LMP7707/LMP7708/LMP7709 Precision, CMOS Input, RRIO, Wide Supply Range Decompensated Amplifiers General Description Features The LMP7707/LMP7708/LMP7709 devices are single, dual, and quad low offset voltage, rail-to-rail input and output precision amplifiers which each have a CMOS input stage and a wide supply voltage range. The LMP7707/LMP7708/ LMP7709 are part of the LMP® precision amplifier family and are ideal for sensor interface and other instrumentation applications. These decompensated amplifiers are stable at a gain of 6 and higher. The guaranteed low offset voltage of less than ±200 µV along with the guaranteed low input bias current of less than ±1 pA make the LMP7707/LMP7708/LMP7709 ideal for precision applications. The LMP7707/LMP7708/LMP7709 are built utilizing VIP50 technology, which allows the combination of a CMOS input stage and a supply voltage range of 12V with rail-to-rail common mode voltage capability. The LMP7707/ LMP7708/LMP7709 are the perfect choice in many applications where conventional CMOS parts cannot operate due to the voltage conditions. The unique design of the rail-to-rail input stage of each of the LMP7707/LMP7708/LMP7709 significantly reduces the CMRR glitch commonly associated with rail-to-rail input amplifiers. Both sides of the complimentary input stage have been trimmed, thereby reducing the difference between the NMOS and PMOS offsets. The output swings within 40 mV of either rail to maximize the signal dynamic range in applications requiring low supply voltage. The LMP7707 is offered in the space saving 5-Pin SOT23 package, the LMP7708 is offered in the 8-Pin MSOP and the quad LMP7709 is offered in the 14-Pin TSSOP package. These small packages are ideal solutions for area constrained PC boards and portable electronics. Unless otherwise noted, typical values at VS = 5V. ±200 µV (max) ■ Input offset voltage (LMP7707) ■ Input offset voltage (LMP7708/LMP7709) ±220 µV (max) ±200 fA ■ Input bias current 9 nV/√Hz ■ Input voltage noise 130 dB ■ CMRR 130 dB ■ Open loop gain −40°C to 125°C ■ Temperature range 14 MHz ■ Gain bandwidth product (AV =10) ■ Stable at a gain of 10 or higher 715 µA ■ Supply current (LMP7707) 1.5 mA ■ Supply current (LMP7708) 2.9 mA ■ Supply current (LMP7709) 2.7V to 12V ■ Supply voltage range ■ Rail-to-rail input and output Applications ■ ■ ■ ■ ■ ■ High impedance sensor interface Battery powered instrumentation High gain amplifiers DAC buffer Instrumentation amplifier Active filters Open Loop Frequency Response 20203764 Increased Bandwidth for Same Supply Current at AV> 10 LMP® is a registered trademark of National Semiconductor Corporation. © 2007 National Semiconductor Corporation 202037 www.national.com LMP7707/LMP7708/LMP7709 Precision, CMOS Input, RRIO, Wide Supply Range Decompensated Amplifiers June 2007 LMP7707/LMP7708/LMP7709 Junction Temperature (Note 3) Soldering Information Absolute Maximum Ratings (Note 1) If Military/Aerospace specified devices are required, please contact the National Semiconductor Sales Office/ Distributors for availability and specifications. ESD Tolerance (Note 2) Human Body Model 2000V Machine Model Charge Device Model VIN Differential Supply Voltage (VS = V+ – V−) Voltage at Input/Output Pins Input Current Storage Temperature Range 235°C Wave Soldering Lead Temp. (10 sec) 260°C (Note 1) Temperature Range (Note 3) Supply Voltage (VS = V+ – V−) 1000V ±300 mV 13.2V V++ 0.3V to V− − 0.3V 10 mA −65°C to +150°C 3V Electrical Characteristics Infrared or Convection (20 sec) Operating Ratings 200V +150°C −40°C to +125°C 2.7V to 12V Package Thermal Resistance (θJA) (Note 3) 5-Pin SOT23 8-Pin MSOP 14-Pin TSSOP 265°C/W 235°C/W 122°C/W (Note 4) Unless otherwise specified, all limits are guaranteed for TA = 25°C, V+ = 3V, V− = 0V, VCM = V+/2, and RL > 10 kΩ to V+/2. Boldface limits apply at the temperature extremes. Symbol VOS Parameter Input Offset Voltage TCVOS Input Offset Voltage Drift (Note 7) IB Input Bias Current (Notes 7, 8) IOS Input Offset Current CMRR Common Mode Rejection Ratio Conditions Min (Note 6) Typ (Note 5) Max (Note 6) LMP7707 ±37 ±200 ±500 LMP7708/LMP7709 ±56 ±220 ±520 ±1 ±5 ±0.2 ±1 −40°C ≤ TA ≤ 85°C ±50 −40°C ≤ TA ≤ 125°C ±400 40 0V ≤ VCM ≤ 3V LMP7707 86 80 130 0V ≤ VCM ≤ 3V LMP7708/LMP7709 84 78 130 86 82 98 2.7V ≤ V+ ≤ 12V, Vo = V+/2 CMVR Input Common-Mode Voltage Range CMRR ≥ 80 dB −0.2 3.2 CMRR ≥ 77 dB −0.2 3.2 RL = 2 kΩ (LMP7707) VO = 0.3V to 2.7V 100 96 114 RL = 2 kΩ (LMP7708/LMP7709) VO = 0.3V to 2.7V 100 94 114 RL = 10 kΩ VO = 0.2V to 2.8V 100 96 124 www.national.com 2 μV/°C pA dB Power Supply Rejection Ratio Open Loop Voltage Gain μV fA PSRR AVOL Units dB V dB VO Parameter Output Swing High Output Swing Low IO Output Short Circuit Current (Notes 3, 9) Conditions Typ (Note 5) Max (Note 6) RL = 2 kΩ to V+/2 LMP7707 40 80 120 RL = 2 kΩ to V+/2 LMP7708/LMP7709 40 80 150 RL = 10 kΩ to V+/2 LMP7707 30 40 60 RL = 10 kΩ to V+/2 LMP7708/LMP7709 35 50 100 RL = 2 kΩ to V+/2 LMP7707 40 60 80 RL = 2 kΩ to V+/2 LMP7708/LMP7709 45 100 170 RL = 10 kΩ to V+/2 LMP7707 20 40 50 RL = 10 kΩ to V+/2 LMP7708/LMP7709 20 50 90 Sourcing VO = V+/2 VIN = 100 mV 25 15 42 Sinking VO = V+/2 VIN = −100 mV (LMP7707) 25 20 42 25 15 42 V+/2 Sinking VO = VIN = −100 mV (LMP7708/ LMP7709) IS Supply Current Min (Note 6) Units mV from V+ mV mA LMP7707 0.670 1.0 1.2 LMP7708 1.4 1.8 2.1 LMP7709 2.9 3.5 4.5 mA SR Slew Rate (Note 10) VO = 2 VPP,10% to 90% 5.1 V/μs