CN1564457A - Optimized design method of microwave amplifying circuit - Google Patents

Abstract

Improving efficiency of designing circuits and reducing complexity for match between input and output networks is realized through Pi type or T type network or degraded Pi type or T type network. Transition matrix is utilized to describe and calculate optimal input and output networks of amplification circuit. Active circuit unit containing transistor in negative feedback is described and calculated through scattering matrix, which then transformed to transition matrix and scattering matrix for integral amplification circuit. Standing wave ratio of input and output, gain, noise temperature and stability coefficient are calculated and optimized through scattering matrix for integral amplification circuit. Advantages of precision, high efficiency and convenience of the method is proved by execution of computer.

Description

A kind of Optimization Design of microwave amplifiercation circuit

Technical field

The Optimization Design of microwave amplifiercation circuit belongs to the design and the optimisation technique field of microwave amplifiercation circuit.

Background of invention

The development of simulation standard to digital standard passed through in the development of present mobile communication, and digital mobile communication system has also been gone through frequency division multiplexing (FDMA), and time division multiplexing (TDMA) is to the evolution of code division multiple access (CDMA) radio communication.The CDMA mode can exchange signal to noise ratio for more effectively to realize bandwidth.In other words, the signal to noise ratio that improves receiving system is equivalent to the increase number of users, so the amplifier LNA and the front end system of research low-noise factor are one of emphasis of microwave current telecommunication circuit area research.Communication must be satisfied special requirement with LNA, for example low standing-wave ratio index, and for example CDMA2000 receiver standard code, the amplifier input vswr is lower than 1.5, and the output standing-wave ratio must be lower than 2.

Below introduce the definition and the explanation thereof of amplifying circuit common counter respectively:

1 voltage standing wave ratio (VSWR)

(for example [Guillermo Gonzalez is outstanding, and Bai Xiaodong translates, microwave transistor amplifier analysis and design for the Amplifier Design document for present document, publishing house of Tsing-Hua University, Beijing, 2003], [Tri.T breathes out work, the design of solid microwave amplifier, National Defense Industry Press, Beijing, 1988]) or be not described in and how design in the Amplifier Design and optimize the standing-wave ratio index, perhaps only provide a kind of special case, and can not be used to optimize the voltage standing wave ratio of circuit.Standing-wave ratio is the amount of a kind of easy measurement in the electromagnetism engineering, it has illustrated the transmission and the reflected energy distribution state of microwave electromagnetic field energy, for example standing-wave ratio 1.5 mean input energy 4% be reflected back toward reference source, 8.2% of standing-wave ratio 1.8 meaning intakes are reflected back toward the reference source end; Desirable standing-wave ratio is 1, promptly means load and information source conjugate impedance match, and power transfer reaches 100%.

Standing-wave ratio 1.0 1.2 1.5 1.6 1.7 1.8 2.0 3.0

Energy 0.00% 0.83% 4.00% 5.33% 6.72% 8.16% 11.11% 25.00% is reflected

The standing-wave ratio loss also determines the error rate of digital circuit, and side circuit can not reach the perfect match state that standing-wave ratio equals 1, needs the designer to optimize this index to minimum.

2 gains

Gain is meant the signal power P that obtains in the load of amplifying circuit _oPower P with signal source output _iRatio, this value could claim that greater than 1.0 this circuit is a signal amplification circuit, is commonly referred to amplifying circuit (or amplifier):

Ga＝(P _o)/（P _i)＝|S ₂₁| ²???????????????????????????????????(1)

Gain is represented with decibel value

Gain＝10*Log(Ga)＝20*Log|S ₂₁|(dB)????????????????????(2)

Under other parameter constant situation of circuit, require yield value should try one's best greatly.

3 noise factors and equivalent noise temperature

Noise factor is a kind of tolerance that circuit is introduced the additive noise power size of describing.It is the decibel numerical value of noise factor parameter.Noise factor is defined as input signal power S _iWith input noise power N _iRatio removes output signal power S _oWith output noise power N _o, promptly

$\mathrm{Nfactor}=\frac{\mathrm{Si}/\mathrm{Ni}}{\mathrm{So}/\mathrm{No}}---\left(3\right)$

Result with dB value (noise factor logarithm 10 times) expression just is called noise factor, is designated as NF.

NF＝10*Log(Nfactor)(dB)????????????????????????????????(4)

Desirable noise-free signal transmission network is not owing to there is an additional internal noise, the input and output signal to noise ratio identical, i.e. Nfactor=1, being write as the noise factor form is NF=0dB.Under all the other parameter constant situations of circuit.The noise factor of amplifier should be as far as possible little.

Also with equivalent noise temperature statement noiseproof feature, noise temperature and noise factor close and are in circuit:

NT＝(10 ^NF/10-1)*290(K)?????????????????????????????????(5)

Noise temperature is fit to describe low-noise factor, and its unit is absolute temperature scale K.

4 stability coefficient K

Stability is that load and source change the characteristic of following amplifier reflected energy less than intake.Be not more than intake steady operation condition by the output energy, derive a kind of COEFFICIENT K of describing stability:

$K=\frac{1-{\left|S_{11}\right|}^2-{\left|S_{22}\right|}^2+{\left|\mathrm{\&Delta;}\right|}^2}{2|{\mathrm{<figure-callout\; id="12"\; label="S"\; filenames="A20041000346700192.png"\; state="\{\{state\}\}">S</figure-callout>}}_{12}\&CenterDot;S_{21}|}---\left(6\right)$

Δ=S ₁₁* S ₂₂-S ₁₂* S ₂₁S is transistorized collision matrix, S _IjBe matrix element (i, j=1,2, the capable j column element of i).The necessary and sufficient condition that amplifying circuit is stable:

Some transistor K＞1 is the unconditional stability transistor, needn't consider the K value during design, and its design can be with reference to the present invention.K＜1 is called the latent instability transistor, need utilize matching network to make integrated circuit satisfy formula (7).

