CS5016 CS5014 CS5012A Semiconductor Corporation 16, 14 & 12-Bit, Self-Calibrating A/D Converters Features General Description CMOS A/D Converters • Monolithic Microprocessor Compatible The CS5012A/14/16 are 12, 14 and 16-bit monolithic analog to digital converters with conversion times of 7.2µs, 14.25µs and 16.25µs. Unique self-calibration circuitry insures excellent linearity and differential nonlinearity, with no missing codes. Offset and full scale errors are kept within 1/2 LSB (CS5012A/14) and 1 LSB (CS5016), eliminating the need for calibration. Unipolar and bipolar input ranges are digitally selectable. Parallel and Serial Output Inherent Track/Hold Input • True 12, 14 and 16-Bit Precision Times: • Conversion CS5016 16.25 µs The pin compatible CS5012A/14/16 consist of a DAC, conversion and calibration microcontroller, oscillator, comparator, microprocessor compatible 3-state I/O, and calibration circuitry. The input track-and-hold, inherent to the devices’ sampling architecture, acquires the input signal after each conversion using a fast slewing on-chip buffer amplifier. This allows throughput rates up to 100 kHz (CS5012A), 56 kHz (CS5014) and 50 kHz (CS5016). CS5014 14.25 µs CS5012A 7.20 µs Calibration Maintains Accuracy • Self Over Time and Temperature • Low Power Dissipation: • Low Distortion HOLD CS 1 RD 21 22 An evaluation board (CDB5012/14/16) is available which allows fast evaluation of ADC performance. 150 mW ORDERING INFORMATION: Pages 2-45, 2-46, & 2-47 BP/UP RST A0 23 24 32 BW INTRLV CAL 33 34 35 EOT 37 EOC SCLK SDATA 38 39 40 2 3 CLKIN 20 4 CLOCK GENERATOR CONTROL 5 6 REFBUF VREF 29 28 CALIBRATION MEMORY 7 MICROCONTROLLER 8 + 9 12 13 AIN 26 + CHARGE REDISTRIBUTION DAC 14 + 15 16 COMPARATOR AGND 27 + 17 18 STATUS REGISTER 19 25 VA+ 30 VA- Crystal Semiconductor Corporation P.O. Box 17847, Austin, TX 78760 (512) 445 7222 FAX: (512) 445 7581 11 VD+ 36 VD- 10 DGND D0 (LSB) CS5016 D1 D2 (LSB) CS5014 D3 D4 (LSB) CS5012A D5 D6 D7 D8 D9 D10 D11 D12 D13 D14 D15 (MSB) 31 TST Copyright Crystal Semiconductor Corporation 1995 (All Rights Reserved) MAR ’95 DS14F6 2-7 CS5012A CS5012A ANALOG CHARACTERISTICS (TA = TMIN to TMAX; VA+, VD+ = 5V; VA-, VD- = -5V; VREF = 2.5V to 4.5V; fclk = 6.4 MHz for -7, 4 MHz for -12; Analog Source Impedance = 200Ω) CS5012A-K Parameter* Min Specified Temperature Range Typ Max CS5012A-B Min 0 to +70 Typ Max CS5012-T Min Typ Max Units -40 to +85 -55 to +125 °C ±1/4 ±1/8 ±1/4 ±1/8 LSB12 ∆LSB12 Accuracy Linearity Error Drift (Note 1) (Note 2) ±1/4 ±1/8 Differential Linearity Drift (Note 1) (Note 2) ±1/4 ±1/2 ±1/32 ±1/4 ±1/2 ±1/32 ±1/4 ±1/2 LSB12 ∆LSB12 ±1/32 Full Scale Error Drift (Note 1) (Note 2) ±1/4 ±1/2 ±1/16 ±1/4 ±1/2 ±1/16 ±1/4 ±1/8 Unipolar Offset Drift (Note 1) (Note 2) ±1/4 ±1/2 ±1/16 ±1/4 ±1/2 ±1/16 ±1/4 ±1/2 LSB12 ∆LSB12 ±1/16 Bipolar Offset Drift (Note 1) (Note 2) ±1/4 ±1/2 ±1/16 ±1/4 ±1/2 ±1/16 ±1/4 ±1/2 LSB12 ∆LSB12 ±1/16 Bipolar Negative Full-Scale Error(Note 1) Drift (Note 2) ±1/4 ±1/2 ±1/16 ±1/4 ±1/2 ±1/16 ±1/4 ±1/2 LSB12 ∆LSB12 ±1/16 Total Unadjusted Error Drift ±1/4 ±1/4 ±1/4 ±1/4 ±1/4 ±1/4 LSB12 ∆LSB12 92 88 dB dB 0.008 % 73 13 dB dB 45 90 µVrms µVrms (Note 1) (Note 2) ±1/2 ±1/2 ±1/2 ±1/2 LSB12 ∆LSB12 Dynamic Performance (Bipolar Mode) Peak Harmonic or Spurious Noise Full Scale, 1 kHz Input Full Scale, 12 kHz Input (Note 1) 84 84 Total Harmonic Distortion 92 88 84 84 0.008 Signal-to-Noise Ratio 1 kHz, 0 dB Input 1 kHz, -60 dB Input (Note 1) Noise Unipolar Mode Bipolar Mode (Note 3) 72 73 13 45 90 92 88 84 84 0.008 72 73 13 45 90 72 Notes: 1. Applies after calibration at any temperature within the specified temperature range. 2. Total drift over specified temperature range since calibration at power-up at 25 °C 3. Wideband noise aliased into the baseband. Referred to the input. * Refer to Parameter Definitions (immediately following the pin descriptions at the end of this data sheet). Specifications are subject to change without notice. 2-8 DS14F6 CS5012A CS5012A ANALOG CHARACTERISTICS (continued) CS5012A-K Parameter* Min Specified Temperature Range Typ Max CS5012A-B Min Typ Max CS5012-T Min Typ Max Units 0 to +70 -40 to +85 -55 to +125 °C Aperture Time 25 25 25 ns Aperture Jitter 100 100 100 ps Analog Input Input Capacitance Unipolar Mode CS5012 CS5012A Bipolar Mode CS5012 CS5012A (Note 4) 275 103 165 72 375 137 220 96 275 103 165 72 375 137 220 96 275 103 165 72 375 137 220 96 pF pF pF pF 12.25 µs µs 3.75 µs µs Conversion & Throughput Conversion Time Acquisition Time Throughput -7 (Notes 5 and 6) -12 -7 -12 (Note 6) -7 -12 (Note 6) 100 62.5 7.2 12.25 2.5 3.0 7.2 12.25 2.8 3.75 2.5 3.0 2.8 3.75 100 62.5 3.0 kHz kHz 62.5 Power Supplies DC Power Supply Currents IA+ IA(CS5012) ID+ (CS5012A) ID+ IDPower Dissipation Power Supply Rejection Positive Supplies Negative Supplies (Note 7) (Note 7) 12 -12 3 6 -3 19 -19 6 7.5 -6 12 -12 3 6 -3 19 -19 6 7.5 -6 12 -12 3 19 -19 6 -3 -6 mA mA mA mA mA 150 250 150 250 150 250 mW (Note 8) 84 84 84 84 84 84 dB dB Notes: 4. Applies only in track mode. When converting or calibrating, input capacitance will not exceed 15 pF. 5. Measured from falling transition on HOLD to falling transition on EOC. 6. Conversion, acquisition, and throughput times depend on CLKIN, sampling, and calibration conditions. The numbers shown assume sampling and conversion is synchronized with the CS5012A/14/16 ’s conversion clock, interleave calibrate is disabled, and operation is from the full-rated, external clock. Refer to the section Conversion Time/Throughput for a detailed discussion of conversion timing. 7. All outputs unloaded. All inputs CMOS levels. 8. With 300 mV p-p, 1 kHz ripple applied to each analog supply separately in bipolar mode. Rejection improves by 6 dB in the unipolar mode to 90 dB. Figure 13 shows a plot of typical power supply rejection versus frequency. DS14F6 2-9 CS5014 CS5014 ANALOG CHARACTERISTICS (TA = TMIN to TMAX; VA+, VD+ = 5V; VA-, VD- = -5V; VREF = 4.5V; CLKIN = 4 MHz for -14, 2 MHz for -28; Analog Source Impedance = 200Ω) CS5014-K Parameter* Min Specified Temperature Range Typ CS5014-B Max Min 0 to +70 Typ CS5014-S, T Max -40 to +85 Min Typ Max Units -55 to +125 °C LSB14 LSB14 ∆LSB14 Accuracy (Note 1) ±1/4 (Note 2) ±1/8 ±1/8 ±1/4 ±1/2 ±1/8 Differential Linearity Drift (Note 1) (Note 2) ±1/4 ±1/2 ±1/32 ±1/4 ±1/2 ±1/32 ±1/4 ±1/2 LSB14 ∆LSB14 ±1/32 Full Scale Error Drift (Note 1) (Note 2) ±1/2 ±1/4 ±1 ±1/2 ±1/4 ±1 ±1/2 ±1/2 ±1 LSB14 ∆LSB14 (Note 1) ±1/4 ±3/4 ±1/4 ±3/4 ±1/4 (Note 2) ±1/4 ±3/4 ±1 LSB14 LSB14 ∆LSB14 (Note 1) ±1/4 (Note 2) ±1/4 ±3/4 ±1 LSB14 LSB14 ∆LSB14 ±1 ±1.5 LSB14 LSB14 ∆LSB14 Linearity Error K, B, T S Drift Unipolar Offset K, B, T S Drift Bipolar Offset K, B, T S Drift Bipolar Negative Full-Scale Error(Note 1) K, B, T S Drift (Note 2) Total Unadjusted Error Drift ±1/2 (Note 1) (Note 2) ±1/2 ±1/4 ±1/2 ±1/2 ±1/4 ±3/4 ±1/4 ±3/4 ±1/4 ±1/2 ±1/2 ±1 ±1/2 ±1 ±1/2 ±1/2 ±1.5 ±1/4 ±1/4 ±1/2 ±1 ±1/2 ±1 ±1 ±1 ±1 LSB14 ∆LSB14 98 dB dB dB dB Dynamic Performance (Bipolar Mode) Peak Harmonic or Spurious Noise Full Scale, 1 kHz Input Full Scale, 12 kHz Input (Note 1) K, B, T S K, B, T S 94 98 94 98 84 87 84 87 Total Harmonic Distortion Signal-to-Noise Ratio 1 kHz, 0 dB Input 0.003 (Notes 1 and 9) K, B, T S 1 kHz, -60 dB Input Noise Unipolar Mode Bipolar Mode 82 84 94 85 84 80 0.003 82 84 82 80 87 0.003 % 84 23 23 23 dB dB dB 45 90 45 90 45 90 µVrms µVrms (Note 3) Notes: 9. A detailed plot of S/(N+D) vs. input amplitude appears in Figure 26 for the CS5014 and Figure 28 for the CS5016. * Refer to Parameter Definitions (immediately following the pin descriptions at the end of this data sheet). Specifications are subject to change without notice. 2-10 DS14F6 CS5014 CS5014 ANALOG CHARACTERISTICS (continued) CS5014-K Parameter* Min Specified Temperature Range Typ CS5014-B Max Min Typ CS5014-S, T Max Min Typ Max Units 0 to +70 -40 to +85 -55 to +125 °C Aperture Time 25 25 25 ns Aperture Jitter 100 100 100 ps Analog Input Input Capacitance Unipolar Mode Bipolar Mode (Note 4) 275 165 375 220 275 165 375 220 275 165 375 220 pF pF 14.25 28.5 µs µs 3.75 5.25 µs µs Conversion & Throughput Conversion Time -14 (Notes 5 and 6) -28 Acquisition Time -14 -28 (Note 6) Throughput -14 -28 (Note 6) 55.6 27.7 14.25 28.5 3.0 4.5 14.25 28.5 3.75 5.25 3.0 4.5 3.75 5.25 55.6 27.7 3.0 4.5 55.6 27.7 kHz kHz Power Supplies DC Power Supply Currents IA+ IAID+ ID- (Note 7) Power Dissipation (Note 7) Power Supply Rejection Positive Supplies Negative Supplies (Note 8) DS14F6 9 -9 3 -3 19 -19 6 -6 9 -9 3 -3 19 -19 6 -6 9 -9 3 -3 19 -19 6 -6 mA mA mA mA 120 250 120 250 120 250 mW 84 84 84 84 84 84 dB dB 2-11 CS5016 CS5016 ANALOG CHARACTERISTICS (TA = TMIN to TMAX; VA+, VD+ = 5V; VA-, VD- = -5V; VREF = 4.5V; CLKIN = 4 MHz for -16, 2 MHz for -32; Analog Source Impedance = 200Ω; Synchronous Sampling.) CS5016-J, K Parameter* Min Specified Temperature Range Typ Max CS5016-A, B Min 0 to +70 Typ Max -40 to +85 CS5016-S, T Min Typ Max Units -55 to +125 °C Accuracy Linearity Error J, A, S K, B, T Drift (Note 1) 0.002 0.003 0.001 0.0015 ±1/4 (Note 2) Differential Linearity (Note 10) 16 0.002 0.003 0.001 0.0015 ±1/4 16 0.002 0.0076 %FS 0.001 0.0015 %FS ±1/4 ∆LSB16 16 Bits ±2 ±2 ±1 ±3 ±3 ±2 ±2 ±1 ±3 ±3 ±2 ±2 ±2 ±4 ±3 LSB16 LSB16 ∆LSB16 ±1 ±1 ±1 ±2 ±3/2 ±1 ±1 ±1 ±3 ±3 ±1 ±1 ±2 ±4 ±3 LSB16 LSB16 ∆LSB16 ±2 ±3/2 ±1 ±1 ±2 ±2 ±2 ±1 ±1 ±2 ±4 ±2 (Note 2) ±1 ±1 ±1 LSB16 LSB16 ∆LSB16 Bipolar Negative Full-Scale Error(Note 1) J, A, S K, B, T Drift (Note 2) ±2 ±2 ±1 ±3 ±3 ±2 ±2 ±2 ±3 ±3 ±2 ±2 ±2 ±5 ±3 LSB16 LSB16 ∆LSB16 Full Scale Error J, A, S K, B, T Drift Unipolar Offset (Note 2) J, A, S K, B, T Drift Bipolar Offset (Note 1) (Note 1) (Note 2) J, A, S K, B, T Drift (Note 1) Dynamic Performance (Bipolar Mode) Peak Harmonic or Spurious Noise Full Scale, 1 kHz Input Full Scale, 12 kHz Input Total Harmonic Distortion Full Scale, 1 kHz Input (Note 1) J, A, S 96 K, B, T 100 J, A, S 85 K, B, T 85 J, A, S K, B, T Signal-to-Noise Ratio 1 kHz, 0 dB Input (Notes 1 and 9) J, A, S K, B, T 1 kHz, -60 dB Input J, A, S K, B, T Noise Unipolar Mode Bipolar Mode 100 104 88 91 96 100 85 85 0.002 0.001 87 90 90 92 30 32 100 104 88 91 92 100 82 85 0.002 0.001 87 90 90 92 30 32 84 90 100 104 88 91 dB dB dB dB 0.002 0.001 % % 90 92 30 32 dB dB dB dB 35 70 µVrms µVrms (Note 3) 35 70 35 70 Notes: 10. Minimum resolution for which no missing codes is guaranteed * Refer to Parameter Definitions (immediately following the pin descriptions at the end of this data sheet). Specifications are subject to change without notice. 