GBWP Gain Bandwidth Product AV = 10 13 MHz THD+N Total Harmonic Distortion + Noise f = 1 kHz, AV = 10, VO = 2.5V, 0.024 % RL = 10 kΩ en Input-Referred Voltage Noise f = 1 kHz 9 nV/ in Input-Referred Current Noise f = 100 kHz 1 fA/ 5V Electrical Characteristics (Note 4) Unless otherwise specified, all limits are guaranteed for TA = 25°C, V+ = 5V, V− = 0V, VCM = V+/2, and RL > 10 kΩ to V+/2. Boldface limits apply at the temperature extremes. Symbol VOS Parameter Input Offset Voltage TCVOS Input Offset Voltage Drift (Note 7) IB Input Bias Current (Notes 7, 8) Conditions Min (Note 6) Typ (Note 5) Max (Note 6) LMP7707 ±37 ±200 ±500 LMP7708/LMP7709 ±32 ±220 ±520 ±1 ±5 ±0.2 ±1 μV μV/°C −40°C ≤ TA ≤ 85°C ±50 −40°C ≤ TA ≤ 125°C ±400 3 Units pA www.national.com LMP7707/LMP7708/LMP7709 Symbol LMP7707/LMP7708/LMP7709 Symbol Parameter IOS Input Offset Current CMRR Common Mode Rejection Ratio Conditions Min (Note 6) Typ (Note 5) 0V ≤ VCM ≤ 5V LMP7707 88 83 130 0V ≤ VCM ≤ 5V LMP7708/LMP7709 86 81 130 86 82 100 Max (Note 6) 40 fA dB PSRR Power Supply Rejection Ratio 2.7V ≤ V+ ≤ 12V, VO = V+/2 CMVR Input Common-Mode Voltage Range CMRR ≥ 80 dB −0.2 5.2 CMRR ≥ 78 dB −0.2 5.2 RL = 2 kΩ (LMP7707) VO = 0.3V to 4.7V 100 96 119 RL = 2 kΩ (LMP7708/LMP7709) VO = 0.3V to 4.7V 100 94 119 RL = 10 kΩ VO = 0.2V to 4.8V 100 96 130 AVOL VO Open Loop Voltage Gain Output Swing High Output Swing Low IO IS Output Short Circuit Current (Notes 3, 9) Supply Current dB 60 110 130 RL = 2 kΩ to V+/2 LMP7708/LMP7709 60 120 200 RL = 10 kΩ to V+/2 LMP7707 40 50 70 RL = 10 kΩ to V+/2 LMP7708/LMP7709 40 60 120 RL = 2 kΩ to V+/2 LMP7707 50 80 90 RL = 2 kΩ to V+/2 LMP7708/LMP7709 50 120 190 RL = 10 kΩ to V+/2 LMP7707 30 40 50 RL = 10 kΩ to V+/2 LMP7708/LMP7709 30 50 100 40 28 66 Sourcing VO = V+/2 VIN = 100 mV (LMP7708/LMP7709) 38 25 66 Sinking VO = V+/2 VIN = −100 mV (LMP7707) 40 28 76 Sinking VO = V+/2 VIN = −100 mV (LMP7708/ LMP7709) 40 23 76 V dB RL = 2 kΩ to V+/2 LMP7707 Sourcing VO = V+/2 VIN = 100 mV (LMP7707) Units mV from V+ mV mA LMP7707 0.715 1.0 1.2 LMP7708 1.5 1.9 2.2 LMP7709 2.9 3.7 4.6 mA SR Slew Rate (Note 10) VO = 4 VPP, 10% to 90% 5.6 V/μs GBWP Gain Bandwidth Product AV = 10 14 MHz www.national.com 4 THD+N Parameter Total Harmonic Distortion + Noise Conditions Min (Note 6) f = 1 kHz, AV = 10, VO = 4.5V, Typ (Note 5) Max (Note 6) 0.024 Units % RL = 10 kΩ en Input-Referred Voltage Noise f = 1 kHz 9 nV/ in Input-Referred Current Noise f = 100 kHz 1 fA/ ±5V Electrical Characteristics (Note 4) Unless otherwise specified, all limits are guaranteed for TA = 25°C, V+ = 5V, V− = −5V, VCM = 0V, and RL > 10 kΩ to 0V. Boldface limits apply at the temperature extremes. Symbol VOS Parameter Input Offset Voltage TCVOS Input Offset Voltage Drift (Note 7) IB Input Bias Current (Notes 7, 8) IOS Input Offset Current CMRR Common Mode Rejection Ratio Conditions Min (Note 6) Typ (Note 5) Max (Note 6) LMP7707 ±37 ±200 ±500 LMP7708/LMP7709 ±37 ±220 ±520 ±1 ±5 ±0.2 1 −40°C ≤ TA ≤ 85°C ±50 −40°C ≤ TA ≤ 125°C ±400 40 −5V ≤ VCM ≤ 5V LMP7707 92 88 138 −5V ≤ VCM ≤ 5V LMP7708/LMP7709 90 86 138 2.7V ≤ V+ ≤ 12V, V- = 0V, VO = V+/2 86 82 98 CMVR Input Common-Mode Voltage Range CMRR ≥ 80 dB −5.2 5.2 CMRR ≥ 78 dB −5.2 5.2 RL = 2 kΩ (LMP7707) VO = −4.7V to 4.7V 100 98 121 RL = 2 kΩ (LMP7708/LMP7709) VO = −4.7V to 4.7V 100 94 121 RL = 10 kΩ (LMP7707) VO = −4.8V to 4.8V 100 98 134 RL = 10 kΩ (LMP7708/LMP7709) VO = −4.8V to 4.8V 100 97 134 5 μV/°C pA dB Power Supply Rejection Ratio Open Loop Voltage Gain μV fA PSRR AVOL Units dB V dB www.national.com LMP7707/LMP7708/LMP7709 Symbol LMP7707/LMP7708/LMP7709 Symbol VO Parameter Output Swing High Output Swing Low IO Output Short Circuit Current (Notes 3, 9) IS Supply Current Conditions Min (Note 6) Typ (Note 5) Max (Note 6) RL = 2 kΩ to 0V LMP7707 90 150 170 RL = 2 kΩ to 0V LMP7708/LMP7709 90 180 290 RL = 10 kΩ to 0V LMP7707 40 80 100 RL = 10 kΩ to 0V LMP7708/LMP7709 40 80 150 RL = 2 kΩ to 0V LMP7707 90 130 150 RL = 2 kΩ to 0V LMP7708/LMP7709 90 180 290 RL = 10 kΩ to 0V LMP7707 40 50 60 RL = 10 kΩ to 0V LMP7708/LMP7709 40 60 110 Sourcing VO = 0V VIN = 100 mV (LMP7707) 50 35 86 Sourcing VO = 0V VIN = 100 mV (LMP7708/LMP7709) 48 33 86 Sinking VO = 0V VIN = −100 mV 50 35 84 Units mV from V+ mV from V– mA LMP7707 0.790 1.1 1.3 LMP7708 1.7 2.1 2.5 LMP7709 3.2 4.2 5.0 mA SR Slew Rate (Note 10) VO = 9 VPP, 10% to 90% 5.9 V/μs GBWP Gain Bandwidth Product AV = 10 15 MHz THD+N Total Harmonic Distortion + Noise f = 1 kHz, AV = 10, VO = 9V, 0.024 % RL = 10 kΩ en Input-Referred Voltage Noise f = 1 kHz 9 nV/ in Input-Referred Current Noise f = 100 kHz 1 fA/ Note 1: Absolute Maximum Ratings indicate limits beyond which damage to the device may occur. Operating Ratings indicate conditions for which the device is intended to be functional, but specific performance is not guaranteed. For guaranteed specifications and the test conditions, see the Electrical Characteristics Tables. Note 2: Human Body Model, applicable std. MIL-STD-883, Method 3015.7. Machine Model, applicable std. JESD22-A115-A (ESD MM std. of JEDEC) Field-Induced Charge-Device Model, applicable std. JESD22-C101-C (ESD FICDM std. of JEDEC). Note 3: The maximum power dissipation is a function of TJ(MAX), θJA. The maximum allowable power dissipation at any ambient temperature is PD = (TJ(MAX) - TA)/ θJA . All numbers apply for packages soldered directly onto a PC board. Note 4: Electrical table values apply only for