +5 V 6 5%, VSS = –5 V 6 5%, AGND = DGND = 0 V, fCLK is internal, unless MIN to TMAX unless otherwise noted.) AD7870A–SPECIFICATIONS otherwise noted. All specifications T J1 Units Test Conditions/Comments DYNAMIC PERFORMANCE Signal-to-Noise Ratio3 (SNR) @ +25°C TMIN to TMAX Total Harmonic Distortion (THD) 70 70 –80 dB min dB min dB max Peak Harmonic or Spurious Noise –80 dB max VIN = 10 kHz Sine Wave, f SAMPLE = 100 kHz Typically 71.5 dB for 0 < VIN ≤ 50 kHz VIN = 10 kHz Sine Wave, fSAMPLE = 100 kHz Typically –86 dB for 0 < VIN < 50 kHz VIN = 10 kHz, fSAMPLE = 100 kHz Typically –86 dB for 0 < VIN < 50 kHz Intermodulation Distortion (IMD) Second Order Terms Third Order Terms Track/Hold Acquisition Time –80 –80 2 dB max dB max µs max Parameter 2 DC ACCURACY Resolution Minimum Resolution for Which No Missing Codes Are Guaranteed Integral Nonlinearity Bipolar Zero Error Positive Full-Scale Error4 Negative Full-Scale Error4 12 Bits 12 ± 1/2 ±5 ±5 ±5 Bits LSB typ LSB max LSB max LSB max ANALOG INPUT Input Voltage Range Input Current ±3 Volts ± 500 µA max REFERENCE OUTPUT REF OUT @ +25°C fa = 9 kHz, fb = 9.5 kHz, fSAMPLE = 50 kHz fa = 9 kHz, fb = 9.5 kHz, fSAMPLE = 50 kHz 2.99 3.01 ± 60 ±1 V min V max ppm/°C max mV max 2.4 0.8 ± 10 ± 10 10 V min V max µA max µA max pF max VDD = 5 V ± 5% VDD = 5 V ± 5% VIN = 0 V to VDD VIN = VSS to VDD 4.0 0.4 V min V max ISOURCE = 40 µA ISINK = 1.6 mA ± 10 15 µA max pF max CONVERSION TIME External Clock (fCLK = 2.5 MHz) Internal Clock 7.6/8 8/10 µs min/µs max µs min/µs max POWER REQUIREMENTS VDD VSS IDD ISS Power Dissipation +5 V –5 13 6 95 V nom V non mA max mA max mW max REF OUT Tempco Reference Load Sensitivity (∆REF OUT/∆I) LOGIC INPUTS Input High Voltage, VINH Input Low Voltage, VINL Input Current, IIN Input Current (12/8 CLK Input Only) Input Capacitance, CIN5 LOGIC OUTPUTS Output High Voltage, VOH Output Low Voltage, VOL DB11–DB0 Floating State Leakage Current Floating-State Output Capacitance5 Reference Load Current Change (0 µA–500 µA) Reference Load Should Not Be Change During Conversion ± 5% for Specified Performance ± 5% for Specified Performance Typically 8 mA Typically 4 mA Typically 60 mW NOTES 1 Temperature range is as follow: J Version: 0°C to +70°C. 2 VIN (pk-pk) = ± 3 V. 3 SNR calculation includes distortion and noise components. 4 Measured with respect to internal reference and includes bipolar offset error. 5 Sample tested @ +25°C to ensure compliance. Specifications subject to change without notice. –2– REV. 0 AD7870A TIMING CHARACTERISTICS1, 2 (V Parameter t1 t2 t3 t4 t5 t6 3 t7 4 t8 t9 t10 t115 t126 t13 t14 DD = +5 V 6 5%, VSS = –5 V 6 5%, AGND = DGND = 0 V. See Figures 9 and 10.) Limit at TMIN, TMAX (J Version) Units Conditions/Comments 50 0 60 0 70 57 5 50 0 0 100 370 135 100 10 100 ns min ns min ns min ns min ns max ns max ns min ns max ns min ns min ns min ns min ns max ns min ns min ns max CONVST Pulse Width CS to RD Setup Time (Mode 1) RD Pulse Width CS to RD Hold Time (Mode 1) RD to INT Delay Data Access Time after RD Bus Relinquish Time after RD HBEN to RD Setup Time HBEN to RD Hold Time SSTRB to SCLK Falling Edge Setup Time SCLK Cycle Time SCLK to Valid Data Delay. CL = 35 pF SCLK Rising Edge to SSTRB Bus Relinquish Time after SCLK NOTES 1 Timing specifications in bold print are 100% production tested. All other times are sample tested at +25 °C to ensure compliance. All input signals are specified with tr = tf = 5 ns (10% to 90% of 5 V) and timed from a voltage level of 1.6 V. 2 Serial timing is measured with a 4.7 kΩ pull-up resistor on SDATA and SSTRB and a 2 kΩ pull-up on SCLK. The capacitance on all three outputs is 35 pF. 3 t6 is measured with the load circuits of Figure 1 and defined as the time required for an output to cross 0.8 V or 2.4 V. 4 t7 is defined as the time required for the data lines to change 0.5 V when loaded with the circuits of Figure 2. 5 SCLK mark/space ratio (measured from a voltage level of 1.6 V) is 40/60 to 60/40. 