PARAMETRIC GENERATION OF ALTERNATING CURRENTS

• Document Type: CREST
• Collection: General CIA Records
• Document Number (FOIA) /ESDN (CREST): CIA-RDP80-00809A000600370626-6
• Release Decision: RIPPUB
• Original Classification: C
• Document Page Count: 11
• Document Creation Date: December 22, 2016
• Document Release Date: September 26, 2011
• Sequence Number: 626
• Publication Date: January 29, 1951
• Content Type: REPORT

Sanitized Copy Approved for Release 2011/09/27: CIA-RDP80-00809A000600370626-6
?1
INFORMATION FROM
FOREIGN DOCUMENTS OR RADIO BROADCASTS CD NO.
CLASSIFICATION CONFIDENTIAL
COUNTRY USSR DATE OF
INFORMATION 1939
SUBJECT Scientific - Electricity, alternating currents
HOW DATE DIST.021 Jan 1951
PUBLISHED Photostat of monthly periodical
DATE
PUBLISHED Not given (possibly 1939)
SUPPLEMENT TO
LANGUAGE Russian REPORT NO.
TN9 DOCUMENT CONTAIC] INFORMATION ANNNCTIMD TNN NATIONAL DINNNIM
OF TNN UNITED ]TAT[] ?ITNIN TNN MNAMIN OF NINIONAN9 ACT ND
II. ]. C.. ]I AND ]:. Al AMNMOND. ITS TNAIINNI]]ION ON TNN N80UTWN
0/ ITS CONTNNIN IN All MAANNN TO AM DIA0IMORIT.MO FIN]OM IN !NO.
NINITND AT LAW. REPRODUCTION OF THIN FORM ID FRONINITID.
Elektrichestvo, No 11, pp 67-76.
fFigures are appended]
1. There are two basic methods by which mechanical energy can be converted
into electrical energy: (a) through the movement of electrical currents in a
magnetic field and (b) through the movement of electric charges in an electric
field. Both of these methods are used in electromagnetic and electrostatic ma-
chines in practice, but the role of the first is immeasurably more important,
since electrostatic machines up to this time have been more important as an ap-
purtenance of the physical laboratory than as an engineering tool.
In the first mechanical generators of electrical energy, the magnetic
field was ^onstant, aid this apparently left its mark upon.the further develop-
ment of electric' machine construction.
The principle. of self-excitation of dynamos, the so-called Siemens'
dynamoelectric principle, which almost completely determined the development of
present-day dc and ac electric machines with supplementary excitation, calls
for a magnetic exciting field which is constant in time. The generation of
direct and alternating currents with the help of a constant magnetic "field is
very simple and convenient, and this, together with the above-mentioned historical
reason, forced electric machine construction almost exclusively into the use of
constant fields for excitation. Even in electrostatic machines, self-excitation,
of the Toepler-Wimehurst type, for example, or supplementary excitation still in-
volves constant exciting fields.
It is also possible in principle to use alternating-magnetic or electric
exciting fields for mechanical generation of electric currents. However, this is
not as simple as the use of constant fields, since for correct utilization of the
laws of induction and mechanical forces, synchronid.m must be maintained between
the movement of the current (or charge) conductors and the field variation. While
NSRB DISTRIBUTION
Sanitized Copy Approved for Release 2011/09/27: CIA-RDP80-00809A000600370626-6
CLASSIFICATION CONFIDENTIAL
 Sanitized Copy Approved for Release 2011/09/27: CIA-RDP80-00809A000600370626-6
the maintenance of this synchronism for supplementary excitation is difficult
(synchronism of phase must also be maintained), it is still more difficult for
self-excitation of an ac circuit. A number of designs for self-excitation of
an ac circuit have been propo,.ed to solve this problem, e.g., Rudenberg [1J,
Barkhausen [2J, Professor M. V. Shuleykin C3J. In all the above designs,
however, the authors circumvented the main difficulty, i.e., maintenance of
phase synchronism, by using commutation (collector rings), thereby changing thek
alternating into a direct process.
