US2491347A - Precision radio tuner - Google Patents

Description

Dec. 13, 1949 s, Y, WHITE PRECISION RADIO TUNER Original Filed Dec. 8, 1942 my.. .mh m VW Y T N T IY. MA J m .y

V.. BW 7 .m -w 4 ,mm m e ,mm mm m. w m m w MSSS Patented Dec. 13, 1949 PRECISION RADIO TUNER Sidney Y. White, Bayside, N. Y., assignor to Victor S. Johnson, Chicago, Ill'.; Victor S, Johnson, Jr., administrator, de bonis non of said Victor S.. Johnson, deceased Original application December 8, 1942, Serial No.

468,195, now Patent No. 2,451,643, dated Octoiler 19, 1948.

Divided' and this application November 21, 1946, Serial No. 711,438

2 Claims. l

The present application is a division of my pending application Serial No. 468,195, led December 8, 1942, for Precision radio apparatus, now Patent No. 2,451,643, issued Oct. 19, 1948. The present invention relates to'important improvements originally disclosed in said application. The complete original disclosure of said application is made a part of the present specilication by reference. Features of that disclosure not claimed herein are claimed in Serial No. 468,195 and in the following divisions thereof:

Serial No. 506,372, led October 15, 1943, for Radio apparatus, now Patent No. 2,407,359; Serial No. 725,685, led October l5, 1943, for Radio apparatus as a continuation of Serial No. 506,373, now abandoned; Serial No. 506,374, filed October 15, 1943, for Electrical condensers, now Patent No. 2,438,592; Serial No. 506,375, filed October 15, 1943, for Method of lining up uni-controlled tuned radio apparatus, now Patent No. 2,422,381; Serial No. 717,139, led October 15, 1943, for Method of making inductance coils as a continuation of Serial No. 506,376, both of said applications now abandoned; Serial No. 506,377, led

October 15, 1943, for Method of temperature compensating tuned circuits, now Patent No. 2,407,360; and Serial No. 711,437, filed November 21, 1946, for Precision radio apparatus.

The present invention, in common with Serial No. 468,195 and the other divisions thereof, relates to radio apparatus, and is particularly, although not exclusively, concerned with mobile radio apparatus, suitable for meeting the exacting requirements of military service on land, at sea and in the air.

The primary object of the present invention is to provide a core tuned radio apparatus in which a straight line relationship between frequency change and core movement is extended over a very wide range.

To this end it is an important feature of the present invention that provision is made of an inductance coil of solenoidal form suitable for use at ultra-high frequencies, and a tuning core assembly comprising a core supporting member operable axially of the coil, a relatively large ferromagnetic core mounted on the core supporting member, and a relatively small non-ferromagnetic conductive core mounted on the core supporting member and separated from the ferromagnetic core by a xed substantial distance somewhat less than the length of the coil.

Other objects and advantages will hereinafter appear.

being taken on the line of Fig. 3 looking inA the direction of the arrows;

Fig. 2 is a plan view of the coil assembly oi' Fig. l;

Fig. 3 is a rear end view showing the coil assembly of Figs. l and 2, together with an associated tank condenser;

Fig. 4 is a view in side elevation of the coil and condenser assembly of Fig. 3;

Fig. 5 is a bottom view of the coil and con denser assembly shown in Fig. 3;

Fig. 6 is a longitudinal sectional view showing details of the condenser of Figs. 3 to 5, inclusive;

Fig. 7 is a fragmentary view showing a composite core cooperating with the coil, the core sections being composed, for instance, of iron 'and copper respectively; and

Fig. 8 is a graph illustrating the various relationships of tuning core movement in mils to the frequency in megacycles of circuits tuned and controlled by various diameter cores.

In the transmitter and receiver originally disclosed in Serial No. 468,195 tuned circuit assemblies are employed for core tuning. The tuning means comprises axially aligned inductance coils and an axially movable rod extending through the several coils and having xed upon it, at properly spaced intervals for cooperation with the respective coils, suitable tuning core members.

In the drawing disclosure is made of a tuned circuit assembly as vemployed. in the illustrative apparatus of Serial No. 468,195. The illustrative assembly employs a cylindrical coil form and a solenoidal coil wound thereon.

