What affects the speed of high-speed operational amplifiers?

High-speed operational amplifiers (English: High Speed ​​OpAmps), referred to as high-speed op amps. High-speed operational amplifiers are mainly used in high-performance data acquisition systems in instrumentation, telecommunications, laboratories and medical systems.

As we all know, operational amplifiers are the basic building blocks for building analog circuits. They are used for a variety of signal conditioning tasks, such as voltage amplification, filtering, and mathematical operations. Of course, one of the important characteristics of an operational amplifier is speed, which is why a distinction is made between general-purpose operational amplifiers and high-speed operational amplifiers. Ideally, operational amplifiers have the characteristic of infinite input impedance at all frequencies, but in reality their speed is limited.

There are two important concepts that determine high-speed op amps: they are related to the speed of the op amp, namely bandwidth and slew rate. These two concepts are difficult to understand, especially how they relate to each other. In the following, microcontroller development engineers will introduce issues related to "bandwidth and slew rate".

1. What factors affect the speed of high-speed operational amplifiers?

So what causes op amps to have finite speed in the first place? This happens because real life op amps are limited by the finite impedance at a node. The impedance at a node is determined by the resistance and capacitance at the node. As the frequency increases, the capacitance behaves more like a "short", resulting in lower impedance and therefore lower gain. Eventually, the signal starts to "lose". It is this point that limits how fast an op amp can operate. The following figure shows how an op amp responds to a step signal change, where both slew rate and bandwidth have an impact on the overall settling time of the signal.

Now, let's dig a little deeper and try to understand slew rate and bandwidth conceptually.

(1) Bandwidth

We design an op amp under DC bias. So we are basically consuming quiescent power to make it "ready" to accept small signals or signals of small amplitude. When broken down using the Fourier transform, these frequencies can give you a very different sum of frequencies from small to large. This is the domain of "small signals", hence the bandwidth. The higher the bandwidth, the higher frequency the op amp is able to amplify, and therefore has higher speed. Electrically speaking, the frequency at which the signal gain is 1/sqrt[2], or 0.707 of the ideal value, is the bandwidth of the op amp. This is the maximum frequency at which the op amp can operate with the expected behavior. For example: a gain bandwidth product of 350kHz, i.e. a closed loop gain of 1, would give a bandwidth of 350kHz. A gain of 2 would give a bandwidth of 175kHz, and so on. The higher the closed loop gain, the slower the op amp will be - the product of gain and bandwidth being constant.

(2) Slew rate

Now, let's say the signal gets really big. For example, it goes to 1V instead of 1-2mV. Well, the op amp gets confused. As we know, op amps are designed to handle small signals and operate within their bandwidth, and now we are in the large signal region. In this case, the op amp will saturate, one of them will have full current and the other will have zero. This is also called tail current, and then it is used to "transmit" the 2V signal to the next stage. At this point, it is impossible to change the voltage immediately because it would require infinite current to charge the capacitance that is "inherent" to the system. If capacitors are used to compensate, they can be as high as 10pF or so. And we don't have infinite current either, which will cause slew rate. So, what is the reason for this large signal change? It is because when the power in the system is turned on or when the input from the previous stage is power cycled or switched. In these cases, we need to do large signal analysis.

Let's talk more about the formula for slew rate. When the op amp is in large signal mode, all biases of the op amp are fully saturated, which is why we need to go back to Coulomb's law, which states that q = CV

Or I = CdV/dt, so dV/dt = I/C, which is the textbook formula for slew rate.

Taking the Runshi high-speed operational amplifier as an example, the slew rate is 160mV/us, that is, the operational amplifier needs 1us to increase its output by 160mV.

2. Choose bandwidth or conversion rate?

If it continues to run, the 2V signal will be in a "slew limited" state until one side of the differential pair is depleted, then once the current starts to build up on the depleted differential side, it will enter the "bandwidth" region, so, settling time = slew time + BW response time.

Slew rate is the maximum rate at which the op amp can respond to large changes in the input signal, and bandwidth is the maximum rate at which it can respond to small changes in the signal. Together, the two determine the overall settling time of the step response. Some applications are more demanding on bandwidth, and their slew rate requirements are not so stringent, and this may be the case when the only real place where slew rate comes in handy is during startup. However, some applications, such as motor drives, require the op amp to be fully on or off, and the slew rate requirements are more stringent at this time. It all comes down to transferring electrical information from one phase to another. We are limited by the amount of current we need to do so, which creates the slew rate. In the large signal region, it is the slew rate; in the small signal region, it is the bandwidth. So, for a high-speed op amp, we may need to choose a high bandwidth and a high slew rate.