The matrix description of microwave network

If circuit incoming wave vector is A, the output wave vector is B. $A=\left[\begin{array}{c}\mathrm{ao}\\ \mathrm{ai}\end{array}\right],B=\left[\begin{array}{c}\mathrm{bo}\\ \mathrm{bi}\end{array}\right].,$ a _i, a _oBe respectively input, output incident wave, b _i, b _oBe respectively the reflected wave of input, output, then the collision matrix S of microwave network is defined as:

B＝S·A????????????????????????????????(8)

In microwave regime, port character utilizes scattering parameter to be described because scattering parameter can explicit definition the energy and the reflected energy of transmission, can obtain by actual measurement;

The method of describing the equivalence of microwave network in addition has impedance matrix z and transfer matrix $\left[\begin{array}{cc}a& b\\ c& d\end{array}\right],$

V＝Z·I，z＝Z/Z0????????????????????????(9)

Wherein Z0=50 Ω is a normalized impedance.

$\left[\begin{array}{c}v1\\ i1\end{array}\right]=\left[\begin{array}{cc}a& b\\ c& d\end{array}\right]\left[\begin{array}{c}v2\\ {i2}^{\&prime;}\end{array}\right]---\left(10\right)$

V1, v2 are respectively the input and output terminal voltage, and i1 is an input current, and i2 ' is reverse output current value.

The statement of these matrixes is not independently, all can derive all the other expression forms from any one.For example concern just like down conversion between impedance matrix z and the collision matrix S:

[s]＝([z]-1)([z]+1) ^-1；????????????????(11)

[z]＝(1+[s])(1-[s]) ^-1；????????????????(12)

Collision matrix [s] and transfer matrix $\left[\begin{array}{cc}a& b\\ c& d\end{array}\right]$ Between concern just like down conversion:

$\left[s\right]=\frac{1}{a+b+c+d}\left[\begin{array}{cc}a+b-c-d& 2a\\ 2& -a+b-c+d\end{array}\right];---\left(13\right)$

$\left[\begin{array}{cc}a& b\\ c& d\end{array}\right]=\frac{1}{2s_{21}}\left[\begin{array}{cc}1+s_{11}-s_{22}-\left|s\right|& 1+s_{11}+s_{22}+\left|s\right|\\ 1-s_{11}-s_{22}+\left|s\right|& 1-s_{11}+s_{22}-\left|s\right|\end{array}\right];---\left(14\right)$

Impedance matrix z and transfer matrix $\left[\begin{array}{cc}a& b\\ c& d\end{array}\right]$ Between transformation relation:

$\left[\begin{array}{cc}a& b\\ c& d\end{array}\right]=\frac{1}{z_{21}}\left[\begin{array}{cc}{z}_{11}& \left|z\right|\\ 1& z_{22}\end{array}\right];---\left(15\right)$

This paper utilizes these matrix formulations and conversion as analysis tool.

Conventional design and optimization low noise amplifier rely on a kind of method that is called the Smith circle diagram, step is at first to make 2 figure, represent input reflection coefficient circle diagram and output reflection coefficient circle diagram respectively, on input reflection coefficient circle, make wait noise factor justify and etc. gain circle, select both suitable points in occuring simultaneously by hand then, or rely on some other standard and determine wherein certain reflection coefficient point, standing-wave ratio then only relies on than higher gain and guarantees, because matched well meaning transmission of power is bigger, it is bigger to gain.Owing to used artificial judgement, can not guarantee I/O reflection coefficient result's global optimum in this method.

Routine is utilized the standing-wave ratio that can not calculate Fig. 1 in the method for designing of accompanying drawing 2 equivalent electric circuits.Standing-wave ratio does not obtain the reason place of discussing during actual design that Here it is, and to bidirectional transistor model (most situation), coupling relates to the plural equation of separating two couplings fully, and can't judge existence of solution in advance; For the latent instability transistor, standing-wave ratio total conjugated coupling means the maximum transmitted gain, and the transistorized maximum gain of latent instability is insignificant.So never appropriate methods is come the constraint of precise design standing-wave ratio, what rely on only is experience and trial and error method.

So Amplifier Design generally adopts the eda software optimisation technique at present, the eda software optimizational function all utilizes the target function of giving the target function weight to carry out vector search.But there is following difficulty in this design: 1 automatic majorization function is limited, easily is absorbed in a Local Extremum and termination automatically; 2 can not guarantee global optimum; 3 performance eda software optimizational functions need the designer to have abundant Amplifier Design experience, and different active devices is needed different design experiences; 4 result of calculations are big to circuit variable number dependence, and the independent variable number is less.

Summary of the invention

The objective of the invention is to seek a kind of microwave amplifier that guarantees and to design the method for the standing-wave ratio of realization regulation, gain, noise factor index, also must satisfy stable prerequisite simultaneously; And this method should be able to realize the design of global optimum.

By amplifier being divided into fan-in network, transistor network, 3 parts of output network, measure typical scattering parameter S on the transistor working frequency range to be designed; And be converted into transfer matrix Tt, then determine the transfer matrix Ti of fan-in network, and output network To, and determine amplifying circuit transfer matrix Tc=f (Ti, Tt, To), thereby obtain the every desired value of amplifying circuit, as standing-wave ratio, gain, noise factor expression formula.The transfer matrix of determining the input and output network can be with any one realization of following 2 kinds of methods: 1 sets the S parameter of input and output matching network, obtain Ti and To[Guillermo Gonzalez work by matrix conversion, Bai Xiaodong translates, microwave transistor amplifier is analyzed and design, publishing house of Tsing-Hua University, Beijing, 2003]; 2 determine input impedance matrix zi and output impedance matrix zo by the circuit topology and the component values of input and output network, are Ti and To by matrix conversion then; Select circuit topology hereinafter to be described in detail.

