Experience summary of common problems in amplifier circuit design

Modern integrated op amps (op amp) and instrumentation Amplifiers (In-amp) offer a number of benefits to design engineers compared to discrete devices. Although many clever, useful and attractive circuits are provided. This is often the case because a hasty assembly of the circuit ignores some very basic problems, causing the circuit to fail to achieve its intended function-or it may not work at all. This article will discuss some of the most common application problems and give practical solutions.

DC bias current loop is missing for AC coupling

One of the most frequently encountered application problems is the DC (DC) circuit where the bias current is not provided in an AC (AC) coupled op amp or in an instrumentation amplifier circuit. In Figure 1, a capacitor is connected to the parallel input of the OP amp for AC coupling, a simple way to isolate the DC component of the input voltage (VIN). This is especially useful in high-gain applications where the low DC voltage at the OP amp input limits the dynamic range and even causes the output to saturate. However, there is a problem with capacitive coupling at the high impedance input and without DC access to the current at the same-phase input.

　　

Figure 1: Error op amp ac-Coupled input

In fact, the input bias current flows into the coupled capacitor and charges it until it exceeds the rated value of the common-mode voltage of the amplifier input circuit or limits the output. Depending on the polarity of the input bias current, the capacitor is charged to the positive or negative voltage of the power supply. The amplifier 's closed-loop DC gain amplifies the bias voltage.

This process can take a long time. For example, a FET input amplifier , when the bias current of 1 PA is coupled with a 0.1μf capacitor, its charging rate i/c to 10–12/10–7=10μv/s, or 600μv per minute. If the gain is 100, then the output drifts to 0.06 V per minute. As a result, general laboratory testing (using an AC-coupled oscilloscope) does not detect this problem, and the circuit does not cause problems until a few hours later. Obviously, it is very important to avoid this problem altogether.

　　

Figure 2: The correct dual-supply op amp ac-Coupled Input method

Figure 2 shows a simple solution to this common problem. Here, a resistor is connected between the input of the OP amp and the ground, providing a ground loop for the input bias current. In order to minimize the offset voltage caused by the input bias current, when using a bipolar op amp , the bias current of its two inputs should be equal, so the resistor value of the R1 should be set equal to the parallel resistance of R2 and R3.

It should be noted, however, that the resistor R1 always introduces some noise into the circuit, so there is a tradeoff between the circuit input impedance, the size of the input coupling capacitor, and the Johnson noise caused by the resistor. Typical resistor values are typically between 100,000ω~1 mω.

Similar problems can occur in the instrumentation amplifier circuitry. Figure 3 shows an instrumentation amplifier circuit that uses two capacitors for AC coupling, and does not provide a return path to the input bias current. This problem is common in instrumentation amplifier circuits that use dual power supplies (Figure 3a) and single supply (Figure 3b).

　　

Figure 3: Example of an AC-coupled instrumentation amplifier that does not work

This problem can also occur in the transformer coupling amplifier circuit, 4, if the transformer secondary circuit does not provide a DC-to-ground circuit, the problem arises.

　　

Figure 4: Non-working transformer-coupled instrumentation amplifier circuit

Figure 5 and Figure 6 show a simple solution to these circuits. Here, a high resistance resistor (RA,BR) is connected between each input and the ground. This is a simple and practical solution for dual-power instrumentation amplifier circuits.

　　

Figure 5. A high resistance resistor is connected between each input and ground to provide the required bias current loop.

A. Dual power supply. B. Single power supply.

These two resistors provide a discharge circuit for the input bias current. In the dual supply example shown in Figure 5, the reference ends of the two inputs are grounded. In the single-supply example shown in Figure 5b, the reference end of the two inputs or ground (VCM ground) or a bias voltage, usually half the maximum input voltage.

The same principle can also be applied to the transformer coupling input circuit (see Figure 6), unless the transformer secondary has an intermediate tap, it can be grounded or connected to VCM.

