Kraus Helix Idea
Kraus Helix Idea
Kraus Helix Idea
263
peaks. If the speech were derived from a local micro- CONCLUSIONS
phone, a fixed bias might suffice for this. If a distant In closing, the following comments be made
microphone were employed, however, the bias would on the data which have been presented: may
have to be derived automatically, since the polarity of (1) The computed swing ratios tend to be equal to or
the speech wave would be uncertain. Then, also, the higher than the experimentally determined values for
frequencies on the line are assumed to be biased by phase modulation, and lower than the experimental
shifting the carrier frequencies, so as to center the values for frequency modulation. This is probably
peaks of the instantaneous frequencies in the band. largely because the computation was based on rms
Physical means for obtaining these biases could doubt- speech voltages, while the experimental method was
less be devised, if desired. based on measured peak speech voltages.
IV. SUMMARY OF RESULTS
(2) Of the four methods of volume regulation as-
sumed in analyzing the experimental data, the method
Table II gives the swing ratios for the several kinds of regulating to constant voltage peaks applied to the
of automatic level adjustment described above, derived phase modulator results in the greatest swing ratios,
as described from ratios of speech peaks which will be and therefore appears the most efficient in regard to
exceeded by 10 per cent of the voices. The computed use of the modulator. This is of practical importance,
swing ratio is included in the table for comparison. since the greater the efficiency, the higher the carrier
frequency at which the modulator may operate, and
TABLE II the fewer the required number of stages of subsequent
SWING RATIOS FOR SPEECH frequency multiplication.
(3) The 3,000-cps low-pass filter had little effect on
No 3,000-cps the swing ratios, except in the case of the moving-coil
Transmitter Type of Regulation Filter Filter microphone and phase modulation.
PM FM PM FM (4) The moving-coil microphone without any filter
Fl Computed in Section 2 1.5 0.6 gave swing ratios substantially the same as the Fl
carbon microphone for phase modulation, but only
Fl A-Constant input volume 1.1 1.0 1.0 0.9 about half as great for frequency modulation.
Fl B-Constant peak level of
speech 1.5 0.8 1.4 0.7 A general conclusion is that the swing ratios are quite
Fl C-Constant phase peaks 1.5 1.2 1.4 1.1 dependent on the microphone and its circuits, on the
Fl D-Spread centered 1.4 0.9 1.3 0.8
kind of volume regulation employed, and probably on
Moving-coil A-Constant input volume 1.1 0.5 0.8 0.5 other features of the circuit. For accurate results, the
Moving-coil B-Constant peak level of
speech 1.7 0.4 1.0 0.4 swing ratios should therefore be determined for the
Moving-coil C-Constant phase peaks 1.7 0.6 1.0 0.6 particular circuits which are to be used. The figures
Moving-coil D-Spreads centered 1.3 0.5 0.8 0.5
derived here will, however, serve to indicate roughly the
Note: PM = phase modulation. FM = frequency modulation. values to be expected.
Summary-The helix is a fundamental form of antenna of which maintained in the axial mode over wide frequency ranges because of
loops and straight wires are limiting cases. When the helix is small a natural adjustment of the phase velocity of wave propagation on the
compared to the wavelength, radiation is maximum normal to the helix. The terminal impedance is relatively constant over the same
helix axis. Depending on the helix geometry, the radiation may, in frequency range because of the large initial attenuation of waves on
theory, be elliptically, plane, or circularly polarized. the helix. The conditions for circular polarization are analyzed, and
When the helix circumference is about 1 wavelength, radiation the importance of the array factor in determining the radiation pat-
may be maximum in the direction of the helix axis and circularly tern of a long helix is discussed.
polarized or nearly so. This mode of radiation, called the axial or
beam mode, is generated in practice with great ease, and may be INTRODUCTION
dominant over a wide frequency range with desirable pattern, im-
pedance, and polarization characteristics. The radiation pattern is A HELIX is a fundamental geometric form. It has
applications in many branches of physics and
* Decimal classification: R125.1 XR326.61. Original manuscript engineering. For example, in mechanical systems
received by the Institute, June 7, 1948. Presented in part, 1948 IRE the helix or coil spring is a familiar structure; in electri-
National Convention, New York, N. Y., March 23, 1948.
