Design Goals

Design Description

This design uses discrete op amps to implement an instrumentation amplifier (IA) design using space-grade (SP) components for use in space applications. The circuit converts a differential signal to a single-ended output signal. Linear operation of an instrumentation amplifier depends upon linear operation of its building block: op amps. An op amp operates linearly when the input and output signals are within the respective input common-mode and output swing ranges of the device. The supply voltages used to power the op amps define these ranges.

Design Notes

1. Use low-tolerance resistors to achieve high DC CMRR performance. Mismatching of resistors can also lead to errors in gain and output accuracy.
2. All resistors and capacitors must be verified space-grade for this design.
3. R_g sets the gain of the input stage. R_1a and R_1b can be used to set the gain of the second stage (see Design Steps).
4. R_f1 and R_f2 are nominally matched in this design. In general, R_f1 and R_f2 do not need to be matched – it may be desirable in some cases to have R_f1 and R_f2 unmatched so that the top amplifier and bottom amplifier in the input stage have different gains. For example, if V_cm is not at mid-supply but is closer to one of the rails, R_f1 and R_f2 can be tuned so that neither of the input stage amplifiers run out of headroom.
5. Integrated instrumentation amplifiers normally have a fixed minimum gain. In addition to using an IA in high-gain configurations, constructing a discrete IA like this affords the flexibility to achieve any gain less than 1 V/V.
6. High-value resistors can degrade the phase margin of the circuit and introduce additional noise in the circuit.
7. Add an isolation resistor to the output stage to drive large capacitive loads.
8. Linear operation is contingent upon the input common-mode and the output swing ranges of the discrete op amps used. For best performance, choose V_cm = (V₊ + V_–) / 2 (mid-supply).
9. C₂ along with R₃ || R₄ forms a low-pass filter with a corner frequency of 147.16 Hz.
10. The V_ref pin must be supplied by a low-impedance reference that can sink and source current, such as a buffer. Using a high-impedance reference, such as a resistor divider with no buffer, may result in a mismatch and degradation of CMRR.
11. V_{out_min} is chosen as 0.2 V for this design to avoid nonlinearities associated with the output of LMP7704-SP swinging too close to the rail. If this design is done with a different op amp, be sure to check the data sheet to determine the minimum and maximum output values allowed.
12. The LMP7704-SP supply voltage of 10 V was selected according the derating specifications provided by the National Aeronautics and Space Administration (NASA) in document EEE-INST-002 (April 2008) and the European Cooperation for Space Standardization (ECSS) in document ECSS-Q-ST-30-11C Rev.1 (4 October 2011). The documents specify an 80% and 90% derating of the absolute maximum supply voltage for linear ICs, respectively.
13. This design can be implemented with a single 4-channel LMP7704-SP or a similar device. See Design Alternative Op Amp for a wider supply op amp (36 V). Note that the listed alternative device meets TID = 50 krad(Si).

Design Steps

1. Calculate the output voltage V_out for this circuit using the following equation:
   
   In this equation, V_d = V₂ – V₁ is the differential input voltage, V_ref is set by R₃ and R₄ to level shift the output, and it is assumed that R_1a = R_2a and R_1b = R_2b. Integrated instrumentation amplifiers normally fix R_f1, R_f2, R_1a, R_2a, R_1b, and R_2b, leaving only R_g to set the gain of the circuit. In this discrete implementation, the designer has the freedom to alter all of these resistors, but the transfer function can be simplified by using standard values, such as R_f1 = R_f2 = R_1a = R_1b = R_2a = R_2b = 10kΩ, and using only R_g to set the gain. In this case, R_g can be calculated using the following simplified equation:
   
2. Set V_ref. For this design, V_ref has been set as shown in the following equation so that a symmetric input voltage range of –50mV to +50mV results in an output voltage range of 0.2V to 5V.
   
   
   R₄ = 13.83kΩ ≈ 13.8kΩ (standard value)
   Note: The magnitudes of R₃ and R₄ were chosen such that R₃ || R₄ is close to 10kΩ so that the low-pass filter formed by R₃ || R₄ and C₂ is close to the common low-pass filter with R = 10kΩ and C = 100nF.
   
   
3. Choose R_g to set the required gain using the simplified transfer function.
   
   R_g = 425Ω ≈ 427Ω (standard value)
   This corresponds to a gain of: