• 沒有找到結果。

CHAPTER 2 The Architecture and Decide the Specifications

G

GA = PAVN / PAVS

ed from the network / Power available from the

(2.2)

GA物理意義:前級考慮最低雜訊的匹配,即 ;後級共軛匹配,

Γ

低雜訊放大器的設計,通常將輸出端設定為共軛匹配即 ,以達

到後 當的

態 本論文

穩定性(stability)描述了放大器電路是否能避免振盪的程度,為放大器 設計

2.9 Stability of two-port network 示意圖

T物理意義:前級與後級皆不共軛匹配,即Γs Γin*ΓLΓout*

= Power deliver source

in* s =Γ Γ Γ *

= out ,適合用於LNA的設計。

L

out* L=Γ Γ

端最低的反射功率及最大的輸出增益。輸入端則選擇適 ,使得 Γ *

=

Γ ,以達到輸入端較低的反射係數且產生較低雜訊的最佳狀 ;因此 即採用Available power gain,G

Γs

in s

A的功率增益定義來設計低雜訊放大 器。

步驟中一項重要的考慮因素。而穩定性可以由電晶體的S(S-parameter) 參數、各匹配網路以及終端阻抗型式…等等原因來決定並得之一個滿足無 條件穩定(unconditionally stable)的電路,可以在電路輸入及輸出兩端,接 上任何被動元件,都能保持穩定工作而不致引發振盪,如圖 2.7 所示;而 在已知頻率下,依反射係數的定義,達到無條件穩定的條件為:

Z

L

V

S

+

- Transistor

Г

OUT

CHAPTER 2 The Architecture and Decide the Specifications

Γs <1 (2.3)

ΓL <1 (2.4)

) 0 ] (Re[

<1 Z 1 22

12

11 21 >

Γ

+ Γ

=

Γ IN

L L

IN S

S

S S (2.5)

以及,

) 0 ]>

(Re[ UT

1

12 <

=

Γ O

S

OUT S S Z

S .6)

上式中的所有係數均都已對特 阻 進行正規化的動作。討論至

,我們可得到以下的結論:

(2

11 21

Γ ΓS 1 S

22 +

性 抗Z0

1

S11 < 、S22 <1且 k>1、∆ <1 (2.7)

其中,

(2.8)

21 12

2 2

22 2

11

S S 2

S S

k 1− − + ∆

=

21 12 22

11S S S

S −

=

(2.9)

符合上述條件,則稱該網路必為無條件穩定。

在射頻電路當中,有一些非理想的特性,包括存在於任何環境中的雜

訊,而這會對電 度的干 響到所要 信號;且再來是

路本身的非線性應,使之無法避免信號間的相互干擾,而或因為元件本 身的限制,影響電路的性能。

2.3.2.1 諧波失真(Harmonic Distortion) 如

2.3.3 非線性效應

路造成相當程 擾,並影 的

在非線性系統中,若輸入之信號為fc在經過放大器後,除了主要的信 外,還會產生多次項的諧波(註.1),其示意如圖2.8所示:

CHAPTER 2 The Architecture and Decide the Specifications

大器輸入與輸出之頻譜示意圖 圖2.10 非線性放

2.3.2.2 增益壓縮點(Gain Compression)

增益壓縮點(1dB Compression Point)(註.2)(註.4) (註.5),因為非線性效 應的關係,放大器的增益在小訊號時為線性,但功率加大時增益就會逐漸

小。為了表示放大器的工作範圍,我們通常以輸出增益比線性增益小1dB 的 出功 ,稱為1dB增益壓縮點 其示意如圖2-9所示。另外,一般放大 的1–dB compression point OIP3 10dB,定義如下式:

時 輸 率 ,

器 皆低於 大約為

1 ) dB ( G ) dB (

G1dB = 0 − (2.10) (

P1dB dBm)=G1dB(dB)+IP1dB(dBm) (2.11)

當S11 <1、S22 <1 (2.12)

圖2.11 1dB增益壓縮點示意圖

G0: 線性放大的增益量 P1dB: 1dB 增益壓縮點

IP1dB: 在1dB 增益壓縮點時的輸入訊號功率

VIN VOUT

VIN

VOUT

fc fc 2fc 3fc

f f

1dB Compression point

CHAPTER 2 The Architecture and Decide the Specifications 2.3.2.3 交互調變(Intermodulation)

在多個頻 時由於電路

件的非線性效應,會造成不同頻率的信號產生交互調變(Intermodulation)

