Femtoampere Input Bias Current Electrometer Amplifier ADA4530-1 Data Sheet FEATURES PIN CONNECTION DIAGRAM +IN 1 8 -IN GRD 2 7 GRD IC 3 6 OUT V- 4 5 V+ NOTES 1. IC = INTERNAL CONNECTION. THIS PIN MUST BE CONNECTED TO V- OR LEFT UNCONNECTED. 13405-001 ADA4530-1 Low input bias current 20 fA maximum at TA = 25C (guaranteed at production test) 20 fA maximum at -40C < TA < +85C 250 fA maximum at -40C < TA < +125C (guaranteed at production test) Low offset voltage: 50 V maximum over specified CMRR range Offset voltage drift: 0.13 V/C typical, 0.5 V/C maximum Integrated guard buffer with 100 V maximum offset Low voltage noise density: 14 nV/Hz at 10 kHz Wide bandwidth: 2 MHz unity-gain crossover Supply voltage: 4.5 V to 16 V (2.25 V to 8 V) Operating temperature: -40C to +125C Long-term offset voltage drift (10,000 hours): 0.5 V typical Temperature hysteresis: 1.5 V typical Figure 1. APPLICATIONS Laboratory and analytical instrumentation: spectrophotometers, chromatographs, mass spectrometers, and potentiostatic and amperostatic coulometry Instrumentation: picoammeters and coulombmeters Transimpedance amplifier (TIA) for photodiodes, ion chambers, and working electrode measurements High impedance buffering for chemical sensors and capacitive sensors GENERAL DESCRIPTION The ADA4530-1 also offers low offset voltage, low offset drift, and low voltage and current noise needed for the types of applications that require such low leakages. Rev. B The ADA4530-1 operates over the -40C to +125C industrial temperature range and is available in an 8-lead SOIC package. 1000 -40C TO +125C LIMIT VSY = 10V VCM = VSY/2 RH < 10% 100 -40C TO +85C LIMIT 10 1 0.1 IB+ IB- 0.01 0.001 0 10 20 30 40 50 60 70 80 90 100 110 120 130 TEMPERATURE (C) 13405-202 It provides ultralow input bias currents that are production tested at 25C and at 125C to ensure the device meets its performance goals in user systems. The integrated guard buffer isolates the input pins from leakage in the printed circuit board (PCB), minimizes board component count, and enables easy system design. The ADA4530-1 is available in an industrystandard surface-mount 8-lead SOIC package with a unique pinout optimized to prevent signals from coupling between the sensitive input pins, the power supplies, and the output pin while enabling easy routing of the guard ring traces. To maximize the dynamic range of the system, the ADA4530-1 has a rail-to-rail output stage that can typically drive to within 30 mV of the supply rails under a 10 k load. IB (fA) The ADA4530-1 is a femtoampere (10-15 A) level input bias current operational amplifier suitable for use as an electrometer that also includes an integrated guard buffer. It has an operating voltage range of 4.5 V to 16 V, enabling it to operate in conventional 5 V and 10 V single supply systems as well as 2.5 V and 5 V dual supply systems. Figure 2. Input Bias Current (IB) vs. Temperature, VSY = 10 V Document Feedback Information furnished by Analog Devices is believed to be accurate and reliable. However, no responsibility is assumed by Analog Devices for its use, nor for any infringements of patents or other rights of third parties that may result from its use. Specifications subject to change without notice. No license is granted by implication or otherwise under any patent or patent rights of Analog Devices. Trademarks and registered trademarks are the property of their respective owners. One Technology Way, P.O. Box 9106, Norwood, MA 02062-9106, U.S.A. Tel: 781.329.4700 (c)2015-2017 Analog Devices, Inc. All rights reserved. Technical Support www.analog.com ADA4530-1 Data Sheet TABLE OF CONTENTS Features .............................................................................................. 1 Input Resistance.......................................................................... 34 Applications ....................................................................................... 1 Input Offset Voltage ................................................................... 34 Pin Connection Diagram ................................................................ 1 Insulation Resistance ................................................................. 34 General Description ......................................................................... 1 Guarding ...................................................................................... 35 Revision History ............................................................................... 3 Dielectric Relaxation .................................................................. 35 Specifications..................................................................................... 4 Humidity Effects......................................................................... 37 5 V Nominal Electrical Characteristics ..................................... 4 Contamination ............................................................................ 38 10 V Nominal Electrical Characteristics ................................... 6 Cleaning and Handling ............................................................. 39 15 V Nominal Electrical Characteristics ................................... 8 Solder Paste Selection ................................................................ 39 Absolute Maximum Ratings .......................................................... 10 Current Noise Considerations ...................................................... 40 Thermal Resistance .................................................................... 10 Layout Guidelines ........................................................................... 43 ESD Caution ................................................................................ 10 Physical Implementation of Guarding Techniques................ 43 Pin Configuration and Function Descriptions ........................... 11 Guard Ring .................................................................................. 43 Typical Performance Characteristics ........................................... 12 Guard Plane ................................................................................. 43 Main Amplifier, DC Performance ............................................ 12 Via Fence ..................................................................................... 44 Main Amplifier, AC Performance ............................................ 21 Cables and Connectors .............................................................. 44 Guard Amplifier ......................................................................... 27 Electrostatic Interferance .......................................................... 44 Theory of Operation ...................................................................... 29 Photodiode Interface ...................................................................... 45 ESD Structure.............................................................................. 29 DC Error Analysis ...................................................................... 45 Input Stage ................................................................................... 29 AC Error Analysis ...................................................................... 45 Gain Stage .................................................................................... 30 Noise Analysis............................................................................. 46 Output Stage ................................................................................ 30 Design Recommendations ........................................................ 47 Guard Buffer ............................................................................... 30 Design Example .......................................................................... 47 Applications Information .............................................................. 31 Power Supply Recommendations ................................................. 50 Input Protection .......................................................................... 31 Power Supply Considerations ................................................... 50 Single-Supply and Rail-to-Rail Output ................................... 31 Long-Term Drift ......................................................................... 51 Capacitive Load Stability ........................................................... 31 Temperature Hysteresis ............................................................. 51 EMI Rejection Ratio ................................................................... 32 Outline Dimensions ....................................................................... 52 High Impedance Measurements ................................................... 33 Ordering Guide .......................................................................... 52 Input Bias Current ...................................................................... 33 Rev. B | Page 2 of 52 Data Sheet ADA4530-1 REVISION HISTORY 5/2017--Rev. A to Rev. B Changes to Features Section and General Description Section ....... 1 Changed Offset Voltage Parameter to Input Offset Voltage Parameter, Table 1 ............................................................................. 4 Changed Offset Voltage Parameter to Input Offset Voltage Parameter, Table 2 ............................................................................. 6 Changed Offset Voltage Parameter to Input Offset Voltage Parameter, Table 3 ............................................................................. 8 Changes to EMI Rejection Ratio Section and Figure 102 ..........32 Moved Figure 114 ............................................................................38 Changes to Current Noise Considerations Section ....................41 Added Long-Term Drift Section, Temperature Hysteresis Section, Figure 136, Figure 137, and Figure 138; Renumbered Sequentially ......................................................................................51 Changes to Ordering Guide ...........................................................52 3/2016--Rev. 0 to Rev. A Changed DNC Pin to IC Pin ........................................Throughout Changes to Figure 1 .......................................................................... 1 Changes to Figure 3 and Table 6 ................................................... 10 Changes to Figure 29 ...................................................................... 15 Changes to Theory of Operation Section .................................... 28 Changes to Humidity Effects Section and Figure 112................ 36 Added Power Supply Recommendations Section, Power Supply Considerations Section, Table 16, and Figure 133 to Figure 135 ......................................................................................... 49 10/2015--Revision 0: Initial Version Rev. B | Page 3 of 52 ADA4530-1 Data Sheet SPECIFICATIONS 5 V NOMINAL ELECTRICAL CHARACTERISTICS Supply voltage (VSY) = 4.5 V, common-mode voltage (VCM) = VSY/2, TA = 25C, unless otherwise specified. Typical specifications are equal to the average of the distribution from characterization, unless otherwise noted. Minimum and maximum specifications are tested in production, unless otherwise noted. Table 1. Parameter 1 INPUT CHARACTERISTICS Input Bias Current 2, 3 Symbol Test Conditions/Comments IB RH < 50% -40C < TA < +85C, RH < 50% -40C < TA < +125C, RH < 50% RH < 50% -40C < TA < +125C, RH < 50% Input Offset Current3 IOS Input Offset Voltage2, 4 VOS Offset Voltage Drift2, 4 VOS/T Input Voltage Range Common-Mode Rejection Ratio IVR CMRR Large Signal Voltage Gain AVO Input Resistance Input Capacitance OUTPUT CHARACTERISTICS Output Voltage High RIN CIN Output Voltage Low Short-Circuit Current Source Sink Closed-Loop Output Impedance POWER SUPPLY Power Supply Rejection Ratio Supply Current VOH VOL Min RL = 10 k to VCM -40C < TA < +125C RL = 2 k to VCM -40C < TA < +125C RL = 10 k to VCM -40C < TA < +125C RL = 2 k to VCM -40C < TA < +125C Max Unit <1 20 20 250 20 150 40 50 70 150 300 0.5 2.8 3 fA fA fA fA fA V V V V V V/C V/C V dB dB dB dB dB T pF <1 +8 +9 VCM = 1.5 V to 3 V VCM = 1.5 V to 3 V, 0C < TA < 125C VCM = 1.5 V to 3 V, -40C < TA < 0C VCM = 0 V to 3 V 0C < TA < 125C -40C < TA < 0C VCM = 1.5 V to 3 V -40C < TA < +125C VCM = 0 V to 3 V RL = 2 k to VCM, VOUT = 0.2 V to 4.3 V -40C < TA < +125C -40C < TA < +125C Typ +0.13 -0.7 0 92 90 73 120 120 114 143 >100 8 4.47 4.46 4.4 4.38 4.49 4.45 10 30 30 40 100 120 V V V V mV mV mV mV ISC ZOUT f = 1 MHz, AV = 1 PSRR VSY = 4.5 V to 16 V -40C < TA < +125C IOUT = 0 mA -40C < TA < +125C ISY DYNAMIC PERFORMANCE Slew Rate Gain Bandwidth Product SR GBP Unity-Gain Crossover UGC RL = 10 k, CL = 10 pF, AV = 1 VIN = 10 mV rms, RL = 10 k, CL = 10 pF, AV = 100 VIN = 10 mV rms, RL = 10 k, CL = 10 pF, AVO = 1 Rev. B | Page 4 of 52 130 130 15 -30 20 mA mA 150 dB dB mA mA 0.9 1.3 1.5 1.4 2 V/s MHz 2 MHz Data Sheet Parameter 1 -3 dB Closed-Loop Bandwidth ADA4530-1 Symbol f-3dB Phase Margin M Settling Time to 0.1% tS EMI Rejection Ratio of +IN EMIRR NOISE PERFORMANCE Peak-to-Peak Voltage Noise Voltage Noise Density eN p-p eN Current Noise Density Total Harmonic Distortion + Noise Bandwidth = 90 kHz Bandwidth = 500 kHz GUARD BUFFER Guard Offset Voltage2, 4, 5 IN THD + N VGOS Guard Offset Voltage Drift2, 4 VGOS/T Output Impedance Output Voltage Range -3 dB Bandwidth ZGOUT f-3dBGUARD Test Conditions/Comments VIN=10 mV rms, RL = 10 k, CL = 10 pF, AV = 1 VIN = 10 mV rms, RL = 10 k, CL = 10 pF, AVO = 1 VIN = 0.5 V step, RL = 10 k, CL= 10 pF, AV = -1 VIN = 100 mV peak, f = 400 MHz VIN = 100 mV peak, f = 900 MHz VIN = 100 mV peak, f = 1800 MHz VIN = 100 mV peak, f = 2400 MHz Min f = 0.1 Hz to 10 Hz f = 10 Hz f = 1 kHz f = 10 kHz f = 0.1 Hz AV = 1, f = 1 kHz, VIN = 0.5 V rms VCM = 1.5 V to 3 V VCM = 1.5 V to 3 V, 0C < TA < 125C VCM = 1.5 V to 3 V, -40C < TA < 0C VCM = 0.1 V to 3 V 0C < TA < +125C -40C < TA < 0C VGOS < 150 V VIN = 10 mV rms, CL = 10 pF Typ 6 Max 62 Degrees 5 s 50 60 80 90 dB dB dB dB 4 80 16 14 0.07 V p-p nV/Hz nV/Hz nV/Hz fA/Hz 0.003 0.0045 % % 15 0.18 1.4 1 0.1 100 120 250 150 1 7 3 5.5 These specifications represent the performance for 5 V 10% power supplies. All specifications are measured at the worst case 4.5 V supply voltage. The maximum specifications at -40C < TA < +85C and -40C < TA < 0C are guaranteed from characterization. 3 RH is relative humidity (see the Humidity Effects section for more information). 4 The typical specifications are equal to the average plus the standard deviation of the distribution from characterization. 5 The guard offset voltage is the voltage difference between the guard output and the noninverting input. 1 2 Rev. B | Page 5 of 52 Unit MHz V V V V V/C V/C k V MHz ADA4530-1 Data Sheet 10 V NOMINAL ELECTRICAL CHARACTERISTICS VSY = 10 V, VCM = VSY/2, TA = 25C, unless otherwise noted. Typical specifications are equal to the average of the distribution from characterization, unless otherwise noted. Minimum and maximum specifications are tested in production, unless otherwise noted. Table 2. Parameter 1 INPUT CHARACTERISTICS Input Bias Current 2, 3 Symbol Test Conditions/Comments IB RH < 50% -40C < TA < +85C, RH < 50% -40C < TA < +125C, RH < 50% RH < 50% -40C < TA < +125C, RH < 50% Input Offset Current3 IOS Input Offset Voltage2, 4 VOS Offset Voltage Drift2, 4 VOS/T Input Voltage Range Common-Mode Rejection Ratio IVR CMRR Large Signal Voltage Gain AVO Input Resistance Input Capacitance OUTPUT CHARACTERISTICS Output Voltage High RIN CIN Output Voltage Low Short-Circuit Current Source Sink Closed-Loop Output Impedance POWER SUPPLY Power Supply Rejection Ratio Supply Current VOH VOL Min RL = 10 k to VCM -40C < TA < +125C RL = 2 k to VCM -40C < TA < +125C RL = 10 k to VCM -40C < TA < +125C RL = 2 k to VCM -40C < TA < +125C Max Unit <1 20 20 250 20 150 40 50 70 150 300 0.5 2.8 8.5 fA fA fA fA fA V V V V V V/C V/C V dB dB dB dB dB T pF <1 +8 +9 VCM = 1.5 V to 8.5 V VCM = 1.5 V to 8.5 V, 0C < TA < 125C VCM = 1.5 V to 8.5 V, -40C < TA < 0C VCM = 0 V to 8.5 V 0C < TA < 125C -40C < TA < 0C VCM = 1.5 V to 8.5 V -40C < TA < +125C VCM = 0 V to 8.5 V RL = 2 k to VCM, VOUT = 0.5 V to 9.5 V -40C < TA < +125C -40C < TA < +125C Typ +0.13 -0.7 0 105 100 87 125 125 114 150 >100 8 9.96 9.94 9.93 9.75 9.97 9.87 15 70 40 60 170 250 V V V V mV mV mV mV ISC ZOUT f = 1 MHz, AV = 1 PSRR VSY = 4.5 V to 16 V -40C < TA < +125C IOUT = 0 mA -40C < TA < +125C ISY DYNAMIC PERFORMANCE Slew Rate Gain Bandwidth Product SR GBP Unity-Gain Crossover UGC -3 dB Closed-Loop Bandwidth f-3dB Phase Margin M RL = 10 k, CL = 10 pF, AV = 1 VIN = 10 mV rms, RL = 10 k, CL = 10 pF, AV = 100 VIN = 10 mV rms, RL = 10 k, CL = 10 pF, AVO = 1 VIN = 10 mV rms, RL = 10 k, CL = 10 pF, AV = 1 VIN = 10 mV rms, RL = 10 k, CL = 10 pF, AVO = 1 Rev. B | Page 6 of 52 130 130 15 -30 20 mA mA 150 dB dB mA mA 0.9 1.3 1.5 1.4 2 V/s MHz 2 MHz 6 MHz 62 Degrees Data Sheet Parameter 1 Settling Time to 0.1% ADA4530-1 Symbol tS EMI Rejection Ratio of +IN EMIRR NOISE PERFORMANCE Peak-to-Peak Voltage Noise Voltage Noise Density eN p-p eN Current Noise Density Total Harmonic Distortion + Noise Bandwidth = 90 kHz Bandwidth = 500 kHz GUARD BUFFER Guard Offset Voltage2, 4, 5 IN THD + N VGOS Guard Offset Voltage Drift2, 4 VGOS/T Output Impedance Output Voltage Range -3 dB Bandwidth ZGOUT f-3dBGUARD Test Conditions/Comments VIN = 1 V step, RL = 10 k, CL = 10 pF, AV = -1 VIN = 100 mV peak, f = 400 MHz VIN = 100 mV peak, f = 900 MHz VIN = 100 mV peak, f = 1800 MHz VIN = 100 mV peak, f = 2400 MHz Min f = 0.1 Hz to 10 Hz f = 10 Hz f = 1 kHz f = 10 kHz f = 0.1 Hz AV = 1, f = 1 kHz, VIN = 2 V rms VCM = 1.5 V to 8.5 V VCM = 1.5 V to 8.5 V, 0C < TA < 125C VCM = 1.5 V to 8.5 V, -40C < TA < 0C VCM = 0.1 V to 8.5 V 0C < TA < 125C -40C < TA < 0C VGOS < 150 V VIN = 10 mV rms, CL = 10 pF Typ 6 2 Rev. B | Page 7 of 52 Unit s 50 60 80 90 dB dB dB dB 4 80 16 14 0.07 V p-p nV/Hz nV/Hz nV/Hz fA/Hz 0.0015 0.0025 % % 15 0.18 1.4 1 0.1 100 120 250 150 1 7 8.5 5.5 These specifications represent the performance for 10 V 10% power supplies. All specifications are measured at the 10 V supply voltage. The maximum specifications at -40C < TA < +85C and -40C < TA < 0C are guaranteed from characterization. 