US2470425A - Low-frequency drift oscillator - Google Patents

Description

May 17,1949. J. F. BELL LOW FREQUENCY DRIFT OSCILLATOR Original Filed Feb. 13, 1943 2 Sheets-Sheet l INVENTOR.

JOHN F. BELL BY H/s Arron/yer:

May 17, 1949. BELL 2,470,425

LOW FREQUENCY DRIFT OSCILLATOR Original Filed Feb. 13, 1943 2 Sheets-Sheet 2 T M. L T. L 3 4 m E n 6 W N I R O F T T -MA H 0.4m J H w a w w w .4. m m n m H a 4 T I m w r m m 5 9 .r A l W 8 7 m M I L m 6. F 8 M F 0. D B I y 6 F m 4 3 r I w w 0 0 I u o m m n m w m w m kfimw y 655th. 2 hkk msziumo ww fifi aqwkumQ Patented May 17, 1949 UNITED STATES PATENT OFFICE LOW-FREQUENCY DRIFT OSCILLATOR John F. Bell, Chicago, 111., assignor to Zenith Radio Corporation, a corporation of Illinois 1 Claim. 1

This invention relates to a low frequency drift oscillator having a variable tuned circuit, the present application being a division of my copending application Serial No. 475,779 which has issued to United States Letters Patent No. 2,437,345 granted to me on March 9, 1948, and assigned to the same assignee as the present application.

One of the objects of the invention is to provide an improved low frequency drift variable tuned circuit and particularly an oscillator circuit which maintains a practically constant frequency notwithstanding considerable variations of temperature.

A further object of the invention is to provide an improved stable frequency tuned circuit including a variable inductor which has a very low change of inductance with change of temperature.

A further object of the invention is to provide an improved stable frequency tuned circuit including a variable inductor in which the ratio of change of inductance with change of temperature is substantially proportional to the inductance for which the inductor may be set.

A further object of the invention is to provide an improved tuned circuit including a variable inductor in which the ratio of change of inductance with change of temperature is small and is substantially proportional to the inductance for which the inductor may be set, together with capacitor means having a temperature co-efficient, that is, a ratio of change of capacity with change of temperature, which substantially corrects for changes of inductance and capacitance with temperature to provide a tuned circuit having substantially negligible frequency drift with changes of temperature.

A further object of the invention is to provide an improved tuned circuit which can be tuned to any desired frequency within an extended range of frequencies and which will maintain any desired frequency to which it may be tuned within close limits notwithstanding considerable change of temperature.

Other objects, advantages and capabilities of the invention will appear from the following description of a preferred embodiment thereof, taken in conjunction with the accompanying drawings, in which:

Figure 1 is a longitudinal sectional view of my improved variable inductor;

Fig 2 is a side elevation thereof;

Fig. 3 is an end elevation thereof;

Fig. 4 is a perspective view of one of the clips to which the inductor coil is connected;

Fig. 5 is a plan view of the coil and a form in which it may be mounted;

Fig. 6 is a similar view showing another coil form;

Fig. 7 is a Wiring diagram of an oscillator circuit in accordance with my invention;

Fig. 8 is a similar diagram illustrating compensating means for initial drift due to heating up of the tube;

Fig, 9 is a frequency-dial marking curve of the oscillator shown in Fig. 7 or Fig. 8; and

Fig. 10 is a frequency drift-frequency curve showing the negligible drift of my improved circuit.

My invention includes an improved variable inductor which is arranged to have a relatively low drift of inductance in all its settings and in which the drift of inductance with temperature change is substantially proportional to the inductance for which the inductor is set within its operating range. I find that if such an inductor is incorporated as the variable element of a variable tuned circuit, the frequency drift has a straight line relation to temperature change and can be corrected by capacitor means having a suitable temperature coefficient.

My invention includes a tuned circuit embodying such an inductor and such capacitor means, My improved variable inductor, and consequently the tuned circuit, are arranged for accurate reproduction of tuning so that the same frequency may be attained accurately when the tuning means of the inductor is located in the same dial position.

