Note: Descriptions are shown in the official language in which they were submitted.
PHASE CONTROL REFLECTOR ELEMENT 12~417
TECHNICAL FIELD
05 This invention relates to a phase control reflector element operable
at microwave frequency.
Phased reflector arrays are useful for a wide range of applications.
They find application in beam shaping and beam steering - ie used in
10 conjunction with a transmitter they may be utilised to vary either
the shape of main beam and sidelobes, or the directi~n of the main
beam. This is attained by selection and variation of the phase
inserted by each array element. They may also be used in beam
selection - ie they may be used to direct radiation incident from one
15 of several selected directions on to a receiver. They also find
application in signal modulation. The phase inserted by each reflec-
tor element may be varied coherently in a time dependent manner to
achieve frequency modulation. Alternatively, reflector elements
capable of lndependent polarisation control may be used in conjunction
20 with an analyser to effect amplitude modulation or gating.
BACKGROUND
A prior art phased array, for frequencies in the range 3 to ô GHz,
25 comprises an array of horn-fed receiving antennae arran8ed back-to-
back with a similar array of transmitting antennae each having a horn
output. Corresponding receiving and transmitting antennae are coupled
in pairs via respective phase-shift networks. This typical transceiver
array is costly, bulky and of appreciable weight. It would have a
~; 30 volume in the region of 1 m3 for example.
.~:
i~ 35
, ~ 1 -
, - _
' ~
'';~
` - 2 - l~S417
There is at present a need for phased arrays operable at higher fre-
quency, especially at microwave frequencies in the 3 to 100 GHz range.
A prior art array is a very unattractive option because of its cost
and bulk.
05
DISCLOS~RE OF THE INVENTION
The invention is intended to provide phase control elements that are
robust, lightweight, compact and relatively inexpensive to manufacture.
These elements and arrays are intended for microwave radiation in the
3 to 100 GHz frequency range.
According to the invention there is provided a phase control element
including:
(1) a dipole,
(2) a substantially lossless dielectric member disposed adjacent
to the dipole and srranged to couple radiation strongly to it,
and
(3) a variable reactance arranged as a substantially lossless
load to the dlpole, whereby radiation incident on the dipole is
reradiated with a pha~e variable in accordance with load reactance
sign and magnitude.
The material of the dielectric member is chosen to have low dielectric
losses in that the microwave power it ab~orbs is small compared with
that coupled to or from the dipole through the dielectric member. The
term "substantially lossless dielectric member" shall be construed
accordingly.
'
'~
~,:
'''
~ 2 -
, I
~':' ''
', ~:.~
1~5417
-- 3 --
An additional resistive contribution to the load impedance results
from the non-ideal properties of the load. Some small resistive
contribution is unavoidable. It is required that as much as possible
of the radiation incident upon the dipole be reflected. Power absorbed
05 by the load will be low, and hence the reflectivity will be high,
provided that either the impedance of the load is comparable in magni-
tude with the impedance of the dipole and the resistive part of the
load impedance is small compared with the reactive part, or that the
impedance of the load is either very high or very low in magnitude
compared with the dipole impedance. In this connection, microwave
theory conventionally treats open and short circuits as extremes of
reactances; the expressions "reactance", "reactive" and similar terms
shall sccordingly be construed to apply inter alia to open and short
circuits.
It is of particular advantage that the dipole and its load may be
constructed in planar form. The dielectric member may occupy a volume
in the region of 10 3 m3 and the dipole and load lO 7 m3, a comb-ination
three orders of magnitude smaller than prior art devices. It is also
an advantage that the dipole couples to radiation substantially only
to one side by virtue of the strongly coupling dielectric member.
This simplifies efficient matching to a microwave field.
The phase control element may be of hybrid construction. The dipole
may be formed of metal deposited upon the surface of a substrate of
insulating dielectric material. The load in this case would comprise
discrete components bonded to form a network shunting the dipole.
