LTC1604 High Speed, 16-Bit, 333ksps Sampling A/D Converter with Shutdown DESCRIPTION FEATURES n n n n n n n n n n n A Complete, 333ksps 16-Bit ADC 90dB S/(N+D) and -100dB THD (Typ) Power Dissipation: 220mW (Typ) No Pipeline Delay No Missing Codes over Temperature Nap (7mW) and Sleep (10W) Shutdown Modes Operates with Internal 15ppm/C Reference or External Reference True Differential Inputs Reject Common Mode Noise 5MHz Full Power Bandwidth 2.5V Bipolar Input Range 36-Pin SSOP Package The LTC(R)1604 is a 333ksps, 16-bit sampling A/D converter that draws only 220mW from 5V supplies. This high performance device includes a high dynamic range sample-and-hold, a precision reference and a high speed parallel output. Two digitally selectable power shutdown modes provide power savings for low power systems. The LTC1604's full-scale input range is 2.5V. Outstanding AC performance includes 90dB S/(N+D) and -100dB THD at a sample rate of 333ksps. The unique differential input sample-and-hold can acquire single-ended or differential input signals up to its 15MHz bandwidth. The 68dB common mode rejection allows users to eliminate ground loops and common mode noise by measuring signals differentially from the source. APPLICATIONS n n n n n n Telecommunications Digital Signal Processing Multiplexed Data Acquisition Systems High Speed Data Acquisition Spectrum Analysis Imaging Systems The ADC has P compatible,16-bit parallel output port. There is no pipeline delay in conversion results. A separate convert start input and a data ready signal (BUSY) ease connections to FlFOs, DSPs and microprocessors. L, LTC and LT are registered trademarks of Linear Technology Corporation. TYPICAL APPLICATION 10F 2.2F 10 + 3 36 AVDD VREF 5V 10F 5V + 35 AVDD 9 10F + LTC1604 4096 Point FFT 0 10 DVDD -20 SHDN 33 + 7.5k 4.375V 1.75X 2.5V REF 47F CS 32 P CONTROL LINES CONVST 31 RD 30 BUSY 27 OVDD 29 + 1 AIN+ DIFFERENTIAL ANALOG INPUT 2.5V + 2 AIN - - OGND 28 16-BIT SAMPLING ADC AGND 5 AGND 6 OUTPUT BUFFERS B15 TO B0 AGND 7 -40 -60 -80 -100 -120 16-BIT PARALLEL BUS -140 11 TO 26 AGND VSS 8 D15 TO D0 5V OR 3V 10F AMPLITUDE (dB) 4 REFCOMP CONTROL LOGIC AND TIMING fSAMPLE = 333kHz fIN = 100kHz SINAD = 89dB THD = -96dB DGND 0 20 40 60 80 100 120 140 160 FREQUENCY (kHz) 1604 TA01 34 1604 TA02 + 10F -5V 1604fa 1 LTC1604 ABSOLUTE MAXIMUM RATINGS PACKAGE/ORDER INFORMATION AVDD = DVDD = OVDD = VDD (Notes 1, 2) ORDER PART NUMBER TOP VIEW Supply Voltage (VDD) ..................................................6V Negative Supply Voltage (VSS) ..................................-6V Total Supply Voltage (VDD to VSS) ............................12V Analog Input Voltage (Note 3) ..........................(VSS - 0.3V) to (VDD + 0.3V) VREF Voltage (Note 4)................... -0.3V to (VDD + 0.3V) REFCOMP Voltage (Note 4).......... -0.3V to (VDD + 0.3V) Digital Input Voltage (Note 4)..................... -0.3V to 10V Digital Output Voltage .................. -0.3V to (VDD + 0.3V) Power Dissipation .............................................. 500mW Operating Temperature Range LTC1604C ................................................ 0C to 70C LTC1604I..............................................-40C to 85C Storage Temperature Range .................. -65C to 150C Lead Temperature (Soldering, 10 sec)................... 300C AIN+ 1 36 AVDD AIN- 2 35 AVDD VREF 3 34 VSS REFCOMP 4 33 SHDN AGND 5 32 CS AGND 6 31 CONV AGND 7 30 RD AGND 8 29 OVDD DVDD 9 28 OGND DGND 10 27 BUSY D15 (MSB) 11 26 D0 D14 12 25 D1 D13 13 24 D2 D12 14 23 D3 D11 15 22 D4 D10 16 21 D5 D9 17 20 D6 D8 18 19 D7 LTC1604CG LTC1604IG LTC1604ACG LTC1604AIG G PACKAGE 36-LEAD PLASTIC SSOP TJMAX = 125C, JA = 95C/W Consult factory for Military grade parts. CONVERTER CHARACTERISTICS PARAMETER With Internal Reference (Notes 5, 6) CONDITIONS l Resolution (No Missing Codes) Integral Linearity Error (Note 7) Transition Noise (Note 8) MIN LTC1604 TYP 15 16 l MAX MIN 16 1 4 Offset Error (Note 9) Offset Tempco (Note 9) Full-Scale Error Internal Reference External Reference Full-Scale Tempco IOUT(Reference) = 0, Internal Reference 0.05 16 0.5 0.7 l LTC1604A TYP MAX Bits 2 0.7 0.125 0.05 0.5 0.125 0.125 15 LSB LSB 0.125 0.5 0.25 0.25 UNITS % ppm/C 0.25 0.25 15 % % ppm/C ANALOG INPUT SYMBOL PARAMETER CONDITIONS VIN Analog Input Range (Note 2) 4.75 VDD 5.25V, -5.25 VSS -4.75V, VSS (AIN-, AIN+) AVDD IIN Analog Input Leakage Current CS = High CIN Analog Input Capacitance Between Conversions During Conversions tACQ tAP tjitter Sample-and-Hold Acquisition Delay Time Jitter CMRR MIN TYP MAX 2.5 l UNITS V 1 A 43 5 pF pF Sample-and-Hold Acquisition Time 380 ns Sample-and-Hold Acquisition Delay Time -1.5 ns Analog Input Common Mode Rejection Ratio -2.5V < (AIN- = AIN+) < 2.5V 5 psRMS 68 dB 1604fa 2 LTC1604 DYNAMIC ACCURACY (Note 5) SYMBOL PARAMETER CONDITIONS S/N 5kHz Input Signal 100kHz Input Signal Signal-to-Noise Ratio S/(N + D) Signal-to-(Noise + Distortion) Ratio THD Total Harmonic Distortion Up to 5th Harmonic LTC1604 MIN TYP MAX l 90 90 5kHz Input Signal 100kHz Input Signal (Note 10) l 90 89 5kHz Input Signal 100kHz Input Signal l -100 -94 LTC1604A MIN TYP MAX 87 84 UNITS 90 90 dB dB 90 89 dB dB -100 -94 -88 dB dB SFDR Spurious Free Dynamic Range 100kHz Input Signal 96 96 dB IMD Intermodulation Distortion fIN1 = 29.37kHz, fIN2 = 32.446kHz -88 -88 dB 5 5 MHz 350 350 kHz Full Power Bandwidth Full Linear Bandwidth (S/(N + D) 84dB INTERNAL REFERENCE CHARACTERISTICS (Note 5) PARAMETER