My improved inductor is illustrated in detail in Figs. 1 to 6. The materials of which it is made are selected so that no changes of parameter result when the temperature of the inductor is raised or lowered and then returned to initial temperature. The metal parts of the inductor fulfill this requirement to a high degree. The coil forms In and II are of dielectric material selected to meet this requirement. I find that a coil form it] of steatite has satisfactory expansion properties. Likewise, the coil form ll of glass meets these requirements. Other dielectric substances have the same properties but the two mentioned have proved to be perfectly satisfactory.

My improved inductor comprises a main frame or base l2 which is preferably of metal, cast iron being a suitable material. The coil form I0 is mounted between two upstanding lugs 13 and I4 which extend upwardly from the base l2. The

lug I3 is provided with an opening l5 which receives a metal bushing l6 having a coned face against which one end of the form Ill abuts as shown in Fig. 1.

The bushing I6 is rigidly held in desired adjusted position by means of a binding screw l! which extends through a slot IS in the upper end of the lug l3 as shown in Fig. 3. The other end of the form l8 abuts against the coned end of a sleeve 19 which is slidably mounted in an opening in the lug 14. The sleeve 13 abuts against a washer 20 carried by a gasket 2| of rubber or other insulating material. The gasket 2| is held compressed by means of an internally threaded cap 22 which is mounted on a threaded nipple 28 extending from the lug 14 away from the form I.

A slug shaft 24 is slidably mounted in the bush ing l6 and in a bearing 25 carried by an annulus 2 5 of compressed rubber or other dielectric loeatedi'nan enlargement 2T intl'iesl'eeve l9 remote from the form Ill.

The slug shaft 24 is biased towards the left as viewed in Fig. 1 by means of a coil spring 28 which abuts against a nut 25 threaded on the end of the shaft 24 and against a washer 3&1 of

insulating material" which abuts against the-face of the cap 22. It is to'be noted that the shaft 24 is electrically connected to the base: l2 through the bushing IE, but the shaft is insulated from the lug l4: to avoid an electrical loop constituted by the; shaft 24 and the frame l2 which would 'r'esult'in excessive losses and uncertain operation,

The shaft 24 is arranged to be moved axially by means-of, a knob 3i which is mounted on a shaft 32 which carries a worm 33. The shaft 321s otatably mounted ina bearing 34 formed on aweb 35 which extends upwardly from the frame l2. The web 35 is provided with a bore 35 which extendsinto the opening. of the bearing 34. The

bore 36 contains-a pressure pin 3'! adapted to be The nner end of theshaft 32 is supported by means of a ball 39. received in a recess in the end of the shaft 32' and. in a recess carried by a plug which is threaded into a boss 4| carried by the'frame l2.

The plug 40 is locked in position by means of lock'nut42. The shaft 32' has a shoulder which abuts against the inner side of the bearing 34.

The worm 33 meshes with a worm gear 43 rigidly carried by a sleeve 44' which also rigidly carries; a pinion 45 at the level of the shaft 24. The sleeve 44'is rotatably mounted ona post 46 which projects upwardly'from the base l2. The pinion 45 meshes with a rack 47 formed on the right-hand end ofshaft 24 as viewed in Fig. 1. Consequently; rotation of the knob 3 effects axial movement of the shaft 2 4 in one direction or the other.

Aft the-right-hand end of the inductor as viewed in Fig; 1, the base I2 is provided with an upstanding lug 48 which has an opening through 7 which the right-hand end of' the shaft 24' extends: A key plate 49 adjustably mountedon the 48 bears against theshaft 24 to limit lateral and rotary motion thereof. Beyond the lug 48 4 a scale 50 of rectangular form and a securing nut 5|.

The longitudinal member of the scale 50 is provided with suitable divisions and numerals which are arranged to cooperate with an index mark 52 on the web 35. Also mounted on the lug 48 is anan'gular resilient clip 53which cooperates with the set screw 38 to prevent accidental loss or loosening thereof.

