Analysis of the Causes of Operational Amplifier Oscillation and Self-excitation

Aguilera

Analysis of the Causes of Operational Amplifier Oscillation and Self-excitation [Copy link]

From my personal experience, the most effective method is: 1. Is the resistance of the amplifier circuit too large? For example, I have seen a circuit that amplifies 10 times, and the resistors are 100k and 1M. (Even textbooks say so). The intrinsic circuit current is too small, the anti-interference ability is weak, and the output signal is more likely to couple in. 2. When the output trace is too long, it is not only easy to self-excite, but also the signal is easily interfered. It is best to use shielding. If there is no condition, you can consider connecting a small resistor in series with the output. Connecting a small resistor in series does not completely solve the phase problem, but also solves the abnormal operation of the op amp caused by excessive capacitive load. 3. The amplification factor should not be too large. If the output signal is too strong and the input signal is too weak, it is easy to couple. 4. Although the capacitor connected in parallel across the impedance is effective in theory, the capacitor affects the output phase consistency and the discrete type of the capacitor is too large. In the scenario where phase consistency is emphasized, the phase impact needs to be considered. 5. Grounding, an eternal topic. Reasons for the self-excitation of the op amp: 1. The loop gain is greater than 1 2. The phase difference between the signals before and after the feedback is more than 360 degrees, which means that positive feedback can be formed. Refer to "Principles of Automatic Control" and "Circuit Design Based on Operational Amplifiers and Analog Integrated Circuits" The cause of self-oscillation is mainly because the integrated operational amplifier is composed of multiple stages of DC amplifiers. Since the output of each stage of the amplifier and the input of the next stage of the amplifier have output impedance, input impedance and distributed capacitance, there is an RC phase shift network between the stages. When the signal passes through each RC network, an additional phase shift will be generated. In addition, the external bias resistor and input capacitor of the operational amplifier, the output resistor of the operational amplifier and the capacitive load feedback capacitor, as well as the common internal resistance of the multi-stage operational amplifier through the power supply, and even the distributed inductance on the power supply line, poor grounding and other couplings can all form additional phase shifts. As a result, the signal output by the operational amplifier is superimposed through the negative feedback loop to increase the additional phase shift to 180 degrees, and if the feedback amount is large enough, the negative feedback will eventually be transformed into positive feedback, thereby causing oscillation. To be more specific 1. The op amp may be caused by distributed capacitance and inductance 2. The op amp drives a capacitive load. 3. It may be caused by too deep feedback Solution: 1. In-loop compensation The feedback resistor of the operational amplifier is connected in parallel with the feedback capacitor: The small capacitor is called a phase-shifting capacitor. It prevents the operational amplifier from self-excitation. Generally, the value is from 0 to tens of picofarads or hundreds of picofarads, depending on the operating frequency and the model of the operational amplifier To put it simply, the larger the capacitance added, the narrower the bandwidth To prevent oscillation, Rf and the input capacitance and stray capacitance of the op amp form a pole. If the pole is within the frequency range of the op amp, it may cause the op amp to oscillate. After adding Cf, Cf and Rf produce a zero point to offset the pole. Generally, the value of Cf>Ci, where Ci is the input capacitance of the op amp and the total stray capacitance of the input pin. 2. External loop compensation method: Connect a small resistor in series at the output end of the op amp and then connect it to the next stage. The resistor can be between a dozen ohms and several dozen ohms. The specific value is related to the input capacitance of the next stage circuit. You can try different resistance values to obtain a stable output. PS: 1. The power supply is stable. It is best to connect capacitors such as 0.1uf and 10uf in parallel. 2. The amplification factor cannot be too large, and the number of amplification stages should not exceed four. Before the experiment or test, if you connect an oscilloscope to the output of the op amp, you can sometimes see a high-frequency waveform that is close to a sine wave, and occasionally low-frequency oscillations. Different methods can be used to solve the problem according to the principle of oscillation: (1) Is the feedback polarity connected incorrectly or is the negative feedback too strong? If the negative feedback is connected incorrectly to positive feedback, oscillation is very likely to occur. In addition, the stronger the negative feedback, the easier it is to self-excite. (2) If the output end is connected to a capacitive load, the capacitive load increases the phase shift of the circuit, making it more prone to self-excitation. Another RC link can be used to compensate for the phase shift. If the compensation is good, the self-oscillation will be eliminated. (3) The stray capacitance of the wiring is too large. When the input circuit is high impedance, the lag link composed of the stray capacitance and resistance between the wiring and the ground or between the wiring will make the component unstable. For this reason, a capacitor CF can be connected in parallel at both ends of Rf (feedback resistor), or an RC branch can be connected in parallel at the input end of the op amp. Both links are of the nature of lead correction, that is, the phase lead effect they produce will likely offset the phase lag effect of the stray capacitance mentioned above, thereby stabilizing the op amp. (4) Insufficient power wiring bypass measures. The power lead has not only a certain resistance, but also a certain inductance and distributed capacitance. Therefore, when many op amps are connected to the same power line, they will affect each other through these factors. The solution is to connect a few tens of uF electrolytic capacitors and a 0.01uF ceramic capacitor in parallel between the positive and negative power terminals on the printed circuit board socket and the ground. If the op amp is used as a broadband amplifier, a capacitor with low inductance must be selected. The old man in the research institute told me that analog electronics is an art; some people also say that analog electronics is a magic trick that is hard to grasp; in fact, I think that analog electronics is a gambling technique, just betting big or small. In the process of hands-on, we can discover the rules, combine theory with practice, and develop in a spiral way.