Background of the Invention
The invention relates to low noise miniature transformers that are useful in hybrid integrated circuits, and more particularly to planar transformers having multiple, series-connected primary windings and a non-overlapping secondary winding imprinted on various layers of a circuit board so as to minimize electromagnetic radiation and electric field coupling between primary and secondary windings, with a ferrite core formed by two ferrite sections, each including a thin, flat plate separated from the other by two spacer sections and also supported by the circuit board.
There are a variety of electronic products which are powered by rechargeable batteries. The functional operation of some electronic products, including some powered by rechargeable batteries, is made unreliable by the presence of electrical noise. FCC regulations limit the amount of electromagnetic interference that can be radiated or conducted to the AC main power line by electronic devices. In some products, the presence of electrical noise causes sounds which are distracting to the user. Portable cellular telephones are an example of a noise-sensitive product. Some such products contain integral battery chargers that convert AC line voltage to low DC voltage levels at which conventional integrated circuit components can operate. It is important that the integral battery charger not introduce sufficient electrical noise to interfere with the reliable operation of the portable cellular telephone, not interfere with other products sharing the charger line voltage, or not violate FCC standards which are applied to electrical products to protect consumers from potential problems. There is an unmet need for a smaller, high efficiency, high frequency, low noise transformer that can be incorporated as part of an AC-to-DC converter or other circuit in a small hybrid integrated circuit package, such as a DIP package, or in a power connecter assembly. Hybrid electronic integrated circuits often are constructed using transfer molded packaging techniques to produce low cost electrical/electronic functions in a single conventional package, such as a dual-in-line package (DIP). Inclusion of magnetic components such as transformers in hybrid integrated circuits has always presented a major challenge, because the transformer cores required usually have large cross-sectional areas. Such large cross-sectional areas are inconsistent with the need to provide circuit functions in small packages.
Summary of the Invention
Accordingly, it is an object of the invention to provide an improved, reduced size, hybrid integrated circuit transformer that is particularly suitable for low noise applications.
It is another object of the invention to provide a hybrid integrated circuit transformer that can be included on a small printed circuit board as a component of a complete battery charger circuit and is easily encapsulated in a small power connector.
It is another object of the invention to provide a miniature integrated circuit transformer which can be encapsulated within a package having a thickness of as little as approximately 0.2 inches.
It is another object of the invention to provide an improved planar hybrid integrated circuit transformer which can be incorporated in a thin conventional DIP package or the like.
Briefly described, and in accordance with one embodiment thereof, the invention provides a planar transformer assembly including a first insulative layer having a top surface and a bottom surface, a first spiral winding on the top surface, circumscribing a magnetic flux path area, a second spiral winding on the bottom surface circumscribing the magnetic flux path area, generally concentric with the first spiral winding, in non-overlapping relation to the first spiral winding, and a ferrite core assembly including a first core section including a thin, flat first ferrite plate and a second core section including a thin, flat second ferrite plate. In one embodiment an elongated first ferrite spacer section separates first ends of the first and second ferrite plates, and an elongated second ferrite spacer section separates second ends of the first and second ferrite plates. An elongated passage is defined by inner surfaces of the first and second ferrite plates and extends entirely through the ferrite core assembly. The first insulative layer passes entirely through the passage and extends outward from opposite sides of the ferrite core assembly. A securing element clamps the first and second ferrite plates and the first and second end spacer sections together.
In one described embodiment of the invention, a plurality of laminated insulative layers are provided, wherein the first spiral winding is a subwinding of a primary winding including a plurality of series-connected spiral subwindings formed on various surfaces, respectively, of the various insulative layers forming a multi-layer printed circuit board, the spiral subwindings being connected in series by conductive vias passing through the various insulative layers. The second spiral winding is a secondary winding and surrounds the plurality of series-connected spiral subwindings of the primary winding. The secondary winding is formed on a second insulative layer adjacent to the first spiral subwinding. A primary current supply circuit is connected to a first primary winding terminal that is connected to the first spiral subwinding. A switch connected to a second primary winding terminal is operated to produce zero voltage switching of the primary current at a resonant frequency of the primary circuit. This arrangement separates the secondary winding as much as possible from the primary winding terminal on which large flyback voltages occur, reducing electric field gradients and the resulting electromagnetic radiation.
Brief Description of the Drawings
FIG. 1 is a partial perspective view of a planar transformer of the invention.
FIG. 2 is an exploded partial perspective view of a preferred embodiment of the planar transformer of the present invention.
FIG. 3 is a partial section view of the planar transformer of FIG. 2.
FIG. 4 is a perspective view of the planar transformer of FIG. 2 showing portions of the printed circuit board extending from opposite sides of the ferrite core and carrying associated circuitry.
FIG. 5 is a schematic circuit diagram of the planar transformer of FIG. 4 and a resonant primary winding circuit thereof.
FIGS. 6A-G are scale copies of the metal traces on each of the layers of the printed circuit board of the embodiment of FIG. 5.
FIG. 7 is a section view diagram useful in explaining one aspect of the invention.
FIG. 8 is a partial exploded perspective view of another embodiment of the planar transformer of the invention including conductive shields that contribute to electrically isolating the secondary windings from the primary winding.
FIG. 9 is a perspective view of another embodiment of the planar transformer of the invention.
FIG. 10 is a section view illustrating encapsulation of the embodiment of FIG. 9 in a molded DIP package.
Detailed Description of the Preferred Embodiments
FIG. 1 shows a partial exploded view of one embodiment of the invention disclosed in the above-identified parent application, Ser. No. 969,508, now U.S. Pat. No. 5,353,001, incorporated herein by reference, in which printed circuit board 11 has a "tongue" section 11A on which windings 12 and 13 are formed generally concentrically, as spiral windings, around a common magnetic flux path area 14. This embodiment of the invention includes a ferrite core structure including an upper section 180 having two post or spacer sections 180A and 180B, one at each end. The two spacer sections 180A and 180B each can have the same width dimension 39 as the upper and lower core half sections 180 and 200. The lower ferrite core half section 200 is a thin flat plate. Spacer sections 180A and 180B are integral with the thin ferrite plate constituting the lower part of ferrite core half section 180.
Core half sections 180 and 200 thereby form a shallow, rectangular passage 44, the height of which is 0.025 inches, as indicated by arrows 32. The width 33 of passage 44 is 0.73 inches. The total thickness of the ferrite core thus formed is 0.305 inches, its length is 1.0 inches, and its width is 0.63 inches, as indicated by dimension lines 38, 37, and 39, respectively. By making the cross-sectional areas of plates 180 and 200 thin and wide, the necessary inductance of typically 800 microhenrys for primary winding 12 is achieved, and the cross-sectional areas are made large enough to keep the magnetic flux density low enough to prevent saturation of the core material. Power loss due to saturation of the core is thereby avoided. The tongue section 11A of printed circuit board 11 with all of the spiral windings printed thereon extends into rectangular hole 44. The width 39 of ferrite core 180, 200 completely subtends or confines all of the generally concentric windings 12, 13, etc. Note that the core 180,200 also could be formed of a single block of ferrite material with hole 44 milled out.
The structure shown in FIG. 1 can be thought of as having a gap 32 of 0.025 inches in the ferrite core 180,200. Gap 32 can be thought of as a gap in the center post of a "E" shaped ferrite core half section 180A,200A as a "limiting case" shown in FIG. 7, where the lengths of the center posts 48 are zero. That is, in the embodiment of FIG. 1, the post structure can be considered to be "infinitely shallow" so the lengths of such posts is zero. In the embodiment of FIG. 1 magnetic flux passes through gap 32 in the center magnetic flux area 14 of the primary and secondary windings, causing mutual inductive coupling therebetween.
FIG. 2 shows a partial exploded view of a preferred embodiment of the planar transformer 10A, with the various insulating layers of a multi-layer printed circuit board 11 omitted for clarity. However, FIG. 3 shows a partial section view of six insulative layers 111-1 . . . 111-5 and 111-21 of a printed circuit board 111 (See FIG. 4) that extends entirely through the passage 44A,B defined by upper core 18 and lower core 20 in FIG. 2. (Note that identical or similar reference numerals are used to designate corresponding parts in the embodiment of FIG. 1 and the embodiment of FIGS. 2-4.)
