US 3179773 A
Abstract available in
Claims available in
Description (OCR text may contain errors)
April 20, 1965 K. v. KEELEY, SR 3,179,773
HIGH SPEED CURRENT INTERRUPTING ELECTRIC FUSES Filed Sept. 24. 1962 FIG. 2. FIG. 3.
INVEN-TOR United States Patent Ofiice 3,179,773 7 HIGH SPEED CURRENT INTERRUPTING ELECTRIC FUSES Kedric V. Keeley, Sr., 1600 Wildwood Drive, Los Angeles 41, Calif.
Filed Sept. 24, 1962, Ser. No. 225,774 6 Claims. (Cl. 200-120) My invention is an improvement in overcurrent protection fusible link type electric fuses and, more particularly, in a form of construction of such a fuse which will reliably interrupt excess current with sufficient speed to adequately protect modern-day semiconductors and which form of construction is adaptable to economic manufacture.
The phrases fusible link and fusible conductor and the word link have one and the same meaning in this specification. The phrase effective diameter means the diameter of a theoretical circle which circle has the same area as an actual cross section.
The use of semiconductors has become an important part of the electronic field and conventional fuses will often not interrupt excessive current with suflicient speed to prevent the excess current from damaging semiconductors located in the circuit. Some prior art fuses of l ampere rating or less will interrupt excessive current with sufiicient speed to protect semiconductors provided the circuit voltage source is sufficiently low that negligible vaporization of the link occurs within the fuse during the current-interrupting process. More particularly, a fuse of my invention will reliably interrupt excess current even though the potential of the voltage source is suflicient to vaporize the entire portion of the link and regardless of normal current rating.
A usable figure of merit of the speed of operation of a current-interrupting fuse is the length of time required for the fuse to interrupt excessive current when said current is suddenly increased from the maximum pure D.C. continuous rated current value of the fuse to 10 times this value, While the potential of the current source remains at 50 volts or higher. In this figure of merit, smaller numbers are the more desirable. For the purpose of this specification, the figure of merit just described will be called the 10X rating. For adequate protection of semiconductors, a fuse with a 10X rating on the order of milliseconds is desirable.
Another object'of my invention is to provide an improved type of fuse which will reliably and consistently produce a x rating of 5 milliseconds or less. Another object of my invention is to provide a form of economical construction of an improved fuse, the normal rating, the 10X rating and other ratings of which can be predictably controlled in the manufacturing process. Further objects and advantages of my invention will become more apparent as the following description proceeds and the features of novelty and usefulness characterizing the invention will be pointed out in greater detail.
For a better understanding of my invention, the accompanying drawing has been included in which FIG. 1 is an elevational view, partly in section, of an improved fuse embodying my invention and employing a single fusible link.
FIG. 2 is a section on line 2-2 of FIGURE 1.
FIG. 3 is a section on line 3-3 of FIGURE 1.
Referring to FIG. 1, my invention consists of a fusible conductor 1 of effective diameter x, a portion 2 of conductor l and of length y being suspended in a gaseous atmosphere in chamber 4 and a portion 3 of conductor 1 and of length 1 being encased in a fine textured solid insulating material 5. The fusible conductor 1 is connected at its distant ends to the usual terminals 6 and 7 3,179,773 Patented Apr. 20, 1965 and terminals 6 and 7 are rigidly spaced apart by the usual insulating means 8.
Typical operation of a fusible link type, overcurrent electric fuse, when the potential of the current source remains sufiiciently high to cause an appreciable are within the fuse, is as follows:
Excessive current causes a hot spot to develop at some point along the fusible link. Metallic separation of the fusible link occurs at this hot spot, due to melting and/or evaporation, and an electric arc is formed bridging said separation. This arc conducts current while the high heat of the arc current evaporates an additional portion of the fusible link and said arc continues to conduct current until the arc length becomes too great for the voltage to maintain the are over the length of evaporated fusible link or until a suitable arc path no longer exists, due to vaporization of the fusible link to a point within some form of insulating arc-suppression material. The speed of operation is, therefore, dependent upon the rate of fusible link evaporization and on the length of conductor necessary to be evaporated before an arc-suppression point is reached.
