5. Impedance Matching and Tuning

1 / 58

# 5. Impedance Matching and Tuning - PowerPoint PPT Presentation

5. Impedance Matching and Tuning. Apply the theory and techniques of the previous chapters to practical problems in microwave engineering. Impedance matching is the 1 st topic. Figure 5.1 (p. 223) A lossless network matching an arbitrary load impedance to a transmission line.

I am the owner, or an agent authorized to act on behalf of the owner, of the copyrighted work described.

## 5. Impedance Matching and Tuning

Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author.While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server.

- - - - - - - - - - - - - - - - - - - - - - - - - - E N D - - - - - - - - - - - - - - - - - - - - - - - - - -
Presentation Transcript
5. Impedance Matching and Tuning
• Apply the theory and techniques of the previous chapters to practical problems in microwave engineering.
• Impedance matching is the 1st topic.

Figure 5.1 (p. 223)A lossless network matching an arbitrary load impedance to a transmission line.

Impedance matching or tuning is important since
• Maximum power is delivered when the load is matched to the line, and power loss in the feed line is minimized.
• Impedance matching sensitive receiver components improves the signal-to-noise ratio of the system.
• Impedance matching in a power distribution network will reduce the amplitude and phase errors.
Important factors in the selection of matching network.
• Complexity
• Bandwidth
• Implementation
• Ajdustability
5.1 Matching with Lumped Elements
• L-section is the simplest type of matching network.
• 2 possible configurations

Figure 5.2 (p. 223)L-section matching networks. (a) Network for zL inside the 1 + jx circle. (b) Network for zL outside the 1 + jx circle.

Analytic Solution
• For Fig. 5. 2a, let ZL=RL+jXL. For zL to be inside the 1+jx circle, RL>Z0. For a match,
• Removing X

Figure 5.3b (p. 227)(b) The two possible L-section matching circuits. (c) Reflection coefficient magnitudes versus frequency for the matching circuits of (b).

5.2 Single Stub Tuning

Figure 5.4 (p. 229)Single-stub tuning circuits. (a) Shunt stub. (b) Series stub.

• d: from the load to the stub position.
• B or X provided by the shunt or series stub.
• For the shunt-stub case,
• Select d so that Y seen looking into the line at d from the load is Y0+jB
• Then the stub susceptance is chosen as –jB.
• For the series-stub case,
• Select d so that Z seen looking into the line at d from the load is Z0+jX
• Then the stub reactance is chosen as –jX.
Shunt Stubs
• Ex 5.2 Single-Stub Shunt Tuning

ZL=60-j80

Figure 5.5a (p. 230)Solution to Example 5.2. (a) Smith chart for the shunt-stub tuners.

Figure 5.5b (p. 231)(b) The two shunt-stub tuning solutions. (c) Reflection coefficient magnitudes versus frequency for the tuning circuits of (b).

To derive formulas for d and l, let ZL= 1/YL= RL+ jXL.
• Now d is chosen so that G = Y0=1/Z0,
• To find the required stub length, BS = -B.

for open stub

for short stub

Series Stubs
• Ex 5.3 Single Stub Series Tuning

ZL = 100+j80

Figure 5.6a (p. 233)Solution to Example 5.3. (a) Smith chart for the series-stub tuners.

Figure 5.6b (p. 232)(b) The two series-stub tuning solutions. (c) Reflection coefficient magnitudes versus frequency for the tuning circuits of (b).

To derive formulas for d and l, let YL= 1/ZL= GL+ jBL.
• Now d is chosen so that R = Z0=1/Y0,
• To find the required stub length, XS = -X.

for short stub

for open stub

5.3 Double-Stub Tuning
• If an adjustable tuner was desired, single-tuner would probably pose some difficulty.

Smith Chart Solution

• yL add jb1 (on the rotated 1+jb circle)  rotate by d thru SWR circle(WTG)  y1  add jb2  Matched
• Avoid the forbidden region.

Figure 5.7 (p. 236)Double-stub tuning. (a) Original circuit with the load an arbitrary distance from the first stub. (b) Equivalent-circuit with load at the first stub.

Figure 5.9a (p. 238)Solution to Example 5.4. (a) Smith chart for the double-stub tuners.

Ex. 5.4 ZL = 60-j80

Open stubs, d = λ/8

Figure 5.9b (p. 239)(b) The two double-stub tuning solutions. (c) Reflection coefficient magnitudes versus frequency for the tuning circuits of (b).

