JPH0696941A - Partial gap magnetic core device - Google Patents

Abstract

PURPOSE: To obtain an appropriate manufacturing cost and improved performance characteristics by using a core provided with an air gap which is magnetically and physically shielded. CONSTITUTION: An E-type core structure 21 comprises mating segments 22 and 23 on which a gap 24 is formed by a recessed surface 26 and a plane surface 27. Said gap 24 is surrounded by a core material and a wall 25 is formed around whole circumference of the gap. At this point, the cubic volume and other parameter of this magnetic core device maintains the inductance in a large current, and accordingly, the operating characteristics relating to the whole gap core can be optimized. Also, the depth of the gap is set at least at 0.5 ×10<-3> inch. As a result, the breakage caused by the difference in thermal expansion of not material and the heating caused by the magnetic field coupling can be prevented.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a device having a core structure in which electric wires are wound. The associated core is generally made of ferrite or other soft magnetic material and has at least 1
It consists of segments that together form a magnetic closed loop at one, partially gapped, mating surface, thereby leading to a structure known as a "partial gap" or "variable gap" core.

[0002]

BACKGROUND OF THE INVENTION Wire wound core devices serve a variety of functions. One of the important applications is in power supplies, where they can serve as transformers for changing the voltage level as well as isolating the output network from the input network. Also as an inductor,
They can take the form of choke coils for AC ripple reduction and other forms for noise reduction. Moreover, they can serve for magnetic energy storage, for example in power supplies using flyback conversion circuits.

A design limitation on the device is that the inductance is significantly dependent on the current-eventually leading to magnetic saturation-as a result of non-linear effects. The reduction of inductance in the presence of current with a large DC component is a major design consideration of the device. Related prior art is discussed in connection with FIG. This figure shows the general form of the relationship of these two parameters for the three types of structures, each based on inductance and DC coordinates on a logarithmic scale. Curve 10 is a graphic for a gapless loop core. Curve 11 shows the shape of the relationship for a gapped structure-a loop core having at least one gap extending across the entire core cross section. Curve 1
2 relates to such a loop core that is "partially gapped" -as a result of the associated regions joining the surfaces such that one or both contours of the surface are partially gapped and partially Draws a general structure-general relationship that is intended to result in a continuous mating bond. Curve 10 drops sharply to reach a low value at 14 with a high initial inductance value drawn to peak at a current point 13 of near zero. 1
The actual inductance value in 4 is too low to be meaningful for the expected device, but is not really zero-close to the value of the air core structure. This steep reduction in inductance is at too little current to meet the requirements of many devices. For example, operating as a choke, it may not be able to provide sufficient smoothing at all expected power.

Curve 11 is otherwise structurally curve 10.
It is based on an all-gap structure that is the same as the structure of. The presence of such non-magnetic gaps, such as "air" gaps, is not magnetically saturable and will therefore be substantially dependent on the inductance at a constant or near constant inductance plateau at -15. Everything. This full gap structure is characterized by a significant inductance persistence at higher currents, which reaches an inductance value of almost zero only at 16. However, the full gap structure is also characterized by a reduced initial inductance value 17. For many purposes, this low initial inductance value is insufficient for the intended application. The desirable properties of gapped and non-gapped core structures have, to some extent, been
It is known that in one type of structure it can have the other. Sometimes referred to as a "partial gap,""steppedgap," or "non-linear" core, this latter takes the form of a core with a reduced cross-section at some point. This is often due to the core interface having a central contact area surrounded by a circumferential gap,
It is realized by joining core segments with a reduced cross section.

Curve 12 depicts the inductance-current characteristic for a partial gap core with a gap depth close to that of the structure of curve 11. Its initial or "at near zero current" inductance of value 18 approaches that of the ungapped core of curve 10. As the current increases, the inductance decreases to a plateau value of 19 and finally decreases to a low, near zero value 20 at current levels close to that of the decrease for the gapped structure of curve 11. The partial-gap “non-linear” structure met several performance requirements. Their use led not only to maintaining the inductance value at high current, which is characteristic of the gapped structure, but also to an initial inductance value close to that of the gapless structure. However, they do possess some properties of the total gap structure that can be defective. From a performance standpoint,
The stepped gap structures, like their full gap counterparts, generate an edging magnetic field in the gap. The coupling of the edging magnetic field and the windings surrounding it results in a reduction of the inductance per unit volume during heating,
Generally, this leads to a decrease in overall performance. This problem is exacerbated for planar devices and as miniaturization increases. Structural deformation can cause further problems. As an example, the use of a winding made of a conductor of rectangular or elliptical cross section (having a large radial dimension), eg a spiral winding, is effective in reducing the DC electrical resistance, but with a larger magnetic field. Practically excluded due to coupling and the resulting increase in heating and power loss.