CONDITIONS MIN TYP MAX VREF Output Voltage IOUT = 0 2.475 2.500 2.515 VREF Output Tempco IOUT = 0 15 ppm/C VREF Line Regulation 4.75 VDD 5.25V -5.25V VSS -4.75V 0.01 0.01 LSB/V LSB/V VREF Output Resistance 0 |IOUT| 1mA 7.5 k REFCOMP Output Voltage IOUT = 0 DIGITAL INPUTS AND DIGITAL OUTPUTS 4.375 UNITS V V (Note 5) SYMBOL PARAMETER CONDITIONS VIH High Level Input Voltage VDD = 5.25V l VIL Low Level Input Voltage VDD = 4.75V l 0.8 V IIN Digital Input Current VIN = 0V to VDD l 10 A CIN Digital Input Capacitance VOH High Level Output Voltage VOL Low Level Output Voltage MIN VDD = 4.75V, IOUT = -10A VDD = 4.75V, IOUT = -400A l VDD = 4.75V, IOUT = 160A VDD = 4.75V, IOUT = 1.6mA l TYP MAX 2.4 UNITS V 5 pF 4.5 V V 4.0 0.05 0.10 0.4 V V IOZ Hi-Z Output Leakage D15 to D0 VOUT = 0V to VDD, CS High l 10 A COZ Hi-Z Output Capacitance D15 to D0 CS High (Note 11) l 15 pF ISOURCE Output Source Current VOUT = 0V -10 mA ISINK Output Sink Current VOUT = VDD 10 mA 1604fa 3 LTC1604 POWER REQUIREMENTS (Note 5) SYMBOL PARAMETER CONDITIONS MIN TYP MAX UNITS VDD Positive Supply Voltage (Notes 12, 13) 4.75 5.25 V VSS Negative Supply Voltage (Note 12) -4.75 -5.25 V IDD Positive Supply Current Nap Mode Sleep Mode CS = RD = 0V CS = 0V, SHDN = 0V CS = 5V, SHDN = 0V l 18 1.5 1 30 2.4 100 mA mA A ISS Negative Supply Current Nap Mode Sleep Mode CS = RD = 0V CS = 0V, SHDN = 0V CS = 5V, SHDN = 0V l 26 1 1 40 100 100 mA A A PD Power Dissipation Nap Mode Sleep Mode CS = RD = 0V CS = 0V, SHDN = 0V CS = 5V, SHDN = 0V l 220 7.5 0.01 350 12 1 mW mW mW TYP MAX UNITS 2.45 2.8 TIMING CHARACTERISTICS SYMBOL (Note 5) PARAMETER CONDITIONS MIN fSMPL(MAX) Maximum Sampling Frequency l tCONV Conversion Time l tACQ Acquisition Time tACQ+CONV Throughput Time (Acquisition + Conversion) t1 CS to RD Setup Time (Note 11) 333 1.5 kHz s l 480 ns l 3 s (Notes 11, 12) l 0 ns t2 CS to CONVST Setup Time (Notes 11, 12) l 10 ns t3 SHDN to CS Setup Time (Notes 11, 12) l 10 ns l 40 t4 SHDN to CONVST Wake-Up Time CS = Low (Note 12) t5 CONVST Low Time (Note 12) t6 CONVST to BUSY Delay CL = 25pF t7 Data Ready Before BUSY 400 ns ns 36 l 80 60 ns ns l 32 ns ns t8 Delay Between Conversions (Note 12) l 200 ns t9 Wait Time RD After BUSY (Note 12) l -5 ns t10 Data Access Time After RD CL = 25pF CL = 100pF t11 Bus Relinquish Time l l LTC1604C LTC1604I l l 40 50 60 ns ns 45 60 75 ns ns 50 60 70 75 ns ns ns t12 RD Low Time (Note 12) l t10 ns t13 CONVST High Time (Note 12) l 40 ns t14 Aperture Delay of Sample-and-Hold 2 ns 1604fa 4 LTC1604 TIMING CHARACTERISTICS (Note 5) The l denotes specifications that apply over the full operating temperature range. Note 1: Absolute Maximum Ratings are those values beyond which the life of a device may be impaired. Note 2: All voltage values are with respect to ground with DGND, OGND and AGND wired together unless otherwise noted. Note 3: When these pin voltages are taken below VSS or above VDD, they will be clamped by internal diodes. This product can handle input currents greater than 100mA below VSS or above VDD without latchup. Note 4: When these pin voltages are taken below VSS, they will be clamped by internal diodes. This product can handle input currents greater than 100mA below VSS without latchup. These pins are not clamped to VDD. Note 5: VDD = 5V, VSS = -5V, fSMPL = 333kHz, and tr = tf = 5ns unless otherwise specified. Note 6: Linearity, offset and full-scale specification apply for a singleended AIN+ input with AIN- grounded. Note 7: Integral nonlinearity is defined as the deviation of a code from a straight line passing through the actual endpoints of the transfer curve. The deviation is measured from the center of the quantization band. Note 8: Typical RMS noise at the code transitions. See Figure 17 for histogram. Note 9: Bipolar offset is the offset voltage measured from -0.5LSB when the output code flickers between 0000 0000 0000 0000 and 1111 1111 1111 1111. Note 10: Signal-to-Noise Ratio (SNR) is measured at 5kHz and distortion is measured at 100kHz. These results are used to calculate Signal-to-Nosie Plus Distortion (SINAD). Note 11: Guaranteed by design, not subject to test. Note 12: Recommended operating conditions. Note 13: The falling CONVST edge starts a conversion. If CONVST returns high at a critical point during the conversion it can create small errors. For best performance ensure that CONVST returns high either within 250ns after conversion start or after BUSY rises. TYPICAL PERFORMANCE CHARACTERISTICS Integral Nonlinearity vs Output Code Differential Nonlinearity vs Output Code 2.0 1.0 100 1.5 0.8 90 0.6 80 0.4 70 DNL (LSB) 0.5 0.0 -0.5 0.2 SINAD (dB) 1.0 INL (LSB) S/(N + D) vs Input Frequency and Amplitude 0.0 -0.2 40 30 20 -1.5 -0.8 10 -2.0 -32768 -1.0 0 16384 32767 0 -32768 -16384 CODE 0 16384 VIN = -40dB 50 -0.6 -16384 VIN = -20dB 60 -0.4 -1.0 VIN = 0dB 32767 1k 10k 100k FREQUENCY (Hz) CODE 1604 G11 1M 1604 G01 1604 G10 Signal-to-Noise Ratio vs Input Frequency SIGNAL-TO-NOISE RATIO (dB) 90 80 70 60 50 40 30 20 10 10k 100k FREQUENCY (Hz) 1M 1604 G03 0 0 -10 -20 -30 -40 -50 -60 -70 -80 THD 3RD -90 -100 -110 1k 2ND 10k 100k INPUT FREQUENCY (Hz) 1M 1604 G04 SPURIOUS-FREE DYNAMIC RANGE (dB) AMPLITUDE (dB BELOW THE FUNDAMENTAL) 100 0 1k Spurious-Free Dynamic Range vs Input Frequency Distortion vs Input Frequency -10 -20 -30 -40 -50 -60 -70 -80 -90 -100 -110 1k 10k 100k INPUT FREQUENCY (Hz) 1M 1604 G05 1604fa 5 LTC1604 TYPICAL PERFORMANCE CHARACTERISTICS Power Supply Feedthrough