Noise shaping in low noise amplifiers using active feedback and pole-zero adjustment

Abstract

In conventional methods, noise calculation of low noise amplifier circuits are performed by eliminating frequency devices or frequency parameters such as inductances and capacitances and only considering the quiescent point. Based on these noise calculation methods, noise figure is evaluated at low frequencies, and this diminishes the accuracy of the calculations. Therefore, in this paper, after extracting full equations of transfer functions of gain and noise, a system of equations is set up. By adjusting different circuit parameters, the frequency responses are adjusted so that in a desired frequency range, the gain assumes its maximum and noise its minimum. In fact, in this way, somehow a noise/gain shaping is realized. Moreover, other circuit characteristics such as input/output matching and power consumption are introduced to the system of equations, and their optimized values are obtained. The proposed method can be applied to different structures of low noise amplifiers; furthermore, it enables the designer to perform noise/gain shaping freely and obtain the most precise trustworthy results by exact investigation of circuit equations and proper placement of the circuit elements. As an example, a wideband low noise amplifier with a complementary structure is designed based on the proposed method, and its post-layout results are obtained as follows: the frequency bandwidth ranges from 0.025 to 2.5 GHz, input/output matching is better than −10 dB, small-signal gain is 13.5 dB, minimum value of noise figure is 1.8 dB, the intercept point of the first and third harmonics is −4.5 dBm, power consumption is 11.17 mW at a bias voltage of 1.2 V, and the core area of the chip is 0.023 mm2.