Yet another disadvantage of prior art non-linear structures is that, like all-gap structures, some of them are lost during use as well as during fabrication. The entry of potted molding material into the exposed perimeter gap can cause physical failure. Curing or crystallization of the potted material is accompanied by a volume change, which can result in a bond separation.
During use, interface failure can be caused by localized heating due to resistive losses in the windings and electromagnetic losses in the core. Failure is due to differences in thermal expansion within different regions of the potted material that fill, or even due to uniform expansion of the potted material different from that in the contact region of the core.
Can be triggered. Even if no physical bond failure occurs, the difference in thermal expansion of potted material in the gap can lead to an unpredictable or undesirable effective temperature coefficient of inductance for the device. The above is discussed in many articles and texts. For example, Soft Ferrites Properties and Applications , EC
See Snelling, sec.ed., at pp.20-21, 139-143 and 274-283.

[0007]

SUMMARY OF THE INVENTION Broadly speaking, the present invention overcomes the deficiencies of prior art structures by the use of a core with an air gap that is magnetically and physically shielded. The advantageous results of the resulting partial gap device include both reasonable fabrication cost and improved performance characteristics. The structure of the present invention can advantageously replace conventional full gap devices. As discussed in the "Design Requirements" section, volume and other parameters can be optimized to maintain the inductance at high currents and thereby approach the operating characteristics associated with full gap cores. The continuous surface contour made possible at the interface avoids the practical problems associated with the reduced cross-section of prior art partial gap interface. The variable gap core characterizing the overall structure of the present invention, in fact, also retains the performance advantages of the earlier variable gap core structures. Therefore, its use can be made with both a high initial (at zero current) inductance characteristic of a gapless structure and a level of inductance maintained at high currents characteristic of a gapped structure.

Secondly, the joining of the core segments by the surfaces defining the areas of the peripheral contact with each other to enclose the gapped area overcomes the technically recognized drawbacks of the non-linear core device described above. The discussion conveniently defines a mating surface symmetric about a center that is wholly enclosed within the surrounding contact area and that has the same contour and volume as those of the core portions joined thereby. It relates to a recess located in the center of one. Various requirements may dictate the resulting variations in the position and shape of the gap, rather than the use gap of multiple gaps. The approach reduces the edging magnetic field to improve performance efficiency with reduced eddy current losses and increased inductance-while minimizing heating due to coupling with the surrounding windings-can be avoided altogether. -. In turn, the increased inductance to volume ratio may allow for further miniaturization. For potted material of cross-linked polymer which has not yet cured, for example, when depending on the closeness of the ambient contact, the viscosity values usually reached avoid avoiding entry into the gap, and thus during assembly and Sufficient to prevent bond failure both during use. This exclusion can be further ensured by a bond bond in the contact-bonding area before potting.

The characteristics outlined are generally advantageous in multiple sets of conductive wound core devices. Inductance / current characteristics as well as interface stability are advantageous in AC devices-transformers and inductors. With prior art non-linear core devices, particular interest concerns energy storage inductors as well as chokes for DC devices such as power supplies. The implication includes a tolerance related to a device design that was previously considered a malfunction. As an example, the essential decoupling of closely spaced devices and leads by the avoidance of edging fields is a pancake shape of rectangular or elliptical cross section, with implications including reduced volume and reduced cost. Allows free use of windings. For convenience, the controversial terminology generally refers to a coil that is generally based on a particular shape, such as a core of predominantly circular cross-section-in general,
With respect to inductors when used for certain devices, such as choke coils. Inventive advances are generally of more widespread value throughout the spectrum of "wound" core devices. In all of the above examples, the inventive advances are helpful in terms of magnetic and physical shielding. For the same reason, the windings can be of any desired cross section-constant or variable in size and / or shape.