-- 3 --
12~?5417
-- 4 --
The phase control element may be of integrated construction, ie the
dipole may be provided with a substrate of substantially lossless
semiconductor material. The substrate may alternatively be a composite
body having 8 surface of such semiconductor material. If the latter,
05 the impedance components may be formed as components integral with the
semiconductor material. Alternatively, the substrate may be of insula-
ting dielectric material, and the phase control element may include in
its construction a supporting layer of semiconductor material, the
dipole being located between the dielectric member and this layer. In
this alternative, heat sinking can be provided without much difficulty.
The semiconductor material layer may be backed with metal, or with a
thin layer of elecerically insulating dielectric material with a metal
coating. This alternative is therefore to be preferred for high power
applications, as in this case efficient heat extraction is important.
The invention exploits the following principle. A variable reactance
shunts the dlpole. This dipole reradlates with unchanged polarisation,
but with a phase shift given by a complex reflectivity Rv:
(GA- GL) - ~(BA+ BL)
V (GA+ GL) + ~BA+ L~
where GA + ~BA is the admittance of the dipole as a radiating source
and GL + ~Bl is the load admittance. RV i8 the voltage reflectivity.
It will be observed that RV has unit modulus as long as the load con-
ductance GL i9 zero. This ldeal case depends on the impedance compo-
nents being lossless and no power absorption occurrlng in the dipole
metal and the dielectric member. The phase shift of the reradiated
signal relative to the incident signal in the general case is:
3~ - arctan ((BA+ BL)/(GA- GL)) - arctan ((BA+ BL)/(GA~ GL))
In the los61ess GL ~ ca~e the phase shift becomes
- 2 arctan ((BA+ BL)/GA)
- 4 -
~ !
1295417
If BL is variable over a range from a large negative to a large posi-
tive value, a phase variation of nearly -~ to ~ can be obtained. This
degree of phase control requires a load to be variable from inductive
to capacitive.
05
Where the phase control element comprises a single dipole, the element
will only couple to radiation having a polarisation component parallel
to the dipole. Power reradiated by this dipole will in turn only be
polarised parallel to the dipole.
The network may include, for example, a plurality of switch-selectable
impedance components, each component comprising the combination of a
reactance and a control switch.
As a further example, the phase control element may comprise a crossed
pair of orthogonal dipoles, one dipole load being either an open cir-
cuit or a short circuit, the other dipole load being an anti-parallel
pair of diodes. In this construction the losd impedance i8 dependent
upon lncident radiation power level. At low levels the load impedance
is high. At high levels, however, the diodes conduct and the load
; impedance is low.
A more versatile embodiment of this invention comprises a crossed pair
of orthogonal dipoles, each having independently controllable loads.
In this construction, each dipole is configured and srrsnged to serve
ss an inductive load shunting the other. This construction allows
separate phase shifts to be spplied for esch of two orthogonal polari-
sations - the polarisation directions parallel to each dipole. Thus
if the incident polarisation has circular polarisation (of either hand)
or else is plane polarised at + 45 to the dipoles, then selection of
the phase shifts for each dipole permits the reradiated polarisation
to be likewise either circular of either hand or plane polarised at
+ 45 - ie polarisation change i8 also possible.
- 5 -
, ~ .
... .
1295417
-- 6 --
Arrays may be constructed incorporating many similar single or crossed
dipoles. A common dielectric member may be used.
05 BRIEF INTRODUCTION TO THE DRAWINGS
Of the drawings accompanying this specification:
Figures 1 and 2 show in plan and cross-section, respectively, a
single-dipole phase control reflector element of
the invention;
Figures 3 and 4 show in plan and cross-section, respectively, a
crossed-dipole phase control element;
Figures 5 and 6 are each plan sections of the control element shown
in figures 3 and 4 above, showing in detail different
control circuit configurations;
Figure 7 is a cross-section of an FM phase modulator compri-
sing a 6ingle cro66ed dipole phase-control element;
Figure 8 is a cross-section of a beam direction control device
incorporating an array of dipoles;
Figures 9 and 10 show, in plan, two alternative constructions of a
crossed-dipole phase control element;
Figure 11 is a cross-section of a duplexed radar including an
array of crossed dipoles each as shown in either one
of the preceding figures 9 and 10;
Figure-12 shows a phase control element incorporating a shorted
transmission line and a varactor diode reactive load;
Figure 13 shows a crossed-dipole phase control element incor-
porating varactor diodes; and
30 Figure 14 is a sectional view of a directionally controllable
transmitter.