factory testing conditions at the temperature indicated. Factory testing conditions result in very limited self-heating of the device. Note 5: Typical values represent the most likely parametric norm as determined at the time of characterization. Actual typical values may vary over time and will also depend on the application and configuration. The typical values are not tested and are not guaranteed on shipped production material. Note 6: Limits are 100% production tested at 25°C. Limits over the operating temperature range are guaranteed through correlations using the Statistical Quality Control (SQC) method. Note 7: This parameter is guaranteed by design and/or characterization and is not tested in production. Note 8: Positive current corresponds to current flowing into the device. Note 9: The short circuit test is a momentary test. Note 10: The number specified is the slower of positive and negative slew rates. www.national.com 6 LMP7707/LMP7708/LMP7709 Connection Diagrams 5-Pin SOT23 8-Pin MSOP 20203702 Top View 14-Pin TSSOP 20203703 Top View 20203704 Top View Ordering Information Package 5-Pin SOT23 8-Pin MSOP 14-Pin TSSOP Part Number LMP7707MF LMP7707MFX LMP7708MM LMP7708MMX LMP7709MT LMP7709MTX Package Marking Transport Media 1k Units Tape and Reel AH4A 3k Units Tape and Reel 1k Units Tape and Reel AJ4A 3.5k Units Tape and Reel LMP7709MT 7 94 Units/Rail 2.5k Units Tape and Reel NSC Drawing MF05A MUA08A MTC14 www.national.com LMP7707/LMP7708/LMP7709 Typical Performance Characteristics Unless otherwise specified, TA = 25°C, VCM = VS/2, RL > 10 kΩ connected to (V++V-)/2 Offset Voltage Distribution TCVOS Distribution 20203736 20203741 Offset Voltage Distribution TCVOS Distribution 20203737 20203742 Offset Voltage Distribution TCVOS Distribution 20203738 www.national.com 20203743 8 LMP7707/LMP7708/LMP7709 Offset Voltage vs. Temperature CMRR vs. Frequency 20203706 20203750 Offset Voltage vs. Supply Voltage Offset Voltage vs. VCM 20203707 20203710 Offset Voltage vs. VCM Offset Voltage vs. VCM 20203708 20203709 9 www.national.com LMP7707/LMP7708/LMP7709 Input Bias Current vs. VCM Input Bias Current vs. VCM 20203730 20203746 Input Bias Current vs. VCM Input Bias Current vs. VCM 20203731 20203747 Input Bias Current vs. VCM Input Bias Current vs. VCM 20203748 www.national.com 20203749 10 Supply Current vs. Supply Voltage (Per Channel) 20203745 20203711 Sinking Current vs. Supply Voltage Sourcing Current vs. Supply Voltage 20203713 20203712 Output Voltage vs. Output Current Slew Rate vs. Supply Voltage 20203717 20203716 11 www.national.com LMP7707/LMP7708/LMP7709 PSRR vs. Frequency LMP7707/LMP7708/LMP7709 Open Loop Frequency Response Open Loop Frequency Response 20203714 20203715 Small Signal Step Response, AV = 10 Large Signal Step Response, AV = 10 20203719 20203718 Small Signal Step Response, AV = 100 Large Signal Step Response, AV = 100 20203726 www.national.com 20203720 12 Open Loop Gain vs. Output Voltage Swing 20203727 20203752 Output Swing High vs. Supply Voltage Output Swing Low vs. Supply Voltage 20203733 20203735 Output Swing High vs. Supply Voltage Output Swing Low vs. Supply Voltage 20203732 20203734 13 www.national.com LMP7707/LMP7708/LMP7709 Input Voltage Noise vs. Frequency LMP7707/LMP7708/LMP7709 THD+N vs. Frequency THD+N vs. Output Voltage 20203728 20203729 Crosstalk Rejection Ratio vs. Frequency (LMP7708/LMP7709) 20203753 www.national.com 14 LMP7707/LMP7708/LMP7709 The LMP7707/LMP7708/LMP7709 devices are single, dual and quad low offset voltage, rail-to-rail input and output precision amplifiers each with a CMOS input stage and the wide supply voltage range of 2.7V to 12V. The LMP7707/ LMP7708/LMP7709 have a very low input bias current of only ±200 fA at room temperature. The wide supply voltage range of 2.7V to 12V over the extensive temperature range of −40°C to 125°C makes either the LMP7707, LMP7708 or LMP7709 an excellent choice for low voltage precision applications with extensive temperature requirements. The LMP7707/LMP7708/LMP7709 have only ±37 µV of typical input referred offset voltage and this offset is guaranteed to be less than ±500 µV for the single and ±520 µV for the dual and quad over temperature. This minimal offset voltage allows more accurate signal detection and amplification in precision applications. The low input bias current of only ±200 fA along with the low give the LMP7707/ input referred voltage noise of 9 nV/ LMP7708/LMP7709 superior qualities for use in sensor applications. Lower levels of noise introduced by the amplifier mean better signal fidelity and a higher signal-to-noise ratio. The LMP7707/LMP7708/LMP7709 are stable for a gain of 6 or higher. With proper compensation though, the LMP7707, LMP7708 or LMP7709 can be operational at a gain of ±1 and still maintain much faster slew rates than comparable fully compensated amplifiers. The increase in bandwidth and slew rate is obtained without any additional power consumption. National Semiconductor is heavily committed to precision amplifiers and the market segment they serve. Technical support and extensive characterization data is available for sensitive applications or applications with a constrained error budget. The LMP7707 is offered in the space saving 5-Pin SOT23 package, the LMP7708 comes in the 8-pin MSOP and the LMP7709 is offered in the 14-Pin TSSOP package. These small packages are ideal solutions for area constrained PC boards and portable electronics. 