Referring to FIGS. 2 and 3, upper core section 18 includes a thin flat plate having enlarged, downwardly extending end spacer portions 18A and 18B which abut similar enlarged and upwardly extending end spacer portions 20A and 20B of an essentially identical bottom core section 20. Core sections 18 and 20 can be composed of H63B ferrite material, available from FDK America, Inc., of Irving, Tex. The length and width of each of core section 18 and 20 are 1.0 inches, and 0.63 inches respectively. The thickness of the central portion of each core section is 0.135 inches and the thickness of the enlarged end sections is 0.162 inches. The dimensions of the gap 44A,B defined by upper core section 18 and lower core section 20 are 0.73 inches by 0.63 inches by 0.054 inches.
Primary winding 12 includes five series-connected primary subwindings 12-1 . . . 12-5 (only three of which are illustrated in FIG. 2, for convenience) and an auxiliary primary winding 21 (see FIGS. 3 and 5) which can be formed on the various surfaces of printed circuit board layers 111-1 . . . 111-5, respectively. Each of the five primary subwindings, which are collectively referred to by numeral 12, is generally coextensive with the others, so that they precisely overlap as indicated by dashed lines 26 and 27. One terminal of each of the primary subwindings 12-1, 12-2 etc. is connected to a terminal of another of the primary subwindings on an adjacent layer by a conductive via such as 50A, 50B or 50C. Each of the primary subwindings 12-1, 12-2 etc. circumscribes or surrounds an inner flux path region 14. The terminals of primary winding 12 are designated by numerals 30A and 30F. As shown in FIGS. 2 and 3, secondary winding 13 is formed on the top surface of printed circuit board layer 111-1 and is precisely aligned in a non-overlapping fashion with the five primary subwindings 12. The outer end of secondary winding 13 constitutes terminal 13A, and the inner end 13C is connected by conductive via 49 to metal trace 13B which constitutes the other secondary winding terminal. The inner portion of secondary winding 13 is generally coextensive with the outer portion of each of the five primary subwindings 12. The auxiliary power winding 21 is formed on the upper surface of insulative layer 111-21 (FIG. 3).
As shown in FIG. 4, a slightly narrowed portion 111A of printed circuit board 111 on which primary winding 12 and secondary winding 13 are formed extends entirely through the gap 44A,44B defined by the inner surfaces of upper core section 18 and lower core section 20 and the enlarged spacer portions 18A,B and 20A,B. Numerals 57 and 58 generally designate various components, including some of the components indicated in FIG. 5 on the left and right portions of printed circuit board 111 extending beyond the opposed sides of core sections 18 and 20.
Core sections 18 and 20 are clamped together by a pair of inwardly biased spring clips 31. The entire assembly, including printed circuit board 111 and planar transformer 10A thereon, then is inserted in a suitable "package", for example, an enlarged housing for a male electrical plug, with conductors 42 and 43 leading to a rechargeable nickel cadmium battery in a mobile telephone unit or the like which is powered by the male electrical plug. Complete details of a "smart" battery charger circuit on printed circuit board 111 are set forth in U.S. patent application Ser. No. 063,163 filed May 11, 1993 and entitled "COMPACT LOW NOISE LOW POWER DUAL MODE BATTERY CHARGING CIRCUIT" which is incorporated herein by reference. Suitable potting material and/or features of the inside of the package prevent spring clips 31 from coming loose, e.g., as a result of vibration. Conductors 59 engage two prongs, respectively, of the male electrical plug, and are connected to supply household current to the input of the battery charger formed by printed circuit board 111 and planar transformer 10A thereon.
FIG. 8 shows another diagram similar to FIG. 2, in which a plurality of conductive shields 16-1 . . . 16-5 are disposed between the five primary subwindings 12 and the one secondary winding 13, and are generally coextensive with the turns of secondary winding 13. Each shield has a narrow gap 16A to prevent flow of eddy currents, and a lead connected by the other shield leads by a suitable via, indicated by dotted line 25. The shields leads are electrically connected to a low impedance conductor carrying an AC ground voltage.
