Self-Protected Semiconductor Protection Component

1. Field of the Invention

The present invention relates to semiconductor protection components, realized in monolithic form. Such protection components are intended to clip a voltage applied to their terminals when this voltage exceeds a predetermined threshold. During the clipping phase, the voltage across the component is maintained at the predetermined threshold when the component is of the avalanche diode type or drops to a practically zero value when the protection component is a four-layer component of the break over type, such as a gateless triac.

2. Discussion of the Related Art

In FIG. 1 a protection component 1 is conventionally parallel connected to a line between a supply source 2 and supply input terminals 3 and 4 of a circuit 5 to be protected.

As indicated above, such components can be realized in various ways. FIG. 2 shows a structure of an avalanche diode formed on an N-type semiconductor substrate 10. The front surface area of this substrate includes a P-type region 11 and the rear surface area an N-type overdoped region 12. The upper surface of the substrate is coated with an isolating layer 13, usually silicon oxide, which serves to delineate the area where region 11 is formed. The upper surface is coated with a metallization 15 contacting region 11.

Conventionally, the lower surface of the substrate is mounted on a base 16 serving as a heat sink, by means of a solder preform 17 and metallization 15 is directly welded to a connection "nail".

A protection component as illustrated in FIG. 2 is intended to absorb overcurrents or overvoltages between metallization 15 and base 16. The size of this component is devised so as to absorb an overload up to a predetermined energy limit. However, there always exists an over-current level liable to destroy the component. Usually, such a destruction results from the occurrence of a shorting region 18 between the upper and lower metallizations. Then, once overvoltage is over, the component remains shorted and, in the case of a circuit such as shown in FIG. 1, circuit 5 is no longer fed between its terminals 3 and 4 because the source 2 is shorted. This situation constitutes a double drawback. To set circuit 5 into operation again, it is necessary to disconnect component 1 and, if it is desired to protect the circuit, component 1 must be replaced by a new component. This operation is always critical and, in some applications, such as for circuits used in satellites, requires the provision of a double amount of components and switching circuits that adversely affect the protection function quality.

To solve this problem, the practice is to use over-sized protection components capable of absorbing high overvoltages which are very unlikely to occur.

An object of the present invention is to provide a monolithic semiconductor protection component that remains operational even after the occurrence of a normally destroying overvoltage.

To achieve this object and others, the invention provides a monolithic semiconductor protection component connected between a first semiconductor region disposed on a first surface of a semiconductor chip and a second semiconductor region disposed on a second surface. The first region is divided into several non-overlapping areas. Each area is connected to a common electrode through a fuse.

According to an embodiment of the invention, each fuse is sized so as to open under normal supply conditions of the circuit to which the component is connected, when a shorting occurs between the area corresponding to the fuse and the second region.

According to an embodiment of the invention, each fuse is made of a gold filament.

According to an embodiment of the invention, each fuse is made of a metallization.

According to an embodiment of the invention, the protection component is an avalanche diode. The first region is a region of the first conductivity type formed in a first surface of a substrate of the second conductivity type, and the second region is a region of the second conductivity type formed in the second surface of the substrate.

According to an embodiment of the invention, the protection component includes two first regions, each of which forms a protection component between itself and the second region, and each first region is divided into non-overlapping areas. Each area of the first group of areas is connected by a fuse to a first electrode. Each area of the second group of areas is connected by a fuse to a second electrode.

In other words, the present invention provides a multi-cellular protection component such that, when a cell is destroyed by an overvoltage, a fuse associated with this cell is destroyed but the component continues operating due to the remaining good other cells.

The protection circuit according to the invention can be associated with an alarm detector which indicates that a very high overvoltage has occurred on the line to be protected. Then, it will be possible to change the component during a maintenance operation at a convenient time, the component being still operational and its associated circuit still operating. In case of a five-cell component, it can be assumed that the component will still be operative even if three overvoltages have destroyed three individual cells.

The foregoing and other objects, features, aspects and advantages of the invention will become apparent from the following detailed description of the present invention which should be read in conjunction with the accompanying drawings.

FIGS. 1 and 2, above described, are intended to illustrate the state of the art and the problem encountered;

FIGS. 3A, 3B and 3C illustrate an embodiment of the present invention in three successive states, before the occurrence of a destroying overvoltage, during its occurrence, and shortly after the occurrence of the overvoltage; and

FIG. 4 shows another embodiment of the invention constituting a bidirectional protection component.

In the various drawings showing cross sectional views of semiconductor components, as is conventional in integrated circuit representation, the sizes of the various layers are not drawn to scale; more particularly, their thicknesses are arbitrarily enlarged to simplify drawings and to facilitate legibility of the figures. Those skilled in the art will be able to select the various thicknesses in order to obtain the desired characteristics.