The inventive method is characterized in that: it be a kind of by an input matching network, comprise degenerative transistor and output matching network microwave amplifiercation circuit that constitutes of cascade successively, when being realized by calculation procedure, it comprises following steps successively:

Initialization:

Set the prototype circuit of input matching network and output matching network, this prototype circuit is by having internal structure or not having ∏ type that the electronic circuit piece of internal structure forms or 2 port electric networks of T type.Perhaps import the involution form circuit of this prototype circuit, they are ∏ type or the T type circuit that reduce by 1 or 2 parts, also are 2 port electric networks, but have only 2 kinds of different structures altogether; These input and output matching networks can corresponding a kind of the equivalent electric circuit form with input information source impedance and output loading, i.e. the describing method that adopts during conventional design, and above-mentioned input and output network is described and is deposited in the computer program with impedance matrix;

Degenerative transistor is described and is deposited in the computer program with collision matrix;

Target setting function and desired value thereof and their calculation procedure deposit computer in; Described target function is standing wave ratio of input voltage VSWR _In, output voltage standing-wave ratio VSWR _Out, the some projects among gain G ain, equivalent noise temperature NT, the stability coefficient K or their combination, the computing formula of each target function is as follows:

${\mathrm{VSWR}}_{\mathrm{in}}=\frac{1+\left|S_{c11}\right|}{1-\left|S_{c11}\right|},---\left(16\right)$

${\mathrm{VSWR}}_{\mathrm{out}}=\frac{1+\left|S_{c22}\right|}{1-\left|S_{c22}\right|},---\left(17\right)$

Gain＝10×Log(|S _c21| ²)，????????????????????????????(18)

$\mathrm{NT}={\mathrm{NT}}_{\min }+\frac{{4R}_n{T}_0}{Z_0}\frac{{|S_{i22}-{\mathrm{\&Gamma;}}_{\mathrm{opt}}|}^2}{(1-{\left|S_{i22}\right|}^2)(1-{\left|{\mathrm{\&Gamma;}}_{\mathrm{opt}}\right|}^2)},.---\left(19\right)$

$K=\frac{1-{\left|S_{c11}\right|}^2-{\left|S_{c22}\right|}^2+{\left|\mathrm{\&Delta;}\right|}^2}{2\left|S_{c12}{S}_{c21}\right|},\mathrm{\&Delta;}=S_{c11}*S_{c22}-{\mathrm{<figure-callout\; id="12"\; label="S\; c"\; filenames="A20041000346700192.png"\; state="\{\{state\}\}">S</figure-callout>}}_c*S_{c21,}---\left(20\right)$

Noise resistance R wherein _n, minimal noise temperature NT _Min, best information source reflection coefficient Γ _OptAll provide by transistor data bool; Described Sc matrix is the collision matrix of whole objective circuit, $\mathrm{Sc}=\left[\begin{array}{cc}{S}_{c11}& S_{c12}\\ S_{c21}& S_{c22}\end{array}\right],$ It is by the transition matrix Tc conversion of whole objective circuit, and transition matrix Tc=Ti * Tt * To, wherein, Ti is the normalized impedance matrix of input matching network $z_i=\left[\begin{array}{cc}{Z}_{i11}& Z_{i12}\\ Z_{i21}& Z_{i22}\end{array}\right]$ Transition matrix; Tt is that transistor active circuit collision matrix St transforms the transfer matrix that obtains; To is the normalized impedance matrix of output matching network $z_0=\left[\begin{array}{cc}{Z}_{o11}& Z_{o12}\\ Z_{o21}& Z_{o22}\end{array}\right]$ Transfer matrix;

Set information source impedance and load:

The value Z0 of information source impedance and load is 50 Ω, and the partiting dc capacitor value of the signal flow inbound port of input matching network and the signal output port of output matching network is calculated according to the capacitive reactance on the radio frequency operation frequency, is taken as 50pF;

Determine the progression of amplifier, circuit at different levels design respectively;

Determine the form of single-level circuit: input matching network have high pass characteristic and circuit must have a bypass with the requirement of introducing direct current biasing under, determine the circuit topology of input, output matching network and as the impedance component type and the variate-value thereof of variable, described impedance component is inductance or electric capacity or resistance or microstrip line or their polyphone or parallel circuits;

Select transistor and negative feedback value: calculate transistorized stability coefficient K value according to the transistor collision matrix earlier, if K＜1, introduce negative feedback network, this negative feedback network by electric capacity, resistance, microstrip line element or their combination through the polyphone or be formed in parallel, but the quantity of element is less than or equal to 3, setting member variate-value, calculating K successively again, only o'clock just can determine the component values of negative feedback network, obtain comprising negative feedback network at interior collision matrix to K＞0.6 $\mathrm{St}=\left[\begin{array}{cc}{S}_{t11}& S_{t12}\\ S_{t21}& S_{t22}\end{array}\right];$

Calculate following each amount according to above-mentioned specific input, output matching circuit:

Calculate normalization input impedance matrix according to input matching network $z_i=\left[\begin{array}{cc}{Z}_{i11}& Z_{i12}\\ Z_{i21}& Z_{i22}\end{array}\right];$

Calculate normalization output impedance matrix according to output matching network $z_o=\left[\begin{array}{cc}{Z}_{o11}& Z_{o12}\\ Z_{o21}& Z_{o22}\end{array}\right];$

Above-mentioned Z ₁Be transformed to transfer matrix $\mathrm{Ti}=\left[\begin{array}{cc}{A}_i& B_i\\ C_i& D_i\end{array}\right]=\frac{1}{z_{i21}}\left[\begin{array}{cc}{z}_{i11}& \left|z_i\right|\\ 1& z_{i22}\end{array}\right];$

Above-mentioned St is transformed to transfer matrix:

$\mathrm{Tt}=\left[\begin{array}{cc}{A}_t& B_t\\ C_t& D_t\end{array}\right]=\frac{1}{{2s}_{21}}\left[\begin{array}{cc}1-s_{i11}-s_{t22}+\left|s_t\right|& 1+s_{t11}+s_{t22}+\left|s_t\right|\\ 1-s_{t11}-s_{t22}+\left|s_t\right|& 1-s_{t11}+s_{t22}-\left|s_t\right|\end{array}\right];$

Above-mentioned z _oBe transformed to transfer matrix $\mathrm{To}=\left[\begin{array}{cc}{A}_o& B_o\\ C_o& D_o\end{array}\right]=\frac{1}{z_{o21}}\left[\begin{array}{cc}{z}_{o11}& \left|z_o\right|\\ 1& z_{o22}\end{array}\right];$