The initial prototypes of the invention included shields such as 16-1 . . . 16-5. Surprisingly, however, it was discovered that when they are eliminated, the electrical noise or interference, which results partly from coupling between the primary subwindings 12 and secondary winding 13, nevertheless is acceptably low to meet the specifications for which a marketable product is being designed.
As explained in detail in the above-mentioned Somerville application, optimum overall circuit performance is obtained by combination of the planar transformer shown in FIGS. 2 and 3 and a resonant primary circuit as described in the above-identified Somerville application Ser. No. 063,163, especially if the turn off times of MOSFET 35 are adjusted in accordance with changes in the inductance of the primary subwindings 12-1 . . . 12-5 occurring as a result of changes in the loading of the secondary winding 13. This results in reduced power dissipation in MOSFET 35 in FIG. 5 and also ensures that the fly-back voltage V.sub.CR returns to zero before MOSFET 19 is turned on, ensuring that "zero voltage switching" is achieved despite variations in the inductance of the primary winding 12 caused by varying load conditions. This substantially reduces the generation of noise harmonics and the like.
FIG. 5 is a simplified schematic diagram that shows the arrangement of the battery charger of the above-identified Somerville application Ser. No. 063,163. The components most closely related to the resonant primary circuit of the planar transformer 10A are shown. More specifically, the series connections of primary subwindings 12-1 . . . 12-5 and the location of secondary winding 13 are shown in FIG. 5. Switching MOSFET 35, a full wave rectifier D1-4, isolation barrier capacitors 54A,55A and 54B,55B also are shown. The auxiliary primary winding 21 used to generate power for circuitry on the primary side of planar transformer also is shown.
In the circuit of FIG. 5, the 110 volt 60 hertz line current is rectified by the full wave rectifier including diodes D1, D2, D3 and D4 to produce the waveform V.sub.28 on conductor 28, which is connected to terminal 30A of primary winding 12. The other terminal 30F of primary winding 12 is connected to the drain electrode of MOSFET 35 and one terminal of flyback capacitor C.sub.R. Terminal 30A of primary winding 12 also is connected to the other terminal of flyback capacitor C.sub.R on which the 120 hertz full wave rectified signal is produced. The flyback voltage V.sub.CR occurs on conductor 30F. MOSFET 35 is turned on and off at a 500 kilohertz to 1 megahertz rate, so that primary winding 12 is energized only for a part of the time which is centered about the times of occurrence of the peak levels of the voltage waveform on conductor 28. A feed-back loop such as described in U.S. Pat. No. 5,111,131 involving primary winding 12 adjusts the "on" time of MOSFET 35 to achieve a relatively constant primary winding current level. The primary winding 12 has a smaller average inductance during heavy loading of the secondary winding than during light loading. Accordingly, the off time of MOSFET 19 is adjusted to be less during high current loading conditions than during low current loading conditions, reducing power dissipation in the primary winding because the magnetic flux has time to return to zero for each cycle of the primary winding resonant circuit.
In the presently preferred embodiment of the invention, isolation barrier capacitors as described in U.S. Pat. No. 5,111,131 couple control signals produced by circuitry on the output or secondary winding side of printed circuit board 111 to the input circuitry on the input or primary winding side of printed circuit board 111. In FIG. 5, a pair of capacitors 54A,55A are connected in series to provide one isolation barrier, and capacitors 54B and 55B are connected in series to form another isolation barrier. This pair of isolation barriers is used to convey control information from the secondary side of planar transformer to the primary side thereof.
FIGS. 6A-6G show scale views of the various patterns of printed circuit board 111 fabricated to implement the various planar transformer windings and subwindings (including the additional primary winding transformer 21 which rectifies the primary current to produce a power supply signal utilized to power the circuitry on the primary winding side of planar transformer 10A) connected in the order shown in FIG. 5. More specifically, FIG. 6A shows the pattern for secondary winding 13, which appears on top of the top layer 111-1. FIG. 6B shows the pattern for primary subwinding 12-1, and also shows one pair of plates of isolation barrier capacitors 55A and 55B and their leads 61A and 61B to the secondary winding circuitry. Note that the leads 61A and 61B pass outside of the region 14 of concentrated magnetic field flux. FIG. 6C shows the primary subwinding pattern for primary winding 12-2 and also shows two of "floating" isolation barrier capacitor plates which are provided to meet UL regulations unrelated to the present invention. FIG. 6D shows the pattern for auxiliary primary winding 21. FIG. 6E shows the pattern for primary subwinding 12-3. FIG. 6F shows the pattern for primary subwinding 12-4. FIG. 6G shows the pattern for primary subwinding 12-5.