The problem of self-excitation of an ac circuit in its purest form can
be formulated in the following manner: We have an electrical ac circuit in
which there are no apparent sourcas of emf, i.e., magnetic or electric fields.
Can we, by changing only the position of the circuit components, its inductance
or capacitance, excite electric currents in it? Obviously, communtation, i.e.,
changing the dc resistance (conversion of ac into dc) is excluded.
As was theoretically shown by Poincare [4J in 1907, it is impossible
to self-excite electric currents in systems without capacitance, in which (dur-
ing motion) only the inductance changes, providing tkat communtation is not used.
Such self-excitation is possible, however, in special cases with capacitance
present when the natural oscillations of the system are used. Rayleigh [5J
pointed out 'this possibility even earlier in connection with similar effects in
acoustics (the Melde experiment). Although electrical engineers knew of cases
of self-excitation of an ac circuit (for example, an induction generator when
the properly selected capacitance is connected into its circuit), these were con-
sidered undesirable effects and until recently the problem of self-excitation of
an ac circuit had not been studied systematically.
2. The physical side of the effects occurring in self-excitation of an ac
circuit can be clarified very simple from the following considerations [11J.
Suppose an electrical circuit consists of a capacitance C, inductance L, and do
resistance R (Figure 1). Let the induction coil be constructed so that it admits
of a periodic change of inductance (Figure 2) and let an infinitesimal current
"i" flow through it at some moment. This infinitesimal current could be obtained
as a result of some change inductions, parasitic currents, or atmospheric discharges;
in principle it might appear in the circuit even if the whole system were isolated
from external effects because of the so-called fluctuations of electrons in the
conductors.
Here, in contrast to the case of dynamoelectric self-excitation, the
current polarity is not important. Let us assume that at this moment, the in-
ductance of he circuit is changed by 4 L by an external force. This requires
the work 2 i 24L and, consequently, the energy of the system is increased by
this amount. After this change of the system, we leave it alone. Since the in-
ductance is connected to a capacitance,1oscillation will occur if the right con-
ditions prevail in the circuit. After o.: a period T of natural oscillations
of this circuit, all the magnetic energy of the system is converted into electrical
energy of the condenser, and the current becomes equal to zero. If at this moment
the inductance was decreased to its initial value, no work would be required since
no current is flowing. After this operation, we again leave the circuit alone,
and then the electrical energy of the condenser will again be converted into mag-
netic energy, and the current will again increase and reach its maximum after the
following 1/4 period of natural oscillations of the circuit. At this time, we
can again increase the inductance ,D L. expending some work to introduce energy
into the system, and thus again repeat the entire cycle of changes. If in chang-
ing the inductance, we have introduced more energy into the system than was ex-
pended in charging the condenser, the store of energy in the system will be increased
and, consequently, the current strength and the charge on the condenser' should be
increased with each new cycle. Thus, the oscillations in the circuit will continue
to increase and it will be self-excited. A simple mechanical analogy to the above
process is the rocking of a swing,
Sanitized Copy Approved for Release 2011/09/27: CIA-RDP80-00809A000600370626-6
 Sanitized Copy Approved for Release 2011/09/27: CIA-RDP80-00809A000600370626-6
CONFIDENTIAL
Sanitized Copy Approved for Release 2011/09/27: CIA-RDP80-00809A000600370626-6
It is easy to derive the energy conditions necessary for self-excitation
of an ac circuit. The losses in the system for one cycle of changing the in-
ductance can be expressed as
and since the energy introduced into the system during this time is equal to
1/2 124L, the condition for self-excitation is defined by the following in-
1nax + Lain
Thus, if we change the inductance with a frequency which is approxi-
mately twice that of the natural frequency of the system, then when the rela-
tive magnitude of the change of inductance is greater than a definite value
which depends on the damping of the natural oscillations of the system, an al-
ternating current should emerge and increase in the system. The frequency of
this alternating current will be approximately equal to the natural frequency of
the system, i.e., it willbe close to half the frequency of the change of induct-
ance. Obviously, these considerations also apply to the case when the capaci--
tanceis.changeliin the circuit.