Calculate microwave amplifiercation circuit global transfer matrix $\mathrm{Tc}=\left[\begin{array}{cc}{A}_c& B_c\\ C_c& D_c\end{array}\right]=\mathrm{Ti}\&times;\mathrm{Tt}\&times;\mathrm{To};$

Is the global transfer matrixing collision matrix

$\mathrm{Sc}=\left[\begin{array}{cc}{S}_{c11}& S_{c12}\\ S_{c21}& S_{c22}\end{array}\right]=\frac{1}{A_c+B_c+C_c+D_c}\left[\begin{array}{cc}{A}_c+B_c-C_c-D_c& 2A_c\\ 2& -A_c+B_c-C_c+D_c\end{array}\right];$

Calculate standing wave ratio of input voltage VSWR according to formula (1)～(5) _In, output voltage standing-wave ratio VSWR _Out, gain G ain, equivalent noise temperature NT, stability coefficient K value;

Whether the target function value that determining step (8) obtains meets design requirement, if satisfy, then they parameter values as input, output matching network; Otherwise, the value of respective impedance element in modification input, the output matching network, repeating step (4)～(9) are only to VSWR _In, VSWR _Out, when Gain, NT, K satisfy the index of setting till.

Experimental data

Utilize this paper method to design, according to the selected component parameters of the result who optimizes, and on the complex media substrate actual fabrication CDMA amplifier circuit, input vswr VSWR _InDesign result is 1.11, and measurement result is 1.2; Output standing-wave ratio VSWR _OutDesign result is 1.42, and measurement result is 1.5; Gain Cain such as design result are 1.42, and stability coefficient K measurement result is 1.5; Noise factor NF design result is 0.71, and measurement result is 0.89; Measurement result is consistent with design load: design process of the present invention is consuming time less than 2 minutes.

Utilize the microwave circuit Amplifier Design software ADS of design software Agilent company commonly used to compare design, at the parameter variation range of each element all under the same terms, utilize its random optimization algorithm to carry out the following input vswr VSWR of design objective that obtains after 11 minutes the calculating _InBe 1.57; Output standing-wave ratio VSWR _OutBe 2.39; Gain G ain is 19.95; Stability coefficient K is 1.52; Noise factor NF is 1.10.

Description of drawings

Fig. 1; The present invention is about utilizing the electric theory diagram of cascade network design microwave amplifier;

Fig. 2; The electric theory diagram of the existing conventional design microwave amplifier that goes out from Fig. 1 equivalence;

Fig. 3; T type network calculations key diagram;

Fig. 4; ∏ type network calculations key diagram;

Fig. 5; T type (∏ type) the network 1 calculation specifications figure that degenerates;

Fig. 6; T type (∏ type) the network 2 calculation specifications figure that degenerate;

Fig. 7; Be used for amplifier embodiment fan-in network T type electric network figure;

Fig. 8; Degeneration T type (∏ type) the electric network figure that is used for amplifier embodiment output network;

Fig. 9-1; The Amplifier Design schematic diagram of the inventive method;

Fig. 9-2; The Amplifier Design flow chart of the inventive method.

Specific embodiment

The cdma mobile communication of one of circuit theory of the present invention and embodiment is described in detail as follows in conjunction with each accompanying drawing with amplifying circuit specific design method:

Fig. 1 has illustrated composition, structure, electric parameter and their witness mark (face) of the prototype circuit of this paper utilization.Amplifying circuit integral body is by input matching network (16), and the transistor active circuit is (17) partly, and output matching network (18) cascade combines.Need to prove 16,18 element topology is the form that conveniently shows as Fig. 1 as an illustration, concrete circuit standard and topological form have detailed description in Fig. 3 Fig. 4 and text of the present invention, one of emphasis of the present invention is the account form and the selection mode of these topological structures.

What need especially to pay close attention to is that the present invention brings forward among Fig. 1 11 and 12 the plane of reference, and this is the measurement of reflection-factor point of integrated circuit in design neutralization actual measurement.Only after having calculated the reflection coefficient characteristic that comprises 11 and 12 plane of references, just can obtain circuit reflection coefficient index accurately.

By 11 plane of references of Fig. 1 eye right into reflection coefficient be defined as Γ ci (19), 11 plane of references of Fig. 1 eye right into voltage standing wave ratio be defined as VSWR _In12 plane of references of Fig. 1 eye left into reflection coefficient be defined as Γ co (114), 12 plane of references of Fig. 1 eye left into voltage standing wave ratio be defined as VSWR _Out

110 plane of references of Fig. 1 and direction see into reflection coefficient be defined as Γ s, 111 plane of references of Fig. 1 and direction see into reflection coefficient be defined as Γ in, 112 plane of references of Fig. 1 and direction see into reflection coefficient be defined as Γ out, 113 plane of references of Fig. 1 and direction see into reflection coefficient be defined as Γ L.

Information source impedance Z 0 (14) and load Z0 (15) are 50 ohm at general provision if not otherwise specified, and this value can and not influence science of the present invention and implementation result according to the change of design needs, for example can change to 75 ohm.In order to narrate conveniently, this paper will adopt 50 ohm of standards in discussing.

Fig. 2 describes be prototype circuit of the present invention through the schematic block circuit diagram after the equivalence, i.e. the model that adopts in the base amplifier design document.Can see the compared to Figure 1 quantity of part and not change of annexation except the inner topology equivalence is input matched impedance (218) and output matched impedance (219), this is to derive from Fig. 1 because of Fig. 2.Whole design process is than the internal relation explanation of parameter and gain index.Implement only to have made things convenient for Γ s after the equivalence, Γ in, Γ out, the description and the design of Γ L variable, and only reduce to variable that 218 and 219 two equiva lent impedances are cost.

The present invention utilizes composition, structure, the topology of Fig. 1 to design for prototype, and the effect of Fig. 2 only is to prove that prototype of the present invention can equivalence be conventional prototype, with the corresponding relation of awareness network, no longer Fig. 2 is further calculated and narrates in addition.

