Note: Descriptions are shown in the official language in which they were submitted.
1~37SS~ `
,. ` ~
, . .
Back round of the Invention
g
The pres~nt invention relates in general to a device
for the accurate and rapid measurement of complex reflection
coefficients hy the measurement o~ power levels at two
detectors. The device described herein is constructed in
a compact arrangement in either stripline or waveguide.
The device of this invention provides for the accurate and
rapid measurement of complex reflection coefficients and,
more particularly, the accurate measurements of low voltage
standing wave ratios over waveguide bandwidths, or with regard
to a stripline construction, over multi-octave bandwidths.
With regard to the measurement of VSWR probably the most
common approach is to use a directional co~lpler. Ilowever, in
order to provide an accurate measurement, in particular of
low VSWR, there is the requirement of a precise mechanical
construction of the directional coupler. Therefore, such
devices have to be made within quite tight tolerances.
Accordingly, one ohject of the present invention is to
provide a device that makes either complex reflection coef-
ficient measurements or low VSWR measureme~ts without the needof a coupler or the like that requires tight construction
tolerances. ~ '
One of the oldest techniques for the measurement of
complex reflection coefficients ~both amplitude and phase)
is the slotted line apparatus. In this apparatus, the measure-
ment is accomplished by means of a mechanical procedure ~hich
is quite time consuming and not at all suitable for computer-
ization of the measured data.
Accordingly, another object of the present invention is
to provide a device which will allow for the rapid computerized
evaluation of data. In accordance with the invention this is
accomplished by means of direct measurement of two power
levels. These power levels are then operated upon in accord-
ance with the theory of this invention to provide readings of
complex reflection coefficient or ~SWR.
,, I
,,
, . , . . ~ .
. . ~ . ,.
.~_.... ... .
,
, - ' ' '
~L37SS4
-3-
In another technique, the reflection coefficient is
obtainable from the power measured at three fixed probes
spaced one-eighth wavelength apart along a slotted line.
See, for example, the article by ~. J. Duffin, "Three Probe
~lethod of Impedance ~leasurement", Wireless Éngineer, Vol. 29,
pp. 317-320 (Dec. 1952). However, with this technique, the
reflection coefficient is obtained with a high de~ree ofaccuracy
only over a narrow frequency range for which the one-eighth
wavelength condition is substantlally satisfied.
Accordingly, a further object of this invention is to
provide a device for measuring the complex reflection co-
efficient accurately by the simple measurement of power
levels over a full waveguide band~idth, or in the case of
stripline or coax over a multi-octave bandwidth.
One prior art article by l~enry J. Riblet "A Swept-
Frequency 3-Centimeter Impedance Indicator", Proceedings of
the I.R.E., Vol. 36, pp. 1493-1499, Dec. 1948 describes a
device for determining the complex reflection coefficient in
waveguide from the powers coupled to two detectors. However,
with this technique, the measurement accuracy is very sensi-
tive to the actual phase of the reflected signal. ~`
Thus, another object of the present invention is to provide
a device for the measurement of reflection coéfficients
wherein the measurement -accuracy is essentially independent of
both amplitude and phase provided that the VSI~R is less than 2/1.
An article by ~lenry L. Bachman, "A Waveguide Impedance ~
Meter for the Automatic Display of Complex Reflection Co- _
efficients", IF~F, Trans. on MTT, Vol. 3, pp. 27-30, Jan. 1955
describes a waveguide impedance meter for the automatic display Y
30 of complex re~lection coefficient~ l-lowever, the circuit ~
configuration requires many waveguide componcnts. Further- _
more, it assumes for its operation a frequency independent
90 phase shifter for which no practical realization exists. ~t
Also, the actual bandwidth is limited to 10%. In another
. :
'' '' ~ ,.'
- :
, '
1~3~SS4
-- 4 --
article by Glenn F. Engen, "An Improved Circuit for Implementing
the Six-Port Technique of Microwave Measurements", IEEE Trans.
on MTT, Vol. MTT-25, pp. 1080-1083, Dec. 1977 there is described
a number of circuits for the accurate determination of-the
reflection coefficient over wide band-widths from measurements
of power levels at four detectors. Again, however, these
circuits require many components and an elaborate calibration
procedure must be performed before measurement~ can be taken.