3 RH is relative humidity (see the Humidity Effects section for more information). 4 These typical specifications are equal to the average plus the standard deviation of the distribution from characterization. 5 The guard offset voltage is the voltage difference between the guard output and the noninverting input. 1 Max V V V V V/C V/C k V MHz ADA4530-1 Data Sheet 15 V NOMINAL ELECTRICAL CHARACTERISTICS VSY = 16 V, VCM = VSY/2, TA = 25C, unless otherwise noted. Typical specifications are equal to the average of the distribution from characterization, unless otherwise noted. Minimum and maximum specifications are tested in production, unless otherwise noted. Table 3. Parameter 1 INPUT CHARACTERISTICS Input Bias Current 2, 3 Symbol Test Conditions/Comments IB Input Offset Current IOS RH < 50% -40C < TA < +85C, RH < 50% -40C < TA < +125C, RH < 50% RH < 50% -40C < TA < +125C, RH < 50% Input Offset Voltage2, 4 VOS Offset Voltage Drift2, 4 VOS/T Input Voltage Range Common-Mode Rejection Ratio IVR CMRR Large Signal Voltage Gain AVO Input Resistance Input Capacitance OUTPUT CHARACTERISTICS Output Voltage High RIN CIN Output Voltage Low Short-Circuit Current Source Sink Closed-Loop Output Impedance POWER SUPPLY Power Supply Rejection Ratio Supply Current VOH VOL Min RL = 10 k to VCM -40C < TA < +125C RL = 2 k to VCM -40C < TA < +125C RL = 10 k to VCM -40C < TA < +125C RL = 2 k to VCM -40C < TA < +125C Max Unit <1 20 20 250 20 150 40 50 70 150 300 0.5 2.8 14.5 fA fA fA fA fA V V V V V V/C V/C V dB dB dB dB dB T pF <1 +8 +9 VCM = 1.5 V to 14.5 V VCM = 1.5 V to 14.5 V, 0C < TA < 125C VCM = 1.5 V to 14.5 V, -40C < TA < 0C VCM = 0 V to 14.5 V 0C < TA < 125C -40C < TA < 0C VCM = 1.5 V to 14.5 V -40C < TA < +125C VCM = 0 V to 14.5 V RL = 2 k to VCM, VOUT = 0.5 V to 15.5 V -40C < TA < +125C -40C < TA < +125C Typ +0.13 -0.7 0 110 105 93 130 125 114 155 >100 8 15.93 15.9 15.72 15.58 15.95 15.78 25 115 70 100 280 420 V V V V mV mV mV mV ISC ZOUT f = 1 MHz, AV = 1 PSRR VSY = 4.5 V to 16 V -40C < TA < +125C IOUT = 0 mA -40C < TA < +125C ISY DYNAMIC PERFORMANCE Slew Rate Gain bandwidth Product SR GBP Unity-Gain Crossover UGC -3 dB Closed-Loop Bandwidth f-3 dB Phase Margin M RL = 10 k, CL = 10 pF, AV = 1 VIN = 10 mV rms, RL = 10 k, CL = 10 pF, AV = 100 VIN = 10 mV rms, RL = 10 k, CL = 10 pF, AVO = 1 VIN = 10 mV rms, RL = 10 k, CL = 10 pF, AV = 1 VIN =10 mV rms, RL = 10 k, CL = 10 pF, AVO = 1 Rev. B | Page 8 of 52 130 130 15 -30 20 mA mA 150 dB dB mA mA 0.9 1.3 1.5 1.4 2 V/s MHz 2 MHz 6 MHz 62 Degrees Data Sheet Parameter 1 Settling Time to 0.1% EMI Rejection Ratio of +IN NOISE PERFORMANCE Peak-to-Peak Voltage Noise Voltage Noise Density Current Noise Density Total Harmonic Distortion + Noise Bandwidth = 90 kHz Bandwidth = 500 kHz GUARD BUFFER Guard Offset Voltage4, 5 ADA4530-1 Symbol tS EMIRR Test Conditions/Comments VIN = 1 V step, RL = 10 k, CL = 10 pF, AV = -1 VIN = 100 mV peak, f = 400 MHz VIN = 100 mV peak, f = 900 MHz VIN = 100 mV peak, f = 1800 MHz VIN = 100 mV peak, f = 2400 MHz eN p-p eN eN eN IN THD + N f = 0.1 Hz to 10 Hz f = 10 Hz f = 1 kHz f = 10 kHz f = 0.1 Hz AV = 1, f = 1 kHz, VIN = 4.5 V rms VGOS Guard Offset Voltage Drift2, 4 VGOS/T Output Impedance Output Voltage Range -3 dB Bandwidth ZGOUT f-3 dB GUARD Min VCM = 1.5 V to 14.5 V VCM = 1.5 V to 14.5 V, 0C < TA < 125C VCM = 1.5 V to 14.5 V, -40C < TA < 0C VCM = 0.1 V to 14.5 V 0C < TA < +125C -40C < TA < 0C VGOS < 150 V VIN = 10 mV rms, CL = 10 pF Typ 6 Max 50 60 80 90 dB dB dB dB 4 80 16 14 0.07 V p-p nV/Hz nV/Hz nV/Hz fA/Hz 0.0012 0.003 % % 15 0.18 1.4 1 0.1 100 120 250 150 1 7 14.5 5.5 These specifications represent the performance for 15 V 1 V power supplies. All specifications are measured at the worst case 16 V supply voltage. The maximum specifications at -40C < TA < +85C and -40C < TA < 0C are guaranteed from characterization. 3 RH is relative humidity (see the Humidity Effects section for more information). 4 These typical specifications are equal to the average plus the standard deviation of the distribution from characterization. 5 The guard offset voltage is the voltage difference between the guard output and the noninverting input. 1 2 Rev. B | Page 9 of 52 Unit s V V V V V/C V/C k V MHz ADA4530-1 Data Sheet ABSOLUTE MAXIMUM RATINGS Table 4. Parameter Supply Voltage Input Voltage Input Current1 Differential Input Voltage Output Short-Circuit Duration to GND Storage Temperature Range Operating Temperature Range Junction Temperature Range Lead Temperature (Soldering, 60 sec) Electrostatic Discharge (ESD) Human Body Model2 Field Induced Charged Device Model (FICDM)3 Rating 17 V (V-) - 0.3 V to (V+) + 0.3 V 10 mA 0.7 V Indefinite -65C to +150C -40C to +125C -65C to +150C 300C THERMAL RESISTANCE JA is specified for the worst case conditions, that is, a device soldered in a circuit board for surface-mount packages using a standard 4-layer JEDEC board. Table 5. Thermal Resistance 4 kV 1.25 kV Package Type 8-Lead SOIC The input pins have clamp diodes to the power supply pins. Limit the input current to 10 mA or less whenever input signals exceed the power supply rail by 0.3 V. 2 Applicable Standard ESDA/JEDEC JS-001-2012. 3 Applicable Standard JESD22-C101-E (ESD FICDM standard of JEDEC). 1 Stresses at or above those listed under Absolute Maximum Ratings may cause permanent damage to the product. This is a stress rating only; functional operation of the product at these or any other conditions above those indicated in the operational section of this specification is not implied. Operation beyond the maximum operating conditions for extended periods may affect product reliability. ESD CAUTION Rev. B | Page 10 of 52 JA 122 JC 41 Unit C/W Data Sheet ADA4530-1 PIN CONFIGURATION AND FUNCTION DESCRIPTIONS +IN 1 8 -IN GRD 2 7 GRD IC 3 6 OUT V- 4 5 V+ NOTES 1. IC = INTERNAL CONNECTION. THIS PIN MUST BE CONNECTED TO V- OR LEFT UNCONNECTED. 13405-003 ADA4530-1 Figure 3. Pin Configuration Table 6. Pin Function Descriptions Pin No. 1 2 3 4 5 6 7 8 Mnemonic +IN GRD IC V- V+ OUT GRD -IN Description Noninverting Input. Guard. Internal Connection. This pin must be connected to V- or left unconnected. Negative Supply Voltage. Positive Supply Voltage. Output. Guard. Inverting Input. Rev. B | Page 11 of 52 ADA4530-1 Data Sheet TYPICAL PERFORMANCE CHARACTERISTICS MAIN AMPLIFIER, DC PERFORMANCE TA = 25C, unless otherwise noted. 100 80 VSY = 4.5V VCM = VSY/2 590 CHANNELS x = 2.31V = 5.48V 90 70 40 60 VOS (V) NUMBER OF AMPLIFIERS 80 VSY = 4.5V VCM = VSY/2 574 CHANNELS 60 50 40 30 20 0 -20 20 -40 0 25 50 75 100 125 TEMPERATURE (C) Figure 4. Input Offset Voltage Distribution, VSY = 4.5 V Figure 7. Input Offset Voltage (VOS) vs. Temperature, VSY = 4.5 V 100 80 VSY = 10V VCM = VSY/2 590 CHANNELS x = 2.27V = 5.45V 90 80 70 VSY = 10V VCM = VSY/2 574 CHANNELS 60 40 60 VOS (V) NUMBER OF AMPLIFIERS -25 13405-007 40 VOS (V) -60 -50 13405-004 36 32 28 24 20 16 8 12 4 0 -4 -8 -12 -16 -20 -24 -28 -32 -36 0 -40 10 50 40 30 20 0 -20 20 -40 0 25 50 75 100 125 TEMPERATURE (C) Figure 5. Input Offset Voltage Distribution, VSY = 10 V Figure 8. Input Offset Voltage (VOS) vs. Temperature, VSY = 10 V 100 80 VSY = 16V VCM = VSY/2 590 CHANNELS x = 2.15V = 5.47V 90 80 70 VSY = 16V VCM = VSY/2 574 CHANNELS 60 40 VOS (V) 60 50 40 30 20 0 -20 20 -40 Figure 6. Input Offset Voltage Distribution, VSY = 16 V -60 -50 -25 0 25 50 75 100 125 TEMPERATURE (C) Figure 9. Input Offset Voltage (VOS) vs. Temperature, VSY = 16 V Rev. B | Page 12 of 52 13405-009 40 13405-006 VOS (V) 36 32 28 24 20 16 12 8 4 0 -4 -8 -12 -16 -20 -24 -28 -32 0 -36 10 -40 NUMBER OF AMPLIFIERS -25 13405-008 VOS (V) -60 -50 13405-005 40 36 32 28 24 16 20 12 8 4 0 -4 -8 -12 -16 -20 -24 -28 -32 -40 0 -36 10 Data Sheet ADA4530-1 120 120 VSY = 4.5V VCM = VSY/2 574 CHANNELS -40C TA 0C x = -0.29V/C = 0.42V/C 80 100 NUMBER OF AMPLIFIERS 60 40 20 -0.3 -0.2 -1.0 0 0.1 0.2 0.3 0.4 0.5 TCVOS (V/C) Figure 13. Input Offset Voltage Drift Distribution, 0C TA 125C, VSY = 4.5 V 120 120 VSY = 10V VCM = VSY/2 574 CHANNELS -40C TA 0C x = -0.29V/C = 0.42V/C 80 100 NUMBER OF AMPLIFIERS 100 60 40 80 VSY = 10V VCM = VSY/2 574 CHANNELS 0C TA 125C x = -0.025V/C = 0.107V/C 60 40 2.8 13405-014 2.4 1.6 2.0 0.8 1.2 0 0.4 -0.8 -0.4 -1.2 -2.0 -1.6 -2.8 -2.4 TCVOS (V/C) 0 -0.5 Figure 11. Input Offset Voltage Drift Distribution, -40C TA 0C, VSY = 10 V -0.4 -0.3 -0.2 -1.0 0 0.1 0.2 0.3 0.4 0.5 TCVOS (V/C) 13405-011 20 20 0 -0.4 13405-010 2.8 Figure 10. Input Offset Voltage Drift Distribution, -40C TA 0C, VSY = 4.5 V NUMBER OF AMPLIFIERS 40 0 -0.5 13405-013 2.4 1.6 2.0 0.8 1.2 0 0.4 -0.8 -0.4 -1.2 -2.0 -1.6 -2.8 TCVOS (V/C) Figure 14. Input Offset Voltage Drift Distribution, 0C TA 125C, VSY = 10 V 120 120 VSY = 16V VCM = VSY/2 574 CHANNELS -40C TA 0C x = -0.29V/C = 0.41V/C 80 100 NUMBER OF AMPLIFIERS 100 NUMBER OF AMPLIFIERS 60 20 -2.4 0 80 60 40 80 VSY = 16V VCM = VSY/2 574 CHANNELS 0C TA 125C x = -0.024V/C = 0.107V/C 60 40 20 2.4 1.6 2.0 0.8 0 -0.5 13405-015 TCVOS (V/C) 1.2 0 0.4 -0.8 -0.4 -1.2 -2.0 -1.6 -2.8 -2.4 0 2.8 20 Figure 12. Input Offset Voltage Drift Distribution, -40C TA 0C, VSY = 16 V -0.4 -0.3 -0.2 -1.0 0 0.1 TCVOS (V/C) 0.2 0.3 0.4 0.5 13405-012 NUMBER OF AMPLIFIERS 100 VSY = 4.5V VCM = VSY/2 574 CHANNELS 0C TA 125C x = -0.025V/C = 0.107V/C Figure 15. Input Offset Voltage Drift Distribution, 0C TA 125C, VSY = 16 V Rev. B | Page 13 of 52 ADA4530-1 Data Sheet 60 10 8 0 -20 -40 0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 VCM (V) 2 0 -2 -4 -6 VSY = 10V -8 27 CHANNELS TA = 25C -10 0 500 1000 13405-016 -60 4 2500 3000 3500 4000 4500 Figure 19. VOS Long-Term Drift 60 2.5 VSY = 10V TA = 25C 20 0 -20 -40 VSY = 10V 590 CHANNELS TA = 25C 0 1 2 3 4 5 6 7 8 9 10 VCM (V) 1.5 1.0 0.5 0 13405-017 -60 2.0 0 1 Figure 17. Input Offset Voltage (VOS) vs. Common-Mode Voltage (VCM), VSY = 10 V 3 4 5 Figure 20. VOS Warm-Up Time 60 0 VSY = 10V VCM = 400mV -20 PREFERRED COMMON-MODE RANGE -40 SMALL SIGNAL CMRR (dB) 40 2 TIME AFTER POWER-ON (Minutes) 13405-220 INPUT OFFSET VOLTAGE (V) PREFERRED COMMON-MODE RANGE 40 VOS (V) 2000 TIME (Hours) Figure 16. Input Offset Voltage (VOS) vs. Common-Mode Voltage (VCM), VSY = 4.5 V 20 VOS (V) 1500 0 -20 -60 PREFERRED COMMON-MODE RANGE -80 -100 -120 -140 -160 -40 VSY = 16V 590 CHANNELS TA = 25C -60 0 1 2 3 4 5 6 7 8 9 VCM (V) 10 11 12 13 14 15 16 13405-018 -180 Figure 18. Input Offset Voltage (VOS) vs. Common-Mode Voltage (VCM), VSY = 16 V Rev. B | Page 14 of 52 -200 0 1 2 3 4 5 6 7 8 9 10 VCM (V) Figure 21. Small Signal CMRR vs. Common-Mode Voltage (VCM) 13405-221 VOS (V) 20 6 13405-219 PREFERRED COMMON-MODE RANGE INPUT OFFSET VOLTAGE (V) 40 VSY = 4.5V 590 CHANNELS TA = 25C Data Sheet ADA4530-1 20 20 15 10 IB+ (fA) 0 -5 -10 -10 -15 -15 -20 0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 VCM (V) Figure 22. Inverting Input Bias Current (IB-) vs. Common-Mode Voltage (VCM), VSY = 4.5 V, TA = 85C 0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 VCM (V) Figure 25. Noninverting Input Bias Current (IB+) vs. Common-Mode Voltage (VCM), VSY = 4.5 V, TA = 85C 20 20 VSY = 10V 27 CHANNELS TA = 85C 15 VSY = 10V 27 CHANNELS TA = 85C 15 10 PREFERRED COMMON-MODE RANGE 0 0 -5 -5 -10 -10 -15 -15 1 2 3 4 5 6 7 8 9 10 VCM (V) -20 13405-023 -20 0 PREFERRED COMMON-MODE RANGE 5 IB+ (fA) 5 Figure 23. Inverting Input Bias Current (IB-) vs. Common-Mode Voltage (VCM), VSY = 10 V, TA = 85C 0 1 2 3 4 5 6 7 8 9 10 VCM (V) 13405-026 10 Figure 26. Noninverting Input Bias Current (IB+) vs. Common-Mode Voltage (VCM), VSY = 10 V, TA = 85C 20 20 VSY = 16V 27 CHANNELS TA = 85C 15 VSY = 16V 27 CHANNELS TA = 85C 15 10 10 PREFERRED COMMON-MODE RANGE 0 0 -5 -5 -10 -10 -15 -15 2 4 6 8 VCM (V) 10 12 14 16 -20 13405-024 -20 0 PREFERRED COMMON-MODE RANGE 5 IB+ (fA) 5 Figure 24. Inverting Input Bias Current (IB-) vs. Common-Mode Voltage (VCM), VSY = 16 V, TA = 85C 0 2 4 6 8 VCM (V) 10 12 14 16 13405-027 IB- (fA) 0 -5 -20 IB- (fA) PREFERRED COMMON-MODE RANGE 5 13405-022 IB- (fA) 10 PREFERRED COMMON-MODE RANGE 5 VSY = 4.5V 27 CHANNELS TA = 85C 15 13405-025 VSY = 4.5V 27 CHANNELS TA = 85C Figure 27. Noninverting Input Bias Current (IB+) vs. Common-Mode Voltage (VCM), VSY = 16 V, TA = 85C Rev. B | Page 15 of 52 ADA4530-1 Data Sheet 300 VSY = 4.5V 27 CHANNELS TA = 125C 200 IB+ (fA) 0 -100 -200 -200 -300 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 VCM (V) Figure 28. Inverting Input Bias Current (IB-) vs. Common-Mode Voltage (VCM), VSY = 4.5 V, TA = 125C -300 0 0.5 1.0 1.5 2.0 3.0 2.5 3.5 4.0 4.5 VCM (V) Figure 31. Noninverting Input Bias Current (IB+) vs. Common-Mode Voltage (VCM), VSY = 4.5 V, TA = 125C 300 VSY = 10V 27 CHANNELS TA = 125C 200 PREFERRED COMMON-MODE RANGE 100 -100 -100 IB+ (fA) 0 -200 -200 -300 -300 -400 -400 -500 -500 2 3 4 5 6 7 8 9 10 VCM (V) -600 13405-029 -600 1 PREFERRED COMMON-MODE RANGE 100 0 0 VSY = 10V 27 CHANNELS TA = 125C 200 Figure 29. Inverting Input Bias Current (IB-) vs. Common-Mode Voltage (VCM), VSY = 10 V, TA = 125C 0 1 2 3 4 5 6 7 8 9 10 VCM (V) 13405-032 300 IB- (fA) 0 -100 0 PREFERRED COMMON-MODE RANGE 100 13405-028 IB- (fA) 200 PREFERRED COMMON-MODE RANGE 100 VSY = 4.5V 27 CHANNELS TA = 125C 13405-031 300 Figure 32. Noninverting Input Bias Current (IB+) vs. Common-Mode Voltage (VCM), VSY = 10 V, TA = 125C 300 300 VSY = 16V 27 CHANNELS TA = 125C 200 VSY = 16V 27 CHANNELS TA = 125C 200 100 PREFERRED COMMON-MODE RANGE PREFERRED COMMON-MODE RANGE 0 IB+ (fA) IB- (fA) 100 0 -100 -200 -300 -100 -400 -200 0 2 4 6 8 VCM (V) 10 12 14 16 -600 13405-030 -300 Figure 30. Inverting Input Bias Current (IB-) vs. Common-Mode Voltage (VCM), VSY = 16 V, TA = 125C 0 2 4 6 8 VCM (V) 10 12 14 16 13405-033 -500 Figure 33. Noninverting Input Bias Current (IB+) vs. Common-Mode Voltage (VCM), VSY = 16 V, TA = 125C Rev. B | Page 16 of 52 Data Sheet ADA4530-1 1000 120 -40C TO +125C LIMIT VSY = 4.5V VCM = VSY/2 RH < 10% 100 100 NUMBER OF AMPLIFIERS -40C TO +85C LIMIT 1 0.1 IB+ I B- 40 10 20 30 40 50 60 70 80 90 100 110 120 130 TEMPERATURE (C) 0 -250 -200 -150 -100 -50 0 50 100 150 200 250 IB- (fA) 13405-019 0 13405-234 0.001 Figure 37. Inverting Input Bias Current Histogram, TA = 125C, VSY = 10 V Figure 34. Input Bias Current (IB) vs. Temperature, VSY = 4.5 V 120 1000 -40C TO +125C LIMIT VSY = 10V VCM = VSY/2 RH < 10% 100 VSY = 10V VCM = VSY/2 590 CHANNELS TA = 125C x = -74.59fA = 23.66fA 100 NUMBER OF AMPLIFIERS -40C TO +85C LIMIT 10 1 0.1 IB+ IB- 80 60 40 20 0.01 0 10 20 30 40 50 60 70 80 90 100 110 120 130 TEMPERATURE (C) 0 -250 -200 -150 -100 13405-235 0.001 -50 0 50 100 150 200 250 IB+ (fA) 13405-020 IB (fA) 60 20 0.01 Figure 38. Noninverting Input Bias Current Histogram, TA = 125C, VSY = 10 V Figure 35. Input Bias Current (IB) vs. Temperature, VSY = 10 V 160 1000 VSY = 16V VCM = VSY/2 RH < 10% 100 VSY = 10V VCM = VSY/2 590 CHANNELS TA = 125C x = 33.9fA = 17.9fA -40C TO +125C LIMIT 140 NUMBER OF AMPLIFIERS -40C TO +85C LIMIT 10 1 0.1 IB+ IB- 0.01 120 100 80 60 40 20 0.001 0 10 20 30 40 50 60 70 80 90 100 110 120 130 TEMPERATURE (C) 13405-236 IB (fA) 80 0 -150 -120 -90 -60 -30 0 30 60 90 IOS (fA) Figure 39. Input Offset Current Histogram Figure 36. Input Bias Current (IB) vs. Temperature, VSY = 16 V Rev. B | Page 17 of 52 120 150 13405-239 IB (fA) 10 VSY = 10V VCM = VSY/2 590 CHANNELS TA = 125C x = -40.69fA = 24.54fA 0.0001 0.01 0.1 1 10 100 LOAD CURRENT (mA) 10 1 -40C +25C +85C +125C VSY = 10V 0.1 0.01 0.001 0.0001 0.01 0.1 1 10 100 LOAD CURRENT (mA) 13405-038 OUTPUT VOLTAGE LOW (VOL) TO SUPPLY RAIL (V) Figure 40. Output Voltage Low (VOL) to Supply Rail vs. Load Current (ILOAD), VSY = 4.5 V 10 1 -40C +25C +85C +125C VSY = 16V 0.1 0.01 0.001 0.0001 0.01 0.1 1 LOAD CURRENT (mA) 10 100 13405-039 OUTPUT VOLTAGE LOW (VOL) TO SUPPLY RAIL (V) Figure 41 Output Voltage Low (VOL) to Supply Rail vs. Load Current (ILOAD), VSY = 10 V Figure 42. Output Voltage Low (VOL) to Supply Rail vs. Load Current (ILOAD), VSY = 16 V VSY = 4.5V 0.1 0.01 0.001 0.01 0.1 1 10 100 LOAD CURRENT (mA) 13405-040 0.001 1 -40C +25C +85C +125C Figure 43. Output Voltage High (VOH) to Supply Rail vs. Load Current (ILOAD), VSY = 4.5 V 10 1 -40C +25C +85C +125C VSY = 10V 0.1 0.01 0.001 0.01 0.1 1 10 100 LOAD CURRENT (mA) 13405-041 0.01 10 Figure 44. Output Voltage High (VOH) to Supply Rail vs. Load Current (ILOAD), VSY = 10 V 10 1 -40C +25C +85C +125C VSY = 16V 0.1 0.01 0.001 0.01 0.1 1 LOAD CURRENT (mA) 10 100 13405-042 0.1 OUTPUT VOLTAGE HIGH (VOH) TO SUPPLY RAIL (V) VSY = 4.5V OUTPUT VOLTAGE HIGH (VOH) TO SUPPLY RAIL (V) 1 -40C +25C +85C +125C OUTPUT VOLTAGE HIGH (VOH) TO SUPPLY RAIL (V) 10 Data Sheet 13405-037 OUTPUT VOLTAGE LOW (VOL) TO SUPPLY RAIL (V) ADA4530-1 Figure 45. Output Voltage High (VOH) to Supply Rail vs. Load Current (ILOAD), VSY = 16 V Rev. B | Page 18 of 52 80 70 60 50 RL = 2k 30 RL = 10k 10 -25 0 25 50 75 100 125 TEMPERATURE (C) 225 200 VSY = 10V 175 150 125 100 RL = 2k 75 50 RL = 10k 25 0 -50 -25 0 25 50 75 100 125 TEMPERATURE (C) 360 VSY = 16V 300 270 240 210 180 RL = 2k 150 120 90 60 RL = 10k 30 0 -50 -25 0 25 50 75 100 125 TEMPERATURE (C) Figure 48. Output Voltage Low (VOL) to Supply Rail vs. Temperature, VSY = 16 V 80 70 RL = 2k 60 50 40 30 RL = 10k 20 10 0 -50 -25 0 25 50 75 100 125 TEMPERATURE (C) 225 200 VSY = 10V 175 RL = 2k 150 125 100 75 50 RL = 10k 25 0 -50 -25 0 25 50 75 100 125 TEMPERATURE (C) Figure 50. Output Voltage High (VOH) to Supply Rail vs. Temperature, VSY = 10 V 13405-045 OUTPUT VOLTAGE LOW (VOL) TO SUPPLY RAIL (mV) Figure 47. Output Voltage Low (VOL) to Supply Rail vs. Temperature, VSY = 10 V 330 VSY = 4.5V Figure 49. Output Voltage High (VOH) to Supply Rail vs. Temperature, VSY = 4.5 V 13405-044 OUTPUT VOLTAGE LOW (VOL) TO SUPPLY RAIL (mV) Figure 46. Output Voltage Low (VOL) to Supply Rail vs. Temperature, VSY = 4.5 V 90 13405-047 0 -50 100 360 330 VSY = 16V 300 270 RL = 2k 240 210 180 150 120 90 RL = 10k 60 30 0 -50 -25 0 25 50 TEMPERATURE (C) 75 100 125 13405-048 20 OUTPUT VOLTAGE HIGH (VOH) TO SUPPLY RAIL (mV) 40 OUTPUT VOLTAGE HIGH (VOH) TO SUPPLY RAIL (mV) 90 VSY = 4.5V 13405-046 100 OUTPUT VOLTAGE HIGH (VOH) TO SUPPLY RAIL (mV) ADA4530-1 13405-043 OUTPUT VOLTAGE LOW (VOL) TO SUPPLY RAIL (mV) Data Sheet Figure 51. Output Voltage High (VOH) to Supply Rail vs. Temperature, VSY = 16 V Rev. B | Page 19 of 52 ADA4530-1 Data Sheet 0 -40C +25C +85C 1.2 +125C VCM = VSY/2 VSY = 10V VCM = 400mV -20 PSRR- PSRR+ -40 SMALL SIGNAL PSRR (dB) ISY PER AMPLIFIER (mA) 1.4 1.0 0.8 0.6 0.4 -60 PREFERRED INPUT VOLTAGE RANGE -80 -100 -120 -140 -160 0.2 2 4 6 8 10 11 14 16 SUPPLY VOLTAGE (V) Figure 52. Supply Current (ISY) per Amplifier vs. Supply Voltage (VSY) VCM = VSY/2 1.0 0.8 0.6 0.4 4.5V 10V 16V -25 0 25 50 75 100 125 TEMPERATURE (C) 13405-050 ISY PER AMPLIFIER (mA) 1.2 0 -50 0 1 2 3 4 5 6 7 8 9 10 VCM (V) Figure 54. Small Signal PSRR vs. Common-Mode Voltage (VCM) 1.4 0.2 -200 Figure 53. Supply Current (ISY) per Amplifier vs. Temperature Rev. B | Page 20 of 52 13405-254 0 13405-049 -180 0 Data Sheet ADA4530-1 MAIN AMPLIFIER, AC PERFORMANCE VSY = 4.5 V to 16 V, data taken at VSY = 10 V, TA = 25C, unless otherwise noted. 