2-12 DS14F6 CS5016 CS5016 ANALOG CHARACTERISTICS (continued) CS5016-J, K Parameter* Min Specified Temperature Range Typ Max CS5016-A, B Min Typ Max CS5016-S, T Min Typ Max Units 0 to +70 -40 to +85 -55 to +125 °C Aperture Time 25 25 25 ns Aperture Jitter 100 100 100 ps Analog Input Input Capacitance Unipolar Mode Bipolar Mode (Note 4) 275 165 375 220 275 165 375 220 275 165 375 220 pF pF 16.25 32.5 µs µs 3.75 5.25 µs µs Conversion & Throughput Conversion Time -16 (Notes 5 and 6) -32 Acquisition Time -16 -32 (Note 6) Throughput -16 -32 (Note 6) 16.25 32.5 3.0 4.5 16.25 32.5 3.75 5.25 50 26.5 3.0 4.5 3.75 5.25 50 26.5 3.0 4.5 50 26.5 kHz kHz Power Supplies DC Power Supply Currents IA+ IAID+ ID- (Note 7) Power Dissipation (Note 7) Power Supply Rejection Positive Supplies Negative Supplies (Note 8) DS14F6 9 -9 3 -3 19 -19 6 -6 9 -9 3 -3 19 -19 6 -6 9 -9 3 -3 19 -19 6 -6 mA mA mA mA 120 250 120 250 120 250 mW 84 84 84 84 84 84 dB dB 2-13 CS5012A, CS5014, CS5016 SWITCHING CHARACTERISTICS (TA = TMIN to TMAX; VA+, VD+ = 5V ±10%; VA-, VD- = -5V ±10%; Inputs: Logic 0 = 0V, Logic 1 = VD+; CL = 50 pF, BW = VD+) Parameter Symbol CS5012A CLKIN Frequency: Internally Generated: Externally Supplied: CS5014/5016 CLKIN Frequency: Internally Generated: Externally Supplied: Min Typ Max Units 1.75 100 kHz 100 kHz - 6.4 4.0 MHz MHz MHz 1.75 1 100 kHz 100 kHz - 4 2 MHz MHz MHz MHz 40 - 60 % fCLK -7 -12 fCLK -14, -28, -14, -28, -16 -32 -16 -32 CLKIN Duty Cycle Rise Times: Any Digital Input Any Digital Output trise - 20 1.0 - µs ns Fall Times: Any Digital Input Any Digital Output tfall - 20 1.0 - µs ns thpw 1/fCLK+50 - tc ns tc 49/fCLK+50 57/fCLK 65/fCLK - 53/fCLK+235 61/fCLK+235 69/fCLK+235 ns ns ns tdd - 40 100 ns tepw 4/fCLK-20 - - ns HOLD Pulse Width Conversion Time: CS5012A CS5014 CS5016 Data Delay Time EOC Pulse Width (Note 11) Set Up Times: CAL, INTRLV to CS Low A0 to CS and RD Low tcs tas 20 20 10 10 - ns ns Hold Times: CS or RD High to A0 Invalid CS High to CAL, INTRLV Invalid tah tch 50 50 30 30 - ns ns Access Times: CS Low to Data Valid A, B, J, K S, T A, B, J, K S, T tca - 90 115 90 90 120 150 120 150 ns ns ns ns Output Float Delay: K, B CS or RD High to Output Hi-Z T tfd - 90 90 110 140 ns ns tpwl tpwh - 2/fCLK 2/fCLK - ns ns RD Low to Data Valid tra Serial Clock Pulse Width Low Pulse Width High Set Up Times: SDATA to SCLK Rising tss 2/fCLK-50 2/fCLK - ns Hold Times: SCLK Rising to SDATA tsh 2/fCLK-100 2/fCLK - ns Notes: 11. EOC remains low 4 CLKIN cycles if CS and RD are held low. Otherwise, it returns high within 4 CLKIN cycles from the start of a data read operation or a conversion cycle. 2-14 DS14F6 CS5012A, CS5014, CS5016 t rise t fall 90% 90% 10% 10% Rise and Fall Times t pwl t pwh SCLK t sh t ss SDATA Serial Output Timing t ca CS t ra RD t ah t as A0 t fd D0-D15 Hi-Z Hi-Z t cs t ch CAL, INTRLV Read and Calibration Control Timing t hpw HOLD t epw tc EOC t dd Output Data LAST CONVERSION DATA VALID NEW DATA VALID Conversion Timing DS14F6 2-15 CS5012A, CS5014, CS5016 DIGITAL CHARACTERISTICS (TA = TMIN to TMAX; VA+, VD+ = 5V ±10%; VA-, VD- = -5V ±10%) Parameter Symbol Min Typ Max Units High-Level Input Voltage VIH 2.0 - - V Low-Level Input Voltage VIL - - 0.8 V High-Level Output Voltage (Note 12) VOH (VD+) - 1.0V - - V Low-Level Output Voltage Iout = 1.6mA VOL - - 0.4 V Input Leakage Current Iin - - 10 µA 3-State Leakage Current IOZ - - ±10 µA Digital Output Pin Capacitance Cout - 9 - pF Notes: 12. Iout = -100 µA. This specification guarantees TTL compatibility (VOH = 2.4V @ Iout = -40 µA). RECOMMENDED OPERATING CONDITIONS (AGND, DGND = 0V, see Note 13) Parameter DC Power Supplies: Positive Digital Negative Digital Positive Analog Negative Analog Analog Reference Voltage Symbol Min Typ Max Units VD+ VDVA+ VA- 4.5 -4.5 4.5 -4.5 5.0 -5.0 5.0 -5.0 VA+ -5.5 5.5 -5.5 V V V V VREF 2.5 4.5 (VA+) - 0.5 V Analog Input Voltage: (Note 14) AGND VREF Unipolar VAIN VAIN -VREF VREF Bipolar Notes: 13. All voltages with respect to ground. 14. The CS5012A/14/16 can accept input voltages up to the analog supplies (VA+ and VA-). It will output all 1’s for inputs above VREF and all 0’s for inputs below AGND in unipolar mode and -VREF in bipolar mode. V V ABSOLUTE MAXIMUM RATINGS (AGND, DGND = 0V, all voltages with repect to ground.) WARNING: Operation at or beyond these limits may reult in permanent damage to the device. Normal operation is not guaranteed at these extremes. Parameter DC Power Supplies: Positive Digital Negative Digital Positive Analog Negative Analog Input Current, Any Pin Except Supplies Analog Input Voltage Symbol Min Max Units (Note 15) VD+ VDVA+ VA- -0.3 0.3 -0.3 0.3 6.0 -6.0 6.0 -6.0 V V V V (Note 16) Iin - ±10 mA VINA (VA-) - 0.3 (VA+) + 0.3 V VIND -0.3 (VA+) + 0.3 V TA -55 125 °C -65 150 °C (AIN and VREF pins) Digital Input Voltage Ambient Operating Temperature Storage Temperature Tstg Notes: 15. In addition, VD+ should not be greater than (VA+) + 0.3V. 16. Transient currents of up to 100 mA will not cause SCR latch-up. 2-16 DS14F6 CS5012A, CS5014, CS5016 AIN VREF AGND C C/2 CS5012A: Bit 11 CS5014: Bit 13 CS5016: Bit 15 MSB Bit 10 Bit 12 Bit 14 C/4 C/8 Bit 9 Bit 11 Bit 13 Bit 8 Bit 10 Bit 12 C tot S1 C/X CS5012A X = 2048 C/X CS5014 X = 8192 CS5016 X = 32768 Bit 0 LSB Dummy = C + C/2 + C/4 + ... + C/X Figure 1. Charge Redistribution DAC THEORY OF OPERATION The CS5012A/14/16 family utilize a successive approximation conversion technique. The analog input is successively compared to the output of a D/A converter controlled by the conversion algorithm. Successive approximation begins by comparing the analog input to the DAC output which is set to half-scale (MSB on, all other bits off). If the input is found to be below half-scale, the MSB is reset to zero and the input is compared to one-quarter scale (next MSB on, all others off). If the input were above half-scale, the MSB would remain high and the next comparison would be at three-quarters of full scale. This procedure continues until all bits have been exercised. A unique charge redistribution architecture is used to implement the successive approximation algorithm. Instead of the traditional resistor network, the DAC is an array of binary-weighted capacitors. All capacitors in the array share a common node at the comparator’s input. Their other terminals are capable of being connected to AIN, AGND, or VREF (Figure 1). When the device is not calibrating or converting, all capacitors are tied to AIN forming Ctot. Switch S1 is closed and the charge on the array, Qin, tracks the input signal Vin (Figure 2a). When the conversion command is issued, switch S1 opens as shown in Figure 2b. This traps charge Qin on the comparator side of the capacitor array and creates a floating node at the comparator’s input. The conversion algorithm operates on this fixed charge, and the signal at the analog input pin is ignored. In effect, the entire DAC capacitor array serves as analog memory D . C tot S1 AIN Qin Qin + Vin - C tot S1 VREF To MCU + Vfn - AGND To MCU (1-D) C tot -Q in = Vin C tot Figure 2a. Tracking Mode DS14F6 V in D = VREF for V fn = 0V Figure 2b. Convert Mode 2-17 CS5012A, CS5014, CS5016 during conversion much like a hold capacitor in a sample/hold amplifier. Auto-zeroing enhances power supply rejection at frequencies well below the conversion rate. The conversion consists of manipulating the free plates of the capacitor array to VREF and AGND to form a capacitive divider. Since the charge at the floating node remains fixed, the voltage at that point depends on the proportion of capacitance tied to VREF versus AGND. The successive-approximation algorithm is used to find the proportion of capacitance, termed D in Figure 2b, which when connected to the reference will drive the voltage at the floating node (Vfn) to zero. That binary fraction of capacitance represents the converter’s digital output. To achieve complete accuracy from the DAC, the CS5012A/14/16 use a novel self-calibration scheme. Each bit capacitor, shown in Figure 1, actually consists of several capacitors which can be manipulated to adjust the overall bit weight. An on-chip microcontroller adjusts the subarrays to precisely ratio the bits. Each bit is adjusted to just balance the sum of all less significant bits plus one dummy LSB (for example, 16C = 8C + 4C + 2C + C + C). Calibration resolution for the array is a small fraction of an LSB resulting in nearly ideal differential and integral linearity. This charge redistribution architecture easily supports bipolar input ranges. If half the capacitor array (the MSB capacitor) is tied to VREF rather than AIN in the track mode, the input range is doubled and is offset half-scale. The magnitude of the reference voltage thus defines both positive and negative full-scale (-VREF to +VREF), and the digital code is an offset binary representation of the input. DIGITAL CIRCUIT CONNECTIONS Calibration Master Clock The ability of the CS5012A/14/16 to convert accurately clearly depends on the accuracy of their comparator and DAC. The CS5012A/14/16 utilize an "auto-zeroing" scheme to null errors introduced by the comparator. All offsets are stored on the capacitor array while in the track mode and are effectively subtracted from the input signal when a conversion is initiated. Sampling Clock The CS5012A/14/16 can be applied in a wide variety of master clock, sampling, and calibration conditions which directly affect the devices’ conversion time and throughput. The devices also feature on-chip 3-state output buffers and a complete interface for connecting to 8-bit and 16-bit digital systems. Output data is also available in serial format. The CS5012A/14/16 operate from a master clock (CLKIN) which can be externally supplied or internally generated. The internal oscillator is activated by externally tying the CLKIN input low. Alternatively, the CS5012A/14/16 can be synchronized to the external system by driving the CLKIN pin with a TTL or CMOS clock signal. HOLD HOLD EOT CS5012A/14/16 Master Clock (Optional) CLKIN Figure 3a. Asynchronous Sampling 2-18 CS5012A/14/16 Master Clock (Optional) CLKIN Figure 3b. Synchronous Sampling DS14F6 CS5012A, CS5014, CS5016 All calibration, conversion, and throughput times directly scale to CLKIN frequency. Thus, throughput can be precisely controlled and/or maximized using an external CLKIN signal. In contrast, the CS5012A/14/16’s internal oscillator will vary from unit-to-unit and over temperature. The CS5012A/14/16 can typically convert with CLKIN as low as 10 kHz at room temperature. Initiating Conversions A falling transition on the HOLD pin places the input in the hold mode and initiates a conversion cycle. Upon completion of the conversion cycle, the CS5012A/14/16 automatically return to the track mode. In contrast to systems with separate track-and-holds and A/D converters, a sampling clock can simply be connected to the HOLD input (Figure 3a). The duty cycle of this clock is not critical. It need only remain low at least one CLKIN cycle plus 50 ns, but no longer than the minimum conversion time or an additional conversion cycle will be initiated with inadequate time for acquisition. Microprocessor-Controlled Operation Sampling and conversion can be placed under microprocessor control (Figure 4) by simply gating the devices’ decoded address with the write strobe for the HOLD input. Thus, a write cycle to the CS5012A/14/16’s base address will initiate a conversion. However, the write cycle must be to RD the odd address (A0 high) to avoid initiating a software controlled reset (see Reset below). The calibration control inputs, CAL, and INTRLV are inputs to a set of transparent latches. These signals are internally latched by CS returning high. They must be in the appropriate state whenever the chip is selected during a read or write cycle. Address lines A1 and A2 are shown connected to CAL and INTRLV in Figure 4 placing calibration under microprocessor control as well. Thus, any read or write cycle to the CS5012A/14/16’s base address will initiate or terminate calibration. Alternatively, A0, INTRLV, and CAL may be connected to the microprocessor data bus. Conversion Time/Throughput Upon completing a conversion cycle and returning to the