In A.C. operation, some prior art fuses are able to achieve full interruption of overcurrent when and because the voltage of the A.C. source drops toward zero in its regular cyclic variations. Since most A.C. sources have a frequency of 60 cycles per second, which gives a halfcycle time of approximately 8 milliseconds, it is not feasible to depend on the cyclic zero periods of the A.C. voltage in A.C. circuits to bring about overcurrent protection for semi-conductors. Therefore, when considering 10 ratings in this specification, no attempt is made to differentiate between A.C. and D.C. current since the current-interrupting periods under consideration are shorter than one-half cycle of 60 c.p.s. alternating current.
Prior art (Alford, Pat. 2,159,649) has established that the relative speed with which a fusible link disintegrates increases as the maximum continuous current density per cross section of link is increased. Alford further recognizes that said continuous current density can be increased by shortening the length of the fusible link, in which case conductive heat dissipation from the link to the more massive materials at the ends of the link is increased and permits the link to be operated at a greater continuous current. Alford also shows that continuous rated current density in a fusible link can be increased by using a plurality of links of smaller cross-sectional area as compared to a single link of larger cross-sectional area.
Prior art (Alford, Pat. 2,159,649) has also established that continuous current density in a fusible link can be increased by completely surrounding said link with a liquid or solid material capable of conducting the heat generated by the electric current away from the link. I have discovered, however, that if the full length of a small fusible link is encased in insulating material, not of fiuid nature, the point at which the hot spot develops along said fusible link and the current value at which said hot spot will develop in the fusible link is inconsistent and is difiicult to predict in manufacture because the cooling effect of the insulating material along the link depends upon the intimacy of physical contact between the covering of insulating material and the link. Should the insulating covering material be of granular nature or of solid nature possessing voids near the link, which voids are of a dimensional size on the order of the effec tive diameter of the link or greater, the link hot spot will develop at such a point where conductive cooling is the least. For this reason, difliculties are experienced in manufacturing a link fuse of predictable rating when the entire length of said link is encased in a solid or granular material and when the fusible conductor is of saver/a relatively small cross-sectional area, on the order of 30 square mils or less.
I have further discovered that a link-type fuse, having a fusible conductor of cross-sectional area on the order of 30 square mils or less, can be improved in manufacturing control of the characteristics of the hot spot by suspending a portion of the length of the fusible conductor in a gaseous atmosphere, the length of said portion being not less than times greater than the largest cross-sectional dimension of said fusible conductor. Under these condi tions, the effective cooling of the link by the gas is always less than the effective cooling of the link by any contacting solid material, and the intimacy of contact between the gas and conductor is naturally consistent and therefore the hot spot will always develop near the center point of the portion of length of link suspended in the gaseous atmosphere, and the current required to crease a disruptive hot-spot in this construction can be accurately and consistently predicted.
When a disruptive arc is caused to occur in a fuse link suspended in a gaseous atmosphere and when an adjacent portion of the link is surrounded by and in intimate contact with a solid insulating arc-suppression material, the length of link in the gaseous atmosphere will become vaporized following which the action of the arc is described in prior art by conflicting statements. In Pat. 462,452, Rice states that metallic vapors expelled from the insulating material extinguishes the arc while Trent, in Pat. 1,484,198, states that fusion within the covered section of the link is merely delayed. Such a delaying action would, of course, lengthen the time required for complete interruption of the current and is not compatible with fast action.
I have discovered that when the cross section of the link is maintained at 30 square mils or less the presence of the insulating material contributes no detectable delay to the operating speed of the fuse and that the tendency of the voltage source to maintain the arc inside of the insulating material is greatly reduced by such a small conductive cross section. I have also discovered that with a link of cross section of 30 square mils or less the ability of the arc suppression material to suppress the arc is best when the largest parts of the texture of the arc suppression material are smaller than the effective diameter of the ink, and if the effective diameter of the fuse link is maintained on the order of 6 mils or less and if the length of link covered by the insulating fine textured material is not less than 10 times the link diameter, reliable, consistent, and predictable extinguishment of the arc can be effected.