20203721 FIGURE 1. Isolating Capacitive Load INPUT CAPACITANCE CMOS input stages inherently have low input bias current and higher input referred voltage noise. The LMP7707/LMP7708/ LMP7709 enhances this performance by having the low input bias current of only ±200 fA, as well as a very low input re. In order to achieve this a ferred voltage noise of 9 nV/ large input stage has been used. This large input stage increases the input capacitance of the LMP7707/LMP7708/ LMP7709. The typical value of this input capacitance, CIN, for the LMP7707/LMP7708/LMP7709 is 25 pF. The input capacitance will interact with other impedances such as gain and feedback resistors, which are seen on the inputs of the amplifier, to form a pole. This pole will have little or no effect on the output of the amplifier at low frequencies and DC conditions, but will play a bigger role as the frequency increases. At higher frequencies, the presence of this pole will decrease phase margin and will also cause gain peaking. In order to compensate for the input capacitance, care must be taken in choosing the feedback resistors. In addition to being selective in picking values for the feedback resistor, a capacitor can be added to the feedback path to increase stability. CAPACITIVE LOAD The LMP7707/LMP7708/LMP7709 devices can each be connected as a non-inverting voltage follower. This configuration is the most sensitive to capacitive loading. The combination of a capacitive load placed on the output of an amplifier along with the amplifier’s output impedance creates a phase lag which in turn reduces the phase margin of the amplifier. If the phase margin is significantly reduced, the response will be either underdamped or it will oscillate. In order to drive heavier capacitive loads, an isolation resistor, RISO, as shown in the circuit in Figure 1 should be used. By using this isolation resistor, the capacitive load is isolated from the amplifier’s output, and hence, the pole caused by CL is no longer in the feedback loop. The larger the value of RISO, the more stable the output voltage will be. If values of RISO are sufficiently large, the feedback loop will be stable, independent of the value of CL. However, larger values of RISO result in reduced output swing and reduced output current drive. 20203744 FIGURE 2. Compensating for Input Capacitance Using this compensation method will have an impact on the high frequency gain of the op amp, due to the frequency dependent feedback of this amplifier. Low gain settings can, again, introduce instability issues. DIODES BETWEEN THE INPUTS The LMP7707/LMP7708/LMP7709 have a set of anti-parallel diodes between the input pins, as shown in Figure 3. These diodes are present to protect the input stage of the amplifier. At the same time, they limit the amount of differential input voltage that is allowed on the input pins. A differential signal larger than one diode voltage drop might damage the diodes. The differential signal between the inputs needs to be limited to ±300 mV or the input current needs to be limited to ±10 mA. Exceeding these limits will damage the part. 15 www.national.com LMP7707/LMP7708/LMP7709 Application Information LMP7707/LMP7708/LMP7709 HIGH IMPEDANCE SENSOR INTERFACE Many sensors have high source impedances that may range up to 10 MΩ. The output signal of sensors often needs to be amplified or otherwise conditioned by means of an amplifier. The input bias current of this amplifier can load the sensor’s output and cause a voltage drop across the source resistance as shown in Figure 5, where VIN + = VS – IBIAS*RS The last term, IBIAS*RS, shows the voltage drop across RS. To prevent errors introduced to the system due to this voltage, an op amp with very low input bias current must be used with high impedance sensors. This is to keep the error contribution by IBIAS*RS less than the input voltage noise of the amplifier, so that it will not become the dominant noise factor. The LMP7707/LMP7708/LMP7709 have very low input bias current, typically 200 fA. 20203725 FIGURE 3. Input of the LMP7707 TOTAL NOISE CONTRIBUTION The LMP7707/LMP7708/LMP7709 have very low input bias current, very low input current noise and very low input voltage noise. As a result, these amplifiers are ideal choices for circuits with high impedance sensor applications. Figure 4 shows the typical input noise of the LMP7707/ LMP7708/LMP7709 as a function of source resistance. The total noise at the input can be calculated using Equation 1. (1) Where: eni is the total noise on the input. en denotes the input referred voltage noise ei is the voltage drop across source resistance due to input referred current noise or ei = RS * in et is the thermal noise of the source resistance The input current noise of the LMP7707/LMP7708/LMP7709 is so low that it will not become the dominant factor in the total noise unless source resistance exceeds 300 MΩ, which is an unrealistically high value. As is evident in Figure 4, at lower RS values, the total noise is dominated by the amplifier’s input voltage noise. Once RS is larger than a few kilo-Ohms, then the dominant noise factor becomes the thermal noise of RS. As mentioned before, the current noise will not be the dominant noise factor for any practical application. 