Referring to Figs. 2 to 5 the means for supporting the coil and its associated condenser is shown as comprising a generally rectangular shaped plate |66 molded of ceramic insulating material and having formed therein three cylindrical holes |61 and a pair of elongated slots |68. Centrally of the block in its front and rear it is provided with arcuate shaped portions |69 and |10 from which depend short tapered tongues |1|, and between the arcuate portions |169 and |10 the middle portion of the plate is undercut in an arcuate shape as indicated at |12. The entire `plate is nished to the shape shown by a molding operation and is then baked at a high temperature.

Referring to Figs. l and 3 to 5, the coil supporting form |13 is shown as comprising a generally cylindrical shaped tube composed of the same ceramic material of which the plate |66 is formed. The coil form has a spiral shaped groove |14 ground therein adapted to accommodate the coil 35 which is herein shown as comprising a thin metallic, ribbon |15 of two turns (see Fig. 1), which may be heated when applied to the coil form, so that it may develop tension throughmshrinkage as it cools. The coil form is also longitudinally slotted as at |16, the slot being tapered to accommodate the tongues |1| so that the slot |16 and tongues |1I provide means for locating the coil form in a definite position on the supporting plate |66. A material which will glaze is applied to the portions of the coil form and plate |66 which are to be brought into contact with each other and the members then baked to glaze the material which thereupon unites the supporting blocks and coil form into a unitary rigid stable assembly. The ceramic material is preferably of such a nature that its surface .contains a large number of small particles which project beyond the general surface level and puncture the skin of the ribbon |15 in numerous places, thereby entirely preventing any slippage of the ribbon on the coil form. The result is that the coil is maintained tightly in engagement with the coil form at all times and does not change in shape due to any changes in temperature or humidity. In other words, the coil and coil form are, as it were, locked together throughout the full length of the coil and the size and shape of the coil remain at all times the same as those of the coil form. This arrangement obviates any non-cyclic variation in distance between one turn of the coil and another and also any non-cyclic variations in the diameter of the coil so that once the coil is wound, its 'inductance thereafter is not subject to non-cyclic variations due to temperature or aging. The ribbon of the coil is desirably a semielastic material such as sterling silver. Such material combines with high conductivity a softness permitting ready penetration by the coil form crystals and an elasticity capable of maintaining the required tension. Pure silver has been found unsuitable because it does not havethe required elasticity.

Referring to Figs. 2, 3 and 5, the powdered iron slug I is secured against the lower surface of the block |66 by means of a pair of screws |11 which pass through the slots |68.` The inner face |18 of the slug I5 is arcuate in shape so that it may be moved inwardlyv into engagement with the surface of the coil form |13. The slug I5 may be adj-usted for controlling the slope of the tuning curve of the oscillator. The left-hand end of the ribbon is soldered to an inwardly extending tongue |19 formed on a metallic coil terminal |80 which has a flat portion I8I held against the lower face of the block |16 by a threaded hexagon head screw |62. The width of the tongue |19 is substantially equal to that of the groove |14 in the coil form so that it engages the sides of the groove and thereby prevents the coil terminal |80 from rotating when the screw |82 is tightened up. Coil terminal |80 is also provided with a depending lug |83 whose lower edge is provided with an arcuate surface |84, adapted to engage and be soldered to a metallic cylindrical coating or thin sleeve |85 secured to the outer peripheral surface at one end of a thin tube |86 formed of insulat- 4 ing material (see Figs. 4 and 6). A similar but somewhat smaller metallic coating or sleeve |81 is provided near the other end of tube |86 and the interior of the tube is provided with a thin metallic coating or sleeve |88, so that the entire unit forms an electrical condenser.

