US 1910957 A
Description (OCR text may contain errors)
May 23, 1933- F. B'. L LEWELLYN REACTIVE ELEMENT HAVING STABLE TEMPERATURE REACTANCE CHARACTERISTICS Filed JamA 15,' 1932 2 Sheets-Sheet l /NVENTOR .5. LEM-LVN y A T TORNE V May 23, 1933. F `B LLEWELLYN 1,910,957
REAGTIVE ELEMENT HAVING STABLE TEMPERATURE REACTANCE CHARACTERISTICS Filed Jan. 13, 1952 2 Sheets-Sheet 2 FIG. 5
A TTORNEV Patented May 23, 193:.l`
UNITED STATES PATENT, OFFICE FREDERICK B. LLEWELLYN, 0F MONTCLAIR, NEW JERSEY, ASSIGNOR T0 BELL TELE- PHQNE MBRATORIES, INCORIdORATED, OF NEW' YORK,
NEW" YRBL N. Y., A. CORPORATION 0F REACTIVE ELEMENT HAVING STABLE TEMPERATURE-REACTANCE CHARACTERISTICS Appncanpn mea January is, i932.V serial No. 586,266.
This invention relates to reactive circuit elements and particularly to means :tor predetermining the eliect of variation oit temperature on the reactances of such elements.
It is an object of the invention to stabilize the reactive properties of such elements under conditions of temperature change, that is, to insure that their variations of reactance with temperature change will proceed in accordance With a, desired definite and normal physical law. In practical embodiments of the invention, the temperature coeflicient of reactance may be either substantially zero or have a definite finite value.
It is another object of the invention to achieve adjustment of the temperature co- 'efcient olf such reactances by means vvhich are simple, systematic, capable of uniform y and gradual variation, and which have a minimum ell'ect on other characteristics of the reactive elements.
In the cases of both inductors and condensers of conventional types the reactance varies in one direction with a change in an axial dimension and in the other direction by a change in a transverse dimension. In an inductor, these dimensions are-respectively the length and radius of the induc` tance coil, or more specifically, the length andcross-section of vthe coil. Ina condenser, these dimensions are respectively the interelectrode spacings and the radii or areas of the electrodes.
Consistently with this physical law, the invention is expressible in two generic forms, each of which may, in turn, be divided according as the principle is to be applied to an inductor or to a condenser. Y
In one form, the variations of the two dimensions with changes in temperature are caused, by simple mechanical means, to mutually compensate to any reasonable degree so as to result in a desired overall temperature coeicient of reactance.
In the other form, lthe reactances are made stalole by insuring either a relative invariability of thedimensions aHecting the conditions of temperature, as by integrally securing the coil or electrodes to a support which yhas a substantially zero coeliicient of expansion with change of temperature, or which. has a definite coeicient of expansion with changefof temperature. The use of fused quartz has been foundto be particularly effective, especially when the coil or electrode is in the form of a thinv conductive coating secured to the quartz With an intimacy such as would result, for example, from the use of a cementing or a sputtering, process.
F or va better understanding of the invention, reference is made to the following detailed description taken in connection with the accompanying drawings in which:
Figs. l and 2 illustrate, respectively by a broken perspective and by a detail of one element, an inductance so mounted as to have a temperature-reactance characteristic which may be easily varied as desired;
lig. 3 illustrates an inductance coil which is so mounted as to be invariable in the dimensions which affect its reactanceso that the coil is immune from the effect of react-V substantially immune from the effects of erature variations on its react'ance.
Tue structural form of the inductance coil organization illustrated by Figs. l and 2,
whereby thev variationsof` the 'two signifi- 4cant dimensions with changes in temperature are caused to mutually compensateto any reasonable degree, is predicated on the following formula expressing the law of variation of the inductance of a single layer solenoid coilvr in terms of such variations of dimension.
length term exerts a preponderating effect. c
For instance, if the ends of the coil were held a iixed distance apart it is apparent from the formula that ,the coeflicient of inductance of the solenoid coil would be approximately twice the linear coeiiicient of Vexpansion of the conductor constituting the coil since the -`terin is small and since the variation in the radius of theV coil is of course a direct function of the change in linear length of the conductor making up the coil. l
Now, imagine that instead of being fixed 'the two ends of the coil are fastened to a form which has a coefficient of expansion equal to twice that of the conductor. With this arrangement, the' overall coeiiicient of the coil will consist only of the coefficient corresponding' to the small correction factor l since the remaining two Iterms completely mutually compensate for each other. Therefore, a suitable selection of the mateyrial of the form, as determined byits temperature coeiiicient of expansion, will yield a coil whose overall coefficient is zero. Analogouslythe coeiiicient may be, within reason, somewhat greater than zero in either direction.