(註.3) (註.4) (註.5) 落在我們所要的

頻道之內, 率(BER),

其示意如圖2.10 所示 個鄰近干擾訊號,經過非

線性放大器時,其 在所要的頻道內,惡化

了訊號品質。故三階截斷點 IIP3及OIP3分別為

輸入及輸出的三階截斷點 ,一般而

言,交會點越高,或者IIP3及OIP3越大,代表線性

道的系統或是無線通訊的環境之下,訊號在接收 元

的作用,若交互調變的非線性失真信號

會造成訊號的相互干擾,增加訊號解調後的位元錯誤

,而無法由濾波器濾除的兩 三階交互調變失真(IM3)信號會落

(IP3),如圖2.11所示,其中

(input or output third-order intercept point) 度越好。

圖2.12 三階非線性現象示意圖

圖2.13 三階截斷點(IP3)示意圖

PIN(dBm) POUT

(dBm) PIP3 (dBm)

IMD (dBc)

X 2X

IM3 (dBm)

1

3 PIP3 = POUT+IMD/2 Fundamental

IM3 X

CHAPTER 2 The Architecture and Decide the Specifications 註.1 當輸出訊號壓縮 ω1與3ω1之振幅大小為何?證明之。

解:

We assum Acosω1t

And we can express the Vout = K0+K1Vin+K2Vin2+K3Vin3+… by use

Taylor series,now xpression about Vout,

out = ( K0+0.5 K2A2) + ( K1A+0.75 K3 A3)cosω1 1dB時,其ω1、2

VIN VOUT

e that Vin =

we instead of Vin in that e we can find that

V t + 0.5 K2A2cos2ω1t

+ 0.25 K3A3cos3ω1t + … now we can draw a diagram abou

So, t PIN(dBm) versus POUT(dBm)

the slope of fundamental|A|the slope of 2nd|A|2 So wh rade 1dB (Pin),we can know the

Ps. Actuality,We now that our gain will compression because nonlin can know that K3 will be negative .

Ps1. 2nd is denote the Second Harmonic 3rd is denote the Third Harmonic

in it,we can find

and the slope of 3rd|A|3

en our input signal deg

magnitude of 2nd will degrade 2dB and the magnitude of 3rd will degrade 3dB .

can k earity,so we

VIN

Fundamental

2nd 3rd VOUT

PIN(dBm)

1 1

1

2 3

1

Saturated power PSAT

POUT(dBm)

CHAPTER 2 The Architecture and Decide the Specifications

註.2 當輸入一弦波訊號 Vin = A 明

之。

解:

We assume that Vin = Acosω t

And we can express the Vout = K0+K1Vin+K2Vin2+K3Vin3+… by use Taylor series,now we instead of Vin in that expression about Vout,So

Vout = K0 + K1Vin + K2Vin2 + K3Vin3 + …

θ)/2

5 K2A2) + (K1A+0.75 K3 A3)cosω1

cosω1t 時,其輸出訊號之大小為何?證

1

= K0 + K1(Acosω1t) + K2(Acosω1t)2 + K3(Acosω1t)3 + … Because,cos2θ = (1+cos2

cos3θ = (3cosθ+cos3θ)/4

So,we can find a result that Vout can express ,

Vout = (K0+0. t + 0.5 K2A2cos2ω1t

+ 0.25K3A3cos3ω1t + …

CHAPTER 2 The Architecture and Decide the Specifications 註.3 當輸入一弦波訊號 Vin = Acosω1t + Acosω2t 時,其輸出訊號之大小

為何?證明之。

We assume that Vin = Acosω1t + Acosω2t

And we can express the Vout = K0+K1Vin+K2Vin2+K3Vin3+… by use Taylor series,now we instead of Vin in that expression about Vout,So

Vout = K0 + K1Vin + K2Vin2 + K3Vin3 + …

= K0 + K1(Acosω1t + Acosω2t) + K2(Acosω1t + Acosω2t)2 + K3(Acosω1t + Acosω2t)3 + …

Because,(cosα+cosβ)2 = cos 2α + 2cosαcosβ + cos2β

= cos2α + cos(α+β) + cos(α-β) + cos2β

= 1+ (cos2α + cos2β)/2 + cos(α+β) + cos(α-β) (cosα+cosβ)3 = (cosα + cosβ)(cos 2α + 2cosαcosβ + cos2β) = cos3α + 3cos2αcosβ + 3cosαcos2β + cos3β in it,

3cosα(cosαcosβ) = 3[cos(2α+β) + 2cos(β) + cos(2α-β)]/4 3cosβ(cosαcosβ) = 3[cos(α+2β) + 2cos(α) + cos(-α+2β)]/4 So,

+ 3[cos(α+2β) + 2cos(α) + cos(-α+2β)]/4 + cos3β

/2 K2A2[cos2ω1t + cos2ω2t] 2nd 3rd (2ω2-ω1)t] IM3

(cosα+cosβ)3 = cos3α + 3[cos(2α+β) + 2cos(β) + cos(2α-β)]/4

So,

Vout = (K0 + K2A2) DC + (K1A + 3/4K3A3)(cosω1t + cosω2t) 1st + 1

+ 1/4 K3A3[cos3ω1t + cos3ω2t]