The coil form If! shown in Figs, 1, 2 and 5, is suitably of steatite or similar material, and it may be of generally dumbbell form. The coil 54; which is relatively short, is tightly wound in a helical grooveformed in the central part of the form l0, its ends being soldered to clips 55 which are mounted on the enlarged ends of the form I llby means-of screws 55.

To avoid excessive drift due to expansion and contraction of the coil 54 with temperature changes, I prefer to make this coil of invar wire which is copper coated and preferably also silver coated to provide a low resistance layer. The coil 54 is tightly wound upon the form I!) at a low temperature so that it is always in firm contact with its groove in the form. The expansion of the coil I0 is consequently the expansion of thematerial of which the'form [Bis made.

As shown in Fig. 6, I may employ a class form H upon which are mounted clips 51 for connection of the ends of the coil. Between these clips the coil is provided with a helical groove 58 for the reception of the tightly wound coil 54.

The shaft 24 carries an iron slug 59 and a conductive slug 60. To attain my objects of keeping inductance drift with temperature low, and keeping the drift proportional to the inductance, I employ these two slugs so as to get a large travel of the slug shaft. I use the term slug todesignate generally a body movable relative to the coil to effect a change of its inductance.

I use powdered iron of such characteristics that, upon temperature change, the inductance always returns to its initial value upon the return of the temperature to its initial value. When such iron is utilized, the permeability is sufficiently high to make the'slug of satisfactorily small diameter. The slug 59 may suitably consist of this powdered iron cemented to slug form by a suitable binder such as a thermal setting plastic. The slug 59 is suitably cemented to the shaft 24.

The slug 55 consists primarily of copper. To minimize-the drift at one. position of adjustment of the shaft 24, I provide at the outer end of this slug an annulus 6! of metal having a smaller thermal expansion than that of copper. Various metals having different thermal expansions may be used to attain any desired correction of drift which may be necessary. In the present embodiment of the invention the annulus 5! may suitably be' of invar or of other similar metal having a very small thermal expansion. The composite slug 60 is-secu-red to the shaft 24 by means of a pin 62. Annulus 6| is plated with copper and preferably also with silver, so that it has high frequency electrical properties of cooper or silver.

While I have referred to the element 50 as a slug, it must be understood that only the outer surface layer is effective electrically, and at high frequencies this layer is extremely thin. Thus, the invar annulus 6.! has no electrical function. It merely provides a support for the copper or silver surfaceand imparts to the copper or silver cylinder a much lower thermal expansion than ported. The slug 60, 6! may be hollow or it may have a core which is non-metallic. Since this slug has a conductive surface, currents are induced on this surface due to the field of the coil, which currents set up opposing fields which in effect decrease the apparent inductance of the coil. Therefore, as the slug 60, 6| is inserted further into the coil, the portion of the conductive surface lying in the coil becomes greater, and hence the apparent inductance of the coil decreases.

The shaft 24 is suitably of stainless steel. The portion of the shaft between the two slugs 59 and 60 may be plated or coated with a nonmagnetic metal of high conductivity such as cop-. per or silver, or both. Instead of so coating this portion of the shaft, I may provide a sleeve extension 63 of the slug 60 which extends up to the slug 59. When the shaft 24 is thus coated it has substantially no efiect upon the electrical qualities of the coil 54.

The slugs 59 and 6!! are in such spaced relation that as one leaves the coil 54 electrically, the other enters the coil electrically. Actually the slugs are spaced slightly less than the physical length of the coil. It is to be noted that the slug 60 is shown with a slight conical formation on its inner end, but the ends of the slugs may be varied in shape to some extent without affecting the operation of the inductor provided the spacing of the slugs is made correct as hereinafter explained.

The electrical properties of my improved inductor can best be understood with reference to a tuned circuit in which it is incorporated. In Fig. 7 I show a wiring diagram of a Colpitts oscillator including the coil 54 tuned by the slugs 59 and 89 to determine the frequency of oscillation.

The B battery is connected through a resistor 64 and through a choke coil 55 to the grid side of the coil 54, the other side thereof being connected to the plate of the tube 56 by a conductor Bl.