Accordingly, another object of the present invention
is to provide a device for the measurement of a reflection
coefficient wherein only a simple calibration procedure is
necessary and wherein the calculation of the coefficient can
be accomplished by relatively simple formulas.
In accordance with one aspect of the invention, there
is provided an apparatus for measuring the reflection coefficient
of a device by detecting normalized power levels, comprising
main coupling means having means for receiving a test signal and
means for coupling to the device being measured, first and
second secondary coupling means associated with the main coupling
means, and first and second detectors associated, respectively,
with the first and second secondary coupling means. The
apparatus further includes means coupling each secondary coupling
means to the main coupling means with the normalized power
- levels Pl and P2 respectively associated with the first and
second detectors being provided in accordance with the following
equations:
Pl~l+lrl2+21rlcOs~
P2 = 1 + I rl2 + 2 Irlsin~
where ¦ rl is the magnitude of the reflection coefficient and
t h
lS 1 s p ase.
~ .
4 ~,~
'''
1137554
- 4A -
According to another aspect of the invention, there
is provided a ~our port apparatus for measuring the reflection
coefficient of a device by detecting power levels, comprising
main coupling means having means de~ining a first port for
receiving a test signal and means defining a second port for
coupling to the device being measured, and first and second
secondary coupling means having means defining third and fourth
ports, respectively of the apparatus. First and second
detectors are connected respectively at the third and fourth
ports. The apparatus further includes means coupling each
secondary coupling means to the main coupling means such that
when feeding a signal into one secondary port the signals
from the first and second ports are the same and when feeding
the other secondary port the signals from the first and
second ports differ only in a phase difference of 90..
1~37SS~
Brief l)escri~tioll of the Draw,in~s
Numerous other objects, features and advantages of the
invention should now become apparent upon a reading of the
following description of this invention taken in conjunction
with the accompanying drawings in which:
FIG. l schematically depicts the apparatus of the
present invention for measuring ref'lection coefficient;
- FIG. 2 shows a diagram in the complex gamma plane useful
in determining the reflection coefficient;
FIG. 2A shows the gamma plane diagram as~ociated with a
prior technique;
FI~. 3 schematically depicts a coupling structure in
accordance with the present invention including a main arm
and one secondary arm;
FIG. 4 is a perspective view of an apparatus of the ~'
present invention constructed in waveguide; r
FIGS. 5-7 show one specific waveguide version of the
present invention;
FIGS. 8 and 9 are cross-sectional views showing an
20 alternate waveguide version in accordance with the invention; ~.
FIG. 10 schematically depicts a stripline version of the
present invention showing the connections to the two r~ ;
detectors, the device being tested and the signal source;
FIG. llA is a plan view of a stripline version of the
', 25 present invention showing the circuit etching; ~
FIG. llB is a perspective view that is partially cut ~'
away showing the device constructed in stripline including
all connectors;
FIGS. 12 and 13 show still further embodiments of the
invention in the form of an optical apparatus for measuring
reflection coefficient. L_
_
.-,, _ .
,
.
: : .
1137554
-6
, _ .
Exposition
FIr.. 1 schematically depicts the hasic circuit configura-
tion in accordance with the presen~ ;nvention including a
signal source 10, detectors 11 and 12, reference detector 14,
and the test device 16. All of these components are asso-
ciated with the basic device of this invention disclosed in
box 18 and shown in more detail hereinafter in different t
embodiments including'a waveguide version and a stripline
version of the invention. For the waveguide version the
reference detector 14 is preferahly used with a directional
coupler to monitor the incident power to device 1~. The
reference detector 14 may be of conventional design as may be
the detectors 11 and 12. As is apparent from FIG. 1 the
device 18 is basically a four-port device with the `ports
numbered 1, 2, 3 and 4 as shown in FIG. 1. ~-'
It is a purpose of the invention to determine the r
parameters of the device enabling power to be coupled to
detectors 11 and 12 in a way that the reflection coefficient
r is unambiguously determined from the powers measuled at the
' 20 detectors at ports 3 and 4. In'accordance with the invention
the accuracy of measurement is insensitive to the actual ampli- ~l
tude and phase of the reflection coefficient provided that the L_
VSWR is less than 2/1. The VSWR of most com~onents being p
tested satisfy this condition.
