2. Description Of The Prior Art: An electrical choke is a DC energy storage inductor. For a toroidal shaped inductor the stored energy is W=1/2 [ (B2Al)/ (2toir)], where B is the magnetic flux density, A the effective magnetic area of the core, Im the mean magnetic path length, and po the permeability of the free space and u. r the relative permeability in the material.

By introducing a small air gap in the toroid, the magnetic flux in the air gap remains the same as in the ferromagnetic core material. However, since the permeability of the air (u-1) is significantly lower than in the typical ferromagnetic material (1l ~several thousand) the magnetic field strength (H) in the gap becomes much higher than in the rest of the core (H=B/FL). The energy stored per unit volume in the magnetic field is W=1/2 (BH), therefore we can assume that it is primarily concentrated in the air gap. In other words, the energy storage capacity of the core is enhanced by the introduction of the gap. The gap can be discrete or distributed.

A distributed gap can be introduced by using ferromagnetic powder held together with nonmagnetic binder or by partially crystallizing an amorphous alloy.

In the second case ferromagnetic crystalline phases separate and are surrounded by nonmagnetic matrix. This partial crystallization method is achieved by subjecting an amorphous metallic alloy to a heat treatment. Specifically, there is

provided in accordance with that method a unique correlation between the degree of crystallization and the permeability values. In order to achieve permeability in the range of 100 to 400, crystallization is required of the order of 10% to 25% of the volume. The appropriate combination of annealing time and temperature conditions are selected based on the crystallization temperature and or the chemical composition of the amorphous metallic alloy. By increasing the degree of crystallization the permeability of the core is reduced. The reduction in the permeability results in increased ability of the core to sustain DC bias fields and increased core losses.

A discrete gap is introduced by cutting the magnetic core and inserting a nonmagnetic spacer. The size of the gap is determined by the thickness of the spacer. Typically, by increasing the size of the discrete gap, the effective permeability is reduced and the ability of the core to sustain DC bias fields is increased. However, for DC bias excitation fields of 100 Oe and higher, gaps of the order of 5-10 mm are required. These large gaps reduce the permeability to very low levels (10-50) and the core losses increase, due to increased leakage flux in the gap.

For power factor correction applications in power equipment and devices there is a need for a small size electrical choke with low permeability (50-300), low core losses, high saturation magnetization and which can sustain high DC bias magnetic fields.

SUMMARY OF THE INVENTION The present invention provides an electrical choke having in combination a distributed gap, produced by annealing the core of the choke, and a discrete gap produced by cutting the core. It has been discovered that use in combination of a distributed gap and a discrete gap results in unique property combinations not readily achieved by use of a discrete gap or a distributed gap solely. Surprisingly, magnetic cores having permeability ranging from 80 to 120, with 95% or 85% of

the permeability remaining at 50 Oe or 100 Oe DC bias fields, respectively are achieved. The core losses remain in the range of 100 to 150 W/kg at 1000 Oe excitation and 100 kHz BRIEF DESCRIPTION OF THE DRAWINGS The invention will be more fully understood and further advantages will become apparent when reference is made to the following detailed description of the preferred embodiments of the invention and the accompanying drawings in which: Figure 1 is a graph showing the percent of the initial permeability of an annealed Fe-based magnetic core as a function of the DC bias excitation field; Figure 2 is a graph showing, as a function of the DC bias excitation field, the percent of the initial permeability of an Fe-based amorphous metallic alloy core, the core having been cut, and having had inserted therein a discrete spacer having a thickness of 4.5 mm; Figure 3 is a graph showing, as a function of the DC bias excitation field, the percent of initial permeability of an Fe-base core having a discrete gap of 1.25 mm and a distributed gap; and Figure 4 is a graph showing, as a function of discrete gap size, empirically derived contour plots of the effective permeability for the combined discrete and distributed gaps, the different contours representing permeability values for the distributed gap.

DETAILED DESCRIPTION OF THE INVENTION The important parameters in the performance of an electric choke are the percent of the initial permeability that remains when the core is excited by a DC field, the value of the initial permeability under no external bias field and the core losses. Typically, by reducing the initial permeability, the ability of the core to sustain increasing DC bias fields and the core losses are increased.

A reduction in the permeability of an amorphous metallic core can be achieved by annealing or by cutting the core and introducing a non magnetic spacer. In both cases increased ability to sustain high DC bias fields is traded for high core losses.

The present invention provides an electrical choke having in combination a distributed gap, produced by annealing or by using ferromagnetic powder held together by binder, and a discrete gap produced by cutting the core. The use in combination of the distributed and discrete gaps increases the ability of the core to sustain DC bias fields without a significant increase in the core losses and a large decrease of the initial permeability. These unique properties of the choke are not readily achieved by use of either a discrete or a distributed gap solely.