N ow, Vconsistently with the disclosure in Figs. l and 2, imagine the form on which the conductor lis Wound to consist of a cylinder cut into fourquadrants or the like so arranged with expansion bushings that they ma be moved radially to press against the hehx formed by the conductor wire. When the. form exerts no pressure on the helix, which may be substantiallyy realized by rinterposing a thin sheet of sponge rubber or the like between' the formand the wire,
there is the condition that the coefficient corresponding to the first term of the for- C mula at the right of the equal sign, is a function of the linear coefficient of the con' is large it is thus quite possible to securean overall coefficient which is negative.
-, Nowlet it be imagined that the bushings are actuated so as to movethe quadrants outwardly until the form presses tightly against the helix. If the coefiicient of the i form is larger Vthan the coefficient of the conductor, a conditionr/which is easily realizable in practice, the dimensions of the coil will be strictly a'function ofthe dimensions of the form as effected by its coeli'icient of expansion, the conductor being stretched an amount governed by the increase in dimensions of such form. Under these conditions, consistently with the relations given in the formula, the overall coeilicient may be expected to be positive.
Adjustment of the between these two limiting conditions will correspondingly adjust the overall teniperaturel coefficient over the range including both` negative and positive values. In an experimental coil it was found that the coef-v icient of inductance could be varied between +60 parts in a million and 30 parts in a million per degree centigrade. perimental coil was placed in an oscillation circuit with a condenser which had anegative coelicient of about 50 parts in amillion. Without great difliculty it was found possible to obtain an overall coefficient for the' oscillator whose frequencyfwas deter.- mined bythe natural frequency of said oscillation circuit which was very small.
An inductance coil organization as last described is illustrated in Figs. 1 and 2. The coil is represented b reference numeral `l, which coil is Woun `ona forni which is comprised principally by the four quadrantal masses (herein to be called quadrants) indicated by reference numeral 2. Since these quadrants are adaptable for movement bushings for pressures.V
The eX- relatively to the coil, the coil vis separated therefrom by a tubing, made of hard rubber or the like, 3, on which the coil is iinmediately Wound and which tlierefore'separates the. quadrants from the coil and insures that their movements do not result in abrasion Between the quadrants and the tube 3 is a layer of sponge rubber or the like 4 through which pressure may be transmitted from the quadrants to the inside surface of the hard rubber tubing and thence to the coil itself. In order that the dimensions of the hard rubber tube 3 may accurately reflect theniovements of the quadrant in a it functions only as a ,wear member for the coil, said tube is slotted at 5.'
The quadrants are actuated in a radial direction bycorrespondingly shaped cam members 6 which in turn are responsive to the screw-fed movement of frustum members 7, which are constrained to both move simultaneously toward or away from the center ofthe form by screw 8, which isoppositely threaded at the respective extremities to insure this relative movement. The screw is propelled by knurled head or the t like 9. The cam members vare constrained to a fixed axial position by engagement of a lug 10 at the mid-section of the-screwmember in slot 11 of such cam members. Fig. 2 illustrates better than the broken perspectiveof Fig. 1 the relation of the quadrants and their actuating cam members.
It is evident thatthe screw, frustums and cam members cooperate effectively as an eX- pansible bushing organization. The mass 'of the quadrants ma be ,made large as com- .L pared with that o all of the other elements taken together andv therefore to be more significant in respect to the effect of temperature-responsive expansions of the form as a whole. It ispresumed that the movement of the quadrants is small and that the velastic reactiony of the coil itself with that of the sponge rubberand hard rubber tube is sufiicient to restore the quadrantsto such a position as to accommodate them to the positions of the frustums when propelled out" wardly, that is, when moved so as to relieve the coil of stress.