The simple considerations above show that with a periodic change of
one of the parameters of the oscillatory circuit (capacitance or inductance) and
observance of the condition (11) (as closer inspection shows, the condition (11)
takes on the form
my n or WOL74R (3)
for the case of a harmonic change of the parameter), ever-increasing currents
must emerge in the circuit, i.e., the circuit will be self-excited. If the para-
meters of the system are left constantly independent of the currents or voltages,
or, in other words, the system is left linear (i.e., can be described by linear
differential equations), the current or charge would continue to increase without
limit until the insulation of the system broke down or the power of the motor
rotating the variable inductance (or capacitance) became insufficient.
 Sanitized Copy Approved for Release 2011/09/27: CIA-RDP80-00809A000600370626-6
A stationary state could not be established in such a "linear" system
and consequently it could not become a current generator. To do this, we must
introduce into the system factors which depend on the current and which will
limit its further increase. Only then can we obtain a stationary state and
produce an alternator. This is completely analogous to dynamos with the usual
self-excitation, in which the stationary state is also provided by the "non-
linearity" of the magnetization curare for iron.
Since self-excitation in our case is provided by a change of a para-
meter of the system, it has been given the name "parametric," and generation
based on it is called parametric. We especially emphasize that the frequency
of the alternating current generated is exactly equal to half the frequency of
the parameter change, so that we deal here with a synchronous electric machine.
From the above, it is clear that the problem of parametric generation
of alternating current is inseparably connected with the problem of self-ex-
citation of an ac circuit, which in turn is connected in principle with the
excitation of oscillations by a periodic change of the parameters of the oscil-
latory system. This last very important problem has recently attracted the in-
terest of many researchers [6, 7, 8, 9, 10, 15_7, both in the Soviet Union
and abroad. Beginning in 1927, it has undergone systematic theoretical and ex-
perimental development by a number of workers under the direction of Academician
L. I. Mandel'shtam and the author in the Scientific-Research Institute of the
first Moscow State University, the former Central Radio Laboratory, the former
Leningrad Electrophysical Institute and then in-the Leningrad Industrial Insti-
tute. As a result of these works t10, 11, 12, 13, 14J, effects of parametric
self-excitation in an ac circuit were obtained for the first time using a periodic
mechanical change of both inductance [10J and capacitance L 11] in the absence
of any apparent sources of electric or magnetic fields (exciters, storage bat-
teries, permanent magnets, etc.). In addition, a theory of the processes was
developed (self-excitation conditions, magnitude of stationary amplitude, etc.).
Naturally, the establishment of a new principle of self-excitation of
alternating current has pushed the problem of using it for generation of alter-
nating currents to the forefront.
L Sections 3 and 4 are omitted here. They deal with a theoretical treatment
of the nonlinear differential equations to which, according to the author, the
theory of parametric generation of alternating current reduces. -7
5. In conformance with the two possible methods of parametric generation,
namely, periodic change of inductance or capacitance, there can be two essen-
tially different new types of electric machines which we shall distinguish (a)
the " inductive" parametric alternator (PMI) and (b) the "capacitive" parametric
alternator (PMYe).
In the first case, we deal with a synchronous electric machine, externally
similar to ma ?iines of the normal inductor type, the only essential differences
being that, ou one hand, there are no exciting windings, a fact which is very im-
portant in design, and, on the other, capacitance is required. Therefore, there
is no doubt about the possibility of obtaining power of the same order as from
ordinary alternators from the "inductive" parametric alternator. The chief in-
terest in work with inductive parametric alternators lies in determining to what
degree and under what conditions the new principle of parametric self-excitation
might prove useful and advantageous in practice. This new principle makes excit-
ing windings unnecessary, but also carries the condition that the ac circuit must
be oscillatory and approximately tuned in resonance to half the frequency of the
inductance change, i.e., it makes capacitance necessary.