An article by Glenn F. ~ngen, "The Six-Port Reflectometer: r
An Alternate Network Analyser", IF.EE Trans. on ~TT, Vol. ~ITT-25, ~,
~p. i075-1080, Dec. 1977 discloses a diagram in the complex r k
plane. FI~. ~ herein shows a complex r plane diagram. At a
particular frequency detectors 11 and 12 (see PI~l. 1) are $
30 represented by points ql ancl qz in the complex plane. The L
measured power ratios Pl=Pl/PR and Y2 P2/PR~ corre p g
respectively to detectors 11 and 12, determine circles of
known radius abou~ points ql and q2. It is the intersection
between these circles that determines the complex reflection ~~
35 coefficient r. In FI~. 2 the point ql is located at the
,,.", ~,.
~ ~ , .
, . .. . ... . .
~3~5S4
,~ 7
coordinate -1 on the x axis while the point ~2 is located at
the coordinate -j on the jy axis. A reflection coefficient
corresponding to VSlYR less than 2/1 falIs within the shaded
circle C shown in FI~l. 2. From FI~. 2 it can be seen that for
5 such a reflection coefficient the circles ahout points ql and
q2 intersect nearly at right angles and thus the determination
of r is unambiguous and the measurement accuracy is substan-
tially insensitive to the phase of the incident signal.
In the previous Riblet article, su~ra, the theory of ~ ;
10 operation de~eloped differently from the theory described
herein-. In this connection reference is no~ made to FI~. 2A
which shows the points ql and q2 both on the jy axis at +jy
and -jy. The equations used in connection with the diagram
of FIG. 2A do not permit the determination of both r and ~
15 unambiguously and accurately. In this regard there had been ~;
an assumption in the equations wherein the second equation r
such as our equation (5) used a -cos9 term instead of our +sinO
term. From FIG. 2A it is noted that there are two points of
intersection within the unit circle C which points are symetric-
20 al about the jy axis. With this criteria of FIG. 2A any changes
in the radius of the circles shown in dotted lines represent a
substantial change in the magnitude of r. With reference to t-
FIG. 2A it is noted that there are two intersections between F
the circles. This means that there are essentially two dif-
25 ferent phase values thus making the determination ambiguous.
Thus, the arrangement in the previous Riblet article was
sensitivé to the phase of the incident signal. ~
In connection with the present invention and the diagram ~r
of FIG. 2, along with the schematic diagram of FIG. 1, if ~
30 the amplitude of the incident signal from source 10 is .
represented by the parameter ¦A I , then: _
PREF = ICREFI2 ¦AI2 . (1)
P~ = Ic l 2 ~ 2 I r q l 2 (2)
P = ¦C 12 IAI I r q 12 (3) _
.
,i,"............................ . . . ..
. - ,
. . . . . .
:... . : .
,. - . . . , ~
.. . ~ .
-8-
-
where PREF, Pl and P2 are the powers measured at the
reference detector 14, detector 11, and detector 12, respect- '
- , ively, while ¦CREF¦2, ¦C 12 and ¦C ¦2 are coupling factors.
The coupling factors ~RE~ ¦ Z ~ ¦ C ¦ 2 and ¦C ¦ 2 may be equated
because any difference in these factors can be referenced o,ut
in a computer or microprocessor based measurement system by
taking a reference reading with a load on the output. Neverthe-
less, it is desirable in practice to have these coupling factors
' approximately equal. IJnder this restriction and by emploYin~
known mathem~tical formulas the following e~uations can be
derived from equations (1), (2) and (3).