In this circuit, a small offset voltage error is generated due to mismatch between the two input resistors and/or mismatch of input bias currents at both ends. To minimize the offset error, a resistor can be connected between the two inputs of the instrumentation amplifier (i.e. bridging between two resistors) with a resistance of approximately 1/10 of the first two resistors (but still large compared to the differential source impedance).

　　

Figure 6: Correct instrumentation amplifier Transformer input Coupling method

Provides reference voltages for instrumentation Amplifiers , op amps , and ADCs

Figure 7 shows a single-supply circuit for a single-ended input analog-to-digital converter (ADC) that is driven by an instrumentation amplifier . The reference voltage of the amplifier provides a bias voltage corresponding to the 0 differential input, while the reference voltage of the ADC provides a scale factor. A simple RC low-pass antialiasing filter is usually used between the output of the instrumentation amplifier and the input of the ADC to reduce the out-of-band Noise. Design engineers often want to use simple methods, such as resistive voltage divider, to provide reference voltages for instrumentation amplifiers and ADCs. As a result, errors occur when using some instrumentation amplifiers .

　　

Figure 7: Typical single-supply circuit for instrumentation amplifier -driven ADCs

Correctly provides the reference voltage of the instrumentation amplifier

It is generally assumed that the reference input of the instrumentation amplifier is of high impedance because it is an input terminal. So the design engineer generally want to be in the reference terminal pin of the instrumentation amplifier to a high impedance source, such as a resistor voltage divider. This can cause serious errors in the use of certain types of instrumentation amplifiers (see Figure 8).

　　

Figure 8. Incorrectly using a simple resistor divider to directly drive the reference voltage pin of the 3 op amp instrumentation Amplifier

For example, the popular instrumentation amplifier design configuration uses the three OP amp structures shown. Its total signal gain is

　　

The gain of the reference voltage input is 1 (if input from a low-impedance voltage source). However, in the circuit shown, the reference input terminal pin of the instrumentation amplifier is directly connected to a simple voltage divider. This will change the symmetry of the subtraction circuit and the voltage divider ratio of the converter. This also reduces the common-mode rejection ratio and gain accuracy of the instrumentation amplifier . However, if the access to the R4, then the equivalent resistance of the resistor will be reduced, the reduction of the resistance value is equal to the voltage divider from the two parallel branch view of the past resistance (three KΩ), the circuit is represented as a supply voltage of half the low impedance voltage source is added to the original value R4, the accuracy of the subtraction circuit remains unchanged.

This method cannot be used if the instrumentation amplifier is in a closed, single-package form (an IC). Also consider that the temperature coefficients of the voltage divider should be consistent with those in the R4 and subtraction resistors. Finally, the reference voltage is not adjustable. On the other hand, if you try to reduce the resistance of the voltage divider, the increased resistance size can be ignored, which increases the consumption of the supply current and the power consumption of the circuit. In any case, this clumsy approach is not a good design solution.

Figure 9 shows a better solution for adding a low-power op- amp buffer between the voltage divider and the instrumentation amplifier reference input. This eliminates the problem of impedance matching and temperature coefficient matching, and it is easy to adjust the reference voltage.

　　

Figure 9: Reference voltage input for driving an instrumentation amplifier using a low output impedance op amp

PSR performance should be guaranteed when a voltage divider is used from the supply voltage to provide a reference to the amplifier

A frequently overlooked problem is that any noise, transients, or drift of the power supply voltage vs will be directly added to the output by a reference input that is attenuated by the divider ratio. Practical solutions include bypass filtering and even reference voltages generated using a precision reference voltage IC, such as ADR121, instead of VS voltage divider.

This is important to consider when designing circuits with instrumentation Amplifiers and op amps . The power supply voltage rejection technology is used to isolate the amplifier from the AC Sound, noise, and any transient voltage changes in its supply voltage. This is important because many of the actual circuits are contained, connected, or present in an environment that only provides a non-ideal supply voltage. In addition, the AC signal in the power line will be fed back into the circuit and will cause parasitic oscillations under appropriate conditions.