In Figure 1 there is shown as a function of the DC bias excitation field the percent of initial permeability for an annealed Fe base magnetic core. The core, composed of an Fe-B-Si amorphous metallic alloy, was annealed using an appropriate annealing temperature and time combination. Such an annealing temperature and time can be selected for an Fe-B-Si base amorphous alloy, provided its crystallization temperature and or chemical composition are known.

For the core shown in Figure 1, the composition of the amorphous metallic alloy was Fe80Bsis and the crystallization temperature was Tx=507 °C. This crystallization temperature was measured by Differential Scanning Calorimetry (DSC). The annealing temperature and time were 480 °C and lhr, respectively and the annealing was performed in an inert gas atmosphere. The amorphous alloy was crystallized to a 50% level, as determined by X-ray diffraction. Due to the partial crystallization of the core, its permeability was reduced to 47. By choosing appropriate temperature and time combinations, permeability values in the range of 40 to 300 and higher are readily achieved. Table 1 summarizes the annealing temperature and time combinations and the resulting permeability values. The permeability was measured with an induction bridge at 10 kHz frequency, 8-turn jig and 100 mVac excitation.

TABLE 1 Annealing Conditions Permeability DC Bias lOKHz Core loss (W/h ; g) @ioKHZ 50 Oe 80 Oe @ioo kHz, 0.035 T 450 C/4hrs 191 14 8 450 C/4hrs 213 11 7 450 C/7hrs 121 20 12 450 C/8hrs 212 13 7 450 C/8hrs 218 11 7 450 C/1 Ohrs 207 12 7 19 450 C/10hrs 212 15 8 12 450 C/6hrs 203 18 10 14 460 C/4hrs 124 24 15 460 C/4hrs 48 74 41 470 C/15min 500 6 1 2.5 470 C/30min 145 17 8 13 470 C/1hr 189 15 6 10 470 C/lhr 132 23 11 14 470 C/2hrs 45 78 41 470 C/2hrs 47 76 40 53 470C/3. 5hrs 45 75 37 480 C/15min 43 75 35 65 480 C/15min 44 40 32 56 480 C/1 hrs 46 77 37 480 C/1 hrs 47 81 38 47 490 C/15min 46 76 37 490 C/15min 46 80 38 490C/30min 46 82 39 490 C/30min 46 78 36 AlloyFe80B11 Si9 Tx=508 C As illustrated by Figure 1, 80% of the initial permeability was maintained at 50 Oe while 30% of the initial permeability was maintained at 100 Oe. The core loss was determined to be 650 W/kg at 1000 Oe excitation and 100 kHz Figure 2 depicts, as a function of the DC bias excitation field, the percent of the initial permeability of an Fe base amorphous core, the core having been cut with an abrasive saw and having had inserted therein a discrete plastic spacer

having a thickness of 4.5 mm. The initial permeability of the Fe base core was 3000 and the effective permeability of the gapped core was 87. The core retained 90% of the initial permeability at 100 Oe. However, the core losses were 250W/kg at 1000 Oe excitation and 100 kHz.

Figure 3 depicts, as a function of the DC bias excitation field, the percent of initial permeability of an Fe base core having, in combination, a discrete gap of 1.25 mm and a distributed gap. The amorphous Fe base alloy can be partially crystallized using an appropriate annealing temperature and time combination, provided its crystallization temperature and or chemical composition are known.

The example shown in Figure 3 had a composition consisting essentially of Fe80Bl, Si9 and a crystallization temperature Tx=507 °C. The annealing temperature and time were 430 °C and 6.5 hr, respectively and the annealing was performed in an inert gas atmosphere. This annealing treatment reduced the permeability to 300. Subsequently, the core was impregnated with an epoxy and acetone solution, cut with an abrasive saw to produce a discrete gap and provided with a plastic spacer of 1.25 mm, which was inserted into the gap. Impregnation of the core is required to maintain the mechanical stability and integrity thereof core during and after the cutting. The final effective permeability of the core was reduced to 100. At least 70 % of the initial permeability was maintained under 100 Oe DC bias field excitation. The core loss was 100 W/kg at 1000 Oe excitation and 100 kHz.

Figures 1,2 and 3 illustrate that in order to improve the DC bias behavior of an Fe base amorphous core while, at the same time, keeping the initial permeability high and the core losses low, a combination of a discrete and distributed gaps is preferred.