t Department of Electrical Engineering, Ohio State University, cal systems, a helical conductor or inductor is a common
Columbus, Ohio. type of circuit element; and in many dynamic phenom-
264. PROCEEDINGS OF THE I.R.E. MarCh
ena, particles follow helical paths. Recently the helix TRANSMISSION AND RADIATION MODES OF HELICES
has been applied as a beam antenna.'-6 The dimensions of a helix are very conveniently il-
In considering the helix as an antenna, it is important lustrated by a diameter versus spacing chart or, as in
that it be regarded, not as a unique or special form of an- Fig. 2, by a circumference versus spacing chart. On this
tenna, but rather as a basic type of which the more chart, the dimensions of a helix may be expressed either
familiar loop and straight-wire antennas are merely spe- in rectangular co-ordinates by the spacing Sx and cir-
cial cases.2 Thus, a helix of fixed diameter collapses to cumference irDx or in polar co-ordinates by the length
a loop as the spacing between turns approaches zero, of one turn Lv, and the pitch angle a.
and, on the other hand, a helix of fixed spacing straight-
ens into a linear conductor as the diameter approaches PITCH ANGLE, oh
zero. It is the purpose of this paper to discuss the helical lA
antenna from this general point of view including the a
SPACING IN WAVELENGTHS, SA
TID Fig. 2-Circumference versus spacing chart for helices showing re-
lL gions for normal radiation mode (shaded) and axial or beam mode
(cross hatched).
T, QQOQQQQQQQ (b)
N PN
uu-_.z
O -0- _+0+
- + -
(c)
T, T2 T3
Fig. 3-Approximate charge distribution on helices for Fig. 4-Sinusoidal field variation for small helices.
different transmission modes.
on low-frequency inductors. It is also the dominant mode A first-order transmission nmode, designated T,, has
in the traveling-wave tube.9-'3 Since the adjacent re- adjacent regions of maximum positive and negative elec-
gions of positive and negative charge are separated by tric charge approximately one-half turn apart or near
an appreciable axial distance, a substantial axial com- the opposite -ends of a diameter, as suggested in Fig.
ponent of the electric field is present, and in the travel- 3(b) for the case of a small pitch angle. This mode is
ing-wave tube this field interacts with the electron important when the length of one turn is of the order of
stream. If the criterion Lx < 2 is arbitrarily selected as a the wavelength (L--X). It is found that the radiation
boundary for the To transmission mode, the region of from helices of this turn length and of a number of turns
helix dimensions for which this mode is important is (n > 1) is usually a maximum in the direction of the helix
shown by the shaded area in Fig. 2. axis and is circularly polarized, or nearly so.' 2 This type
Theoretically, it is of interest to examine some of the of radiation pattern is referred to as the axial or beam
possible radiation patterns associated with the To trans- mode of radiation Ra. This radiation mode occurs for a
mission mode. Only the simplest radiation case will be wide range of helix dimensions and, being associated with
considered. This occurs when the helix is very short so the T, transmission mode, the combined designation ap-
that nL<<X and the assumption is made that the cur- propriate to this region of helix dimensions is TIRa. The
rent on the helix is uniform in magnitude and in-phase axial type of radiation is discussed further in a later sec-
along its length.'4 Referring to Fig. 3(a), the length is tion (Axial Radiation Mode).
much less than that between adjacent regions of maxi- Still higher-order transmission modes, designated T2,
mum positive and negative charge. Theoretically, it is T3, etc., will have the approximate charge distributions
possible to approximate this condition with a standing suggested in the one-turn views of Fig. 3 (c) for the case of
wave on a small end-loaded helix. rfhe terminal imped- a small pitch angle. For these modes to exist, the length of
ance of such a small helix would be sensitive to fre- one turn must generally be at least one wavelength.'5
quency and the radiation efficiency would be low. How- The normal Rn and axial R. radiation modes are, in
ever, let us assume that appreciable radiation can be reality, special cases for the radiation patterns of helical
obtained. The maximum radiation is then normal to the antennas. In the general case, the maximum radiation is
helix axis for all helix dimensions, provided only that neither at 0 = 0 nor at 0=900 but at some intermediate
nL<<X. Hence, this condition is referred to as a normal value, the pattern being conical or multilobed in form.2,4
radiation mode R.. Referring to Fig. 4, any compo- THE NORMAL RADIATION MODE
nent E of the distant electric field perpendicular to the
radius vector is given approximately by E=k sin 0, The direction of maximum radiation is always normal
where k is a constant. The radiation is, in general, el- to the helix axis when the helix is small (nL<<X). Refer-
liptically polarized, but for particular helix dimensions ring to Fig. 5(a), the helix is coincident with the polar or
y axis. At a large distance r from the helix, the electric
9 R. Kompfner, "The traveling-wave tube as amplifier at micro- field may have, in general, two components Ep and Eo,
waves," PROC. I.R.E., vol. 35, pp. 124-127; February, 1947.