1 1/ REF ¦r+l¦2 = 1 + Ir12 +2 IrI
P2= P2/PR~ = Ir+jl2 = 1 + Irl2 +2 Irl~in~ (5) 1-
Equations (4) and (5) are derived by assuming that the points
in FI~. 2 are ql = -1 and q2 = -i as before. Equations (4)
and (5) are in a sense the basic equations for determining r
¦ r ¦ and ~ in terms of measured power ratios Pl/PREF, P2/PREF. j
These power ratios may be normalized to normalized powers Pl L
and P2 . Thus, the, magnitude of r iS given by__he expression: t
Irl2 = PllPz ~(PllP2)2 (Pl-1)2 (P ~ )2 (6) ~ '
Equations (4) and (5) have been derived by assuming the
values of ql and q2 defined above. This allows for the unam-
biguous determination of r for the case of small VSWR which is
' 25 of most practical interest. Now, with this initial determina-
tion a circuit is now constructed to yield these parameters., ,
In order to,accomplish this it is best to express the powers ~
Pl and P2 in terms of the S matrix'elements of the network ~ ' :
shown in ~IG. 1. The following S matrix relationships apply: , ~:
~ = {S + s s r} {S * + s * s * r*
~ 13 23 12 13 23 12 . _
~ . . .
P = {S + s s r} {S * + s * s * r*} . ~,. .
2 14 24 12 14 24 12
!
-r ~_
,, .~ .. ~ .. .
' ' '
.~ ' ' ' '~
:
: . , . : .
. .h ~ , .
, ' .
~ ' ,
1~37554
. .
Therefore
~,= lsl3J2 + ls23l2 lsl2rl2 ~ (S~3S23*s~2*r* ~ Sl,*S~Sl2r) ,
F I IS .l2 + IS l2 lSI2rl2 + tS,4S24*S,2*r* + S~4*S~4SI2r)
l~ith the above S matrix relationships the subscripts for the
S matrix elements correspond to the numbered ports shown in
FIG. 1. Thus, for example, the term S13 refers to coupling
between the input port and detector 11. If the relationshipq
of equations (1) - (5) are a~plied to the above S matrix
relationships then;
,S13 S23 (7)
S14 = ej S24 (8)
S12 = 1 (9)
A further condition which in practice need not hold e~actly
is ¦S ¦ = ¦S ¦. From the reciprocity principle equation (7)
denotes that the signal divides equally out of ports 1 and 2
when feeding port 3, and further equation t8) denotes that
signals from ports 1 and 2 should differ only in a phase
difference of 90 when feeding port 4.
'' The condition represented in equation (9) is impossible
to satisfy exactly. However, an approximate condition can be
derived by making the coupling to detectors ll and 12 weak
(30dB or so). Under that condition the equations (4) and (5)
are satisfied to a high degree of accuracy if one remembers
that ~ also includes the phase shift due to S . This phase
term may be removed by a,n appropriate shift of the reference
plane. The weak coupling condition which can ~e statcd as
lS,2l < 1 (ln)
has some practical advanta~es. For a ty~ical input power of
~lOdBm and 30dB coupling, the power to the detectors will be
-20dBm. This is large enough so that the si~nal won't be
noisy but not so large as to be in the highly non-linear
region of the detectors. Furthermore, because of the weak c~up-
ling, the errors introduced by detector mismatches and imperfectcomponents are minimized.
~f~
:
.' ,, ' ` :
,
.
~ . - .
113~S54
- 10-
Waveguide Embodiment
FIG. 3 schematically depicts a representation of the ! '
device 18 of FI~. 1 showing the main ports 1 and 2 as
identified in FrG. 1 and also showing ports 3 and 3' which
5 may, for example, be associated with port 3 coupled to
detector' 11 in PIG. 1. This schematic diagram of FIG. 3
may be considered as including waveguide sections 22 and 24.
Equations ~7) and (8) can be relatively simply satisfied
with this representativé structure by coupling the ~ain arm
10 22 to the secondary arm 24 throu~h an appropriately positia~ed
hole 26 schematically represented in FIG. 3. The desired
30 dB coupling is readily obtainable with such a single hole.
In the specific version discussed hereina-Çter in FIGS. 4-7
the secondary arm for detector 11 is located on the top wall of
15 main arm 22. In this case there are provided a pair o~ ~
coupling holes, one for each detector. As descri~ed herein- r
after, the unused secondary arm ports are terminated in
loads.
A condition of equations (7) and (8) is that the power
20 divides equally out of the main arm ports 1 and 2 when feeding L
either detector port. By applying the Bethe small hole ~
coupling theory, it can be determined under what circumstances L
this condition is satisfied preferably independent of frequency.