[0010]

EXAMPLES The terminology mostly uses terms that are well known to today's researchers for devices that are of particular interest at this point. A general Hososaki was explained. That is, the device of the present invention relies on inductive coupling for current flow along a helical conductor path that surrounds the relevant region of the magnetic core. The device of the present invention has the common feature of partially contacting the core mating surface-all with a magnetic core that is partially continuous unless at least one gap is provided. Consistent with common usage, "continuous" is (1)
Physically continuous, like a toroidal core,
(2) Or, although often described as "no gap", it may actually refer to a core segment that is mated, etc., that is simply approaching the annulus to the extent that the mating surface is completely smooth. (For many purposes, “no gap” refers to an effective gap of 0.5 × 10 ⁻³ inches or less.) The mating surfaces are “partial gap”, “stepped gap” and “variable gap”. Variously referred to, causing a "non-linear" structure (for structures in which the inductance represents a current-dependent shape, such as the shape of curve 12 in FIG. 1). For the same reason, the description cites "the wire is wound" in the familiar way. As with common usage, the use of this term intends a device that, regardless of how it is made, acts like a prototype core when really surrounded by a spiral winding of wire. To do.
In fact, many "wire-wound" structures, as currently manufactured, rely on deposited or printed conductor segments, not literal "windings." It is believed that the commonly used uses of the present invention will take the form of this type of structure. Terms such as "winding" and "coil" are to be interpreted accordingly.

Design Requirements To a large extent, design requirements are well known. Important parameters at least in basic conditions-volume,
The configuration, etc.-has the same effect as for the early partial gap device. Good citation, ie 27
Soft Ferrites cited above on pages 4-283
Properties and Applications- for one type of device
For choke coils and storage inductors-consider said design. The same requirements for other types of equipment are:
It is discussed elsewhere in the same text. As such, the detailed design is not considered an integral part of this disclosure.
Instead, the debate has become a more general term. All structures of the present invention have a common feature-a core with a shielded gap. As such, one or both of the shielded "gaps" are at least 1
It is defined as a three-dimensional crevice surrounded by the core, such as caused by the joining of core surfaces in a topology leading to the two said gaps. In order to qualify as a "gap", it is
In other respects it is necessary to be at the functionally inevitable level for winding current values beyond that which characterizes a gapless structure consisting of the same design parameters. For many purposes, this translates into one side / both sides of the cross section that defines the minimum gap that is operationally significant. 0.5 × minimum
Experimental studies based on a 10 ⁻³ inch gap depth show that the operational implications are a centrally located gap in an area as small as 10% of that of the faying surface and −5 of that of the faying surface. As small as% -attributed to a non-centered gap in a smaller area. This study was confirmed with a matched RM10 core, which has a low profile. The M10 core structure is based on IEC Publication 431-1983 (Ge
neva, Switzerland) and JIS C 2516-1990 (Tokyo, Japan)
Is defined by The core of this research is
It was modified to have a height of 50% of the standard when compared to the 0.366 inch standard, or a height of 0.183 inch per half core. Gap depth may be related to increasing inductance and other parameters according to the following formula: That is,

[Equation 1]

In the above equation, L = inductance N = number of winding turns A = cross-sectional area of magnetic circuit l = magnetic path length of circuit body excluding gap area β = ratio of contact area in gap area to overall area A μ ₀ = Air permeability μ _l = Relative gradual permeability of magnetic material in the circuit body μ _m = Relative gradual permeability of magnetic material in the region around the gap δ = Gap depth

The relative magnetic conductivities μ _l and μ _m are:
It is a non-linear function of the magnetic field strength of the H magnetic fields H _l and H _m . The magnetic fields H _l and H _m are then related to the NI or DC amperage due to the magnetic susceptibility which depends on the magnetic flux distribution between the gap and the contact wall. Therefore, the formula cannot be solved analytically. Many methods, such as finite component analysis, can be used to obtain a solution with reasonable accuracy. Nevertheless, this equation yields μ _l and μ _m
Related to the approximations for the value of, it can serve as a valid starting point for the experimental investigation of the changes in L as a function of the values of δ, β and I _dc .