Sanitized Copy Approved for Release 2011/09/27: CIA-RDP80-00809A000600370626-6
 Sanitized Copy Approved for Release 2011/09/27: CIA-RDP80-00809A000600370626-6
1
These facts force us to assume that the region of high frequencies
would be especially advantageous for parametric alternators, since the absence
of rotor windings permits the use of high angular velocities, while the cost
and size of the condensers required for tuning to resonance are decreased
with an increase of frequency. On the basis of these considerations, an ex-
perimental model of a 1-kw, 600-cycle, 3,OjO.r1m alternator. was. des%nedwiththehelp
of Engineer M. M. Verbitskiy and succesetuily produced in the shops of the
Leningrad Industrial Institute. The rotor of the alternator is a duraluminum
disk, on the periphery of which are placed 24 blocks of iron, each separated
by layers of insulating material. The stator in turn consists of 24 open
magnetic circuits encompassing, one common winding placed symmetrically on the
periphery. When the rotor turns, the magnetic circuits of the stator are al-
ternately closed and open, which provides a periodic change of the circuit in-
ductance.
Tests made by scientific workers A. G. Rzyakin and A. M. Martynov of
the parametric machines group, aided by mechanic P. K. Loncha, showed that
when self-excitation conditions were fulfilled (obtained by tuning the circuit
to the desired frequency by connecting the proper capacitance in the circuit),
the alternator was excited and could produce power of one kw at an efficiency
of about 70% for prolonged periods. A number of characteristics of this al-
ternator are given in Figure 4 and an oscillogram of the current in one case is
shown in Figure 3, b.
Among the interesting design features of this alternator are its low
weight (about 15 kilograms without the base), low copper weight (0.76 kilograms)
and a low "machine constant"
Testing revealed that some parts could be improved, which would result
in increased efficiency and lower alternator weight.
On the basis of data obtained in the production and testing of this ex-
perimental model, a more powerful model of a parametric alternator (40-60 kva,
1,600 cycles, 3,000 rpm) was designed with the help of Engineer M. M. Verbitskiy
and Professor A. A. Alekseyev. This alternator design was given to the "Elektro-
sila" Plant imeni Kirov for production.
The characteristics of inductive parametric alternators indicate that
it can best be used in those branches of engineering where high frequencies
(500-2,000 cycles) are necessary or advantageous, e.g., for supplying induction
furnaces and vacuum-tube radio transmitters, and also wherever a high dc voltage
is required (this would be obtained by rectifying the ac voltage generated by
the parametric alternator). Since capacitance is necessary in all these cases,
the condenser used as an integral part of the parametric alternator is not excess
baggage from the standpoint of increasing the size, weight, and cost. The more
powerful inductive parametric alternator now under construction is designed for
supplying induction furnaces.
Along with all the structural advantages gained by the use of the para-
metric principle of self-excitation which eliminates the exciting windings, the
need for a magnetic field exciter is eliminated. This fact, which is of little
importance in ordinary high-power synchronous machines, may prove very advantageous
with respect to weight, size, and cost for small generators, especially when
the problem of reducing weight is very important. In addition, the problem of
the condenser is much more easily solved in the generation or low power.
Sanitized Copy Approved for Release 2011/09/27: CIA-RDP80-00809A000600370626-6
 Sanitized Copy Approved for Release 2011/09/27: CIA-RDP80-00809A000600370626-6
"1
6. The second possible method of parametric generation of alternating cur-
rent was implemented in the so-called "capacitive" parametric alternator (PMYe).