Modern op amps and Instrumentation Amplifiers offer a fairly low frequency supply voltage rejection (PSR) capability as part of their design. This is what most engineers seem to take for granted. Many modern op amps and Instrumentation Amplifiers have a PSR indicator above 80~100db, which can attenuate the change in supply voltage to 1/10,000~1/100,000. Even the most moderate of the 1/100-DB PSR of the amplifier isolation can inhibit the power supply as well. However, high-frequency bypass capacitors are always required (as shown in the 1~7) and often play an important role.

In addition, when a design engineer uses a simple supply voltage resistor divider and uses an op amp buffer to provide a reference voltage to the instrumentation amplifier , any change in the supply voltage is passed through the circuit without attenuation directly into the instrumentation amplifier Output stage. Therefore, unless a low-pass filter is provided, the usual good PSR performance of the IC is lost.

In Figure 10, a large capacitor is added to the output of the voltage divider to filter out changes in the supply voltage and to ensure PSR performance. The -3 db poles of the filter are determined by resistor r1/r2 parallel and capacitor C1. The -3 db pole should be set at 1/10 of the lowest useful frequency.

　　

Figure 10: Reference-side decoupling circuitry to ensure PSR Performance

The CF trial value shown above can provide a –3 db pole frequency of approximately 0.03 Hz. A small capacitor (0.01μf) connected to the R3 ends minimizes the noise of the resistor.

It takes time for the filter to charge. According to the trial value, the rise time of the reference input should be several times the time constant (here t=r3cf= 5 s), or 10~15s.

The circuit in Figure 11 has been further improved. Here, the op- amp buffer acts as an active filter, which allows the use of capacitors with a much smaller capacitance to decouple the same large power supply. In addition, the active filter can be used to raise the Q value to accelerate the on-time.

　　

Figure 11: Connecting the OP amp buffer to the reference input pin of the active filter driver Instrumentation amplifier

Test Result: The 6 v reference voltage of the instrumentation amplifier is filtered by applying a V supply voltage with the displayed component value. Set the gain of the instrumentation amplifier to 1 and modulate the power supply with a frequency-varying 1 vp-p sine signal. Under these conditions, the AC signal is not visible on the oscilloscope as the frequency is reduced to about 8 Hz. When a low-amplitude input signal is applied to the instrumentation amplifier , the test supply voltage range for the circuit is 4 V to + V. The conduction time of the circuit is approximately 2 s.

Decoupling of single-supply op amp Circuit

Finally, a single-supply op amp circuit needs to bias the common-mode input voltage amplitude to control the forward swing and negative swing of the AC signal. When a bias voltage is provided from the supply voltage using a divider, proper decoupling is required to ensure the performance of the PSR.

A common but incorrect method is to use the kω/100 KΩ resistor divider (plus 0.1μf bypass capacitor) to provide a VS/2 to the OP amp 's co-phase input. Decoupling the power supply with such a small capacitance is often not enough, since the Poles are only a few Hz. The circuit appears unstable ("low frequency oscillation"), especially when driving inductive loads.

Figure 12 (inverting input) and Figure 13 (same-phase input) show the VS/2 bias circuit that achieves the best decoupling results. In both cases, the bias voltage is added to the co-phase input, fed back to the reverse input to ensure the same bias voltage, and the unit DC gain is biased to the same output voltage. The coupling capacitor C1 reduces the low-frequency gain from BW3 to unity gain.

　　

Figure 12: The correct power-supply decoupling scheme for the single-supply co-phase input amplifier circuit. Medium Frequency gain =1+r2/r1

As shown, when using the kω/100 KΩ resistor divider, a good experience is that to obtain the 0.3 Hz –3 DB cutoff frequency, the C2 should be selected to be the smallest 10ωf. and 100μf (0.03 Hz) is actually enough for all the circuits.

　　

Figure 13: The correct decoupling circuit for a single-supply inverting input amplifier , if gain =–r2/r1

Experience summary of common problems in amplifier circuit design