track mode, the CS5012A/14/16 require time to acquire the analog input signal before another conversion can be initiated. The acquisition time is specified as six CLKIN cycles plus 2.25 µs (1.32 µs for the CS5012A -7 version only). This adds to the conversion time to define the converter’s maximum throughput. The conversion time of the CS5012A/14/16, in turn, depends on the sampling, calibration, and CLKIN conditions. RD CONCLK HOLD ADDR VALID RD RD WR HOLD ADDR VALID AN Address Bus Addr Dec A3 AN CS CS5012A/14/16 A2 CAL A1 INTRLV A0 A0 A3 CS CS5012A/14/16 A2 Figure 4a. Conversions Asynchronous to CLKIN DS14F6 Addr Dec Address Bus CAL A1 INTRLV A0 A0 Figure 4b. Conversions under Microprocessor Control 2-19 CS5012A, CS5014, CS5016 1 / Throughput HOLD Input 1 / Throughput (64 + N cycles) HOLD Input Conversion EOC Output Conversion (49 + N cycles) EOC Output Acquisition * Acquisition (15 cycles) EOT Output EOT Output * Dashed line: CS & RD = 0 Solid line: See Figure 9 Synchronization Uncertainty (4 cycles) CS5012A N = 0 CS5014 N = 8 CS5016 N = 16 Figure 5b. Synchronous (Loopback Mode) Figure 5a. Asynchronous Sampling (External Clock) Asynchronous Sampling Synchronous Sampling The CS5012A/14/16 internally operate from a clock which is delayed and divided down from CLKIN (fCLK/4). If sampling is not synchronized to this internal clock, the conversion cycle may not begin until up to four clock cycles after HOLD goes low even though the charge is trapped immediately. In this asynchronous mode (Figure 3a), the four clock cycles add to the minimum 49, 57 and 65 clock cycles (for the CS5012A/14/16 respectively) to define the maximum conversion time (see Figure 5a and Table 1). To achieve maximum throughput, sampling can be synchronized with the internal conversion clock by connecting the End-of-Track (EOT) output to HOLD (Figure 3b). The EOT output falls 15 CLKIN cycles after EOC indicating the analog input has been acquired to the CS5012A/14/16’s specified accuracy. The EOT output is synchronized to the internal conversion clock, so the four clock cycle synchronization uncertainty is removed yielding throughput at [1/64]fCLK for the CS5012A, [1/72]fCLK for CS5014 and [1/80]fCLK for CS5016 where fCLK is the CLKIN frequency (see Figure 5b and Table 1). Conversion Time Min Sampling Mode Throughput Time Max Min Max CS5012A Synchronous (Loopback) 49 t clk -7 49 t clk 49 t clk Asynchronous -12,-24 49 t clk 53 t clk + 53 t clk + 64 t clk 235 ns N/A 235 ns N/A 64 t clk 59 t clk + 59 t clk + 1.32 µs 2.25 µs CS5014 Synchronous (Loopback) 57 t clk Asynchronous 57 t clk 57 t clk 61 t clk + 235 ns 72 t clk 72 t clk N/A 67 t clk + 2.25 µs CS5016 Synchronous (Loopback) 65 t clk Asynchronous 65 t clk 65 t clk 69 t clk + 235 ns 80 t clk 80 t clk N/A 75 t clk + 2.25 µs Table 1. Conversion and Throughput Times (tclk = Master Clock Period) 2-20 DS14F6 CS5012A, CS5014, CS5016 Also, the CS5012A/14/16’s internal RC oscillator exhibits jitter (typically ± 0.05% of its period), which is high compared to crystal oscillators. If the CS5012A/14/16 is configured for synchronous sampling while operating from its internal oscillator, this jitter will directly affect sampling purity. The user can obtain best sampling purity while synchronously sampling by using an external crystal-based clock. Reset Upon power up, the CS5012A/14/16 must be reset to guarantee a consistent starting condition and initially calibrate the devices. Due to the CS5012A/14/16’s low power dissipation and low temperature drift, no warm-up time is required before reset to accommodate any self-heating effects. However, the voltage reference input should have stabilized to within 5%, 1% or 0.25% of its final value, for the CS5012A/14/16 respectively, before RST falls to guarantee an accurate calibration. Later, the CS5012A/14/16 may be reset at any time to initiate a single full calibration. Reset overrides all other functions. If reset, the CS5012A/14/16 will clear and initiate a new calibration cycle mid-conversion or mid-calibration. +5V R RST C CS5012A/14/16 Figure 6. Power-on Reset Circuit eliminate the possibility of inadvertent software reset. The EOC output remains high throughout the calibration operation and will fall upon its completion. It can thus be used to generate an interrupt indicating the CS5012A/14/16 is ready for operation. While calibrating, the HOLD input is ignored until EOC falls. After EOC falls, six CLKIN cycles plus 2.25 µs (1.32 µs for the CS5012A -7 version only) must be allowed for signal acquisition before HOLD is activated. Under microprocessor-independent operation (CS, RD low; A0 high) the CS5014’s and CS5016’s EOC output will not fall at the completion of the calibration cycle, but EOT will fall 15 CLKIN cycles later. Initiating Calibration Resets can be initiated in hardware or software. The simplest method of resetting the CS5012A/14/16 involves strobing the RST pin high for at least 100 ns. When RST is brought high all internal logic clears. When it returns low, a full calibration begins which takes 58,280 CLKIN cycles for the CS5012A (approximately 9.1 ms with a 6.4 MHz clock) and 1,441,020 CLKIN cycles for the CS5016, CS5014 and CS5012 (approximately 360 ms with a 4 MHz CLKIN). A simple power-on reset circuit can be built using a resistor and capacitor, and a Schmitt-trigger inverter to prevent oscillation (see Figure 6). The CS5012A/14/16 can also be reset in software when under microprocessor control. The CS5012A/14/16 will reset whenever CS, A0, and HOLD are taken low simultaneously. See the Microprocessor Interface section (below) to DS14F6 All modes of calibration can be controlled in hardware or software. Accuracy can thereby be insured at any time or temperature throughout operating life. After initial calibration at power-up, the CS5012A/14/16’s charge-redistribution design yields better temperature drift and more graceful aging than resistor-based technologies, so calibration is normally only required once, after power-up. The first mode of calibration, reset, results in a single full calibration cycle. The second type of calibration, "burst" cal, allows control of partial calibration cycles. Due to an unforeseen condidtion inside the part, asynchronous termination of calibration may result in a sub-optimal result. Burst cal should not be used. 2-21 CS5012A, CS5014, CS5016 The reset calibration always works perfectly, and should be used instead of burst mode. The CS5012’s and CS5012A/14/16’s very low drift over temperature means that, under most circumstances, calibration will only need to be performed at power-up, using reset. The CS5012A/14/16 feature a background calibration mode called "interleave." Interleave appends a single calibration experiment to each conversion cycle and thus requires no dead time for calibration. The CS5012A/14/16 gathers data between conversions and will adjust its transfer function once it completes the entire sequence of experiments (one calibration cycle per 2,014 conversions in the CS5012A and one calibration per 72,051 conversions in the CS5012, CS5014 and CS5016). Initiated by bringing both the INTRLV input and CS low (or hard-wiring INTRLV low), interleave extends the CS5012A/14/16’s effective conversion time by 20 CLKIN cycles. Other than reduced throughput, interleave is totally transparent to the user. Interleave calibration should not be used intermittently. The fact that the CS5012A/14/16 offer several calibration modes is not to imply that the devices need to be recalibrated often. The devices are very stable in the presence of large temperature changes. Tests have indicated that after using a single reset calibration at 25 °C most devices exhibit very little change in offset or gain when exposed to temperatures from -55 to +125 °C. The data indicated 30 ppm as the typical worst case total change in offset or gain over this temperature range. Differential linearity remained virtually unchanged. System error sources outside of the A/D converter, whether due to changes in 2-22 temperature or to long-term aging, will generally dominate total system error. Microprocessor Interface The CS5012A/14/16 feature an intelligent microprocessor interface which offers detailed status information and allows software control of the self-calibration functions. Output data is available in either 8-bit or 16-bit formats for easy interfacing to industry-standard microprocessors. Strobing both CS and RD low enables the CS5012A/14/16’s 3-state output buffers with either output data or status information depending on the status of A0. An address bit can be connected to A0 as shown in Figure 4b thereby memory mapping the status register and output data. Conversion status can be polled in software by reading the status register (CS and RD strobed low with A0 low), and masking status bits S0-S5 and S7 (by logically AND’ing the status word with 01000000) to determine the value of S6. Similarly, the software routine can determine calibration status using other status bits (see Table 2). Care must be taken not to read the status register (A0 low) while HOLD is low, or a software reset will result (see Reset above). Alternatively, the End-of-Convert (EOC) output can be used to generate an interrupt or drive a DMA controller to dump the output directly into memory after each conversion. The EOC pin falls as each conversion cycle is completed and data is valid at the output. It returns high within four CLKIN cycles of the first subsequent data read operation or after the start of a new conversion cycle. DS14F6 CS5012A, CS5014, CS5016 PIN D0 STATUS BIT S0 STATUS END OF CONVERSION D1 D2 S1 S2 RESERVED LOW BYTE/HIGH BYTE D3 S3 END OF TRACK D4 D5 D6 D7 S4 S5 S6 S7 RESERVED TRACKING CONVERTING CALIBRATING DEFINITION Falls upon completion of a conversion, and returns high on the first subsequent read. Reserved for factory use. When data is to be read in an 8-bit format (BW=0), indicates which byte will appear at the output next. When low, indicates the input has been acquired to the devices specified accuracy. Reserved for factory use. High when the device is tracking the input. High when the device is converting the held input. High when the device is calibrating. Table 2. Status Pin Definitions To interface with a 16-bit data bus, the BW input to the CS5012A/14/16 should be held high and all data bits (12, 14 and 16 for the CS5012A, CS5014 and CS5016 respectively) read in parallel on pins D4-D15 (CS5012A), D2-D15 (CS5014), or D0-D15 (CS5016). With an 8-bit bus, the converter’s result must be read in two portions. In this instance, BW should be held low and the 8 MSB’s obtained on the first read cycle following a conversion. The second read cycle will yield the remaining LSB’s (4, 6 or 8 for the CS5012A/14/16 respectively) with 4, 2 or 0 trailing zeros. Both bytes appear on pins D0-D7. The upper/lower bytes of the same data will continue to toggle on subsequent reads until the next conD15 D14 D13 D12 D11 D10 Status (A0=0) CS5012A Data (A0=1) X X version finishes. Status bit S2 indicates which byte will appear on the next data read operation. The CS5012A/14/16 internally buffer their output data, so data can be read while the devices are tracking or converting the next sample. Therefore, retrieving the converters’ digital output requires no reduction in ADC throughput. Enabling the 