The volume is chosen with the desire to obtain device functionality. So, this hope is an action that approaches that of an all-gap structure. The surrounding wall thickness is made as thin as possible. In the above example, the main purpose of the retained contact area of the final core interface is the avoidance of edging fields and the physical integrity of the interface. A wall thickness of 10 ^-2 inches is functionally sufficient for magnetic field shielding. The minimum wall thickness to avoid mechanical failure is primarily a matter of physical stability and manufacturing convenience. Experimentally, a ferrite with a wall thickness of 10 ^-2 was found to be sufficient for the studied structure. In examples where the device is adapted to act like some early partial gap devices, for example where high inductance values in low current operation are particularly important, the surrounding wall thickness is considered in the previous section. It is likely to be larger than the minimum value. Inductance maintained at high current decreases somewhat in the above example, as shown in FIG. However, the centrally located gap, which is as small as 38% of the total cross-sectional area of the junction, is a functionally significant increase in the inductance maintained at increased current for the structure studied, and is therefore intended. Qualified for use with the device. Due to the low profile core devices studied, eg the RM10 design as cited above, such a minimum gap area with a gap depth of 19 × 10 ⁻³ inches is associated with a corresponding non-gap structure. At a current approximately four times greater than that of the device, the device has a significant inductance (region 19 in FIG. 1).
Equivalent to the plateau value).

Although the initial design parameters are generally computable, the practical requirements require some trial and error. For example, the deviation from perfect smoothness of the surface is
(For example, compare the first curve 10 of FIG. 1 for inductance vs. current for a truly gap-free core with the curve 40 of FIG. 4 based on relevant properties for a real structure including the joining of two “smooth” surfaces. Yes) requires some empirical method. As experience gains, the design of new structures may even be able to dispense with theoretical requirements overall. Structures include those that have the greatest advantage for closely spaced, low profile, small sized devices where temperature rise is particularly significant. From this point of view, the device dimensions, which consist of a fraction of an inch and are equally spaced, derive in particular from the avoidance of heating by edging field coupling.

Fabrication This heating-based requirement is further well understood with respect to devices that meet this requirement. The requirements of the core configuration are alleviated somewhat considering the inventive nature of the increased inductance per unit volume, especially with windings of rectangular or oval cross section. The structure for inclusion in an ordinary wiring board network is a well known ferrite construction,
For example, ferrites based on manganese-zinc or nickel-zinc can be used. The required magnetic properties and fabrication requirements can be in any of various alternatives-eg other ferrites as well as elemental metals or alloys). Recent advances in the assembly of the main types of the structure can be of benefit. US patent application Ser. No. 07 / 710,736, filed May 31, 1991, discloses suitable core structures made from core segments by adhesive bonding. It was mentioned that it was appropriate for the coil winding to take a form other than the literal wire winding form. US Patent Application No. filed on February 14, 1992
07 / 835,793 discloses joining partial windings, which leads to the functioning of the windings. Other fabrication approaches, some used commercially and others disclosed in the literature, may be useful.

Drawings Figure 1 was discussed in the previous discussion. The three curve shapes provided, namely the shapes of curves 10, 11 and 12, represent the general shape of inductance vs. DC current L vs. I _dc when plotted in log-log coordinates. These curves are
Corresponds to no-gap and partial-gap core structures. For the purposes of discussion, the axis intercept value is closer to said value since it is based on log coordinates, but is treated as a zero value for the other coordinate axes. However, 14, 16
The final inductance values at 20 and 20 approach non-zero values such as those obtained from air core devices where the study-based structure is effectively altered by magnetic saturation of the core. The value of I _{dc at which} the value 14 is reached varies-that is, the intensity of the decrease in curve 10 is the imperfections of the surface for otherwise identical structures including the joining of non-perfect "smooth" mating surfaces. Decreases with increasing −.

