US3486088A - Regenerative gate thyristor construction - Google Patents

Description

Dec. 23, 1969 GRAY ET AL 3,486,088

REGENERATIVE GATE THYRISTOR CONSTRUCTION File-d May 22, 1968 4 Sheets-Sheet 1 FIG. 1

N V N[ p /2\ /22 V9 N //v VEN 70/35 Donald 1'. gray John ff -fuzclzzzzg's CZIZdrew 772. ZZZoiziert Dec. 23, 1969 GRAY ETAL 3,486,088

REGENERATIVE GATE THYRISTOR CONSTRUCTION Filed May 22, 1968 4 Sheets-Sheet 2 '70 70 O p; N

Dec. 23, 1969 D. GRAY ETAL 3,486,088

REGENEHATIVE GATE THYRISTOR CONSTRUCTION Filed May 22, 1968 4 Sheets-Sheet 5 FIG. 3 F1663 I16 ll 84 9 .94 [76 1/4 58 52 \5 I I J60 Dec. 23, 1969 GRAY ETAL 3,486,088

REGENERATiVE GATE THYRISTOR CONSTRUCTION Filed May 22, 1968 4 Sheets-Sheet 4 nited States Patent 3,486,088 REGENERATIVE GATE THYRISTOR CGNSTRUCTION Donald I. Gray, St. Charles, John H. Hutchings, Geneva,

and Andrew M. Wohlert, St. Charles, 111., assignors to National Electronics, Inc, Geneva, IlL, a corporation of Illinois Filed May 22, 1968, Ser. No. 731,193 Int. Cl. H011 11/00, /00 US. Cl. 317-235 19 Claims ABSTRAUI OF THE DISCLOSURE A thyristor construction having a current carrying takeoif or other arrangement adjacent to the cathode emitter in the line of current fiow from the cathode to the trigger ing gate. The take-off voltage is utilized to provide at least one additional firing point on the face of the thyristor with each additional firing point being located adjacent to an edge portion of the emitter. This results in substantially reduced gate power, greatly improved di/dt, and maximum utilization of available device area. In a modified form of the invention, an area of weakness is deliberately formed in the device to provide preferential firing adjacent to the take-off in the event that the device is subject to excess voltage.

This invention relates to improvements in semiconductor devices known as thyristors which are defined in JEDEC standards as bistable semiconductor devices comprising three or more junctions, which can be switched from the oil-state to the on-state or vice versa, such switching occurring within at least one quadrant of the principal voltage-current characteristic.

When a thyristor is initially turned on, only a small area adjacent to the gate becomes conductive. This on state of the thyristor gradually spreads until the Whole junction is turned on.

Since only a very small area of the thyristor is initially on, the current density in this area is high, raising the possibility of burnout. Burnout may occur when a thyristor turns on from the forward blocking condition unless the current buildup thruogh the device is limited. Most presently available power thyristors have high switching losses since they turn on to full area conduction at a relatively slow rate.

In general, whether a thyristor is a low or high current device, it will tolerate a relatively low increase of current with time (di/dt) when considering industry needs. When the di/dt exceeds the device tolerance, catastrophic failure can occur. Even at levels of di/dt below the failure point, the switching heat generated in the device limits its usefulness, particularly at high repetition rates.

Saturable and non-saturable reactors that limit the current after turn-on have been used to minimize the difi'iculties of poor di/dt and poor area utilization. The saturable reactor gives the conducting region an opportunity to increase in area before the steep current increase occurs. One of the problems with this ararngement is that current spreading will not occur if the current density is too low during the reactor delay period. Also, device turnon characteristics vary over a wide range, and it is difficult to predict how a particular device will operate in this circuit. Thus, the current during reactor delay, the delay time relationship at the reactor, and the ability of the device to spread at low currents, are variable that must be considered in this kind of application. One big advantage of this kind of current limiting is that it allows reasonable protection to the device for voltage breakover.

"ice

Non-saturable air core reactors have also been used extensively to limit the current to safe values of di/dt. This type of reactor is simpler for the circuit designer to incorporate into his circuit, and for the manufacturer to pro-evaluate the devices going into that circuit, however, this type of reactor limits the di/dt at all current levels.