A voltage regulator tube 58 which may suitably be a VR-l5ll tube, is connected to the common point of the resistance 64 and choke coil 65 and to ground so as to establish substantially constant voltage on the plate of the tube 66.

The grid-'to-ground condenser 69 may suitably be several times larger than the plate-to-ground condenser 10. Last said condenser may have a value of around I00 ,LL/L farads. The condenser 10 may suitably be a ceramic condenser having a suitable temperature coerlicient, usually a negative temperature coeflicient, sufficient to correct for frequency drift due to changes in capacity in the circuit with temperature and due also to changes of inductance in the inductor including the coil 54. This inductor has only a slight change of inductance with temperature change and that change of inductance with temperature change is substantially proportional to the inductance for which the inductor is set. Consequently, this slightly change of inductance can be compensated for, together with drift due to capacity, by a condenser '19 of correct temperature coefficient.

The correct temperature coefficient to be selected for the condenser 19, can be determined empirically in the manner set forth in United States Letters Patent No. 2,371,790 granted to me on March 20, 1945, and assigned to the same assignee as the present application. In short, the method consists in installing a condenser 19 of known temperature coefilcient and determining the frequency, with one setting lOf the slug shaft at two considerably different ambient temperatures.

From the frequency change it can be determined that an additional change of capacity of determined amount in capacitor 10 would bring the drift to zero. To this change of capacity is algebraically added the actual change introduced by the condenser 10 and from this result it is a matter of elementary calculation to determine hat the condenser 10 should, for correction of ambient temperature, have a definite temperature coefficient. Correction for ambient temperature drift is thus attained by substituting a condenser iii of the same capacity and having the desired temperature coefficient.

In Fig. 9 I have shown a graph of oscillator frequency against dial markings derived from such a circuit. The part of the curve corresponding to between 0 and 12 on the dial marking corresponds to the movement of the iron slug from the position shown in Fig. 1, that is, when it is fully inside the coil 54, to a position in which it just emerges from this coil. Change of frequency has a straight-line relation to dial marking for the greater parts of its length, but adjacent the 16 mc. point, the curve tends to flatten out as shown by the dotted line H.

The slope of the curve depends primarily upon the diameter of the slug. If the slu 59 were made of larger diameter, the straight-line part of the curve between 13 mc. to 16 mc. would be steeper. The same applies to the conductive slug 60 with respect to the portion of the curve between about 17 mc. and 20 me. The diameters of the two slugs are thus correlated so that they provide two portions of the curve of the same slope.

Correct alignment of the two curves is attained by accurate spacing of the two slugs. Thus, if the slug 60 were moved slightly to the right, portion of the curve of Fig. 9 between 17 me. and 20 mo. would be moved downwardly on this figure. Correct spacing places the two portions of the curve in alignment. It is to be noted that below around 1'7 mc. point, the upper part of the curve, due to the slug 6d alone, tends to flatten out as shown by the dotted line 12. The two dotted lines H and 12 combine to form the complete straight line designated as 15.

It is to be understood that it is within the scope of my invention to construct an inductance in which the inductance varies non-linearly in a desired fashion upon adjustment of the two relatively movable parts thereof. That is, as set forth above, the two slugs 59 and 60 may be made of such relative diameters, various sections of each of the two slugs 59 and 69 may be made of different relative diameters, and the spacing between the two slugs 59 and 69 may be such as to obtain a very large variety of non-linear curves of induction upon change in adjustment of the two relatively movable parts.

In Fig. 10 I have shown the frequency drift of a circuit as described above, expressed in cycles, against the frequency expressed in megacycles. From this curve, it is obvious that the frequency drift lies within the limits of 10.01% of the frequency for a variation of i47.5 C., these limits being indicated by the dotted lines 13 on Fig. 10.

It is to be noted that when the slug 59 is within the coil 54, the frequency diminishes 20 cycles per temperature rise of one degree centigrade.