The coil terminal |89 for the other end of the coil is similar in construction to coil terminal |80, except that its parts are reversed, and corresponding parts of the two terminals are designated by the same reference numerals. The tongue |19 of terminal |89 is secured and soldered to the other end of the ribbon |15 and the arcuate surface |84 of its depending lug |83 is in engagement with and soldered to the coating |81 of the condenser. The mid-tap 53 o f the coil is soldered to a tongue |90 formed on the center terminal |9I Whose main body portion is fiat and is threaded to receive the securing screw |92. The tongue |90 extends substantially the full width of the spiral groove in the coil form, thus preventing rotation of coil tap |9| when the screw |92 is tightened. The upper ends of the hexagon securing screws |82 and |92 are rounded ofi as indicated in Figs. 3 and 4 at |93, thereby providing switch contacts for the coil and condenser assembly. The condenser is of fixed value and is connected across the ends of the oscillator coil. The securing screws |82 and |92 form switch contacts.

At V125 mc., if we use concentrated circuit elev ments of the L-C type, the coil used can be little more than 2 inches of wire. We can, therefore, use no leads whatsoever, in a disciplined circuit, as we want all the wire possible on the coil obeying a single set of rules of expansion and vibration.

Since such high sustained accuracy is sought for, no structure or material can be used except of the most unchanging nature. Physically, glass, quartz and ceramic are most suitable and have good retrace characteristics of dielectric constant and physical size when-varied with temperature. No structure can be employed where there is the slightest possibility of any permanent change to any degree, either electrical or mechanical.

The type of tuning employed is of the core type.

The tuned circuit must, therefore, be designed with the requirements of core tuning in mind. lit is basic, however, that before we can tune the circuit over a range, the circuit Without such tuning means must in itself maintain a fixed frequency to a high order of accuracy. It must also allow trimming, tracking and aligning with a precalibrated dial having great length and accuracy of resetting. It must have no wiring at all.

The concentration of over of the inductance is actually in the coils of the tuned circuit assemblies described where it is capable of being acted on by a core.

The diameter of the coil is chosen to be about 405 mils in the present instance for use with a. 375 mil core. Considerable difficulty is had in the ceramic artin making thin walled tubes beyond a certain minimum thickness of wall. Maximum tuning ranges obtainable with core tuning are reached where the core substantially fills up the coil, but the core must still freely pass through the bore of the coil form. If we chose this same ratio with a mil core, the wall thickness would be less than 5 mils, an impracticable figure for quantity production in the present state of the ceramic art.

The coil form is made with grooves for the conductor, to have the thick lands to support the thin grooves during firing, and also to guide in the winding.

Since the conductor chosen must have high conductivity, its thermal coemcent of expansion must also be high, atleast two or three times that of the coil form. A spiral winding inherently has no strength of its own, so it must be the mechanical slave of the coil form. This means the wire must be wound under silicient tension and have enough elasticity to cling to the form at the most adverse temperature.

The cross-section of the conductor is a very thin strap, rather wide. If large, round conductors are used, such as #i4 round wire, the current tends to hug the coil form as it is the smallest diameter of the turn. Any good conductor has a` large temperature coeflicient of resistance, however, and if the temperature be raised the resulting increased resistance causes a redistribution of the current, causing the diameter of the mean current path to be increased. This markedly increases the inductance, since diameter of the current path is square in the formula for the inductance coil, and great changes in frequency result.

By using a very thin strap of the order of three mils in thickness, this effect is minimized and a disciplined current path results. Instead of using pure silver, sterling silver is used for greater vtoughness and elasticity and may be wound on the form quite hot by passing a heavy current through it while winding, in which case it shrinks on the form'. Tension may be used also, suflicient to stress it nearly half way to its elastic limit so it hugs the coil form like a rubber band.

It isof great advantage to use ceramics of the low loss type such as Alsimag 196 because of the presence on the surface of minute sharp crystal structures which apparently pierce the skin of any unhardened metal pressed firmly against them. Repeated temperature cycling of these coils from -40 to -l-21'7 F. show no creepage of the winding, since each unit length is captured by its adjacent crystals and held firmly in place.

'I'he length of coil chosen must depend in part upon the tuning curve desired and upon the length of core travel most easily obtained with a desirable dial mechanism. A coil 375 mils long, measured center of winding strap to center of winding strap gives an active core movement of about 250 mils for 25% tuning range.