In fact, there are'two situations which lead to equal power
25 division independent of frequency: ~t
Case I: ~= 90 (cos~=o); d or d' = a/2 ~refer to figure 3)
i.e. the waveguides are at right angles and the hole is centered ~
in one of the waveguides. ' _'
Case II: 9~90; d = o, d' '= a/2 or d = a/2, d' = o 'v
30 i e. the waveguides'are not at right angles but the hole is .
centered in one waveguide and against the wall of the other.
The condition represented by equation (8) cannot be _
satisfied with Case II and thus this case is considered to ' ~'
hot be u'seful. The condition of equation t7) is possible _
35 to satisfy with Case I. This is accomplished by centering
..... .
~r
' ";'
. '
.
~ ,
'' '
.
~, . . . . .
'
, 1~37S54 1'
- 1 1 -
the hole in one of the secondary arms with this hole also
being off-center in the main arm. For example, in FIGS. 8 and
'` 9,'reference is made to hole 26A which i5 centered'in the
secondary arm 24~ but off-center in the main arm 22 as denoted
in FIG. 9. From symmetry and reciprocity it is known that
S13=S31=S32=S23. Under Case I equation (8) is somewhat more
difficult to satisfy. The other hole, such,as the hole 26B
shown in FIGS. 8 and 9 is off,center in the secondary arm, '
such as arm 2.4B, and is centered in the main arm. l~ithin a
common proportionality factor'the S matrix elements S14, S24
are given by the Bethe theory as: ~
14 ae ~2 sin a + i g ~m(l) cos~d (11) ~,
24 ae ~ sin a - j ~a am(l)cos~d (12) r
where Ag is the guide wavelength, ~ is the wavelength, _ is
the waveguide width, d is the distance of the center of-the
~hole from the waveguide wall in the secondary arm, ae and
am(l) are electric and magnetic polarizeabilities of the hole.
It is apparent from equations (11) and (12) that the phase ~ -
of S24 witll respect to S14 may be varied by varying the ~:
distance d of the hole from the wall of the secondary arm.
The phase difference is'given by
~= 2 Tan 1 aA m cot ~d
= 2 Tan ~ A A tAC a ~e a (13)
where A is the cut-off wavelength of the waveguide. In ,
the'followi,ng it is assumed that circular holes are used in
which case am(l~ = ae. These holes give the largest coupling
for a given value of ~. If ~ = 90 then
d = a~ Tan~l A~ (14)
J
~r
...... . ,~.. . ... ... . . . .
~375S4
-
-llA-
~his formula determines the required hole location in terms
- of a and ~. For a lYR-90 waveguide at a center frequency of
9.375 GHz, the dimension d = .225". Ideally, the phase
difference O should be 90 and independent of frequency
S (wavelength). From equation (13) this will not be the
case in practice, hol~ever, it l~ill track 90 fairly closely
over a waveguide band. It can readily be shown that dO = 0
- for some frequency within the waveguide band.
It follows from equation (13)
'~
.' . .
. ~
: ;,
~..
. L
. ~,
.
;,...... -
,;. . .. .
,. . .
, .
: ', '' - '
: . ..
.... .
` 1:137554
-12-
that d~ = O if d(A d~ )= Thus,
d(~2-~4!Ac )= 2~ ~4~ /Ac = -- > ~ c/
For a l~'R90 waveguide this will occur at 9.273 GHz. The vari-
ation in ~ is expected to be small over the waveguide band
of 8.2 to 12.4 GHz.
~ IG. 4 is~a perspective view showing the basic outer
construction of the device of this invention including
the main arm-22 and a pair of secondary arms 24A 'and 24B
which are both disposed at a 90 angle to the main arm 22.
All of these arms may include conventional flanges so that
the device can be connected to other means such as the test
device, the detectors and the input signal source. For the
sake of simplicity these components are not shown in detail
but are referred to in the schematic diagram of FIG. 1. The
main guide arm 22 has flanges at both ends while each of the
secondary arms has a flange only at one end. The closed ends
of the secondary arms are terminated by means of a load 30 L
depicted in the drawings.