6 t6 SDATA will drive higher capacitive loads but this will add to t 12 since it increases the external RC time constant (4.7 kΩiCL) and hence the time to reach 2.4 V. Specifications subject to change without notice. a. High-Z to VOH b. High-Z to VOL a. VOH to High-Z Figure 1. Load Circuits for Access Time REV. 0 b. VOL to High-Z Figure 2. Load Circuits for Output Float Delay –3– AD7870A Storage Temperature Range . . . . . . . . . . . . –65°C to +150°C Lead Temperature (Soldering, 10 sec) . . . . . . . . . . . . +300°C Power Dissipation (Any Package) to +75°C . . . . . . . 450 mW Derates above +75°C by . . . . . . . . . . . . . . . . . . . . 10 mW/°C ABSOLUTE MAXIMUM RATINGS* VDD to AGND . . . . . . . . . . . . . . . . . . . . . . . . . –0.3 V to +7 V VSS to AGND . . . . . . . . . . . . . . . . . . . . . . . . . +0.3 V to –7 V AGND to DGND . . . . . . . . . . . . . . . . . –0.3 V to VDD +0.3 V VIN to AGND . . . . . . . . . . . . . . . . . . . . . . . . . –15 V to +15 V REF OUT to AGND . . . . . . . . . . . . . . . . . . . . . . . 0 V to VDD Digital Inputs to DGND . . . . . . . . . . . . –0.3 V to VDD +0.3 V Digital Outputs to DGND . . . . . . . . . . . –0.3 V to VDD +0.3 V Operating Temperature Range Commercial (J Version) . . . . . . . . . . . . . . . . . . . 0°C to +70°C *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 listed in the operational sections of this specification is not implied. Exposure to absolute maximum rating conditions for extended periods may affect device reliability. CAUTION ESD (electrostatic discharge) sensitive device. Electrostatic charges as high as 4000 V readily accumulate on the human body and test equipment and can discharge without detection. Although the AD7870A features proprietary ESD protection circuitry, permanent damage may occur on devices subjected to high energy electrostatic discharges. Therefore, proper ESD precautions are recommended to avoid performance degradation or loss of functionality. WARNING! ESD SENSITIVE DEVICE PIN CONFIGURATION ORDERING GUIDE DIP Model Temperature Range SNR (dBs) Relative Accuracy (LSB) Package Option* AD7870AJN 0°C to +70°C 70 min ± 1/2 typ N-24 *N = Plastic DIP. –4– REV. 0 AD7870A PIN DESCRIPTION Pin No. Pin Mnemonic 1 2 3 RD INT CLK 4 DB11/HBEN 5 DB10/SSTRB 6 DB9/SCLK 7 DB8/SDATA 8–11 DB7/LOW– DB4/LOW Function 12 DGND 13-16 DB3/DB11– DB0/DB8 17 18 19 VDD AGND REF OUT 20 21 22 VIN VSS 12/8/CLK 23 CONVST 24 CS Read. Active low logic input. This input is used in conjunction with CS low to enable the data outputs. Interrupt, Active low logic output indicating converter status. See timing diagrams. Clock input. An external TTL-compatible clock may be applied to this input pin. Alternatively, tying this pin to VSS enables the internal laser-trimmed clock oscillator. Data Bit 11 (MSB)/High Byte Enable. The function of this pin is dependent on the state of the 12/8/CLK input (see below). When 12-bit parallel data is selected, this pin provides the DB11 output. When byte data is selected, this pin becomes the HBEN logic input. HBEN is used for 8-bit bus interfacing. When HBEN is low, DB7/LOW to DB0/DB8 become DB7 to DB0. With HBEN high, DB7/LOW to DB0/DB8 are used for the upper byte of data (see Table I). Data Bit 10/Serial Strobe. When 12-bit parallel data is selected, this pin provides the DB10 output. SSTRB is an active low open-drain output that provides a strobe or framing pulse for serial data. An external 4.7 kΩ pull-up resistor is required on SSTRB. Data Bit 9/Serial Clock. When 12-bit parallel data is selected, this pin provides the DB9 output. SCLK is the gated serial clock output derived from the internal or external ADC clock. If the 12/8/CLK input is at –5 V, then SCLK runs continuously. If 12/8/CLK is at 0 V, then SCLK is gated off after serial transmission is complete. SCLK is an open-drain output and requires an external 2 kΩ pull-up resistor. Data Bit 8/Serial Data. When 12-bit parallel data is selected, this pin provides the DB8 output. SDATA is an