vs Ripple Frequency 0 fSAMPLE = 333kHz fIN1 = 29.3kHz fIN2 = 32.4kHz -20 AMPLITUDE (dB) -40 -60 -80 -100 -120 -140 AMPLITUDE OF POWER SUPPLY FEEDTHROUGH (dB) 0 20 40 60 80 100 120 140 160 FREQUENCY (kHz) 80 fSAMPLE = 333kHz VRIPPLE = 10mV -20 -40 -60 -80 AVDD -100 V SS -120 0 Input Common Mode Rejection vs Input Frequency COMMON MODE REJECTION (dB) Intermodulaton Distortion 70 60 50 40 30 20 10 0 1k 10k 100k INPUT FREQUENCY (Hz) 1604 G06 1M 1k 10k 100k INPUT FREQUENCY (Hz) 1604 G07 1M 1604G09 PIN FUNCTIONS AIN+ (Pin 1): Positive Analog Input. The ADC converts the difference voltage between AIN+ and AIN- with a differential range of 2.5V. AIN+ has a 2.5V input range when AIN- is grounded. OGND (Pin 28): Digital Ground for Output Drivers. AIN- (Pin 2): Negative Analog Input. Can be grounded, tied to a DC voltage or driven differentially with AIN+. RD (Pin 30): Read Input. A logic low enables the output drivers when CS is low. VREF (Pin 3): 2.5V Reference Output. Bypass to AGND with 2.2F tantalum in parallel with 0.1F ceramic. REFCOMP (Pin 4): 4.375 Reference Compensation Pin. Bypass to AGND with 47F tantalum in parallel with 0.1F ceramic. AGND (Pins 5 to 8): Analog Grounds. Tie to analog ground plane. DVDD (Pin 9): 5V Digital Power Supply. Bypass to DGND with 10F tantalum in parallel with 0.1F ceramic. DGND (Pin 10): Digital Ground for Internal Logic. Tie to analog ground plane. D15 to D0 (Pins 11 to 26): Three-State Data Outputs. D15 is the Most Significant Bit. BUSY (Pin 27): The BUSY output shows the converter status. It is low when a conversion is in progress. Data is valid on the rising edge of BUSY. OVDD (Pin 29): Digital Power Supply for Output Drivers. Bypass to OGND with 10F tantalum in parallel with 0.1F ceramic. CONVST (Pin 31): Conversion Start Signal. This active low signal starts a conversion on its falling edge when CS is low. CS (Pin 32): The Chip Select Input. Must be low for the ADC to recognize CONVST and RD inputs. SHDN (Pin 33): Power Shutdown. Drive this pin low with CS low for nap mode. Drive this pin low with CS high for sleep mode. VSS (Pin 34): -5V Negative Supply. Bypass to AGND with 10F tantalum in parallel with 0.1F ceramic. AVDD (Pin 35): 5V Analog Power Supply. Bypass to AGND with 10F tantalum in parallel with 0.1F ceramic. AVDD (Pin 36): 5V Analog Power Supply. Bypass to AGND with 10F tantalum in parallel with 0.1F ceramic and connect this pin to Pin 35 with a 10 resistor. 1604fa 6 LTC1604 FUNCTIONAL BLOCK DIAGRAM 10F 2.2F 10 + 3 VREF 5V 10F + 36 AVDD 5V 35 9 AVDD 10F + 10 DVDD DGND SHDN 33 4 REFCOMP + 7.5k 1.75X 4.375V CS 32 CONTROL LOGIC AND TIMING 2.5V REF P CONTROL LINES CONVST 31 RD 30 BUSY 27 47F OVDD 29 + 1 AIN+ DIFFERENTIAL ANALOG INPUT 2.5V 2 AIN- OGND 28 + 16-BIT SAMPLING ADC - AGND 5 AGND 6 OUTPUT BUFFERS B15 TO B0 AGND 7 16-BIT PARALLEL BUS D15 TO D0 11 TO 26 AGND VSS 8 5V OR 3V 10F 1604 TA01 34 + 10F -5V TEST CIRCUIT Load Circuits for Access Timing Load Circuits for Output Float Delay 5V 5V 1k DN 1k DN 1k CL (A) Hi-Z TO VOH AND VOL TO VOH DN DN 1k CL (B) Hi-Z TO VOL AND VOH TO VOL 1604 TC01 (A) VOH TO Hi-Z CL CL (B) VOL TO Hi-Z 1604 TC02 1604fa 7 LTC1604 APPLICATIONS INFORMATION CONVERSION DETAILS The LTC1604 uses a successive approximation algorithm and internal sample-and-hold circuit to convert an analog signal to a 16-bit parallel output. The ADC is complete with a sample-and-hold, a precision reference and an internal clock. The control logic provides easy interface to microprocessors and DSPs. (Please refer to the Digital Interface section for the data format.) Conversion start is controlled by the CS and CONVST inputs. At the start of the conversion the successive approximation register (SAR) resets. Once a conversion cycle has begun it cannot be restarted. During the conversion, the internal differential 16-bit capacitive DAC output is sequenced by the SAR from the Most Significant Bit (MSB) to the Least Significant Bit (LSB). Referring to Figure 1, the AIN+ and AIN- inputs are acquired during the acquire phase and the comparator offset is nulled by the zeroing switches. In this acquire phase, a duration of 480ns will provide enough time for the sample-and-hold capacitors to acquire the analog signal. During the convert phase the comparator zeroing switches open, putting the comparator into compare mode. The input switches connect the CSMPL capacitors to ground, transferring the differential analog input charge onto the SAMPLE The A/D converter is designed to interface with microprocessors as a memory mapped device. The CS and RD control inputs are common to all peripheral memory interfacing. A separate CONVST is used to initiate a conversion. Internal Clock The A/D converter has an internal clock that runs the A/D conversion. The internal clock is factory trimmed to achieve a typical conversion time of 2.45s and a maximum conversion time of 2.8s over the full temperature range. No external adjustments are required. The guaranteed maximum acquisition time is 480ns. In addition, a throughput time (acquisition + conversion) of 3s and a minimum sampling rate of 333ksps are guaranteed. The LTC1604 operates on 5V supplies, which makes the device easy to interface to 5V digital systems. This device can also talk to 3V digital systems: the digital input pins (SHDN, CS, CONVST and RD) of the LTC1604 recognize 3V or 5V inputs. The LTC1604 has a dedicated output supply pin (OVDD) that controls the output swings of the digital