20203759 FIGURE 5. Noise Due to IBIAS USAGE OF DECOMPENSATED AMPLIFIERS This section discusses the differences between compensated and decompensated op amps and presents the advantages of decompensated amplifiers. In high gain applications decompensated amplifiers can be used without any changes compared to standard amplifiers. However, for low gain applications special frequency compensation measures have to be taken to ensure stabilitiy. Feedback circuit theory is discussed in detail, in particular as it applies to decompensated amplifiers. Bode plots are presented for a graphical explanation of stability analysis. Two solutions are given for creating a feedback network for decompensated amplifiers when relatively low gains are required: A simple resistive feedback network and a more advanced frequency dependent feedback network with improved noise performance. Finally, a design example is presented resulting in a practical application. The results are compared to fully compensated amplifiers (National Semiconductors LMP7701/LMP7702/LMP7704). COMPENSATED AMPLIFIERS A (fully) compensated op amp is designed to operate with good stability down to gains of ±1. For this reason, the compensated op amp is also called a unity gain stable op amp. Figure 6 shows the Open Loop Response of a compensated amplifier. 20203758 FIGURE 4. Total Input Noise www.national.com 16 LMP7707/LMP7708/LMP7709 202037aa 202037a9 FIGURE 6. Open Loop Frequency Response Compensated Amplifier (LMP7701) FIGURE 7. Open Loop Frequency Response Decompensated Amplifier (LMP7707) This amplifier is unity gain stable, because the phase shift is still < 180°, when the gain crosses 0 dB (unity gain). Stability can be expressed in two different ways: Phase Margin This is the phase difference between the actual phase shift and 180°, at the point where the gain is 0 dB. Gain Margin This is the gain difference relative to 0 dB, at the frequency where the phase shift crosses the 180°. As shown in Figure 7, the reduced internal compensation moves the first pole to higher frequencies. The second open loop pole for the LMP7707/LMP7708/LMP7709 occurs at 4 MHz. The extrapolated unity gain (see dashed line in Figure 7) occurs at 14 MHz. An ideal two pole system would give a phase margin of > 45° at the location of the second pole. Unfortunately, the LMP7707/LMP7708/LMP7709 have parasitic poles close to the second pole, giving a phase margin closer to 0°. The LMP7707/LMP7708/LMP7709 can be used at frequencies where the phase margin is > 45°. The frequency where the phase margin is 45° is at 2.4 MHz. The corresponding value of the open loop gain (also called GMIN) is 6 times. Stability has only to do with the loop gain and not with the forward gain (G) of the op amp. For high gains, the feedback network is attenuating and this reduces the loop gain; therefore the op amp will be stable for G > GMIN and no special measures are required. For low gains the feedback network attenuation may not be sufficient to ensure loop stability for a decompensated amplifier. However, with an external compensation network decompensated amplifiers can still be made stable while maintaining their advantages over unity gain stable amplifiers. The amplifier is supposed to be used with negative feedback but a phase shift of 180° will turn the negative feedback into positive feedback, resulting in oscillations. A phase shift of 180° is not a problem when the gain is smaller than 0 dB, so the critical point for stability is 180° phase shift at 0 dB gain. The gain margin and phase margin express the margin enhancing overall stability between the amplifiers response and this critical point. DECOMPENSATED AMPLIFIERS Decompensated amplifiers, such as the LMP7707/LMP7708/ LMP7709, are designed to maximize the bandwidth and slew rate without any additional power consumption over the unity gain stable op amp. That is, a decompensated op amp has a higher bandwidth to power ratio than an equivalent compensated op amp. Compared with the unity gain stable amplifier, the decompensated version has the following advantages: 1. 2. 3. EXTERNAL COMPENSATION FOR GAINS LOWER THAN GMIN. This section explains how decompensated amplifiers can be used in configurations requiring a gain lower than GMIN. In the next sections the concept of the feedback factor is introduced. Subsequently, an explanation is given how stability can be determined using the frequency response curve of the op amp together with the feedback factor. Using the circuit theory, it will be explained how decompensated amplifiers can be stabilized at lower gains. A wider closed loop bandwidth. Better slew rate due to reduced compensation capacitance within the op amp. Better Full Power Bandwidth, given with Equation 2. (2) FEEDBACK THEORY Stability issues can be analyzed by verifying the loop gain function GF, where G is the open loop gain of the amplifier and F is the feedback factor of the feedback circuit. The feedback function (F) of arbitrary electronic circuits, as shown in Figure 8, is defined as the ratio of the input and output signal of the same circuit. Figure 7 shows the frequency response of the decompensated amplifier. 17 www.national.com LMP7707/LMP7708/LMP7709 (20 dB). This is shown as the dashed line in Figure 9. The resistor choice of RF = R1 = 2 kΩ makes the inverse feedback equal 2 V/V (6 dB), shown in Figure 9 as the solid line. The intercept of G and 1/F represents the frequency for which the loop gain is identical to 1 (0 dB). Consequently, the total phase shift at the frequency of this intercept determines the phase margin and the overall system stability. In this system example 1/F crosses the open loop gain at a frequency which is larger than the frequency where GMIN occurs, therefore this system has less than 45° phase margin and is most likely instable. 