Here we deal with a synchronous electric machine, in which mechanical work is
transformed into electrical energy by a periodic change of the capacitance con-
nected in the oscillatory circuit, the latter being tuned to approximately half
the frequency of the capacitance change. The basic part of this alternator is
the variable capacitance, which can be made in the same way as the rotating vari-
able condensers used in radio engineering in the form of a collection of stator
and rotor platds of the proper form, so designed that when the rotor turns with
the desired speed the capacitance changes within wide limits from Cmax to Cmin?
The capacitive parametric alternator is so different essentially from
ordinary electric machines that electrical engineers will first of all doubt
whether. power of the order in which they are interested can be obtained from it.
Actually, the power Pc which can be generated by changing the capacitance of the
oscillatory circuit according to the law-.
C-1=Co ?(1+m?cos 2wt)
W 1 T
PC=- _a w r g2sin 2 wt?dt,
Co T .1 0
so that with a harmonic change of current we will have in the best case
Pc- 2 mwgo/2Co mwCo UZmax (17)
from which it follows that the power which can be generated will be the greater,
the higher the frequency W, the greater the percentage modulation of the capaci--
?tam a m, and the greater the energy in the condenser
1/2 0U2max.
Thus, the problem of generating high power depends on the possibility
of accumulating sufficient energy in the variable condenser. Since it is impos-
sible in practice to construct a condenser which would have high capacitance and
?still permit rapid changes in capacitance over wide limits, the only possible
methdd of solving this problem, as is seen from formula (17) for P , lies in in-
creasing the voltage Umgx However, the dielectric strength of air under normal
conditions (30,000 v/cm) is far too low to obtain any degree of power.
To illustrate, we calculate the power which can be obtained from an ex-
perimental model of a capacitive parametric alternator. The stator of the vari-
able condenser, which is the main part of the alternator, consists of 26 disks
with diameter 28.5 cm, each divided into 16 sections, while the rotor has 25
such disks. Adjacent disks are 2 mm distant and taus
Cmax = 6420 cm, while Cmin = 3860 cm,
-6-
CONFIDENTIAL
Sanitized Copy Approved for Release 2011/09/27: CIA-RDP80-00809A000600370626-6
C0 = 1/2 C f Cmin - 5140 cm and m = 0.25.
For 3,000 rpm, the frequency of capacitance change is 2w=5,000 cycles.
Taking as the permissible voltage U - 5,000 v, we obtain for the power:
Pc = 22.4 w,
a negligible quantity from the standpoint of-power. The only method by which
the power can be increased is to increase the permissible voltage, i.e., to
 Sanitized Copy Approved for Release 2011/09/27: CIA-RDP80-00809A000600370626-6
-1
increase the dielectric strength of the medium. There are two ways in which this
can be done. One of these is to place the variable condenser in a gaseous medium
under high pressure. Since the breakdown voltage of gases, wherever Paschen's
law holds, increases with an increase of pressure p as a function of pd (where d
is the distance between electrodes or, in our case, between adjacent condenser
plates), it is possible by increasing the pressure to increase the working voltage
U For example, if the pressure is increased to 20 atm, the breakdown voltage
isaincreased 16 times, and. consequently the power is inereased,.approximptely-:250.;
ub iglQ g
times (for the same, safety factor),. Therefore, under.these conditi~o~.gi op
of a capacitive alternator would, generate about 5.6,icy,.a,value already+qf,inter at
for engineering. -Even higher voltages could be-obtained with a further ini?'rease
of pressure.
According to I. M. Gol'dman's recent data f20_7, gases under pressures
of 70-80 kg/sq cm can sustain voltage gradients up to 106 v/cm. This would per-
mit us to obtain power of the order of 22 kw from the model capacitive alternator
considered. If we use this data as a basis, calculation shows that it is possible
to construct capacitive parametric alternators producing high power (tens of thou-
sands of kilowatts) whose size would not exceed that of the ordinary high-power
alternators. We should keep in mind, however, that the quantitative data cited
for breakdown voltages at high pressures applies to experiments with gases at rest,
while in the capacitive parametric generator we deal with breakdown voltages in a
gaseous medium which is in very rapid random motion. In addition, we also have
here a nonuniform electric field which, according to available data f22-7, may
substantially affect the permissible voltages.