output pins (D0 to D15, BUSY) and allows the part to talk to either 3V or 5V digital systems. The output is two's complement binary. HOLD ZEROING SWITCHES CSMPL AIN- DIGITAL INTERFACE 3V Input/Output Compatible CSMPL AIN+ summing junctions. This input charge is successively compared with the binary-weighted charges supplied by the differential capacitive DAC. Bit decisions are made by the high speed comparator. At the end of a conversion, the differential DAC output balances the AIN+ and AIN- input charges. The SAR contents (a 16-bit data word) which represent the difference of AIN+ and AIN- are loaded into the 16-bit output latches. HOLD SAMPLE HOLD HOLD +CDAC + -CDAC COMP - +VDAC Power Shutdown -VDAC 16 SAR OUTPUT LATCHES t t t D15 D0 1604 F01 Figure 1. Simplified Block Diagram The LTC1604 provides two power shutdown modes, Nap and Sleep, to save power during inactive periods. The Nap mode reduces the power by 95% and leaves only the digital logic and reference powered up. The wake-up time from Nap to active is 200ns. In Sleep mode all bias 1604fa 8 LTC1604 APPLICATIONS INFORMATION currents are shut down and only leakage current remains (about 1A). Wake-up time from Sleep mode is much slower since the reference circuit must power up and settle. Sleep mode wake-up time is dependent on the value of the capacitor connected to the REFCOMP (Pin 4). The wake-up time is 160ms with the recommended 47F capacitor. SHDN t3 CS 1604 F02a Figure 2a. Nap Mode to Sleep Mode Timing Shutdown is controlled by Pin 33 (SHDN). The ADC is in shutdown when SHDN is low. The shutdown mode is selected with Pin 32 (CS). When SHDN is low, CS low selects nap and CS high selects sleep. SHDN t4 CONVST 1604 F02b Figure 2b. SHDN to CONVST Wake-Up Timing Timing and Control Conversion start and data read operations are controlled by three digital inputs: CONVST, CS and RD. A falling edge applied to the CONVST pin will start a conversion after the ADC has been selected (i.e., CS is low). Once initiated, it cannot be restarted until the conversion is complete. Converter status is indicated by the BUSY output. BUSY is low during a conversion. CS t2 CONVST t1 RD We recommend using a narrow logic low or narrow logic high CONVST pulse to start a conversion as shown in Figures 5 and 6. A narrow low or high CONVST pulse prevents the rising edge of the CONVST pulse from upsetting the critical bit decisions during the conversion time. Figure 4 shows the change of the differential nonlinearity error versus the low time of the CONVST pulse. As shown, if CONVST returns high early in the conversion (e.g., CONVST low time <500ns), accuracy is unaffected. Similarly, if CONVST returns high after the conversion is over(e.g., CONVST low time >tCONV), accuracy is unaffected. For best results, keep t5 less than 500ns or greater than tCONV. 1604 F03 Figure 3. CS top CONVST Setup Timing CHANGE IN DNL (LSB) 4 3 2 tCONV tACQ 1 0 0 400 800 1200 1600 2000 2400 2800 CONVST LOW TIME, t5 (ns) 1604 F04 Figure 4. Change in DNL vs CONVST Low Time. Be Sure the CONVST Pulse Returns High Early in the Conversion or After the End of Conversion Figures 5 through 9 show several different modes of operation. In modes 1a and 1b (Figures 5 and 6), CS and RD are both tied low. The falling edge of CONVST starts the conversion. The data outputs are always enabled and data can be latched with the BUSY rising edge. Mode 1a shows operation with a narrow logic low CONVST pulse. Mode 1b shows a narrow logic high CONVST pulse. In mode 2 (Figure 7) CS is tied low. The falling edge of CONVST signal starts the conversion. Data outputs are 1604fa 9 LTC1604 APPLICATIONS INFORMATION t CONV CS = RD = 0 t5 CONVST t6 t8 BUSY t7 DATA DATA (N + 1) D15 TO D0 DATA N D15 TO D0 DATA (N - 1) D15 TO D0 1604 F05 Figure 5. Mode 1a. CONVST Starts a Conversion. Data Outputs Always Enabled (CONVST = ) CS = RD = 0 tCONV t8 t5 t13 CONVST t6 t6 BUSY t7 DATA (N - 1) D15 TO D0 DATA DATA N D15 TO D0 DATA (N + 1) D15 TO D0 1604 F06 Figure 6. Mode 1b. CONVST Starts a Conversion. Data Outputs Always Enabled (CONVST = ) t13 CS = 0 tCONV t5 t8 CONVST t6 BUSY t9 RD t 12 t 11 t 10 DATA DATA N D15 TO D0 1604 F07 Figure 7. Mode 2. CONVST Starts a Conversion. Data is Read by RD 1604fa 10 LTC1604 APPLICATIONS INFORMATION CS = 0 t8 t CONV RD = CONVST t6 t 11 BUSY t 10 t7 DATA (N - 1) D5 TO D0 DATA DATA N D15 TO D0 DATA N D15 TO D0 DATA (N + 1) D15 TO D0 1604 F08 Figure 8. Mode 2. Slow Memory Mode Timing t CONV CS = 0 t8 RD = CONVST t6 t 11 BUSY t 10 DATA DATA N D15 TO D0 DATA (N - 1) D15 TO D0 1604 F09 Figure 9. ROM Mode Timing in three-state until read by the MPU with the RD signal. Mode 2 can be used for operation with a shared data bus. DIFFERENTIAL ANALOG INPUTS In slow memory and ROM modes (Figures 8 and 9) CS is tied low and CONVST and RD are tied together. The MPU starts the conversion and reads the output with the combined CONVST-RD signal. Conversions are started by the MPU or DSP (no external sample clock is needed). Driving the Analog Inputs In slow memory mode the processor applies a logic low to RD (= CONVST), starting the conversion. BUSY goes low, forcing the processor into a wait state. The previous conversion result appears on the data outputs. When the conversion is complete, the new conversion results appear on the data outputs; BUSY goes high, releasing the processor and the processor takes RD (=CONVST) back high and reads the new conversion data. In ROM mode, the processor