Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                

Phase Shifters Simplified 2

Download as pdf or txt
Download as pdf or txt
You are on page 1of 94

MICROWAVE PHASE SHIFTERS

Phase shifters are two port components that provide variable phase shift by
changing a control dc bias. An ideal phase shifter is a matched lossless structure.
Phase shifters can be classified as analog or digital.
Analog phase shifters change continuously the phase of the output signal by a
corresponding continuous variation of the control dc bias.
Phase shifters are called digital when the phase of the output signal can be
changed only in a range of few discrete values.
Microwave phase shifters find extensive applications in phased array antenna
systems. Phase control of the signal fed to the various radiating elements of the
array allows the direction of the radiated beam to be scanned electronically.
BINARY PHASE SHIFTERS

A binary phase shifter is a two state two port network. Therefore binary phase shifters
are particular cases of the digital ones.

φ0 , φc0 +Δφ

Δφ is the phase difference between the output signals in the two states and
is conventionally called the differential phase shift. For φ0 we assume φ0 = 0.
Tipically in a digital phase shifter the number of states is an integer power
of two, that is 2N, where the integer number N is the number of bits.
Digital phase shifter are realized by cascading N cells, each one composed by
a binary (two state) phase shifter. The schematic of a 4-bit phase shifter is
shown in figure. For such a structure the phase of the output signal can take
the following values: 0°, 22.5°, 45°, 67.5°, 90°, 112.5°, 135°, 157.5°, 180°,
202.5°, 225°, 247.5°, 270°, 292.5°, 315°, 337.5° .

0°,180° 0°,90° 0°,45° 0°,22.5°


REALIZATION OF PHASE SHIFTERS

Several kinds of devices can be used for the realization of microwave and millimeter
phase shifters:

a) Ferrite components

b) Semiconductor devices (p-i-n diodes, varactors, field effect transistors (MESFET))

c) MEMS components
SWITCHED LINE PHASE SHIFTERS

The switched line phase shifter is basically a circuit in which the phase shift is obtained
by switching between two transmission line sections of different lengths. To this end two
SPDT switches are used.

Z0, l1

1 SPDT SPDT 2

Z0, l2
Let’s assume β as the propagation constant of each transmission line. When the
transmission is through the line with geometric length l1, the phase shift of the signal
is φ1 = βl1 . If the transmission is through path length l2, the phase shift of the signal is
φ2 = βl2. Thus for the phase difference between the two switching states Δφ we have:

Δφ= φ1 - φ2 = β(l1- l2)


Since β = ω/c, being c the light speed, we can obtain for φ1 and φ2:

φ1 = ω l1 /c , φ2 = ω l2 /c Δφ = ω/c (l1- l2)

If ω0 is the working frequency, the behavior of φ1 and φ2 versus frequency is plotted


in figure.

φ1 , φ2
φ1
It is worth noting that for a phase
shifter designed for the working
φ2
frequency ω0 the phase error
increases linearly with | ω - ω0|. Δφ

ω0 ω
The switched line phase shifter with the two SPDT switches realized by means of
series mounted diodes.

D1 D3

D2 D4
The switched line phase shifter with the two SPDT switches realized by means of
shunt mounted diodes.

D1 D3

λ/4 λ/4
λ/4 λ/4
D2 D4
THE LOADED LINE PHASE SHIFTER
The loaded line phase shifter is composed by a transmission line
loaded with a pair of switchable circuits. The input admittance of each
circuit is Yi , and can assume two values: Y1 and Y2. This phase shifter
can be modeled as shown in the figure.

1 Yi ZC , θ Yi 2

Since the phase shifter is a lossless matched


two port - network we have : S 21 = e − j
If Yi = jBi (i = 1, 2 ), for Bi = B1 we have  = 1
and for Bi = B2 we have  = 2

A common choice for  is  =  2.


Defining  = 1 − 2 we obtain for Z C and Bi :

Z C = Z 0 cos  2
1 1
B1 = tan  2 , B2 = − tan  2 = − B1
Z0 Z0
THE LOADED LINE PHASE SHIFTERS WITH
THE DIODES SERIES MOUNTED

ZC , θ

Z0 , θ1 Z0 , θ1

Z0 , θ2 Z0 , θ2
THE LOADED LINE PHASE SHIFTERS WITH
THE DIODES SHUNT MOUNTED

ZC , θ

Z0 , θ1 Z0 , θ 1

Z0 , θ2 Z0 , θ2
REFLECTION TYPE PHASE SHIFTERS

This phase shifter makes use of a 3-dB 90° divider, with two of its ports
terminated in identical reflective networks. The reflection coefficient Γ
of each reflective network is controllable. The branch-line divides the
input signal equally between ports 3 and 4, but with a phase difference
of 90°. Port 2 is uncoupled to port 1. The signals reflected back from the
two terminations add up at port 2, and no signal returns to port 1.