As thevironi slug, is: withdrawn fromthe coil, the frequency drift. increases slightly and.v then. it decreases until at a point. where the slug 59 is almost out of the coil, the frequency drift is 0.- As the slug 6%) is drawn into the coil, the fre quency drift decreases somewhat and then. increases. The dotted line Hi represents the frequency drift curve obtained with the slug 60 completely of copper. Since this curve provides excessive increase of frequency with temperature increase between and 21 mc., undoubtedly. due to expansion ofthe copper slug,. I modified the slug.60 by replacing a substantial part of its outer end by-the-invar element 6] and I findthat with this-invar. element,.the frequency drift between wand 21 1110., is practically zero.

It must be emphasized'that the full line curve 3.! shown in Fig. l0.can, for all.practical purposes, be. regarded as a straight line. The drift there indicated, amounting to less that cycles in L about. 15,000,000. cycles for. each degree centigrade, is negligible for all practical purposes. Thedial may be set to a desired frequency with theaidof. the Vernier scale on the knob 3| with anaccuracy of about one kilocycle. At the worst part ofthe curve shown in Fig. 10, one kilocycle corresponds to drift due to 30 C., so that the maximum frequency drift is of substantially the same order as the accuracy. of setting the dial and its effect is completely negligible. The dial can be setto a desired. position and a station of the corresponding wave length can be received with afrequency. shift of the circuit of less than 30 cycles in about 15 kilocycles for each. degree centigrade change in temperature conditions.

It has been explained. that the. condenser 10 having an appropriate temperature coefficient, substantially compensates" for drift due to changes of'capacity and inductance'due to change of ambient temperature. If desired, th tuned circuit may be also corrected for frequency changes dueto' warming up of the tube 661 This isdone in the manner" described'in-detail in my aforesaid PatentNo. 2,371,790, and it will therefore be referred to only briefly herein with reference toFig. 8.

For'this' result, I connected in parallel'with the heater element of the discharge device 66, a resistor 16, which resistor is connected to ground bya lead ll, so that both the resistor 76' and the heater of device 56; start to heat" up together when. switch 82 is'closed. Instead of the single capacitor was in Figure 7, I employ a capacitor 10" having'a capacitor 18 connected in parallel thereto through a ground connection, the two capacitors having the same total capacity as'a capacitor in of Figure 7'. The temperature-coefiicients of the two capacitors m, 18' ar such that their combined change in capacity with change in temperature is that of the single capacitor W.

The. capacitor 18 is connected to. ground through. a leacl'l l and a portion of lead ET; the lead '19 being connected to lead Tl at point 80. The lead fl is preferably much larger than lead 79'. The lead iTis arranged so that itserves as a.temperature divider, that is, it is, a heat conductor analogous to the slide wire of a voltage divider, temperature being regarded as analogous to voltage. In other words Iprovide. a temperature gradientalong. conductor Ti. At its upper end which is connected to resistor 16, the temperaturemf conductor. T! is. at a. maximum,.and atv its lower. end whichis connected to-ground, or: tothe chassis supporting. the. circuit. elements,

the. temperature is. substantially that of. the chassis. Intermediate. points on conductor. 1.! have intermediate temperatures which increase from the lower totheupper end. The position of the point. of connection. 801 on. conductor 11 determines. the maximumtemperature which the capacitor lszobtains, andthe. length of the con ductor 19 determines approximately the rate. at which. capacitorv i8 heats. up When switch 82: is closed, conductor i9, hence,.acting.as a heat conductor to conduct heat from the temperature divider lead 11 to capacitor. 18. The point 8H and the efiective. length. of conductors 79 are. deter.- mined ina manner set forth in my. aforesaid Patent No. 2,371 790 so that the capacitor 18 heats upas a. discharge device 66 heats up, and at. a. rate which compensatesv for the change. of frequency due to. the change of temperature of the device 66. during heating. With. thiscompensating system, the. frequency remains. sub.- stantially constant from the: instant when device 69 first becomes effective to the normal operating temperature of the device.