To clarify experimentally how all these factors affect the dependency
of the working voltage on pressure and to what degree the pressure increase could
be used to increase the power of the capacitive parametric alternator, we con-
structed an experimental model of a variable condenser and rotated it in a medium
under high pressure. The stator of the condenser is its system of fixed plates
while the rotor is its system of movable plates. Since the plates of the stator
and rotor have 16 sections, each symmetrically spaced along the periphery, at
3,000 rpm the frequency of the capacitance change is 800 cycles and, consequently,
the frequency of the alternating current generated is 400 cycles. The condenser
was placed in a steel casing into which dry air could be pumped with the aid of a
compressor.
To fulfill the self-excitation conditions for the variable condenser,
a coil with inductance L equals 27.4 h was connected in the circuit. This coil
was selected to tune the circuit consisting of the condenser, the coil, and the
load resistance to resonance at a frequency of about 400 cycles. The schematic
diagram of the generator is shown in Figure 5. C denotes the variable condenser
described above, F is a spherical discharger connected in parallel with C, L is
the coil for tuning the circuit to resonance at the desired frequency, R is a
lamp rheostat, and R2 is a rheostat of the Rustrat type. Since the logarithmic
decrement of the circuit without load was equal to 0.035, and the experimentally
measured modulation factor m was equal to 0.34, the generator was easily self-
exci,:ed and a high load resistance could be used.
The results of 'some measurements made by A. G. Rzyankin and A. M. Mar-
tynov showed that the breakdown voltage in the variable condenser increased
smoothly and continuously with the increase of gas (air) pressure when the rotor
turned at 3,000 rpm. Thus, the working voltage of the generator and consequently
its rower also could be increased. Because of poor packing at the point where the
rotor shaft left the casing, the pressure could not be increased above 12-13 kg/sq
cm; in addition, low accuracy in the production of the condenser plates and in
their assembly considerably reduced the breakdown voltage. The useful power ob-
tained at a pressure of 12,5 kg/sq cm was about 560 w at a voltage Umax ft 21 kv.
The electrical efficiency was above 92%. At the same time, measurements were made
of the mechanical losses as a function of the pressure in the medium containing
the rotating condenser. An oscillogram of the current produced by this generator
is shown :n Figure 3, a. .
-7-
CONFIDENTIAL
Sanitized Copy Approved for Release 2011/09/27: CIA-RDP80-00809A000600370626-6
 Sanitized Copy Approved for Release 2011/09/27: CIA-RDP80-00809A000600370626-6
-I
7. While the experiments made with the laboratory model of a capacitive
parametric alternator showed that the power generated increases with increased
gas pressure and that, consequently, considerable power could be generated by
following this method, they also underscored the fact that the pressure in-
crease was unavoidably accompanied by an increase in mechanical losses and also
by liberation of heat within the condenser when the condenser rotor turned in
an atmosphere of compressed gas. The last factor not only reduced the efficiency
of the generator, but also required special measures to draw off the heat lib-
erated. The use of a gaseous medium under high pressure, therefore, did not
prove technically advantageous. Various methods could be employed for the solu-
tion of this problem; we could, for example, try to use gases or vapors of
technical substances whose dielectric strength is considerably higher than that
of air. On the other hand, we could give most of our attention to decreasing
losses in the rotation of the condenser, using for this purpose a gas with a
considerably lower specific weight than air. Even more important, we could try
to find methods which would allow us to recover partially the energy released
in the compressed gas in the form of heat.