takes RD (=CONVST) low, starting a conversion and reading the previous conversion result. After the conversion is complete, the processor can read the new result and initiate another conversion. The differential analog inputs of the LTC1604 are easy to drive. The inputs may be driven differentially or as a single-ended input (i.e., the AIN- input is grounded). The AIN+ and AIN- inputs are sampled at the same instant. Any unwanted signal that is common mode to both inputs will be reduced by the common mode rejection of the sample-and-hold circuit. The inputs draw only one small current spike while charging the sample-and-hold capacitors at the end of conversion. During conversion the analog inputs draw only a small leakage current. If the source impedance of the driving circuit is low, then the LTC1604 inputs can be driven directly. As source impedance increases so will acquisition time (see Figure 10). For minimum acquisition time with high source impedance, a buffer amplifier should be used. The only requirement is that the amplifier driving the analog input(s) must settle after the small current spike before the next conversion 1604fa 11 LTC1604 ACQUISITION TIME (s) APPLICATIONS INFORMATION 10 LT(R)1007: Low Noise Precision Amplifier. 2.7mA supply current, 5V to 15V supplies, gain bandwidth product 8MHz, DC applications. 1 LT1097: Low Cost, Low Power Precision Amplifier. 300A supply current, 5V to 15V supplies, gain bandwidth product 0.7MHz, DC applications. 0.1 LT1227: 140MHz Video Current Feedback Amplifier. 10mA supply current, 5V to 15V supplies, low noise and low distortion. 0.01 1 10 100 1k SOURCE RESISTANCE () 10k 1604 F10 LT1360: 37MHz Voltage Feedback Amplifier. 3.8mA supply current, 5V to 15V supplies, good AC/DC specs. Figure 10. tACQ vs Source Resistance LT1363: 50MHz Voltage Feedback Amplifier. 6.3mA supply current, good AC/DC specs. starts (settling time must be 200ns for full throughput rate). LT1364/LT1365: Dual and Quad 50MHz Voltage Feedback Amplifiers. 6.3mA supply current per amplifier, good AC/ DC specs. Choosing an Input Amplifier Choosing an input amplifier is easy if a few requirements are taken into consideration. First, to limit the magnitude of the voltage spike seen by the amplifier from charging the sampling capacitor, choose an amplifier that has a low output impedance (<100) at the closed-loop band-width frequency. For example, if an amplifier is used in a gain of +1 and has a unity-gain bandwidth of 50MHz, then the output impedance at 50MHz should be less than 100. The second requirement is that the closed-loop bandwidth must be greater than 15MHz to ensure adequate smallsignal settling for full throughput rate. If slower op amps are used, more settling time can be provided by increasing the time between conversions. The best choice for an op amp to drive the LTC1604 will depend on the application. Generally applications fall into two categories: AC applications where dynamic specifications are most critical and time domain applications where DC accuracy and settling time are most critical. The following list is a summary of the op amps that are suitable for driving the LTC1604. More detailed information is available in the Linear Technology databooks, the LinearViewTM CD-ROM and on our web site at: www.linear-tech. com. Input Filtering The noise and the distortion of the input amplifier and other circuitry must be considered since they will add to the LTC1604 noise and distortion. The small-signal bandwidth of the sample-and-hold circuit is 15MHz. Any noise or distortion products that are present at the analog inputs will be summed over this entire bandwidth. Noisy input circuitry should be filtered prior to the analog inputs to minimize noise. A simple 1-pole RC filter is sufficient for many applications. For example, Figure 11 shows a 3000pF capacitor from AIN+ to ground and a 100 source resistor to limit the input bandwidth to 530kHz. The 3000pF capacitor also acts as a charge reservoir for the input sample-and-hold and isolates the ADC input from sampling glitch sensitive circuitry. High quality capacitors and resistors should be used since these components can add distortion. NPO and silver mica type dielectric capacitors have excellent linearity. Carbon surface mount resistors can also generate distortion from self heating and from damage that may occur during soldering. Metal film surface mount resistors are much less susceptible to both problems. LinearView is a trademark of Linear Technology Corporation. 1604fa 12 LTC1604 APPLICATIONS INFORMATION 100 ANALOG INPUT 3000pF 1 AIN+ 2.500V 2 3 4 R1 7.5k 3 VREF AIN- LTC1604 VREF 4.375V 4 REFCOMP REFERENCE AMP REFCOMP R2 12k 47F 47F 5 BANDGAP REFERENCE AGND 1604 F11 R3 16k 5 AGND LTC1604 Figure 11. RC Input Filter 1604 F12a Figure 12a. LTC1604 Reference Circuit Input Range The 2.5V input range of the LTC1604 is optimized for low noise and low distortion. Most op amps also perform well over this same range, allowing direct coupling to the analog inputs and eliminating the need for special translation circuitry. Some applications may require other input ranges. The LTC1604 differential inputs and reference circuitry can accommodate other input ranges often with little or no additional circuitry. The following sections describe the reference and input circuitry and how they affect the input range. Internal Reference The LTC1604 has an on-chip, temperature compensated, curvature corrected, bandgap reference that is factory trimmed to 2.500V. It is connected internally to a reference amplifier and is available at VREF (Pin 3) (see Figure 12a). A 7.5k resistor is in series with the output so that it can be easily overdriven by an external reference or other circuitry (see Figure 12b). The reference amplifier gains the voltage at the VREF pin by 1.75 to create the required internal reference voltage. This provides buffering between the VREF pin and the high speed capacitive DAC. The reference amplifier compensation pin (REFCOMP, Pin 4) must be bypassed with a capacitor to ground. The reference amplifier is stable with capacitors of 22F or greater. For the best noise performance a 47F ceramic or 47F tantalum in parallel with a 0.1F ceramic is recommended. 