20203796 FIGURE 8. Op Amp with Resistive Feedback. (a) Noninverting (b) Inverting The feedback function for a three-terminal op amp as shown in Figure 8 is the feedback voltage VA – VB across the op amp input terminals relative to the op amp output voltage, VOUT. That is (3) GRAPHICAL EXPLANATION OF STABILITY ANALYSIS Stability issues can be observed by verifying the closed loop gain function GF. In the frequencies of interest, the open loop gain (G) of the amplifier is a number larger than 1 and therefore positive in dB. The feedback factor (F) of the feedback circuit is an attenuation and therefore negative in dB. For calculating the closed loop gain GF in dB we can add the values of G and F (both in dB). One practical approach to stabilizing the system, is to assign a value to the feedback factor F such that the remaining loop gain GF equals one (unity gain) at the frequency of GMIN. This realizes a phase margin of 45° or greater. This results in the following requirement for stability: 1/F > GMIN. The inverse feedback factor 1/F is constant over frequency and should intercept the open loop gain at a value in dB that is greater than or equal to GMIN. The inverse feedback factor for both configurations shown in Figure 8, is given by: 202037a2 FIGURE 9. 1/F for RF = R1 and Open Loop Gain Plot RESISTIVE COMPENSATION A straightforward way to achieve a stable amplifier configuration is to add a resistor RC between the inverting and the non-inverting inputs as shown in Figure 10. (4) The closed loop gain for the non-inverting configuration (a) is: 20203797 FIGURE 10. Op Amp with Compensation Resistor between Inputs (5) This additional resistor RC will not affect the closed loop gain of the amplifier but it will have positive impact on the feedback network. The inverse feedback function of this circuit is: The closed loop gain for the inverting configuration (b) is: (6) For stable operation the phase margin must be equal to or greater than 45° . The corresponding closed loop gain GMIN, for a non-inverting configuration, is (9) Proper selection of the value of RC results in the shifting of the 1/F function to GMIN or greater, thus fulfilling the condition for circuit stability. The compensation technique of reducing the loop gain may be used to stabilize the circuit for the values given in the previous example, that is GMIN = 20 dB and RF = R1 = 2 kΩ. A resistor value of 250 Ω applied between the amplifier inputs shifts the 1/F curve to the value GMIN (20 dB) as shown by the dashed line in Figure 11. This results in overall stability for the circuit. This figure shows a combination of the open and closed loop gain and the inverse feedback function. (7) For an inverting configuration: (8) If R1 and RF and are chosen so that the closed loop gain is lower than the minimum gain required for stability, then 1/F intersects the open loop gain curve for a value that is lower than GMIN. For example, assume the GMIN is equal to 10 V/V www.national.com 18 However, adding RC results in reduced loop gain and increased noise gain. The noise gain is defined as the inverse of the feedback factor, F. The noise gain is the gain from the amplifier input referred noise to the output. In effect, loop gain is traded for stability. 2. The ideal closed loop gain retains the same value as the circuit without the compensation resistor RC. LEAD-LAG COMPENSATION This section presents a more advanced compensation technique that can be used to stabilize amplifiers. The increased noise gain of the prior circuit is prevented by reducing the low frequency attenuation of the feedback circuit. This compensation method is called Lead-Lag compensation. Lead-lag compensation components will be analyzed and a design example using this procedure will be discussed. The feedback function in a lead-lag compensation circuit is shaped using a resistor and a capacitor. They are chosen in a way that ensures sufficient phase margin. Figure 13 shows a Bode plot containing: the open loop gain of the decompensated amplifier, a feedback function without compensation and a feedback function with lead-lag compensation. 202037a3 FIGURE 11. Compensation with Reduced Loop Gain The technique of reducing loop gain to stabilize a decompensated op amp circuit will be illustrated using the non-inverting input configuration shown in Figure 12. 20203798 FIGURE 12. Closed Loop Gain Analysis with RC The effect of the choice of resistor RC in Figure 12 on the closed loop gain can be analyzed in the following manner: Assume the voltage at the inverting input of the op amp is VX. Then, (10) Where G is the open loop gain of the op amp. 202037a5 FIGURE 13. Bode Plot of Open Loop gain G and 1/F with and without Lead-Lag Compensation (11) Combining Equation 10, Equation 11, and Equation 9 produces the following equation for closed loop gain, The shaped feedback function presented in Figure 13 can be realized using the amplifier configuration in Figure 14. Note that resistor RP is only used for compensation of the input voltage caused by the IBIAS current. R P can be used to introduce more freedom for calculating the lead-lag components. This will be discussed later in this section. (12) 19 www.national.com LMP7707/LMP7708/LMP7709 By inspection of Equation 12, RC does not affect the ideal closed loop gain. In this example where RF = R1, the closed loop gain remains at 6 dB as long as GF >> 1. The closed loop gain curve is shown as the solid line in Figure 11. The addition of RC affects the circuit in the following ways: 1. 