3state outputs while the CS5012A/14/16 is converting will not introduce conversion errors. Connecting CMOS logic to the digital outputs is recommended. Suitable logic families include 4000B, 74HC, 74AC, 74ACT, and 74HCT. D9 D8 D7 D6 D5 D4 D3 D2 D1 D0 X X X X X X S7 S6 S5 S4 S3 S2 S1 S0 B8 B7 B6 B5 B4 B3 B2 B1 B0 0 0 0 0 B11 B10 B9 CS5014 B13 B12 B11 B10 B9 B8 B7 B6 B5 B4 B3 B2 B1 B0 0 0 CS5016 B15 B14 B13 B12 B11 B10 B9 B8 B7 B6 B5 B4 B3 B2 B1 B0 CS5012A X X X X X X X X B11 B3 B10 B2 B9 B1 CS5014 X X X X X X X X B13 B5 B12 B4 CS5016 X X X X X X X X B15 B7 B14 B6 "X" Denotes High Impedance Output B8 B0 B7 0 B6 0 B5 0 B4 0 B11 B10 B3 B2 B9 B1 B8 B0 B7 0 B6 0 B13 B12 B5 B4 B11 B10 B3 B2 B9 B1 B8 B0 8- or 16-Bit Data Bus 16-Bit Bus (BW=1) 8-Bit Bus (BW=0) Figure 7. CS5012A/14/16 Data Format DS14F6 2-23 CS5012A, CS5014, CS5016 Microprocessor Independent Operation The CS5012A/14/16 can be operated in a standalone mode independent of intelligent control. In this mode, CS and RD are hard-wired low. This permanently enables the 3-state output buffers and allows transparent latch inputs (CAL and INTRLV) to be active. A free-running condition is established when BW is tied high, CAL is tied low, and HOLD is continually strobed low or tied to EOT. The CS5012A/14/16’s EOC output can be used to externally latch the output data if desired. With CS and RD hard-wired low, EOC will strobe low for four CLKIN cycles after each conversion. Data will be unstable up to 100 ns after EOC falls, so it should be latched on the rising edge of EOC. +5V INTRLV BW A0 Sampling Clock RST Reset EOC Latching Output D15 HOLD CS RD CAL CS5012A CS5014 CS5016 12-Bit Data Out D4 Figure 8. Microprocessor-Independent Connections Serial Output All successive-approximation A/D converters derive their digital output serially starting with the MSB. The CS5012A/14/16 present each bit to the SDATA pin four CLKIN cycles after it is derived and can be latched using the serial clock output, SCLK. Just subsequent to each bit decision SCLK will fall and return high once the bit information on SDATA has stabilized. Thus, the rising edge of the SCLK output should be used to clock the data from the CS5012A/14/16 (See Figure 9). ANALOG CIRCUIT CONNECTIONS Most popular successive-approximation A/D converters generate dynamic loads at their analog 2-24 connections. The CS5012A/14/16 internally buffer all analog inputs (AIN, VREF, and AGND) to ease the demands placed on external circuitry. However, accurate system operation still requires careful attention to details at the design stage regarding source impedances as well as grounding and decoupling schemes. Reference Considerations An application note titled "Voltage References for the CS501X Series of A/D Converters" is available for the CS5012A/14/16. In addition to working through a reference circuit design example, it offers several built-and-tested reference circuits. During conversion, each capacitor of the calibrated capacitor array is switched between VREF and AGND in a manner determined by the successive-approximation algorithm. The charging and discharging of the array results in a current load at the reference. The CS5012A/14/16 include an internal buffer amplifier to minimize the external reference circuit’s drive requirement and preserve the reference’s integrity. Whenever the array is switched during conversion, the buffer is used to pre-charge the array thereby providing the bulk of the necessary charge. The appropriate array capacitors are then switched to the unbuffered VREF pin to avoid any errors due to offsets and/or noise in the buffer. The external reference circuitry need only provide the residual charge required to fully charge the array after pre-charging from the buffer. This creates an ac current load as the CS5012A/14/16 sequence through conversions. The reference circuitry must have a low enough output impedance to drive the requisite current without changing its output voltage significantly. As the analog input signal varies, the switching sequence of the internal capacitor array changes. The current load on the external reference circuitry thus varies in response with the analog input. Therefore, the external reference must not exhibit significant DS14F6 CS5012A, CS5014, CS5016 peaking in its output impedance characteristic at signal frequencies or their harmonics. A large capacitor connected between VREF and AGND can provide sufficiently low output impedance at the high end of the frequency spectrum, while almost all precision references exhibit extremely low output impedance at dc. The magnitude of the current load on the external reference circuitry will scale to the CLKIN frequency. At full speed, the reference must supply a maximum load current of 10 µA peak-to-peak (1 µA typical). For the CS5012A an output impedance of 15 Ω will therefore yield a maximum error of 150 mV. With a 2.5V reference and LSB size of 600 mV, this would insure better than 1/4 LSB accuracy. A 1 µF capacitor exhibits an imCS5016: 60 CS5014: 52 CS5012A: 44 CLKIN 62 54 46 64 56 48 66 58 50 68 60 52 70 62 54 72 64 56 74 66 58 pedance of less than 15 Ω at frequencies greater than 10 kHz. Similarly, for the CS5014 with a 4.5V reference (275µV/LSB), better than 1/4 LSB accuracy can be insured with an output impedance of 4Ω or less (maximum error of 40 µV). A 2.2 µF capacitor exhibits an impedance of less than 4Ω at frequencies greater than 5kHz. For the CS5016 with a 4.5V reference (69µV/LSB), better than 1/4 LSB accuracy can be insured with an output impedance of less than 2Ω (maximum error of 20 µV). A 20 µF capacitor exhibits an impedance of less than 2Ω at frequencies greater than 16 kHz. A high-quality tantalum capacitor in parallel with a smaller ceramic capacitor is recommended. 76 68 60 78 70 62 80/0 72/0 64/0 2 2 2 4 4 4 6 6 6 8 8 8 10 10 10 12 12 12 EOC Status LSB Determined Coarse Charge EOT MSB - 1 MSB - 2 MSB Determined Determined Determined Fine Charge t HOLD t d d SCLK SDATA LSB+2 LSB+1 LSB MSB MSB - 1 Notes: 1. Synchronous (loopback) mode is illustrated. After EOC falls the converter goes into coarse charge mode for 6 CLKIN cycles, then to fine charge mode for 9 cycles, then EOT falls. In loopback mode, EOT trips HOLD which captures the analog sample. Conversion begins on the next rising edge of CLKIN. If operated asynchronously, EOT will remain low until after HOLD is taken low. When HOLD occurs the analog sample is captured immediately, but conversion may not begin until four CLKIN cycles later. EOT will return high when conversion begins. 2. Timing delay td (relative to CLKIN) can vary between 135 ns to 235 ns over the military temperature range and over ± 10% supply variation 3. EOC returns high in 4 CLKIN cycles if A0 = 1 and CS = RD = 0 (Microprocessor Independent Mode); within 4 CLKIN cycles after a data read (Microprocessor Mode); or 4 CLKIN cycles after HOLD = 0 is recognized on a rising edge of CLKIN/4. Figure 9. Serial Output Timing DS14F6 2-25 CS5012A, CS5014, CS5016 Peaking in the reference’s output impedance can occur because of capacitive loading at its output. Any peaking that might occur can be reduced by placing a small resistor in series with the capacitors (Figure 10). The equation in Figure 10 can be used to help calculate the optimum value of R for a particular reference. The term "fpeak" is the frequency of the peak in the output impedance of the reference before the resistor is added. +V ee Vref C1 1.0 µF 0.1 µF Internal Charge Error (LSB's) CS5014 +200 +50 +12.5 0 0 0 VREF 29 REFBUF 30 VA- R CS5012A CS5014 CS5016 -5V The CS5012A/14/16 can operate with a wide range of reference voltages, but signal-to-noise performance is maximized by using as wide a signal range as possible. The recommended reference voltage is between 2.5 and 4.5 V for the CS5012A and 4.5 V for the CS5014/16. The CS5012A/14/16 can actually accept reference voltages up to the positive analog supply. However, the buffer’s offset may increase as the reference voltage approaches VA+ thereby increasing external drive requirements at VREF. A 4.5V reference is the maximum reference voltage recommended. This allows 0.5V headroom for the internal reference buffer. Also, the buffer enlists the aid of an external 0.1 µF ceramic capacitor which must be tied between its output, REFBUF, and the negative analog supply, VA-. For more information on references, consult the applicaCS5016 C2 0.01 µF 28 1 2π (C1 + C2) fpeak Figure 10. Reference Connections R = tion note: Voltage References for the CS501X Series of A/D Converters. For an example of using the CS5012A/14/16 with a 5 volt reference, see the application note: A Collection of Application Hints for the CS501X Series of A/D Converters. Analog Input Connection The analog input terminal functions similarly to the VREF input after each conversion when switching into the track mode. During the first six CLKIN cycles in the track mode, the buffered version of the analog input is used for pre-charging the capacitor array. An additional period is required for fine-charging directly from AIN to CS5012A Pre-Charge -200 -50 -12.5 -400 -100 -25.0 0.5 1.0 Fine-Charge 1.5 2.0 2.5 Acquisition Time (us) (Delay from EOC) Figure 11. Internal Acquisition Time 2-26 DS14F6 CS5012A, CS5014, CS5016 obtain the specified accuracy. Figure 11 illustrates this operation. During pre-charge the charge on the capacitor array first settles to the buffered version of the analog input. This voltage is offset from the actual input voltage. During fine-charge, the charge then settles to the accurate unbuffered version. The acquisition time of the CS5012A/14/16 depends on the CLKIN frequency. This is due to a fixed pre-charge period. For instance, operating the CS5012A -12, CS5014 -14 or CS5016 -16 version with an external 4 MHz CLKIN results in a 3.75 µs acquisition time: 1.5 µs for pre-charging (6 clock cycles) and 2.25 µs for fine-charging. Fine-charge settling is specified as a maximum of 2.25 µs for an analog source impedance of less than 200 Ω. (For the CS5012A -7 version it is specified as 1.32 µs.) In addition, the comparator requires a source impedance of less than 400 Ω around 2 MHz for stability, which is met by practically all bipolar op amps. Large dc source impedances can be accommodated by adding capacitance from AIN to ground (typically 200 pF) to decrease source impedance at high frequencies. However, high dc source resistances will increase the input’s RC time constant and extend the nec- CS5012A/14/16 HOLD Input Convert Channel N essary acquisition time. For more information on input applications, consult the application note: Input Buffer Amplifiers for the CS501X Family of A/D Converters. During the first six clock cycles following a conversion (pre-charge) in unipolar mode, the CS5012A is capable of slewing at 20V/µs and the CS5014/16 can slew at 5V/µs. In bipolar mode, only half the capacitor array is connected to the analog input so the CS5012A can slew at 40V/µs, and the CS5014/16 can slew at 10V/µs. After the first six CLKIN cycles, the CS5012A will slew at 1.25V/µs in unipolar mode and 3.0V/µs in bipolar mode, and the CS5014/16 will slew at 0.25V/µs in unipolar mode and 0.5V/µs in bipolar mode. Acquisition of fast slewing signals (step functions) can be hastened if the step occurs during or immediately following the conversion cycle. For