:~.
: - 6 -
' ~ ~ `,i,l
, ~ .
, . ... . . . .
417
-- 7 --
DESCRIPTION OF EMBODIMENTS
Embodiments of the invention will now be described, by way of example
only, with reference to the accompanying drawings.
05
Figures 1 and 2 show an example of a single dipole phase control
reflector element 1 of the invention. This element comprises a
single dipole 3 formed of metal deposited upon a substrate 5 of
substantially lossless dielectric material, for ex~mple silicon semi-
conductor material. In this embodiment, the substrate 5 acts both asa dipole support and as a dielectric member for coupling radiation to
the dipole 3. The dipole 3 is divided into two limbs 3a, 3b of equal
or nearly equal length. A local impedance network 7, located in the
vicinity of the centre of the dipole 3 is connected between the two
limbs 3a, 3b. This network 7 includes a shorted transmission line 9
to serve as an inductive load. The network 7 also includes a plurality
of switch-selectable impedance components 11, 13 each of which, in this
example, comprises a capacitor llc, 13c and a PIN-diode switch lls, 13s.
With appropriate values of inductance and capacitance, operation of the
~witches 118, 13s, provides a nett load across the dipole 3 that can be
either lnductive or capacitive. Each of the capacitors 11c and 13c is
or i9 not connected across the dipole 3 according to whether its
corresponding diode switch lls or 13s is short or open circuit respec-
tively. This provides four reactance possibilities selectable by a
two-bit instruction. Control lines 15, 17, 19 are provided for bias
control. Control line 15 is common to both diodes lls and 13s, whilst
lines 17 and 19 are connected one to each diode lls and 13, respec-
tively. Bias voltages voltages applied between controls lines 15 and
17, 15 and 19 switch the diodes 11s and 13s, which in turn connect the
capacitors llc and 13c across the dipole 3. Spurious coupling between
the dipole 3 and the control lines 15, 17 and 19 is minimised by
arran8ing the lines to lie in a direction orthogonal to the dipole 3.
~; 35
- 7 -
,'
129~;417
-- 8 --
Whereas impedance network 7 comprises a fixed inductance with switch-
able capacitors, it is also possible to employ a switchable inductance
with a fixed capacitor.
05 Consideration will now be given to those factors that determine the
choice of length for the dipole 3. At resonance the length "R~" of
the dipole and absolute wavelength A of the radiation are related
by the formula:
Q~ = ~A /~(E1 ~ E2) ~ eff
~ A ~ for ~1 E2 (1)
(See Brewitt-Taylor et al "Planar Antennas on a diectric surface"
Electronics Letters Vol. 17 No. 20 pages 729-731 (October 1981)).
where El and E2 are the dielectric constants of the media each side
of the dipole. For silicon El ~ 12, and air c2 ~ 1. The symbol A
represents the wavelength of radlation measured in the dielectric
substrate medium. Thls formula assumes lowest-mode resonance - so
cslled "half-wavelength" resonance by analogy to resonance in a free-
standing dipole. At this wavelength, the next higher order resonance
corresponds to a length three times this value. The length of dipole
Q is chosen within this range:
Q~ < Q < 3Q~ (2)
The formula (1) given above is theoretical in that it assumes a dipole
length to width aspect ratio approaching infinity. However, this
formula may be considered a reasonable approximation for a dipole with
a 10:1 aspect ratio. The formula can be modified by a simple geometric
factor to take account of dipole shape and aspect in more general cases.
!
1 ~
- 8 -
~!
:
.. . .
9 129~;41~
Attenuation 108s due to the resistivity of the mounting substrate or
dielectric member is given approximately by the ratio (Z/PS) where Z
is the characteristic impedance snd p the sheet resistivity. For a
silicon substrate (Z ~ 100 n) of nominal thickness 400 ~m, a resisti-
05 vity of 100 n.cm corresponds to an attenuation loss of approximately5%, an àcceptable value. The antenna dipole impedance and radiation
polar dia8ram are also sensitive to the substrate resistivity, but
for the dipole described this effect is small for substrate resisti-
vities of 100 n.cm and higher.