Analytic Solution
• To the left of the first stub in Fig. 5.7b,

Y1 = GL + j(BL+B1) where YL = GL + jBL

• To the right of the 2nd stub,
• At this point, Re{Y2} = Y0
Since GL is real,
• After d has been fixed, the 1st stub susceptance can be determined as
• The 2nd stub susceptance can be found from the negative of the imaginary part of (5.18)
B2 =
• The open-circuited stub length is
• The short-circuited stub length is
5.4 The Quarter-Wave Transformer
• Single-section transformer for narrow band impedance match.
• Multisection quarter-wave transformer designs for a desired frequency band.
• One drawback is that this can only match a real load impedance.
• For single-section,
Figure 5.10 (p. 241)A single-section quarter-wave matching transformer. at the design frequency f0.

where t = tanβl = tanθ, θ = π/2 at f0.

• The reflection coefficient
• Since Z12 = Z0ZL,
The fractional bandwidth is
• Ex. 5.5 Quarter-Wave Transformer Bandwidth

ZL = 10, Z0 = 50, f0= 3 GHz, SWR ≤ 1.5

Figure 5.12 (p. 243)Reflection coefficient magnitude versus frequency for a single-section quarter-wave matching transformer with various load mismatches.

5.5 The Theory of Small Reflection

Single-Section Transformer

Figure 5.13 (p. 244)Partial reflections and transmissions on a single-section matching transformer.
Multisection Transformer
• Assume the transformer is symmetrical,
5.6 Binomial Multisection Matching Transformer
• The response is as flat as possible near the design frequency.  maximally flat
• This type of response is designed, for an N-section transformer, by setting the first N-1 derivatives of |Γ(θ)| to 0 at f0.
• Such a response can be obtained if we let
Note that |Γ(θ)| = 0 for θ=π/2, (dn |Γ(θ)|/dθn ) = 0 at θ=π/2 for n = 1, 2, …, N-1.
• By letting f  0,
Γn must be chosen as
• Since we assumed that Γn are small, ln x ≈ 2(x-1)/(x+1),
• Numerically solve for the characteristic impedance  Table 5.1
The bandwidth of the binomial transformer
• Ex. 5.6 Binomial Transformer Design

Figure 5.15 (p. 250)Reflection coefficient magnitude versus frequency for multisection binomial matching transformers of Example 5.6 ZL = 50Ω and Z0 = 100Ω.

5.7 Chebyshev Multisection Matching Transformer

Chebyshev Polynomial

• The first 4 polynomials are
• Higher-order polynomials can be found using
Properties
• For -1≤x ≤1, |Tn(x)|≤1  Oscillate between ±1  Equal ripple property.
• For |x| > 1, |Tn(x)|>1  Outside the passband
• For |x| > 1, |Tn(x)| increases faster with x as n increases.
• Now let x = cosθ for |x| < 1. The Chebyshev polynomials can be expressed as

More generally,

Design of Chebyshev Transformers
• Using (5.46)
• Letting θ = 0,
Once θm is known,

Ex 5.7 Chebyshev Transformer Design

Γm = 0.05, Z0 = 50, ZL = 100

Use Table 5.2

Figure 5.17 (p. 255)Reflection coefficient magnitude versus frequency for the multisection matching transformers of Example 5.7.

Figure 5.18 (p. 256)A tapered transmission line matching section and the model for an incremental length of tapered line. (a) The tapered transmission line matching section. (b) Model for an incremental step change in impedance of the tapered line.

Figure 5.19 (p. 257)A matching section with an exponential impedance taper. (a) Variation of impedance. (b) Resulting reflection coefficient magnitude response.

Figure 5.20 (p. 258)A matching section with a triangular taper for d(In Z/Z0/dz. (a) Variation of impedance. (b) Resulting reflection coefficient magnitude response.

Figure 5.21 (p. 260)Solution to Example 5.8. (a) Impedance variations for the triangular, exponential, and Klopfenstein tapers. (b) Resulting reflection coefficient magnitude versus frequency for the tapers of (a).

Figure 5.22 (p. 262)The Bode-Fano limits for RC and RL loads matched with passive and lossless networks (ω0 is the center frequency of the matching bandwidth). (a) Parallel RC. (b) Series RC. (c) Parallel RL. (d) Series RL.

Figure 5.23 (p. 263)Illustrating the Bode-Fano criterion. (a) A possible reflection coefficient response. (b) Nonrealizable and realizable reflection coefficient responses.