In further regard to normal current density in a fuse link, while materials are available which would permit the operation of a fuse link at or above an incipient red heat temperature, such operation is undesirable because the link material can actually boil away at these temperature values which would reduce the life of the fuse, and such high temperature operation would also otter certain fire hazards when the fuse is operated at its normal rated current.
I have also discovered that the rated continuous current density in a fusible link can be made relatively high without producing red heat by using certain materials of relatively good conductivity and operating the hottest spot in the link at a temperature on the order of 400 C., which is well below incipient red heat. The better conductivity gives a low resistance value of the fusible link and consequently permits a relatively greater current to flow for operation at a given temperature. The use of materials of relatively good conductivity also enhances the characteristics of a fuse in that less superfluous resistance is introduced into any circuit in which such a fuse is employed.
Alford, in Pat. 2,159,649, points out that the time required for a fusible link to evaporate is dependent upon the mass of the link, the heat required to raise said link to the evaporation point, and the latent heat of vaporization of the material. The mass of material required for a fusible link of given rating is directly related to its specific gravity and inversely related to tie maximum permissible current density. The permissible maximum current density is, in turn, inversely dependent on the resistivity and thus said mass is directly dependent on the product of specific gravity times specific volume resistivity of the link material. The heat required to bring a link to the vaporization point is directly dependent on the number of degree temperature by which the link must be raised before the boiling point is reached, and on the specifiic heat of the material, and on the latent heat of fusion. The heat required for actual vaporization is dependent on the latent heat of vaporization.
Through numerous calculations, which in turn have been confirmed by numerous tests, I have discovered that the suitability of materials for use as a fast-acting fusible link can be determined by the Total Heat Capacity Formula where R is the specific volume resistivity of the material in micro-ohm centimeters at 20 C.,
G is the specific gravity of the material at 20 C.,
H is the specific heat of the material in calories per gram at 20 C.,
T is the temperature in degree centigrade at which the material vaporizes,
H is the latent heat of fusion of the material in calories per gram, and
I is the latent heat of vaporization of the material in calories per gram.
In this calculation, low numbers are more desirable than high numbers; for example, the calculation for gold in accordance with the above formula results in a figure of approximately 25,000 While for tungsten the calculation in accordance with the formula gives a figure of approximately 150,000. In carefully conducted experiments, tungsten as a fuse link material wa found to require several times longer to interrupt current overloads than was required by gold fuse links of comparable normal current ratings. I Since it is advantageous to establish the normal operating temperature of a fuse link on the order of 400 C., there are certain materials which, while they would be otherwise satisfactory as a fast-acting fuse link material, do not possess sufficient strength to provide self-support at 400 C. when suspended in a gaseous atmosphere and are, therefore, not suitable for use as fuse link materials in a fuse of my invention. Again, on conducting numerous tests, it was found that materials with a melting point near or below 400 C. were not satisfactory link materials for use in a fast-acting link fuse of my invention.
Furthermore, certain materials which were otherwise thought to be satisfactory as fuse link material for my invention were found to possess considerable tendency at 400 C. to combine with ordinary air; therefore, if these materials are used in a fuse of my invention, the gaseous atmosphere must be of a special nature as to be inert with respect to the link material, and air must be sealed out.
I have invented an improved fuse employing the facts and discoveries given above, which fuse has a l0 rating of 5 milliseconds or less and all rating of which fuse can be accurately reproduced in manufacture.
Referring to FIGURE 1, the material of which fusible conductor 1 is made has a Total Heat Capacity Formula figure, as previously described, of less than 50,000 and is preferably on the order of 25,000.