Curve 11 shows the relationship between inductance and current for a full gap structure, starting with an initial inductance 17 for zero current, maintaining a constant or plateau value for the main part of the curve for region 15 and finally Decrease to reach a minimum inductance value of 16. Similar to the decrease point, the plateau value changes as the gap is changed. Increasing the size of the gap leads to a decrease in the inductance as the current value increases. This relationship is also known-a valid reference is the Soft Ferrites text cited above (see Figure 9.13 on page 277 and related text). Curve 12 is
It represents a partial gap structure, starting with an inductance value at zero current at 18, then decreasing to a plateau value of 19 and finally decreasing to a small inductance at 20 (air core value). As discussed, the shape of the relationship, as represented by curve 12, can be made closer to curve 10 or 11. In short, the ratio of (the ratio of contact surfaces at the bond, as well as the reduction in gap depth,
β = contact area / overall area)
Results in a property relationship that more closely approximates the gapless structure of. The actual inductance value at zero current mainly depends on β. On the contrary, for example, the decrease of β is determined by the curve 10
In the area before the interception of and, the characteristic approaches the shape of the curve 11. The magnitude of the inductance at the plateau value 19 decreases, and the current value 20 at the time of the decrease increases according to the gap depth δ.

FIG. 2 is a perspective view of a cross section of a mating E-shaped core structure 21 that has the same cross section as the structure used in the experiment in which most of the reported data was measured. It is then composed of mating segments 22 and 23 which together define a gap 24, which in this example is the result of the joining of the concave surface 26 and the planar surface 27. As with other contemplated structures, the gap 24 is surrounded by the core material, thereby forming a wall 25 around its entire perimeter, including the surface portion of the structure 21 that is removed in the cut below.

FIG. 3 contains information drawn on two partial gap structures that provide a substantial elimination of the edging field. The coordinate is the ordinate and the inductance index AL = L / N.
² (L is the inductance in nano-henry, N is the number of turns), and the abscissa is the amperage number NI. The two experimental structures used the same core. That is, the first structure is the structure of curve 30 and has 26 turns of spiral cross-section conductor spiral winding, and the second structure is three turns of rectangular cross-section conductor spiral ( Pancake type) with winding.
The substantial coincidence of curves 30 and 31 is clear evidence of no edging field. Because the magnetic field is
Because it couples more strongly with the pancake winding (when current is applied to reach the same number of ampere turns), which leads to lower plateau values for its structure-the structure of curve 31. is there.

FIG. 4 shows the inductance L in microhenry on the ordinate, the abscissa on four RM10 (FIG. 5C) core structures with the same design, but with the same gaps and the same gap, but with the same gap and size. 2 is a logarithmic-logarithmic chart of the amperage NI. Curve 40 relates to these quantities for a gapless structure, curve 41 for the total gap at a depth of 20 × 10 ⁻³ inches, curve 4
2 reports the measured quantity for a 20x10 ^-3 inch deep cylindrical gap surrounded by a ^{40x10 -3} inch wall,
Curve 43 is the same as the structure of curve 42, but for a structure having a wall thickness of 31 × 10 ^-3 inches. As discussed, the characteristics of the gapless structure of curve 40 get closer as the relative contact area increases, while the characteristics of the total gap structure get closer as the area decreases.

5A-5F are perspective views of the core loop segment currently in use. As before, the illustrated segment surface can be mated with a contoured surfaced segment, such as a mirror image segment or, as an alternative, a planar (or gap-free) surface segment. The illustrated structure, as well as a number of alternatives, are disclosed in detail in the Soft Ferrites cited above. The drawings correspond to each structure as follows. That is,
5A-U type core, 5B-E type core, 5C-RM core, 5
D-low profile core, 5E-EP core, 5F-pot core. As shown, the entire structure shown comprises an exemplary centrally located recess and the cross-sectional shape of the core leg surrounding it.

6A-6D are cross-sectional perspective views of core segments representative of a large number of gap configurations, all of which have a segment with a contour,
Alternatively, it may be joined to a flat mating surface. FIG. 6A shows −
Cavity 60 in a portion or wall 62 that is not a recess in this example
And 61-indicates multiple cavity gaps. Figure 6B
Depends on a stepped gap consisting of gap regions 64 and 65 defined by walls 66. FIG. 6C shows a structure with a gap 67 having a varying depth to be enclosed within a wall 68. FIG. 6D shows the structure with an annular gap 67 enclosed within the wall 68 and then surrounding the non-recessed region 69.