1/F is moved to a higher gain, resulting in overall system stability. This example, represented by Figure 8 and Figure 9, is generic in the sense that the GMIN as specified did not distinguish between inverting and non-inverting configurations. LMP7707/LMP7708/LMP7709 Note that the constraint 1/F ≥ Gmin needs to be satisfied only in the vicinity of the intersection of G and 1/F; 1/F can be shaped elsewhere as needed. Two rules must be satisfied in order to maintain adequate phase margin. Rule 1 The plot of 1/F should intersect with the plot of the open loop gain at a value larger than GMIN. At that point, the open loop gain G has a phase margin of 45°. The location f2 in Figure 15 illustrates the proper intersection point for the LMP7707/LMP7708/ LMP7709 using the circuit of Figure 14. The intersection of G and 1/F at the op amp's second pole location is the 45° phase margin reference point. Rule 2 The 1/F pole (see Figure 15) should be positioned at the frequency that is at least one decade below the intersection point f2 of 1/F and G. This positioning takes full advantage of the 90° of phase lead brought about by the 1/F pole. This additional phase lead accompanies the increase in magnitude of 1/F observed at frequencies greater than the 1/F pole. 20203765 FIGURE 14. LMP7707 with Lead-Lag Compensation for Inverting Configuration The inverse feedback factor of the circuit in Figure 14 is: The resulting system has approximately 45° of phase margin, based upon the fact that the open loop gain's dominant pole and the second pole are more than one decade apart and that the open loop gain has no other pole within one decade of its intersection point with 1/F. If there is a third pole in the open loop gain G at a frequency greater than f2 and if it occurs less than a decade above that frequency, then there will be an effect on phase margin. (13) The pole of the inverse feedback function is located at: (14) The zero of the inverse feedback function is located at: DESIGN EXAMPLE The input lead-lag compensation method can be applied to an application using the LMP7707, LMP7708 or LMP7709 in an inverting configuration, as shown in Figure 14. (15) The low frequency inverse feedback factor is given by: (16) The high frequency inverse feedback factor is given by: (17) From these formulas, we can tell that 1. The 1/F's zero is located at a lower frequency compared to 1/F's pole. 2. The intersection point of 1/F and the open loop gain G is determined by the choice of resistor values for RP and RC if the values of R1 and RF are set before compensation. 3. This procedure results in the creation of a pole-zero pair, the positions of which are interdependent. 4. This pole-zero pair is used to: — Raise the 1/F value to a greater value in the region immediately to the left of its intercept with the A function in order to meet the Gmin requirement. — Achieve the preceding with no additional loop phase delay. 5. The location of the 1/F zero is determined by the following conditions: — The value of 1/F at low frequency. — The value of 1/F at the intersection point. — The location of 1/F pole. www.national.com 202037ab FIGURE 15. LMP7707 Open Loop Gain and 1/F Lead-Lag Feedback Network. Figure 15 shows that GMIN = 16 dB and f2 (intersection point) = 2.4 MHz. A gain of 6 dB (or a magnitude of –1) is well below the LMP7707’s GMIN. Without external lead-lag compensation, the inverse feedback factor is found using Equation 4 which applies to both inverting and non-inverting configurations. Unity gain implementation for the inverting configuration means RF = R1, and 1/F = 2 (6 dB). 20 This method uses bode plot approximation. Some fine-tuning may be needed to get the best results. Calculations: As described in Step 1, use Equation 17. 202037a7 FIGURE 16. Bench Results for Lead- Lag Compensation The top waveform shows the output response of a uncompensated LMP7707 using no external compensation components. This trace shows ringing and is unstable (as expected). The middle waveform is the response of a compensated LMP7707 using the compensation components calculated with the described procedure. The response is reasonably well behaved. The bottom waveform shows the response of an overcompensated LMP7707. Finally, Figure 17 compares the step response of the compensated LMP7707 to that of the unity gain stable LMP7701. The increase in dynamic performance is clear. (18) Now substitute RF/R1 = 1 into the equation above since this is a unity gain, inverting amplifier, then (19) According to Step 2 use Equation 14 (20) which leads to: (21) Choose a value of RF that is below 2 kΩ, in order to minimize the possibility of shunt capacitance across high value resistors producing a negative effect on high frequency operation. If RF = R1 = 1 kΩ, then RF // R1 = 500 Ω. For simplicity, choose RP = 0 Ω . The value of RC is derived from Equation 19 and has a value of RC = 250 Ω. This is not a standard value. A value of RC = 330 Ω is a first choice (using 10% tolerance