This paper proposes a feedforward class AB CMOS operational amplifier with an input stage controlled by a minimum current selection technique to stabilize the transconductance and an output stage controlled by a floating current source.
1 input level design
1.1 Track-to-rail op amp input stage circuit analysis
Usually the op amp input stage uses a differential input mode. In CMOS processes, differential amplifiers can be implemented with PMOS or NMOS differential pairs. However, the usual differential pair cannot meet the requirements of the rail-to-rail common-mode input. Therefore, the method commonly used in practice is to use NMOS and PMOS complementary differential pairs. A simple rail-to-rail input stage structure is shown in Figure 1.
The circuit working range can be divided into three areas:
(1) When VCM is close to VSS, the NMOS differential pair is turned off, and the PMOS differential pair is in operation, gm=gmP;
(2) When VCM is close to VDD, the PMOS differential pair is turned off, and the NMOS differential pair is in operation, gm=gmN;
(3) When the VCM is at the intermediate value, both differential pairs work simultaneously, gm=gmP+gmN.
However, there is an important problem with this structure, that is, the total transconductance of the input circuit is not constant over the entire common mode input range, and the variation is twice as shown in FIG. Transconductance changes can cause signal distortion and have a large impact on the gain of the loop and the frequency compensation of the op amp. Therefore, the transconductance of the input stage is required to remain constant throughout the common mode input range.
There are currently four methods for constant transconductance:
(1) Using a 3x current mirror bias loop to keep the sum of the square roots of the tail current constant to obtain a constant transconductance; the disadvantage of this method is that it is overly dependent on the ideal square law model, working in the MOS tube in the strong inversion layer and weak The inversion layer is not universal.
(2) Using a Zener diode to make the sum of the gate and source voltages of the P and N differential input pairs constant; the drawback of this technique is that the MOS performance of the diode connection is a function of the voltage across it, so the gm common mode input range remains There are some changes.
(3) Use level shifting to shift the PMOS transconductance curve to the left or the NMOS transconductance curve to the right; the biggest drawback of this method is the need to adjust because its performance changes with process, temperature changes, and optimal DC level changes. .
(4) The maximum/minimum current selection method selects only a pair of differential pairs with a large current as the output when the circuit operates. Although the circuit design is more complicated, its output current is continuous, independent of the square law model, and the transconductance stability is good. The MOS tube can work in all areas. In this paper, the input stage of the op amp is designed by the method of minimum current selection.
Wuxi Juxingyao Trading Co., Ltd , https://www.juxingyao.com