b4
a1 INPUT 1 4 Γ
a4
a3
b2 OUTPUT 2 3 Γ
b3
b4
a1 INPUT 1 4 Γ
a4
a3
b2 OUTPUT 2 3 Γ
b3

j 1
S 41 = S 23 = − , S 31 = S 24 = −
2 2

b1 = S14 ΓS 41a1 + S13 ΓS31a1 = (S14 S 41 + S13 S31 )Γa1

 − j  2  − 1  2   1 1
b1 =   +   Γa1 = − +  Γa1 = 0
 2   2    2 2
b4
a1 INPUT 1 4 Γ
a4
a3
b2 OUTPUT 2 3 Γ
b3

j 1
S 41 = S 23 = − , S 31 = S 24 = −
2 2
b2 = S 24 ΓS 41a1 + S 23 ΓS31a1 = (S 24 S 41 + S 23 S 31 )Γa1

 − 1  − j   − j  − 1  j
b2 =   +   Γa1 = 2 Γa1 = ja1 Γ
 2  2   2  2  2
j j
Γ = Γe  b2 = ja1 Γ = j Γ e a1

If the argument  of the reflection coefficient Γ can be


changed, the circuit acts as a phase shifter. To this end,
a p - i - n diode is utilized in each reflective network, so
obtaining a binary phase shifter. If a varactor diode is used,
an analog phase shifter can be realized.
Moreover, if the reflecting networks are lossless
structures, we have :

j
Γ =1  b2 = je a1
Realization of reflective termination 1

Z0 , θ Z0 , Δφ/2 Z0 , θ Z0 , Δφ/2

The diode is modeled by an open circuit in the OFF state,


thus if  = 0 Γ can be written as : Γ = −e − j

The diode is modeled by a short circuit


in the ON state, thus if  = 0 Γ can Z0 , θ

be written as : Γ = −1
Realization of reflective termination 2

Z0 , θ Z0 , Δφ/2 Z0 , θ

The diode is modeled by an open circuit in the OFF state,


thus if  = 0 Γ can be written as : Γ = 1

The diode is modeled by a short Z0 , θ Z0 , Δφ/2


circuit in the ON state and if  = 0
Γ can be written as : Γ = e − j
EXAMPLE OF A DIGITAL PHASE SHIFTER
SWITCHED NETWORK PHASE SHIFTER
This structure may be considered as a conceptual evolution of the switched line
phase shifter configuration. When the input signal, originally passing through
network 1, is switched to pass through network 2, a differential phase shift φ2 - φ1
is obtained. The most commonly used networks are the high-pass and the
low-pass circuits.

network 1

Φ1

Φ2

network 2
The final configuration

jX jX

JB

-jX -jX

-jB
THE DISTRIBUTED PHASE SHIFTER
INTRODUCTION TO DISTRIBUTED PHASE SHIFTERS

lg

Z0 , φ


 =   lg = lg l g =  Ll Cl  l g , where
v
Ll = inductance per unit length
Cl = inductance per unit length
If Cl is a function of a bias voltage Vb , Cl = Cl (Vb ), also  is
a function of Vb ,  =  (Vb ). Thus the key point of this phase
shifter is to modify the phase of the output signal by
changing Vb .
DISTRIBUTED PHASE SHIFTERS
The distributed phase shifter is composed by transmission line sections periodically
loaded with variable capacitors Cb. Each transmission line section can be approximated
by a lumped inductance L and two lumped capacitances C.

ZC , θ ZC , θ

Cb Cb Cb

ZC , θ C C
L L

Cb C C Cb C C Cb

We can define a unit cell for this periodic


structure which consists of a section of a
Cb/2 ZC , θ Cb/2
transmission line and two shunt variable
capacitors Cb /2 to ground.

Cb/2 C C Cb/2
L

ZC , θ C C

L
Ll = inductance per unit length =
l
2C
Cl = inductance per unit length =
l
Ll
 =  Ll Cl = , Z C =
Cl
L

Cb/2 C C Cb/2

A synthetic transmission line can be defined, whose


capacitance per unit length C t is increased due to the
loading capacitors. Ct is given by :

Cb + 2C Cb
Ct = = + Cl
l l
The inductance per unit length of the synthetic
transmission line is unchanged from that of the original
line Ll . The phase shift of the unit cell t can be written as :

 Cb 
t =  t  l =  Ll Ct  l =   l  Ll  + Cl 
 l 

It is worth noting that if Cb = 0 we obtain  =  t =  Ll Cl


that is  for the unloaded cell. Cb is a function of the
bias voltage Vb . Therefore if we have for Vb the two values
Vb1 and Vb 2 , the correspond ing values for Cb are Cb1 and Cb 2 .
Hence we have obtained a binary phase shifter, with
the following two values for t :
 Cb1   Cb 2 
t1 =   l  Ll  + Cl  , t 2 =   l  Ll  + Cl 
 l   l 