10 J. R. Pierce and L. M. Field, "Traveling-wave
tubes," PROC. as shown.
I.R.E., vol. 35, pp. 108-111; February, 1947. Two limiting cases of the small helix are: (1) the short
11 J. R. Pierce, 'Theory of the beam-type traveling-wave tube,"
PROC. I.R.E., vol. 35, pp. 111-123; February, 1947. electric dipole of Fig. 5 (b), a = 900, and (2) the small loop
12 L. J. Chu and D. Jackson, "Field theory of traveling-wave of Fig. 5(c), a= 00. In the case of the short electric di-
tubes," PROC. I.R.E., vol. 36, pp. 853-863; July, 1948.
13 C. C. Cutler, "Experimental determination of helical-wave pole, E = 0 everywhere and the distant electric field
properties," PROC. I.R.E., vol. 35, pp. 230-233; February, 1948.
14 It is assumed here that the phase velocity on the helical con- 15 The phase velocity along the helical conductor for T1 and higher
ductor is approximately that of light. The in-phase condition requires modes may differ considerably from that of light. It is often less, and,
an infinite phase velocity, but this can be approximated by consider- as shown later, it may be a function of the helix pitch angle and
ing only short helices nL<KX. diameter.
266 PROCEEDINGS OF THE I.R.E. March
IY the length of one "turn" is now given by L = S+irD, the
far field pattern will be independent of the number of
turns. Hence, to simplify the analysis, only a single
turn will be considered. The electric field components at
a large distance are then given by (1) and (2). The
operator j in (1) and its absence in (2) indicates that
ER and Eo are in time phase quadrature. Taking the
Helix Dipole Loop
ratio of the magnitudes of E0 and Eo, we have
(a) (b) (c) Eo SX
Fig. 5-Relation of field components to helix, dipole, and loop. Ek 2irA (3)
has only an Eo component. On the other hand, with the Introducing the relation between the area and diameter
small loop, Eo =0 everywhere and the distant electric of the loop, A = rD2/4, (3) becomes
field has only an E4 component. By the retarded poten- Eo 2SX
tial method, it may be shown that Eo at a large distance (4)
from a short electric dipole (r>>X>>s) is given by'6"7 E, ir2D2
jco[I]s sin 0 j60-r[I] sin 0 s In the general case, both Eo and EK have values and
Eo _- (1) the electric field is elliptically polarized. Since Eo and E,*
47recr' r X
are in time phase quadrature, either the major or the
where minor axis of the polarization ellipse will lie in a plane
s =length of short dipole through the polar or y axis (see Fig. 5(a)). Let us assume
w = 27rf that the y axis is vertical and that observations of the
r=distance from origin field are confined to the equatorial or x-z plane. The
c= velocity of light (in free space) ratio of the major to minor axes of the polarization el-
E = dielectric constant of medium (free space) lipse is conveniently designated as the axial ratio (A.R.).
and [I] = retarded value of the current =10 exp [jw Let us define the axial ratio in this case as
(t-r/c) ]. Eo 2SX
In an analogous way, E4 at a large distance from a A.R. = -=(5 (5)
short magnetic dipole or from the equivalent small loop E,g 7r2D2
(r>>A>>D) is
Thus, in the extreme case when EO =0, the axial ratio
E
1207r2 [I] sinG A is infinite and the polarization ellipse becomes a straight
r 2(2) Vertical line indicating linear vertical polarization. At
the other extreme, when Eo =0, the axial ratio is zero
where and the polarization ellipse becomes a straight horizon-
A= area of loop = 7rD2/4 tal line indicating linear horizontal polarization.