open-drain serial data output which is used with SCLK and SSTRB for serial data transfer. Serial data is valid on the falling edge of SCLK while SSTRB is low. An external 4.7 kΩ pull-up resistor is required on SDATA. Three-state data outputs controlled by CS and RD. Their function depends on the 12/8/CLK and HBEN inputs. With 12/8/CLK high, they are always DB7–DB4. With 12/8/CLK low or –5 V, their function is controlled by HBEN (see Table I). Digital Ground. Ground reference for digital circuitry. Three-state data outputs which are controlled by CS and RD. Their function depends on the 12/8/CLK and HBEN inputs. With/12/8/CLK high, they are always DB3–DB0. With 12/8/CLK low or –5 V, their function is controlled by HBEN (see Table I). Positive Supply, +5 V ± 5%. Analog Ground. Ground reference for track/hold, reference and DAC. Voltage Reference Output. The internal 3 V reference is provided at this pin. The external load capability is 500 µA. Analog Input. The analog input range is ± 3 V. Negative Supply, –5 V ± 5%. Three Function Input. Defines the data format and serial clock format. With this pin at +5 V, the output data format is 12-bit parallel only. With this pin at 0 V, either byte or serial data is available and SCLK is not continuous. With this pin at –5 V, either byte or serial data is again available but SCLK is now continuous. Convert Start. A high to low transition on this input puts the track/hold into its hold mode and starts conversion. This input is asynchronous to the CLK and independent of CS and RD. Chip Select. Active low logic input. The device is selected when this input is active. Table I. Output Data for Byte Interfacing HBEN DB7/LOW DB6/LOW DB5/LOW DB4/LOW DB3/DB11 HIGH LOW LOW LOW LOW LOW DB7 DB6 DB5 DB4 REV. 0 –5– DB2/DB10 DB1/DB9 DB0/DB8 DB11 (MSB) DB10 DB9 DB8 DB3 DB1 DB0 (LSB) DB2 AD7870A CONVERTER DETAILS The AD7870A is a complete 12-bit A/D converter, requiring no external components apart from power supply decoupling capacitors. It is comprised of a 12-bit successive approximation ADC based on a fast settling voltage-output DAC, a high speed comparator and SAR, a track/hold amplifier, a 3 V buried Zener reference, a clock oscillator and control logic. INTERNAL REFERENCE The AD7870A has an on-chip temperature compensated buried Zener reference that is factory trimmed to 3 V ± 10 mV. Internally it provides both the DAC reference and the dc bias required for bipolar operation. The reference output is available (REF OUT) and capable of providing up to 500 µA to an external load. Figure 4. Analog Input ANALOG INPUT Figure 4 shows the AD7870A analog input. The analog input range is ± 3 V into an input resistance of typically 15 kΩ. The designed code transitions occur midway between successive integer LSB values (i.e., 1/2 LSB, 3/2 LSBs, 5/2 LSBs . . . FS 3/2 LSBs). The output code is twos complement binary with 1 LSB = FS/4096 = 6 V/4096 = 1.46 mV. The ideal input/output transfer function is shown in Figure 5. The maximum recommended capacitance on REF OUT for normal operation is 50 pF. If the reference is required for use external to the AD7870A, it should be decoupled with a 200 Ω resistor in series with a parallel combination of a 10 µF tantalum capacitor and a 0.1 µF ceramic capacitor. These decoupling components are required to remove voltage spikes caused by the AD7870A’s internal operation. Figure 3. Reference Circuit Figure 5. Bipolar Input/Output Transfer Function TRACK-AND-HOLD AMPLIFIER The track-and-hold amplifier on the analog input of the AD7870A allows the ADC to accurately convert an input sine wave of 6 V peak-peak amplitude to 12-bit accuracy. The input bandwidth of the track/hold amplifier is much greater than the Nyquist rate of the ADC even when the ADC is operated at its maximum throughput rate. The 0.1 dB cutoff frequency occurs typically at 500 kHz. The track/hold amplifier acquires an input signal to 12-bit accuracy in less than 2 µs. The overall throughput rate is equal to