In any coil to be used with. a core the inside of the coil form must be left free to pass the core. Most methods of terminating coils use rivets, eyelets. or passing the conductor through holes in the form, all; of which would interfere with core movement. Some structure outside the simule cylindrical coil form' is. therefore, reouirefl. This takes the form of the plate or block |66 with its associated terminal blocks |80, |89 and |9| (see Figs. 3, 4 and 5).

The block |66 is preferably glazed to the coil form. Plastic cements are undesirable because of cold iiow and change with age, but a good glaze in the joint fired at 1700?, lik-really makes the two pieces unitary.

Plate |66 allows use of massive structures such as blocks |80 and |89 to be employed to give a rigid and definite termination of the inductance at either end. These blocks are given large crosssection so that they will have a minimum possible inductan'ce, and the tongues |19 provide exact termination of the inductance wound on the form, in that the take-ofi of the current is normal to the axis of the coil. Each tongue |19, being the full width of groove |14, provides a rigid nonturning structure when the contact screws |82 are tightened up. Shaping of these blocks to include the turned up portion |83 (Fig. 4) allows a cylindrical type of condenser to be used for tuning the circuit.

A number of assemblies have been assembled and tested using the condenser of Fig. 6, which is a commercial form where the capacity may be formed between the inner plate |88 and the two bands |81 and |85 forming two condensers in series, or the outer band |85 may be continued around the end of the hollow cylinder joining the inside plate |88l forming a single condenser hrough the ceramic body |88.

When this condenser is laid in the cradle formed by the connecting blocks |03 it will be seen that an absolute minium inductance return path closing the physical separation between the ends of the coil proper has been achieved.

It is found to be a considerable advantage in this self-contained structure that rounded contacts |93 can be used as a switch in the case of multi-band apparatus. There is a real problem in switching ultra-high frequency circuits where the switch is placed within the tuned circuit. A coil in the broadcast band may easily have an R. F. resistance of 5 ohms, or 5000 milliohms. A satisfactory commercial type of small switch may have contact resistances of 5 to 40 milliohms. which is negligible in proportion to 5000 milliohms. A two-turn coil such as shown in Fig. 4, however, may have a total R. F. resistance in the entire tuned circuit of only 40 milliohms, and consequently the contact resistance of any practical form of switch, which of necessity must be small because of the small physical dimensions of these circuits, becomes a substantial portion of the total resistance. It is an advantageous feature of the present invention that each coil carries its own tank condenser with it, allowing-switching` of the charging current to the electrodes of the tubes' only, a much easier matter.

Provision of these contacts also allows desirable slipping of the whole tuned circuit assembly J axially.

Provision of the plate |66 also allows for the use of a solid block trimmer I5 as shown in Fig 3.

In a practical receiver using this type of gear, the problem of injecting the oscillator voltage into the mixer circuit is desirably accomplished bv coaxial mounting of the two assemblies, a distance between them being chosen to give the desired amount of oscillator injection into the mixer grid circuit, for instance.

Tracking between the oscillator and R. F. circuits may be accomplished by any one of several means, for example, by selection of the diameters of the cores for tuning the circuits. One circuit tunes over a greater percentage range than the other circuit. depending on whether the oscillator is run at a higher or lower frequency than the signal. The ratio of the two diameters of the cores is a function of the frequency separation desired to give the necessary intermediate frequency, which may well be anywhere between one megacycle and twenty megacycles.

The tuned circuit assembly shown makes provision for a single unit that has in effect fastening means. tuning means. switch, tank condenser, trimming, tracking and aligning means in a single simple structure, so that all the frequency determining elements are well within a cubic inch,

and under temperature, vibration and shock, all

travel together. There is no iniiuence of the in frequency to avoid confusion.

7 chassisuponthefrequency. Thereisthusprovided asingle universal unit that can be usedfor transmitter, receiver, wave trap, or any of the numerous uses to which tuned circuits can beput.

The Q of these assemblies is found to be quite high without'the core. If measured in air without any associated apparatus, the Q is about '100. When measured in the coil holder and with an oscillator tube assembly attached andA with the tube in place but not lit, the Q exceeds 400.