instance, channel selection in multiplexed applications should occur while the CS5012A/14/16 is converting (see Figure 12). Multiplexer settling is thereby removed from the overall throughput equation, and the CS5012A/14/16 can convert at full speed. Convert Channel N+1 CS5012A/14/16 EOC Output MUX Address Address N CS5012A/14/16 Analog Input Address N + 1 MUX Settling to Channel N + 1 Address N + 2 Address N + 3 MUX Settling to Channel N + 2 Figure 12. Pipelined MUX Input Channels DS14F6 2-27 CS5012A, CS5014, CS5016 Analog Input Range/Coding Format The reference voltage directly defines the input voltage range in both the unipolar and bipolar configurations. In the unipolar configuration (BP/UP low), the first code transition occurs 0.5 LSB above AGND, and the final code transition occurs 1.5 LSB’s below VREF. Coding is in straight binary format. In the bipolar configuration (BP/UP high), the first code transition occurs 0.5 LSB above -VREF and the last transition occurs 1.5 LSB’s below +VREF. Coding is in an offset-binary format. Positive full scale gives a digital output of all ones, and negative full scale gives a digital output of all zeros. The BP/UP mode pin may be switched after calibration without having to recalibrate the converter. However, the BP/UP mode should be changed during the previous conversion cycle, that is, between HOLD falling and EOC falling. If BP/UP is changed at any other time, one dummy conversion cycle must be allowed for proper acquisition of the input. Grounding and Power Supply Decoupling The CS5012A/14/16 use the analog ground connection, AGND, only as a reference voltage. No dc power currents flow through the AGND connection, and it is completely independent of DGND. However, any noise riding on the AGND input relative to the system’s analog ground will induce conversion errors. Therefore, both the analog input and reference voltage should be referred to the AGND pin, which should be used as the entire system’s analog ground reference point. 2-28 The digital and analog supplies to the CS5012A/14/16 are pinned out separately to minimize coupling between the analog and digital sections of the chip. All four supplies should be decoupled to their respective grounds using 0.1 µF ceramic capacitors. If significant low-frequency noise is present on the supplies, 1 µF tantalum capacitors are recommended in parallel with the 0.1 µF capacitors. The positive digital power supply of the CS5012A/14/16 must never exceed the positive analog supply by more than a diode drop or the device could experience permanent damage. If the two supplies are derived from separate sources, care must be taken that the analog supply comes up first at power-up. The system connection diagram in Figure 36 shows a decoupling scheme which allows the CS5012A/14/16 to be powered from a single set of ± 5V rails. As with any high-precision A/D converter, the CS5012A/14/16 require careful attention to grounding and layout arrangements. However, no unique layout issues must be addressed to properly apply the device. The CDB5012/14/16 evaluation board is available for the CS5012A/14/16, which avoids the need to design, build, and debug a high-precision PC board to initially characterize the part. The board comes with a socketed CS5012A/14/16, and can be quickly reconfigured to simulate any combination of sampling, calibration, CLKIN, and analog input range conditions. DS14F6 CS5012A, CS5014, CS5016 Power Supply Rejection CS5012A/14/16 PERFORMANCE The CS5012A/14/16’s power supply rejection performance is enhanced by the on-chip self-calibration and an "auto-zero" process. Drifts in power supply voltages at frequencies less than the calibration rate have negligible effect on the CS5012A/14/16’s accuracy. This is because the CS5012A/14/16 adjust their offset to within a small fraction of an LSB during calibration. Above the calibration frequency the excellent power supply rejection of the internal amplifiers is augmented by an auto-zero process. Any offsets are stored on the capacitor array and are effectively subtracted once conversion is initiated. Figure 13 shows power supply rejection of the CS5012A/14/16 in the bipolar mode with the analog input grounded and a 300 mVp-p ripple applied to each supply. Power supply rejection improves by 6 dB in the unipolar mode. The plot in Figure 13 shows worst-case rejection for all combinations of conversion rates and input conditions in the bipolar mode. 90 Power Supply Rejection (dB) 80 Differential Nonlinearity One source of nonlinearity in A/D converters is bit weight errors. These errors arise from the deviation of bits from their ideal binary-weighted ratios, and lead to nonideal widths for each code. If DNL errors are large, and code widths shrink to zero, it is possible for one or more codes to be entirely missing. The CS5012A/14/16 calibrate all bits in the capacitor array to a small fraction of an LSB resulting in nearly ideal DNL. Histogram plots of typical DNL of the CS5012A/14/16 can be seen in Figures 14, 16, 17. Figure 15 illustrates the DNL of the CS5012 for comparison with the CS5012A (Figure 14). A histogram test is a statistical method of deriving an A/D converter’s differential nonlinearity. A ramp is input to the A/D and a large number of samples are taken to insure a high confidence level in the test’s result. The number of occurrences for each code is monitored and stored. A perfect A/D converter would have all codes of equal size and therefore equal numbers of occurrences. In the histogram test a code with the average number of occurrences will be considered ideal (DNL = 0). A code with more or less occurrences than average will appear as a DNL of greater or less than zero LSB. A missing code has zero occurrences, and will appear as a DNL of -1 LSB. 70 Integral Nonlinearity 60 Integral Nonlinearity (INL; also termed Relative Accuracy or just Nonlinearity) is defined as the deviation of the transfer function from an ideal straight line. Bows in the transfer curve generate harmonic distortion. The worst-case condition of bit-weight errors (DNL) has traditionally also defined the point of maximum INL. 50 40 30 20 1 kHz 10 kHz 100 kHz 1 MHz Power Supply Ripple Frequency Figure 13. Power Supply Rejection DS14F6 Bit-weight errors have a drastic effect on a converter’s ac performance. They can be analyzed as step functions superimposed on the input signal. 2-29 CS5012A, CS5014, CS5016 +1 DNL (LSB) +1/2 0 -1/2 -1 0 2,048 4,095 Codes Figure 14. CS5012A Differential Nonlinearity Plot +1 DNL (LSB) +1/2 0 -1/2 -1 0 2,048 4,095 Codes Figure 15. CS5012 Differential Nonlinearity Plot +1 DNL (LSB) +1/2 0 -1/2 -1 0 8,192 16,383 Codes Figure 16. CS5014 Differential Nonlinearity Plot +1 DNL (LSB) +1/2 0 -1/2 -1 0 32,768 65,535 Codes Figure 17. CS5016 Differential Nonlinearity Plot 2-30 DS14F6 CS5012A, CS5014, CS5016 Since bits (and their errors) switch in and out throughout the transfer curve, their effect is signal dependent. That is, harmonic and intermodulation distortion, as well as noise, can vary with different input conditions. Designing a system around characterization data is risky since transfer curves can differ drastically unit-to-unit and lot-to-lot. The CS5012A/14/16 achieves repeatable signalto-noise and harmonic distortion performance using an on-chip self-calibration scheme. The CS5012A calibrates its bit weight errors to a small fraction of an LSB at 12-bits yielding peak distortion below the noise floor (see Figure 19). The CS5014 calibrates its bit weights to within ±1/16 LSB at 14-bits (±0.0004% FS) yielding peak distortion as low as -105 dB (see Figure 22). The CS5016 calibrates its bit weights to within ±1/4 LSB at 16-bits (±0.0004% FS) yielding peak distortion as low as -105 dB (see Figure 24). Unlike traditional ADC’s, the linearity of the CS5012A/14/16 are not limited by bit-weight errors; their performance is therefore extremely repeatable and independent of input signal conditions. Quantization Noise The error due to quantization of the analog input ultimately dictates the accuracy of any A/D converter. The continuous analog input must be represented by one of a finite number of digital codes, so the best accuracy to which an analog input can be known from its digital code is ±1/2 LSB. Under circumstances commonly encountered in signal processing applications, this quantization error can be treated as a random variable. The magnitude of the error is limited to ±1/2 LSB, but any value within this range has equal probability of occurrence. Such a probability distribution leads to an error "signal" with an rms value of 1 LSB/√12. Using an rms signal value of FS/√8 (amplitude = FS/2), this relates to ideal 12, 14 and 16-bit signal-to-noise ratios of 74, 86 and 98 dB respectively. DS14F6 Equally important is the spectral content of this error signal. It can be shown to be approximately white, with its energy spread uniformly over the band from dc to one-half the sampling rate. Advantage of this characteristic can be made by judicious use of filtering. If the signal is bandlimited, much of the quantization error can be filtered out, and improved system performance can be attained. FFT Tests and Windowing In the factory, the CS5012A/14/16 are tested using Fast Fourier Transform (FFT) techniques to analyze the converter’s dynamic performance. A pure sinewave is applied to the CS5012A/14/16, and a "time record" of 1024 samples is captured and processed. The FFT algorithm analyzes the spectral content of the digital waveform and distributes its energy among 512 "frequency bins." Assuming an ideal sinewave, distribution of energy in bins outside of the fundamental and dc can only be due to quantization effects and errors in the CS5012A/14/16. If sampling is not synchronized to the input sinewave, it is highly unlikely that the time record will contain an integer number of periods of the input signal. However, the FFT assumes that the signal is periodic, and will calculate the spectrum of a signal that appears to have large discontinuities, thereby yielding a severely distorted spectrum. To avoid this problem, the time record is multiplied by a window function prior to performing the FFT. The window function smoothly forces the endpoints of the time record to zero, thereby removing the discontinuities. The effect of the window in the frequency-domain is to convolute the spectrum of the window with that of the actual input. Figure 18 shows an FFT computed from an ideal 12-bit sinewave. The quality of the window used for harmonic analysis is typically judged by its highest side-lobe level. The Blackman-Harris window used for testing the CS5014 and CS5016 has a maximum side-lobe level of -92 dB. Fig2-31 CS5012A, CS5014, CS5016 0.0 0.0 Sampling Rate: 100kHz Full Scale: 9Vp-p S/N+D: 73.6dB S/N+D: 73.9 dB -20.0 -20.0 -40.0 Signal Amplitude Relative to Full Scale -60.0 (dB) -40.0 Signal Amplitude Relative to -60.0 Full Scale (dB) -80.0 -80.0 -100.0 -100.0 -120.0 -120.0 dc fs /2 dc Input Frequency Figure 18. Plot of Ideal 12-bit ADC Signal Amplitude Relative to Full Scale (dB) Sampling Rate: 100kHz Full Scale: 9Vp-p S/N+D: 72.9dB -40.0 -60.0 -80.0 -100.0 -120.0 12.0 dc Input Frequency (kHz) 50.0 Figure 20. FFT Plot of CS5012A with 12 kHz Full-Scale Input ures 21 and 23 show FFT plots computed from an ideal 14 and 16-bit