A method suggested by Academician L. I. Mandel'shtam and the author
called for the use of superheated steam as the gas under high pressure, the
specific weight of which is considerably less than the specific weight of air
(for example, at pressure of 20 atm and a temperature of 3500 C, the specific
weight of steam is only 1/3 the specific weight of air for the same pressure
and a temperature of 200 C), and the use of a steam turbine as the motor driving
the variable condenser. The superheated steam would enter.the turbine from the
chamber in which the variable condenser is located. This would be advantageous,
in that (a) the losses in driving the condenser would be considerably lower than
in air, and (b) the heat liberated in the condenser would be used to heat the
steam and thus the energy expended in it would be almost completely recovered.
An important prerequisite for this solution was that the superheated
steam have sufficient dielectric strength. Since there was no available data
on this problem, we had to set up special. investigations. According to the pre-
liminary results obtained by A. G. Rzyankin f23J, the dielectric strength of
steam at pressures of about 10 kg/sq cm was no less than that of air, other con-
ditions being equal. Since a preliminary study of the feasibility of the steam-
turbine capacitive-parametric-alternator combination had already given promising
results both from the standpoint of efficiency and design, this method of creat-
ing a capacitive parametric alternator seemed quite possible.
However, there still remained a number of design difficulties, connected
..:mainly with the problem of satisfactorily tapping the high voltage from the steam
chamber in which the rotating condenser is located. The potentialities and
regions c' application opened by this method can be judged decisively only after
proper preliminary studies are made and laboratory models tested.
8. Another possible method for increasing the power of the capacitive
generator by increasing the permissible voltage is to use a vacuum as an in-
sulator. A vacuum is in many respects an ideal insulator, and therefore en-
gineers and physicists have long searched for ways to solve this problem both
here (Academician N. N. Semenov) and abroad (the Metro-Vickers Laboratory and
others). Until recently, however, this problem remained unsolved. The main-
tenance of a sufficiently high stable vacuum in a medium containing large masses
of metal was an especially difficult task in practice, especially when these
large masses move. at high speed (as is the case in electrical machines). The
results of recent works [18, 19] on the use of a vacuum as an insulator give
reason to believe that this problem has already passed from the stage of pure
laboratory studies to the field of practical engineering use. The 100-kv elec-
trostatic generator in a vacuum (power of 2 kw) constructed by crump, operating
on the ordinary electrostatic machine principle, shows that a high-power capaci-
tive parametric alternator using a vacuum as an insulator is feasible.
Sanitized Copy Approved for Release 2011/09/27: CIA-RDP80-00809A000600370626-6
 Sanitized Copy Approved for Release 2011/09/27: CIA-RDP80-00809A000600370626-6
Judging from the voltage gradients in a vacuum (up to 1.1.106v/cm)
obtained by Trump f197, it seems possible to design a capacitive parametric
alternator which will compare in power with ordinary alternators. The advan-
tages to be gained by using a vacuum as an insulator are tremendous, e.g., there
will be no losses in the dielectric, no mechanical losses in rotation, no need
for drawing off heat, etc. All of these will undoubtedly increase the efficiency
and in addition will be very beneficial from the design standpoint.
Another advantage of a vacuum in the capacitive generator is the possibility
of obtaining very hign, almost maximum, efficiencies (above 99%). Moreover, the
electric losses in the variable condenser in a vacuum or an atmosphere of com-
pressed gas will remain low even when the frequency is increased, so that the
efficiency can be made very high even for capacitive alternators operating at
higher frequencies. Since the power of the capacitive alternator is proportional
to the frequency, its most efficient application will be in the high-frequency
field.
During operation of the generator, a constant stable vacuum (or a constant
vapor or gas pressure) must be maintained in it, and this requires special equip-
ment (air pumps, compressors, steam generators, etc.,), the importance of which with
respect to weight, size, and need for additional power will decrease as the power of
the generator itself is increased. Therefore, the capacitive generator, in contrast
to the Inductive generator, will be more advantageous the greater its power.
We should note that in the capacitive generator the voltage on the variable
condenser, and consequently on the coil connected in series with it, is much greater
than the voltage of an ordinary alternator of the same power. Therefore it might be
considered as a combination of an alternator and a step-up autotransformer. This
must be taken into consideration when the capacitive alternator is compared with
ordinary alternators from the standpoint of technology and economics.