Fig. 3 is a T type impedance network prototype figure.It is a kind of two ports topology form of input and output network.Special emphasis proposes is wherein impedance component 32, impedance component 33, and impedance component 34 is not only represented elements such as inductance, electric capacity, resistance and microstrip line, can also be by polyphone, the parallel circuits of elements such as electric capacity, inductance, resistance, microstrip line are formed arbitrarily.For example realizing 32,33,34 among Fig. 3 respectively by 3 50 ohm resistance, is a way of realization of this figure prototype.Fig. 7 is the used circuit of this paper embodiment, also is the specific implementation of Fig. 3.

T type impedance network prototype shown in Figure 3 is 2 port networks, and signal can be exported from the another one port from any one input the port of 31A, 31B or 36A, 36B composition.

Fig. 4 is a ∏ type impedance network prototype figure.It is a kind of two ports topology form of input and output network.Special emphasis proposes is wherein impedance component 42, impedance component 43, and impedance component 44 only plays the signal effect, and they can be made up of unit elements such as electric capacity, inductance, resistance, microstrip line arbitrarily or their polyphone, parallel circuits.For example realizing 42,43,44 among Fig. 4 by 3 50 ohm resistance, is a way of realization of this figure prototype.

The ∏ type impedance network prototype of Fig. 4 is 2 port networks, and signal can be exported from the another one port from any one input the port of 41A, 41B or 46A, 46B composition.

Fig. 5 is a kind of T type (perhaps ∏ type) impedance network prototype (50) figure of degeneration, the T type impedance network prototype that it is equivalent to Fig. 3 cancelled 33 or the ∏ type impedance network prototype of Fig. 4 cancelled 43 parts.The electronic circuit that circuit block 52 and 53 equally also is made up of electric capacity, inductance, resistance and microstrip line, the form among Fig. 5 of being depicted as only plays a kind of signal effect of equivalence.

Fig. 6 is a kind of T type (perhaps ∏ type) impedance network prototype (60) figure of degeneration, the T type impedance network prototype that it is equivalent to Fig. 3 cancelled 32 or the ∏ type impedance network prototype of Fig. 4 cancelled 44 parts.The electronic circuit that circuit block 62 and 63 equally also is made up of electric capacity, inductance, resistance and microstrip line, the form among Fig. 6 of being depicted as only plays a kind of signal effect of equivalence.

Fig. 7 is the realization circuit of input matching network among the 800MHz frequency range amplifier embodiment, and the circuit of this figure is a realization of Fig. 3 circuit prototype.Parts 32,33,34 are realized with element C1, L1, L2 among Fig. 7 respectively in the corresponding diagram 3.Syndeton belongs to aforesaid T type network prototype.This network determines by 3 variablees, i.e. the parameter value of C1, L1, L2.

Fig. 8 is the realization circuit of output matching network among the 800MHz frequency range amplifier embodiment, and the circuit of this figure is a realization of Fig. 6 circuit prototype.Parts 62,63 are realized with element C2, L3 among Fig. 8 respectively in the corresponding diagram 6.Syndeton belongs to aforesaid involution form network prototype 60.This network determines by 2 variablees, i.e. the parameter value of C2, L3.

Fig. 9-the 1st, single-level circuit design cycle schematic diagram of the present invention, Fig. 9-the 2nd, the program flow diagram of in computer, realizing.Text description is as follows:

Step 1: determine amplifier progression;

Step 2: according to selecting match circuit topology (promptly from Fig. 3-Fig. 6, selecting a kind of) and variate-value (C among corresponding diagram 7-Fig. 8 in example ₁, L ₁, L ₂, C ₃, L ₃Numerical value), and transistor and negative feedback value, the match network topologies that chooses need satisfy allows the interior signal of working band by stoping the out of band signal transmission, and the feedback quantity size had both suitably improved stability also must satisfy attainable requirement;

Step 3: the normalized impedance matrix that from specific match circuit topology, calculates input circuit $z_i=\left[\begin{array}{cc}{Z}_{i11}& Z_{i12}\\ Z_{i21}& Z_{i22}\end{array}\right]$ Normalized impedance matrix with output circuit $z_o=\left[\begin{array}{cc}{Z}_{o11}& Z_{o12}\\ Z_{o21}& Z_{o22}\end{array}\right];$ And introduced the transistor active circuit collision matrix after the negative feedback

$\mathrm{St}=\left[\begin{array}{cc}{S}_{t11}& S_{t12}\\ S_{t21}& S_{t22}\end{array}\right].$

Step 4: zi is changed into transfer matrix $\mathrm{Ti}=\left[\begin{array}{cc}{A}_i& B_i\\ C_i& D_i\end{array}\right]=\frac{1}{z_{i21}}\left[\begin{array}{cc}{z}_{i11}& \left|z_i\right|\\ 1& z_{i22}\end{array}\right],$ The collision matrix St of transistor active circuit is converted to $\mathrm{Tt}=\left[\begin{array}{cc}{A}_t& B_t\\ C_t& D_t\end{array}\right]=\frac{1}{{2s}_{21}}\left[\begin{array}{cc}1+s_{t11}-s_{t22}-\left|s_t\right|& 1+s_{t11}+s_{t22}+\left|s_t\right|\\ 1-s_{t11}-s_{t22}+\left|s_t\right|& 1-s_{t11}+s_{t22}-\left|s_t\right|\end{array}\right],$ Zo is changed into transfer matrix

$\mathrm{To}=\left[\begin{array}{cc}{A}_o& B_o\\ C_o& D_o\end{array}\right]=\frac{1}{z_{o21}}\left[\begin{array}{cc}{z}_{o11}& \left|z_o\right|\\ 1& z_{o22}\end{array}\right].$

Step 5: the transfer matrix of computing amplifier integral body $\mathrm{Tc}=\left[\begin{array}{cc}{A}_c& B_c\\ C_c& D_c\end{array}\right]=\mathrm{Ti}\&times;\mathrm{Tt}\&times;\mathrm{To},$ Transfer matrix Tc with amplifier integral body is converted to the collision matrix form again