The conventional formula for calculating the effective permeability of a gapped choke is not applicable for a core having in combination a discrete and a distributed gap. Figure 4 depicts, as a function of the discrete gap size, empirically derived contour plots of the effective permeability for a core having

combined discrete and distributed gaps. The different contours represent the various values of the distributed gap (annealed) permeability.

Table 2 displays various combinations of annealed permeability and discrete gap sizes. The corresponding effective permeability, percent permeability at 100 Oe and core losses are listed, as well as the cutting method and the type of the spacer material.

TABLE 2 Annealed Perm Spacer (mm) Effective Perm % Perm @ 50 Oe % Perm @ 100 Oe Core loss (W/kg) 300 1.25 107.2 93.4 74.4 87 300 1.25 103.4 91.6 74.6 91 300 1.25 101.5 93.1 74.6 86 300 1.25 97.3 93.6 77.6 100 300 1.25 97 94 78 34* 300 1.5 96 94 79 34* 300 2 87 94 82 40* 300 2.5 81 94 84 45* 300 3 75 95 86 51* 300 4.5 65 97 91 63* 300 8.25 53 98 93 68* 300 12.75 43 99 96 79* 300 1.25 105.2 92 72.4 86 1000 3.75 88.3 97.1 88.3 115 1000 3.75 85. 3 97.2 89.4 109 250 0.5 129.3 82.3 50.4 105 250 0.75 111.8 84.4 58.7 170 250 1.5 91.8 92.5 73.4 212 450 0.5 177.5 89.9 18.3 108 450 0.75 158.9 91.9 33.3 101 450 1.5 118.8 95.9 77 110 450 2.25 100 95.7 86.4 96 350 1.5 104 95 78 110 350 1.5 105 94 77 117 350 1.5 103 95 79 114 350 1.5 104 95 79 115 350 1.5 99 95 79 112 450 2.25 94 97 87 98 450 2.25 95 95 81 111 450 2.25 94 96 83 105 Continue-Table 2 Annealed Pertn Spacer (mm) Effective Perm % Perm (@ 50 Oe % Perm @ 100 Oe Core loss (W/kg) 450 2. 25 96 95 82 120 580 3 89 97 85 106<BR> 580 3 89 97 90 103<BR> 580 3 92 98 90 110<BR> 580 3 89 97 88 104<BR> 2500. 75 110 85 58 89<BR> 2500. 75 91 93 74 101**<BR> 2500. 75 118 82 57 89***<BR> 2500. 75 124 82 54 99*** 250 0. 75 117 84 57 89*** 250 0. 75 115 85 58 90*** Continue-Table 2 Cutting Method Spacer Type abrasive saw plastic abrasive saw plastic abrasive saw plastic abrasive saw plastic abrasive saw plastic abrasive saw plastic abrasive saw plastic abrasive saw plastic abrasive saw plastic abrasive saw plastic abrasive saw plastic abrasive saw plastic abrasive saw plastic abrasive saw plastic abrasive saw plastic abrasive saw plastic abrasive saw plastic abrasive saw plastic abrasive saw plastic abrasive saw plastic abrasive saw plastic abrasive saw plastic abrasive saw plastic abrasive saw plastic abrasive saw plastic abrasive saw plastic abrasive saw plastic abrasive saw plastic abrasive saw plastic abrasive saw plastic abrasive saw plastic abrasive saw plastic abrasive saw plastic abrasive saw plastic abrasive saw plastic wire edm plastic water jet plastic abrasive saw ceramic abrasive saw plastic abrasive saw ceramic abrasive saw plastic Continue-Table 2 Core loss was measured at 1000 Oe excitation field and 100 kHz with the exception of * Excitation field 500 Oe ** Excitation field 850 Oe *** Excitation field 900 Oe

Two different types of spacer material, plastic and ceramic, were evaluated. No difference was observed in the resulting properties. Typically the magnetic core is placed in a plastic box. Since a plastic spacer can be used for the gap, the spacer can be molded directly into the plastic box.

Several methods for cutting the cores were evaluated, including an abrasive saw, wire electro-discharge machining (wire edm), and water jet. All these methods were successful. However, there were differences in the quality of the cut surface finish, with the wire edm being the best and the water jet the worst. From the results in Table 2, it was concluded that the wire edm method produced cores exhibiting the lowest losses and the water jet method the highest, with all other conditions being equal. The abrasive method produced cores with satisfactory surface finish and core losses. From the above results it was concluded, that the finish of the cut surface of the core is important for achieving low core losses.

Having thus described the invention in rather full detail, it will be understood that such detail need not be strictly adhered to but that further changes and modifications may suggest themselves to one skilled in the art, all falling within the scope of the invention as defined by the subjoined claims.