120 60 100 100 50 80 80 60 60 40 40 -20 -40 -60 10k -20 100k -60 10M 1M AV = +10 20 10 AV = +1 0 -10 -40 VSY = 10V RL = 10k 30 -20 1k FREQUENCY (Hz) 10k Figure 55. Open-Loop Gain and Phase Margin vs. Frequency 10k CLOSED-LOOP OUTPUT IMPEDANCE () VSY = 10V VCM = VSY/2 RL = 10k CL = 10pF 80 60 50 40 30 20 1M 10M FREQUENCY (Hz) 100 AV = +10 10 AV = +100 1 AV = +1 0.1 0.01 0.001 1k 10k 100k 1M 10M FREQUENCY (Hz) Figure 59. Closed-Loop Output Impedance vs. Frequency Figure 56. CMRR vs. Frequency 60 60 PSSR+ PSSR- VSY = 10V VCM = VSY/2 RL = 10k CL = 10pF 40 VSY = 10V VIN = 100mV p-p AV = +1 RL = 10k 50 OVERSHOOT (%) 50 30 20 40 OS- 30 20 OS+ 10 10 0 10k 100k 1M FREQUENCY (Hz) 10M 13405-057 PSRR (dB) VSY = 10V VCM = VSY/2 1k 0.0001 100 13405-056 10 100k 10M 0 0 30 60 90 120 150 180 210 240 LOAD CAPACITANCE (pF) Figure 60. Small Signal Overshoot vs. Load Capacitance Figure 57. PSRR vs. Frequency Rev. B | Page 21 of 52 270 13405-060 CMRR (dB) 70 0 10k 1M Figure 58. Closed-Loop Gain vs. Frequency 100 90 100k FREQUENCY (Hz) 13405-058 0 10pF 10pF 100pF 100pF 40 13405-059 0 CLOSED-LOOP GAIN (dB) 20 PHASE MARGIN (Degrees) 20 VSY = 10V RL = 10k CL = 10pF AV = +100 13405-055 OPEN-LOOP GAIN (dB) 120 ADA4530-1 Data Sheet 80 1.0 60 0.5 0 -20 -1.0 -1.5 -40 -2.0 -80 TIME (2s/DIV) 13405-061 -60 -2.5 TIME (2s/DIV) Figure 61. Small Signal Transient Response, VSY = 4.5 V Figure 64. Large Signal Transient Response, VSY = 4.5 V 6 80 60 4 2 20 -20 VOLTAGE (V) VOLTAGE (mV) 40 0 VSY = 4.5V VIN = 2.75V p-p AV = +1 RL = 10k CL = 10pF RS = 1k -0.5 13405-064 20 0 VSY = 4.5V VIN = 100mV p-p AV = +1 RL = 10k CL = 10pF RS = 1k VOLTAGE (V) VOLTAGE (mV) 40 VSY = 10V VIN = 100mV p-p AV = +1 RL = 10k CL = 10pF RS = 1k VSY = 10V VIN = 8.25V p-p AV = +1 RL = 10k CL = 10pF RS = 1k 0 -2 -40 -4 TIME (2s/DIV) -6 TIME (5s/DIV) Figure 62. Small Signal Transient Response, VSY = 10 V Figure 65. Large Signal Transient Response, VSY = 10 V 80 10 8 60 6 40 VOLTAGE (V) -20 VSY = 16V VIN = 100mV p-p AV = +1 RL = 10k CL = 10pF RS = 1k VSY = 16V VIN = 14.25V p-p AV = +1 RL = 10k CL = 10pF RS = 1k 2 0 -2 -4 -40 -6 -60 TIME (2s/DIV) -10 TIME (20s/DIV) Figure 66. Large Signal Transient Response, VSY = 16 V Figure 63. Small Signal Transient Response, VSY = 16 V Rev. B | Page 22 of 52 13405-066 -8 -80 13405-063 VOLTAGE (mV) 4 20 0 13405-065 -80 13405-062 -60 Data Sheet ADA4530-1 6 VSY = 4.5V VIN = 450mV AV = -10 RL = 10k CL = 10pF 2 VOUT -1.0 -1.2 -1.4 0 1.0 -0.2 0.5 -0.4 0 -0.6 -0.5 -0.8 1 -1.0 0 -1.2 -1 TIME (2s/DIV) 0.6 VIN 3 0.4 0.2 1 0 -0.2 -1 -0.4 -0.6 -0.8 -1.0 VSY = 10V VIN = 900mV AV = -10 RL = 10k CL = 10pF VOUT -1.2 -1.4 -3 -5 -7 -1.6 TIME (2s/DIV) Figure 71. Negative Overload Recovery, VSY = 10 V 2 12 21 15 12 -3 VOUT 9 -5 6 -6 3 -7 0 INPUT VOLTAGE (V) VIN 1 VIN 8 18 OUTPUT VOLTAGE (V) VSY = 16V VIN = 1.5V AV = -10 RL = 10k CL = 10pF -1 0 4 -1 0 -2 -4 VSY = 16V VIN = 1.5V AV = -10 RL = 10k CL = 10pF VOUT -3 -3 -8 TIME (2s/DIV) 13405-069 INPUT VOLTAGE (V) 5 0.8 24 0 -4 7 1.0 Figure 68. Positive Overload Recovery, VSY = 10 V -2 -2.5 1.2 TIME (2s/DIV) 1 -2.0 OUTPUT VOLTAGE (V) 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0 -1 INPUT VOLTAGE (V) VOUT -1.5 Figure 70. Negative Overload Recovery, VSY = 4.5 V OUTPUT VOLTAGE (V) VIN -1.0 TIME (2s/DIV) 13405-068 INPUT VOLTAGE (V) VSY = 10V VIN = 900mV AV = -10 RL = 10k CL = 10pF VOUT -1.4 Figure 67. Positive Overload Recovery, VSY = 4.5 V 0.2 0 -0.2 -0.4 -0.6 -0.8 -1.0 -1.2 -1.4 -1.6 -1.8 -2.0 -2.2 -2.4 -2.6 -2.8 -3.0 -3.2 VSY = 4.5V VIN = 400mV AV = -10 RL = 10k CL = 10pF Figure 69. Positive Overload Recovery, VSY = 16 V -4 Figure 72. Negative Overload Recovery, VSY = 16 V Rev. B | Page 23 of 52 -8 -12 TIME (2s/DIV) OUTPUT VOLTAGE (V) -0.8 3 1.5 13405-072 -0.6 4 0.2 13405-071 VIN 5 OUTPUT VOLTAGE (V) -0.4 VIN 13405-067 INPUT VOLTAGE (V) -0.2 2.0 0.4 OUTPUT VOLTAGE (V) 0 2.5 0.6 13405-070 7 INPUT VOLTAGE (V) 0.2 TIME (1s/DIV) ERROR BAND POST GAIN = 20 TIME (1s/DIV) Figure 76. Positive Settling Time to 0.1%, VSY = 4.5 V ERROR BAND POST GAIN = 20 TIME (1s/DIV) OUTPUT ERROR BAND POST GAIN = 20 TIME (1s/DIV) INPUT VOLTAGES (250mV/DIV) Figure 77. Positive Settling Time to 0.1%, VSY = 10 V VSY = 16V DUT AV = -1 RL = 10k CL = 10pF ERROR BAND POST GAIN = 20 TIME (1s/DIV) OUTPUT VOLTAGE (5mV/DIV) INPUT INPUT VSY = 16V DUT AV = -1 RL = 10k CL = 10pF OUTPUT ERROR BAND POST GAIN = 20 13405-275 INPUT VOLTAGES (250mV/DIV) Figure 74. Negative Settling Time to 0.1%, VSY = 10 V OUTPUT VSY = 10V DUT AV = -1 RL = 10k CL = 10pF OUTPUT VOLTAGE (5mV/DIV) OUTPUT OUTPUT VOLTAGE (5mV/DIV) INPUT INPUT 13405-277 INPUT VOLTAGES (250mV/DIV) VSY = 10V DUT AV = -1 RL = 10k CL = 10pF 13405-274 INPUT VOLTAGES (250mV/DIV) Figure 73. Negative Settling Time to 0.1%, VSY = 4.5 V TIME (1s/DIV) Figure 78. Positive Settling Time to 0.1%, VSY = 16 V Figure 75. Negative Settling Time to 0.1%, VSY = 16 V Rev. B | Page 24 of 52 OUTPUT VOLTAGE (5mV/DIV) ERROR BAND POST GAIN = 20 OUTPUT 13405-278 OUTPUT OUTPUT VOLTAGE (5mV/DIV) INPUT INPUT VSY = 4.5V DUT AV = -1 RL = 10k CL = 10pF OUTPUT VOLTAGE (5mV/DIV) VSY = 4.5V DUT AV = -1 RL = 10k CL = 10pF 13405-276 INPUT VOLTAGES (250mV/DIV) Data Sheet 13405-273 INPUT VOLTAGES (250mV/DIV) ADA4530-1 Data Sheet ADA4530-1 0.1 10 90kHz LOW-PASS FILTER 500kHz LOW-PASS FILTER VSY = 4.5V AV = +1 RL = 10k f = 1kHz VSY = 4.5V AV = +1 RL = 10k VIN = 0.5V rms THD + N (%) THD + N (%) 1 0.01 0.1 0.01 100 1k 10k 100k FREQUENCY (Hz) 0.001 0.001 13405-279 0.001 10 1 Figure 82. THD + N vs. Amplitude, VSY = 4.5 V 10 VSY = 10V AV = +1 RL = 10k VIN = 2V rms VSY = 10V AV = +1 RL = 10k f = 1kHz 1 THD + N (%) THD + N (%) 90kHz LOW-PASS FILTER 500kHz LOW-PASS FILTER 0.1 AMPLITUDE (V rms) Figure 79. THD + N vs. Frequency, VSY = 4.5 V 0.1 0.01 13405-282 90kHz LOW-PASS FILTER 500kHz LOW-PASS FILTER 0.01 0.1 0.01 0.001 100 1k 10k 100k FREQUENCY (Hz) 0.0001 0.001 13405-280 0.001 10 90kHz LOW-PASS FILTER 500kHz LOW-PASS FILTER 0.1 1 10 AMPLITUDE (V rms) Figure 83. THD + N vs. Amplitude, VSY = 10 V Figure 80. THD + N vs. Frequency, VSY = 10 V 1 0.01 13405-283 90kHz LOW-PASS FILTER 500kHz LOW-PASS FILTER 10 VSY = 16V AV = +1 RL = 10k VIN = 4.5V rms VSY = 16V AV = +1 RL = 10k f = 1kHz 1 THD + N (%) THD + N (%) 0.1 0.1 0.01 0.01 100 1k 10k FREQUENCY (Hz) 100k Figure 81. THD + N vs. Frequency, VSY = 16 V 0.001 0.001 0.01 0.1 1 AMPLITUDE (V rms) Figure 84. THD + N vs. Amplitude, VSY = 16 V Rev. B | Page 25 of 52 10 13405-284 0.001 10 13405-281 90kHz LOW-PASS FILTER 500kHz LOW-PASS FILTER ADA4530-1 Data Sheet 3 100 10 1 0.1 1 10 100 1k 10k 100k 1M 10M 100M FREQUENCY (Hz) Figure 85. Voltage Noise Density vs. Frequency, VSY = 10 V VSY = 10V NOISE = 4V p-p 2 1 0 -1 -2 -3 TIME (1s/DIV) Figure 86. 0.1 Hz to 10 Hz Noise Rev. B | Page 26 of 52 13405-286 INPUT REFERRED VOLTAGE (V) VSY = 10V AV = +1 RL = 10k CL = 10pF 13405-285 VOLTAGE NOISE DENSITY (nV/Hz) 1000 Data Sheet ADA4530-1 GUARD AMPLIFIER TA = 25C, unless otherwise noted. 100 120 VSY = 4.5V VCM = VSY/2 590 CHANNELS TA = 25C x = -3.68V = 12.35V 80 50 VOS (V) 60 0 -50 40 VOS (V) -150 -50 13405-091 -10 0 10 20 30 40 50 60 70 80 90 100 -30 -20 -50 -40 -100 -90 -80 -70 -60 0 0 25 50 100 125 Figure 90. Input Offset Voltage (VOS) vs. Temperature, VSY = 4.5 V 120 100 VSY = 10V VCM = VSY/2 590 CHANNELS TA = 25C x = -3.8V = 12.4V 100 VOS (V) 80 50 60 0 -50 40 VOS (V) -150 -50 13405-092 -10 0 10 20 30 40 50 60 70 80 90 100 -30 -20 -50 -40 -100 -90 -80 -70 -60 0 -25 0 25 50 75 100 125 TEMPERATURE (C) 13405-095 VSY = 10V VCM = VSY/2 574 CHANNELS -100 20 Figure 91. Input Offset Voltage (VOS) vs. Temperature, VSY = 10 V Figure 88. Input Offset Voltage Distribution, VSY = 10 V 100 120 VSY = 16V VCM = VSY/2 590 CHANNELS TA = 25C x = -3.86V = 12.2V 100 50 VOS (V) 80 60 0 -50 40 VSY = 16V VCM = VSY/2 574 CHANNELS -100 VOS (V) 100 13405-093 0 10 20 30 40 50 60 70 80 90 -20 -10 0 -40 -30 20 -100 -90 -80 -70 -60 -50 NUMBER OF AMPLIFIERS 75 TEMPERATURE (C) Figure 87. Input Offset Voltage Distribution, VSY = 4.5 V NUMBER OF AMPLIFIERS -25 13405-094 VSY = 4.5V VCM = VSY/2 574 CHANNELS -100 20 Figure 89. Input Offset Voltage Distribution, VSY = 16 V -150 -50 -25 0 25 50 75 100 125 TEMPERATURE (C) Figure 92. Input Offset Voltage (VOS) vs. Temperature, VSY = 16 V Rev. B | Page 27 of 52 13405-096 NUMBER OF AMPLIFIERS 100 ADA4530-1 Data Sheet 160 120 VSY = 4.5V VCM = VSY/2 574 CHANNELS -40C TA 0C x = 0.26V/C = 1.14V/C 80 140 NUMBER OF AMPLIFIERS NUMBER OF AMPLIFIERS 100 60 40 120 100 VSY = 4.5V VCM = VSY/2 574 CHANNELS 0C TA 125C x = 0.014V/C = 0.168V/C 80 60 40 20 Figure 93. Input Offset Voltage Drift Distribution, -40C TA 0C, VSY = 4.5 V -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1.0 TCVOS (V/C) Figure 96. Input Offset Voltage Drift Distribution, 0C TA 125C, VSY = 4.5 V 160 120 VSY = 10V VCM = VSY/2 574 CHANNELS -40C TA 0C x = 0.26V/C = 1.14V/C 80 140 NUMBER OF AMPLIFIERS 100 NUMBER OF AMPLIFIERS -0.8 13405-100 TCVOS (V/C) 0 -1.0 13405-097 6 4 5 2 3 0 1 -2 -1 -4 -3 -5 -7 -6 0 7 20 60 40 120 100 VSY = 10V VCM = VSY/2 574 CHANNELS 0C TA 125C x = 0.017V/C = 0.168V/C 80 60 40 20 Figure 94. Input Offset Voltage Drift Distribution, -40C TA 0C, VSY = 10 V -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1.0 TCVOS (V/C) Figure 97. Input Offset Voltage Drift Distribution, 0C TA 125C, VSY = 10 V 120 160 VSY = 16V VCM = VSY/2 574 CHANNELS -40C TA 0C x = 0.27V/C = 1.14V/C 80 140 NUMBER OF AMPLIFIERS 100 NUMBER OF AMPLIFIERS -0.8 13405-101 TCVOS (V/C) 0 -1.0 13405-098 6 4 5 2 3 0 1 -2 -1 -4 -3 -5 -7 -6 0 7 20 60 40 120 100 VSY = 16V VCM = VSY/2 574 CHANNELS 0C TA 125C x = 0.02V/C = 0.168V/C 80 60 40 20 Figure 95. Input Offset Voltage Drift Distribution, -40C TA 0C, VSY = 16 V -0.8 -0.6 -0.4 -0.2 0 0.2 TCVOS (V/C) 0.4 0.6 0.8 1.0 13405-102 6 4 5 2 0 -1.0 13405-099 TCVOS (V/C) 3 0 1 -2 -1 -4 -3 -5 -7 -6 0 7 20 Figure 98. Input Offset Voltage Drift Distribution, 0C TA 125C, VSY = 16 V Rev. B | Page 28 of 52 Data Sheet ADA4530-1 THEORY OF OPERATION The ADA4530-1 is an operational amplifier designed to interface with the extremely high impedance sensors used in electrometer applications. The guard pins are connected to the power supplies through Diode D5 and Diode D6. During ESD events, the transient current flows from the input pins through one of the antiparallel diodes and harmlessly into the supplies through one of the power supply diodes. During normal operation, the guard buffer (BUF1) forces the voltage across the antiparallel diodes to 0 V. Resistor R1 shields the guard buffer from potentially large capacitances connected to the guard pins. Its value is nominally 1 k. A metal-oxide semiconductor field effect transistor (MOSFET) input stage eliminates the gate leakage currents associated with legacy junction gate field effect transistor (JFET) electrometers. The ADA4530-1 achieves extremely low input bias currents while simultaneously providing robust protection against ESD damage. A unique ESD diode structure provides protection while also allowing the diodes to be guarded to minimize leakage currents to the input pins. The ADA4530-1 integrates the precision buffer that guards internal ESD diode leakage paths. The output of this guard buffer also connects to external pins, allowing the user to guard external components against leakage currents. INPUT STAGE The input stage comprises a P-channel metal-oxide semiconductor (PMOS) differential pair (M1, M2), folded cascode transistors (M5 to M12), and current source (I1). The ADA4530-1 achieves its high performance specifications by using low voltage MOS devices for its differential inputs. These low voltage MOS devices offer better 1/f noise and bandwidth per unit current compared to high voltage devices. The input stage is isolated from the high system voltages with proprietary protection circuitry. This regulation circuitry protects the input devices from the high supply voltages in which the amplifier can operate. The input bias current is determined by the accuracy of the guard voltage applied across the ESD diodes. The offset voltages of the amplifier and guard buffer set the accuracy of the guard voltage and, therefore, the input bias current. The ADA4530-1 uses Analog Devices, Inc., DigiTrim(R) technology to achieve superior performance. The proprietary high voltage protection circuitry in the ADA4530-1 operates to minimize the common-mode voltage changes seen by the amplifier input stage for most of the input common-mode range. This circuitry results in excellent disturbance rejection when operating in this preferred input common-mode range. The performance benefits of operating within this preferred range are shown in the VOS vs. VCM graphs (see Figure 16 to Figure 18), the small signal CMRR vs. VCM graph (see Figure 21), and the small signal PSRR vs. VCM graph (see Figure 54). DigiTrim trims the offset voltage of the amplifier and guard buffer to reject changes in the common-mode voltage, power supply voltage, and temperature. This technique significantly improves VOS, CMRR, PSRR, and offset voltage temperature coefficient (TCVOS) specifications. Figure 99 shows the simplified schematic of the ADA4530-1. The amplifier uses a three-stage architecture with a fully differential input stage to achieve excellent dc performance specifications. ESD STRUCTURE The input devices are protected from large differential input voltages by the antiparallel ESD diodes (D1 to D4). The diodes can conduct significant current when the differential voltage exceeds 700 mV. The user must ensure that the current flowing into the input pins is limited to the absolute maximum of 10 mA. The input ESD structure consists of Diode D1 to Diode D6. The noninverting input is coupled to the guard pins (GRD) by the D1 and D2 antiparallel diodes. The inverting input is coupled to the guard pins by the D3 and D4 antiparallel diodes. V+ HIGH VOLTAGE PROTECTION M19 M20 M17 M18 D5 I1 +IN M1 D1 D2 M11 M12 M9 M10 M22 M2 GRD C2 C1 M3 M4 R1 OUT V1 BUF1 GRD D3 D4 C3 M7 -IN M8 M5 D6 M6 I2 HIGH VOLTAGE PROTECTION V- Figure 99. Simplified Schematic Rev. B | Page 29 of 52 M16 M13 M14 13405-299 M21 M15 ADA4530-1 Data Sheet GAIN STAGE The second stage of the amplifier comprises an n-channel metal-oxide semiconductor (NMOS) differential pair (M3, M4) and folded cascode transistors (M13 to M20). The amplifier features nested Miller compensation (C1 to C3). OUTPUT STAGE The ADA4530-1 features a complementary common-source output stage consisting of the M21 and M22 transistors. These transistors are configured in a Class AB topology and are biased by the voltage source, V1. This topology allows the output voltage to be within tens of millivolts of the supply rails, achieving a rail-to-rail output swing. The output voltage is limited by the output impedance of the transistors. The output voltage swing is a function of the load current and can be estimated using the output voltage to the supply rail vs. load current graphs (see Figure 40 to Figure 45). GUARD BUFFER The guard buffer (BUF1) is a