the conversion time plus the track/hold amplifier acquisition time. For a 2.5 MHz input clock the throughput rate is 10 µs max. BIPOLAR OFFSET AND FULL SCALE ADJUSTMENT In most digital signal processing (DSP) applications, offset and full-scale errors have little or no effect on system performance. Offset error can always be eliminated in the analog domain by ac coupling. Full-scale error effect is linear and does not cause problems as long as the input signal is within the full dynamic range of the ADC. Some applications will require that the input signal span the full analog input dynamic range. In such applications, offset and full-scale error will have to be adjusted to zero. Where adjustment is required, offset error must be adjusted before full-scale error. This is achieved by trimming the offset of the op amp driving the analog input of the AD7870A while the input voltage is 1/2 LSB below ground. The trim procedure is as follows: apply a voltage of 0.73 mV (–1/2 LSB) at V1 in Figure 6 and adjust the op amp offset voltage until the ADC output code flickers between 1111 1111 1111 and 0000 0000 0000. Gain error can be adjusted at either the first code transition (ADC negative full scale) or the last code transition (ADC positive full scale). The trim procedures for both cases are as follows (see Figure 6). The operation of the track/hold is essentially transparent to the user. The track/hold amplifier goes from its tracking mode to its hold mode at the start of conversion. The track-to-hold transition occurs on the falling edge of CONVST. –6– REV. 0 AD7870A The ideal input/output transfer function is shown in Figure 8. The output can be converted to natural binary by inverting the MSB. Figure 6. Full-Scale Adjust Circuit Positive Full-Scale Adjust Apply a voltage of 2.9978 V (FS/2 – 3/2 LSBs) at V1. Adjust R2 until the ADC output code flickers between 0111 1111 1110 and 0111 1111 1111. Negative Full-Scale Adjust Apply a voltage of –2.9993 V (–FS/2 +1/ 2 LSB) at V1 and adjust R2 until the ADC output code flickers between 1000 0000 0000 and 1000 0000 0001. UNIPOLAR OPERATION A typical unipolar circuit is shown in Figure 7. The AD7870A REF OUT is used to offset the analog input by 3 V. The analog input range is determined by the ratio of R3 to R4. The minimum range with which the circuit will work is 0 V to +3 V (R3 = 0, R4 = O/C). The resistor values are given in Figure 7 for input ranges of 0 V to +5 V and 0 V to +10 V. R5 and R6 are included for offset and full-scale adjust only and should be omitted if adjustment is not required. Figure 8. Unipolar Transfer Function UNIPOLAR OFFSET AND FULL-SCALE ADJUSTMENT When absolute accuracy is required, offset and full-scale error can be adjusted to zero. Offset must be adjusted before full scale. This is achieved by applying an input voltage of (1/2 LSB) to Vl and adjust R6 until the ADC output code flickers between 1000 0000 0000 and 1000 0000 0001. For full-scale adjustment, apply an input voltage of (FS–3/2 LSBs) to Vl and adjust R5 until the output code flickers between 0111 1111 1110 and 0111 11111111. TIMING AND CONTROL The AD7870A is capable of one basic interfacing mode. In this mode (Mode 1), the CONVST line is used to start conversion and drive the track/hold into its hold mode. At the end of conversion the track/hold returns to its tracking mode. It is principally intended for digital signal processing and other applications where precise sampling in time is required. In these applications, it is important that the signal sampling occurs at exactly equal intervals to minimize errors due to sampling uncertainty or jitter. For these cases, the CONVST line is driven by a timer or some precise clock source. DATA OUTPUT FORMATS Figure 7. Unipolar Circuit REV. 0 The AD7870A offers a choice of three data output formats, one serial and two parallel. The parallel data formats are a single, 12-bit parallel word for 16-bit data buses and a two-byte format for 8-bit data buses. The