What does a simple triangle symbol mean?

Looking at Figure 1, which triangle represents an op amp? Which triangle represents a comparator? Which triangle represents an instrumentation amplifier? Answer:

They all are!

Figure 1. Op amps, instrumentation amplifiers, and comparators.

So, what are the differences and why should we care? From Table 1, we can see that some of the characteristics are very different, but what do they mean for circuits and systems?

Table 1. Comparison of op amps, comparators, and instrumentation amplifiers

Let’s take a look at how everyone got into trouble…

Op amps have enormous gain. We were taught in school that we start our analysis with the difference between the two inputs equal to zero. But in real life, this is not possible. If the open-loop gain is one million, then to get 5 V at the output, there must be 5 μV at the input. To make the circuit usable, we need to apply feedback, a control signal that feeds back to the input to cancel the original stimulus when the output is about to go too high—i.e., negative feedback. When used as a comparator, without feedback, the output will go straight to one rail or the other. With positive feedback, the output will be driven farther in the same direction. Therefore, op amps need negative feedback. In fact, when some op amps are used as comparators without feedback, the supply current can be 5 to 10 times higher than the maximum on the data sheet.

However, for comparators, positive feedback is what we need. Without feedback, if one input to a comparator slowly exceeds the level of the other input, the output will start to change slowly. If there is noise in the system, such as ground bounce, the output may reverse, which is of course undesirable in a control system. But then it starts to reverse, producing oscillatory behavior, sometimes called chattering (see Figure 5 in MT-083). The article "Eliminating Comparator Instability with Hysteresis" by Reza Moghimi fully describes the benefits of adding positive feedback (also called hysteresis).

Figure 2. The classic three-op-amp instrumentation amplifier

For an instrumentation amplifier, feedback is already internal and adding feedback only creates an imprecise gain. Figure 2 shows a typical way to build an instrumentation amplifier using an op amp.

Note: Every op amp has feedback. We start by using the standard negative feedback diagram (see Figure 3), with the instrumentation amplifier as G and the desired gain of 10, which means a feedback factor of 0.1. Next, the instrumentation amplifier is chosen to have a fixed gain of 100. Using Equation 1, the actual closed-loop gain will be 9.09, which is almost a 10% error. Therefore, it does not make sense to use the instrumentation amplifier as an op amp and add feedback to it.

Figure 3. Classic feedback schematic

An op amp requires negative feedback; a comparator requires positive feedback; an instrumentation amplifier does not require any feedback.

For op amps, referring to Equation 1, the higher the open-loop gain (AVOL ₎ , the more accurate the closed-loop gain will be. Most op amps have an open-loop gain between 100,000 and 10 million, but some older high-speed op amps can be as low as 3000. As mentioned previously, the higher the open-loop gain, the smaller the closed-loop gain error.

For a comparator, if the logic swing of the output is 3 V and you need a 1 mV threshold, the minimum gain is 3000. Higher gains will make the uncertainty window smaller, but if the gain is too high, microvolt noise can trigger the comparator.

For instrumentation amplifiers, the concept of open-loop gain does not apply.

Capacitors are often added to circuits to limit bandwidth. Examining Figure 4, at first glance it appears that R1 and C1 form a low-pass filter. This is not the way to go and can result in oscillation. The feedback factor for an inverting amplifier is R2/R1, but in Figure 4, it is R2/(R1 // Xc). As frequency increases, the feedback factor increases, so the noise gain increases at a rate of +20 dB/decade, while the op amp open-loop gain decreases at a rate of –20 dB/decade. They cross at 40 dB, which, according to control system theory, is a sure bet for oscillation. The correct way to limit the bandwidth of the circuit is to place a capacitor across R2.

Figure 4. Attempt to reduce op amp bandwidth.

Comparators typically do not have a negative feedback network, so a simple R and C low-pass filter in front of the comparator in Figure 5 works well. R _HYS should be much larger than R7, so that the two divide the output swing to provide a small amount of positive feedback (hysteresis). If the comparator has built-in hysteresis, such as the LTC6752 or ADCMP391, then R7 and R _HYS are not used .

Figure 5. Comparator with LPF and hysteresis

For instrumentation amplifiers, it is perfectly acceptable to place capacitors at the inputs, as shown by C4 in Figure 6. The graphic in Chapter 5 of the Analog Devices Instrumentation Guide shows a good thing to do whenever you use an instrumentation amplifier. If you lay out the printed circuit board with appropriate traces and pads to allow for the addition of two resistors and three capacitors, you can start with 0Ω resistors and no capacitors and measure the system performance. The common-mode roll-off and normal-mode roll-off can be set independently by adjusting the values ​​of the five components (see the Guide for details).

Figure 6. RFI filter before instrumentation amplifier

The output of an op amp or instrumentation amplifier swings from nearly one rail to the other. Depending on whether the output stage uses a common-emitter or common-source configuration, the output can get to within 25 mV to 200 mV of either rail. This is considered a rail-to-rail output. If the op amp is powered by +15 V and –15 V, it is not convenient to interface with digital circuits. A poor solution is to put diode clamps on the output to protect the digital inputs from damage. But instead, the op amp is damaged by excessive current. There are more complicated ways to interface op amps with digital logic, but why bother? Just use a comparator.