The effective diameter x, of fusible link 1 is not greater than 6 mils. The length y of portion 2 of fusible conductor- 1 is many times greater than its effective diameter x such that The length z of portion 3 of fusible conductor 1 within encasing material 5 is many times greater than the effective diameter x of fusible conductor 1 such that Encasing insulating material 5 is of a non-fluid nature and its texture is finer than the effective diameter x of fusible conductor '1. The nature of portion 2 of fusible conductor 1 and the gaseous atmosphere in chamber 4 are such that they will not combine when portion 2 of the fusible conductor 1 is operated at a temperature on the order of 400 C.
The operation of a fuse of my invention is as follows: When an overcurrent flows through fusible conductor 1, the cooling effect of the encasing insulating material on portion 3 and the cooling effect of the supporting means of portion 2 of fusible conductor 1 confines the development of the hot spot in fusible conductor 1 near the center part of portion 2. The hot spot causes portion 2 of fusible conductor 1 to separate and an electric arc is developed across said separation of portion 2 of fusible conductor 1. When the power source of said overcurrent has sufficient voltage to maintain the electric arc throughout the evaporation of portion 2 of fusible conductor 1, such evaporation progresses to the point where the electric arc bridges substantially the length of portion 2 of fusible conductor 1. At this point of progress, the electric arc has difficulty following the vaporization of the small fusible conductor 1 into the body of encasing insulating material 5 and the are therefore becomes quickly extinguished, resulting in the complete interruption of the current.
If the power source of .the overcurrent does not have sufficient voltage to maintain the arc to fully evaporate portion 2 of fusible conductor .1, then of course the arc becomes extinguished before vaporization of portion 2 is complete and current interruption occurs in suflicient time to provide the desired protection of the circuit.
The time required for full interruption of the current as described above is controllably small and if the initial overcurrent mentioned above is 10 times more than the maximum continuous current rating of the fusible conductor 1., the time required for complete interruption of the overcurrent even though the potential of the voltage source is more than 50 volts, is not more than 5 milliseconds, and greater overcurrent values produce even quicker interruptions.
On making tests using various values of source voltage and subjecting fuses of my invention to sudden large current over loads .1 have discovered that as said source voltage was increased from a value of approximately 50 volts, the time integral of the fuse let-through current remained fairly constant and satisfactorily low until the source voltage was increased above a certain value after which said current-time integral increased rapidly with increases in source voltage. This point of discontinuity in the relationship between source voltage and currenttime integral occurs at a relatively high value of voltage compared to the dimensions used on the fuses of my invention being tested and it has been established, therefore, that a fuse of my invention possesses characteristics favorable towards use in circuit miniaturization. For instance, in a fuse of my invention in which the fusible link 1 had an effective diameter x of .002" and the length z of portion 3 encased in insulating material 5 was .15" long and the length y of portion 2 in the chamber 4 was .125" long, the source voltage required to produce the discontinuity described above was on the order of 250 v. Thus a fuse of my invention and of this small size could reliably be employed to protect semiconductors in circuits employing source voltages up to 250 v.
The actual normal continuous current rating of fusible conductor 1 is dependent upon the material of which said fusible conductor 1 is made, is further dependent upon the cross-sectional area of fusible conductor 1, and is further dependent upon the length of portion 2 of fusible conductor 1.
If the length y of portion 2 of fusible conductor 1 is made at least 10 times greater than the effective diameter x of fusible conductor 1, then the normal current rating of the resulting fuse can be easily controlled in manufacture since small percentage variations in the dimensions of fusible conductor 1 and in the length of portion 2 of fusible conductor 1, each causes only similarly small percentage variations in the final normal continuous current rating and other ratings of the fuse.