These additional circuit elements and circuit compromises represent a trade-off in circuit performance versus device performance and do not represent any improvement in the basic thyristor. Changes in the thyristor design itself have, however, been proposed. Specifically, gating schemes other than the side-fire type have been proposed for enhancing di/dt. The center-fire gate has been suggested and used in conventional devices in an attempt to obtain high di/dt, however, with a center-fire gate, the device will not turn on equally all around. On the contrary, there will still be a localized high current density area at the initiating spot with the consequent danger of burnout.

Multiple gates have been considered in an attempt to increase di/dt and reduce the turn-on losses. The multiple gate approach will give reduced turn-on losses at low di/dt levels, and if the initiating spots are on opposite sides of the emitter, total turn-on time at such levels can be reduced by nearly half. Multiple gates operate best when there is isolation between the gates or when separate gate supplies are employed. It has been found, however, that multiple gates will not significantly increase di/dt since one of the gates will always initiate first and set the stage for the following current to concentrate at the initiating point.

A ring gate has also been used, however, this arrangement has low sensitivity when compared with the sidefire gate, and it has not been used to any great extent.

Thyristors having extended emitter lips have also been used as an alternative to saturable and non-saturable reactors for purposes of limiting the current after turn-on.

The instant invention provides a new approach in thyristor design, and it is the main object of the invention to provide a thyristor device which provides greatly improved di/dt, which operates with substantially reduced gate power, and which accomplishes maximum utilization of available device area.

It is a further object of this invention to provide means for operation of thyristor devices which enables safe operation when the devices are subject to excess voltage.

These and other objects of this invention will appear hereinafter and for purposes of illustration, but not of limitation, specific embodiments of the invention are shown in the accompanying drawings in which:

FIGURE 1 is a schematic cross-sectional view of a conventional thyristor provided with test leads;

FIGURE 2 is a schematic cross-sectional view of a thyristor characterized by the features of this invention;

FIGURE 3 is a plan view of a particular connection design which can be employed in a thyristor characterized by the features of this invention;

FIGURE 4 is a schematic cross-sectional view of the thyristor shown in FIGURE 3;

FIGURES 5, 6 and 7 comprise schematic cross-sectional views of thyristors characterized by dip emitter characteristics in accordance with one form of the invention;

FIGURES 8 and 9 are plan views of interdigitated thyristors contemplated as alternative forms of the invention;

FIGURE 10 is a circuit diagram illustrating certain principles involved in the design of this invention;

FIGURE 11 is a circuit diagram illustrating conventional thyristors connected in parallel and modified in accordance with the principles of this invention;

FIGURES 12 and 13 are plan views of modified forms of the invention; and

FIGURE 14 is a circuit diagram illustrating constructions of the invention connected in series.

The invention will be described with reference to reverse blocking triode thyristors; it will be understood, however, that all thyristor types falling within the previously set forth definition are contemplated. Reference will be made herein to vertically layered devices; however, all configurations of thyristors, for example, a planar thyristor, having anode and cathode contacts on the same side, can be designed in accordance with principles of this invention.

With reference to FIGURE 1, the thyristor construction shown is characterized by a p-n-p-n body 12 with anode backing contact 14 and cathode backing contact 16 on the opposite faces of the body. A side gate 18 is applied adjacent the cathode emitter 19 in the conventional manner.

The emitter 19 defines a lip 20 which extends laterally beyond the contact 16. When a gate signal triggers the device, cathode current flows through this lip. Current flowing through this lip generates voltage across the lip which can be measured by means of the leads 24 and 26. As will appear, utilization of the voltage that appears along the lip 20 is an important factor in establishing the improvements of this invention.

The location of initial turn-on in a conventional thyristor is illustrated at 22 in FIGURE 1. As previously indicated, various means have been employed to limit the current until the turned on area is sufliciently large to avoid burnout. The actual initial turn-on area is extremely small and the illustration at 22 should not be considered representative of the true size of this area. Because of the extremely small area of initial turn-on, the current must be kept quite low since otherwise the current density would become dangerously lnigh.

FIGURE 2 illustrates a structure which also includes the semi-conductor body 12, contacts 14 and 16 and gate 18, but which also includes means for avoiding problems previously encountered. Specifically, in this construction, a lead is applied to the emitter lip, and this lead extends to a connection 32 at a separate location on the face. This arrangement has been found to provide vast improvements in turn-on characteristics since turn-on activity occurs at 34 in the conventional manner; however, the connection at 32 provides an additional area 36 which is triggered substantially simultaneously with the area 34.