unity-gain amplifier that creates a low impedance replica of the input common-mode voltage. The buffer input is connected to the noninverting input (IN+). The noninverting input voltage is approximately equal to the input common-mode voltage when the main amplifier feedback loop is settled. The guard buffer uses a three-stage architecture similar to that of the amplifier. The guard buffer uses a rail-to-rail output stage that allows the guard voltage to swing within 100 mV of the supply rails. Because the guard buffer output follows the input common-mode voltage, this output swing limits the effectiveness of the guard buffer at low input common-mode voltages. This limit can be seen as a significant increase in the input bias current at low common-mode voltages shown in the input bias current vs. common-mode voltage graphs (see Figure 22 to Figure 33). For this reason, it is not recommended to operate the circuit with an input common-mode voltage of less than 100 mV from the V- supply rail. The guard buffer output voltage can be degraded from excessive loading. The 1 k output resistance adds 1 V of guard voltage error per 1 nA of load current. It is possible to drive the guard offset voltage out of its specifications with a few tens of nanoamperes of load current. For this reason, it is not recommended to drive anything except insulation resistance (see the Insulation Resistance section and the Guarding section) with the guard buffer. If more drive strength is needed, the guard voltage can be buffered with a low offset, low input bias current op amp such as the ADA4661-2. Rev. B | Page 30 of 52 Data Sheet ADA4530-1 APPLICATIONS INFORMATION INPUT PROTECTION When either input of the ADA4530-1 exceeds one of the supply rails by more than 300 mV, the input ESD diodes become forward-biased and large amounts of current begin to flow through them. Without current limiting, this excessive fault current causes permanent damage to the device. If the inputs are expected to be subject to overvoltage conditions, insert a resistor in series with each input to limit the input current to 10 mA maximum. However, consider the resistor thermal noise effect on the entire circuit. SINGLE-SUPPLY AND RAIL-TO-RAIL OUTPUT The ADA4530-1 is a single-supply amplifier with an input voltage range (IVR) from V- to V+ - 1.5 V. The amplifier has a small keep alive input stage that allows it to function properly when the input common-mode voltage is greater than the specified IVR. This feature allows the ADA4530-1 to start up and recover quickly in certain types of circuits where the IVR is violated at power-up. The ac and dc performance of this keep alive stage is poor; do not rely upon this keep alive stage for normal use. VSY = 8V VIN = 8.3V AV = +1 RL = 10k CL = 10pF 8 6 4 2 0 -2 -4 -6 VIN VOUT -8 13405-300 The ADA4530-1 is suited for applications requiring very low input bias current and low offset voltage, including, but not limited to, preamplifier applications, for a wide variety of current output transducers (such as photodiodes and photomultiplier tubes), spectrometry, chromatography, and high impedance buffering for chemical sensors. 10 -10 TIME (200s/DIV) Figure 100. No Phase Reversal CAPACITIVE LOAD STABILITY The ADA4530-1 can safely drive capacitive loads of up to 250 pF in any configuration. Driving larger capacitive loads than specified can cause excessive overshoot, ringing, or oscillation. A heavy capacitive load reduces phase margin and causes the amplifier frequency response to peak. Peaking corresponds to overshooting or ringing in the time domain. Therefore, it is recommended that external compensation be used if the ADA4530-1 must drive a load exceeding 250 pF. This compensation is particularly important in the unity-gain configuration, which is the worst case for stability. A quick and easy way to stabilize the op amp for capacitive load drive is by adding a series resistor, RISO, between the amplifier output terminal and the load capacitance, as shown in Figure 101. RISO isolates the amplifier output and feedback network from the capacitive load. However, with this compensation scheme, the output impedance as seen by the load increases, and this reduces gain accuracy. +VSY RISO VIN ADA4530-1 -VSY VOUT CL 13405-201 The ADA4530-1 has ultralow input bias currents that are production tested at 25C and 125C to ensure the device meets its performance goals in a system application. An integrated guard buffer is provided to minimize input pin leakage in a PCB design, minimize board component count, and enable ease of system design. The guard buffer output pins are also strategically placed next to the input pins to enable easy routing of the guard ring and to prevent coupling between the inputs, power supplies, and the output pin. Figure 100 shows the input and output waveforms of the ADA4530-1 configured as a unity-gain buffer with a supply voltage of 8 V. The output tracks the input voltage over the entire range until the output voltage is clamped at its maximum output swing. The amplifier still operates even when the signal is outside the specified input voltage range (-8 V IVR +6.5 V); this is due to the keep alive stage. Additionally, the amplifier output does not phase reverse. It is not recommended to apply an input voltage that is outside of the input voltage range. VOLTAGE (V) The ADA4530-1 is a single, electrometer grade, complementary metal-oxide semiconductor (CMOS) operational amplifier with femtoampere input bias current and ultralow offset voltage. It operates over a wide supply voltage range of 4.5 V (or 2.25 V dual supply) to 16 V (or 8 V dual supply). It is a single-supply amplifier; its input voltage range includes the lower supply rail and has a rail-to-rail output. The ADA4530-1 also achieves a low offset voltage of 40 V maximum and offset voltage drift of 0.5 V/C maximum. Figure 101. Stability Compensation with Isolating Resistor, RISO Rev. B | Page 31 of 52 ADA4530-1 Data Sheet EMI REJECTION RATIO Figure 102 shows the phase margin of the ADA4530-1 with different values of output isolating resistors and capacitive loads. Figure 103 shows the frequency response with 1 nF capacitive load and different isolating resistors. Circuit performance is often adversely affected by high frequency electromagnetic interference (EMI). When the signal strength is low and transmission lines are long, an op amp must accurately amplify the input signals. However, all op amp pins--the noninverting input, inverting input, positive supply, negative supply, and output pins--are susceptible to EMI signals. These high frequency signals are coupled into an op amp by various means, such as conduction, near field radiation, or far field radiation. For example, wires and PCB traces can act as antennas and pick up high frequency EMI signals. 80 70 60 Amplifiers do not amplify EMI or RF signals due to their relatively low bandwidth. However, due to the nonlinearities of the input devices, op amps can rectify these out of band signals. When these high frequency signals are rectified, they appear as a dc offset at the output. 50 1000 10000 LOAD CAPACITANCE (pF) To describe the ability of the ADA4530-1 to perform as intended in the presence of electromagnetic energy, the electromagnetic interference rejection ratio (EMIRR) of the noninverting pin is specified in Table 1, Table 2, and Table 3 of the Specifications section. A mathematical method of measuring EMIRR is defined as follows: Figure 102. Phase Margin vs. Load Capacitance with Various Output Isolating Resistors 100 100 VSY = 10V AV = +1 CL = 1nF PHASE 80 40 GAIN 20 -20 -40 1k 20 GAIN (RISO = 0) PHASE (RISO = 0) GAIN (RISO = 301) PHASE (RISO = 301) GAIN (RISO = 732) PHASE (RISO = 732) 10k PHASE (Degrees) 40 0 EMIRR = 20log(VIN_PEAK/VOS) 60 where VIN_PEAK is the peak amplitude of the input voltage. Figure 104 shows the typical EMIRR vs. frequency performance for each of the specified supply voltages. 120 0 VSY = 4.5V TO 16V 110 -20 100k 1M 100 -40 10M 13405-303 GAIN (dB) 60 80 FREQUENCY (Hz) Figure 103. Frequency Response with CL = 1 nF and Various Isolating Resistors 90 80 70 60 50 40 VSY = 4.5V VSY = 10V VSY = 16V 30 20 10M 100M 1G FREQUENCY (Hz) Figure 104. EMIRR vs. Frequency Rev. B | Page 32 of 52 10G 13405-304 40 100 13405-302 RISO = 301 RISO = 499 RISO = 732 EMIRR (dB) PHASE MARGIN (Degrees) VSY = 10V AV = +1 Data Sheet ADA4530-1 HIGH IMPEDANCE MEASUREMENTS The ADA4530-1 is typically used in two kinds of circuits: a buffer and a transimpedance amplifier (TIA). Buffer circuits are useful for measuring voltage output sensors with high output resistance. Some example sensors include pH probes and reference electrodes (RE) in coulometry control loops. TIA circuits are useful for converting the signal from a current output sensor to an output voltage. Some example sensors include photodiodes and ion chambers. The following sections describe some of the most important error sources when using the ADA4530-1 in these circuits. Simplified models with error sources are provided for the buffer (see Figure 105) and the TIA (see Figure 106). The buffer circuit models the voltage output sensor as a voltage source (VSRC) with an output resistance (RSRC). The voltage on the A terminal is sensed by Pin 1 of the ADA4530-1 in a noninverting gain configuration (or a unity-gain configuration). The B terminal is driven to a suitable reference voltage (signal ground in this case). If all error sources are ignored, the output of the circuit is as follows: R VOUT = VSRC 1 + F RS VOLTAGE SENSOR ADA4530-1 A 1 +IN RIN RSRC IB+ 8 VSRC OUT VOUT 6 RSHUNT -IN VOS RF 13405-310 RS B If all error sources are ignored, the output of the circuit is as follows: VOUT = ISRC x RF V+ IRF RF RSHUNT CURRENT SENSOR ADA4530-1 A 8 RIN ISRC RSRC IB- 6 1 B VOS VOUT 13405-311 The ADA4530-1 is designed to maximize the performance of very high impedance circuits. Its performance advantages make it useful for circuit impedances ranging from 100 M to over 10 T. Measurements of high impedance circuits are subject to a number of error sources. General information about making measurements from high resistance sources can be found in the Low Level Measurements Handbook, sixth edition (Keithley Instruments, Inc., 2004). Figure 106. TIA Circuit INPUT BIAS CURRENT The input bias current of the amplifier is a major error source in high impedance electrometer circuits. Like other semiconductor amplifiers, the input bias current of the ADA4530-1 has an exponential dependence on temperature. The input bias current of the ADA4530-1 increases by a factor of 2.5 for every 10C increase in temperature. Refer to the input bias current vs. temperature graphs (see Figure 34 to Figure 36) for typical temperature performance. Notice that the exponential diode currents cease to be the dominant contributor to the input bias current at temperatures below 60C to 70C. The residual 100 aA to 200 aA (aA = 10-18 A) bias currents are dominated by other leakage paths that are highly sensitive to environmental conditions. These vanishingly small bias currents require highly controlled laboratory conditions to measure. Most practical applications are dominated by other errors, and the ADA4530-1 input bias current can be considered to be zero for temperatures less than 70C. The input bias current of the ADA4530-1 can only be guaranteed to 20 fA due to the measurement limitations of a production environment, even though the achievable input bias currents are more than an order of magnitude lower. The input bias current affects the buffer circuit by loading down the voltage sensor. The input bias current is forced to flow through the output resistance of the sensor, which creates an error voltage, VERR. VERR = IB+(RSRC) Figure 105. Voltage Buffer Circuit The TIA circuit models the current output sensor as a current source (ISRC) with a shunt resistance (RSRC). The current from the A terminal is connected to the inverting input pin of the ADA4530-1 and the feedback resistor (RF). The B terminal and the noninverting input of the amplifier are driven to a suitable reference voltage (signal ground in this case). The negative feedback of the circuit suppresses any voltage changes at the A terminal. This suppression is accomplished by forcing all current through the feedback resistor. The magnitude of this voltage error can be significant with very high impedance sensors operating at high temperature. For example, the input bias current can generate a maximum voltage error of 25 mV from a 100 G sensor operating at 125C. The input bias current affects the TIA circuit by summing together with the sensor current. Both of these currents flow through the feedback resistor to generate the output voltage as follows: Rev. B | Page 33 of 52 VOUT = (ISRC + IB-)RF ADA4530-1 Data Sheet The magnitude of the input bias current limits how small of a signal current can be resolved accurately. For example, if the acceptable error level is 10%, the minimum measurable signal current is 2.25 pA for a circuit operating at 125C. ISRC = IB-(1/err - 1) where err is the error level. 1 2.25 pA = 250 fA - 1 0. 1 equal to a few fA, which is the same magnitude as the input bias current itself. These uncertainties make it impossible to calculate input resistances higher than a few hundred teraohms. The input resistance affects the buffer circuit by loading down the voltage sensor. This resistance acts as a voltage divider so the voltage measured by the amplifier is some fraction of the unloaded voltage of the sensor. This voltage drop is calculated as follows: VA = VSRC INPUT RESISTANCE The input resistance of the amplifier is another error source that must be considered. Input resistance typically has two components: differential and common mode. The differential input resistance is suppressed by the negative feedback of the circuit. The ADA4530-1 has enough gain that the differential input resistance is much too large to measure. The common-mode input resistance (hereafter referred to as input resistance) is a more important error source. The input resistance is equal to the change in input bias current relative to the change in input voltage. This change is not caused by a physical resistance inside the ADA4530-1; it is the result of a complex relationship between the accuracy of the guard voltage across the ESD structures and the input common-mode voltage; that is, the input resistance changes with common mode voltage. It is also possible for the input resistance to be negative. Negative input resistance means the input bias current decreases as the common-mode voltage increases. The input resistance, RIN, can be approximated by calculating the slope of the input bias current vs. common-mode voltage graphs (see Figure 22 to Figure 33). For example, the noninverting input resistance can be calculated at 125C from Figure 32. The input bias current changes by approximately 20 fA for commonmode voltages from 4 V to 6 V. RIN = RIN = VCM I B + 2V 20 fA RIN RIN + RSRC Consider the previous example of a 100 G sensor operating at 125C. The 100 T input resistance causes the measured voltage to equal 99.9% of the actual voltage, a 0.1% gain error. The input resistance has much less of an effect on the TIA circuit. The input common-mode voltage does not change in this circuit; therefore, the error created is vanishingly small. The input resistance affects the noise gain of the circuit, which changes the input offset voltage error (see the Photodiode Interface section for more information). INPUT OFFSET VOLTAGE The input offset voltage of the amplifier affects the buffer circuit by adding directly to the voltage output of the sensor. This error is typically much smaller than other errors. The input offset voltage affects the TIA circuit in a different manner. The burden voltage of the TIA is equal to the input offset voltage. This burden voltage appears between the A and B terminals. An error current is created by applying this burden voltage across the sensor shunt resistance. For sensors with low output resistances such as photodiodes, this error can be significant. Consider a sensor with a 1 G output resistance. The 50 V maximum offset voltage of the ADA4530-1 creates a 50 fA error current. INSULATION RESISTANCE = 100 T The slope of the curves in the input bias current vs. commonmode voltage graphs increases rapidly outside the preferred common-mode range (see Figure 22 to Figure 33). The input resistance drops rapidly outside this range. This drop in input resistance must be considered before operating these circuits with input voltages close to the V- power supply. Like the input bias current, the input resistance has a strong temperature dependence. At lower temperatures, the amplifier input resistance is so high that it is dominated by other error sources. It is important to recognize the limitations of calculating input resistance at lower temperatures. Measurement uncertainties make it difficult to accurately calculate the IB term. Consider the 85C input bias current vs. common-mode voltage graphs (see Figure 22 to Figure 27): the measurement uncertainties are The ADA4530-1 has such low input bias current and such high input resistance that the insulation resistance of the materials that are used to construct the circuit is often the largest error source. Any insulators with finite resistance