The temperature effect of the core is a difilcult thing to handle, since the amount of frequency which is attributable to the core varies from a small amount to quite a large amount. It will thus be seen that the core could be compensated for thermally by the choice of a suitable temperature coefficient condenser only at one core position, and'consequently only at one frequency. It is, therefore, highly desirable that the core have almost no temperature coeillcient of its own, as otherwise either incomplete or very elaborate compensating means would have to be used. Cores of the ferrous oxide type, while having very low losses at frequencies in excess of 100 mc., have marked temperature coeiilcients of both permeability and losses. Certain cores of the carbonyl produced type, however, are found to have a very high order of stability, both in regard to losses and permeability, when used at these frequencies. Great care is 'necessary in insulating and binding these spherical particles together, and particles must be chosen having differentiated internal structure, such as being formed of a plurality of concentric shells. The losses are markedly higher than those in the ferrous oxide type, but the cores are still decidedly usable. Total thermal drifts of to 20 kc. at 100,000 kc. can be reproducibly obtained with this type core, when heated to several hundred degrees.

There are numerous advantages in having a straight line relationship between the amount of displacement of a core which controls the frequency of a resonant circuit and the resonant frequency produced. For example, it may bedesired to cause an increment in frequency of exactly one megacycle with a core movement of exactly 10 mils (.010"), and to have this relationship hold over as wide a tuning range as possible as, for example, from 110 mc. to 135 mc.

Among the advantages of such an arrangement are that the use of an essentially linear dial and movement (not shown)is possible, and that it permits one resonant circuit to be readily tracked, with another in cases where a definite relationship, such as a constant frequency difference, is to be maintained between the resonant frequencies of the two circuits. In tuning a circuit over this frequency range it has been found practically 4desirable, because of physical limitations, to arrange for a total core movement of the orderv of V4 inch, as giving the best balanced design in this instance.

Fig. 8 shows a generalized series of curves for .the purpose of discussing design parameters in connection with the straight line frequency relationship. All curves except A have been displaced If we consider the case where the coil is of the construction shown in Fig. 4, and if the coil is essentially square, i. e., its diameter and length are about the same, and if this diameter be about .410 inch, then we may make up a series of cores of different diameters and observe their effect A core the diameter of a pencil lead, for instance, would give us curve E of Fig. 8, an S-shaped curve with nor approximately straight section in the middle. Curve D represents a somewhat larger core with greater tuning range, but still no substantially straight portion in the center. A somewhat larger diameter gives us curve C having a small apparently straight portion, while further increase gives us curve B with a substantially larger central straight portion.

Curve A represents the largest practical diameter that can be conveniently produced commercially without forcing us to thin the walls of the coil form beyond practical limits. This might have an approximately straight line frequency over a tuning range of about 17% in the frequency range shown. This shows the principle that the last few percent in tuning range increase that we obtain by steadily increasing the diameter of the core shows up as substantially straight line frequency (SLF) A coil stretched out over too much length is in general undesirable as the ends are too far apart physically, forcing' the use of a long return path on the outside of the coil which gives us unwanted inductance in the tuned circuit, since only that portion of the wire in the circuit which is wound on the coil form will be affected by the core used, and any external inductance will lessen the tuning range and consequently the SLF range may possibly reach.

The iron cores we have available at present that can be used in these frequency ranges consist in general of two types, the oxide type, and

-the carbonyl type consisting of spheres of the order of 5 microns in diameter, and often having differentiated internal structure. Either type, when insulated and compressed into a core by the use of a binding agent, has an apparent permeability of the order of 2 when used in circuits of this nature, and we often wish to extend the SLF portion of the curve to the maximum possible to be reached with this iron.

Curve A represents an actual case of the use of such cores, and as seen it covers a range of about 17%' Whichis substantially straight line frequency. It will be noted that the abscissa shows a total core movement of about l@ inch or 500 mils,` and that the SLF portion covers approximately 200 mils in the center of the curve.

There are a minimum of four effects taking place as a ferrous core is introduced into such a coil for producing the curve A shown in Fig. 8.l

'I'hey are; the true permeability effect alone; the

eddy current eect which tends to cancel the .permeability eil'ect; the fact that the core is a capacity by itself -and is also effective in gradually capacity coupling one part of the circuit to another; and a resistance loss which is present in all apparatus. The curves shown, therefore,

are the result of summation of these four effects.