The shorted length of transmission line 9 is typically of length
between Aeff/32 and A ff/8, and is therefore inductive.
A more versatile variant of the àbove control element 1 is shown in
plan and section in figures 3 and 4. This element 1 comprises a pair
of crossed orthogonal dipoles 3 and 3' patterned from a common layer
of metal deposited upon the surface of a thin layer 21 of semiconductor
silicon - a layer 21 of thickness typically between A/100 and A/4,
where A is the chosen signal wavelength measured in silicon. A pro-
tective oxide coating 23 is provided between the metal and the silicon,to prevent the formation of undesirable intermetallic compounds. The
silicon layer 21 is backed by a thin coating of beryllia 25 and a metal
coating 27, to facilitate heat sinking. The dipole~ 3 and 3' are
mounted ad~acent to, or ~ust above, the surface of a dielectric member
5 of insulating dielectric material. The dielectric constant of this
insulating material 5 is chosen so that the dipoles couple substan-
tially only to radiation incident via the material 5.
; 30
_ 9 _
~ ~'
- 10 _ 129S41~
Each of the dipole limbs 3a, 3b, 3'a, 3'b has a respective slot 4a,
4b, 4'a, 4'b. Each slotted dipole portion serves as a shorted trans-
mission line such as 9 shunted across a respective dipole limb 3a, 3b,
3'a or 3'b, each limb being approximately A/4 in length. The shorted
05 line length, ie the length of each slot, is less, typically in the
range A/32 to At8, and so each shorted line presents an inductive
load. These parallel inductive loads across ehe dipoles 3 and 3' are
complimented by switch-selectable impedance components 11, 13 and 11',
13'. Each of the switch-selectable impedance components 11, 11', 13
and 13' comprises a capacitor 11c, 11'c, 13c or 13'c and a PI~-diode
switch 11s, 11's, 13s or 13's respectively.
The loaded dipoles 3 and 3' couple independently to their own polari-
sations. The phase-shifts inserted in the re-radiated fields are con-
trolled by the impedance components 11, 13, 11' and 13', and are inde-
pendent.
Consider incident radiation plane polariRed at 45 to the dipoles 3
and 3', inducing in-phase currents. The re-radiated fields are sub-
~ect to phase-shifts ~ and ~ for the horizontal and vertical dipoles
3' snd 3 respectively. If ~ = ~, the resulting radiation is plane
polarised at 45 (ie. parallel to the incident field). If, on the
other hand, ~ = ~ + ~, the re-radiated field is then plane polarised
at -45 (ie orthogonally to the incident field). If ~ /2,
circular polarisation of either hand is re-radiated. In each case,
the re-radiated field is shifted in phase by ~ relative to the incident
field. This demonstrates independent control of phase and polarisation.
;
1 35
:
- 10 _
, ~ . .
11 12g5~17
Control line connection to the PIN-diodes lls, ll's, 13s and 13's
may be made via resistive layer connections. It is also possible to
position low frequency semiconductor devices beneath the antenna metal,
to provide logic functions or drive to the PIN diodes lls, ll's, 13s
05 and 13's. Here electrical power may be ~upplied either through fur-
ther transmission lines or via resistive connections.
Where large amounts of microwave power are to be controlled by the
relay elements, the current supply needed for the PI~ diodes lls to
13's is increased (typically to about 10 mA for a diode capable of
controlling 10 W of microwave power). For the crossed dipoles 3, 3',
it may be inconvenient to supply the current for all the control
diodes through resistive connections because of energy dissipation.
