ADC100 Precision 22 Bit Integrating A/D Converter THALER CORPORATION • 2015 N. FORBES BOULEVARD • TUCSON, AZ. 85745 • (520) 882-4000 FEATURES · 22-BIT RESOLUTION · ±10.48 INPUT RANGE · 1ppm/°C MAX. SCALE FACTOR ERROR APPLICATIONS · TEST EQUIPMENT · DATA ACQUISITION · SCIENTIFIC INSTRUMENTS · 2 ppm MAX. LINEARITY ERROR · MEDICAL INSTRUMENTS · AUTO ZERO · SEISMOLOGICAL EQUIPMENT · BUS COMPATIBLE · ROBOTIC SYSTEMS · INTERNAL CLOCK and REFERENCE · LOW POWER CONSUMPTION (0.4 WATTS) · WEIGHING SYSTEMS DESCRIPTION ADC100 is a high performance 22-bit A/D converter based on a patented architecture which provides outstanding performance (accuracy) comparable to the best digital meters. The ADC100 is available in two operating temperature ranges, -25°C to +85°C and -55°C to +125°C. "M" versions are screened for high reliability and quality. ADC100 offers 3 ppm max. linearity error and 1ppm/°C max. scale factor error over the military temperature range. It also has excellent offset stability at 2 ppm max. which the user can auto zero if desired. Type Temperature Operating Range Max. Scale Factor Deviation ADC100C -25°C to +85°C 60ppm ADC100CA -25°C to +85°C 30ppm ADC100M -55°C to +125°C 100ppm ADC100's compatibility with popular microcomputer buses increases its ease of application in smart systems. An on-board microprocessor controls all internal functions of the ADC100. Thaler designers have minimized external connections to greatly reduce the problem often encountered when applying ADC's. Operating from ±15VDC and a +5VDC power supply, ADC100 is packaged in a hermetically sealed 40pin ceramic DIP package. Precision test equipment, scientific and medical instruments, and data acquisition systems are primary application areas for the unusually high resolution and accuracy of this ADC. ADC100DS REV. E MAR 00 MAXIMUM RATING SPECIFICATIONS ADC100 ADC100 MODEL PARAMETER MIN MAX UNITS TEMPERATURE Operating Storage POWER SUPPLY -55 0 125 160 °C °C +14 -14 +4 +16 -16 +6 VDC VDC VDC VEE 0 VCC VDD VCC VEE VDD INPUTS Analog Inputs Digital Inputs NOTES: EXTERNAL CONNECTIONS (TOP VIEW) N.C. 1 40 ANALOG LOW N.C. 2 39 ANALOG HIGH N.C. 3 38 N.C. Vee (-15V) 4 37 N.C. Vee (+15V) 5 36 N.C. Vdd (+5V) 6 35 GND 7 34 N.C. 8 33 N.C. N.C. 9 32 N.C. N.C. 10 31 N.C. N.C. 11 30 N.C. N.C. 12 29 D0 13 28 AUTO ZERO RESET N.C. D1 14 27 N.C. D2 15 26 N.C. D3 16 25 N.C. D4 17 24 STATUS 1 D5 18 23 STATUS 0 D6 19 22 CONVERT D7 20 21 OUTPUT ENABLE ADC100 INTEGRATION CAPACITOR 1. Power Supply Decoupling The ADC100 has internal 0.1µF decoupling capacitors for all power supply inputs. The internal decoupling capacitors are adequate for applications with relatively short power supply leads (approx. 5") or if additional capacitors are located on a circuit board. For applications with long power supply leads an external capacitor of 10 mF on the +/- 15V inputs and 33 mF on the +5V input is recommended. 2. Ground The ground connection (pin 7) should be made as solid as possible since ground noise can result in a loss of accuracy. Use of a ground plane is a good approach to maintain the full accuracy of the ADC100. 3. External Components A 0.68 µF polystyrene integration capacitor must be connected to pins 34 and 35 with a lead length not exceeding 2". 4. Analog Inputs In order to avoid differential noise pickup it is recommended to use parallel adjacent lines for the analog inputs (pins 39, 40) on PC boards and shielded lines outside of the PC connections. ADC100DS REV. E MAR 00 ELECTRICAL SPECIFICATIONS ADC100 (Vps = +/- 15V, + 5V, T = 25 Deg. C.) ADC100CA ADC100C MODEL PARAMETER MIN TYP MAX MIN TYP ADC100M MAX MIN TYP MAX ACCURACY 22 1 Resolution Input Equivalent Noise Offset without Auto Zero Offset with Auto Zero Full Scale Noise (.1-10Hz) @ 10V Nonlinearity Normal Mode Rejection 1 * * 4 1 100 * * 2 0.5 50 6 * * * * 3 60 * 2 * * * Bits µV ppm ppm ppm µVpp ppm dB TEMPERATURE STABILITY 0.2 1.0 Offset Full Scale 0.1 0.5 * * ppm/o C ppm/o C TIME STABILITY .1 2 Offset Full Scale 2 * * * * ppm/month ppm/24 hrs. ERROR ALL SOURCES 24 hrs, +/- 1 Deg. C Amb. 90 days, +/- 5 Deg. C Amb. 1 year, +/- 5 Deg. C Amb. CONVERSION TIME WARM-UP TIME .0007, 2 .0010, 2 .0015, 2 .0005, 2 .0008, 2 .0013, 2 * * * %, +/- Counts %, +/- Counts %, +/- Counts 320 * * ms 5 * * minutes POWER SUPPLY REJECTION 80 80 +/- 15 VDC 5 VDC * * * * dB dB ANALOG INPUT CHARACTERISTICS Input Range Bias Current Input Impedance -10.485760 1.2 200 10.485755 3 * 15.5 15.5 5.5 * * * * * * * * * * * * * * V nA GΩ * * * * * * V V V * * * * * POWER SUPPLY VOLTAGES +15 V -15 V 5v 14.5 14.5 4.5 POWER SUPPLY CURRENTS +15 V -15 V 5v 15 15 5 23 24 42 * * * * * * mA mA mA DIGITAL INPUTS Low High 0.8 4.0 * * * V V * V V * V V * V V 125 oC * DIGITAL OUTPUTS Low High 0.8 4.0 * * * AUTO ZERO INPUT Low High * 0.8 4.0 * * CONVERT INPUT Low High TEMPERATURE RANGE 0.8 4.0 -25 * Same as ADC100C Note: 1) 60 Cycle 2) ( Max-Min Value) - Noise(.1-10Hz) * * * 85 * * -55 ADC100DS REV. E MAR 00 THEORY OF OPERATION The timing control circuitry governs the counters that measure the integration time in both directions. The ADC100's on-board microprocessor is used to calculate the results of the integration equation above. It is also used to perform error corrections and to control the built-in-auto-zero function. Note that the mP automatically performs an auto-zero function at start-up, but it is recommended, to achieve maximum accuracy, that an auto-zero be performed again after the ADC100 is fully warmed up. When the µP detects a convert signal, it lowers the status lines to indicate that the ADC is involved in a conversion. When it detects a change in slope direction, the µP will collect the counts for the integration time. When sufficient counts have been collected, the µP performs the calculations described above. When the calculations are complete, the µmP places the most significant byte in the output buffer and raises the S0 flag. When another pulse is placed on the convert line, the middle byte is placed on the output, the S0 flag is lowered and the S1 flag raised. When the last pulse is placed in the convert line, the least significant byte is placed in the output buffer and both status flags are high indicating that the ADC100 is ready for another conversion. Status line summary: In the ADC100 block diagram (see Figure 1), Vhi and Vlow are the inputs. Both are buffered and fed into a differential, voltage controlled, single output current source. This current is added to the reference current at the input of the op amp integrator. The output of the integrator is fed into a Schmitt trigger, which in turn, is fed into the ADC's timing control circuitry. When the integrator output actuates the Schmitt trigger, the timing circuit changes the direction of the reference current source and the integrator begins integrating in the opposite direction. This continues until the Schmitt trigger is actuated again by the integrator and reverses the direction of the reference current. The equation for integration times are: Tp= VXC I ref + I inp Tm= VXC -I ref + I inp V = Voltage C= Integration Capacitor Value I ref = Reference Current I inp = Input Current Resolving these equations produces: I inp = I ref Tp - Tm Tp + Tm Tp = Time Positive Tm = Time Negative S1 S0 0 0 1 1 0 1 0 1 Conversion in progress. Conversion complete. MSB in output. Middle byte in output register. LSB in output. Ready for next conversion. Vhi Auto Zero Switch Differential Voltage Controlled Current Source Schmitt Trigger Vlow +15V Bidirectional Reference Current Source Data Output ï ï Output Buffer Output Enable FIGURE 1. BLOCK DIAGRAM Microprocessor Auto Zero Convert Current Directional Switch ï Timing