that come in contact with the high impedance conductor contribute to the error current. Some examples include the PCB laminate material, cable, and connector insulation. The physical insulation resistance is distributed across the entire contact surface of the high impedance conductor, and it can end at several different conductors at different potentials. It is useful to make a simple model where all of these resistance paths are lumped into a single resistor. This lumped element is shown as RSHUNT in the voltage buffer circuit (see Figure 105). The insulation resistance affects the buffer circuit in the same way as the amplifier input resistance. This resistance acts as a voltage divider so that the voltage measured by the amplifier is some fraction of the unloaded voltage of the sensor. Rev. B | Page 34 of 52 Data Sheet ADA4530-1 The effect on the insulation resistance on the TIA circuit depends on the leakage path. The insulation resistance between the A terminal and B terminal of the current sensor affects the circuit in the same way as the amplifier input resistance. This error is extremely small because the voltage across the insulation is equal to the offset voltage of the amplifier. A much more significant error is created from insulation paths to conductors with significantly different potentials. This type of leakage path is shown as a lumped element, RSHUNT, in the TIA circuit (see Figure 106). In this example, the leakage path is created from the positive supply voltage (V+) to the A terminal. If the positive supply voltage is 5 V relative to signal ground, 500 fA flows through the insulation resistance of 10 T. This large error dominates the amplifier input bias current and input resistance errors over the entire temperature range. Leakage paths to high voltages can also affect the buffer circuit with equally ruinous results. GUARDING High source impedances and low error requirements can create insulation resistance requirements that are unrealistically high. Fortunately, a technique called guarding can reduce these requirements to a reasonable level. The concept of guarding is to surround the high impedance conductor with another conductor (guard) that is driven to the same voltage potential. If there is no voltage across the insulation resistance (between high impedance conductor and guard), there cannot be any current flowing through it. The ADA4530-1 uses guarding techniques internally, and it has a very high performance guard buffer integrated. The output of this buffer is made available externally to simplify the implementation of guarding at the circuit level. The voltage buffer circuit (see Figure 105) has been modified to show the implementation of the guard (see Figure 107). In this model, a conductor (VGRD) is added, and it completely separates the high impedance (A) node from the low impedance (B) node at a different voltage. The insulation resistance is modeled as two resistances: all of the insulation between the A conductor and the guard conductor (RSHUNT1), and all of the insulation between the guard conductor and the B conductor (RSHUNT2). The ADA4530-1 guard buffer then drives this guard conductor (through Pin 2 and Pin 7) to the A terminal voltage. If the A node and VGRD node are exactly the same voltage, no current flows through the RSHUNT1 insulation resistance. In practice, the voltage across RSHUNT1 cannot be 0 V, the guard buffer offset voltage contributes to the difference in voltage potential between the A node and VGRD node. For the ADA4530-1, this offset voltage is trimmed to provide offsets less than 100 V when the input common-mode voltage is 1.5 V from the supply rails. The guard buffer offset voltage and drift are specified in Table 1, Table 2, and Table 3. For example, assume that the voltage sensor produces an output of 1 V. Without guarding, the 10 T insulation resistance creates an error current of 100 fA. With the guard, the voltage across the insulation resistance is limited to 100 V. The guard limits the error current to 0.01 fA. In this example, the guard reduces the error by a factor of 104 to an insignificant level. VOLTAGE SENSOR ADA4530-1 A 1 RSHUNT1 RSRC 2 VGRD 8 VSRC VOUT 6 7 RSHUNT2 RF RS B 13405-313 This error is significant because it is very difficult to maintain high insulation resistance values in glass epoxy (such as FR-4) PCB materials. Resistance values of 10 T to 100 T are achievable. A 10 T insulation resistance creates a 1% error with the 100 G sensor used in previous examples. Insulation resistance does not have an exponential temperature dependence like the amplifier errors previously discussed in the Input Bias Current section and the Input Resistance section, which makes insulation resistance the dominate error source at lower temperatures (less than 70C). Figure 107. Voltage Buffer Circuit with Guard DIELECTRIC RELAXATION Dielectric relaxation (also known as dielectric absorption or soakage) is a property of all insulating materials that can limit the performance of electrometer circuits that need to settle to a few femtoamperes. Dielectric relaxation is the delay in polarization of the dielectric molecules in response to a changing electric field. This delay is a property of all insulating materials. The magnitude and the time constant of the delay depend on the specific dielectric material. The delays in some materials can be minutes or even hours. Dielectric relaxation is a problem for electrometer circuits because small displacement currents flow through the insulator in response to the polarization of the molecules. Delays in polarization cause delays in the dissipation of these currents, which can dominate the settling time in these circuits. In the context of capacitors, dielectric relaxation is called dielectric absorption. Capacitors are specified with a test that measures the residual open-circuit voltage after a specific charge/discharge cycle. For electrometer circuits, it is more useful to consider the short-circuit currents produced from step changes in a test voltage. A simple lumped circuit model of an insulator is connected to the test voltage source (see Figure 108). The majority of the dielectric polarizes instantly; this is modeled as Capacitor C1. A small percentage of the dielectric polarizes slowly with a time constant of 2, modeled as Capacitor C2 and Resistor R2. Rev. B | Page 35 of 52 ADA4530-1 Data Sheet INSULATOR MODEL RS R2 12 50 8 25 4 0 0 -25 -4 -50 -8 -75 -12 -100 -16 C2 13405-326 VSRC -125 Figure 108. Dielectric Relaxation Model Test Circuit 0 The current step response (ISRC) of the insulator to a voltage step is shown in Figure 108. A large initial current charges Capacitor C1 with a fast time constant. This time constant, 1, equals the source resistance RS x C1 (see Figure 109). Long after Capacitor C1 is charged, a much smaller current continues to flow, which charges Capacitor C2. The time constant of charging is not affected by the external circuitry; it depends only on the material properties of the insulator. The magnitude of the current depends on the magnitude of the voltage change across the insulator. 0.5 1.0 1.5 2.0 2.5 ELAPSED TIME (Hours) 3.0 3.5 -20 4.0 Figure 110. Glass Epoxy Dielectric Relaxation Performance An alternative PCB laminate to consider is Rogers 4350B. Rogers 4350B is a ceramic laminate designed for RF/microwave circuits. Rogers 4350B is compatible with standard PCB production techniques and is widely available. The measurement results for the Rogers 4350B material are shown in Figure 111. This material requires less than 20 sec to dissipate the dielectric relaxation current to less than 1 fA. Based on its superior performance, it is recommended to use Rogers 4350B laminates with the ADA4530-1 in the highest performance applications. All of the critical characterization measurements of the ADA4530-1 are taken using Rogers 4350B. 125 10 VSRC ISRC 100 Figure 109. Step Response of Dielectric Relaxation Model The dielectric relaxation performance was measured for a variety of PCB laminates using the test circuit in Figure 108. An electrometer grade source measurement unit (SMU), Keithley 6430, applies a 100 V test stimulus and measures the resulting current. The large alternating polarity test voltage distinguishes the small dielectric relaxation current from the input offset current of the SMU. APPLIED VOLTAGE (V) TIME 13405-314 2 = R2 x C2 8 75 6 50 4 25 2 0 0 -25 -2 -50 -4 -75 -6 -100 -8 -125 The first PCB laminate tested is the industry-standard FR-4 glass epoxy. The measurement results are shown in Figure 110. The glass epoxy laminate requires 1 hour to dissipate the dielectric relaxation current to less than 10 fA. This result shows that glass epoxy laminates are unsuitable for the highest performance electrometer circuits. Rev. B | Page 36 of 52 0 20 40 60 80 100 120 140 ELAPSED TIME (Seconds) 160 180 -10 200 Figure 111. Rogers 4350B Dielectric Relaxation Performance CURRENT (fA) 1 = RS x C1 13405-316 VSRC ISRC 16 75 13405-315 C1 20 VSRC ISRC 100 APPLIED VOLTAGE (V) 2 = R2 x C2 ISRC 125 CURRENT (fA) The size of C2 reflects the proportion of slow molecules. The size depends on the material but it is typically 100 to 10,000 times smaller than C1. The size of R2 sets the time constant. Data Sheet ADA4530-1 HUMIDITY EFFECTS The insulation resistance of the materials used to construct circuits is sensitive to moisture. At lower temperatures (<70C) the insulation resistance creates more significant leakage current errors than the amplifier itself. This means that the relative humidity of the air is the most important error source at lower temperatures. The dependence on humidity is evident in the input bias current vs. temperature graphs (see Figure 34 to Figure 36). There is significant deviation in the low temperature measurements due to the difficulty maintaining a consistently low relative humidity at low temperatures. The ADA4530-1 amplifiers are mounted on Rogers 4350B PCBs (glass epoxy boards have poor humidity absorption properties). Figure 112 shows the effective input bias current vs. relative humidity for seven characterization units. Figure 112 is plotted with a split log axis to effectively show the magnitude and polarity of the bias current. The magnitude of the leakage currents changes by more than a factor of 100 across the relative humidity span from 5% to 80%. The effective bias current is much less than 1 fA for typical conditioned environments (RH < 50%). 100 VSY = 10V TA = 25C VCM = VSY/2 7 UNITS 10 To evaluate the humidity sensitivity of insulation resistance, there are two mechanisms which must be considered: adsorption and absorption. 1 IB (fA) 0.1 Adsorption is a process where thin films of molecules adhere to the surface of a material. Water molecules are subject to this process. The magnitude of the effect depends on the insulating material and the relative humidity. Thin films of moisture are conductive, and they act as leakage resistances in parallel with the insulation resistance of the material. Because this is a surface effect, guard ring techniques are effective at reducing it. It is not possible to completely guard all the leakage paths: bulk or surface. A relevant example of this limitation is the molding compound of the SOIC package housing the ADA4530-1. Surface and bulk paths exist from the input pins to all other pins of the package. The nature of the resulting current depends on the specific leakage path: paths to V+ increase the bias current flowing out of the amplifier, paths to V- increase the bias current flowing into the amplifier, and paths to VOUT lower the effective feedback resistance in TIA circuits. Consider the example of a circuit powered from 5 V power supplies with an input common-mode voltage of 0 V. Assume that all leakage resistance between the input and V+ is effectively 100 T. This resistance creates a current equal to 50 fA flowing from V+. Assume that the leakage resistance between the input and V- is effectively 250 T. This resistance creates a current equal to 20 fA flowing to V-. The net current equals -30 fA flowing out of the input pin. All of these leakage currents can be combined with the amplifier input bias current and treated as an effective input bias current. The effective input bias current sensitivity to relative humidity of the ADA4530-1 is characterized for several units. The test amplifiers are configured in TIA and unity buffer circuits with 100 G, hermetically sealed resistors (RX-1M1009FE) as the feedback and source resistors, respectively. These glass bodied resistors have a silicone coating (glass has poor humidity adsorption properties). -1 -10 -100 -1000 0 20 40 60 80 RELATIVE HUMIDITY (%) 13405-112 Absorption is a process where molecules enter the bulk of a material. Water molecules can diffuse into a material and affect the bulk conductivity of that material. Because the leakage paths are through the bulk of the material, guard rings are not effective at reducing it. -0.1 Figure 112. Effective Input Bias Current vs. Relative Humidity The magnitude of the effective input bias current becomes very sensitive to the relative humidity at higher humidity levels (>60%). Some of the units show an exponential dependence on humidity (see the blue curve in Figure 112). Other units show a less predictable dependence; the leakage current magnitude increases rapidly, but the polarity can change. The net leakage current is the sum of the currents sourced from higher voltages (like V+) with the currents sunk by lower voltages (like V-). As the humidity changes, the relative magnitudes of each of these leakage paths can change, which can result in changes in the polarity of the leakage current (see the red and green curves in Figure 112). The response time of these leakage currents depends on the physical process that causes them. Because adsorption is a surface effect, the film thickness rapidly achieves equilibrium with changes in the relative humidity of the air. Because absorption is a bulk diffusion process, it is very slow compared to the adsorption process. These widely different time constants mean that the effective input bias current responds quickly to a step change in relative humidity, but has a very long settling time. The step response of one amplifier to a 50% to 60% relative humidity change is shown in Figure 113. The high frequency response of the initial humidity step (and the overshoot recovery) is on the order of seconds to tens of seconds. Complete settling takes over a week as the moisture slowly diffuses through the PCB insulation and package molding compound. Each data point in Figure 112 was taken after one week of settling time. Rev. B | Page 37 of 52 ADA4530-1 Data Sheet 6 70 4 60 2 VSY = 10V TA = 25C 50 VCM = VSY/2 0 40 -2 30 -4 20 RELATIVE HUMIDITY (%) IB (fA) The rapid adsorption response can change the effective bias current in response to local fluctuations in humidity. These current fluctuations can be much larger than the low frequency current noise of the amplifier and thermal noise of the resistors. The sensitive circuitry can be isolated from these local humidity fluctuations by restricting the airflow around the circuitry with an air baffle. Electrostatic shielding added to reduce interference can also function as an air baffle. Remove or reduce the sources of humidity fluctuations whenever possible. Avoid breathing on the high impedance circuitry, for example. The effective insulation resistance of an electrometer circuit can be substantially degraded if the insulators are contaminated. Solder flux, body oils, dust, and dirt are all possible sources of contamination. Some of these contaminants form a parallel leakage path across the surface of the existing insulator, effectively lowering the insulation resistance. Guarding techniques help to suppress these effects. The effects are more severe when the source of contamination contains ionic compounds. In the presence of humidity, these contaminants act as an electrolyte, which can form a weak battery. Flux residue and body oils are particularly effective at creating these parasitic batteries. As an example, the PCB insulation between two high impedance nodes was purposefully contaminated with a 3 mm drop of rosin mildly activated (RMA) type solder flux. This sample was dried and allowed to stabilize in laboratory conditions (25C, 40% RH) for several days. After this time, the voltage vs. current relationship was measured with an electrometer grade SMU (see Figure 114). 