Comparators can have CMOS totem-pole outputs, or NPN or NMOS open-collector or open-drain outputs. Although open-collector or open-drain outputs require a pull-up resistor, resulting in unequal rise and fall times, they have the following advantages: the comparator is powered at one voltage (such as 5 V) and interfaces with logic at another voltage (such as 3.3 V).

The op amp needs a gain bandwidth that is higher than the highest signal frequency to keep closed-loop errors low. Looking at Equation 1, we know that the gain bandwidth should be 10 to 100 times the highest signal frequency. As mentioned earlier, Equation 1 shows that A _VOL is a function of frequency, which affects closed-loop accuracy. Phase margin is also important and will vary with capacitive loading, so the specification sheet should clearly state the test conditions. To ensure dc accuracy, the offset voltage should be low. For trimmed bipolar op amps, 25 μV to 100 μV is good; for FET-input op amps, 200 μV to 500 μV is good. Auto-zero/chopped/zero-drift op amps are almost always less than 20 μV (maximum) over temperature. Check the data sheets for some typical op amps, such as the OP27, AD8610, or ADA4522.

Figure 7. Bidirectional current sensing with high common-mode swing

Propagation delay is a key specification for comparators. Unlike op amps, which slow down when overdriven, comparators speed up when overdriven. Spec sheets sometimes give a propagation delay for a small amount of overdrive, such as 5 mV, and a different propagation delay for a larger overdrive of 50 mV or even 100 mV.

The most important specification for an instrumentation amplifier is the common-mode rejection ratio (CMRR), since the application requires the extraction of a very small differential-mode signal that sits on top of a large common-mode voltage. Like many specifications, this varies with frequency, and sometimes the DC CMRR or the CMRR at very low frequencies is listed. A plot of CMRR vs. frequency is often provided. For example, this plot will be very important when sensing current in an H-bridge motor driver, as shown in Figure 7.

This is probably the most difficult application for an instrumentation amplifier, because the common-mode voltage changes from near one rail to near the other, and the current reverses rapidly. Gain-bandwidth and slew rate are both important.

Programming here does not mean writing code, it means configuring the device to meet system requirements (although some in-amps do have traditional software programming capabilities through the SPI port and registers).

The op amp needs to be configured for negative feedback. This can be a purely resistive element, but often a resistor is used in parallel with the capacitor to limit the bandwidth. This helps improve the signal-to-noise ratio because the noise is integrated over the entire range, even if we only use a portion of it. It is also possible to use only capacitors and have an integrator or differentiator.

Comparators should always have a little positive feedback to ensure that once the input forces the output to move, the output will move (see Figure 4 and Figure 5). See MT-083 for pictures and calculations. Some comparators have internal hysteresis, but you can usually add more hysteresis if needed. Some comparators with internal hysteresis have a pin to add a resistor to change the amount of hysteresis.

An op amp can be used as a comparator, but it's not ideal and there are some things to be aware of. You have to be a good analog designer to do this well. MT-083 covers some of the considerations, and there are many articles discussing the pros and cons. If you're not afraid of danger, you can check out the references.

Comparators are almost always programmed with resistors. You can add a high value resistor to provide a little positive feedback, and you can use a capacitor to provide AC feedback to avoid adding DC hysteresis. Some comparators have built-in hysteresis, but this can also be increased by adding a small amount of positive feedback.

There are subtle things going on when trying to use an op amp as a comparator. There are quite a few low noise bipolar op amps that have anti-parallel diodes between the inputs. Most comparators have an input common mode range of 80% or more of the total range. However, some low noise bipolar op amps have one or two series diodes between the inputs. This is to prevent the input stage from forming a zener effect with one of the emitter base junctions, causing noise performance to degrade over time.

In a 3.3 V system, if a 5 V op amp is used as a comparator with a power good indicator threshold level of 3 V, then one input at 3 V and the other at 0 V is a problem because the diodes limit the maximum differential voltage allowed at the op amp inputs.

For many applications, the choice of op amp depends on whether the user focuses on DC accuracy, AC accuracy, input offset voltage, gain bandwidth, or supply voltage. In 2020, there are over 700 devices to choose from. The key parameters for comparators are usually propagation delay and supply voltage. The choice is relatively easy, with 122 devices to choose from. The main criteria for instrumentation amplifiers is CMRR vs. frequency, but offset voltage and gain accuracy are also important near DC. Since instrumentation amplifiers are specialized devices, there are "only" 63 to choose from.

Only by choosing the right components can you achieve products and designs that are trouble-free and can be manufactured in large quantities for years to come.

Digital health technology is key to preventing future outbreaks

ADI Nanosensor Technology Enables Ultra-Rapid Detection to Fight Disease

The editor will randomly select 5 lucky winners from the fans who liked the video to receive ADI lucky small prizes