$\mathrm{Sc}=\left[\begin{array}{cc}{S}_{c11}& S_{c12}\\ S_{c21}& S_{c22}\end{array}\right]=\frac{1}{A_c+B_c+C_c+D_c}\left[\begin{array}{cc}{A}_c+B_c-C_c-D_c& {2A}_c\\ 2& -A_c+B_c-C_c+D_c\end{array}\right].$

The voltage standing wave ratio of circuit, gain, noise factor and stability coefficient K design objective can be obtained with following formula:

${\mathrm{VSWR}}_{\mathrm{in}}=\frac{1+\left|S_{c11}\right|}{1-\left|S_{c11}\right|}---\left(21\right)$

${\mathrm{VSWR}}_{\mathrm{out}}=\frac{1+\left|S_{c22}\right|}{1-\left|S_{c11}\right|}---\left(22\right)$

Gain＝10×Log(|S _c21|)????????????????????????????????????????????(23)

$\mathrm{NT}={\mathrm{NT}}_{\min }+\frac{4R_n{T}_0}{Z_0}\frac{{|S_{i22}-{\mathrm{\&Gamma;}}_{\mathrm{opt}}|}^2}{(1-{\left|S_{i22}\right|}^2)(1-{\left|{\mathrm{\&Gamma;}}_{\mathrm{opt}}\right|}^2)}.---\left(24\right)$

$K=\frac{1-{\left|S_{c11}\right|}^2-{\left|S_{c22}\right|}^2+{\left|\mathrm{\&Delta;}\right|}^2}{2\left|S_{c12}{S}_{c21}\right|},(\mathrm{\&Delta;}=S_{c11}*S_{c22}-{\mathrm{<figure-callout\; id="12"\; label="S\; c"\; filenames="A20041000346700192.png"\; state="\{\{state\}\}">S</figure-callout>}}_c*S_{c21})---\left(25\right)$

Step 6: if the above index that the parameter designing of this circuit goes out meets the demands, design process finishes so, circuit parameter undetermined be in the step 2 the circuit parameter value (as the C of Fig. 7 in the example ₁, L ₁, L ₂And the C of Fig. 8 ₂, L ₃Numerical value);

Step 7: otherwise, circuit parameter revised (as the C of Fig. 7 in the example ₁, L ₁, L ₂And the C of Fig. 8 ₂, L ₃Numerical value), turn back to step 2.

With FHC40LG model PHEMT transistor is example, in the S of 830MHz parameter is:

Form 1 FHC40LG model PHEMT transistor 830 frequency range scattering parameters

f(GHz)??|S11|??∠S11??|S21|??∠S21??|S12|??∠S12??|S22|??∠S22

0.83??0.99??-14.88??5.71??164.96??0.01??80.45??0.49??-11.39

|| and ∠ represents the argument of the mould value of complex parameter respectively.Last table has provided transistor collision matrix S.

The noise parameter of 830MHz:

Form 2 FHC40LG model PHEMT transistors 800 frequency range noise parameters

f(GHz)??NTmin???|Γopt|??∠Γopt??????Rn

0.83????30K?????.95??????8????????????0.22

NTmin is transistorized minimal noise temperature; Γ opt is the best plural information source impedance of transistor; Rn is a noise resistance.More than these parameters provide by the handbook of manufacturer.

Design objective by the communication control processor regulation and stipulation is:

The design object and the design result of the amplifying circuit of form 3 embodiment

This routine design result of index Design target typical value

VSWRin?????????＜1.5????????????????1.11

VSWRout????????＜1.8????????????????1.42

Gain???????????＞16dB???????????????19.10

NF?????????????＜1dB????????????????0.71

K??????????????＞1??????????????????2.11

After determining above these data, specific implementation method:

Determine the single-level circuit form;

Calculate transistor according to transistor collision matrix St $K=\frac{1-{\left|S_{t11}\right|}^2-{\left|S_{t22}\right|}^2+{\left|\mathrm{\&Delta;}\right|}^2}{2\left|S_{t12}{S}_{t21}\right|}=0.043<1,$ Therefore introduce negative feedback network Mf, this negative feedback network is made up of the microstrip line of a segment length 6mm, utilizes formula (5) calculating to comprise degenerative transistor K=0.7255 then, and what obtain comprises degenerative collision matrix $\mathrm{St}=\left[\begin{array}{cc}{S}_{t11}& S_{t12}\\ S_{t21}& S_{t22}\end{array}\right];$

Consider following factors: 1): input matching network is high pass characteristic; 2): must be provided with partiting dc capacitor at the signal flow inbound port of input matching network and the signal output port of output matching network; Regulation capacitance numerical value is 50pF; 3): circuit must have a bypass and be used for introducing direct current biasing; Input matching network is selected the prototype of Fig. 3, considers the constraint of fan-in network high pass, adopts the way of realization of Fig. 7, and each element as shown in the figure; Output matching network is selected the prototype of Fig. 6, adopts the way of realization of Fig. 8, and each element as shown in Figure 8; To realize stopping direct current and AC coupled structure.

Carry out parameter optimization according to topology, calculate by the electric network theory formula $\mathrm{zi}=\frac{1}{Z0}\left[\begin{array}{cc}\frac{1}{\mathrm{j\&omega;}{\mathrm{<figure-callout\; id="1"\; label="C"\; filenames="A20041000346700201.png"\; state="\{\{state\}\}">C</figure-callout>}}_1}+\mathrm{j\&omega;}{\mathrm{<figure-callout\; id="2"\; label="L"\; filenames="A20041000346700191.png"\; state="\{\{state\}\}">L</figure-callout>}}_2& \mathrm{j\&omega;}{\mathrm{<figure-callout\; id="2"\; label="L"\; filenames="A20041000346700191.png"\; state="\{\{state\}\}">L</figure-callout>}}_2\\ \mathrm{j\&omega;}{\mathrm{<figure-callout\; id="2"\; label="L"\; filenames="A20041000346700191.png"\; state="\{\{state\}\}">L</figure-callout>}}_2& \mathrm{j\&omega;}({\mathrm{<figure-callout\; id="1"\; label="L"\; filenames="A20041000346700201.png"\; state="\{\{state\}\}">L</figure-callout>}}_1+L_2)\end{array}\right],$ $\mathrm{zo}=\frac{1}{Z0}\left[\begin{array}{cc}\mathrm{j\&omega;}{L}_3& \mathrm{j\&omega;}{L}_3\\ \mathrm{j\&omega;}{L}_3& \frac{1}{\mathrm{j\&omega;}{\mathrm{<figure-callout\; id="2"\; label="C"\; filenames="A20041000346700191.png"\; state="\{\{state\}\}">C</figure-callout>}}_2}+\mathrm{j\&omega;}{L}_3\end{array}\right],$ Further calculate the transfer matrix of fan-in network $\mathrm{Ti}=\left[\begin{array}{cc}\frac{z_{i11}}{z_{i21}}& \frac{\left|z_i\right|}{z_{i21}}\\ \frac{1}{z_{i21}}& \frac{z_{i22}}{z_{i21}}\end{array}\right]\left(\right|z_i|=z_{i11}{z}_{i22}-z_{i21}{z}_{i12}),$ The transfer matrix of output network $\mathrm{To}=\left[\begin{array}{cc}\frac{z_{o11}}{z_{o21}}& \frac{\left|z_o\right|}{z_{o21}}\\ \frac{1}{z_{o21}}& \frac{z_{o22}}{z_{o21}}\end{array}\right]$ (|z _o|＝z _o11z _o22-z _o21z _o12)。Transistor active circuit transfer matrix can obtain by the electric network theory formula equally, is designated as

$\mathrm{Tt}=\frac{1}{2s_{t21}}\left[\begin{array}{cc}1+s_{t11}-s_{t22}-\left|s_t\right|& 1+s_{t11}+s_{t22}+\left|st\right|\\ 1-s_{t11}-s_{t22}+\left|s_t\right|& 1-s_{t11}+s_{t22}-+\left|s_t\right|\end{array}\right];$

Utilize cascade network theorem and complex matrix multiplication to obtain amplifier global transfer matrix T c=Ti * Tt * To (order that attention is multiplied each other).Obtain the whole collision matrix of amplifier by electric network theory $\mathrm{Sc}=\frac{1}{A+B+C+D}\left[\begin{array}{cc}A+B-C-D& 2(\mathrm{AD}-\mathrm{BC})\\ 2& -A+B-C-D\end{array}\right],$ A＝T _t11，B＝T _t12，C＝T _t21，D＝T _t22。

Utilize formula (21)～(25) calculation Design target function value, all parameters only with step 5) gained Sc matrix correlation.

Finish if index meets the demands, otherwise come back to step 2.

Global optimum's parameter value that on the PC of dominant frequency 1400MHz internal memory 256Mb, searched out selected matching type network based on the computer program of above method in 92.3630 seconds consuming time.

Obtain embodiment 1 circuit process optimum parameters table:

Form 4 is optimized the circuit parameter value that obtains

Variable	Parameter value
		????C1	????50pF
????C2	????50pF
		????L1	????90nH
????L2	????23nH
		????L3	????9nH

Respective embodiments 1 circuit design result's standing-wave ratio, noise factor, gain, stability coefficient K key index:

Circuit index after form 5 is optimized

Index Design result

VSWR _in?????????????1.11

VSWR _out????????????1.42

Gain(dB)????????????19.10

NF(dB)??????????????0.71

K???????????????????2.11

For the validity of proof scheme, on the complex media substrate actual fabrication CDMA amplifier circuit, measurement result shows and meets fine with design:

Form 6 circuit measurement result on the 830MHz frequency

The index measurement result

VSWR _in???????????????1.20

VSWR _out??????????????1.50

Gain(dB)??????????????18.20

NF(dB)????????????????0.89

K?????????????????????2.20

The result show this paper design with optimization method be a kind of Optimization Design of amplifier accurately.

Claims (1)

1. the Optimization Design of microwave amplifiercation circuit is characterized in that: it be a kind of by an input matching network, comprise degenerative transistor and output matching network microwave amplifiercation circuit that constitutes of cascade successively, when being realized by calculation procedure, it comprises following steps successively:

(1) initialization:

Set the prototype circuit of input matching network and output matching network, this prototype circuit is by having internal structure or not having ∏ type that the electronic circuit piece of internal structure forms or 2 port electric networks of T type.Perhaps import the involution form circuit of this prototype circuit, they are ∏ type or the T type circuit that reduce by 1 or 2 parts, also are 2 port electric networks, but have only 2 kinds of different structures altogether; These input and output matching networks can corresponding a kind of the equivalent electric circuit form (describing method that adopts during conventional design) with input information source impedance and output loading.Above-mentioned input and output network is described and is deposited in the computer program with impedance matrix;

Degenerative transistor is described and is deposited in the computer program with collision matrix.

Target setting function and desired value thereof and their calculation procedure deposit computer in; Described target function is standing wave ratio of input voltage VSWR _In, output voltage standing-wave ratio VSWR _Out, the some projects among gain G ain, equivalent noise temperature NT, the stability coefficient K or their combination, the computing formula of each target function is as follows:

${\mathrm{VSWR}}_{\mathrm{in}}=\frac{1+\left|S_{c11}\right|}{1+\left|S_{c11}\right|},---\left(1\right)$

${\mathrm{VSWR}}_{\mathrm{out}}=\frac{1+\left|S_{c22}\right|}{1-\left|S_{c22}\right|}---\left(2\right)$

Gain＝10×Log(|S _c21| ²)，?????????????????????????????(3)

$\mathrm{NT}={\mathrm{NT}}_{\min }+\frac{{4R}_n{T}_0}{Z_0}\frac{{|S_{i22}-{\mathrm{\&Gamma;}}_{\mathrm{opt}}|}^2}{(1-{\left|S_{i22}\right|}^2)(1-{\left|{\mathrm{\&Gamma;}}_{\mathrm{opt}}\right|}^2)},.---\left(4\right)$