sinewave multiplied by a Blackman-Harris window. Artifacts of windowing are discarded from the signal-to-noise calculation using the assumption that quantization noise is white. All FFT plots in this data sheet were derived by averaging the FFT results from ten 1024 point time records. This filters the spectral variability that can arise from capturing finite time records without disturbing the total energy outside the fundamental. All harmonics which exist above the noise floor and the -92 dB side-lobes from the Blackman-Harris window are therefore clearly visible in the plots. For more information on FFT’s and windowing refer to: F.J. HARRIS, "On the use of windows for harmonic 0dB 0dB S/(N+D): 86.1 dB Signal Amplitude Relative to Full Scale -20dB -20dB -40dB -40dB -60dB Signal Amplitude Relative to Full Scale -80dB Sampling Rate: 56 kHz Full Scale: 9V p-p S/(N+D): 85.3 dB -60dB -80dB -100dB -100dB -120dB -120dB fs /2 dc Input Frequency Figure 21. Plot of Ideal 14-bit ADC 2-32 50.0 Input Frequency (kHz) Figure 19. Plot of CS5012A with 1 kHz Full Scale Input 0.0 -20.0 1.0 dc 1 kHz 28 kHz Input Frequency Figure 22. CS5014 FFT plot with 1 kHz Full Scale Input DS14F6 CS5012A, CS5014, CS5016 0dB 0dB Sampling Rate: 50 kHz Full Scale: 9V p-p S/(N+D): 92.4 dB S/(N+D): 97.5 dB -20dB -20dB -40dB -40dB -60dB -60dB Signal Amplitude Relative to Full Scale Signal Amplitude Relative to Full Scale -80dB -100dB -120dB -80dB -100dB -120dB dc fs /2 dc 1 kHz 25 kHz Input Frequency Input Frequency Figure 24. CS5016 FFT plot with 1 kHz Full Scale Input Figure 23. Plot of Ideal 16-bit ADC CS5012A High Frequency Performance analysis with the Discrete Fourier Transform", Proc. IEEE, Vol. 66, No. 1, Jan 1978, pp.51-83. This is available on request from Crystal Semiconductor. The CS5012A performs very well over a wide range of input frequencies as shown in Figure 25. The figure depicts the CS5012A-KP7 tested under four different conditions. The conditions include tests with the voltage reference set at 4.5 and at 2.5 volts with input signals at 0.5 dB down from full scale and 6.0 dB down from full scale. The sample rate is at 100 kHz for all cases. The plots indicate that the part performs very well even with input frequencies above the Nyquist rate. Best performance at the higher frequencies is achieved with a 2.5 volt reference. Figures 19, 22, and 24 show the performance of the CS5012A/14/16 with 1kHz full scale inputs. Figure 20 shows CS5012A performance with 12kHz full scale inputs. Notice that the performance CS5012A/14/16 closely approaches that of the corresponding ideal ADC. 75 fs =100 kHz CS5012A-KP7 1. 2. 3. 4. 1 70 Signal to Noise + Distortion (dB) VREF 4.5 2.5 4.5 2.5 Signal FS-0.5dB FS-0.5dB FS-6.0dB FS-6.0dB 2 65 60 4 3 55 0 20 40 60 f s /2 80 100 fs 120 140 160 180 200 Input Frequency (kHz) Figure 25. CS5012A High Frequency Input Performance DS14F6 2-33 CS5012A, CS5014, CS5016 100 dB 0dB 1 kHz 12 kHz 24 kHz S(N+D) 80 dB -40dB Input Frequencies 60 dB Signal Amplitude Relative to Full Scale 40 dB Sampling Rate: 56 kHz Full Scale: 9V p-p S/(N+D): 24.1 dB -20dB -60dB -80dB 20 dB -100dB 0 dB -100 dB -80 dB -60 dB -40 dB -20 dB -120dB 0 dB dc Analog Input Amplitude 1 kHz Figure 27. CS5014 FFT plot with 1 kHz -60 dB Input Figure 26. CS5014 S/(N+D) vs. Input Amplitude (9Vp-p Full-Scale Input) 100 dB 0dB 1 kHz 12 kHz 24 kHz 80 dB S(N+D) 28 kHz Input Frequency -40dB Input Frequency 60 dB Sampling Rate: 50 kHz Full Scale: 9V p-p S/(N+D): 9.6 dB -20dB -60dB Signal Amplitude Relative to Full Scale 40 dB -80dB -100dB 20 dB -120dB 0 dB -100 dB -80 dB -60 dB -40 dB -20 dB 0 dB Analog Input Amplitude Figure 28. CS5016 S/(N+D) vs. Input Amplitude (9Vp-p Full-Scale Input) Signal to Noise + Distortion vs Signal Level As illustrated in Figures 26 - 29, the CS5014/16’s on-chip self-calibration provides very accurate bit weights which yield no degradation in quantization noise with low-level input signals. In fact, quantization noise remains below the noise floor in the CS5016, which dictates the converter’s signal-to-noise performance. CS5016 Noise Considerations All analog circuitry in the CS5016 is wideband in order to achieve fast conversions and high throughput. Wideband noise in the CS5016 integrates to 35 µV rms in unipolar mode (70 µV rms in bipolar mode). This is approximately 1/2 LSB 2-34 dc 1 kHz 25 kHz Input Frequency Figure 29. CS5016 FFT plot with 1 kHz -80 dB Input rms with a 4.5V reference in both modes. Figure 30 shows a histogram plot of output code occurrences obtained from 5000 samples taken from a CS5016 in the bipolar mode. Hexadecimal code 80CD was arbitrarily selected and the analog input was set close to code center. With a noiseless converter, code 80CD would always appear. The histogram plot of the CS5016 has a "bell" shape with all codes other than 80CD due to internal noise. In a sampled data system all information about the analog input applied to the sample/hold appears in the baseband from dc to one-half the sampling rate. This includes high-frequency components which alias into the baseband. Low-pass (anti-alias) filters DS14F6 CS5012A, CS5014, CS5016 Count 5000 Noiseless Converter 4000 CS5016 3000 formed on the charge trapped on the capacitor array at the moment the HOLD command is given. The charge on the array is ideally related to the analog input voltage by Qin = -Vin x Ctot as shown in Figure 2. Any deviation from this ideal relationship will result in conversion errors even if the conversion process proceeds flawlessly. 2000 1000 80CA 80CB 80CC 80CD 80CE 80CF 80D0 Code (Hexadecimal) Counts: 0 11 911 3470 599 9 0 Figure 30. Histogram Plot of 5000 Conversion Inputs from the CS5016 are therefore used to remove frequency components in the input signal which are above one-half the sample rate. However, all wideband noise introduced by the CS5016 still aliases into the baseband. This "white" noise is evenly spread from dc to one-half the sampling rate and integrates to 35 µV rms in unipolar mode. Noise can be reduced by sampling at higher than the desired word rate and averaging multiple samples for each word. Oversampling spreads the CS5016’s noise over a wider band (for lower noise density), and averaging applies a low-pass response which filters noise above the desired signal bandwidth. In general, the CS5016’s noise performance can be maximized in any application by always sampling at the maximum specified rate of 50 kHz (for lowest noise density) and digitally filtering to the desired signal bandwidth. CS5014 and CS5016 Sampling Distortion The ultimate limitation on the CS5014/16’s linearity (and distortion) arises from nonideal sampling of the analog input voltage. The calibrated capacitor array used during conversions is also used to track and hold the analog input signal. The conversion is not performed on the analog input voltage per se, but is actually perDS14F6 At dc, the DAC capacitor array’s voltage coefficient dictates the converter’s linearity. This variation in capacitance with respect to applied signal voltage yields a nonlinear relationship between charge Qin and the analog input voltage Vin and places a bow or wave in the transfer function. This is the dominant source of distortion at low input frequencies (Figures 22 and 24). The ideal relationship between Qin and Vin can also be distorted at high signal frequencies due to nonlinearities in the internal MOS switches. Dynamic signals cause ac current to flow through the switches connecting the capacitor array to the analog input pin in the track mode. Nonlinear onresistance in the switches causes a nonlinear voltage drop. This effect worsens with increased signal frequency as shown in Figures 26 and 28 since the magnitude of the steady state current increases. First noticeable at 1 kHz, this distortion assumes a linear relationship with input frequency. With signals 20 dB or more below full-scale, it no longer dominates the converter’s overall S/(N+D) performance (Figures 31-34). This distortion is strictly an ac sampling phenomenon. If significant energy exists at high frequencies, the effect can be eliminated using an external track-and-hold amplifier to allow the array’s charge current to decay, thereby eliminating any voltage drop across the switches. Since the CS5014/16 has a second sampling function onchip, the external track-and-hold can return to the track mode once the converter’s HOLD input falls. It need only acquire the analog input by the time the entire conversion cycle finishes. 2-35 CS5012A, CS5014, CS5016 0dB 0dB Sampling Rate: 56 kHz Full Scale: 9V p-p S/(N+D): 81.5 dB -20dB -40dB Signal Amplitude Relative to Full Scale Sampling Rate: 56 kHz Full Scale: 9V p-p S/(N+D): 64.6 dB -20dB -40dB -60dB Signal Amplitude Relative to Full Scale -80dB -100dB -60dB -80dB -100dB -120dB -120dB dc 12 kHz 28 kHz dc Input Frequency 28 kHz Input Frequency Figure 31. CS5014 FFT plot with 12 kHz Full Scale Input Figure 32. CS5014 FFT plot with 12 kHz -20 dB Input 0dB 0dB Sampling Rate: 50 kHz Full Scale: 9V p-p S/(N+D): 84.3 dB -20dB Sampling Rate: 50 kHz Full Scale: 9V p-p S/(N+D): 71.9 dB -20dB -40dB -40dB Signal Amplitude Relative to Full Scale 12 kHz Signal Amplitude Relative to Full Scale -60dB -80dB -60dB -80dB -100dB -100dB -120dB -120dB dc 12 kHz 25 kHz dc Figure 33. CS5016 FFT plot with 12 kHz Full Scale Input 12 kHz 25 kHz Input Frequency Input Frequency Figure 34. CS5016 FFT plot with 12 kHz -20 dB Input Clock Feedthrough in the CS5014 and CS5016 Maintaining the integrity of analog signals in the presence of digital switching noise is a difficult problem. The CS5014/16 can be synchronized to the digital system using the CLKIN input to avoid conversion errors due to asynchronous interference. However, digital interference will still affect sampling purity due to coupling between the CS5014/16’s analog input and master clock. The effect of clock feedthrough depends on the sampling conditions. If the sampling signal at the HOLD input is synchronized to the master clock, clock feedthrough will appear as a dc offset at the CS5014/16’s output. The offset could theoretically reach the peak coupling magnitude 2-36 (Figure 35), but the probability of this occurring is small since the peaks are spikes of short duration. Master Clock Int/Ext Freq Internal 2MHz External 2MHz External 4MHz External 4MHz External 4MHz Analog Input Source Impedance 50 50 50 25 200 Clock Feedthrough RMS Peak-to-Peak 15uV 70uV 25uV 110uV 40uV 150uV 25uV 110uV 80uV 325uV Figure 35. Examples of Measured Clock Feedthrough If sampling is performed asynchronously with the master clock, clock feedthrough will appear as an ac error at the CS5014/16’s output. With a fixed DS14F6 CS5012A, CS5014, CS5016 sampling rate, a tone will appear as the clock frequency aliases into the baseband. The tone frequency can be calculated using the equation below and could be selectively filtered in software using DSP techniques. ftone = (N fs - fclk) where N = fclk/fs rounded to the nearest integer The magnitude of clock feedthrough depends on the master clock conditions and the source impedance applied to the analog input. When operating with the CS5014/16’s internally generated clock, the CLKIN input is grounded and the dominant source of coupling is through the device’s substrate. As shown in Figure 35, a typical CS5014/16 operating with their