This paper introduces a method of designing low noise amplifier using ads simulator. Firstly, the main technical and performance indexes of the low noise amplifier are described in general. Then, on the basis of using the 2sc5507 (ne661m04) tube of NEC, the circuit design, optimization and ADS simulation are carried out synchronously according to the indexes of the low noise amplifier. Finally, the design results of the low noise amplifier reach the expected value at the beginning of the design, The circuit design of low noise amplifier is successfully completed.

Performance index and design steps of low noise amplifier

Performance index of low noise amplifier

Frequency range: 2.0 ~ 2.25 GHz; Signal source impedance: 50 Ω; Gain: 10 dB; Noise figure

Design steps

Amplifier stage (select one stage); Transistor selection; Circuit topology; Preliminary circuit design; Advance design system 2005a (ads) software is used for design, optimization and simulation.

Main technical specifications of low noise amplifier

Noise figure and noise temperature of LNA

The noise figure NF of the amplifier can be defined as Formula 1:

Where NF is the noise figure of the microwave component; Sin and N in are the signal power and noise power of the input respectively; Sound and nout are the signal power and noise power of the output respectively. Generally, the noise figure is expressed in decibels. At this time, the noise generated by the amplifier itself is often expressed by the equivalent noise temperature te.

The relationship between noise temperature te and noise coefficient NF is shown in formula (2):

Where t0 is the ambient temperature, usually 293 K.

Power gain, correlation gain and gain flatness of LNA

Power gain usually refers to the measured gain when the source and load are 50 Ω standard impedance. In the actual measurement, the insertion method is commonly used, that is, use the power meter to measure the power P1 given by the source first; Then connect the amplifier to the source and measure the output power P2 of the amplifier with the same power meter. The power gain (g) is equal to P2 divided by P1. The gain in the case of noise best matching is called correlation gain. The best matching point of noise is not the maximum gain point. Generally, the correlation gain is about 2 ~ 4 dB lower than the maximum gain. The power gain will also affect the noise figure of the whole machine. The simplified expression of noise figure of multistage amplifier is given in Formula 3.

Where, NF is the noise coefficient of the whole amplifier; NF1, NF2 and NF3 are the noise coefficients of level 1, 2 and 3 respectively; G1 and G2 are the first and second stage power gains respectively.

When the gains G1 and G2 are large enough, the noise figure of the whole machine is close to the noise figure of the first stage. Therefore, the noise figure of the first stage of multistage amplifier plays a decisive role. Gain flatness refers to the fluctuation of power gain in the working frequency band. It is usually the difference between the maximum gain and the minimum gain, i.e Δ G indicates.

Operating frequency band

The working frequency band not only refers to the frequency band range in which the power gain meets the flatness requirements, but also makes the noise meet the requirements in the whole frequency band, and gives the noise coefficient of each frequency point.

dynamic range

The dynamic range refers to the range of minimum power and maximum power allowed by the input signal of the low noise amplifier. The upper limit of dynamic range is limited by nonlinear index. The lower limit of the dynamic range depends on the noise performance. When the noise figure NF of the amplifier is given, the allowable minimum value of input signal power is calculated as formula (4):

Where, Δ FM is the passband of the microwave system; M is the allowable signal-to-noise ratio or signal identification coefficient of the microwave system; T0 is the ambient temperature 293 K.

Port standing wave ratio

The input matching circuit of LNA is designed according to the best noise, and the result will deviate from the conjugate matching state with the best standing wave ratio. In addition, the gain characteristics of microwave field effect transistors or bipolar transistors generally decrease with the increase of frequency according to the law of 6dB per octave. In order to obtain the flat gain characteristic in the working band, when the input matching circuit and output matching circuit are non power consumption resistant circuits, the low-frequency mismatch method can only be used to reduce the gain to keep the flat gain in the band. Therefore, the port standing wave ratio must increase with the decrease of frequency.

stability

When the mode of the reflection coefficient at the input and output of the amplifier is less than 1 (i.e| Г 1 | 2 |

Amplifier in Г The necessary and sufficient condition for absolute stability on the s input plane is defined as formula (8):

Amplifier in Г The necessary and sufficient condition for absolute stability on the L input plane is defined as formula (9):

Transistor selection

The 2sc5507 (ne661m04) of NEC is adopted, which has the advantages of high frequency, low noise and good low temperature performance. The datasheet of 2sc5507 provides S-parameters of wide frequency band, and the S-parameter model is also selected in ads setting, so the S-parameter model is more accurate. After selecting the device, the S parameter can be used to judge its stability. From the S parameter at VDS = 2 V, id = 5 mA, f = 2.0 GHz, it can be concluded that DC is not absolutely stable and stability design is required.