Because very high voltages are generated simultaneously with high power in
capacitive parametric generators, they may be of importance for long-distance power
transmission. In this case, the fact that it is possible and even more advantageous
to generate very high power in the form of high-frequency currents is very important.
Since high frequencies present a number of advantages for conversion of ac into dc,
it is possible that high-power capacitive parametric generators might be preferred
over ordinary alternators as sources for'long-distance power transmission.
1, R. Rudenberg, Phys Zs, 1908, p 556.
2. M. Barkhausen, Problem der Schwingungserzeugung, 1907, p 37.
3, M. V. Shuleykin, "Self-Excitation of Electrical Oscillation in Circuits of
Dynamos With Collector Rings," Izv S-Peterb. Politekhn Institut, 1913, Vol XX.
4. N. Poinc: "^, L'eclairage electrique, 50, 1907, p 299.
5, Rayleigh, Phil Mag, Apr 1883, pp 223-224; 1887, p 144.
6. M. Brillouin, L'eclairage electrique, Vol XI, p 491.
7, B. van der Pol, Experimental Wireless, 1926, p 343,
8. H. Gunther-Winter, Zs fur Hochfr, Vol XXXVII, 1931, p 172.
I Sanitized Copy Approved for Release 2011/09/27: CIA-RDP80-00809A000600370626-6
 Sanitized Copy Approved for Release 2011/09/27: CIA-RDP80-00809A000600370626-6
Watanabe Y., Saito T., Kaito Journ of the Inst of Elec Eng of Japan, No 506,
Vol LIII, Mar 1933, p 21-
10. L. I. Mandel'shtam, Works of the Conference on Oscillations, 1931; N. D. Papa-
leksi, ibid.
11. L. Mandel'shtam and N. Papaleksi, Zhur Tekh Fiz, Vol III, No 7, 1933.
12. L. Mandel'shtam and N. Papaleksi, Zhur Tekh Fiz, Vol IV, No 1, 1934.
13. V. Lazarev, Zhur Tekh Fiz, Vol IV, No 1, 1934.
14. A. Andronov and M. Leontovich, ZhRFKhO, Vol LIX, 1927, p 430, 442.
15. B. Van der Pol and Strutt, Phil Mag, No 5, 1928, p 18.
16. N. Papaleksi, "Concerning Parametric Generation of Alternating Current,"
Byutleten' Leningradskogo Industrial'nogo Instituta, No 4, 1937.
17. A. Walter, The Physics of Dielectrics, 1932, p 269?
18. Van de Graft, Elektrichestvo, Nov 1935, p 20.
19. Trump, Elektrichestvo, Nov 1935, p 26.
20. I. Goldman, DAN, 1938, Vol XVIII, p 89.
21. W. L. Barr&, PIRE,` 934, No 22, p 210.
22. B. M. Vul and I. M. Goldman, Zhur Tekh Fiz, Vol VI, 1936, p 244.
23. A. G. Rzyankin, Zhur Tekh Fiz, Vol VIII, 1938, p 65.
24. H. Poincare, Methodes nouvelles de la mechanique celeste.
25. Iyapunov [Liapounof7, The General Problem of Stability of Motion, ONTI, 1935.
fAppended figures follow]
Sanitized Copy Approved for Release 2011/09/27: CIA-RDP80-00809A000600370626-6
 Sanitized Copy Approved for Release 2011/09/27: CIA-RDP80-00809A000600370626-6
Figure 1. Equivalent Circuit
of the Inductive Parametric
Generator
(a) Capacitive and (b)
IAN, I
rTV
7
5
I ~ xa.a ;nsrar.e
Figure 4. Characteristics of
Inductive Parametric
Generator
Sanitized Copy Approved for Release 2011/09/27: CIA-RDP80-00809A000600370626-6