The embodiment ,just previously discussed is the one shown
in ~IGS. 8 and 9 wherein the coupling hole is centered in one
secondary arm and off-center in the main arm and is off-
center in the other secondary arm and centered in the main
arm. In FIGS. 8 and 9 the holes Z6A and 26B are shown in ~
their respective walls 32 and 34. As previously indicated r
this embodiment may be constructed in IVR-90 waveguide or could ~.
be constructe'd in other waveguide bands. The holes 2,6A and i
26B may have a"diameter of 0.312" with a wall thickness (walls ,
32, 34) of 0.030" providing a coupling of 30clB + ldB for
the band o~ 8.2-12.4 GHz. In order to achieve nearly the same
coupling to both detectors the same hole size was used in '`.
both secondary arms but the hole was symmetrically located in ~: '
the secondary arm with regard to the top hole and sy~metrically _
located in the main arm with regard to the bottom hole 26B.
~ .
. .. , , , , '' - '; ~.! . , ' ~
.
' ' '
.~ , , .
I ,.~....
,
, 1~37SS4 I:
_~ -13-
Generally, the Bethe theory applies only in the limit of very
small holes. Even at a coupling level of 30dB significant
deviations' from the Bethe theory have been found experimentaliy. ~,
In order to re-establish the equal power division requirement
5 for the non-symmetrical arm and to improve the input match, it ! -
was found to be useful to put a capa,citive button 36 on a '-
- ~ septum 40 facing the bottom hole 26B as indicated in FIG. 8.
With the use of this matching~structure the input VSWR was ii.
less~than 1.07 and the co~upling unbalance less than .2dB ' `'
for the range of 8.2-12.4GHz. r-
In the complex r plane of FIG. 2, the centers for the
concentric circles shown as dotted lines have been at points
and q2 = ~j These points may actually be chosen at
any location along the unit circl~e shown in solid in FIG. 2 - ~,
provided that the points are s~eparated by 90. Because of ' ~:
this~generalization equations~ 7) and ~9) may thus be written r
I in a more general sense as: j
3 e 523 (15)
~ 20 S; = ~e~ l~/2+~}S24~ ~ ~ ( 16) ,~ ~,
,~ where ~is a reference~phase angle. The conditions set forth
, ~ in equations (15) and~(16) lea~to relatively minor modifications ~
of the basic equations (4) and~(5). Thus, in another exampIè r~:
'in accordance with the lnvention the points on the diagram of
25 FIG. 2 may be seIe~cted so that ql = ' ~ and q2 = ~ . For :
this condition reference is now made to the embodiment shown
in FIGS. 5-7 wherein both of the holes 26A and 26B are centered
in the main waveguide arm 22 but are off-center in the secondary 4;
waveguide arms 24A and 24B so that the phase relationship ' ~,~
, 30 of equations (15) and (16) apply. In the embodiment of
FIGS. 5-7 the hole diameter and wall thickness may be the same
,7, .... .
as previously discussed with regard to FIGS. 8 and 9. The
hole,may have,a diameter of .312" while the wall thickness
may be .030~. The distances d and d~ depicted in FIG. 6 r
35 can be found approximately by solving equation (13) with
,.5`, !
;;~ :
: L
~ .
. .
~: , ' . " ': . ,
:. : , . , , ' :
''' , ' ' ' ' . : .
--
' '
- ~ :
. ~ , . - : .