It is to be noted that on increase of temperature, the cast iron frame I2 between the post GB and the lug it expands, thus tendingto move the coil form ill to the left relative to the slugs, thus increasingthe frequency. The much lesser expansion of the material of the coil form [0 tends to have the same effect. The expansion ofthe stainless steel shaft 2'4 tends to move the slugs 59and fifl'to the left, as viewed in Fig. 1". The stainless steel shaft is selected of material which has a thermal expansion intermediate those of the base [2 and the form I9, and such that the expansion of the base l2 and the form H) are offsetv to some extent by the expansion of the shaft 26'.

The expansion of the portion of the shaft 24 between its point of engagement with the pinion 4'5 and the effective slug, that is, the slug which is cooperating with the coil 54, is somewhat less than the sum" of the expansion of the frame between the pin 46 and the abutment i 3" and the expansion of the half of the coil form I i! between this abutment and the coil. Consequently, an increase of temperaturehas the effect of pulling the iron slug 59 out of the coil 54 to a slight extent, thus tending-to increase the frequency; The expansion of the shaft 24 is selected'so that this slight net movement of the slug 59 relative to the coil 54 substantially balances the effect of the simultaneous expansion of the iron slug 59 and'the coil 54;

Since these materials have excellent expansibility and contractibility with change of temperature, the rate of change of inductance with temperature change resulting from these factors is highly constant at any setting of'the inductance.

I may, by proper-selection of thermal expansions. and-dimensions of the coil form, frame and slug shaft, make the drift of inductance with change of temperature, when the slug.59 is 00- operating with the coil, substantially zero. In practice, it is not necessary to compensate to this extent because I find'that if the inductance drift with temperature change is made small, it is substantially proportional to the inductance setting, and it can be balanced by a temperature compensating condenser. Since I prefer to. use such a condenser to compensate for driftdue to changes of capacity of the circuit with change of temperature, it is convenient to employ a condenser of appropriate temperatureicoefficientto #correct foriboth the residuald'rift of theirrductance and the drift of capacity due to change of temperature.

Expansion of the iron slug with increase of temperature tends to decrease the frequency in the case of most iron slug compositions. This effect, however, is minimized since the slug 59 is relatively small. My employment of an iron slug and a slug of conductive metal enables me to have a relatively large travel of the slug shaft, which materially reduces the drift of inductance in frequency due to temperature change.

Expansion of the coil :3 with increase of temperature tends to decrease the frequency. This effect is kept at a minimum by employing, in effect, a coil which has a very low expansibility. In fact, it has the expansibility of the form which is quite low, compared with that of most normal metals. The invar coil 54 is wound upon the form at a very low temperature and under substantial tension. This tension is retained at all times with the result that the wire of the coil is firmly seated in its groove at all times.

The tendency toward increase of frequency resulting from the relative displacement between the coil 54 and the slug 59 owing to the expansion of the coil form [0, the frame [2 and the shaft 24, is arranged so that it substantially nullifies the tendency towards decrease of frequency resulting from the simultaneous expansion of the coil 54 and the slug 59. While this result can be conveniently attained by selecting a shaft 24 of correct thermal expansion, it is obvious that the effect involves the thermal expansion of the coil form [0 and the frame I2 and the same result can be obtained by selecting material for the coil form It or material for the frame l2 so that the three elements I0, l2 and 24 will provide the small displacement of the slug 59 relative to the coil 54 necessary effectively to nullify the frequency change resulting from expansion of the slug 59 and the coil 54.

Expansion of the slug 60 tends to increase the frequency with rise of temperature. During the first part of the travel of the slug this tendency minimizes a tendency toward opposite drift; but a slug 69 completely of copper gives excessive increase of frequency with rise of temperature when the slug is more than half way into the coil. As shown by the portion of the dotted line 14 above the upper dotted line 13 in Fig. 10, the frequency drift exceeds 0.01% of the frequency for a change of temperature of 147.5 C. when more than about three-quarters of the slug 60 is within the coil. By substituting the portion Bl of low thermal expansion for part of the slug 60, it will be seen that the frequency drift when more than about one-third of the slug 60 is within the coil, is reduced to not more than 0.001 of the frequency for a change of temperature of 47.5 C., a drift which is substantially negligible.