To construct a fuse of my invention with a normal current capacity greater than can be achieved through observing the limits of fusible conductor materials and cross-sectional areas described above, a plurality of physically and electrically identical fusible conductors can be employed electrically in parallel physically spaced apart 1 in one encasement or as a plurality of individual single link fuses connected externally in parallel. The operation of a fuse constructed with a plurality of links is similar to the above-described operation of a fuse constructed in accordance with FIG. 1, however, each of the individual fusible conductors operates substantially independently of the other fusible conductors except that each carries its equal share of current. Since these fusible conductors are physically and electrically identical, they operate to interrupt an overcurrent substantially simultaneously with each other. Small variations in the electrical and physical similarity of parallel connected fusible conductors are not of great importance since, should one said fusible conductor separate at its hot spot before any others do, an additional current load would be diverted to the others and a very rapid chain reaction of disintegration of all fusible conductors would occur. Therefore, the ratings of fuses manufactured in accordance with FIG. 1 or with a plurality of links are very similar excepting that fuses manufactured in accordance with a plurality of links Will possess continuous current ratings substantially equal to the rating of one fusible conductor multiplied by the total number of fusible conductors.
As a preferred embodiment of my invention, I have constructed a fuse in accordance with FIG. 1 in which the link 1 consists of a hard-drawn, gold wire of diameter x of 2 mils. The gaseous atmosphere in chamber 4 is ordinary air and the encasing insulating covering 5 is a silicone rubber compound. The overall length of the unit is approximately /8". The length y of portion 2 of the fuse link 1 suspended in the gaseous atmosphere in chamber 4 is .125" and the length 2 of portion 3 of the fuse link 1 encased in the insulating material 5 is .2". The overall diameter of the unit is A" and the unit fits into a fuse holder built for X A" tubular fuses. The long internal length of terminal 7 is used to maintain a overall length and yet limit the link length to .325". This unit possesses a continuous direct current rating of 1.7 amperes and a 10X rating of less time than 4 milliseconds. Its full length resistance, operating at its maximum steady direct curlrent of 1.7 amperes, is approximately .2 ohm. This design was repeatedly tested and found to be reliable protection for semiconductors by interrupting excessive currents from low impedance electric power sources of 300 volts or less.
Prior art discussions concerning the action of fuses have indicated that the ability of prior art fuses to interrupt over-currents have been dependent to some extent on the current capacity of the power source feeding the circuits in which said prior art fuses have been employed. Using a fuse of my invention, I have not found any such dependency. The maximum current capacity of any source is dependent upon the total effective impedance of the circuit, including the internal impedance of the source, and on the voltage generated by the source. I have tested circuits employing a source voltage on the order of 300 volts in which a fuse of my invention was' placed and in which the only other impedance of the circuit beside an open switch was the necessary wiring and the internal impedance of the source. On closing the switch in a circuit of this nature, I have found that the instantaneous current rose to high values but that the voltage of the source, measured across the switch and fuse in adjacent series relationship, did not diminish appreciably and yet my fuse successfully interrupted the current in sufficient time to be deemed satisfactory for the protection of semiconductors.
In these tests it is therefore evident that, as the switch was being closed and high current began to flow, the impedance of the fuse began to rise limiting the peak current to a value well below the maximum capacity of the source; therefore, the use in a circuit of a high current capacity source does not invalidate the usefulness of a fuse of my invention. In further tests, when a semiconductor with a continuous current rating equal to that of the fuse was added to the circuit, the high current resulting when the switch Was closed was interrupted with sufficient speed that the sensitive semiconductor was not damaged.
In my preferred embodiment, I have chosen gold as a link material because of its availability as finely drawn wire, its inherent inertness, its possession of adequate mechanical strength at a temperature near 400 C., its solderability for connection to the terminals, and its low figure in accordance with the Total Heat Capacity Formula previously described. Aluminum was found to be a satisfactory link material except for the difficulty encountered in connecting to the terminals. Copper was found to be a less satisfactory link material because it would deteriorate in ordinary atmosphere and so require a less common gaseous atmosphere and the necessary related hermetic sealing. Silver was found to be an unsatisfactory link material in that it did not possess sufficient mechanical strength to remain suspended intact in an atmosphere of ordinary air at a temperature near 400 C. (This may be due to the known tendency of silver to absorb oxygen at elevated temperatures.)
Silicon rubber was chosen as the arc-suppression, insulating material because it can be easily formed over the fusible conductor and Will withstand high temperatures and exposure to arcing.