Control and Counter -15V Clock Status Lines ADC100DS REV. E MAR 00 CONNECTING THE ADC100 POWER SUPPLIES The power supply lines are connected to pins 4-7. Pin 4 is -15V, pin 5 is +15v, pin 6 is +5V and pin 7 is GND. OUTPUT DATA LINES The output data is available in byte form on pins 13-20. Pin 20 is the Most Significant Bit and pin 13 the Least Significant Bit. The data lines go to a high impedance state when the Output Enable line is at a logic one level. OUTPUT ENABLE (PIN 21) Data is placed on the Output Data Lines by a logic zero on this line. CONVERT (Pin22) This line is used to initiate a conversion cycle and to retrieve the output data. The status lines indicate which function will be executed. The first pulse (transition from logic one to logic zero) starts the conversion cycle. Two subsequent pulses are used to place the lower two bytes on the Output Data Lines. AUTO-ZERO / RESET (Pin 29) A logic zero on this input will autozero the ADC1503 by internally connecting the analog high to analog low. Since the µP is reset the status lines S1 and S0 are tristate before going to the low position. The status lines will remain low until the autozero is complete. INTEGRATION CAPACITOR (Pin 34, 35) A .68 µF polystyrene capacitor must be connected to these pins. Lead length should be as short as possible and not exceed 2". ANALOG INPUTS (Pin 39, 40) Both analog inputs are buffered by op-amps and have a common mode rejection of approximately 80dB. min. To maintain the full accuracy at the ADC it is recommended to keep the input to analog common to less than 0.1VDC. STATUS LINES (Pins 23, 24) These lines indicate the present state of the ADC. When the Convert line receives the first pulse in a conversion cycle the Status Lines go to logic zero, indicating that a conversion cycle is in progress. When the conversion is complete the microprocessor places the MSB of the output data in the output buffer and then raises S0 to a logic one, indicating that the MSB at the output data is available in the output buffer. When the Convert Line is pulsed again the middle byte of the output data is placed in that output buffer and S1 changes to logic one and S0 to logic zero. The third pulse places the LSB of the output data in the buffer and both status lines go to the logic one. The converter is now ready for the next conversion cycle. The table below shows a summary of the status code. S1 S0 Conversion in progress. 0 0 Conversion complete. MSB in output. 0 1 Middle byte in output register. 1 0 LSB in output. Ready for next conversion. 1 1 OUTPUT DATA REPRESENTATION The output data is represented in BOB (Bipolar Offset Binary) format. One LSB is scaled to be exactly 5mV. The table below shows the output data codes for zero and plus-minus full scale input voltage. Input Voltage -10.485760 V 0.0 V +10.485755 V High Byte Output Data Middle Byte Low Byte 00 20 3F 00 00 FF 00 00 FF ADC100DS REV. E MAR 00 TIMING DIAGRAMS CONVERT → → AZ tAZD S1 S0 tTRST → → tAZ Symbol Parameter Min. Typ. Max. Unit tAZD AZ Pulse Width 0.2 tTRST Tristate Time 30 ms tAZ AZ Time 400 ms µs FIGURE 2. AUTO ZERO → → CONVERT tCONZ S1 → tSZ → S0 tCONV Symbol Parameter Min. Typ. Max. Unit tCONZ Convert Pulse 5.0 tSZ Status Delay 8.0 µs tCONV Convert Time 320 ms µs FIGURE 3. CONVERSION ADC100DS REV. E MAR 00 TIMING DIAGRAMS OE tSIR → → tOEDV → MIB LSB → CONVERT MSB → D0 - D7 → tSIR S1 S0 Symbol Parameter tOEDV OE Delay tSIR Status Delay Min. Typ. 3.0 Max. Unit 45 ns µs 40-PIN HYBRID PACKAGE FIGURE 4. DATA OUTPUT INCHES DIM MIN MAX E 1.080 1.100 D 2.075 2.115 A 0.155 0.185 L 0.220 0.240 B2 .100 typ B .018 typ Q .015 .035 C .009 .012 P .012 .018 G1 .890 .910 B1 .040 typ NOTES: 1. GOLD PLATING 60 MICRO INCHES MINIMUM THICKNESS OVER 100 MICRO INCHES NOMINAL THICKNESS OF NICKEL FIGURE 5. MECHANICAL SPECIFICATIONS ADC100DS REV. E MAR 00