[0024] Examples Example 1 the same shape, size, three choke coil configuration and the winding number of windings have been excited to reach the form of data shown in FIG. The cores used for all three are-Fig. 5C
Was a low profile RM10 core with 26 turns wound around the central leg, as shown in FIG. The size of each mating core pair was approximately 1.09 inches x 0.52 inches with a 0.42 inch diameter circular center leg. The first structure is gapless and the second structure is 20 × 10 ⁻³ in the central leg.
A gap was provided at a constant depth of inches, and the third structure was shielded to a depth of 19 x 10 ^-3 inches surrounded by a perimeter wall 34 x 10 ^-3 inches thick as shown. It was equipped with a cylindrical gap. When excited, the measured inductance values are 4470, 174 and 23.
60 microhenry zero current inductance,
Low inductance (points 14, 1, 1) at I _dc = 0.9 amps, 7.0 amps and 5.5 amps, respectively.
With 70 μH corresponding to 6 and 20). Subsequent Note-Both the curve format and the reported values were about the same as for the prior art partial gap structure (with the surrounding gap rather than surrounded).

Example 2 Two flyback transformers are provided, the first one with all gaps, the second one otherwise identical in size (low profile RM10), core configuration, primary and It consists of a partial gap surrounded by a secondary coil structure (circular cross section with a diameter of 0.416 inches, gap depth = 20 × 10 ^-3 inches, β = 0.27), which results in a difference in performance. Operated in a flyback conversion circuit to confirm. Both operated with an average input current of about 2 amps, as resulting from an input at 500 kHz, 40 volts. The transformer with the enclosed partial gap showed a transformer loss of 3 watts, which was about 20% lower than the transformer with full gap, while maintaining all other relevant conversion performance.

Example 3 The data curves measured as shown in FIG. 3 were based on two structures of the same shape, size and composition as the partial gap structure of Example 1. The core used is a depth of 20
A gap of × 10 ^-3 inches was provided. The gap diameter was 0.354 inches, surrounded by a 31 × 10 ⁻³ inch wall (resulting in β = 0.28). The coil in the first structure has a diameter of 17.9 × 10 ^{−3 with} 26 turns.
Constructed from an inch circular cross-section copper wire. The second one is
With long dimensions arranged radially with respect to the core,
It was equipped with three turns of a 0.150 inch × 20 × 10 ⁻³ inch rectangular cross section (pancake shaped) conductor. As recorded in FIG. 3, both curves plateaued with an inductance index value AL of 200-300 nanohenries per match over a range of 6-100 + amperages.
As shown in that figure, the inductance index was nearly identical to support what was assumed to avoid edging fields. (The closer coupling of the pancake winding and the edging field would have led to ineffective operation to reduce inductance.)

Example 4 Four cores of the same shape, construction and size, all given 26 windings, were 500 millivolts and 100
It operated at a frequency of kilohertz, resulting in the inductance characteristics reported in FIG. The first without gap led to the measurement of curve 40. -20 x 10
^-With a constant gap depth of ^-3 inches-the whole gap structure gave rise to the characteristic depicted as curve 41. The two partial gap structures led to the performance of curves 42 and 43. The purpose of the experiment was to ascertain the effect of varying β (contact ratio to total surface at the bond, and thus only the diameter of the gap varying between the two). Curve 42
The structure corresponding to is 40 × 1 for a value of β = 0.35.
It had a cylindrical gap of 20 × 10 ^-3 inches depth when enclosed in a 0 ^-3 inch wall. The second partial gap structure for which the data points for curve 43 are measured is 3
The only difference was that the gap diameter was increased to leave a wall thickness of 1 × 10 ⁻³ inch (β = 0.28). It can be seen that the increased β leads to a partial gap structure that comes closer to that of the ungapped structure for inductance at lower current values. The reduced β resulted in an inductance / ampere frequency ratio that more closely approximated that of the full gap structure.

[Brief description of drawings]

FIG. 1 is a diagram relating these characteristics for prior art (also inventive) partial gap structures as compared to the original full-gap and ungapped structures by logarithmic coordinates of inactance and dc.

FIG. 2 is a cross-sectional view showing an exemplary core structure designed in accordance with the teachings of the present invention.