FIG. 3A shows a protection component of the same type as the component of FIG. 2, modified according to the invention. In these figures, the same elements are designated with the same reference numerals. According to the invention, the P-type region 11 is replaced with several non-overlapping P-type regions 21, each of which is coated with a metallization 25. Each metallization 25 is connected through a fuse 26 to an anode electrode 27. In the embodiment represented, the anode electrode 27 is a connecting lug and each fuse 26 is made of a gold filament, such as a filament of 50 .mu.m in diameter. In practice, these filaments may also be of aluminum and are analogous to the filaments usually used to achieve contacts on metallizations of integrated circuit chips. Thus, under normal operation conditions, the structure of FIG. 3A is a multi-cellular structure (a four-diode structure in the embodiment represented) having the same behavior as the diode of FIG. 2.

In case of a normal overvoltage, the cells operate in avalanche mode in parallel. The energy is evenly distributed among the cells since the junctions are identical and have identical reverse characteristics due to their monolithic structure. The normal operation is therefore the same as for a conventional component.

In case of an excessive overvoltage, resulting in an overcurrent having a value higher than the current flow capacity of the chip, the shorting phenomenon described in relation with FIG. 2 will occur. However, with the structure according to the invention, as is illustrated in FIG. 3B, a shorting region 28 will occur on a single cell. Experiment has shown that this mode of destruction corresponds to the occurrence of a fuse channel in the chip between its upper and lower contacts, shorting the PN junction. This channel associated with a current focusing mechanism is limited to a few hundreds .mu.m in diameter as soon as destruction occurs. Thus, a single cell will be destroyed, as shown by shorting region 28 in FIG. 3B. This destruction phenomenon is liable to occur very rapidly, in approximately a few tens of microseconds. Then, referring back to the mounting diagram of FIG. 1, the current from voltage source 2 flows through the protection component even after the end of the over-voltage. According to the invention, connection filaments 26 are chosen so as to melt under the influence of the source current. As illustrated in FIG. 3C, the whole shorted source current passes in one of the filaments serving as a fuse and the filament will melt and disconnect the damaged cell. Then, the component regains its initial characteristic (its protection function) but will include three parallel diodes only instead of four diodes. The only difference is that the excessive over-voltage values liable to destroy a cell of the component will be slightly lower when the component has the configuration of FIG. 3C than when it has the configuration of FIG. 3A.

In the above description, an embodiment of the present invention has been described in relation with a unidirectional protection diode. The invention also applies to the realization of a bidirectional protection diode. FIG. 4 shows an exemplary embodiment of a bidirectional protection diode according to the invention. FIG. 4 again shows a multicellular diode in which half of the cells are connected to a first electrode 31 and the other half of the cells are connected to a second electrode 32. Base 16 no longer constitutes an electrode but simply a heat sink, and a rear surface conductor enabling current to flow in case of overvoltage between electrodes 31 and 32. For example, in case of a negative overvoltage on electrode 31, the diodes on the left part of FIG. 4 switch in the avalanche mode, and a current flows from electrode 31 to base 16, then from base 16 to electrode 32 through junctions PN which are forward biased in this case. In the same way as in the previous example (the unidirectional protection diode illustrated in FIGS. 3A-3C), in case of excessive overvoltage, one of the filaments 26 connected to electrode 31 or 32 will be destroyed.

The gold filaments 26 are sized so as to be destroyed once a shorting current has flowed during a long time duration with respect to the duration of an overvoltage pulse, such as approximately 10 milliseconds.

In a specific embodiment of the invention, comprising a monolithic chip including a couple of five diodes having avalanche voltages of 27 volts, the surface per diode being 1 mm.sup.2, it can be appreciated that a diode can be destroyed by the occurrence of a 8/20 wave (that is, an overvoltage having a rise time of 8 microseconds and a fall time of 20 microseconds) and a 280-ampere intensity during a shock wave. Then, provided the gold filaments have a diameter of approximately 50 .mu.m, the damaged diode is shorted due to melting of the gold filament by a 3.5-A current for approximately 10 milliseconds, when the supply voltage source has a voltage of approximately 15 V.

Various modifications can be made to the invention which has been described above in a specific embodiment.

In the foregoing, the described fuses 26 are formed by gold filaments connecting a metallized area to an electrode or lug. Those skilled in the art will be able to use materials other than gold to realize calibrated filaments. Additionally, each diode metallization can be connected to an electrode formed by a metal area deposited onto the semiconductor chip by means of a fuse formed on the chip, the fuse being constituted by any known technique, for example in the form of a thin metal deposit of the same material as the material forming the contacts or any other selected material.

The invention has been described above in the case where the protection component is a diode intended to operate in avalanche mode. Those skilled in the art will appreciate that other protection components lend themselves to a multicellular structure according to the invention. Indeed, in other protection components such as thyristors or gateless triacs, the destruction of a protection component corresponds to the shorting of a small area of the protection component.

Having thus described one particular embodiment of the invention, various alterations, modifications, and improvements will readily occur to those skilled in the art. Such alterations, modifications, and improvements as are made obvious by this disclosure are intended to be part of this disclosure though not expressly stated herein, and are intended to be within the spirit and scope of the invention. Accordingly, the foregoing description is by way of example only and is not intended as limiting. The invention is limited only as defined in the following claims and the equivalents thereto.