One w~y of avoiding this problem is to rectify a small amount of the
incident microwave power to provide the dc current for the diodes lls
to 13's snd for any logic and drive transistors included. Only low
level control signals then need be supplied through resistive connec-
tions. Schottky bsrrier dlodes are suitsble ss RF to DC power conver-
ters. In the clrcult shown ln flgure 5, a metal line llm ant two
Schottky-barrier rectifying tiotes llr are connected in ~eries across
a dlpole slot 4'a. The diodes llr are coupled to the microwave field
by the line llm and by a capacitor C connected at 10'a to the dipole
limb 3'a. The rectified output of the tiodes llr is fed to the PIN-
tiote lls via a transistor switch llt ant a bias resistor R. A base-
emitter control current is appliet to the transistor llt via resistors12b and 12e. When a strong ratiation field is incident on the antenna,
a microwave voltage is established across the tiote llr ant the conse-
quent rectified current charges the capacitor C This provides control
current for the tiote lls via bias resi6tor R and transistor llt.
Transistor llt amplifies the control current, whlch is therefore small
comparet to the current taken by the tiode lls when in a conducting
state.
- 11 --
.
,: ,. ...... .
1295417
- 12 -
Another way of bringing in DC power and control signals is via metal
tracks - eg track 29 as indicated in figure 6. These metal tracks may
be disposed in various places around the antenna metal 3, 3'. As they
are coupled capacitatively to the antenna metal, they will always divert
05 some antenna current with the result that the required re-radiated
power is disturbed or dissipated to some extent. However the microwave
impedance of the tracks 29 can be raised, at least over a narrow band-
width, by including eg meanders 31 and capacitors 33 as resonant stops.
Increase in the impedance reduces microwave currents in the tracks and
hence causes the loss of efficiency to be reduced.
An FM-phase modulator incorporating a single crossed dipole reflector
3 is illustrated in figure 7. This modulator is comprised of a dielec-
tric lens 41 on the rear surface of which is mounted the crossed dipole
3. The lens 41 includes within its construction a polarisation selec-
tive mirror 43. A transmitter dipole 45 adjoins the side of the lens
41 and, in conjunction with the mirror 43, illuminates the element 1.
~ypically, the crossed dipole 3 has reactive loads comprising a number
of switch-selectable impedances, together with a co-operative logic
function circuit to permit 3-bit phase-shift selection. The crossed
dipole 3 is arranged with its constituent dipoles inclined at 45 to
the plane of polarisation of the incident radiation directed from the
transmitting dipole 45. The load impedances are chosen so that the
re-radiated field is orthogonally polarised. Thus radiation directed
from the phase control element passes through the mirror 43 without
any appteciable reflection occurring. Phase-shifts of 0, ~/4, ~/2,
3%~4, ~, 5%/4, 3n/2, 7~/4 may be selected and inserted under 3-bit
logic control to provide step-wise discrete phase modulation. These
; phase-shifts could be provided at least approximately by three switch-
able diode-capacitor series arrangements (cf 11s/11c in figure 1).
Since phase is not a linear function of capacitance, the foregoing 1/4
phase-shift intervals would not be reproduced exactly. If exact 1/4
phase-shift intervals were to be necessary, seven diode-capacitor com-
binations would be needed with at most one diode conducting at any time.
` 35
- 12 _
~1
~ ,. . .
- 13 - ~2~5417
Arrays may be constructed incorporating many single or crossed dipoles
and utilising a common substrate. The phase inserted at each dipole
site can then be controlled for various applications - eg beam direc-
tion control. An example of such application is shown in figure 8.
05 Here, an array 47 of four single or crossed dipoles 48 has been
arranged on the rear surface of a dielectric lens 49. Radiation is
directed on to the array from a dipole transmitter 45. Microwave power
is re-radiated from the array and focussed into a beam by the lens 49.
The position of the virtual image I of the transmitter dipole 45 may
be varied, and thus beam direction controlled, by appropriate phase
insertion at each of the dipoles 48.
Another form of crossed-dipole phase control element 1 is shown in
figure 9. In this form of construction, the load impedance across one
of the two dipoles 3, 3' is variable by radiation power level, rather
than by the application of bias from an external circuit, as previously
discussed. The polarisation of radiation reflected from this phase
control element 1 differs for high power level snd low power level
radiation. The impedance network 7~ connected between the two consti-
tuent limbs 3a, 3b of one of the dipoles 3, comprises an anti-parallel
pair of diodes 119 and 13s; ie these diodes are connected in parallel
across the gap between the two limbs 3a and 3b, and are arranged so
that the polarity of one of the diodes 11s is the reverse of that of
the other diode 13s. The diodes 11s and 13s may be of the same type,
for example both may be Schottky barrier diodes.