Fig. 3, illustrates the alternative type of temperature-stable inductance which is distinguished from the one disclosed in Fi s. 1 and 2 by the fact that the dimensions o the g inductance coil are made invariable withv changes in temperature. Of course, in the instance of this alternative type the temperature-reactance coefficient, for a given frequency, is substantially 'zero rather than being adjustable to a desired value., The invention disclosed in Fig.
3 depends for its eflicacy on the fact that sed quartz has a temperature coefficient which equals substantially zero, specifically 0.5 part in a million per degree centigrade. Y The inductance coil 12 is in the form of a very thin conducting strip so attached to the fused quartz tube 13 as to partake of its changes of dimensions or gure .to the exclusion of any change in dimension of the conductor individual to itself. Experiments have indicated that it is quite possible to attach ponducting material (speclfically copper) to a' fused quartz form by the sputtering proeessxso that the resultant conductor is thin enough to exert no appreciable force on thev quartz. An inductance coil so mounted would satisfy the above conditions and thereforerwould haveclosely the temperatureco-y be constituted by a solid fused quartz cylinder. The conductor could even be fastened to the surface of a fused quartz plate in the form of a spiral.
The sputtering process, which is one process by which a conductor may be so intimately associated with the quartz form asvto satisfy the conditions, is old and well known in the art. Instances of patent disclosures in the prior art of this or related processes are Burtis 1,692,074, January 20, 1928; Hitchcock 1,790,148, January 27, 1931 and Hartley 1,565,566, December 15, 1925. Similar disclosurel of`the process may be found in the Journal of the Franklin Institute, vol. 196, 1923, page 751, and in the Journal of the Society of Scientific Instruments for June 1930, page 193. Alternatively to the use of the sputtering process or'other analogous processes of metal deposition, an electroplating process may be used with like effect, this, lof course, necessitating initially providing a conducting course of some kind as by one of the deposition processes.
Fig. 4 discloses a condenser having a stable temperature-reactance coefficient and one which may easilybe varied` between reasonably spaced limits of opposite sign. The coefficient may be adjusted, as in accordance with a method somewhat analogous to that described in connection with Figs. 1 and 2.
In U. S. patent to Gabriel and Thurston 1,722,083, granted July 23, 1929, there is disclosed a relation involving the dimensions of the significant elements of a fixed air condenser and its temperature coeflicient so as toenable one, knowing the temperature coefficient of the respective materials used, to closely predetermine the temperaturecoeflicient of the condenser as a whole.
v ,In such relation theftemperature coefficient of lthe material-of the spacers is a significant item. The condenser disclosed in Fig. 4 may be considered, accordin to one aspect, as a.. condenser of they Ga riel-Thursvton type provided with a convenient means `:for varying at will the temperature coefficient ofthe aggregate of spacers. The condenser is constituted by condenser plates 111V alternating with separating insulating members 15, the whole being confined between end. plates 16 by'bolt 17, washer 18 and nuts 19. The condenser assembly is, therefore, basically the same, as so far described, as conventional forms of drystack condensers. I
An important feature of the uinvention'is the bi-metallicnature of the bolt 17. If the two portions, which are axially aligned, are of materials having different temperature coefficients of expansion, perhaps one being of one sign and the other of the opposite sign, the assembly is such as to permit this holding4 bolt to be adjusted, by backing off one nut 19 and turning up the other, relatively to the condenser plate s asto interpose in expansion-significant relation to said condenser plates the desired relative portions of the bolt. ln this way, the temperature coeiiicient of expansion of the bolt, s0 Jlar as concerns the condenser plates, may be adjusted between the two limits corresponding to the temperature coeiicients o1? the `respective portions et the bolt.
As simulating the Gabriel-Thurston condenser, the invention provides a means for edecting a temperature coeiiicient adjustment that could be effected in the Grabrielr.Thurston condenseronly by choice et spacer material. l-'loweven regardless of the temperature coeilicient ot the other parts of the condenser, the oi-metallic bolt does at least enable an adjustment of aA condenser element which has significance as to the temperature coefficient of the combination as a whole, and, in this aspect, the invention may be used more generally than in the Gabriel-Thurston organization.
lt is a condition necessary to the effective functioning of the bi-metallic bolt that the elasticity oi the combination of plates and spacers be small enough so that the change in axial dimension of the condenser is governed substantially entirely by the change in length of the bolt. rhis condition may most easily be satisiied it the condenser plates and spacers are put under considerable compression by nuts 19 and washers 18. This perhaps is the principal reason for the extension of the spacers l5 nearly to the edges of, the condenser plates.