A differential phase shift per unit length can be so defined :

t1 − t 2   Cb1   Cb 2 
=   Ll  + Cl  − Ll  + Cl  
l   l   l  
The same structure can be used as an analog phase shifter
if Vb , and therefore also Cb , is changed continuosly.
For the synthetic transmission line the characteristic
impedance Z Ct can be defined as follows :

Ll Ll ZC
Z Ct = = =
Ct Cb 1 Cb
+ Cl +1
l Cl l

Since Z Ct is a function of the bias voltage Vb , by changing


Vb a mismatch could occur.
REALIZATION OF THE DISTRIBUTED PHASE SHIFTER BY MEANS OF MEMS SWITCHES
Photo of a 6 cells, Distributed Phase Shifter

200
150
100
Phase [deg]

50
0 180°
-50
-100
DOWN
-150 UP
-200
9 10 10 10
0.0 5.0x10 1.0x10 1.5x10 2.0x10
Frequency [Hz]
APPENDIX
DESIGN EXAMPLE OF A SWITCHED LINE
PHASE SHIFTER
SCHEMATIC 1
SCHEMATIC 2
LOADED LINE PHASE SHIFTER
DESIGN EXAMPLE

tan  2 0.1989
 = 22.5  B1 = = = 0.003978
Z0 50
B2 = − B1 = −0.003978 , Z C = Z 0 cos  2 = 49 
1
X f = j L , X r = j  L + , f = 9.375 GHz
j C
L = 0.2 nH , C = 0.08 pF
1) 1 = 20.98 ,  2 = 169.87
2 ) 1 = 159.02 ,  2 = 28.77
SCHEMATIC 1
SCHEMATIC 2
THE MICROSTRIP PHASE SHIFTER
REFLECTION TYPE PHASE
SHIFTER WITH 90°
DIFFERENTIAL PHASE SHIFT
REFLECTION TYPE PHASE SHIFTER
SHORT- CIRCUITED

Z0 , θ Z0 , Δφ/2
SCHEMATIC 1
SCHEMATIC 2
REFLECTION TYPE PHASE SHIFTER OPEN

Z0 , θ Z0 , Δφ/2
SCHEMATIC 1 OPEN
SCHEMATIC 2 OPEN
REFLECTION TYPE PHASE
SHIFTER WITH
DIRECTIONAL COUPLER
REFLECTION TYPE PHASE SHIFTER
SHORT- CIRCUITED
SCHEMATIC 1
SCHEMATIC 2
REFLECTION TYPE PHASE SHIFTER OPEN
SCHEMATIC 1
SCHEMATIC 2
SWITCHED NETWORK PHASE SHIFTER
jX jX

JB

-jX -jX

-jB
DESIGN EQUATIONS OF THE SWITCHED
NETWORK PHASE SHIFTER

X
Defining : X n = , Bn = Z 0 B
Z0
we find the following equations for X n and Bn :

     
X n = tan   , Bn = sin  
 4   2 

where  is the differential phase shift.


DESIGN EXAMPLE

  
 = 90, X n = tan   = tan (22.5) = 0.4142 
 4 
X = X n Z 0 = 0.4142  50 = 20.71
  
Bn = sin   = sin (45) = 0.707 
 2 
Bn 0.707
B= = = 0.01414
Z0 50
jX jX

JB

DESIGN OF THE LOW - PASS CIRCUIT

20.71
f = 10 GHz , X = 20.71 = L  L =
2 1010

L = 3.296 10 −10 = 0.3296 nH

0.01414
B = 0.01414 = C  C = = 0.225 pF
2 1010
-jX -jX

-jB

DESIGN OF THE HIGH - PASS CIRCUIT

1 1
f = 10 GHz , X = 20.71 =  C=
C 2 1010  20.71
C = 0.76886 10 −12 = 0.76886 pF
1 1 −9
B = 0.01414 =  L= = 1.1255 10
L 2 10  0.01414
10

L = 1.1255 nH
SIMULATION OF THE SWITCHED
NETWORK PHASE SHIFTER
MONOLITHIC SWITCHED NETWORK
PHASE SHIFTER
The Field Effect Transistor (FET) is used for the realization of
switches in monolithic microwave circuits. This device is
modeled by means of the resistor RON in the ON state, and by
means of the capacitor COFF in the OFF state. The value of COFF is
typically large. For this reason the effect of the capacitor is often
included in the design of the low-pass and high-pass filters.

RON
COFF
EXAMPLE OF MONOLITHIC SWITCHED NETWORK PHASE SHIFTER
HIGH-PASS FILTER
HIGH-PASS FILTER

C1+ CF1 C2+ CF2

L1L2/(L1+L2)

You might also like