`,':~... ," , ,
~37ss4
. . , q
~ = 45 and -45 for a WR90 waveguide at a center frequency
of 9.375GHz. The dimensions d and d' are d = .337" and d' =
.563". For the embodiment of ~I~S 5-1 it is noted that the
device has four-fold symmetry. With this embodiment there may
S also be provided a pair of matching ~uttons 36A, 36B on a
septum 40 as depicted, for example in FIG. 6. In the two
embodiments described in FIGS. 5-9 circular holes have been
shown. However, elli~tical shaped coupling holes could also
be provided although there is little to be gained by using the
10 elliptical hole in place of the circular hole. ~-
Stripline Embodiment
IYith règard to the construction of a compact broadband
stripline network, FIG. 10 is a schematic diagram showing a
15 stripline circuit; FIG~S. llA and llB depict an actual device ~-
constructed in stripline and embodying the principles of
this invention. In the interpretation of equations (7) and
(8) the reciprocity principle may be applied so that S31=
513~ S32 S23, S41 S14 and S42=S24- By application of this
20 reciprocity principle to equations (7) and (8) and with .~
- reference to FIG. 10 and the ~ort numbering denoted therein k
power at port 3 divides equally with identical phase out of L
ports 1 and 2. This suggests the use of a broadband power -
divider 51 at port 3. Also, power at port 4 divides equally ri
25 between ports 1 and 2 with a 90 phase difference between the "
signals out of ports 1 and 2. This suggests the use of a
broadband 90 hybrid 54 at port 4. Furthermore, means must
be provided to couple weakly to the main line extending
between ports 1 and 2 so that the weak coupling condition
30 o equation tlO)is satisfied. This may be accomplished h
using a broad~and coupled line couplers 50, 53, 56, 57 and 58 ~
4 as noted in FIG. 10. The first coupler 50 has 30dB coupling L
and is used in association with the reference detector 14. Of
course, each of the coupling lines is also terminated by means
35 of a load L as also shown in FIG. 10. The first power level
~ . .
. , :
.: ' '
, ' ...... . .
,
, r : .
375S4
:` - 1 s -
detector 11 couples by way o~ the power divider 51 to cou~lers
56 and 57. ~s denoted these couplers couple to the main line
with 27dB coupling. In fact, t]le coupling to the line 1-2, is L
by this loose coupling other than at the coupler S0. Thus,
5 the second detector 12 copules to the main coupling line 1-2
by means of the hybrid 54. The couplers 53 and 58 couple the
signals from the main line to the hybrid 54. ~fulti-octave
couplers with this type of coupling level are readily constructed. L
A diagram of a planar stripline device, as far as the etched F
10 circuit is concerned, is shown in FI~. 11. This device covers
the 2-8G}-lz band. If single step coupled line couplers are
used, then in theory a copuling of 27dB+3dB and 30dB+3dB can '
be achieved over this frequency range. I
It may be desirable to have the hyhrid 54 shown in FIG. ~,
15 10 external to the basic coupling structure as indicated by F
the solid line 55. This permits flexibility of choosing a
high quality hybrid appropriate to a na`rrower frequency range
than the operating range of, for example, 2-8GI3z. The quality
of the hybrid 54 largely determines how well the equations
20 (4) and (5) are approximated. The two pairs of 27dB coupl'ers
on each side of the main line 1-2 may be arranged to ~ace each
other rather than in the staggered manner shown in FIG. 10. ~_
In the waveguide embodiment discussed previouslv and with P
reference to FIG. 2 the reference points ql and q2 may be ~r
25 arranged at different locations on the unit circle but always
separated by 90. However, if ¦S ¦<1, then equation (10) ~
no longer holds and the points ql and q2 mav lie outside of ~-
the Ullit circle. This can be useful under some circumstances. ~
For example, if the directional couplers as shown in FIG. 10 ~r,
30 have relatively tight coupling such as 6dB coupling then ql =
-2 and q2 = -2j. This means that the line between points
and q2 will not intersect the unit circle. Thus, there is no --~
'ambiguity in the determination of reflection coefficients for
which ¦ r ¦~ or = 1, which is the case for passive devices. A r
35 similar effect can be achieved by inserting a 6dB attenuator
_
.z, . .
.
- . . . . ~ . .
~ ~37S54
-16-
,~ . .
pad in the main line 1-2 at the symmetry axis connecting
ports 3 and 4 (the coupling values noted in FIG. 10 remain
the same~. With this arrangement the only additional require~
.' - ment is tha.t an additional reference measurement of the
transmitted power has to be made before measurements can be
made.
FIGS. llA and llB show a device constructed in accord-
ance with the schematic diagram of FIG. 10. This device' ~,
,- ' ' includes:a printed circuit'board 60 having the coupler circuits, r1~ the main coupling line and the divider e~ched thereon prefer-
ably in copper. The drawing shows the various conductors 62
that form the stripline'circuit. In FIGS. llA and llB like
. ref,erence characters are used as previously identified in
. .FIG. 10 to identify the different circuits disclosed. 'In
addition to the printed circuit board 60 there is also provided ~.
a plane dielectric board 66 with both boards being sandwiched
between ground planes 68 and 70. FIG. llB also shows connect- r
ors C associated with the circuit such as at both ports 1 and
2 of the main coupling line. These connectors enable
connection to the detectors, the input signal source and the
device under test.
