Research On Key Technology Of Terminal RF Chips And Negative Group Delay Circuits

With the development of communication technology,the requirement of terminal RF circuit integration degree in communication equipment is becoming stricter.The traditional hybrid integration circuit is difficult to meet the needs of miniaturization in modern systems.Under this background,the monolithic integrated circuit has already become a hot spot in recent years because of its high circuit density,stable performance,large-scale production applicability and other advantages.Achieving comprehensive independent intellectual property rights of the terminal RF chip is very important,which not only helps to accelerate the development of Chinese communication field but also enhances the innovation capability and international competitiveness of integrated circuit.In addition,in the field of communication,it has already been known that for the RF signal with bandwidth,the dispersion effect will be produced due to the positive group delay.The negative group delay(NGD)circuits can eliminate this effect and help to optimize the signal integrity,reduce the bit error rate and improve the communication quality.In the academic field,an NGD circuit can be considered to have a characteristic that the group velocity transcends the speed of light,which greatly deepens the understanding of the group velocity.Accordingly,the research of NGD circuit is very important in both theoretical and practical fields.This paper studies the new design methods of terminal RF chip and NGD circuit deeply and systematically.According to the specific needs of circuits,the specific manufacturing processes are used to verify the relevant technology.The main contents are summarized as follows:1.In large signal mode,the non-linear effect in active mixerts will cause serious third inter-modulation products.To solve this problem,a Gilbert active mixer structure with linear transconductance is proposed according to the theory of body-effect and circuit model.In the process of intermediate frequency signal conversion,the source degradation resistance is adopted to convert the source voltage to the current form with high linearity for up-conversion modulation.Based on the analysis of the nonlinear characteristics of the circuit,a transconductance model is established.It is proposed that by using deep well structure,the body effect can be eliminated.The linear transconductance is independent of a fixed resistance and transistor size ratio rather than the bias voltage.This circuit improves the non-linear effect of the transistor when the IF signal is large,ensuring that the transmitter has better linearity in the large signal mode.The proposed circuit is validated by software simulation and integrated into a radio frequency transmitter chip that supports satellite communications systems.The transmitter includes the entire transmit chain from a digital-to-analog converter(DAC)to an RF drive amplifier.The overall test results of the chip show that the RF transmitter can achieve better than-37 dBc adjacent channel rejection ratio(ACPR)and better than 49 dB of local oscillator suppression at 0 dBm output power.Compared with the same type of transmitter,the gain dynamic range is improved by at least 10 dB.2.Aiming at the problem that the traditional double-balanced mixer is too large and the bandwidth of the intermediate frequency is narrow,according to the equivalent circuit model,a balun with tail capacitor is applied in the passive mixer,which can greatly reduce the circuit size and expand the intermediate frequency bandwidth.By building the lumped parameter model,the equivalent circuit model of the balun is deduced while the frequency response function of the model is used to derive the theory that the tail capacitance can reduce the working frequency and reduce the coil length of the balun.Based on the common mode transmission characteristics of the IF signal in the balun,it is proved that the mixer using proposed structure can show a wider bandwidth,which is verified by software simulation and measurement.By using GaAs process,two mixer chips working in 2.5-8 GHz and 6-18 GHz with ultra-wide IF bandwidth are fabricated,the test results show that the conversion losses are less than 10 dB,IF bandwidth reached 3 GHz and 8 GHz respectively.Compared with the traditional traditional mixers,the proposed circuit reduces the chip size to 1/4 and the IF bandwidth is extended to 300%.3.Aiming at the problem that the general NGD circuit is not adjustable and the stability is not good,the theory of stable NGD is proposed by using the step impedance conversion line in the distributed amplifier.Based on the modeling of the transfer function between the input and reverse output ports,the relationship between the NGD time and the voltage gain ratio of the circuit is quantitatively deduced.By calculating the input and output impedance of each stage transistor in the amplifier,the conclusion that the specific voltage gain ratio can be realized by adjusting the characteristic impedances of the transmission lines is obtained.The theory is verified by software simulation and measurement.Finally,through the circuit processing test,an NGD circuit with-1 ns group delay time and adjustable attenuation characteristic is achieved.The time fluctuation is less than 0.2 ns.Compared with other NGD circuits,it can be seen that the circuit has excellent characteristics of group delay time stability and adjustable gain.4.Aiming at the problem that the existing Non-Foster circuits are achieved by lumping method,it is proposed that the NGD circuit based on the step impedance distributed amplifier can realize the equivalent non-Foster series negative capacitance and dual-port matching Non-Foster T-network.Through the analysis of the S-parameter of the ideal series negative capacitance network and the physical link between two circuits,the synthesis method is summarized and the conclusion that the equivalent negative capacitance and matching network can be realized by using NGD circuit is proposed.The theory is verified by software simulation and measurement.By processing and testing,a-2 pF equivalent series negative capacitance and the T-network consists of a-1 pF parallel capacitor and two-1.25 nH series inductors are achieved based on the proposed NGD circuit.The negative capacitance can exhibit a maximum quality factor of 110 in the frequency band of 1.4-1.55 GHz.Compared with other negative capacitance works,this method is based on distributed circuits and is therefore more suitable for high frequency applications.The equivalent component value error of the network is less than 10% in the operating frequency range of 1.35-1.45 GHz and the return losses of the input and output ports of the circuit are both better than 20 dB.