,
- 14 - 1~95417
Alternatively, the diodes lls snd 13s may be of differing types; for
example, one diode lls may be a Schottky barrier diode and the other
diode 13s a PIN diode. When the power level of incident radiation is
low, both diodes 118 and 13s are non-conducting, and the network 7
05 presents a high impedance load to the dipole 3. However, when the
power level of incident radiation is high, both diodes lls and 13s
conduct so the load impedance of the network 7 falls to a low value
compared with the dipole impedance. Hence the phase of radiation
reflected by this dipole 3 differs by approximately ~ for low and
high radiation power levels. The second dipole 3's has an open cir-
cuit load, and is arranged to be orthogonal to the first dipole 3.
At low power level the two dipoles 3, 3' are similarly loaded. Radia-
tion plane polarised at ~/4 to the two dipoles 3, 3' is reflected with
unchanged polarisation. At high power levels, however, the dipole
loads differ and in the ideal situation radiation reflected from one
dipole 3 is ~ out of phase with that reflected from the other dipole
3'. In the practical situation the phase difference will be only
approximately ~. Incident radlation plane polarised parallel to axes
X or Y shown, excites both dipoles 3, 3' equally since the dipoles 3,
3' are oriented at ~/4 or -~/4 to the axes X, Y. The reflected radia-
tion is plane polarised, but parallel to the orthogonal axis Y or X,
respectively, because of the phase-shift.
A variant of this latter form of construction is shown in figure 10.
Here a low impedance load 7', such as a short circuit, is connected
between the limbs 3'a, 3'b of the second dipole 3'. In this case the
reflected radiation is polarised in a direction orthogonal to the inci-
dent radiation at low power levels when the diode impedance is high,
and parallel to the incident radiation when the diode impedance is low.
As is conventional in microwave theory, open and short circuits are
treated and considered as being extreme cases of reactive loads.
;
_ 14 -
- 15 - 12~541~
An array of like crossed dipoles, as in figure 9 or 10, may be utilised
in a radar to couple a transmitter source and one or more receivers
to a common aperture. An example of a duplexed radar is shown in
figure 11. This radar comprises a body of dielectric material 5
05 having a front surface 5a shaped to form a dielectric lens. This
radar also includes an array 1 of crossed dipoles as shown in figure 9,
a receiver Rx, and a transmitter Tx, arranged adjacent to respective
surfaces 5b, 5c and 5d of the dielectric body 5. Surfaces 5c and 5d
are mutually perpendicular, and both are inclined at ~/4 to surface 5b.
The body 5 incorporates an inclined polarisation selective mirror 43.
The mirror 43 is formed by evaporating parallel metal strips on to an
exposed surface (not shown) of the body 5, strip centre spacing being
less than A/4 and strip width being less than interstrip gap width.
Necessarily the body 5 is originally produced in two component parts
(not shown) to allow mirror fabrication prior to assembly. Low-power
level radiation incident upon the surface 5a is focussed on to the
receiver Rx. This radlation is however first converged towards and
reflected at the array of control elements 1, and then reflected a
second time at the polarisation selective mirror 43. The polarisation
of the signal radiation is ùnchanged. The transmitter source Tx is
oriented to launch radiation into the dielectric body 5 with polarisa-
tion such that it can pass through the mirror 43. (The transmitter
output radiation and reflected incident radiation are of mutually
orthogonal polarisation at the mirror 43.) The transmitter output
radiation is of high power level. When the transmitter output radia-
tion is reflected by the array 1 of crossed dipoles each as in figure
9, the polarisation is rotated by ~/2. The outgoing radiation leaving
the surface 5a is therefore polarised parallel to the incoming signal
radiation.
,
_ 15 -
.
- 16 - 12~54i7
A duplexed radar may alternatively be constructed using phase^control
elements 1 as shown in figure 10. In this case either Rx and Tx are
exchanged in position as compared to that shown in figure 11, or
alternatively the polarisation selective mirror 43 is oriented so that
05 its metal strips are perpendicular to their figure 11 direction. The
polarisation of the transmitter ouptut radiation is then unchanged,
whereas the polarisation of the incident signal radiation is changed
upon reflection by the array. As in the previous example, the outgoing
radiation is polarised parallel to the incoming radiation.