Conducting walls ~not shown) may be placed between the ~_
circuit elements,shown in FIG. llB. These,conducting walls ~.
are used for preventing the propagation of higher mo,des in
25 the.apparatus. For example, a conducting wall may extend ~
perpendicular to the main line between couplers such as ~'
couplers 50 and 53 shown in FIG. 11~. Also, walls may extend t:'
between couplers 53 and 58 and 56 and 57. Furthermore,
there may be a conductor wall wi.thin, for example, couplcr
30 50 again extending.in a direction perpendicu,lar to the main ~,~
1 ine .
~, In an alternate emhodiment of the present invention _
the apparatus may be,constructed in coax including a solid ' ''
mass that may be machined to receive a center conductor.
The conducting strips shown in.FIG. llB'would become solid:-
~,,~, 1
.' ~"; .
..~
,: .
~ L3755~:
-16A-
center conductors suspended in troughs or grooves machined
in a solid piece of metal. l~ith such a coax construction ,
there is inherently no or very little problem of high mode ''
propagation. t
The use of an attenuator has previously been discussed.
FIG. llB shows the attenuator A in the main coupling line.'
This attenuator may be of conventional design having leads
' , coupling to the main line 1-2 and including upper and lower , -' f
conductors for conductivelylcontacting the respective ground t'
planes. Similarly, the boards 60 and 67 may be drilled to
receive the loads L, four of which are used in this embodi-
ment. The load L also has a single lead connecting,to the
copper conductor or etching 62.
FIGS. 12 and 13 show an optical device in accordance
~J
with the present invention in two different embodiments. At
frequencies above the,microwave range it is common to use r
quasi-optical techniques. At sufficiently high frequencies
directional couplers may be constructed usîng partia]ly
reflecting mirrors, thin dielectric sheets, or closely spaced
' 20 ' ~^,' '
.. .
----
~; ,.
~i-
,, . .
L
,' ' ' ' '` ''~'.
r
.,
~,,", ~
. i
. . ............ . - .
.
1~l37554
- 17 -
prisms. The coupled line couplers of FIG. 10 may be replaced
by optical couplers as depicted in FIGS. 12 and 13. FIG. 1
shows a diagram of an optical network which uses partially
reflecting mirrors as couplers. In the embodiment of FIG. 12 in
comparison to the one of FIG. 10, the power divider 51 at
port 3 is replaced with a beam splitter66 along with ~/4 path
difference to the two outer mirrors, while the hydrid 54 at port
4 is replaced with a second beam splitter 64. As a result, the
ideal equations only apply at that frequency for which the path
difference is in face ~/4. However, these equations can be
modifiedto give accurate results over a bandwitdth of at least
an octave provided that the actual frequency is known. If the
reflected amplitude and phase varies over the width of the bèam,
then this variation could be recorded by replacing the photo
detectors 70 and 72 with photographic film~ In effect the
reflected amplitude and phase over a two dimensional field of
view will be mapped uniquely onto 2 two-dimensional intensity
fields in such a way that one can be constructed from the other
using simple mathematical formulii. When one speaks of light
waves, one must be concerned about polarization. To completely
characterize the reflected signal, one must know the reflected
amplitude and phase for two independent polarizations (a total
of four quantities). This additional information may be
obtained with a ~imple modification of FIG. 12 as provided in
FIG. 13. In FIG. 13 additional photodetectors 74 and 76 have
been added opposite photodetectors 70 and 72, respectively,
givi~g a symmetrical arrangement. One sort of polarizer is
located in front of detectors 70 and 72 and the other sort
in front of detectors 74 and 76. These polarizers are identified
as polarizers 70', 72', 74' and 76', respectively. In this
arrangement detector 70 and 72 determine r for one kind of
polarization and detectors 74 and 76 determine r for the other
opposite polarization.
, ~ , .
... .