5V ANALOG INPUT VIN LT1019A-2.5 VOUT 1 AIN+ 2 AIN- 3 VREF LTC1604 4 + 10F 0.1F 5 REFCOMP AGND 1604 F12b Figure 12b. Using the LT1019-2.5 as an External Reference The VREF pin can be driven with a DAC or other means shown in Figure 13. This is useful in applications where the peak input signal amplitude may vary. The input span of the ADC can then be adjusted to match the peak input signal, maximizing the signal-to-noise ratio. The filtering of the internal LTC1604 reference amplifier will limit the bandwidth and settling time of this circuit. A settling time of 20ms should be allowed for after a reference adjustment. Differential Inputs The LTC1604 has a unique differential sample-and-hold circuit that allows rail-to-rail inputs. The ADC will always convert the difference of AIN+ - AIN- independent of the common mode voltage (see Figure 15a). The common mode rejection holds up to extremely high frequencies (see Figure 14a). The only requirement is that both inputs 1604fa 13 LTC1604 APPLICATIONS INFORMATION 1 ANALOG INPUT 2V TO 2.7V DIFFERENTIAL 2 AIN+ ANALOG INPUT AIN- 0V TO 5V 2.5V + LTC1604 LTC1450 2V TO 2.7V 3 4 1 AIN+ 2 AIN- 3 VREF - VREF LTC1604 4 REFCOMP REFCOMP 10F 47F 5 5 AGND AGND 1604 F14b 1604 F13 Figure 14b. Selectable 0V to 5V or 2.5V Input Range 80 Full-Scale and Offset Adjustment 70 Figure 15a shows the ideal input/output characteristics for the LTC1604. The code transitions occur midway between successive integer LSB values (i.e., -FS + 0.5LSB, -FS + 1.5LSB, -FS + 2.5LSB,... FS - 1.5LSB, FS - 0.5LSB). The output is two's complement binary with 1LSB = FS - (-FS)/65536 = 5V/65536 = 76.3V. 60 50 40 30 20 10 0 1k 10k 100k INPUT FREQUENCY (Hz) 1M 1604 G14a Figure 14a. CMRR vs Input Frequency can not exceed the AVDD or VSS power supply voltages. Integral nonlinearity errors (INL) and differential nonlinearity errors (DNL) are independent of the common mode voltage, however, the bipolar zero error (BZE) will vary. The change in BZE is typically less than 0.1% of the common mode voltage. Dynamic performance is also affected by the common mode voltage. THD will degrade as the inputs approach either power supply rail, from 96dB with a common mode of 0V to 86dB with a common mode of 2.5V or -2.5V. Differential inputs allow greater flexibility for accepting different input ranges. Figure 14b shows a circuit that converts a 0V to 5V analog input signal with only an additional buffer that is not in the signal path. In applications where absolute accuracy is important, offset and full-scale errors can be adjusted to zero. Offset error must be adjusted before full-scale error. Figure 15b shows the extra components required for full-scale error adjustment. Zero offset is achieved by adjusting the offset applied to the AIN- input. For zero offset error apply 011...111 011...110 OUTPUT CODE COMMON MODE REJECTION (dB) Figure 13. Driving VREF with a DAC 000...001 000...000 111...111 111...110 100...001 100...000 - (FS - 1LSB) FS - 1LSB INPUT VOLTAGE (AIN+ - AIN- ) 1604 F15a Figure 15a. LTC1604 Transfer Characteristics 1604fa 14 LTC1604 APPLICATIONS INFORMATION -5V ANALOG INPUT R3 24k R8 50k 1 AIN+ 2 AIN- R4 100 3 R5 R7 47k 50k R6 24k + 47F 0.1F 4 5 LTC1604 VREF REFCOMP AGND 1604 F15b Figure 15b. Offset and Full-Scale Adjust Circuit -38V (i.e., -0.5LSB) at AIN+ and adjust the offset at the AIN- input until the output code flickers between 0000 0000 0000 0000 and 1111 1111 1111 1111. For full-scale adjustment, an input voltage of 2.499886V (FS/2 - 1.5LSBs) is applied to AIN+ and R2 is adjusted until the output code flickers between 0111 1111 1111 1110 and 0111 1111 1111 1111. BOARD LAYOUT AND GROUNDING Wire wrap boards are not recommended for high resolution or high speed A/D converters. To obtain the best performance from the LTC1604, a printed circuit board with ground plane is required. Layout should ensure that digital and analog signal lines are separated as much as possible. Particular care should be taken not to run any digital track alongside an analog signal track or underneath the ADC. The analog input should be screened by AGND. An analog ground plane separate from the logic system ground should be established under and around the ADC. Pin 5 to Pin 8 (AGNDs), Pin 10 (ADC's DGND) and all other analog grounds should be connected to this single analog ground point. The REFCOMP bypass capacitor and the DVDD bypass capacitor should also be connected to this analog ground plane. No other digital grounds should be connected to this analog ground plane. Low impedance analog and digital power supply common returns are essential to low noise operation of the ADC and the foil width for these tracks should