When both high DC accuracy and high bandwidth are required, it can be difficult to achieve. Engineers are often challenged to develop new applications to meet a wide range of needs. In general, these needs are difficult to meet at the same time. An example is a high-speed, high-voltage operational amplifier (op amp) with high output power and equally good dc accuracy, noise, and distortion performance. It is rare to find an op amp on the market that combines all of these characteristics. Depending on the circuit configuration, there are several effective approaches, including building a composite amplifier or implementing a servo loop around a high-speed amplifier.

Combining two op amps combines the best features of each into one. In this way, the combination of two op amps can achieve higher bandwidth than a single amplifier with the same gain.

The configuration of the composite amplifier is similar to that of the non-inverting amplifier, which has two external operating resistors, R1 and R2. Think of two op amps connected in series as one amplifier. The overall gain (G) is set by the resistor ratio, G = 1 + R1/R2. A change in the resistance ratio of R3 to R4 affects the gain of Amplifier 2 (G2), which also affects the gain or output level of Amplifier 1 (G1). However, R3 and R4 do not change the effective overall gain. If G2 decreases, G1 will increase.

Another characteristic of composite amplifiers is their higher bandwidth. Composite amplifiers have higher bandwidth than individual amplifiers. So, if you use two identical amplifiers with a gain-bandwidth product (GBWP) of 100 MHz and a gain of G = 1, you can improve the –3 dB bandwidth by about 27%. The higher the gain, the more pronounced the effect, but only up to a certain limit. Once the limit is exceeded, it may become unstable. This instability can also occur when the two gains are not evenly distributed. In general, the maximum bandwidth is obtained when the gain of the two amplifiers is equally distributed. Using the above values (GBWP = 100 MHz, G2 = 3.16, G = 10), at an overall gain of 10, the –3 dB bandwidth of the two amplifiers combined can be three times that of a single amplifier.

For an inverting circuit configuration, a DC servo loop using an operational amplifier configured as an integrator is most suitable. For a non-inverting circuit, an operational transconductance amplifier (OTA) based DC servo loop would be the easiest to implement. These two circuits are shown in Figure 1 and Figure 2 below. How to Build a High Speed Amplifier Circuit with High DC Accuracy and High Bandwidth

Figure 1: How a DC Servo Loop for an Inverting Amplifier Configuration Builds a High Speed Amplifier Circuit with High DC Accuracy and High Bandwidth

Figure 2: DC Servo Loop for Non-Inverting Amplifier Configuration

Both circuits are AC coupled whether or not you use decoupling capacitors. I represent the circuit here with decoupling capacitors to emphasize that the equivalent circuit will be AC coupled.

The servo loop actually removes the DC voltage and replaces it with a reference voltage (Vref). The accuracy of the system is limited only by the accuracy of the equipment used in the servo loop and the speed of the loop. In both circuits, you must balance the high-pass bandwidth with the response time of the servo amplifier. If the servo amplifier is too fast or the signal changes too slowly, the signal will be servoed, with disastrous consequences for its integrity. The system will also have an initial stabilization time before accurate measurements can be achieved.

For integrator-based circuits, the output voltage increase of the servo amplifier is directly related to the output of the signal amplifier. With a DC gain of 1-V/V, the signal amplifier’s input will then be seen at the output. The low-pass filter formed by R4 and C3 will limit the bandwidth and minimize noise impact on the signal amplifier. Servo amplifiers are usually precision amplifiers such as the OPA277 or OPA333.

How to Build a High Speed Amplifier Circuit with High DC Accuracy and High Bandwidth

A DC servo loop in a non-inverting configuration behaves the same for the integrator up to the SOTA (sampled OTA) output of the OPA615. The voltage difference between pins 10 and 11 will generate a current output that charges the Chold capacitor. The resulting voltage is then fed to another OTA. The voltage appearing at the B input (pin 3) of the OTA is mirrored as a voltage to the E input and converted to a current through resistor RE. The current is finally mirrored to the C output (pin 12) and plugged into the inverting node of the OPA656. Current will continue to be applied to this node until the voltage across pins 10 and 11 is zero.

Now to add some complexity, SOTA can be used to sample a specific time during which no signal reaches a certain DC value, effectively shifting the entire signal up or down. In this mode, the circuit behaves like a DC restoration circuit. If SOTA is always sampling, DC correction can only be achieved by inserting an RC filter on pin 10. This RC filter has the same effect as the R4, C3 filter in Figure 1.