internal oscillator at 2 MHz and 50 Ω of analog input source impedance will exhibit only 15 µV rms of clock feedthrough. However, if a 2 MHz external clock is applied to CLKIN under the same conditions, feedthrough increases to 25 µV rms. Feedthrough also increases with clock frequency; a 4 MHz clock yields 40 µV rms. Clock feedthrough can be reduced by limiting the source impedance applied at the analog input. As shown in Figure 35, reducing source impedance from 50 Ω to 25 Ω yields a 15 µV rms reduction in feedthrough. Therefore, when operating the CS5014/16 with high-frequency external master clocks, it is important to minimize source impedance applied to the CS5014/16’s input. Also, the overall effect of clock feedthrough can be minimized by maximizing the input range and LSB size. The reference voltage applied to VREF can be maximized, and the CS5014/16 can be operated in bipolar mode which inherently doubles the LSB size over the unipolar mode. Differences between the CS5012A and the CS5012 The differences between the CS5012A and the CS5012 are tabulated in Table 3. The CS5012 is a short-cycled version of the CS5016 A/D converter and includes the same 18-bit calibration circuitry. This calibration circuitry sets the calibration resolution of the CS5012 at 1/64th of an LSB and achieves the near perfect differential linearity performance illustrated by the CS5012 DNL plot in Figure 15. The CS5012A calibration circuitry was modified to provide calibration to 15-bit resolution therefore achieving calibration to 1/8 of an LSB. This reduction in calibration resolution for the CS5012A reduces the time required to calibrate the device (see Table 3) and reduces the size of the total array capacitance. The reduced array capacitance improves the high frequency performance by allowing higher slew rate in the input circuitry. Table 3 documents some other improvements included in the CS5012A. The burst mode calibration was made functional, although it should not be used. The device was also modified so the EOC signal goes low at the end of a reset calibration in either microprocessor or microprocessor-independent mode. The CS5012A was modified to maintain a throughput rate of 64 CLKIN cycles in loopback mode for all frequencies of CLKIN. Schematic & Layout Review Service Confirm Optimum Schematic & Layout Before Building Your Board. For Our Free Review Service Call Applications Engineering. C a l l : ( 5 1 2 ) 4 4 5 - 7 2 2 2 DS14F6 2-37 CS5012A, CS5014, CS5016 CS5012 CS5012A Calibration resolution 15 bits. Results in DNL calibration to 1/8 LSB at 12 bits. 18 bits. Results in DNL calibration to 1/64 LSB at 12 bits. 58,280 CLKIN cycles 2,014 conversions fully functional 1,441,020 CLKIN cycles 72,051 conversions not functional End of calibration indicator EOC falls in either microprocessor or microprocessor-independent mode at the completion of a RESET calibration cycle. EOC falls at the completion of a RESET calibration cycle in microprocessor mode only. In microprocessor-independent mode EOT must be used. EOT falls 15 CLKIN cycles after completion of a RESET calibration. Throughput rate in loopback mode The device acquires and converts a sample in 64 CLKIN cycles for all CLKIN frequencies when in loopback. The device acquires and converts in 64 CLKIN cycles for CLKIN=4MHz, but will require 68 CLKIN cycles at 100kHz throughput. This is due to excess delay on EOT. Input capacitance in fine-charge mode 103pF typical, unipolar mode 72pF typical, bipolar mode 275pF typical, unipolar mode 165pF typical, bipolar mode Calibration time reset: interleave: burst: Slew Rate Unipolar Coarse charge Fine charge Bipolar Coarse charge Fine charge 20V/us 1.5V/us 5V/us 0.25V/us 40V/us 3.0V/us 10V/us 0.5V/us Table 3. Differences Between the CS5012A and CS5012 2-38 DS14F6 CS5012A, CS5014, CS5016 CS CAL INTRLV RD A0 RST X X X X * 0 Function Hold and Start Convert X 0 1 X X * 0 Initiate Burst Calibration 1 0 0 X X * 0 Stop Burst Cal and Begin Track X 0 X 0 X * 0 Initiate Interleave Calibration X X 1 X 0 0 0 1 X X X X 1 X X X X 0 0 X * 1 0 * 0 0 0 X Terminate Interleave Cal X X 0 X X 0 X X X X X X 1 X X * X 0 X 1 X High Impedance Data Bus Reset Reset HOLD Read Output Data Read Status Register High Impedance Data Bus * The status of A0 is not critical to the operation specified. However, A0 should not be low with CS and HOLD low, or a software reset will result. Table 4. CS5012A/14/16 Truth Table +5V Analog Supply 10 Ω 25 VA+ 0.1 µF CS5012A CS5014 CS5016 Analog Signal Source Signal Conditioning 200 Ω 26 1000 pF 0 VREF or ±VREF 28 Voltage Reference 0.01 µF 10 µF 27 0.1 µF -5V Analog Supply 29 0.1 µF AIN 11 VD+ 33 BW BP/UP 24 0.1 µF Mode Select * 20 Clock Source (optional) SDATA 40 SCLK 39 Serial Data Interface (optional) CLKIN Data D0-D15 8 or 16 Processor EOC 38 EOT 37 1 HOLD 35 CAL Control 34 Logic INTRLV 21 CS 22 RD VREF 23 A0 32 RESET Reset 31 AGND Generator TST 10 REFBUF DGND VAVD- 0.1 µF 36 30 10 Ω May be microprocessor or discrete logic. Unused Logic inputs should only be connected to VD+ or DGND. * BW and BP/UP should always be terminated to VD+ or DGND, or driven by a logic gate. For best dynamic S/(N+D) performance. Figure 36. CS5012A/14/16 System Connection Diagram DS14F6 2-39 CS5012A, CS5014, CS5016 HOLD CS5016 (LSB) DATA BUS BIT 0 DATA BUS BIT 1 CS5014 (LSB) DATA BUS BIT 2 DATA BUS BIT 3 CS5012 (LSB) DATA BUS BIT 4 DATA BUS BIT 5 DATA BUS BIT 6 DATA BUS BIT 7 DIGITAL GROUND POSITIVE DIGITAL POWER DATA BUS BIT 8 DATA BUS BIT 9 DATA BUS BIT 10 DATA BUS BIT 11 DATA BUS BIT 12 DATA BUS BIT 13 DATA BUS BIT 14 (MSB) DATA BUS BIT 15 CLOCK INPUT HOLD D0 D1 D2 D3 D4 D5 D6 NC D7 DGND VD+ NC D8 NC D9 D10 D11 D12 D13 D14 D15 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 HOLD D0 D1 D2 D3 D4 D5 D6 D7 DGND VD+ D8 D9 D10 D11 D12 D13 D14 D15 CLKIN 7 8 9 10 11 12 13 14 15 16 17 6 40 39 38 37 36 35 34 33 32 31 30 29 28 27 26 25 24 23 22 21 CS5012A CS5012 CS5014 CS5016 4 2 1 44 SDATA SCLK EOC EOT VDCAL INTRLV BW RST TST VAREFBUF VREF AGND AIN VA+ BP/UP A0 RD CS 42 40 26 28 top view CS5012A CS5012 CS5014 CS5016 18 20 22 24 39 38 37 36 35 34 33 32 31 30 29 SERIAL OUTPUT SERIAL CLOCK END OF CONVERSION END OF TRACK NEGATIVE DIGITAL POWER CALIBRATE INTERLEAVE BUS WIDTH SELECT RESET TEST NEGATIVE ANALOG POWER REFERENCE BUFFER OUTPUT VOLTAGE REFERENCE ANALOG GROUND ANALOG INPUT POSITIVE ANALOG POWER BIPOLAR/UNIPOLAR SELECT READ ADDRESS READ CHIP SELECT SDATA SCLK EOC EOT VDCAL INTRLV BW RST TST VANC REFBUF VREF AGND AIN VA+ BP/UP A0 RD CS CLKIN NOTE: All pin references in this data sheet refer to the 40-pin DIP package numbering. Use this figure to determine pin numbers for 44-pin package. 2-40 DS14F6 CS5012A, CS5014, CS5016 PIN DESCRIPTIONS Power Supply Connections VD+ – Positive Digital Power, PIN 11. Positive digital power supply. Nominally +5 volts. VD- – Negative Digital Power, PIN 36. Negative digital power supply. Nominally -5 volts. DGND – Digital Ground, PIN 10. Digital ground. VA+ – Positive Analog Power, PIN 25. Positive analog power supply. Nominally +5 volts. VA- – Negative Analog Power, PIN 30. Negative analog power supply. Nominally -5 volts. AGND – Analog Ground, PIN 27. Analog ground. Oscillator CLKIN – Clock Input, PIN 20. All conversions and calibrations are timed from a master clock which can either be supplied by driving this pin with an external clock signal, or can be internally generated by tying this pin to DGND. Digital Inputs HOLD – Hold, PIN 1. A falling transition on this pin sets the CS5012A/14/16 to the hold state and initiates a conversion. This input must remain low at least one CLKIN cycle plus 50 ns. CS – Chip Select, PIN 21. When high, the data bus outputs are held in a high impedance state and the input to CAL and INTRLV are ignored. A falling transition initiates or terminates burst or interleave calibration (depending on the status of CAL and INTRLV) and a rising transition latches both the CAL and INTRLV inputs. If RD is low, the data bus is driven as indicated by BW and A0. RD – Read, PIN 22. When RD and CS are both low, data is driven onto the data bus. If either signal is high, the data bus outputs are held in a high impedance state. The data driven onto the bus is determined by BW and A0. DS14F6 2-41 CS5012A, CS5014, CS5016 A0 – Read Address, PIN 23. Determines whether data or status information is placed onto the data bus. When high during the read operation, converted data is placed onto the data bus; when low, the status register is driven onto the bus. BP/UP – Bipolar/Unipolar Input Select, PIN 24. When high, the device is configured with a bipolar transfer function ranging from -VREF to +VREF. Encoding is in an offset binary format, with the mid-scale code 100...0000 centered at AGND. When low, the device is configured for a unipolar transfer function from AGND to VREF. Unipolar encoding is in straight binary format. Once calibration has been performed, either bipolar or unipolar mode may be selected without the need to recalibrate. RST – Reset, PIN 32. When taken high for at least 100 ns, all internal digital logic is reset. Upon being taken low, a full calibration sequence is initiated. BW – Bus Width Select, PIN 33. When hard-wired high, all 12 data bits are driven onto the bus simultaneously during a data read cycle. When low, the bus is in a byte wide format. On the first read following a conversion, the eight MSB’s are driven onto D0-D7. A second read cycle places the four LSB’s with four trailing zeros on D0-D7. Subsequent reads will toggle the higher/lower order byte. Regardless of BW’s status, a read cycle with A0 low yields the status information on D0-D7. INTRLV – Interleave, PIN 34. When latched low using CS, the device goes into interleave calibration mode. A full calibration will complete every 2,014 conversions in the CS5012A, and every 72,051 conversions in the CS5014/16. The effective conversion time extends by 20 clock cycles. CAL – Calibrate, PIN 35. (See Addendum appending this data sheet)) When latched high using CS, burst calibration results. The device cannot perform conversions during the calibration period which will terminate only once CAL is latched low again. Calibration picks up where the previous calibration left off, and calibration cycles complete every 58,280 CLKIN cycles in the CS5012A, and every 1,441,020 CLKIN cycles in the CS5014/16 . If the device is converting when a calibration is signaled, it will wait until that conversion completes before beginning. Analog Inputs AIN – Analog Input, PIN 26. Input range in the unipolar mode is zero volts to VREF. Input range in bipolar mode is -VREF to +VREF. The output impedance of buffer driving this input should be less than or equal to 200 Ω. 2-42 DS14F6 CS5012A, CS5014, CS5016 VREF – Voltage Reference, PIN 28. The analog reference voltage which sets the analog input range. It represents positive full scale for both bipolar and unipolar operation, and its magnitude sets negative full scale in bipolar mode. Digital Outputs D0 through D15 – Data Bus Outputs, PINS 2 thru 9, 12 thru 19. 