US 3731237 A
A high return loss is maintained over the passband of a broadband transformer by reducing parasitic interwinding capacitance. This is accomplished by locating the windings along physically separated regions of a common magnetic core. To compensate for any loss of magnetic coupling at the higher frequencies due to a decrease in core permeability, equal portions of the physically separated windings are connected in parallel.
Claims available in
Description (OCR text may contain errors)
Ilnited States etent 1 Beurrier  BROADBAND TRANSFORMERS  Inventor: Henry Richard Beurrier, Chester Township, Morris County, NJ.
 Assignee: Bell Telephone Laboratories, Incorporated, Murray Hill, NJ.
22 Filed: Feb.3, 1972 21 Appl.No.: 223,117
 U.S. Cl. ..333/24 R, 333/24 C, 333/32  Int. Cl. ..H03h 7/38, H03h 13/00  Field of Search ..333/24, 24 C, 78,
 References Cited UNITED STATES PATENTS 2,990,524 6/1961 O'Meara et al ..333/24 R 2,648,825 8/1953 Eberhardt ..333/78 X 2,470,307 5/1949 Guanella ...333/24 R 2,632,803 3/1953 Cotsworth ..333/77 [4 1 May Ll9'73 FOREIGN PATENTS OR APPLICATIONS 649,767 8/1937 Germany ..333/77 OTHER PUBLICATIONS LecherPulse Transformer in IBM Technical Disclosure Bulletin-Vol. 2, No. 4, Dec. 1959; page 31.
Primary Examiner-Rudolph V. Rolinec Assistant Examiner-Marvin Nussbaum Attorney-R. J. Guenther et a1.
 ABSTRACT A high return loss is maintained over the passband of a broadband transformer by reducing parasitic interwinding capacitance. This is accomplished by locating the windings along physically separated regions of a common magnetic core. To compensate for any loss of magnetic coupling at the higher frequencies due to a decrease in core permeability, equal portions of the physically separated windings are connected in parallel.
4 Claims, 14 Drawing Figures PATENTEU MY 1 I 73 SHEET 1 UF 3 FIG. IC
FREQUENCY TRANSFER CHARACTERISTIC V V in PATENTEDHAY H915 3.731.237
SHEET 2 [If 3 FIG. 20
BROADBAND TRANSFORMERS This application relates to broadband transformers.
BACKGROUND OF THE INVENTION Two of the more important characteristics of a broadband transformer are its passband and its return loss, which defines the magnitude of the energy reflected by the transformer. Typically, the passband of a transformer is limited at the lower end of the band by the self-inductance of its windings, and at the higher end of the band by its parasitic interwinding capacitance and its leakage inductance. In an effort both to increase the self-inductance and to reduce the leakage inductance, it is customary to arrange the conductors in very close proximity, forming a multifilar winding, and to wind the conductors so arranged on a high permeability core. While this increases the self-inductance of the windings, thereby lowering the low frequency end of the passband, it also increases the parasitic interwinding capacitance, thus negating any improvement produced by lowering the leakage inductance. In particular, the increased capacitance converts the transformer to a complex reactive network which tends to lower the upper end of the passband, and to increase the return loss over the passband. Clearly, some means of obtaining a high self-inductance, and a lower leakage inductance without adversely affecting the high frequency characteristics of a transformer is required.
It is, accordingly, the broad object of the present invention to minimize the parasitic interwinding capacitance of transformers.
SUMMARY OF THE DISCLOSURE A transformer, in accordance with the present invention, comprises a pair of physically separated windings, wound on a common magnetic core. To compensate for the loss of magnetic coupling at the higher frequencies, due to a decrease in core permeability, equal portions of the physically separated windings are connected in parallel.
In a first embodiment of a two-winding transformer, the two windings are capacitatively coupled together. In a second embodiment, comprising an autotransformer, the windings are conductively coupled together.
It is an advantage of the present invention that the parasitic interwinding capacitance of the transformer is decreased without adversely affecting its self-inductance or leakage inductance.
These and other objects and advantages, the nature of the present invention, and its various features, will appear more fully upon consideration of the various illustrative embodiments now to be described in detail in connection with the accompanying drawings.
BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1A shows a schematic diagram ofa two-winding transformer;
FIG. 1B shows the transformer of FIG. 1A constructed in accordance with the present invention;
FIG. 1C shows the transfer characteristics of a transformer in accordance with the present invention without and with the paralleling capacitors;
FIGS. 2A, 2B and 2C show schematic diagrams of an autotransformer;
FIG. 2D shows the autotransformer of FIG. 2C constructed in accordance with the present invention;
FIGS. 3A and 3B show schematic diagrams of a twowinding transformer having any arbitrary turns ratio NzM;
FIGS. 3C shows the transformer of FIG. 3B constructed in accordance with the present invention; and
FIGS. 4A, 4B, 4C and 4D show a transformer modified to include three separately wound bifilar windings.
DETAILED DESCRIPTION Referring to the drawings, FIG. 1A shows, schematically, a simple 1:1 turns ratio transformer 10, comprising a primary winding 11 and a secondary winding 12. Typically, windings l1 and 12 are wound on a common magnetic core with a sufficient number of turns to provide the requisite impedance at the lowest frequency of interest.