ADS simulation

Basic principles

In order to make the transistor work in the amplification region, it is necessary to determine the static DC operating point. VDS = 2 V, id = 5 Ma can be obtained from the datasheet of 2sc5507. The placement of various components of the basic simulation schematic diagram is shown in Figure 1.

K1 ： K = stab_ fact （ S） ， stab_ The fact (s) function returns the Rolette stability factor. K》 1, the circuit is absolutely stable, and the stability B1 is 1.

M1: Mu = mu (s), the mu (s) function returns the geometric derivation factor of the load. Mu 1, the circuit is absolutely stable. B1 ： B1 =stab_ meas （ S） ， stab _ The meas (s) function returns a stable quantity. It can be seen from the simulation results that Mu at frequencies above 2.0 GHz is not greater than 1. Before the output stability circuit is added, the transistor output is unstable. Therefore, it is necessary to design its stability and add the output stability circuit.

Increased output stability circuit

On the basis of the original drawing, an RC series circuit can be connected in parallel, or LC or RL circuit can be added. Specifically, it can be designed together with the output matching circuit. Here R1 = 50 Ω, C1 = 2.0pf. The circuit diagram is shown in Figure 2.

After adding the output stability circuit, the simulation result is that when the transistor frequency is between 2 GHz and 3 GHz, Mu “1, it can be seen that the system is absolutely stable.

Optimal noise matching

For LNA, if the input port has a certain mismatch, the phase relationship between various noises in the device can be adjusted to reduce the noise figure. In order to obtain the minimum noise figure, Г S has an optimal value Г Opt, at this time, LNA reaches the minimum noise coefficient, that is, it reaches the best noise matching state. among Г Opt is the best source reflection coefficient (equivalent noise parameter of microwave transistor). When the matching state deviates from the best, the noise figure of LNA will increase. Г Opt can be obtained from the datasheet file of the device. Sopt is the optimal matching coefficient of minimum noise. Using this optimal coefficient, the input matching circuit can be designed. The noise figure simulation circuit is shown in Figure 3.

Sopt = 0.32 / 29.4 (amplitude and angle) from the matching result.

The input matching circuit is designed according to the minimum noise

The design of input matching circuit is shown in Figure 4.

The input reflection coefficient s [1, 1] is set as the conjugate of sopt for 50 Ω matching. The best input matching coefficient is C1 = 1.73 PF and L1 = 5.79 NH. The matching results are shown in Figure 5. From Figure 5, we can get: s (1, 1) = 0.097e – 4 / 4.374.

So far, the input impedance matching designed according to the principle of minimum noise figure has been completed.

The output matching circuit is designed according to the maximum power gain

According to the principle of maximum power gain, the output matching circuit is designed, that is, the output is matched by 50 Ω. Considering the influence of output stability circuit on output impedance, the form of output matching circuit is a little different. The output matching circuit is designed as the input impedance matching method. The circuit diagram is shown in Figure 6.

Optimized L2 = 3.651 NH, L4 = 4.028 NH. The matching result is shown in Figure 7. From the figure: s (1, 1) = 2. 412e – 4 / – 38. 789.

So far, the input-output matching circuit is completed. The simulation results are shown in Figure 8 and Figure 9. Thus, the gain is 16.917 db; Noise figure: 1. 649 dB.

conclusion

The final simulation results show that the gain is 16.917 db; Noise coefficient: 1. 649. The results meet the performance indexes at the beginning of the design, but the working frequency band is small, which is normal. This is because it is difficult to obtain very low noise in the case of wide frequency band, so the working frequency band of low noise amplifier is generally not wide, mostly about 2 GHz.

Responsible editor: GT