[I] =retarded value of the current on the loop. An interesting special case occurs for an axial ratio of
If nL<<X, a helix may be considered, as has been done unity (Eo = E+). This is the case for circular polarization.
by Wheeler,7 to be a combination of a series of loops and Setting the axial ratio in (4) equal to 1, we have
linear conductors as illustrated in Fig. 6. Each turn is
assumed to consist of a short dipole of length S coh- 7rD=V/2SX or 7rDx= V/2Sx. (6)
nected in series with a small loop of diameter D. Fur-
ther, the current on the helix of Fig. 6 is assumed to be This relation was first shown by Wheeler in an equiva-
uniform and in phase over the entire length. The re- lent form.7 For this case, the polarization ellipse becomes
quired end loading is not shown. Provided nL<<X where a circle. The radiation is circularly polarized not only in
all directions in the x-z plane but in all directions in
space except the direction of the y axis, where the
MIs field is zero.
The relation of helix dimensions for circularly polar-
ized radiation normal to the axis as given by (6) is indi-
cated in Fig. 2, and also in Fig. 7 by the curve marked
H- C.P. (Circular Polarization). This curve is accurate only
D in the region for which rD<X and S<KX. This region is
shown to an enlarged scale in Fig. 7. In general, the
Fig. 6-Equivalent form of small helix. radiation is elliptically polarized. If 7rD> x/25X, the
16 See, for example, S. Ramo and J. R. Whinnery, "Fields and major axis of the polarization ellipse is horizontal, while
Waves in Modern Radio," John Wiley and Sons, New York, N. Y., if irD <V2X the major axis is vertical. By varying the
1944; p. 430.
17 Rationalized mks units are used. pitch angle a of a helix of constant turn length L, ori-
1949 Kraus: The IHelical Antenna 267
0
U,-
I
Traveling
z wave (a)
-j
LI
S
r
zC
Helix axis
z
0
-r re
Y
SPACING IN WAVELENGTHS, SN Y< 41 g |~~~Hli (b)
Fig. 11-Diameter versus spacing chart showing region for funda-
mental axial or beam mode of radiation. conduc or
p cos a (25)
X- -_ k-1I
For a=0, we have a loop, and b=-1/p. Hence, the
relation being derived may be applied not only to the In a similar fashion, we obtain for the relative electric
general helix case but also the special case of a loop. field intensity component in the y direction, E,:
Since t and z1 are independent of 0, the first two terms of E, ei (k-1)01 1 -I
-
the exponent may be taken outside the integral. Hence Ey=- jO,+ k 1 (26)
(16) becomes 2j -
2rn If the helix is very long (0>>1), (25) and (26) become
E Eoei(wtt-Azj) sin 6eikGdO (18) very nearly
E101 E101
where Ex= -.i .1 2 and EX,
= (27)
2 2
(3= 2r/X
The ratio of these then gives Er/Ev= -j, which satisfies
k-Irb=Lx (sin a--) (19) the condition for circular polarization. Although these
give the important conditions for circular polarization,
Integrating and introducing limits, we obtain another condition resulting in circular polarization is
obtained when (k 1)O1 = 2wm where m = integer. This
El condition is fulfilled when either the positive or negative
Ex =
k2 -
-
1
(ei2rnk - 1) (20) sign in (k 1) is chosen, but not for both. To summarize
the important conditions :28
where (1) The radiation in the axial direction from a helical
E= Eoei(It-zlz). antenna of any pitch angle (O<a<900) and of an in-
tegral number of one or more turns will be circularly
The expression for the relative electric field intensity polarized if k = 1.
in the y direction Ev is identical to (20) except that it (2) The radiation in the axial direction from a helical
has cos 0 instead of sin 0. From this we obtain antenna of any pitch angle (0 <a< 90) and a large num-
Elk ber of turns, which are not necessarily an integral num-
Ey-(*
Elk
(e j2rnk -1) (21) ber, is nearly circularly polarized if k is nearly 1.