400 300 200 100 0 -100 -200 -300 -400 -100 -80 -60 -40 -20 0 20 40 APPLIED VOLTAGE (mV) 60 80 100 13405-319 In this environment, measurements of the effective input bias current appear to drift with time because the leakage currents depend on the relative humidity for the previous week. This long-term memory due to the absorption process may need to be taken in account in certain circumstances (such as long-term product storage in an unconditioned high humidity environment prior to use). CONTAMINATION OUTPUT CURRENT (fA) In practical applications, the relative humidity of the air changes rapidly with daily and seasonal variations. The effective input bias current response to these humidity changes has two parts. The response due to the adsorption process follows the rapid changes immediately. The response due to the absorption process low-pass filters the humidity changes. This low-pass response causes the effective input bias current to have longterm memory of the relative humidity fluctuations. Figure 114. Current to Voltage Response of RMA Contaminated Insulation -6 10 -8 0 25 50 75 100 125 ELAPSED TIME (Hours) 150 0 175 13405-402 RELATIVE HUMIDITY IB Figure 113. Effective Input Bias Current Transient Response to Humidity Step It is important to note that all electrometer circuits are subject to humidity effects. The legacy circuits constructed with TO-99 packages using air wiring techniques have insulator leakage paths such as the epoxy between the pins and the Teflon(R) standoffs that support the air wired components. The input bias currents of legacy amplifiers are high enough to mask the humidity effects. In summary, the ADA4530-1 can be designed using the specified performance for normal laboratory (<60%) relative humidity conditions. In applications that must operate in uncontrolled or high humidity environments, some additional derating of the input bias current is prudent. Characterize the amount of derating on a per product basis because the net leakage depends on the material types and physical dimensions of the insulators. This contamination formed a weak battery with an open circuit voltage (VBATT) of 15 mV and an output resistance (RBATT) of 300 G. This sort of contamination is disastrous in electrometer circuits because guarding techniques cannot suppress it. A simplified model is made with the contamination battery applied across the A terminal and B terminal of a TIA circuit (see Figure 115). The A terminal and B terminal are both driven to the same voltage, which creates an error current (IBATT) because the open circuit battery voltage is dropped across the output resistance as follows: IBATT = VBATT / RBATT This battery current flows through the feedback resistance, where it is summed with the signal and other error currents in the circuit. The error current in this example is 50 fA. The battery characteristics are subject to the environmental conditions; therefore, the error current drifts with time, temperature, and humidity. Rev. B | Page 38 of 52 Data Sheet ADA4530-1 highly effective. Scrubbing the area around the high impedance insulators with an acid brush also works. Flush the insulators with a final wash of fresh IPA to remove any contaminants suspended in the solvent. IBATT RF ADA4530-1 A 8 RBATT BATTERY MODEL V BATT The residual moisture must completely evaporate before the insulator can be used. This evaporation can take many hours at room temperature; the process can be accelerated by baking the insulators in an oven at elevated temperature. 6 1 B 13405-312 VOUT For detailed cleaning and handling procedures, refer to the ADA4530-1R-EBZ user guide. Figure 115. TIA Circuit with Contamination Battery SOLDER PASTE SELECTION CLEANING AND HANDLING The contamination described in the Contamination section can usually be removed by an appropriate cleaning process. Solvents like isopropyl alcohol (IPA) are effective at removing the residue from solder fluxes and body oils. Use high purity cleanroom grade solvents to ensure that there is no additional contamination from the solvent itself. Solder paste selection can drastically impact the performance of the board if not cleaned properly. Solder flux residue on a PCB degrades the low IB performance of the amplifier. An experiment was performed to evaluate the cleaning procedure for different solder paste types. Table 7 shows the result of the experiment. The recommended cleaning procedure column lists the times required to restore the effective input bias currents to less than 1 fA. The suggested solder paste of choice is the RMA type. Insulators that are severely contaminated benefit from mechanical abrasion in addition to the solvent. Ultrasonic cleaners are Table 7. Recommended Cleaning Procedures for Different Solder Paste Material Solder Paste Type RMA Solder Paste Part Number AIM RMA258-15R Water Soluble SAC305 Shenmao No Clean SAC 305 AMTECH LF4300 1 Recommended Cleaning Procedure1 15 min clean time in an ultrasonic cleaner with fresh IPA, followed by 1.5 hours of bake time at 125C 1.5 hours clean time in an ultrasonic cleaner with fresh IPA, followed by 1.5 hours of bake time at 125C 3 hours clean time in an ultrasonic cleaner with fresh IPA, followed by 3 hours of bake time at 125C Bake time was not optimized and was set equal to the cleaning time. Rev. B | Page 39 of 52 ADA4530-1 Data Sheet CURRENT NOISE CONSIDERATIONS The current noise from an amplifier input pin is important when it flows through an impedance and generates a voltage noise. If the current noise and impedance are large enough, the resulting voltage noise can dominate the other noise sources in the circuit, such as the voltage noise of the resistors and amplifier. For an electrometer amplifier, such as the ADA4530-1, the typical circuit impedances are so large that the current noise of the amplifier can be the most important noise source. To measure current noise, it is necessary to flow the noise current through a test impedance large enough that the resulting noise voltage is larger than the other noise voltages in the circuit. Practically, this test impedance is usually a resistor. All resistors have thermal noise. The value of thermal noise is usually presented as an output referred voltage noise spectral density (NSD), VNRTO. VNRTO = (4kTR) where: k is Boltzmann's constant. T is the temperature in Kelvin. R is the resistance value. The resistor thermal noise can be interpreted as a current NSD by dividing the thermal noise with the resistance value, R, per Ohm's Law. Table 8 shows the thermal noise of a series of resistor values presented as both voltage and current noise. The current noise of a resistor decreases as the resistance increases. This surprising result illustrates that it is necessary to use high valued resistors to measure low levels of current noise. Table 8. Resistor Thermal Noise Resistor Value 1 M 100 M 10 G 1 T Voltage Noise 128 nV/Hz 1.28 V/Hz 12.8 V/Hz 128 V/Hz Current Noise 128 fA/Hz 12.8 fA/Hz 1.28 fA/Hz 128 aA/Hz The measurement setup used to gather the current noise data is shown in Figure 116. The ADA4530-1 is configured as a TIA with a large value feedback resistor, RF. All amplifier current noise from the inverting input flows through Resistor RF to produce a voltage noise at VOUT. CF RF ADA4530-1 The output referred voltage NSD, VNRTO, is sampled by the SR785 high performance dynamic signal analyzer (DSA), and is equal to the root-sum-square of the amplifier current noise multiplied by RF, the resistor thermal noise, and amplifier voltage noise. VNRTO = ((IN-RF)2 + 4kTRF + VN2) where: IN- is the amplifier inverting current noise. 4kTRF is the resistor thermal noise. VN2 is the amplifier voltage noise. Calculate the current noise of the amplifier from VNRTO as follows: IN - = VNRTO 2 - 4kTRF - VN 2 (1) RF For Equation 1 to be valid, the measured noise must be somewhat larger than the resistor thermal noise plus the amplifier voltage noise. In practice, ensure that the resistor current noise is less than or equal to the amplifier current noise. For example, if the amplifier noise is expected to be 2 fA/Hz, use a value of RF that is at least 10 G, according to Table 8. The amplifier voltage noise is not a concern at most frequencies because the resistor thermal noise is much larger than the amplifier voltage noise. At very low frequencies, this assumption is not valid due to the 1/f characteristic of the amplifier voltage noise. It is important to consider the bandwidth limitations of the current noise measurement system shown in Figure 116. The presence of stray capacitance makes it impossible to maintain the high impedances required for the measurement. All stray capacitances that couple the amplifier output to the inverting input can be lumped into a single capacitor, CF, as shown in Figure 116. The current noise must pass through RF to become voltage noise. However, in practice, the current noise passes through the parallel combination of RF and CF to become voltage noise. At frequencies higher than the RFCF pole, most of the noise current flows through the capacitor and current noise calculations at these frequencies are error prone due to the distributed parasitic nature of CF. A good guideline is to set the measurement bandwidth limit equal to the RFCF pole frequency. The measurement bandwidth limitations for high valued resistors can be surprisingly low. Table 9 shows the -3 dB bandwidth of a series of resistor values with a practical minimum stray capacitance value. Table 9. Bandwidth Limitations SR785 DSA 13405-305 VOUT Figure 116. Current Noise Measurement Setup Resistor Value 1 M 100 M 10 G 1 T Rev. B | Page 40 of 52 Capacitor Value 100 fF 100 fF 100 fF 100 fF -3 dB Bandwidth 1.59 MHz 15.9 kHz 159 Hz 1.59 Hz Data Sheet ADA4530-1 Current Noise Density 128 aA/Hz 1.28 fA/Hz 12.8 fA/Hz 128 fA/Hz Bandwidth 1.59 Hz 159 Hz 15.9 kHz 1.59 MHz Table 10 demonstrates the error of the often claimed 0.1 fA/Hz at 10 kHz presented in the specifications of low input bias current amplifiers. Measuring this value requires a 1 T resistor with less than 15.9 aF (15.9 x 10-18 F) of stray capacitance, which is impossible. These kinds of claims are simply shot noise calculations based on the specified input bias currents of a few tens of femtoamperes. Calculate the shot noise of a semiconductor as follows: Shot Noise = (2qIB) where: q is the charge on an electron. IB is the current flowing through a junction. Shot noise calculations are appropriate only for some legacy JFET-based electrometer amplifiers, where only a single junction is connected to the amplifier input pins. Modern high impedance amplifiers have several semiconductor junctions connected to the amplifier input pins. The most significant of these junctions are the ESD diode structures. The input bias currents are equal to the sum of these diode currents. The diode currents are designed to cancel each other, but the shot noise currents are uncorrelated and cannot cancel, which makes calculating the shot noise from the input bias current impossible. Even when appropriate, these shot noise calculations neglect all capacitive coupling effects so that they are valid only at very low frequencies. The gate to source capacitance of the input transistors couples noise currents from sources other than the input junctions for frequencies above a few tens of hertz. This blowback noise effect is present in all amplifiers, and it ensures that the current NSD always increases as frequency increases. It is critically important to use high quality resistors during this measurement. Many high valued resistors designed for high voltage operation are nonlinear at low voltage levels and are not suitable for electrometer work. Inferior resistors can also have their own 1/f noise that corrupts the measurement results. Table 11 lists the resistors used for the characterization of the ADA4530-1. Table 11. Test Resistor Device Numbers Resistor Value 100 M 1 G 10 G 100 G 1 T Manufacturer Vishay Ohmite Ohmite Ohmite Ohmite Device Number RNX050100MDHLB RX-1M1007GE RX-1M1008JE RX-1M1009FE RX-1M100AKE Figure 117 shows the output referred voltage NSD (VNRTO) of the transimpedance test circuit for the test resistors listed in Table 11. The calculated thermal noise for each resistor is represented with the dashed line. The black dashed line represents the 1/f voltage noise of the amplifier. VSY = 10V AV = +1 1T 100 100 100G 10G 10 10 1G 100M 1/f 1 0.001 0.01 0.1 1 10 100 FREQUENCY (Hz) 1k 10k 1 100k INPUT SERIES RESISTOR THERMAL NOISE (V/Hz) Table 10. Measurement Current Noise Density vs. Bandwidth 13405-306 It is useful to construct a table that combines the resistor noise and measurement bandwidth guidelines. Table 10 shows the approximate bandwidth limitations for a variety of input current noise measurements. The complex relationship between current noise, feedback resistance, and bandwidth means that the correct way to characterize the current noise of an electrometer amplifier is by measuring the output NSD with a variety of feedback resistors that cover the entire span of values used in the end applications. Each feedback resistor establishes a boundary for the minimum measurable current noise over a range of frequencies. TOTAL OUTPUT VOLTAGE NOISE DENSITY (V/Hz) Reconsider the example amplifier with 2 fA/Hz of current noise. The required RF value of 10 G also limits the measurement bandwidth to 159 Hz according to Table 9. Figure 117. Transimpedance NSD Referred to Output VNRTO is dominated by the resistor noise for all of the test resistors up 1 T. This means that the current noise contribution from the ADA4530-1 is insignificant relative to the thermal noise of these resistors. Rev. B | Page 41 of 52 ADA4530-1 Data Sheet 100M 1G 600 500 400 300 200 100 0 0 10G 1 0.1 0.01 0.1 1 10 100 FREQUENCY (Hz) 20 30 40 50 60 TEMPERATURE (C) 70 80 90 Figure 119. Current Noise Density vs. Temperature 100G 1T 0.01 0.001 10 1k 10k 100k Figure 120 shows the equivalent noise resistance vs. temperature. From Figure 120, it is easy to determine that the ADA4530-1 contributes less noise than a 1 T resistor for temperatures less than 40C. If the application requires an 85C operation, the ADA4530-1 contributes as much noise as a 30 G resistor. This example illustrates the considerable impact of the temperature in determining the noise performance of an application. Figure 118. Current Noise Spectral Density The current noise of the ADA4530-1 originates from the saturation current of the ESD diodes. The diode saturation current has an exponential dependence on temperature; therefore, it is expected that the current noise tracks this temperature behavior. The current noise of the ADA4530-1 is characterized over temperature using the transimpedance measurement circuit with a 1 T resistor. The measurements are limited to 85C because of the maximum operating temperature of the resistor. Figure 119 shows the current noise density at a frequency of 0.1 Hz for all of the test temperatures. It can be useful to calculate an equivalent noise resistance from the current noise density data in Figure 119. This conversion facilitates comparisons between the current noise generated by the ADA4530-1 and the thermal noise of the feedback resistor used in the circuit. EQUIVALENT NOISE RESISTANCE () 10T f = 0.1Hz 1T 100G 10G 0 10 20 30 40 50 60 TEMPERATURE (C) 70 80 90 13405-309 10 VSY = 10V f = 0.1Hz 700 13405-308 VSY = 10V AV = +1 13405-307 CURRENT NOISE DENSITY (fA/Hz) 100 800 CURRENT NOISE DENSITY (aA/Hz) It is possible to calculate the current noise of the ADA4530-1 with the 1 T resistor. This result is shown in Figure 118. It is impossible to calculate the amplifier current noise for all of the other resistors, because, at those resistor values, the resistor noise is much greater than the amplifier current noise. The current noise densities of each of the test resistors are plotted as dashed lines in Figure 118. The current noise of the ADA4530-1 is below the resistor noise values. Figure 120. Equivalent Noise Resistance vs. Temperature In summary, the excellent noise performance of the ADA4530-1 makes it ideal for electrometer applications. For impedances less than 1 T, the amplifier noise is negligible. Also, unlike other amplifiers, the current noise is fully characterized and is free of excessive blowback noise. Rev. B | Page 42 of 52 Data Sheet ADA4530-1 LAYOUT GUIDELINES In the Guarding section, guarding is introduced as a technique fundamental to high impedance work. The goal of guarding is to completely surround the insulation of the high impedance node with another conductor that is driven to the guard voltage. This ideal is impossible to achieve in practice; however, there are several practical structures that provide good performance. GUARD RING A guard ring is a structure typically used to implement the guarding technique on the surface of the PCB. A simplified layout of the buffer circuit implements the guard ring around the high impedance (A) trace (see Figure 121). The output of the voltage sensor is wired directly to the A and B pads in Figure 121. The guard ring is a filled copper shape that completely surrounds the high impedance (A) trace from the sensor connection to the noninverting input (Pin 1). The guard ring is driven directly from the ADA4530-1 guard buffer (Pin 2) through a thermal relief shape connection. It is not necessary to connect the other guard buffer output (Pin 7). The solder mask is removed from the high impedance trace and the guard trace to ensure that the guard makes electrical contact with any surface leakage paths. For the same reason, avoid printing any silkscreen in this section. The magnitude of this error current is dependent on the area of the exposed high impedance insulation. A gap of 15 mil between the A trace and the guard ring is sufficient. Another simplified layout demonstrates the implementation of a guard ring in the TIA circuit (see Figure 122). The guard ring is implemented in the same manner as the buffer circuit. The primary difference is that the left half of the feedback resistor (RF) and feedback capacitor (CF) are connected to the high impedance node. The guard ring shape is extended around these passive components