$K=\frac{1-{\left|S_{c11}\right|}^2-{\left|S_{c22}\right|}^2+{\left|\mathrm{\&Delta;}\right|}^2}{2\left|S_{c12}{S}_{c21}\right|},(\mathrm{\&Delta;}=S_{c11}*S_{c22}-S_{c12}*S_{c21}),---\left(5\right)$

Noise resistance R wherein _n, minimal noise temperature NT _Min, best information source reflection coefficient Γ _OptAll provide by transistor data bool.Described Sc matrix is the collision matrix of whole objective circuit, $\mathrm{Sc}=\left[\begin{array}{cc}{S}_{c11}& S_{c12}\\ S_{c21}& S_{c22}\end{array}\right],$ It is by the transition matrix Tc conversion of whole objective circuit, and transition matrix Tc=Ti * Tt * To, wherein, Ti is the normalized impedance matrix of input matching network $z_i=\left[\begin{array}{cc}{Z}_{i11}& Z_{i12}\\ Z_{i21}& Z_{i22}\end{array}\right]$ Transition matrix; Tt is that transistor active circuit collision matrix St transforms the transfer matrix that obtains; To is the normalized impedance matrix of output matching network $z_o=\left[\begin{array}{cc}{Z}_{o11}& Z_{o12}\\ Z_{o21}& Z_{o22}\end{array}\right]$ Transfer matrix;

(2) set information source impedance and load:

The value Z0 of information source impedance and load is 50 Ω, and the partiting dc capacitor value of the signal flow inbound port of input matching network and the signal output port of output matching network is calculated according to the capacitive reactance on the radio frequency operation frequency, is taken as 50pF.

(3) determine the progression of amplifier, circuit at different levels design respectively;

Determine the form of single-level circuit: input matching network have high pass characteristic and circuit must have a bypass with the requirement of introducing direct current biasing under, determine the circuit topology of input, output matching network and as the impedance component type and the variate-value thereof of variable, described impedance component is inductance or electric capacity or resistance or microstrip line or their polyphone or parallel circuits;

Select transistor and negative feedback value: calculate transistorized stability coefficient K value according to the transistor collision matrix earlier, if K＜1, introduce negative feedback network, this negative feedback network by electric capacity, resistance, microstrip line element or their combination through the polyphone or be formed in parallel, but the quantity of element is less than or equal to 3, setting member variate-value, calculating K successively again, only o'clock just can determine the component values of negative feedback network, obtain comprising negative feedback network at interior collision matrix to K＞0.6 $\mathrm{St}=\left[\begin{array}{cc}{S}_{t11}& S_{t12}\\ S_{t21}& S_{t22}\end{array}\right];$

(4) calculate following each amount according to above-mentioned specific input, output matching circuit:

Calculate normalization input impedance matrix according to input matching network $z_i=\left[\begin{array}{cc}{Z}_{i11}& Z_{i12}\\ Z_{i21}& Z_{i22}\end{array}\right];$

Calculate normalization output impedance matrix according to output matching network $z_o=\left[\begin{array}{cc}{Z}_{o11}& Z_{o12}\\ Z_{o21}& Z_{o22}\end{array}\right];$

(5) above-mentioned z _iBe transformed to transfer matrix $\mathrm{Ti}=\left[\begin{array}{cc}{A}_i& B_i\\ C_i& D_i\end{array}\right]=\frac{1}{{\mathrm{zi}}_{21}}\left[\begin{array}{cc}{\mathrm{zi}}_{11}& \left|\mathrm{zi}\right|\\ 1& {\mathrm{zi}}_{22}\end{array}\right];$

Above-mentioned St is transformed to transfer matrix:

$\mathrm{Tt}=\left[\begin{array}{cc}{A}_t& B_t\\ C_t& D_t\end{array}\right]=\frac{1}{{2s}_{21}}\left[\begin{array}{cc}1+{\mathrm{st}}_{11}-{\mathrm{st}}_{22}-\left|\mathrm{st}\right|& 1+{\mathrm{st}}_{11}+{\mathrm{st}}_{22}+\left|\mathrm{st}\right|\\ 1-{\mathrm{st}}_{11}-{\mathrm{st}}_{22}+\left|\mathrm{st}\right|& 1-{\mathrm{st}}_{11}+{\mathrm{st}}_{22}-\left|\mathrm{st}\right|\end{array}\right];$

Above-mentioned z _oBe transformed to transfer matrix $\mathrm{To}=\left[\begin{array}{cc}{A}_o& B_o\\ C_o& D_o\end{array}\right]=\frac{1}{{\mathrm{zo}}_{21}}\left[\begin{array}{cc}{\mathrm{zo}}_{11}& \left|\mathrm{zo}\right|\\ 1& {\mathrm{zo}}_{22}\end{array}\right];$

(6) calculate microwave amplifiercation circuit global transfer matrix $\mathrm{Tc}=\left[\begin{array}{cc}{A}_c& B_c\\ C_c& D_c\end{array}\right]=\mathrm{Ti}\&times;\mathrm{Tt}\&times;\mathrm{To};$

(7) be the global transfer matrixing collision matrix $\mathrm{Sc}=\left[\begin{array}{cc}{S}_{c11}& S_{c12}\\ S_{c21}& S_{c22}\end{array}\right]=$ $\frac{1}{A_c+B_c+C_c+D_c}\left[\begin{array}{cc}{A}_c+B_c-C_c-D_c& 2A_c\\ 2& -A_c+B_c-C_c+D_c\end{array}\right];$

(8) calculate standing wave ratio of input voltage VSWR according to formula (1)～(5) _In, output voltage standing-wave ratio VSWR _Out, gain G ain, equivalent noise temperature NT, stability coefficient K value;

(9) whether the target function value that obtains of determining step (8) meets design requirement, if satisfy, then they parameter values as input, output matching network; Otherwise, the value of respective impedance element in modification input, the output matching network, repeating step (4)～(9) are only to VSWR _In, VSWR _Out, when Gain, NT, K satisfy the index of setting till.

Images