Figs. 5 and 6 illustrate a substantially zero temperature coefficient condenser' ot a type analogous to the zero temperatureY coefficient inductance coil illustrated in l41 ig. 3. That is, the condenser plates or electrodes are constituted by thin metallic coatings on fused quartz plates, the coatings being so intimately associated with the plates as to insure that the coating has no Variation of dimension specific to itself but partakes wholly'of the characteristics of the plate on which it is mounted as to change in dimension. A lcondenser so constituted therefore has the same temperature coeilicient as that of the plate.
The coating may beunited with the plate by any of the methods described in connection with Fig. 3, the sputtering process perhaps being of most general availability.
Fig. 5 4illustrates theA organization of a plurality of units, each unit being illustrated in Fig. 6 in which the reference numeral 20 indicates the coating and reference numeral 21 indicates the plate, the corresponding element in Fig. 5 being similarly numbered. The various units are shown as build-up into the resultant multi-unit condenser much after the manner of conventional drystack condensers, and somewhat after the manner of the condenser of Fig.-
4. As being subject to physical variation to adapt it to special environments in Ways not significant for the present purpose, the multi-plate condenser as a whole is shown devoid of much of the detail that might otherwise be illustrated. lt is noted, however, that the unit illustrated by Fig. 6 does not constitute a complete condenser, being rather a constructional unit. Rather, each quartz plate with its own coating and the coat-ing ot the next adjacentA plate constitutes a condenser unit. As illustrated in Fig. 5 (and analogously in Fig. el) the condenser as a whole is adapted to be connected into its associated circuit by electrically connecting alternate sets of electrodes together7 each set effectively constituting the corresponding electrodes o'I' a like number of parallel unitary condensers, each electrode therefore functioning with respect to two such unit condensers.
lVhat is claimed is:
l. A reactance element comprising electrical structure, a support therefor, and
means attaching said electrical structure to said support so as to tend to cause it to change all or" its electrically significant dimensions in accordance with the temperature-responsive changes in dimensions of the support, the relative elasticities of said structure and said support being such that the structure is readily coerced by the support.
2. ln a reactor, in combination, mounting means whereby a given dimension of the re* actor varies with temperature, and means associated with the rst mentioned means whereby a dimension transverse with respect to the iirst dimension varies with temperature, said two means having such relative variation ,characteristics as to mu' tually compensate said changes of dimen sion to a desired degree depending on the desired overall temperature coeicient of reactance for a given frequency.
3. A reactance element com rising electrical structure, a support there or, said support having a fixed and known temperature coeeient of expansion, and means integrally attaching said electrical structurek to said support.
4. The combination specified in claim 2 in which said reactor is a helical inductance coil.
5. The combination specified in claim 2 in which said reactor is a condenser.
6. A temperature-reactance stable induc- I tor comprising/a linear conductor, a support therefor having a fixed and known temperature coefficient of expansion, and :means unitarily attaching said vconductor `throughout its linear extent to said support, the relative elasticities of the conductor and support being such that the dimensions of the conductor in a plane parallel to the surface of the support are responsive substantially wholly to the temperature-dimensional changes in the support.
7. A temperature-reactance stable condenser comprising electrodes, a sufpport for each said electrodes having a xed and known temperature coeilicient of expansion,
l5 and means unitarily attaching each said electrode at ever portion of at least one of its principal surfaces to its support, the relative elasticities of each electrode and its support being such that the dimensions of each said electrode in a plane parallel to thesurface of the support is responsive substantially wholly to the temperature-dimensional changes in the support.
p 8. The structure 4specified in claim 3 inv which the support principally comprises a mass of fused quartz. l
9. The structure specified in claim 6 in which the support comprises a mass of fused quartz. p
10. The structure specified in claim 7 in which the support principally comprises a mass of fused quartz.
v p In witness whereof I hereunto `subscribe my name this 12th day of January, 1932.
i FREDERICK B. LLEWELLYN.