j(k2-1) Let us now investigate the significance of the require-
For circular polarization on the axis, the required condi- ment that k = 1. Referring to (19), k is negative in
tion is the case of interest, since sin a <. 1 and 1/p 1. Thus, for
k= -1 we have
_= + j. (22)
Ey Lx (sin a - = -1
Taking the ratio of Et to Ey as given by (20) and (21), or p
we get Lix Lx
(28)
Ex Lx sin az+ I SA + I
(23)
EX k If p =1, the circular, polarization condition is La-S = 1
or L-S=X. This was first pointed out in footnote refer-
Hence, for circular polarization on the axis of a helix of ence 1. The relation for p in (28) is identical with the
an integral number of turns (n =1, 2, 3, ), k must-
However, as will be shown, nearly circular polariza- and the effect of the reflected wave (2 in Fig. 8) is neglected. The
effect of the reflected wave on the axial ratio is discussed in fQotnote
tion may be obtained provided only that the helix is reference 2, p. 91.
272 PROCEEDINGS OF THE I.R.E. March
value of p required for the fields of each turn of a helix helix by noting the close agreement of measured field
to add in phase in the axial direction as given by (13).29 patterns with array-factor patterns calculated with this
value of p and the poor agreement when other values of
PHASE-VELOCITY COMPARISON p are used.
Several expressions for the required phase-velocity SINGLE-TURN PATTERN
factor p have been derived corresponding to different The pattern of a single turn is an important factor in
conditions. These are summarized for the T, transmis- determining the pattern of short helices. In the case of
sion mode in Table II. Two of the expressions are iden- long helices, the array factor is relatively more impor-
tical, namely, for circular polarization (C.P.), and in- tant, and is usually sufficient to give the approximate
phase fields from each turn. In Table II, 4o is the value main-lobe pattern of the helix. However, it is neverthe-
of 4 at the first null in the radiation pattern, and 4'o is the less necessary that the direction of maximum radiation
value of ^,t at the first null in the array factor. from a single turn be approximately in the axial direc-
TABLE II tion. Accordingly, it is of interest to investigate briefly
the form of the single-turn pattern of helical antennas
Condition Required Phase-Velocity Factor p radiating in the axial mode. Referring to the preceding
(1) and (2) C.P. and in- La I sections, the condition k -1 also results in the single-
phase fields p= =- turn pattern maximum being nearly in the axial direc-
Sx++ Cos a tion. This follows from the fact that when is small the
sin a+ a
2n+1 \cos a
1 equivalent at one instant of time to a broadside array of
SX+1+- sin a-+-(
2n 2n / rDX two short dipoles spaced by about the diameter of the
(4) From first null of L4 helix, as in Fig. 14(b). Since the dipoles are in phase,
measured pattern30 p= the maximum radiation is normal to their plane or in
Po the axial direction. The pattern is also very broad in the
SX\ Cos 4s+1+
2ir axial direction, as indicated. With passage of time, these
equivalent dipoles rotate around the axis, yielding cir-
Curves calculated by the three different methods of cular polarization. If is not small, then it becomes nec-
a
Table II are compared in Fig. 13 with the measured varia- essary to approximate the single turn of the helix by a
tion of the phase velocity as a function of frequency on a square turn with four short linear segments, as was done
seven-turn 120 helix.31 All curves are in general agree- in the pattern calculations of footnote reference 2. A
ment in the region in which p increases with frequency.32 square turn is suggested by the perspective sketch in Fig
In Fig. 13, 300 Mc corresponds to a helix circumference 14(c). Since the wave on the helix is, to a good approxi-
of 0.72 free-space wavelengths and 500 Mc to a helix mation a single, traveling wave, the radiation maximum
circumference of 1.2 free-space wavelengths. It can be is tilted forward from the normal to the conductor. As
effectively demonstrated that p for "maximum directiv- shown in footnote reference 2, it turns out that the tilt
ity" is most probably the one actually occurring on the angle of the radiation maximum for a short segment
r
>~.9
-0
.8
- - --c--Calculated
from measured \Rotation direction
>, .8
- pattern nulls Wave direction
8 .6
300 350 400 450 500 (a)
Frequency in megacycles/sec. ZtCXjS
Axis (b)
lnstantoneous'
Fig. 13-Comparison of variation of measured phase-velocity factor current \\Axis
(p =v/c) with frequency on a seven-turn 12 helix with the calcu- directions