FIG. 3 shows two identical structures according to the inductance index and the coordinates of amperage—one is that of the curve 30,
FIG. 2 is a chart relating those parameters for windings of circular cross section, the other of which is of curve 31, the windings of rectangular cross section, both using the same core. The close agreement of the two curves constitutes experimental evidence supporting the substantial elimination of the edging field.

FIG. 4 shows the coordinates of the inductance and the number of amperes,
It consists of design implications when showing the performance characteristics of an exemplary gapped structure, ungapped structure, and two partial gap structures.

FIG. 5A is a perspective view of a core segment of an exemplary design suitable for use with the teachings of the present invention, eg, as mated with a planar mating surface not shown. B is a perspective view of an exemplary design core segment suitable for use with the teachings of the present invention, eg, as mated with a planar mating surface not shown. C is a perspective view of a core segment of an exemplary design suitable for use with the teachings of the present invention, eg, as mated with a planar mating surface not shown. D is a perspective view of a core segment of an exemplary design suitable for use with the teachings of the present invention, eg, as mated with a planar mating surface not shown. E is a perspective view of a core segment of an exemplary design suitable for use with the teachings of the present invention, eg, as mated with a planar mating surface not shown. F is a perspective view of a core segment of an exemplary design suitable for use with the teachings of the present invention, eg, as mated with a planar mating surface not shown.

FIG. 6A is a cross-sectional view of an unmatched core surface to be joined with a mating surface-eg, a planar mating surface-representing a suitable alternative to those of FIGS. 5A-5F. B is a cross-sectional view of an unmatched core surface to be joined with a mating surface-for example with a planar mating surface-and represents a suitable alternative to those of FIGS. 5A-5F. C is a cross-sectional view of an unmatched core surface to be joined with a mating surface-eg, a planar mating surface-and represents a suitable alternative to those of FIGS. 5A-5F. D is a cross-sectional view of an unmatched core surface to be joined with a mating surface-for example with a planar mating surface-and represents a suitable alternative to those of FIGS. 5A-5F.

[Explanation of symbols]

 22, 23 segment 24 gap 26 concave surface 27 plane surface

Claims (14)

[Claims] 1. A device comprising a magnetic core defining at least one substantially continuous magnetic path, the core comprising at least one set of first windings defining a current path for the core. Thereby providing a coil, the magnetic path comprising a partial gap having a reduced magnetic permeability and an increased saturation flux density, thereby providing an operating characteristic intermediate to that resulting from the use of a full gap structure and an ungapped structure. In a device having, said partial gap consists essentially of a gap that is wholly enclosed within the core to provide physical and magnetic shielding for the gap, whereby the edging magnetic field produced during operation is An apparatus that is substantially unchanged by the gap. 2. A device according to claim 1, wherein the relative magnetic conductivities of the magnetic paths are numerically greater than one. 3. The device according to claim 2, wherein the relative magnetic permeability is greater than 2. 4. The device according to claim 3, wherein the core is
In essence, it comprises one such gap consisting of recesses, at least one of which results from the joining of the core surfaces containing the recesses, said surface being subject to the general enclosure of said gap and the continuity of the magnetic path. A device that is equipped with a surrounding area. 5. The device according to claim 4, wherein the gap is symmetrically arranged with respect to the center. 6. The device according to claim 5, wherein the gap has a cross-sectional shape close to that of the core in the gap region. 7. The apparatus of claim 6, wherein the gap has a varying depth. 8. The device of claim 4, wherein the entire gap region is 95 of the entire core cross section in the gap region.
The device with the highest area percentage. 9. The apparatus according to claim 8, wherein the gap depth is at least 0.5 × 10 ⁻³ inches for a gap area of 5% of the area percentage. 10. The apparatus of claim 4, wherein at least the region containing the gap is potted and potted material is removed from the gap by the peripheral region. 11. The apparatus of claim 10, wherein potting involves infiltration when the fluid is placed in the pot, the viscosity of the fluid being sufficiently high so as not to substantially permeate around the bond. Device. 12. The apparatus according to claim 11, wherein the potted fluid consists essentially of an organic polymeric material that hardens to increase its viscosity with infiltration. 13. The apparatus according to claim 1, wherein the core and the coil function as an inductor. 14. The apparatus of claim 1, including at least two coils wound on the core, which function as a transformer.