data format is controlled by the 12/8/CLK input. A logic high on this pin selects the 12-bit parallel output format only. A logic low or –5 V applied to this pin allows the user access to either serial or byte formatted data. Three of the pins previously assigned to the four MSBs in parallel form are now used for serial communications while the fourth pin becomes a control input for the byte-formatted data. –7– AD7870A during the serial transmission only. In these cases, it can be shut down at the end of conversion to allow multiple ADCs to share a common serial bus. However, some serial systems (e.g., TMS32020) require a serial clock that runs continuously. Both options are available on the AD7870A using the 12/8/CLK input. With this input at –5 V, the serial clock (SCLK) runs continuously; when 12/8/CLK is at 0 V, SCLK is turned off at the end of transmission. Parallel Output Format The two parallel formats available on the AD7870A are a 12-bit wide data word and a two-byte data word. In the first, all 12 bits of data are available at the same time on DB11 (MSB) through DB0 (LSB). In the second, two reads are required to access the data. When this data format is selected, the DB11/HBEN pin assumes the HBEN function. HBEN selects which byte of data is to be read from the AD7870A. When HBEN is low, the lower eight bits of data are placed on the data bus during a read operation; with HBEN high, the upper four bits of the 12-bit word are placed on the data bus. These four bits are right justified and thereby occupy the lower nibble of data while the upper nibble contains four zeros. MODE 1 INTERFACE Conversion is initiated by a low going pulse on the CONVST input. The falling edge of this CONVST pulse starts conversion and drives the track/hold amplifier into its hold mode. INT is normally high and goes low at the end of conversion. This INT line can be used to interrupt the microprocessor. A read operation to the AD7870A accesses the data and the INT line is reset high on the falling edge of CS and RD. Trying to exercise CS and RD during a conversion can cause errors to the conversion in progress. In applications where precise sampling is not critical, the CONVST pulse can be generated from a microprocessor WR line OR-gated with a decoded address. Serial Output Format Serial data is available on the AD7870A when the 12/8/CLK input is at 0 V or –5 V and in this case the DB10/SSTRB, DB9/SCLK and DB8/SDATA pins assume their serial functions. Serial data is available during conversion with a word length of 16 bits; four leading zeros, followed by the 12-bit conversion result starting with the MSB. The data is synchronized to the serial clock output (SCLK) and framed by the serial strobe (SSTRB). Data is clocked out on a low to high transition of the serial clock and is valid on the falling edge of this clock while the SSTRB output is low. SSTRB goes low within three clock cycles after CONVST, and the first serial data bit (the first leading zero) is valid on the first falling edge of SCLK after SSTRB goes low. All three serial lines are open-drain outputs and require external pull-up resistors. Figure 9 shows the timing diagram for a 12-bit parallel data output format (12/8/CLK = +5 V). A read to the AD7870A at the end of conversion accesses all 12 bits of data at the same time. CS and RD control the output three-state drivers. If CS and RD are high, the databus is three-state. If CS and RD are hardwired low, data from the previous conversion will remain on the databus. This data will be updated approximately tCONVERT after the falling edge of CONVST. With CS and RD hardwired low, the INT line will remain low. Serial data is not available for this data output format. The serial clock out is derived from the ADC clock source, which may be internal or external. Normally, SCLK is required Figure 9. Mode 1 Timing Diagram, 12-Bit Parallel Read –8– REV. 0 AD7870A Figure 10. Mode 1 Timing Diagram, Byte or Serial Read The timing