be as wide as possible. In applications where the ADC data outputs and control signals are connected to a continuously active microprocessor bus, it is possible to get errors in the conversion results. These errors are due to feedthrough from the microprocessor to the successive approximation comparator. The problem can be eliminated by forcing the microprocessor into a WAIT state during conversion or by using three-state buffers to isolate the ADC data bus. The traces connecting the pins and bypass capacitors must be kept short and should be made as wide as possible. The LTC1604 has differential inputs to minimize noise coupling. Common mode noise on the AIN+ and AIN- leads will be rejected by the input CMRR. The AIN- input can be used as a ground sense for the AIN+ input; the LTC1604 will hold and convert the difference voltage between AIN+ and AIN-. The leads to AIN+ (Pin 1) and AIN- (Pin 2) should be kept as short as possible. In applications where this is not possible, the AIN+ and AIN- traces should be run side by side to equalize coupling. SUPPLY BYPASSING High quality, low series resistance ceramic, 10F or 47F bypass capacitors should be used at the VDD and REFCOMP pins as shown in Figure 16 and in the Typical Application on the first page of this data sheet. Surface mount ceramic capacitors such as Murata GRM235Y5V106Z016 provide excellent bypassing in a small board space. Alternatively, 10F tantalum capacitors in parallel with 0.1F ceramic capacitors can be used. Bypass capacitors must be located as close to the pins as possible. The traces connecting the pins and the bypass capacitors must be kept short and should be made as wide as possible. 1604fa 15 LTC1604 APPLICATIONS INFORMATION 1 ANALOG INPUT CIRCUITRY + - AIN+ AIN- 2 DIGITAL SYSTEM LTC1604 VREF REFCOMP AGND 3 4 2.2F 47F 5 TO 8 VSS AVDD AVDD DVDD DGND OVDD OGND 34 36 35 9 10F 10F 10F 10F 10 29 28 10F 1604 F16 Figure 16. Power Supply Grounding Practice DC PERFORMANCE 2000 COUNT 1500 1000 500 0 -5 -4 -3 -2 -1 0 1 CODE 2 3 4 5 1604 F17 Figure 17. Histogram for 4096 Conversions 0 fSAMPLE = 333kHz fIN = 4.959kHz SINAD = 90.2dB THD = -103.2dB -20 AMPLITUDE (dB) The noise of an ADC can be evaluated in two ways: signalto-noise raio (SNR) in frequency domain and histogram in time domain. The LTC1604 excels in both. Figure 18a demonstrates that the LTC1604 has an SNR of over 90dB in frequency domain. The noise in the time domain histogram is the transition noise associated with a high resolution ADC which can be measured with a fixed DC signal applied to the input of the ADC. The resulting output codes are collected over a large number of conversions. The shape of the distribution of codes will give an indication of the magnitude of the transition noise. In Figure 17 the distribution of output codes is shown for a DC input that has been digitized 4096 times. The distribution is Gaussian and the RMS code transition noise is about 0.66LSB. This corresponds to a noise level of 90.9dB relative to full scale. Adding to that the theoretical 98dB of quantization error for 16-bit ADC, the resultant corresponds to an SNR level of 90.1dB which correlates very well to the frequency domain measurements in DYNAMIC PERFORMANCE section. 2500 -40 -60 -80 -100 DYNAMIC PERFORMANCE The LTC1604 has excellent high speed sampling capability. Fast fourier transform (FFT) test techniques are used to test the ADC's frequency response, distortions and noise at the rated throughput. By applying a low distortion sine wave and analyzing the digital output using an FFT algorithm, the ADC's spectral content can be examined for frequencies outside the fundamental. Figures 18a and 18b show typical LTC1604 FFT plots. -120 -140 0 20 40 60 80 100 120 140 160 FREQUENCY (kHz) 1604 F18a Figure 18a. This FFT of the LTC1604's Conversion of a Full-Scale 5kHz Sine Wave Shows Outstanding Response with a Very Low Noise Floor When Sampling at 333ksps 1604fa 16 LTC1604 APPLICATIONS INFORMATION 0 Signal-to-Noise Ratio -40 AMPLITUDE (dB) The signal-to-noise plus distortion ratio [S/(N + D)] is the ratio between the RMS amplitude of the fundamental input frequency to the RMS amplitude of all other frequency components at the A/D output. The output is band limited to frequencies from above DC and below half the sampling frequency. Figure 18a shows a typical spectral content with a 333kHz sampling rate and a 5kHz input. The dynamic performance is excellent for input frequencies up to and beyond the Nyquist limit of 167kHz. fSAMPLE = 333kHz fIN = 97.152kHz SINAD = 89dB THD = -96dB -20 -60 -80 -100 -120 -140 0 20 40 60 80 100 120 140 160 FREQUENCY (kHz) Effective Number of Bits 1604 F18b The effective number of bits (ENOBs) is a measurement of the resolution of an ADC and is directly related to the S/(N + D) by the equation: Figure 18b. Even with Inputs at 100kHz, the LTC1604's Dynamic Linearity Remains Robust Total Harmonic Distortion Total harmonic distortion (THD) is the ratio of the RMS sum of all harmonics of the input signal to the fundamental itself. The out-of-band harmonics alias into the frequency band between DC and half the sampling frequency. THD is expressed as: 2 2 2 2 V2 + V3 + V4 +...Vn V1 where V1 is the RMS amplitude of the fundamental frequency and V2 through Vn are the amplitudes of the second through nth harmonics. THD vs Input Frequency is shown in Figure 20. The LTC1604 has good distortion performance up to the Nyquist frequency and beyond. 