Referring now to figure 12, there is shown a further phase control
element S0 of the invention. The dement 50 has two dipole limbs 51a
and 51b connected to respective arms 52a and 52b of a short transmis-
sion line 52. A varactor diode 53 connects dipole limbs 51a and 51b
through the widths of the arms 52a and 52b, and a capacitor 54 termi-
nates the short transmission line 52. A second transmission line 55
having arms 55a and 55b including resistors 56a and 56b is connected to
the ~hort transmi6sion line 52, and provides for DC bias has to be
applled to the varactor 53. Resistors 56a and 56b inhibit microwave
power loss in the line 55.
The figure 12 device operates as follows. The susceptance of the
varactor diode 53 at the microwave frequency depends on the DC bias
voltage across it and also on the magnitude of the microwave voltaRe.
Thus the phase of the radiation reradiated from the element 50 is
controlled by the DC bias voltage across the varactor 53 for the
reasons previously discussed. The phase will depend to some extent on
the magnitude of the incident microwave power because the varactor
susceptance varies with microwave voltage. The phase will be fully
determined by the DC bias under two conditions: either (a) the microwave
voltage is very small as when the phase control element 50 is used in a
microwave receiver or (b) the microwave power level is a fixed quantity
which ls the case when the phase control element 50 i6 used in a trans-
mitter. Thus for practical purpose6, the phase is controlled by the DC
bias voltage across the varactor.
- 16 -
i
.:. .,
1295417
Referring now also to figure 13, there is shown a crossed-dipole phase
control element 60. It is equivalent to a pair of crossed elements 50,
and comprises dipoles 61 and 61' having limbs 61a, 61b, 61'a and 61'b.
These limbs have respective slots 62a, 62b, 62'a and 62'b to provide
05 transmission lines, the latter being terminated by capacitors formed by
overlying patches 63a, 63b, 63'a and 63'b. Four varactor diodes 64a,
64b, 64'a and 64'b are connected between dipole limbs as shown, bridging
the slots 62a, 62b, 62'a and 62'b respectively. The varactor diode
polarities correspond to a bridge rectifier arrangement. Diode bias
connections 65a, 65b, 65'a and 65'b are provided, and include respective
resistors 66a, 66b, 66'a and 66'b for microwave power loss reduction.
The crossed dipole phase control element 60 operates as follows. The
load presented to dipole 61 comprises the terminated transmission lines
formed by slotted dipole limbs 61'a and 61'b, together with varactors
64'a and 64'b. The varactors 64'a and 64'b are preferably equal in the
sense that they have the same dependence of capacitance on voltage. It
19 preferably also arran8ed that the DC bias voltages across varactors
64'a and 64'b are equal. Consequently the microwave currents through
these two varactors will be the same if the microwave voltages across
them are the same. Thus radiation incident on and polarised parallel to
dipole 61 causes currents to flowin ~t, and this will be equally divided
between varactors 64'a and 64'b. No microwave voltage will be developed
across varactors 64a and 64b. Thus, for the reasons previously described
for the circuit of figure 12, the DC bias across the varactors 64'a and
64'b controls the phase of the radiation reradiated from dipole 61
relative to that of the incident radiation. Yaractors 64a and 64b are
preferably also equal, and their DC bias voltages are preferably
arranged to be equal. Thus the DC bias across these varactors controls
the phase of the radiation reradiated from dipole 61', relative to that
for the incident radiation polarised parallel to dipole 61'. If the DC
bias voltages applied to bias connections 65a, 65b, 65'a and 65'b are
respectively V1+ V2, 0, V2 and Vl, the DC voltage is V~ across varactors
64a and 64b and V~ across varactors 64'a and 64'b. Thus application of
bias voltages to these bias connections provides independent control of
the phase of the reradiated radiation for the two polarisations.
_ 17 -
.
,~ .