To minimize the parasitic interwinding capacitance, C, shown in broken line, the windings of transformer 10 are physically separated from each other as illustrated in FIG. 18, wherein windings 11 and 12 are wound on opposite sides of a common magnetic core 23. Since the two windings are no longer in close proximity, coupling between them is dependent upon the core material and tends to fall off at the higher frequencies as the permeability of the core material decreases with increasing frequency. In order to compensate for this, the two windings are connected in parallel by means of capacitors 24 and 25, where capacitor 24 connects one end a of winding 1 l to the corresponding end b of winding 12, and capacitor 25 connects the other end a of winding 11 to the other end b of winding 12. The magnitude of capacitors 24 and 25 are selected such that their impedance is very small relative to the load impedance across winding 12 at some specified frequency of interest. For example, the transfer characteristic, (i.e., the ratio of output voltage V, to the input voltage V,, as a function of frequency) for transformer 10 is typically as illustrated in FIG. 1C. In the absence of capacitors 24 and 25, the transformer passband, as given by curve 30, extends between a lower frequency f and a higher frequency f The fall-off at the high end of the band, indicated by the dashed portion 31 of curve 30, is a result of the loss of coupling due to the decrease in the permeability of core 23.
The second curve 32 shows the added coupling provided by capacitors 24 and 25. By suitably selecting their magnitudes, the coupling thus provided can be made significant at the frequency f, at which the magnetic coupling starts to decrease. This compensation extends the passband upward to frequency f which is on the decreasing portion 33 of the net, overall transfer characteristic. The latter roll-off is due to the parasitic capacitance between turns of the respective windings, which effectively short-circuit the transformer windings, and to leakage inductance.
Thus,'separating the transformer windings provides a simple and convenient means of reducing parasitic interwinding capacitance, while paralleling the two windings provides a means of compensating for any loss of coupling due to a decrease in core permeability. The
overall effect is to improve the return loss at the upper end of the transformer passband.
It will be recognized that in those cases where one end of each winding (i.e., a and b) is connected to a common ground connection, only one capacitor (i.e., 24) is required.
While the above discussion explains the principles of the invention, their application in connection with the particular embodiment described is not of great practical importance. Clearly, a prior art transformer, wherein the windings are wound closely together, has a significant parasitic interwinding capacitance. However, it will be noted that in a simple 1:1 turns ratio transformer the primary voltage V, and the secondary voltage V, are equal. Hence, the net interwinding voltage b across the parasitic capacitance, C, is zero. Thus, while there is a significant parasitic capacitance produced by the transformer structure, the effect of this capacitance upon the operation of this particular transformer is not very great. On the other hand, there are other transformers, now to be considered, where the detrimental effect of winding-towinding capacitance is quite noticeable.
FIG. 2A, now to be described, shows a 2:3 autotransformer 40, wherein one terminal 35 is a tap which divides the transformer winding into two portions, 42 and 41, having a 2:1 turns ratio. In practice, winding 42 is further divided into two equal portions 43 and 44, and then the three equal windings 41, 43 and 44 are wound on a common magnetic core as a trifilar winding, connected as shown in FIG. 2B. It will be noted that in this transformer, the voltages across the three windings are equal and additive such that a voltage v exists across the parasitic interwinding capacitances C l and C and a voltage 2v exist across interwinding capacitance C In order to extend the upper end of the return loss characteristic, in accordance with the present invention, the transformer illustrated in FIG. 2A is first modified, as shown in FIG. 2C, by the addition of a fourth winding 43', having the same number of turns as winding 43. Winding 43 is then conductively connected in parallel with winding 43 by means of conductors 37 and 38. In all other respects, the transformers of FIGS. 2A and 2C are electrically the same. However, modified as shown, it is possible to divide the four windings into two groups 43-44, and 41-43, which can then be physically separated, as shown in FIG. 2D. Accordingly, a 2:3 turns ratio autotransformer, in accordance with the present invention, comprises a pair of two, physically separated bifilar windings 46 and 47, wound on a common magnetic core 49. Using the same identification numerals for the several windings as in FIGS. 2A and 2C, the first bifilar winding 46, comprising one group of windings 43-44, forms the equivalent of a primary winding the second bifilar winding, comprising the other group of windings 41-43, forms the equivalent of a secondary winding. As in the schematic diagram of FIG. 2C, ends a and a of winding 43 are connected, respectively, to ends a and a of winding 43 by means of conductors 37 and 38. The common junction formed by ends a of windings 43 and 43 is also connected to end c of winding 41, forming one terminal 35 of the transformer. The other end of winding 41 is the other terminal 36 of transformer 40. The common junction formed by ends a of windings 43 and 43' is connected to end b of winding 44. End b of said winding is the common terminal for the transformer input and output circuits. It will be noted that while this arrangement retains the interwinding capacitance between the two windings of each bifilar winding, it eliminates parasitic capacitance C which coupled between the largest voltage differential 2v and, hence, was potentially the most harmful.