It will be understood that although I have referred to invar as a prime example of metal having substantially no thermal expansion, it is well known in the arts that many other metals and alloys have the same or similar property, and they may be used instead of invar in the relations herein stated. It will be understood that in some cases a lesser degree of correction than that provided by the invar element 6| is necessary and that the element 6| may in such cases be made of a metal which has a substantial thermal expansion somewhat less than that of copper. It will also be understood that the length of the 10 element 6| and its position on the slug 60, 6! is determined by the position in the tuning range where the correction it provides is necessary.

It must also be understood that it is within the scope of the present invention to provide a slug such as 50, 5|, to correct for a reduction of frequency with rise of temperature, and in such cases an element 6| may be provided at the proper position on the slug 60, 6|, which has a greater thermal expansion than that possessed by the body 65 of the slug.

While it is preferred in the embodiment of the invention illustrated to employ two slugs 59 and 60, it will of course be apparent that each slug exercises its individual tuning effect and it is within the scope of my invention to employ a single iron slug 59 or a single composite slug 60, 6| to provide an inductor having a lesser tuning range.

Although the invention has been described in connection with specific details of preferred embodiments thereof, it must be understood that such details are not intended to be limitative of the invention except in so far as set forth in the accompanying claim.

I claim:

An oscillator circuit comprising a tunable inductor, capacitor means, and an electron discharge device having a heater element, said inductor being arranged to have low inductance drift with temperature changes and one which is substantially proportional to the inductance in all settings of the inductor, said capacitor means including two capacitors having temperature coefficients which compensate for changes of capacity and inductance in said oscillator circuit due to changes of ambient temperature, a resistor heater, means for supplying current to said heater element and to said resistor heater simultaneously, a temperature divider conductor extending from said heater resistor, and a heat conductor lead extending from one of said capacitors to an intermediate point on said temperature divider conductor, said intermediate point of connection of the heat conductor lead and said temperature divider conductor and the effective length of the heat conductor lead being such that heat is transferred from said resistor heater to the capacitor connected to the heat conductor lead at a rate and to an extent to provide capacity change in the circuit to compensate for changes in the frequency which would otherwise result from changes of parameters due to the heating up of the tube.

JOHN F. BELL.

REFERENCES CITED The following references are of record in the file of this patent:

UNITED STATES PATENTS Number Name Date 1,706,837 Bailey Mar. 26, 1929 1,898,792 Shermund Feb. 21, 1933 2,027,521 Drake Jan. 14, 1936 2,028,596 Franklin et al Jan. 21, 1936 2,093,331 Lynn Sept. 14, 1937 2,141,890 Weis Dec. 27, 1938 2,151,752 Ellis Mar. 28, 1939 2,325,279 Schaper July 27, 1943 2,332,868 Nowak Oct. 26, 1943 2,371,790 Bell Mar. 20, 1945 2,407,360 White Sept. 10, 1946 Certificate of Correction Patent No. 2,470,425. May 17, 1949. JOHN F. BELL It is hereby certified that errors appear in the printed specification of the above numbered patent requiring correction as follows:

Column 4, line 28, for the Word class read glass; column 5, line 65, for slightly read slight;

and that the said Letters Patent should be read with these corrections therein that the same may conform to the record of the case in the Patent Oflioe.

Signed and sealed this 18th day of October, A. D. 1949.

THOMAS F. MURPHY,

Assistant Commissioner of Patents.

Certificate of Correction Patent No. 2,470,425. May 17, 1949. JOHN F. BELL It is hereby certified that errors appear in the printed specification of the above numbered patent requiring correction as follows:

Column 4, line 28, for the word class read glass; column 5, line 65, for slightly read slight;

and that the said Letters Patent should be read with these corrections therein that the same may conform to the record of the case in the Patent Oifice.

Signed and sealed this 18th day of October, A. D. 1949.

THOMAS F. MURPHY,

Assistant Commissioner of Patents.

Images