16 98 15 92 14 86 13 80 12 74 11 68 10 62 9 56 8 1k 10k 100k FREQUENCY (Hz) SINAD (dB) where N is the effective number of bits of resolution and S/(N + D) is expressed in dB. At the maximum sampling rate of 333kHz the LTC1604 maintains above 14 bits up to the Nyquist input frequency of 167kHz (refer to Figure 19). EFFECTIVE BITS N = [S/(N + D) - 1.76]/6.02 50 1M 1604 F19 Figure 19. Effective Bits and Signal/(Noise + Distortion) vs Input Frequency AMPLITUDE (dB BELOW THE FUNDAMENTAL) THD= 20Log 0 -10 -20 -30 -40 -50 -60 -70 -80 THD 3RD -90 -100 -110 1k 2ND 10k 100k INPUT FREQUENCY (Hz) 1M 1604 G04 Figure 20. Distortion vs Input Frequency 1604fa 17 LTC1604 APPLICATIONS INFORMATION Intermodulation Distortion If the ADC input signal consists of more than one spectral component, the ADC transfer function nonlinearity can produce intermodulation distortion (IMD) in addition to THD. IMD is the change in one sinusoidal input caused by the presence of another sinusoidal input at a different frequency. If two pure sine waves of frequencies fa and fb are applied to the ADC input, nonlinearities in the ADC transfer function can create distortion products at the sum and difference frequencies of mfa nfb, where m and n = 0, 1, 2, 3, 0 fSAMPLE = 333kHz fIN1 = 29.3kHz fIN2 = 32.4kHz -20 AMPLITUDE (dB) -40 -60 -80 -100 -120 -140 0 20 40 60 80 100 120 140 160 FREQUENCY (kHz) 1604 G06 Figure 21. Intermodulation Distortion Plot etc. For example, the 2nd order IMD terms include (fa - fb). If the two input sine waves are equal in magnitude, the value (in decibels) of the 2nd order IMD products can be expressed by the following formula: IMD ( fa fb) = 20Log Amplitude at (fa fb) Amplitude at fa Peak Harmonic or Spurious Noise The peak harmonic or spurious noise is the largest spectral component excluding the input signal and DC. This value is expressed in decibels relative to the RMS value of a full-scale input signal. Full-Power and Full-Linear Bandwidth The full-power bandwidth is that input frequency at which the amplitude of the reconstructed fundamental is reduced by 3dB for a full-scale input signal. The full-linear bandwidth is the input frequency at which the S/(N + D) has dropped to 84dB (13.66 effective bits). The LTC1604 has been designed to optimize input bandwidth, allowing the ADC to undersample input signals with frequencies above the converter's Nyquist Frequency. The noise floor stays very low at high frequencies; S/(N + D) becomes dominated by distortion at frequencies far beyond Nyquist. 1604fa 18 LTC1604 PACKAGE DESCRIPTION Dimensions in inches (millimeters) unless otherwise noted. G Package 36-Lead Plastic SSOP (0.209) (LTC DWG # 05-08-1640) 0.499 - 0.509* (12.67 - 12.93) 36 35 34 33 32 31 30 29 28 27 26 25 24 23 22 21 20 19 0.301 - 0.311 (7.65 - 7.90) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 0.205 - 0.212** (5.20 - 5.38) 0.068 - 0.078 (1.73 - 1.99) 0 - 8 0.005 - 0.009 (0.13 - 0.22) 0.022 - 0.037 (0.55 - 0.95) *DIMENSIONS DO NOT INCLUDE MOLD FLASH. MOLD FLASH SHALL NOT EXCEED 0.006" (0.152mm) PER SIDE **DIMENSIONS DO NOT INCLUDE INTERLEAD FLASH. INTERLEAD FLASH SHALL NOT EXCEED 0.010" (0.254mm) PER SIDE 0.0256 (0.65) BSC 0.010 - 0.015 (0.25 - 0.38) 0.002 - 0.008 (0.05 - 0.21) G36 SSOP 1196 1604fa Information furnished by Linear Technology Corporation is believed to be accurate and reliable. However, no responsibility is assumed for its use. Linear Technology Corporation makes no representation that the interconnection of its circuits as described herein will not infringe on existing patent rights. 19 LTC1604 TYPICAL APPLICATION Using the LTC1604 and Two LTC1391s as an 8-Channel Differential 16-Bit ADC System 5V 1 2 3 4 5 6 7 CH7 + 8 CH0 V+ D CH1 CH2 V DOUT CH4 DIN CH5 CS CH6 CLK CH7 GND 2 3 4 5 6 7 CH7 - 8 V+ CH1 D CH2 V- CH3 DOUT CH4 DIN CH5 CS CH6 CLK CH7 GND 10 9 DVDD AVDD DGND SHDN 33 + 13 1F 4 REFCOMP 12 + 4.375V 47F 11 7.5k 1.75X 10 2.5V REF CS 32 P CONTROL LINES CONVST 31 RD 30 BUSY 27 OVDD 29 LTC1604 1 AIN 9 2 AIN- + OGND 28 + - 16-BIT SAMPLING ADC OUTPUT BUFFERS B15 TO B0 D15 TO D0 3000pF 1F 16 15 AGND AGND AGND 5 6 7 8 P CONTROL LINES CS 10 1604 TA03 10F -5V DIN 11 16-BIT PARALLEL BUS 34 + + 12 5V OR 3V 10F 11 TO 26 AGND VSS -5V 14 13 CONTROL LOGIC AND TIMING + LTC1391 CH0 + 35 -5V 5V 1 5V 10F 15 3000pF CH0 - AVDD 10F + 36 VREF - 14 CH3 5V 10 3 1F 16 10F + + LTC1391 CH0+ 2.2F CLK 9 RELATED PARTS SAMPLING ADCs PART NUMBER DESCRIPTION COMMENTS LTC1410 12-Bit, 1.25Msps, 5V ADC 71.5dB SINAD at Nyquist, 150mW Dissipation LTC1415 12-Bit, 1.25Msps, Single 5V ADC 55mW Power Dissipation, 72dB SINAD LTC1418 14-Bit, 200ksps, Single 5V ADC 15mW, Serial/Parallel 10V LTC1419 Low Power 14-Bit, 800ksps ADC True 14-Bit Linearity, 81.5dB SINAD, 150mW Dissipation LTC1605 16-Bit, 100ksps, Single 5V ADC 10V Inputs, 55mW, Byte or Parallel I/O PART NUMBER DESCRIPTION COMMENTS LTC1595 16-Bit Serial Multiplying IOUT DAC in SO-8 1LSB Max INL/DNL, Low Glitch, DAC8043 16-Bit Upgrade LTC1596 16-Bit Serial Multiplying IOUT DAC 1LSB Max INL/DNL, Low Glitch, AD7543/DAC8143 16-Bit Upgrade LTC1597 16-Bit Parallel, Multiplying DAC 1LSB Max INL/DNL, Low Glitch, 4 Quadrant Resistors LTC1650 16-Bit Serial VOUT DAC Low Power, Low Gritch, 4-Quadrant Multiplication DACs 1604fa 20 Linear Technology Corporation 1604a LT/TP 1098 REV A 2K * PRINTED IN USA 1630 McCarthy Blvd., Milpitas, CA 95035-7417 (408) 432-1900 FAX: (408) 434-0507 www.linear.com (c) LINEAR TECHNOLOGY CORPORATION 1998