In my preferred embodiment, greater continuous current ratings can be achieved through the use of a larger diameter fusible conductor up to the specified limit of 6 mils or through the use of a lesser length of fusible conductor in the gaseous atmosphere, down to the specified limit of diameters. Greater voltage ratings can be obtained by using longer lengths of arc suppression material.
The claims are:
1. In an improved fusible link type fuse, the improvement comprising: a fusible conductor having an effective and substantially constant diameter x along it full length of not more than 6 mils and of a sufficient effective diameter to carry a normal operative current, with a portion of said fusible conductor of length y being suspended in a gaseous atmospheric compartment of said fuse, and an adjacent portion of said fusible conductor of length 2 being encased in a nonfluid arc-resistant insulating material within said fuse; the length y of said fusible conductor suspended in said gaseous atmosphere being not less than ten times greater than the effective diameter x of said fusible conductor and the length z of said fusible conductor encased in said arc-resistant material being not less than ten times greater than the effective diameter x of said fusible conductor with the texture of the arc-resistant insulating material being finer than the effective diameter of the fusible conductor, said fusible conductor being made of material wherein the value of the following expression is less than 50,000:
R is the specific volume resistivity of the material in micro-ohm centimeters at 20 C., and has a value less than 10;
G is the specific gravity of the material at 20 C.;
H is the specific heat of the material in calories per gram at 20 C.;
T is the temperature in degrees centigrade at which the material vaporizes;
H is the latent heat of fusion of the material in calories per gram; and
H is the latent heat of vaporization of the material in calories per gram.
2. A fuse in accordance with claim 1 wherein the melting point of the fusible conductor is greater than 400 C.
3. A fuse in accordance with claim 1 wherein the material of the fusible conductor, suspended in the geasous atmosphere, and said gaseous atmosphere are both of such a nature that the two have no tendency to combine when the fusible conductor is maintained at a temperature of 400 C.
4. A fuse in accordance with claim 1 wherein the material of the fusible conductor suspended in the gaseous atmosphere and said gaseous atmosphere are both of such a nature that the two have no tendency to combine when the fusible conductor is maintained at a temperature of 400 C. and wherein the melting point of the fusible conductor is greater than 400 C.
5. A fuse in accordance with claim 4 wherein the fusible conductor is formed of gold.
6. A fuse in accordance with claim 4 wherein the fusible conductor is formed of aluminum.
References Cited by the Examiner UNITED STATES PATENTS 856,292 6/07 Phelps 200l3l 1,278,893 9/18 Eustice 20013 5 2,159,649 5/39 Alford 200- 2,768,264 10/56 Jones 200-l44 2,921,250 1/60 Swain 200l20 BERNARD A. GILHEANY, Primary Examiner.
UNITED STATES PATENT OFFICE CERTIFICATE OF CORRECTION Patent No. 5,179,773 April 20, 1965 Kedric V. Keeley, Sr.
It is hereby certified that error appears in the above numbered patent requiring correction and that the said Letters Patent should read as corrected belowo Column 2, line 20, for "evaporization" read vaporization column 3, line 18, for "crease" read create column 4, line 67 for "ratin read ratin s column 6 line 64 for g g 3 Q Q "curlrent" read current Signed and sealed this 8th day of March 1966.
( L) Attest:
ERNEST W. SWIDER EDWARD -J. BRENNER Attesting Officer Commissioner of Patents UNITED STATES PATENT OFFICE CERTIFICATE OF CORRECTION Patent No 3,179,773 I April 20, 1965 Kedric V. Keeley Sr. 7
It is hereby certified that error appears in the above numbered patent requiring correction and that the said Letters Patent should read as corrected below.
Column 2 line 20 for "eva orization" read Va orization p p column 5, l1ne 18, for "crease" read create column 4, line 67 for "ratin read ratin s column 6 line 64 for g g "curlrent" read current Signed and sealed this 8th clay of March 1966.
( L) Attcst:
,IERNEST W. SWIDER EDWARD -J. BRENNER Attesting Officer Commissioner of Patents