For primary winding 12, the largest AC voltage is the flyback voltage V.sub.CR, which has peaks of approximately 500 volts. It occurs on primary terminal conductor 30F, and the smallest voltage changes, with peaks of approximately 167 volts, occur on primary terminal conductor 30A. Therefore, in accordance with the present embodiment of the invention, secondary winding 13 is provided on the printed circuit board layer adjacent to primary subwinding 12-1 on which the smallest voltage changes occur. This causes the physical location of secondary winding 13 to be separated by the maximum possible distance from conductor 30F on which the largest AC voltage peaks of approximately 500 volts occur. This minimizes the electric field gradients between primary winding 12 and secondary winding 13, and therefore reduces electromagnetic noise generation and also reduces the likelihood of current leakage or voltage breakdown between primary winding 12 and secondary winding 13. In accordance with the present invention, the parasitic capacitances 51 between primary subwindings 12-1 . . . 12-5 and secondary winding 13 are minimized by the illustrated non-overlapping structure. This further minimizes electromagnetic noise generation.
In some applications, especially if printed circuit board 111 is very thin, the end spacer portions 18A and 18B and 20A and 20B shown in FIG. 2 can be omitted, in which case the gap 44A,B extends over the entire length and width of the upper and lower ferrite core sections 18 and 20. FIG. 9 shows such an implementation of the planar transformer, in which upper core section 18 is adhesively or otherwise attached to the upper surface of printed circuit board 111, and lower ferrite core section or plate 20 is adhesively or otherwise attached to the bottom surface of printed circuit board 111.
In FIG. 10, numerals 45 designate adhesive attachment of ferrite cores 18 and 20 to the top and bottom surfaces of printed circuit board 111. Gap 44 is equal to the thickness of printed circuit board 111 and extends throughout the entire region between upper core section 18 and lower core section 20. The printed primary winding and secondary winding, and also a shield similar to that described with reference to FIG. 8, can be provided on layers of printed circuit board 111 between ferrite core sections 18 and 20 in the manner previously described. This embodiment of the invention has lower inductance than the previously described embodiments of the invention, but will be suitable for some applications.
The planar transformer of FIG. 9 has several advantages, one being that it is more compact, as the ends of ferrite cores sections 18 and 19 need not extend around opposed edges of the printed circuit board 111. Another advantage is that gap 44 can be thinner, precisely equal to the thickness of the printed circuit board 111, whereas in the embodiments of FIGS. 1 and 2 the gap 44A,B must be slightly greater than the thickness of the printed circuit board because a small amount of clearance tolerance is needed to ensure that the end sections 18A,B and 20A,B properly abut. Furthermore, it is more convenient to adhesively attach ferrite core sections 18 and 20 to the surfaces of printed circuit board 111 in the planar transformer of FIG. 9 than in the earlier described embodiments.
Another important advantage of the planar transformer of FIG. 9 is that it can be more conveniently encapsulated in a molded DIP package, as shown in FIG. 10, wherein numeral 65 designates the plastic injection-molded body of a DIP package in which the planar transformer and associated printed circuit board 111 are encapsulated. Numeral 63 designates the leads of the DIP, and numeral 64 designates conventional wire bonds of leads 63 to wire bond pads (not shown) on a surface of printed circuit board 111. Magnetostrictive forces produced by ferrite core sections 18 and 20 can be absorbed by the material of which printed circuit board 111 is constructed, avoiding associated power loss and stresses on the DIP body material 65. Problems associated with leakage of the injected molding material between the surfaces of printed circuit board 111 and the inner surfaces of ferrite core sections 18 and 20 are avoided.
Thus, a very small, efficient hybrid integrated circuit transformer with low core losses and low electromagnetic interference is achieved by the above described invention.
While the invention has been described with reference to several embodiments thereof, those skilled in the art will be able to make various modifications to the described embodiments without departing from the true spirit and scope of the invention.