to ensure that the entire high impedance node is surrounded by guard. The guard ring is directly driven from the ADA4530-1 guard buffer (Pin 7). CF V+ A C+ GUARD C- V- VOUT Figure 122. TIA Circuit Layout RS The guard voltage in the TIA circuit is nominally equal to the B voltage, which makes it possible to drive the guard ring directly from the B voltage without using the ADA4530-1 guard buffer. When implementing the guard ring this way, do not make any connection to the guard buffer outputs (Pin 2 and Pin 7). V+ C+ B GND ADA4530-1 GUARD PLANE C- A GND ADA4530-1 B RF VOUT RF 13405-321 PHYSICAL IMPLEMENTATION OF GUARDING TECHNIQUES V- 13405-320 GUARD Figure 121. Buffer Circuit Layout There is not a large amount of exposed insulation between the A trace and the guard ring. It is often counterproductive to increase this spacing to try to increase the insulation resistance because the exposed insulator tends to accumulate surface charges generated from piezoelectric or triboelectric effects. These charges are eventually swept across the insulator toward the high impedance conductor. A guard plane is a structure used to implement the guarding technique through the bulk of the PCB. The structure of the guard plane is shown in a cross section of the PCB (see Figure 123). The guard plane is a filled copper shape that is placed directly below the high impedance (A) trace. This plane is connected to the guard ring on the surface layer with vias. If the circuit board is constructed using high performance PCB laminates such as Rogers 4350B, a hybrid stackup is required for mechanical strength. The outside layers are ceramic, whereas the core layers are conventional glass epoxy laminate. It is important to place the guard shield on the boundary of the ceramic and glass epoxy materials to protect the high impedance node from the poor dielectric relaxation characteristics of the glass epoxy materials. Rev. B | Page 43 of 52 ADA4530-1 Data Sheet Traditional electrical interferers are not the only sources of concern. Calculate the displacement current, I, in a capacitor as follows: SOLDER MASK SOLDER MASK GUARD A GUARD ROGERS 4350B I =C GUARD 13405-322 FR-4 Figure 123. Layout Cross Section with Guard Plane VIA FENCE A via fence is an additional structure that guards the lateral leakage paths in the laminate between the guard ring and the guard plane (see Figure 123). The fence is implemented by surrounding the entire guard ring with vias that connect the guard ring to the guard plane (see Figure 121 and Figure 122). CABLES AND CONNECTORS Guarding techniques are required for all high impedance wiring--not just on the PCB. Frequently, the high impedance sensor is not directly mounted on the PCB with the electrometer amplifier and external cables are used to make the connection. The typical way to guard a cable connecting to a current output sensor is by using a coaxial cable. A coaxial cable consists of an inner conductor surrounded with insulation, which is, in turn, surrounded by a braided conductor. Use the inner conductor for the high impedance (A) terminal and the outer braided shield conductor for the low impedance (B) terminal. Conveniently, this arrangement effectively guards the coaxial insulation resistance because the A terminal and B terminal are nominally at the same voltage (when attached to a TIA interface circuit). Voltage output sensors are more problematic because the A terminal and B terminal are not at the same voltage. The typical way to guard the voltage output sensor cable is to use a triaxial cable. A triaxial cable is constructed with an inner conductor with two separate braided conductors. Each of these braided conductors is separated from each other with insulation. Use the inner conductor for the high impedance (A) terminal and the inner braided conductor for the guard (VGRD) connection, and use the outer braided conductor for the low impedance (B) terminal. All the insulation around the inner conductor is completely surrounded by the guard conductor, which keeps the voltage drop across this insulation equal to zero. ELECTROSTATIC INTERFERANCE Very high impedance electrometer circuits are susceptible to interference through capacitive coupling. The amount of capacitance required to couple low frequency signals is surprisingly small. For example, line frequency (60 Hz) interference is coupled (with a -3 dB loss) to a 1 T impedance with only 3 fF of coupling capacitance. V C +V t t (2) The second term in this equation is frequently ignored in most circuits, but it can generate some unusual problems in electrometer circuits. The problem is that the movement of any charged object changes the coupling capacitance between the object and the electrometer, and this change in capacitance injects small currents into the circuit. The ADA4530-1 is so sensitive that it easily detects the movement of a hand or the movement of a piece of paper. These types of effects are not periodic or predictable, and they can appear as erratic dc shifts on the time scales of interest. Both of these types of interference can be reduced by the addition of a shield. A shield is a piece of conductive material placed between the high impedance input and the interference source. This shield must be electrically connected to a low impedance source (such as signal ground). If the shield physically interrupts all of the capacitive coupling paths, all of the displacement current from the interference source is shunted to the low impedance source. The construction of a shield is almost the same as the construction of a guard. Because of this similarity, many guard structures also provide shielding as well. The primary difference is that the dc voltage of the shield is not important, whereas the guard must have a voltage equal to that of the high impedance input. Shields that are driven by the guard buffer have the added benefit of bootstrapping the capacitance between the high impedance input and the shield. The disadvantage of this approach is that the guard buffer output impedance is 1 k, which makes the shield less effective than a signal ground or a chassis ground connection. The most effective systems typically use the box within a box construction: the outer shield is driven with ground and the inner shield is driven with guard. There is another capacitive interference effect that typically cannot be shielded. This displacement current is generated from a change in capacitance with respect to time (the second term of Equation 2). This change is due to the mechanical movement of the circuit components. This movement, which can be caused by mechanical impact or vibration, generates electrical interference. This interference typically appears at unexpected frequencies that are equal to the mechanical resonances of the components. This effect must be considered when using traditional air wiring techniques for large feedback resistors or relays. It is important to ensure solid mechanical connections to Teflon standoffs for this type of construction. Rev. B | Page 44 of 52 Data Sheet ADA4530-1 PHOTODIODE INTERFACE The low input bias current and low input offset voltage makes the ADA4530-1 an excellent choice for signal conditioning photodiodes at extremely low illumination levels. Figure 124 shows the ADA4530-1 configured in a transimpedance amplifier interfacing with a photodiode operating in photovoltaic mode (photodiode is zero biased). A photodiode produces an output current proportional to the illumination level. The amplifier converts the signal current, IPD, into an output voltage with the following equation: More importantly, the shunt resistance decreases by half for every 10C increase in temperature. An error current is created because the amplifier offset voltage is applied across this shunt resistance, resulting in an RTI error (IVOS_RTI) equal to IVOS_RTI = VOS/RSHUNT It is equivalent to think that the shunt resistance increases the DC noise gain (NG), which multiplies the offset voltage to the output. The RTO error due to VOS is equal to VOS_RTO = VOS x Noise Gain VOUT = IPD x RF VOS_RTO = VOS x (1 + RF/RSHUNT) CF The amplifier input resistance and insulation resistance appear in parallel with the photodiode shunt resistance. These additional resistances reduce the effective shunt resistance, but they are much larger than the photodiode shunt resistance and can usually be ignored. RF VOUT 13405-323 IPD PHOTODIODE Figure 124. Transimpedance Amplifier with Photodiode AC ERROR ANALYSIS Figure 125 replaces the photodiode with an equivalent circuit model. IPD is the photo current generated by incident light and is proportional to the light level. The shunt capacitance (CSHUNT) models the depletion capacitance of the diode. This capacitance depends on the area of the photodiode and the voltage bias. The shunt resistance (RSHUNT) represents the voltage vs. current slope of the exponential diode curve near zero bias voltage. CF PHOTODIODE CSHUNT The classic way of analyzing this circuit is by examining the noise gain vs. frequency (see Figure 126). At low frequencies, the noise gain is determined by the ratio of the feedback to the shunt resistance. VOUT RSHUNT 13405-324 IPD RF Photodiode TIA circuits typically require external compensation to give satisfactory dynamic performance. The large feedback resistor (RF) interacts with the large photodiode capacitance (CSHUNT) to create a low frequency pole in the feedback network. Photodiode shunt capacitance, amplifier input capacitance, and trace capacitance are lumped into a single element, CSHUNT. The phase shift due to this pole must be recovered prior to the crossover frequency for the feedback loop to be stable. The usual method to recover this phase shift is to create a zero in the feedback factor with the addition of the feedback capacitor (CF). Figure 125. Transimpedance Amplifier with Photodiode Model NG1 = 1 + DC ERROR ANALYSIS All of the errors described in the High Impedance Measurements section related to TIA circuits are applicable to photodiode interfaces. The inverting input bias current, IB-, sums directly with the photodiode current for a referred to input (RTI) error equal to IB-. This current flows through the feedback resistor, creating a referred to output (RTO) error, VIB_TRO, equal to VIB_RTO = IB- x RF The amplifier offset voltage, VOS, is a major error source in photodiode interface circuits because of the relatively low shunt resistance of large area photodiodes. Typical values are in the range of 1 G to 100 G at 25C. RF RSHUNT The troublesome low frequency pole (which is a zero in the noise gain) occurs at Frequency f1. From this frequency onward, the noise gain increases. If there is no feedback capacitor in the circuit, the noise gain follows the dotted line until it intersects with the amplifier open-loop gain curve. If these curves intersect at the 20 dB/decade slopes shown in Figure 126, the circuit is unstable. The addition of CF adds a zero to the feedback factor (which is a pole in the noise gain) at Frequency f2. Beyond Frequency f2, the noise gain is determined by the ratio of the shunt capacitance to the feedback capacitance. Rev. B | Page 45 of 52 NG2 = 1 + CSHUNT CF ADA4530-1 Data Sheet OPEN-LOOP GAIN GAIN (dB) IF CF = 0fF SIGNAL BANDWIDTH CLOSED-LOOP BANDWIDTH NG2 NOISE GAIN FREQUENCY (Hz) f1 f2 f3 f UGC 13405-325 NG1 Figure 126. Transimpedance Noise Gain vs. Frequency NOISE ANALYSIS For completeness, the noise gain equations are as follows: 2f +1 RF f1 NG( f ) = 1 + 2f R S f +1 2 f1 = Photodiode TIA circuits have four noise sources that must be considered: * * * * 1 RF RSHUNT (CF + CSHUNT ) RF + RSHUNT 1 f2 = RF C F For simplicity, bandwidth limitations are ignored in the noise gain equations. The noise gain starts to roll-off when it intersects with the open-loop gain of the amplifier. This pole frequency (f3) is determined by the unity gain crossover frequency (fUGC) of the amplifier and the high frequency noise gain, NG2, as follows: f3 = fUGC CSHUNT 1 + CF (3) The addition of CF has an impact on the signal frequency response. At low frequencies, the transimpedance gain is equal to RF. As the frequency increases, the impedance of CF drops below RF and starts to reduce this transimpedance gain. This signal gain equation is as follows: 1 Signal Gain ( f ) = RF 2f + 1 f 2 The thermal noise of the feedback resistor (RF) The saturation current noise of the photodiode The current noise of the amplifier The voltage noise of the amplifier The noise contributions of these sources are typically referred to output for analysis. The thermal noise of RF appears directly at the output. This noise is filtered by the feedback capacitance so that its -3 dB bandwidth is the same as the signal bandwidth (f2). The photocurrent of a photodiode, IPD, produces shot noise equal to INPD = (2qIPD) It is a mistake to assume that the noise goes to zero as the diode current goes to zero. Zero net current out of the diode simply means that the saturation current flowing in one direction is at thermal equilibrium with the saturation current flowing in the opposite direction. These currents are uncorrelated and add in a root sum square fashion. This net current noise is equivalent to the thermal noise of a physical resistor with a value of RSHUNT. This convenient fact allows the photodiode to be accurately modeled with a simple resistor, RSHUNT. The thermal noise of RSHUNT is amplified by the ratio of the feedback resistance to the shunt resistance. This noise is also filtered to the signal bandwidth. The current noise of the amplifier flows through the feedback resistor to become a noise voltage at the output. It is subject to the same bandwidth limitations as the previous noise contributors. The voltage noise of the amplifier is multiplied by the noise gain of the circuit to the output. This noise source is significant for two reasons. First, the high frequency noise gain can be high due to the large ratio between the shunt capacitance and the feedback capacitance. Second, the voltage noise bandwidth is much higher than the other contributors. The noise bandwidth is limited only by bandwidth of the amplifier. Rev. B | Page 46 of 52 Data Sheet ADA4530-1 Each of these noise contributors is graphed vs. frequency in Figure 127. A summary of the noise sources and their RTO contributions is shown in Table 12. The total RTO noise adds the contributions of each noise source in root sum square. * * 100 RF * NSD (V/Hz) RSHUNT 10 IN- VN 1 The signal bandwidth increases as the feedback capacitance (CF) decreases. The lower limit for CF is typically limited by one of the following: 0.1 * 1 f3 f2 10 100 1k FREQUENCY * 13405-404 f1 Figure 127. Photodiode TIA RTO Noise Spectral Density * Table 12. Photodiode Interface Noise Sources Noise Source RF Photodiode IN- Amplifier VN Amplifier RTO Noise (4kTRF) (RF/RSHUNT)(4kTRSHUNT) RF x IN- VN x noise gain Noise Bandwidth /2 x f2 /2 x f2 /2 x f2 /2 x f3 DESIGN RECOMMENDATIONS The design goal for a large area photodiode TIA circuit is usually to maximize signal-to-noise ratio (SNR) and minimize dc errors. Increasing the feedback resistor size accomplishes both goals. The signal gain increases directly with RF, whereas the noise increases in a square root fashion. High gains also make the output signal large relative to output voltage errors (such as VOS). The upper limit for RF is typically determined by one of the following: * * The thermal noise of the photodiode (RSHUNT). When RF is significantly larger than RSHUNT, the total noise is dominated by the photodiode and the SNR stops improving. The current noise of the amplifier. When the current noise of the amplifier is larger than the noise of RF, the SNR stops improving. The photodiode noise is higher than the amplifier current noise in nearly all practical photodiodes. The low frequency noise gain due to RSHUNT. When RF is larger than RSHUNT, the noise gain multiplies VOS and TCVOS errors and the signal to error ratio stops improving. Amplifier output swing. The maximum photocurrent multiplied by RF must be less than amplifier swing limitations. Signal bandwidth (or settling time). Signal bandwidth is dependent on RF x CF. Achieving high signal bandwidths with large feedback resistors can require vanishingly small feedback capacitors to implement. The ultimate limitation is due to the parasitic feedback capacitance from the fringing electric fields in the circuit. Parasitic capacitances in the 50 fF to 100 fF range are possible. For example, a 100 fF parasitic capacitance limits the signal bandwidth of a 100 G TIA to 16 Hz. * Parasitic feedback capacitances limit the minimum value of CF to 50 fF to 100 fF. Available component values. Physical components can be found in surface mount packages for values from 0.1 pF to 1 pF in 100 fF increments. Feedback loop stability. CF must be large enough to recover enough phase shift prior to the loop crossover for stable operation. This capacitance value can be a significant consideration for smaller values of RF. Large values (>1 G) tend to be self compensating through the parasitic feedback capacitance. High frequency noise gain. The high frequency noise gain is set by the ratio of CSHUNT to CF. For very large noise gains, it is possible for the amplifier voltage noise to be greater than the feedback resistor noise. DESIGN EXAMPLE In this section, an example TIA circuit is designed using a photometry grade photodiode (Hamamatsu S1226-18BQ). This medium area (1.2 mm2) silicon photodiode is responsive in the ultraviolet (UV) through visible frequency range. The minimum shunt resistance (RSHUNT) is specified at 5 G at 25C. The shunt capacitance (CSHUNT) is specified at 35 pF. The quartz window limits the maximum operating temperature to 60C. Based on the specified minimum shunt resistance and the recommendations in the Design Recommendations section, a value of 10 G is chosen for RF. This example circuit is powered from 5 V with the input common-mode voltage set at 0 V, which allows a maximum photocurrent of approximately 500 pA. An error budget is constructed based on the DC Error Analysis section (see Table 13). The amplifier offset voltage applies the maximum temperature drift limit to the maximum room temperature offset limit. The photo diode shunt resistance limit is reduced by half for every 10C. Rev. B | Page 47 of 52 ADA4530-1 Data Sheet 60C 40 V + 18 V 442 M 23 1.3 mV 20 fA 200 V 1.5 mV 150 fA 10 The total RTI error over the entire temperature range is less than 150 fA, which is equal to 300 ppm of the 500 pA full-scale range. The low input bias current of the ADA4530-1 is not a significant contributor to the total error over temperature. The interaction of the offset voltage with the shunt resistance of the photodiode is the most significant error source. This circuit was constructed as described with a 10 G feedback resistor (Ohmite RX-1M1008JE). The dc error performance was measured over the 25C to 60C temperature range (see Figure 128). The error increases rapidly with temperature as the shunt resistance changes the noise gain exponentially. The total RTI error ranges from +2 fA to -10 fA, considerably lower than the worst case error budget, as expected. 