3-state output pins. Enabled by CS and RD, they offer the converter’s output in a format consistent with the state of BW if A0 is high. If A0 is low, bits D0-D7 offer status register data. EOT – End Of Track, PIN 37. If low, indicates that enough time has elapsed since the last conversion for the device to acquire the analog input signal. EOC – End Of Conversion, PIN 38. This output indicates the end of a conversion or calibration cycle. It is high during a conversion and will fall to a low state upon completion of the conversion cycle indicating valid data is available at the output. Returns high on the first subsequent read or the start of a new conversion cycle. SDATA – Serial Output, PIN 40. Presents each output data bit after it is determined by the successive approximation algorithm. Valid on the rising edge of SCLK, data appears MSB first, LSB last, and each bit remains valid until the next bit appears. SCLK – Serial Clock Output, PIN 39. Used to clock converted output data serially from the CS5012A/14/16. Serial data is stable on the rising edge of SCLK. Analog Outputs REFBUF – Reference Buffer Output, PIN 29. Reference buffer output. A 0.1 µF ceramic capacitor must be tied between this pin and VA-. Miscellaneous TST – Test, PIN 31. Allows access to the CS5012A/14/16’s test functions which are reserved for factory use. Must be tied to DGND. DS14F6 2-43 CS5012A, CS5014, CS5016 PARAMETER DEFINITIONS Linearity Error The deviation of a code from a straight line passing through the endpoints of the transfer function after zero- and full-scale errors have been accounted for. "Zero-scale" is a point 1/2 LSB below the first code transition and "full-scale" is a point 1/2 LSB beyond the code transition to all ones. The deviation is measured from the middle of each particular code. Units in % Full-Scale. Differential Linearity Minimum resolution for which no missing codes is guaranteed. Units in bits. Full Scale Error The deviation of the last code transition from the ideal (VREF-3/2 LSB’s). Units in LSB’s. Unipolar Offset The deviation of the first code transition from the ideal (1/2 LSB above AGND) when in unipolar mode (BP/UP low). Units in LSB’s. Bipolar Offset The deviation of the mid-scale transition (011...111 to 100...000) from the ideal (1/2 LSB below AGND) when in bipolar mode (BP/UP high). Units in LSB’s. Bipolar Negative Full-Scale Error The deviation of the first code transition from the ideal when in bipolar mode (BP/UP high). The ideal is defined as lying on a straight line which passes through the final and mid-scale code transitions. Units in LSB’s. Peak Harmonic or Spurious Noise (More accurately, Signal to Peak Harmonic or Spurious Noise) The ratio of the rms value of the signal to the rms value of the next largest spectral component below the Nyquist rate (excepting dc). This component is often an aliased harmonic when the signal frequency is a significant proportion of the sampling rate. Expressed in decibels. Total Harmonic Distortion The ratio of the rms sum of all harmonics to the rms value of the signal. Units in percent. Signal-to-Noise Ratio The ratio of the rms value of the signal to the rms sum of all other spectral components below the Nyquist rate (excepting dc), including distortion components. Expressed in decibels. Aperture Time The time required after the hold command for the sampling switch to open fully. Effectively a sampling delay which can be nulled by advancing the sampling signal. Units in nanoseconds. Aperture Jitter The range of variation in the aperture time. Effectively the "sampling window" which ultimately dictates the maximum input signal slew rate acceptable for a given accuracy. Units in picoseconds. NOTE: Temperatures specified define ambient conditions in free-air during test and do not refer to the junction temperature of the device. 2-44 DS14F6 CS5012A, CS5014, CS5016 CS5012A Ordering Guide Model CS5012A-KP12 CS5012A-KP7 CS5012A-KL12 CS5012A-KL7 CS5012A-BP12 CS5012A-BP7 CS5012A-BL12 CS5012A-BL7 5962-8967901QA 5962-8967901XA Throughput Conversion Time 63 kHz 12.25 µs 100 kHz 7.20µs 63 kHz 12.25 µs 100 kHz 7.20 µs 63 kHz 12.25 µs 100 kHz 7.20 µs 63 kHz 12.25 µs 100 kHz 7.20 µs 63 kHz 12.25 µs 63 kHz 12.25 µs Maximum DNL ±1/2 LSB ±1/2 LSB ±1/2 LSB ±1/2 LSB ±1/2 LSB ±1/2 LSB ±1/2 LSB ±1/2 LSB ±1/2 LSB ±1/2 LSB Temp. Range 0 to 70 °C 0 to 70 °C 0 to 70 °C 0 to 70 °C -40 to +85 °C -40 to +85 °C -40 to +85 °C -40 to +85 °C -55 to +125 °C -55 to +125 °C Package 40-Pin Plastic DIP 40-Pin Plastic DIP 44-Pin PLCC 44-Pin PLCC 40-Pin Plastic DIP 40-Pin Plastic DIP 44-Pin PLCC 44-Pin PLCC 40-Pin CerDIP 44-Pin Ceramic LCC The CS5012A is recommended for new designs. The following is a list of upgraded part numbers. Discontinued Part Number CS5012-KP24 CS5012-KP12 CS5012-KP7 CS5012-KL24 CS5012-KL12 CS5012-KL7 CS5012-BD24 CS5012-BD12 CS5012-BD7 CS5012-BL24 CS5012-BL12 CS5012-BL7 Equivalent Recommended Device. CS5012A-KP12 CS5012A-KP12 CS5012A-KP7 CS5012A-KL12 CS5012A-KL12 CS5012A-KL7 CS5012A-BP12 CS5012A-BP12 CS5012A-BP7 CS5012A-BL12 CS5012A-BL12 CS5012A-BL7 CS5012-TD24B CS5012-TD12B CS5012-TE24B CS5012-TE12B 5962-897901QA 5962-897901QA 5962-897901XA 5962-897901XA AS PER EOL MAY 22, 2000 DS14F6 CURRENT PRODUCT BEING OBSOLETED REPLACEMENT PRODUCT CS5012A -KP12 CS5012A -KP7 CS5012A -BP12 40 pin DIP 40 pin DIP 40 pin DIP CS5012A -BP7 CS5012A -BP7 CS5012A -BP7 CS5012A -KL12 CS5012A -KL7 CS5012A -BL12 44 pin PLCC 44 pin PLCC 44 pin PLCC CS5012A -BL7 CS5012A -BL7 CS5012A -BL7 2-45 CS5012A, CS5014, CS5016 CS5014 Ordering Guide Model CS5014-KP28 CS5014-KP14 CS5014-KL28 CS5014-KL14 CS5014-BP28 CS5014-BP14 CS5014-BL28 CS5014-BL14 CS5014-SD14 CS5014-TD14 CS5014-SE14 CS5014-TE14 5962-8967401QA 5962-8967402QA 5962-8967401XA 5962-8967402XA Throughput 28 kHz 56 kHz 28 kHz 56 kHz 28 kHz 56 kHz 28 kHz 56 kHz 56 kHz 56 kHz 56 kHz 56 kHz 56 kHz 56 kHz 56 kHz 56 kHz Conversion Time 28.50 µs 14.25 µs 28.50 µs 14.25 µs 28.50 µs 14.25µs 28.50 µs 14.25 µs 14.25 µs 14.25 µs 14.25 µs 14.25 µs 14.25 µs 14.25 µs 14.25 µs 14.25 µs Linearity ±0.5 LSB ±0.5 LSB ±0.5 LSB ±0.5 LSB ±0.5 LSB ±0.5 LSB ±0.5 LSB ±0.5 LSB ±1.5 LSB ±0.5 LSB ±1.5 LSB ±0.5 LSB ±1.5 LSB ±0.5 LSB ±1.5 LSB ±0.5 LSB Temp. Range 0 to 70 °C 0 to 70 °C 0 to 70 °C 0 to 70 °C -40 to +85 °C -40 to +85 °C -40 to +85 °C -40 to +85 °C -55 to +125 °C -55 to +125 °C -55 to +125 °C -55 to +125 °C -55 to +125 °C -55 to +125 °C -55 to +125 °C -55 to +125 °C Package 40-Pin Plastic DIP 40-Pin Plastic DIP 44-Pin PLCC 44-Pin PLCC 40-Pin Plastic DIP 40-Pin Plastic DIP 44-Pin PLCC 44-Pin PLCC 40-Pin CerDIP 40-Pin CerDIP 44-Pin Ceramic LCC 44-Pin Ceramic LCC 40-Pin CerDIP 40-Pin CerDIP 44-Pin Ceramic LCC 44-Pin Ceramic LCC The following is a list of upgraded part numbers. Discontinued Part Number CS5014-SD14B CS5014-TD14B CS5014-SE14B CS5014-TE14B Equivalent Recommended Device 5962-8967401QA 5962-8967402QA 5962-8967401XA 5962-8967402XA AS PER EOL MAY 22, 2000 CURRENT PRODUCT BEING OBSOLETED 2-46 REPLACEMENT PRODUCT CS5014 -KP28 CS5014 -KP14 CS5014 -BP28 40 pin DIP 40 pin DIP 40 pin DIP CS5014 -BP14 CS5014 -BP14 CS5014 -BP14 CS5014 -KL28 CS5014 -KL14 CS5014 -BL28 44 pin PLCC 44 pin PLCC 44 pin PLCC CS5014 -BL14 CS5014 -BL14 CS5014 -BL14 DS14F6 CS5012A, CS5014, CS5016 CS5016 Ordering Guide Model CS5016-JP32 CS5016-JP16 CS5016-KP32 CS5016-KP16 CS5016-JL32 CS5016-JL16 CS5016-KL32 CS5016-KL16 CS5016-AP32 CS5016-AP16 CS5016-BP32 CS5016-BP16 CS5016-AL32 CS5016-AL16 CS5016-BL32 CS5016-BL16 CS5016-SD16 CS5016-TD16 CS5016-SE16 CS5016-TE16 5962-8967601QA 5962-8967602QA 5962-8967601XA 5962-8967602XA Linearity .0030% .0030% .0015% .0015% .0030% .0030% .0015% .0015% .0030% .0030% .0015% .0015% .0030% .0030% .0015% .0015% .0076% .0015% .0076% .0015% .0076% .0015% .0076% .0015% Signal to Noise Ratio 87 dB 87 dB 90 dB 90 dB 87 dB 87 dB 90 dB 90 dB 87 dB 87 dB 90 dB 90 dB 87 dB 87 dB 90 dB 90 dB 87 dB 90 dB 87 dB 90 dB 87 dB 90 dB 87 dB 90 dB Conversion Time 32.50 µs 16.25 µs 32.50 µs 16.25 µs 32.50 µs 16.25 µs 32.50 µs 16.25 µs 32.50 µs 16.25 µs 32.50 µs 16.25 µs 32.50 µs 16.25 µs 32.50 µs 16.25 µs 16.25 µs 16.25 µs 16.25 µs 16.25 µs 16.25 µs 16.25 µs 16.25 µs 16.25 µs Temp. Range 0 to 70 °C 0 to 70 °C 0 to 70 °C 0 to 70 °C 0 to 70 °C 0 to 70 °C 0 to 70 °C 0 to 70 °C -40 to +85 °C -40 to +85 °C -40 to +85 °C -40 to +85 °C -40 to +85 °C -40 to +85 °C -40 to +85 °C -40 to +85 °C -55 to +125 °C -55 to +125 °C -55 to +125 °C -55 to +125 °C -55 to +125 °C -55 to +125 °C -55 to +125 °C -55 to +125 °C Package 40-Pin Plastic DIP 40-Pin Plastic DIP 40-Pin Plastic DIP 40-Pin Plastic DIP 44-Pin PLCC 44-Pin PLCC 44-Pin PLCC 44-Pin PLCC 40-Pin Plastic DIP 40-Pin Plastic DIP 40-Pin Plastic DIP 40-Pin Plastic DIP 44-Pin PLCC 44-Pin PLCC 44-Pin PLCC 44-Pin PLCC 40-Pin CerDIP 40-Pin CerDIP 44-Pin Ceramic LCC 44-Pin Ceramic LCC 40-Pin CerDIP 40-Pin CerDIP 44-Pin Ceramic LCC 44-Pin Ceramic LCC The following is a list of upgraded part numbers. Discontinued Part Number CS5016-SD16B CS5016-TD16B CS5016-SE16B CS5016-TE16B DS14F6 Equivalent Recommended Device 5962-8967601QA 5962-8967602QA 5962-8967601XA 5962-8967602XA 2-47 40 21 40 pin Plastic DIP E1 20 1 D A SEATING PLANE B1 A1 e1 L ∝ C eA B NOTES: 1. POSITIONAL TOLERANCE OF LEADS SHALL BE WITHIN 0.25mm (0.010") AT MAXIMUM MATERIAL CONDITION, IN RELATION TO SEATING PLANE AND EACH OTHER. 2. DIMENSION eA TO CENTER OF LEADS WHEN FORMED PARALLEL. 3. DIMENSION E1 DOES NOT INCLUDE MOLD FLASH. DIM A A1 B B1 C D E1 e1 eA L ∝ MILLIMETERS MIN NOM MAX 3.94 4.32 5.08 0.51 0.76 1.02 0.36 0.46 0.56 1.02 1.27 1.65 0.20 0.25 0.38 51.69 52.20 52.71 13.72 13.97 14.22 2.41 2.54 2.67 15.87 15.24 3.81 3.18 15° 0° - MIN 0.155 0.020 0.014 0.040 0.008 2.035 0.540 0.095 0.600 0.125 0° INCHES NOM MAX 0.170 0.200 0.030 0.040 0.018 0.022 0.050 0.065 0.010 0.015 2.055 2.075 0.550 0.560 0.100 0.105 0.625 0.150 15° - 44 pin PLCC NO. OF TERMINALS MILLIMETERS INCHES DIM MIN NOM MAX MIN NOM MAX A A1 4.20 4.45 4.57 0.165 0.175 0.180 2.29 2.79 3.04 0.090 0.110 0.120 B 0.33 0.41 0.53 0.013 0.016 0.021 E1 E D/E 17.40 17.53 17.65 0.685 0.690 0.695 D1 D1/E1 16.51 16.59 16.66 0.650 0.653 0.656 D D2/E2 14.99 15.50 16.00 0.590 0.610 0.630 e B e A1 D2/E2 A 1.19 1.27 1.35 0.047 0.050 0.053 40 21 MILLIMETERS 40 pin CerDIP E1 20 1 D A SEATING PLANE B1 A1 L ∝ C eA e1 B NOTES: 1. POSITIONAL TOLERANCE OF LEADS SHALL BE WITHIN 0.13mm (0.005") AT MAXIMUM MATERIAL CONDITION, IN RELATION TO SEATING PLANE AND EACH OTHER. 2. DIMENSION eA TO CENTER OF LEADS WHEN FORMED PARALLEL. DIM A A1 B B1 C D E1 e1 eA L ∝ MIN 4.06 0.51 0.38 NOM 0.46 1.27 INCHES MAX MIN NOM MAX 0.230 5.84 0.160 1.27 0.020 0.050 0.56 0.015 0.018 0.022 0.065 1.65 0.050 0.20 0.25 0.30 50.29 52.32 52.57 12.70 14.73 15.37 2.41 2.54 2.67 15.11 15.24 15.37 2.92 5° 3.81 0.008 0.010 0.012 1.980 2.060 2.070 0.500 0.580 0.605 0.095 0.100 0.105 0.595 0.600 0.605 4.06 0.115 0.150 0.160 5° 15° 15° A 28/44 pin CLCC 1 Top View B D4/E4 E2 E1 E B1 e1 D2 D1 D NO. OF TERMINALS 28 DIM 44 MILLIMETERS INCHES MILLIMETERS INCHES MIN NOM MAX MIN NOM MAX MIN NOM MAX MIN NOM MAX A 2.54 3.05 3.43 0.100 0.120 0.135 2.54 3.05 3.43 0.100 0.120 0.135 B 0.33 0.46 0.58 0.013 0.018 0.023 0.33 0.46 0.58 0.013 0.018 0.023 0.51 0.64 0.81 B1 0.02 0.025 0.032 0.51 0.64 0.81 0.02 0.025 0.032 D/E 12.19 12.46 12.70 0.480 0.490 0.500 17.27 17.53 17.78 0.680 0.690 0.700 D1/E1 11.18 11.43 11.68 0.440 0.450 0.460 16.26 16.51 16.76 0.640 0.650 0.660 D2/E2 7.49 D4/E4 10.80 10.92 11.05 0.425 0.430 0.435 15.88 16.00 16.13 0.625 0.630 0.635 e1 1.14 7.62 1.27 7.75 0.295 0.300 0.305 12.57 12.70 12.83 0.495 0.500 0.505 1.40 0.045 0.050 0.055 1.14 1.27 1.40 0.045 0.050 0.055 Smart AnalogTM is a Trademark of Crystal Semiconductor Corporation