As an example of the improvement that can be realized by the transformer structure illustrated in FIG. 2C, an autotransformer constructed in accordance with the prior art had a -30 db return loss characteristic which extended between 25 MHz and MHz. By contrast, an autotransformer constructed in accordance with the present invention had a 30 db return loss which extended between 20 MHz and MHz. (The downward extension of the lower end of the band from 25 to 20 MHz is not fully understood.)
FIGS. 3A, 3B, and 3C show how the principles of the present invention can be extended to include transformers having any arbitrary turns ratio NzM. For example, FIG. 3A shows, schematically, a transformer 50, comprising windings 51 and 52, having any arbitrary turns ratio of NzM. In FIG. 3B, winding 52 is divided into two portions 53 and 54, of which winding portion 54 has N turns and winding portion 53 has (M-N) turns. A capacitor 55, connected between one end 0 of winding 51 and one end b of winding 54, effectively connects these two windings in parallel at the higher end of the transformers passband in the manner explained in connection with FIG. 1C.
The schematic of FIG. 3B is implemented as illustrated in FIG. 3C, which shows the two transformer windings 51 and 52 wound on a common magnetic core 59, but physically separated from each other. Winding 52 can be a simple winding or advantageously, tighter coupling is obtained if at least a portion is wound as a bifilar winding. The latter arrangement is illustrated in FIG. 3C. Assuming, for purposes of illustration, that N (M/2), the (MN) turns of winding 53, and (M-N) turns of winding 54 are bifilar wound. To complete the circuit, end 'a' of winding 53 is connected to end b of winding 54. Capacitor 55, connected between end 0 of winding 51 and the junction of ends b and a of windings 53 and 54, couple the N turns on winding 51 to an equal number of turns on winding 52.
In this embodiment it was assumed that ends b' and c of the transformer windings are both connected to a common junction. If they are not, a second capacitor is connected between these two ends, as illustrated in FIG. 1B.
The above-described technique can be readily extended to form transformers having three or more physically separated windings, disposed along a common magnetic core. To illustrate, FIG. 4A shows a 1:2 turns ratio autotransforrner 60. FIG. 4B shows the transformer winding divided into four equal winding portions 61, 62, 63 and 64. FIG. 4C shows a further modification of the transformer wherein two additional windings 65 and 66, equal to the winding portions 61 through 64, are added and connected, respectively, in parallel with winding portions 62 and 63 by means of conductors 81, 82, 83 and 84.
The windings are then grouped in three groups X, Y and Z, of which group X includes windings 63 and 64,
group Y includes windings 65 and 66, and group Z includes windings 61 and 62. In accordance with the present invention, the three groups are then wound along physically separated regions of a common magnetic core 80, as shown in FIG. 4D. Identifying the winding ends with the numerals 1 through 12, the windings are located and connected as shown therein. Since all the winding portions are equal, i.e., have the same number of turns, they are advantageously wound as bifilar windings. To reduce leakage inductance, the paralleling conductors 81 and 82, and 83 and 84 are shown as twisted pairs.
SUMMARY As explained hereinabove, the high return loss characteristic of a transformer can be improved at the higher frequencies by reducing the parasitic interwinding capacitance. This is done by physically separating the transformer windings. In a two-winding transformer, the primary and secondary windings are readily separated by winding them on different portions of a common magnetic core. To compensate for any decrease in core permeability, coupling capacitors or conductors couple a portion or all of the primary winding in parallel with an equal number of turns on the secondary winding.
In the case of an autotransformer, an additional winding is capacitively or conductively connected in parallel with all or a portion of the common portion of between the physically separated windings.
Thus, in all cases it is understood that the abovedescribed arrangements are illustrative of but a small number of the many possible specific embodiments which can represent applications of the principles of the invention. Numerous and varied other arrangements can readily be devised in accordance with these principles by those skilled in the art without departing from the spirit and scope of the invention.
I claim: 1. A broadband transformer comprising: a plurality of windings, including primary and secondary windings, wound on a common magnetic core; characterized in that:
said primary and secondary windings are physically separated from each other to minimize parasitic interwinding capacitance;
said core has a decreasing permeability at the higher end of the frequency band of interest;
and in that at least a portion of a primary winding is connected in parallel with an equal portion of a secondary winding. 2. The transformer in accordance with claim 1 wherein said winding portions are conductively connected in parallel.
3. The transformer in accordance with claim 1 wherein there are two windings; and
wherein said winding portions are connected in parallel by means of a capacitor at one end and a conductor at the other end.
4. The transformer in accordance with claim 1 wherein said winding portions are connected in parallel by means of a pair of capacitors.
Disclaimer 3,731,237.--Hemy Richard Beurn'er, Chester Township, Morris County, NJ.
BROADBAND TRANSFORMERS. Patent dated May 1, 1973. Disclaimer filed July 16, 1973, by the assignee, Bell Telephone Laborator'ies, I ncorpomted. Hereby enters this disclaimer to claims 1-4 of said patent.
[Oflioz'al Gazette December 4, 1.973.]