20 VSY = 10V VCM = VSY/2 1 0.1 0.1 0 -40 -4 -60 -6 -80 -8 -100 -10 -120 -12 70 OUTPUT VOLTAGE (mV) -2 50 60 13405-401 30 40 TEMPERATURE (C) 100 1k FREQUENCY (Hz) 10k 100k VSY = 10V TA = 25C IPD = 10pA 2 -20 -2 -40 -4 -60 -6 -80 -8 -100 -10 -12 CF = 0fF CF = 300fF -140 0 Figure 128. DC Error vs. Temperature 4 0 -120 20 10 40 -20 10 1 The stability improvement can be seen in the time domain as well. The circuits step response to a 10 pA photocurrent is shown in Figure 130. The uncompensated circuit (red curve) shows considerable (20%) overshoot. The compensated circuit (blue curve) is overdamped. 20 0 CF = 0fF CF = 300fF Figure 129. Transimpedance Gain vs. Frequency 0 RTI ERROR (fA) RTO ERROR (V) 0 2 VSY = 10V VCM = VSY/2 TA = 25C 13405-407 45C 40 V + 10 V 1.25 G 9 450 V 20 fA 200 V 650 V 65 fA 5 10 INPUT REFERRED CURRENT (pA) 25C 40 V 5 G 3 120 V 20 fA 200 V 320 V 32 fA 15 20 25 TIME (ms) 30 35 -14 40 13405-400 Error Source VOS RSHUNT Noise Gain VOS Error RTO IB IB Error RTO Total Error RTO Total Error RTI These low valued capacitors are designed for RF use and are readily available. The 300 fF capacitor eliminates the frequency peaking completely (blue curve) but it reduces the -3 dB bandwidth from 390 Hz to 50 Hz. TRANSIMPEDANCE GAIN (G) Table 13. Photodiode Interface DC Error Budget Figure 130. 10 pA Step Response The ac performance of the circuit was also measured. The circuit was initially constructed without a physical feedback capacitor as a baseline. The transimpedance gain vs. frequency is shown in Figure 129. The 30% frequency peaking seen in the frequency response (red curve) indicates that the feedback loop is marginally compensated with parasitic capacitance. A physical capacitor was added to improve the loop compensation. This capacitor is a 300 fF C0G ceramic in a Size 0805, surface-mount package (AVX UQCFVA0R3BAT2A\500). C0G ceramic capacitors are good candidates for electrometer circuits because they have adequate insulation resistance and dielectric absorption performance. A noise budget is constructed based on the Noise Analysis section. The RTO noise budget is separated into noise sources integrated with a low bandwidth (see Table 14) and those integrated with a high bandwidth (see Table 15). The low frequency noise contributors include the feedback resistance, the shunt resistance and the amplifier current noise. Each of these sources has a -3 dB bandwidth equal to the signal bandwidth (50 Hz); this is equivalent to a noise bandwidth of 79 Hz. The most significant noise source is the photodiode shunt resistance by a large margin. The second most significant source is the feedback resistor. The amplifier current noise is so low that it can be ignored. Rev. B | Page 48 of 52 60C 13.5 V/Hz 442 M 2.8 V/Hz 22 61 V/Hz 0.24 fA/Hz 2.4 V/Hz 62 V/Hz 194 V rms 345 V rms 549 V rms The sole high frequency noise contributor is the amplifier voltage noise, which is multiplied by the high frequency noise gain and band limited only by the amplifier gain. The -3 dB bandwidth of the amplifier is 17 kHz (refer to Equation 3, where f3 = fUGC / NG2 = 2 MHz / 118). The equivalent noise bandwidth is 27 kHz. The high bandwidth is the reason the high frequency noise is significant even though the noise spectral density is much lower than the low frequency noise. Table 15. High Frequency Noise Budget Error Source VN High Frequency Noise Gain VN_RTO High Frequency RMS Total 25C 14 nV/Hz 118 45C 14.5 nV/Hz 118 60C 14.8 nV/Hz 118 1.6 V/Hz 271 V rms 1.7 V/Hz 281 V rms 1.7 V/Hz 286 V rms At low temperatures, the amplifier voltage noise is more significant than any other noise source. This is important because the majority of this noise occurs outside the useful bandwidth of the circuit. For this reason, it is recommended to add a low-pass filter to the output of a photodiode TIA circuit. This filter can be active or passive depending on the needs of the system. A simple resistor capacitor (RC) filter with a -3 dB cutoff of 500 Hz has an insignificant impact on the frequency response of the signal path, but it lowers the integrated noise from 271 V rms to 45 V rms (a 6x reduction). The NSD was measured for this circuit with (blue curve) and without (red curve) the 300 fF CF capacitor (see Figure 131). At low frequencies, the NSD is approximately equal to the noise from the feedback resistor alone (12.8 V/Hz). The value of the low frequency NSD shows that the shunt resistance is much larger than the specified minimum (which is expected). As frequency increases, the resistor noise rolls off at the signal bandwidth (50 Hz). The NSD then plateaus at the amplifier voltage noise level until the bandwidth limitations of the amplifier roll off the NSD toward zero. 1000 10 100 1 0.1 NSD, CF = 0fF NSD, CF = 300fF RMS, CF = 0fF RMS, CF = 300fF 1 10 100 1k FREQUENCY (Hz) 10k 10 100k 13405-405 45C 13.2 V/Hz 1.25 G 4.7 V/Hz 8 37 V/Hz 0.15 fA/Hz 1.5 V/Hz 39 V/Hz VSY = 10V VCM = VSY/2 TA = 25C Figure 131. RTO Noise Spectral Density (25C) The dashed curves show the integration of the NSD across the frequency spectrum. These are useful to calculate the rms noise over a variety of bandwidths. For example, the rms noise over the entire 100 kHz measurement bandwidth is 400 V rms, which is approximately the same as the calculated total noise of 333 V rms. If a postfilter is added with a noise bandwidth of 1 kHz, Figure 131 shows that the integrated noise is 200 V rms (a 2x improvement). The uncompensated circuit (red curves) shows considerably worse noise performance. The frequency peaking due to the marginal loop stability multiplies the noise as well as the signal. In addition, the high frequency noise gain is larger, which adds much more noise outside the signal bandwidth. Both of these effects together generate 1.2 mV rms of total noise. Even if the transient and frequency response of an undercompensated TIA are acceptable, the large noise penalty may not be. Lastly, the NSD was measured for this circuit at 60C (see Figure 132). As expected, the low frequency noise increased as a result of the photodiode shunt resistance. The average low frequency NSD is 22 V/Hz. Removing the contribution of RF gives an RTO contribution of 17 V/Hz, which is equivalent to an RTI current noise of 1.7 fA/Hz. RSHUNT must be approximately 6.5 G at 60C to generate this noise. 100 NSD, CF = 300fF RMS, CF = 300fF VSY = 10V VCM = VSY/2 TA = 60C 10 1 0.1 100 1 10 100 1k FREQUENCY (Hz) 10k Figure 132. RTO Noise Spectral Density (60C) Rev. B | Page 49 of 52 1000 10 100k INTEGRATED VOTLAGE NOISE (V rms) 25C 12.8 V/Hz 5 G 9 V/Hz 2 18 V/Hz 0.07 fA/Hz 700 nV/Hz 22 V/Hz 100 13405-406 Error Source VNRF RSHUNT VNRSHUNT RF/RSHUNT VNRSHUNT_RTO IN- IN-_RTO Low Frequency NSD Total Low Frequency RMS Total VOLTAGE NOISE SPECTRAL DENSITY (V/Hz) Table 14. Low Frequency Noise Budget INTEGRATED VOTLAGE NOISE (V rms) ADA4530-1 VOLTAGE NOISE SPECTRAL DENSITY (V/Hz) Data Sheet ADA4530-1 Data Sheet POWER SUPPLY RECOMMENDATIONS Analog Devices offers a wide range of power management products to meet the requirements of most high performance signal chains. Examples of a single- and dual-supply solution is shown in Figure 133. The ADP2370 and ADP5075, cascaded with the ADP7118 or ADM7170, and the ADP7182 generate clean positive and negative rails. These rails power the ADA4530-1, electrometer amplifier and/or the precision converter in a typical signal chain. ADP5075 INVERTING REGULATOR 3.3V INPUT 15V/12V/6V INPUT +8.5V OR +6V -8.5V OR -6V ADP7182 -7.5V OR -5V LDO +6V ADM7170/ ADP7118 Figure 134 shows the combined PSRR of using the ADP7118 to provide +5 V on +VSY of the ADA4530-1, and the ADP7182 to provide -5 V on-VSY, from a 9 V battery main supply. Figure 135 shows the maximum allowable ripple at the input so that the combined PSRR of the LDO and the amplifier can still attenuate the noise level down to the noise floor. +7.5V OR +5V LDO ADP5070 BOOST AND INVERTING REGULATOR ADP7118 ADP7118 For example, if the main supply to the ADP7118 and ADA4530-1 has a switching noise of 20 mV p-p at 300 kHz, Figure 135 shows that it is below the maximum value of 90 mV p-p. Therefore, the combined PSRR of the system can still attenuate and bring the noise level down to the noise floor, in effect, the 300 kHz noise at the input is not seen at the output of the amplifier. +5V LDO -6V ADP7182 -5V LDO +12V/+10V/+5V LDO 0 Figure 133. Recommended Power Solutions -20 Table 16. Recommended Power Management Devices ADP5070 ADM7170 ADP7118 ADP7182 Description 800 mA, dc-to-dc inverting regulator High voltage, 1.2 MHz/600 kHz, 800 mA, low quiescent current buck regulator 1 A/0.6 A, dc-to-dc switching regulator with independent positive and negative outputs 6.5 V, 500 mA, ultralow noise, high PSRR, CMOS LDO 20 V, 200 mA, low noise, high PSRR, CMOS LDO -28 V, -200 mA, low noise, linear regulator -40 PSRR (dB) Product ADP5075 ADP2370 +PSRR -PSRR -60 -80 -100 -120 10k POWER SUPPLY CONSIDERATIONS For a single-supply application, the ADA4530-1 typically needs a 5 V, 10 V, or 12 V supply, although a 4.5 V to 16 V supply can 10M Figure 134. Positive and Negative PSRR for the ADP7118 and ADP7182 Powering ADA4530-1, VSY = 5 V, +IN = 0 V 1000 ADP7118 ADP7182 100 (mV p-p) The PSRR of the ADA4530-1 is excellent at dc (approximately 150 dB); however, it decreases as frequency increases. To achieve the best performance of the ADA4530-1, a low-noise supply is necessary. If switching supplies are used for input rails, a low dropout regulator (LDO) is essential to attenuate the switching spurs to a level that does not affect the ADA4530-1 output. Switching power supply noise typically spans a frequency range from 300 kHz and up. The switching spurs can effectively be attenuated using the LDO. Additional filtering around the LDO may be necessary, especially when using a switching regulator to generate an intermediate rail. Switching regulators also generate high frequency noise content (>100 MHz), even when running in the 100 kHz range, because of the high dv/dt of the switch node. In this case, ferrite beads can be used, as described in the AN-1120 Application Note and the AN-1368 Application Note. 100k 1M FREQUENCY (Hz) 13405-534 BUCK REGULATOR For a dual-supply application, the ADA4530-1 typically needs a 5 V supply, although in some applications, a 2.5 V to 8 V supply can be used. LDOs like the ADP7118 or ADM7170 are the optimum choices for the positive supply, and the ADP7182 for the negative supply. In addition, if a negative supply is not already available, the ADP5075 or the ADP5070 can generate the negative supply from a positive supply, as shown in Figure 133. 10 1 0.1 0.01 0.01 0.1 1 FREQUENCY (Hz) 10 13405-535 ADP2370 13405-533 12V INPUT also be used. An LDO like the ADM7170 or ADP7118 is ideal to generate the low noise rail. Figure 135. Maximum Allowable Ripple at the Input of the LDOs to Bring Spurs To Noise Floor Level (-120 dB) Rev. B | Page 50 of 52 Data Sheet ADA4530-1 LONG-TERM DRIFT TEMPERATURE HYSTERESIS The stability of a precision signal path over its lifetime or between calibration procedures is dependent on the long-term stability of the analog components in the path, such as op amps, references, and data converters. To help system designers predict the long-term drift of circuits that use the ADA4530-1, Analog Devices measured the offset voltage of multiple units for 10,000 hours (more than 13 months) using a high precision measurement system, including an ultrastable oil bath. To replicate real-world system performance, the devices under test (DUTs) were soldered onto an FR4 PCB using a standard reflow profile (as defined in the JEDEC J-STD-020D standard), as opposed to testing them in sockets. This manner of testing is important because expansion and contraction of the PCB can apply stress to the integrated circuit (IC) package and contribute to shifts in the offset voltage. In addition to stability over time as described in the Long-Term Drift section, it is useful to know the temperature hysteresis, that is, the stability vs. cycling of temperature. Hysteresis is an important parameter because it tells the system designer how closely the signal returns to its starting amplitude after the ambient temperature changes and subsequent return to room temperature. Figure 137 shows the change in input offset voltage as the temperature cycles three times from room temperature to 125C to -40C and back to room temperature. The dotted line is an initial preconditioning cycle to eliminate the original temperature induced offset shift from exposure to production solder reflow temperatures. In the three full cycles, the offset hysteresis is typically only 1.5 V. The histogram in Figure 138 shows that the hysteresis is larger when the device is cycled through only a half cycle, from room temperature to 125C and back to room temperature. 25 MEAN MEAN PLUS ONE STANDARD DEVIATION MEAN MINUS ONE STANDARD DEVIATION 15 10 5 PRECONDITION CYCLE 1 CYCLE 2 CYCLE 3 VSY = 10V 20 0 15 10 5 00 -5 -10 -15 -25 -40 -10 -15 NUMBER OF DEVICES 10,000 13405-136 9000 8000 7000 6000 5000 4000 3000 2000 1000 0 -25 TIME (Hours) 0 20 40 60 TEMPERATURE (C) 80 100 120 Figure 137. Change in Offset Voltage over Three Full Temperature Cycles VSY = 10V 27 UNITS TA = 25C SAMPLE 1 SAMPLE 2 SAMPLE 3 -20 -20 13405-137 -20 -5 Figure 136. Measured Long-Term Drift of the ADA4530-1 Offset Voltage over 10,000 Hours 40 VSY = 10V 35 27 UNITS x 3 CYCLES HALF CYCLE = +26C, +125C, +26C 30 FULL CYCLE = +26C, +125C, +26C, -40C, +26C 25 20 15 10 5 0 40 35 30 25 20 15 10 5 0 -12 -9 -6 -3 0 3 HALF CYCLE FULL CYCLE 6 OFFSET VOLTAGE HYSTERESIS (V) 9 12 13405-138 CHANGE IN OFFSET VOLTAGE (V) 20 25 CHANGE IN OFFSET VOLTAGE (V) The ADA4530-1 have extremely low long-term drift, as shown in Figure 136. The red, blue, and green traces show sample units. Note that the ADA4530-1 has a mean drift over 10,000 hours of less than 0.5 V, or less than 2% of its maximum specified offset voltage of 40 V at room temperature. Figure 138. Histogram Showing the Temperature Hysteresis of the Offset Voltage over Three Full Cycles and over Three Half Cycles Rev. B | Page 51 of 52 ADA4530-1 Data Sheet OUTLINE DIMENSIONS 5.00 (0.1968) 4.80 (0.1890) 1 5 6.20 (0.2441) 5.80 (0.2284) 4 1.27 (0.0500) BSC 0.25 (0.0098) 0.10 (0.0040) COPLANARITY 0.10 SEATING PLANE 1.75 (0.0688) 1.35 (0.0532) 0.51 (0.0201) 0.31 (0.0122) 0.50 (0.0196) 0.25 (0.0099) 45 8 0 0.25 (0.0098) 0.17 (0.0067) 1.27 (0.0500) 0.40 (0.0157) COMPLIANT TO JEDEC STANDARDS MS-012-A A CONTROLLING DIMENSIONS ARE IN MILLIMETERS; INCH DIMENSIONS (IN PARENTHESES) ARE ROUNDED-OFF MILLIMETER EQUIVALENTS FOR REFERENCE ONLY AND ARE NOT APPROPRIATE FOR USE IN DESIGN. 012407-A 8 4.00 (0.1574) 3.80 (0.1497) Figure 139. 8-Lead Small Outline Package [SOIC_N] Narrow Body (R-8) Dimensions shown in millimeters and (inches) ORDERING GUIDE Model 1 ADA4530-1ARZ ADA4530-1ARZ-R7 ADA4530-1ARZ-RL ADA4530-1R-EBZ-BUF ADA4530-1R-EBZ-TIA EVAL-CN0407-SDPZ 1 Temperature Range -40C to +125C -40C to +125C -40C to +125C Package Description 8-Lead Small Outline Package [SOIC_N] 8-Lead Small Outline Package [SOIC_N] 8-Lead Small Outline Package [SOIC_N] Evaluation Board Buffer Configuration for 8-Lead SOIC Evaluation Board Transimpedance Configuration for 8-Lead SOIC Ultrahigh Sensitivity Femtoampere Measurement Platform Z = RoHS Compliant Part. (c)2015-2017 Analog Devices, Inc. All rights reserved. Trademarks and registered trademarks are the property of their respective owners. D13405-0-5/17(B) Rev. B | Page 52 of 52 Package Option R-8 R-8 R-8