components). The value of capacitor C is 2.2 nF. This value is significantly higher than the parasitic capacitances associated with passive components and board layout, and is therefore a good solution. Bench results: For bench evaluation the LMP7707 in an inverting configuration has been verified under three different conditions: • Uncompensated. • Lead-lag compensation resulting in a phase margin of 45°. • Lead lag overcompensation resulting in a phase margin larger than 45°. The calculated components for these three conditions are Condition RC 202037a6 FIGURE 17. Bench Results for Comparison of LMP7701 and LMP7707 The application of input lead-lag compensation to a decompensated op amp enables the realization of circuit gains of less than the minimum specified by the manufacturer. This is accomplished while retaining the advantageous speed versus power characteristic of decompensated op amps. C Uncompensated NA NA Compensated 330 Ω 2.2 nF Overcompensated 240 Ω 3.3 nF 21 www.national.com LMP7707/LMP7708/LMP7709 Figure 16 shows the results of the compensation of the LMP7707. Procedure: The compensation circuit shown in Figure 14 is implemented. The inverse feedback function is shaped by the solid line in Figure 15. The 1/F plot is 6 dB at low frequencies. At higher frequencies, it is made to intersect the loop gain G at frequency f2 with gain amplitude of 16 dB (GMIN), which equals a magnitude of six times. This follows the recommendations in Rule 1. The 1/F pole fp is set one decade below the intersection point (f2 = 2.4 MHz) as given in Rule 2, and results in a frequency fp = 240 kHz. The next steps should be taken to calculate the values of the compensation components: Step 1) Set 1/F equal to GMIN using Equation 17. This gives a value for resistor RC. Step 2) Set the 1/F pole one decade below the intersection point using Equation 14. This gives a value for capacitor C. LMP7707/LMP7708/LMP7709 Physical Dimensions inches (millimeters) unless otherwise noted 5-Pin SOT23 NS Package Number MF05A 8-Pin MSOP NS Package Number MUA08A www.national.com 22 LMP7707/LMP7708/LMP7709 14-Pin TSSOP NS Package Number MTC14 23 www.national.com LMP7707/LMP7708/LMP7709 Precision, CMOS Input, RRIO, Wide Supply Range Decompensated Amplifiers Notes THE CONTENTS OF THIS DOCUMENT ARE PROVIDED IN CONNECTION WITH NATIONAL SEMICONDUCTOR CORPORATION (“NATIONAL”) PRODUCTS. NATIONAL MAKES NO REPRESENTATIONS OR WARRANTIES WITH RESPECT TO THE ACCURACY OR COMPLETENESS OF THE CONTENTS OF THIS PUBLICATION AND RESERVES THE RIGHT TO MAKE CHANGES TO SPECIFICATIONS AND PRODUCT DESCRIPTIONS AT ANY TIME WITHOUT NOTICE. NO LICENSE, WHETHER EXPRESS, IMPLIED, ARISING BY ESTOPPEL OR OTHERWISE, TO ANY INTELLECTUAL PROPERTY RIGHTS IS GRANTED BY THIS DOCUMENT. TESTING AND OTHER QUALITY CONTROLS ARE USED TO THE EXTENT NATIONAL DEEMS NECESSARY TO SUPPORT NATIONAL’S PRODUCT WARRANTY. EXCEPT WHERE MANDATED BY GOVERNMENT REQUIREMENTS, TESTING OF ALL PARAMETERS OF EACH PRODUCT IS NOT NECESSARILY PERFORMED. NATIONAL ASSUMES NO LIABILITY FOR APPLICATIONS ASSISTANCE OR BUYER PRODUCT DESIGN. BUYERS ARE RESPONSIBLE FOR THEIR PRODUCTS AND APPLICATIONS USING NATIONAL COMPONENTS. PRIOR TO USING OR DISTRIBUTING ANY PRODUCTS THAT INCLUDE NATIONAL COMPONENTS, BUYERS SHOULD PROVIDE ADEQUATE DESIGN, TESTING AND OPERATING SAFEGUARDS. EXCEPT AS PROVIDED IN NATIONAL’S TERMS AND CONDITIONS OF SALE FOR SUCH PRODUCTS, NATIONAL ASSUMES NO LIABILITY WHATSOEVER, AND NATIONAL DISCLAIMS ANY EXPRESS OR IMPLIED WARRANTY RELATING TO THE SALE AND/OR USE OF NATIONAL PRODUCTS INCLUDING LIABILITY OR WARRANTIES RELATING TO FITNESS FOR A PARTICULAR PURPOSE, MERCHANTABILITY, OR INFRINGEMENT OF ANY PATENT, COPYRIGHT OR OTHER INTELLECTUAL PROPERTY RIGHT. LIFE SUPPORT POLICY NATIONAL’S PRODUCTS ARE NOT AUTHORIZED FOR USE AS CRITICAL COMPONENTS IN LIFE SUPPORT DEVICES OR SYSTEMS WITHOUT THE EXPRESS PRIOR WRITTEN APPROVAL OF THE CHIEF EXECUTIVE OFFICER AND GENERAL COUNSEL OF NATIONAL SEMICONDUCTOR CORPORATION. As used herein: Life support devices or systems are devices which (a) are intended for surgical implant into the body, or (b) support or sustain life and whose failure to perform when properly used in accordance with instructions for use provided in the labeling can be reasonably expected to result in a significant injury to the user. A critical component is any component in a life support device or system whose failure to perform can be reasonably expected to cause the failure of the life support device or system or to affect its safety or effectiveness. National Semiconductor and the National Semiconductor logo are registered trademarks of National Semiconductor Corporation. All other brand or product names may be trademarks or registered trademarks of their respective holders. Copyright© 2007 National Semiconductor Corporation For the most current product information visit us at www.national.com National Semiconductor Americas Customer Support Center Email: [email protected] Tel: 1-800-272-9959 www.national.com National Semiconductor Europe Customer Support Center Fax: +49 (0) 180-530-85-86 Email: [email protected] Deutsch Tel: +49 (0) 69 9508 6208 English Tel: +49 (0) 870 24 0 2171 Français Tel: +33 (0) 1 41 91 8790 National Semiconductor Asia Pacific Customer Support Center Email: [email protected] National Semiconductor Japan Customer Support Center Fax: 81-3-5639-7507 Email: [email protected] Tel: 81-3-5639-7560