+ 1/2 K2A2[cos(ω2+ω1)t + cos(ω2-ω1)t] IM2 + 3/4 K3A3[cos(2ω1-ω2)t + cos

+ …

CHAPTER 2 The Architecture and Decide the Specifications 註.4 當輸入訊號為單一頻率時,其 1dB 增益壓縮點(P1dB)與三階截斷點 解:

[One tone]:

We assume that Vin = Acosω1t

Vout = K1A --- Ideal (linear)

Vout = K1A + 3/4K3A3(K3 always is negative) --- Actuality (nonlinear) So,

(K1A - 3/4K3A3)|dB – (K1A)|dB = -1dB (take 20 logarithmic) (K1A - 3/4K3A3) / (K1A) = 10-1/20 = 0.89125

An

= 3/4K3A3

A2 = K1/0.75K

So,

10log[AIP32 / ] = 10log[(K1/0.75K3)/( 0.10875K1 / 0.75K3)] = On output ratio increased by 1dB

So,

POIP3 – P1dB =

(IP3out)為何?證明之。

PIP3

A2 = 0.10875K1 / 0.75K3 = AP1dB2 d,

K1A

3 = AIP32 AP1dB2

9.6377dB

10.6377dB 1

3

Fundamental P1dB

10.63dB

IM3(dBm)

PIN(dBm)

?dBc

CHAPTER 2 The Architecture and Decide the Specifications 註.5 當輸入訊號為兩根頻率時,其 1dB 增益壓縮點(P1dB)與三階截斷點

osω2t ( A1≠A2 )

we can express the Vout = K0 + K1Vin + K2Vin2 + K3Vin3 + … by So,

3Vin3 + …

= K0 + K1(A1cosω1t + A2cosω2t) + K2(A1cosω1t + A2cosω2t)2 + osω2t)3 + …

= K0 + K1(A1cosω1t + A2cosω2t) + K2[A12/2 + (A12cos2ω1t)/2 + A1A2 cos(ω1+ω2)t + A1A2 cos(ω1-ω2)t + A22/2 +

(A22cos2ω2t)/2] + K3[(3A13/4 + 3A1A22/2)cosω1t + (3A23/4 + 3A12A2/2)cosω2t + 3[A12A2cos(2ω1+ω2)t]/4 +

3[A1A22cos(ω1+2ω2)t]/4 + 3[A12A2 cos(2ω1-ω2)t]/4 + 3[A1A22cos(ω1-2ω2)t]/4 + A13cos(3ω1)t/4 + A23cos(3ω2)t/4]

+ …

= (K0 + K2 A12/2 + K2 A22/2) DC + 3K3A12A2/2)(cosω2t) 1st + [(K2A12)cos2ω1t]/2 2nd

22)cos2ω2t]/2 2nd 3rd

+ 3K3A12A2[cos(2ω1+ω2)t + cos(2ω1-ω2)t]/4 IM3 ω2)t]/4 IM3

Vout = K

Vout = K tive)

So,

(K1A1 -(take 20

(K1A1 - 13/4 - 3K3A1A22/2) / K1A1) = 10-1/20 = 0.89125 (IP3out)為何?證明之。

解:

[Two tones]:

When Vin = A1cosω1t + A2c Then

using Taylor series,now we instead of Vin in that expression about Vout,

Vout = K0 + K1Vin + K2Vin2 + K K3(A1cosω1t + A2c

+ (K1A1 + 3K3A13/4 + 3K3A1A22/2)(cosω1t) 1st + (K1A2 + 3K3A13/4

+ [(K2A

+ (K3A13)cos(3ω1)t/4

+ (K3A23)cos(3ω2)t/4 3rd + {K2A1A2[cos(ω1+ω2)t + cos(ω1-ω2)t]}/2 IM2 + 3K3A1A22[cos(ω1+2ω2)t + cos(ω1-2

+ …

1A1 --- Ideal (linear)

1A1 + 3K3A13/4 + 3K3A1A22/2 (K3 always is nega ---- Actuality(nonlinear)

3K3A13/4 - 3K3A1A22/2)|dB – (K1A1)|dB = -1dB logarithmic)

3K3A

CHAPTER 2 The Architecture and Decide the Specifications

A12 = 0.10875*4K1 /9 K3 = AP1dB2 And,

K1A1 =

A22 = K

Now we So,

10log[A On outp So,

POIP3 – .40692dB

15.4069 4.76922 dB

e P1dB will d

3/4K3A1A22 (The slope of the third order product is the same) 1/0.75K3 = AIP32

assume A1 = A2 = A,

IP32/ AP1dB2] = 10log[(K1/0.75K3)/( 0.10875*4K1 /9 K3)]

= 14.40692dB ut ratio increased by 1dB

P1dB = 15

Comparison the results of one tone and two tones:

2dB - 10.6377dB =

Now we can find if the fundamental signal are two tones input,th egrade 4.76922 dB

CHAPTER 3 Design Considerations of Radio Frequency Circuit