- 18 - 1 Z 9 ~ 4 1 7
Referring now to figure 14, there is shown a reflecting device 70
arranged for control of the direction of output radiation. The device
70 comprises a multi-element array 71 of four either single or (prefer-
ably) crossed dipole phase control elements 72a to 72d mounted on a
05 planar rear surface 73 of a plano-convex first dielectric lens 74.
The number of elements 72 is not critical. The lens 74 shares a
spherical interface 75 with a concavo-convex second dielectric lens 76
having an outer surface 77. This arrangement provides a composite
lens. If the first and second lens dielectric constants are 1 and ~2
respectively, then 1 is greater than 2 and both are high compared to
that of free space, as will be described. A transmitter 78 is mounted
on a third surface 79 of the first lens 74, and is arranged to irradiate
the array 71 after reflection at a polarisation selective mirror 80.
The dipoles 72 change the radiation polarisation to that transmitted by
the mirror 80. The radiation is refracted at the spherical interface
75 between the lenses 74 and 76. The curvature of the interface 75 is
arranged 80 that each of the dipoles 72a to 72d reflects radiation
incident thereon through a respective region 81a to 81d of the second
lens outer surface 77. The regions 81a to 81d are arranged to be
substantially contiguous as shown. Ray paths 82b and 82c are shown
respectively as arrowed chain and continuous lines for the inner dipoles
72b and 72c. It will be noted that radiation emerging from outer lens
surface 77 is inverted with respect to dipole position in the array 71.
Radiation reflected from the array 71 produces a free space wave front
(not shown) leaving the outer lens surface 77, the wave front direction
being determined by the relative phases of the radiation contributions
traversing the outer lens surface regions 81a to 81d.
- 18 _
,,
lg- lZ95417
Each contribution will have 8 phase comprising a fixed component
determined by that of the output from transmitter 78, and a variable
component determined by the operational state (eg bias condition) of
the corresponding dipole 72. Accordingly, beamforming of the radiation
05 from the outer lens surface 77 may be carried out by appropriate choice
of the dipole loads, eg switching in appropriate capacitors or setting
appropriate varactor bias as described with reference to figures 1 and
12 respectively.
This beamforming technique requries e2 (second lens 76) to be high
compared to that of free space because two conditions governing the
size of regions 81a to 81d are necessary. Firstly, the centre to
centre separation of these regions should be less than A /2 where A
is the radiation free space wavelength. Secondly, the separation
should not be less than the optical resolution provided by the first
and second lenses 74 and 76. This resolution is kA1/2 sin~l, where
k is a number close to 1.2, A1 is the wavelength in the second lens
76, ie A1 = Ao/ ~ , and l i8 the half angle of the cone of converging
radiation illuminating an outer lens surface region 81.
To satisfy both the foregoing conditions, the refractive index n2 of
the dielectric material forming the second lens 76 must exceed n given
by
n O/ 1 k/sin
~ may typically be in the region of 25, in which case n = 2.8 and
n ~ 8; n2 must therefore exceed 2.8 and 2 s n2 must exceed 8. In
addition E1 must be greater than e2 as previously said. These criteria
are not difficult to satisfy in practice at microwave frequencies.
Alumina ceramic has a dielectric constant (c2) ~ 10 and zirconium
titanate stannate (ZTS) has a dielectric constant (c1) of ~ 36, for
example.
- 19 -
- 20 - 12954i7
In order to improve the matching of phase control array 71 to the
combination of lenses 74 and 76, each of the dipoles 72a to 72d may
be provided with a respective small converging lens. Each small lens
may conveniently be inset into the rear surface 73 of the first lens
05 74. The small lenses will be concave or convex according respectively
to whether their lens materials have dielectric constants less or
greater than ~1'
The small or individual phase control element lenses alter the polar
diagrams of their respective dipoles. The composite polar diagram
for the array 71 may accordingly be finely adjusted to a desired
configuration by appropriately varying the individual focussing prop-
erties of the small lenses. Inclusion of these lenses provides an
extra degree of freedom for optimising the phase control array beam
configuration. Optical design to achieve this is well-known in the
art of optics and will not be described in detail.
- 20 -
.
