How to choose analog-signal-chain components: part 2

Signal conditioning can prepare a sensor’s output for digitization.

Figure 1. An op amp in the basic inverting configuration provides a gain of -RFB/RIN.

In part 1 of this series, we looked at a typical analog signal chain that you can use in conjunction with analog-to-digital converters (ADCs) or digital-to-analog converters (DACs). A key building block of the analog signal chain is the operational amplifier (op amp), shown in its basic inverting configuration in Figure 1, where the gain is the negative of the feedback resistor R_FB divided by the input resistor R_IN.

Q: How do we derive the gain?
A: First, note that an ideal op amp has an infinite open-loop gain and infinite input impedance, so during normal operation (when the device is not saturated), both inputs will be at ground. Consequently, the current through R_IN is V_IN/R_IN, and by Kirchoff’s current law, that current must flow through R_FB., so:

and

Q: What is R_B?
A: Real op amps don’t have infinite input impedance and will draw an input bias current^[1], disturbing the voltage at the inverting input. If you are using a CMOS op amp, you might not need R_B because the input bias currents are low. However, for test-and-measurement applications, you’ll probably want to choose amplifiers with bipolar input stages to take advantage of their low noise and other performance benefits^[2]. If the bias currents on the inverting and noninverting inputs are close to the same value (a good bet for input stages fabricated on the same die), you can add an R_B whose value equals that of R_FB and R_IN in parallel.

Q: What are some other op amp configurations?
A: Figure 2a shows a voltage follower, or unity-gain buffer. This configuration is useful if, for example, you have a sensor that needs to drive a low-impedance load. Figure 2b shows a differential amplifier. Here, the output is (V₁–V₂)(R₂/R₁). Note that the output only depends on the difference between the input voltages. If all the resistors in the circuit are equal, giving us a unity-gain difference amplifier, and if V₁ is 1 V and V₂ is 4 V, then the output will be 3 V. Similarly if V₁ is 11 V and V₂ is 14 V, the output is still 3 V. In the latter case, the amplifier rejects the additional 10 V common to both inputs, called the common-mode voltage.

Figure 2. An op amp can be used as a voltage follower (a) or differential amplifier (b).

Q: What is a practical application of the differential amplifier?
A: Figure 3 shows a load powered by a battery, with a differential amplifier used to measure the current flowing through a high-side sense resistor R_SENSE. This current-measuring capability can be used as part of a control loop for diagnostics or to initiate a current-limit function when a threshold is passed.

Figure 3. A differential amplifier can monitor the load current flowing through a high-side sense resistor.

Q: Why can’t we put the sense resistor on the low side?
A: We could do that and use our single-ended amplifier from Figure 1 to monitor the current, as shown in Figure 4. This approach, however, has two problems. First, it creates a floating ground for the load (shown in blue) that differs from the earth/system ground. This may be acceptable for some applications because the voltage across the resistor will be low, generally in the millivolt range. That voltage, however, depends on load current, which may be very noisy if the load contains high-speed logic or power-switching transistors in DC/DC converters, resulting in excessive electromagnetic interference (EMI). The floating ground can also create problems if the load must communicate with other subcircuits in the system — including the current-sense amplifier itself.

Figure 4. A low-side sense resistor creates a floating ground for the load (blue) and is unable to detect load-to-system-ground shorts (red).

The second problem is that the low-side resistor is poorly placed to detect overcurrent fault conditions. For example, if a short to system ground occurs within the load, the short-circuit current will bypass the sense resistor, as shown in red in the figure, and the system will not be able to initiate corrective action. In general, it’s good to use high-side sense resistors whenever possible.

Q: What else can we do with op amps?
A: In part 3, we’ll look at some op-amp-based filters as alternatives to digital filters.