diagram for byte and serial data is shown in Figure 10. INT goes low at the end of conversion and is reset high by the first falling edge of CS and RD. This first read at the end of conversion can either access the low byte or high byte of data depending on the status of HBEN (Figure 10 shows low byte only for example). The diagram shows both a noncontinuously and a continuously running clock (dashed line). AD7870A DYNAMIC SPECIFICATIONS The output spectrum from the ADC is evaluated by applying a sine-wave signal of very low distortion to the VIN input which is sampled at a 100 kHz sampling rate. A Fast Fourier Transform (FFT) plot is generated from which the SNR data can be obtained. Figure 11 shows a typical 2048 point FFT plot of the AD7870AJN with an input signal of 25 kHz and a sampling frequency of 100 kHz. The SNR obtained from this graph is 72.6 dB. It should be noted that the harmonics are taken into account when calculating the SNR. The AD7870A is specified and 100% tested for dynamic performance specifications. These ac specifications are required for signal processing applications such as speech recognition, spectrum analysis and high speed modems. These applications require information on the ADC’s effect on the spectral content of the input signal. Hence, the parameters for which the AD7870A is specified include SNR, harmonic distortion, intermodulation distortion and peak harmonics. These terms are discussed in more detail in the following sections. Signal-to-Noise Ratio (SNR) SNR is the measured signal-to-noise ratio at the output of the ADC. The signal is the rms magnitude of the fundamental. Noise is the rms sum of all the nonfundamental signals up to half the sampling frequency (FS/2) excluding dc. SNR is dependent upon the number of quantization levels used in the digitization process; the more levels, the smaller the quantization noise. The theoretical signal-to-noise ratio for a sine wave input is given by SNR =(6.02N + 1.76) dB (1) where N is the number of bits. Thus for an ideal 12-bit converter, SNR = 74 dB. REV. 0 Figure 11. FFT Plot –9– AD7870A it is the ratio of the rms sum of the individual distortion products to the rms amplitude of the fundamental expressed in dBs. In this case, the input consists of two, equal amplitude, low distortion sine waves. Figure 13 shows a typical IMD plot for the AD7870A. Effective Number of Bits The formula given in (1) relates the SNR to the number of bits. Rewriting the formula, as in (2), it is possible to obtain a measure of performance expressed in effective number of bits (N). N= SNR – 1.76 6.02 (2) The effective number of bits for a device can be calculated directly from its measured SNR. Figure 12 shows a typical plot of effective number of bits versus frequency for an AD7870AJN, with a sampling frequency of 100 kHz. The effective number of bits typically falls between 11.7 and 11.85 corresponding to SNR figures of 72.2 dB and 73.1 dB. Figure 12. Effective Number of Bits vs. Frequency Harmonic Distortion Figure 13. IMD Plot Harmonic distortion is the ratio of the rms sum of harmonics to the fundamental. For the AD7870A, total harmonic distortion (THD) is defined as 2 THD = 20 log 2 2 2 Peak Harmonic or Spurious Noise Peak harmonic or spurious noise is defined as the ratio of the rms value of the next largest component in the ADC output spectrum (up to FS/2 and excluding dc) to the rms value of the fundamental. Normally, the value of this specification will be determined by the largest harmonic in the spectrum, but for parts where the harmonics are buried in the noise floor the peak will be a noise peak. 2 V 2 +V 3 +V 4 +V 5 +V 6 V1 where V1 is the rms amplitude of the fundamental and V2, V3, V4, V5 and V6 are the rms amplitudes of the second through the sixth harmonic. The THD is also derived from the FFT plot of the ADC output spectrum. AC Linearity Plot Intermodulation Distortion With inputs consisting of sine waves at two frequencies, fa and fb, any active device with nonlinearities will create distortion products at sum and difference frequencies of mfa + nfb where m, n = 0, 1, 2, 3, etc. Intermodulation terms are those for which neither m nor n are equal to zero. For example, the second order terms include (fa + fb) and (fa – fb), while the third order terms include (2fa + fb), (2fa – fb), (fa + 2fb) and (fa – 2fb). Using the CCIF standard, where two input frequencies near the top end of the input bandwidth are used, the second and third order terms are of different significance. The second order terms are usually distanced in frequency from the original sine waves while the third order terms are usually at a frequency close to the input frequencies. As a result, the second and third order terms are specified separately. The calculation of the intermodulation distortion is as per the THD specification where When a sine wave of specified frequency is applied to the VIN input of the AD7870A, and several million samples are taken, a histogram showing the frequency of occurrence of each of the 4096 ADC codes can be generated. From this histogram data it is possible to generate an ac integral linearity plot as shown in Figure 14. This shows very good integral linearity performance from the AD7870A at an input frequency of 25 kHz. The absence of large spikes in the plot shows good differential linearity. Simplified versions of the formulae used are outlined below. V (i )–V (o) INL(i ) = × 4096 – i V ( fs)–V (o) where INL(i) is the integral linearity at code i. V(fs) and V(o) are the estimated full-scale and offset transitions and V(i) is the estimated transition for the ith code. –10– REV. 0 AD7870A V(i) the estimated code transition point is derived as follows: V (i ) = – A × Cos noisy code transitions. Other causes of concern are ground loops and digital feedthrough from microprocessors. These are factors that influence any ADC, and a proper PCB layout that minimizes these effects is essential for best performance. π × cum(i ) N where A is the peak signal amplitude, N is the number of histogram samples and cum(i) = ∑in = 0 V(n) occurrences LAYOUT HINTS Ensure that the layout for the printed circuit board has the digital and analog signal lines separated as much as possible. Take care not to run any digital track alongside an analog signal track. Guard (screen) the analog input with AGND. Establish a single point analog ground (star ground) separate from the logic system ground at the AD7870A AGND pin or as close as possible to the AD7870A. Connect all other grounds and the AD7870A DGND to this single analog ground point. Do not connect any other digital grounds to this analog ground point. Low impedance analog and digital power supply common returns are essential to low noise operation of the ADC, so make the foil width for these tracks as wide as possible. The use of ground planes minimizes impedance paths and also guards the analog circuitry from digital noise. Figure 14. AC INL Plot NOISE APPLICATION HINTS Good printed circuit board (PCB) layout is as important as the overall circuit design itself in achieving high speed A/D performance. The AD7870A is required to make bit decisions on an LSB size of 1.465 mV. Thus, the designer has to be conscious of noise both in the ADC itself and in the preceding analog circuitry. Switching mode power supplies are not recommended as the switching spikes will feed through to the comparator causing Keep the input signal leads to VIN and signal return leads from AGND as short as possible to minimize input noise coupling. In applications where this is not possible, use a shielded cable between the source and the ADC. Reduce the ground circuit impedance as much as possible since any potential difference in grounds between the signal source and the ADC appears as an error voltage in series with the input signal. OUTLINE DIMENSIONS Dimensions shown in inches and (mm). 24-Pin Plastic DIP (N-24) REV. 0 –11– –12– PRINTED IN U.S.A. C1628–7.5–2/92