It has been found possible to extend the approximate SLF tuning range of curve A, Fig. 8, somewhat by the use of a composite core in a spiral coil as shown in Fig. 7. This core comprises a relatively large ferromagnetic core 241, a ceramic spacer sleeve 248 somewhat shorter than the length of the rod and a relatively small core 249 of suitable conductive, non-magnetic 9 their spacings may advantageously be as shown in the drawing.

If a large diameter ferromagnetic core were I employed, the curve A of Fig. 8 would be produced as the core is withdrawn from the coil. With the core fully inserted in the coil the inductance would be a maximum and the frequency of the V substantially fixed capacity circuit would be a minimum. As the core is withdrawn, the curve is initially of slight but gradually increasing slope i until the maximum slope is attained, and then the slope continues substantially uniform until in the final portion the slope gradually diminishes. As shown in Fig. 8, this operation covers a range from 100 megacyclesto approximately i 140 megacycles.

If instead of using the ferromagnetic core a non-ferromagnetic conductive core had been l separately employed. the eect of this core would have been opposite to that of the ferromagnetic core. Where the curve A is produced by withdrawal of the ferromagnetic core, a curve of like l character is produced by insertion of the con g ductive core. At the 140 megacycle level, which has the normal frequency of the circuit with the ferromagnetic core entirely withdrawn, the insertion of the non-ferromagnetic conductive core would result in a curve of the character of one of the curves B, C, D or E depending upon the 'i can be started into the coil during withdrawal of the ferromagnetic core just as the slope of the curve A begins to diminish at its upper end. The frequency increasing effect obtained by introduction of the. conductive core is added to thel l diminishing frequency increasing effect obtained pensate for the diminution of the latter and thereby to extend the straight portion of the curve A by a small but important amount. It will be noted that the lower end of curve C, for example, is of increasing slope, while the upper end of curve A is of diminishing slope. The spacing is chosen to cause the conductive core to begin to be effective just at the point where the slope of curve A begins to diminish and the conductive coreis made of such size as to cause the beginnlngof -its characteristic curve .to produce the compensating effect described.

I have described what I believe to be the best embodiments of my invention. I do not wish. however, to be conned to the embodiments shown, but what I desire to cover by Letters Patent is set forth in the appended claims.

I claim: l.. In a tunable radio circuit, in combination. an inductance coil of -solenoidal form suitable for fby withdrawal of the ferromagnetic core to comuse at ultra-high frequencies, and a tuning'core assembly comprising a core supporting member mechanically operable to different settings axially of the coil, a relatively large ferromagnetic core mounted on said core `supporting member in fixed relation thereto for-movement intc and out of the coll through one end thereof, and a relatively small non-ferromagnetic conductive core mounted on the core supporting member and separated from the ferromagnetic core by a fixed substantial distance somewhat less than the length of the coil for movement into and out of the coil through the opposite end thereof, the construction and arrangement being such that an approximately straight line frequency relation between the motion of the assembly and the frequency to'which the coil is tuned by the cores is extended over a range greater than that attainable by either core alone.

2. In a tunable radio circuit, in combination, an lnductance coil of solenoidal form suitable for use at ultra-high frequencies, and a tuning core assembly comprising a core supporting member mechanically operable to different settings axially of the coil, a ferromagnetic core mounted on said core supporting member in fixed relation thereto for movement into and out of the coil through one end thereof, and a non-ferromagnetic conductive core mounted on the core supporting member and separated from the ferromagnetic core by a fixed substantial distance somewhat less than the length of the coil for movement into and out of the coil through the opposite end thereof, said cores being of different lengths and diiIerent diameters, the construction and arrangement being such `that the approxil, mate straight line frequency relation between the motion of the core assembly and the frequency to which the coil is tuned by the coresls 4o.

extended over a range greater than that attainable by either core alone.

SIDNEY Y. WHITE..

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

A l, Number UNITED STATES PATENTS

Images