FIGURES 3 and 4 illustrate a thyristor 40 which comprises one practical form of the invention. In this instance, the gate 42 is positioned opposite emitter lip 44. A wire lead is attached at 46 to the emitter lip, and this lead is then attached at a plurality of points 48 around the periphery of the device. It will be noted that there are ten contact points 48 positioned around the device whereby ten additional triggering areas are simultaneously available upon generation of the regenerative gating signal at 46. Emitter lips 50 are formed opposite each of the contact points 48.

FIGURE 10 provides a circuit diagram which will serve to illustrate the principle of regenerative gate operation.

The circuit diagram illustrates anode contact 14, cathode contact 16 and initiating gate contact 18. The resistance path 52 represents the emitter lip resistance such as shown at 20 in FIGURE 1. The designation of a thyristor at 54 is intended to represent the area initially turned on by the gate signal such as shown at 22 in FIGURE 1 or at 34 in FIGURE 2. The thyristors illustrated at 56 are intended to represent the areas such as shown at 36 in FIGURE 2 which are turned on by the regenerative gate signals. The resistance illustrated at 58 corresponds with the resistance of the individual emitter lips such as shown at 50 in FIGURE 3. Thus, the entire circuit of FIGURE 10 represents a single regenerative gate thyristor.

FIGURE 10 also illustrates a gate return lead 53. Such leads are often used in conventional devices, and the use of such leads is preferred for a thyristor subject to high di/dt circuit conditions.

The resistance 52 is preferably higher than the resistance 58. When the initiating gate signal is applied, current will immediately fiow from the cathode through resistance 52 to the thyristor area 54 for turn-on in this area. The current through thyristor area 54 creates a positive voltage in resistor 52 which will also immediately trigger thyristor areas 56 adjacent resistors 58. Where the resistance of the resistors 58 is lower than the resistance 52, the chances of overburdening of the initial turn-on area are reduced. The regenerative gate system thus provides a plurality of turn-on points which will very quickly develop full switching of the thyristor while at the same time greatly minimizing the danger of burnout.

The regenerative operation relies on the presence of some resistance between the point where the regenerative signal appears and the adjacent emitter portion. The provision of lips 50 extending beyond the contact 16 provides this since the contact will not short the signal and prevent firing of the emitter.

The arrangement of this invention provides an auxiliary regenerative gate supply built into the device which can immediately adjust to the circuit di/dt stress imposed on the device. With this device, it is not necessary for the initiating gate signal to do anything more than start conduction since the regenerative gate can serve the principal gating function. The regenerative gate thyristor can tolerate spurious gate signal or even sine wave firing without harm. This device also exhibits high di/dt ability, reduced losses during switching, and reduced time to full conduction.

The regenerative device uses the initiating gate only to locate the proper spot to trigger the device, and, therefore, problems normally associated with gating, such as the rise rate, pulse width, and amplitude of the pulse, do not pertain. The initiating gate current, after device triggering is accomplished, may even be reversed since the principal high power gating function is done by the regenerative gates.

If the regenerative thyristor is triggered by the gate supply in the proper spot, the current initiation is at the gate edge of the emitter lip. As already noted, the resistance of this lip can be controlled by suitable geometry and diffusion control so that some current limiting occurs. If the initiating lip resistance is made higher than the resistance of the lips gated by the regenerative gates, the current takes a new easier path and thus transfers the anode current away from the initiation spot. Because current fiow is initiated when the first spot is turned on, the lips gated by the regenerative gate are turned on at a voltage which is reduced from that initially across the device. Thus, the maximum temperature reached at any one spot in the device can be controlled and at the same time reduced. It is important to note that after turn-on is established, the various lip resistances are automatically transferred out of the circuit, thus further reducing losses. Multiple spot turn-on reduces the current density in the device, and the time to full area conduction is greatly reduced.

Thyristors with as many as ten regenerative gates have been made and measurements show that all can be made to fire at nearly the same instant. Good load sharing is achieved by utilizing the emitter lateral resistance at each additional spot that is turned on. This reduces the instantaneous temperature and the impedance of the device in proportion to the number of spots turned on.

Ten spot regenerative gate ampere 1000 volt rated devices of the type shown in FIGURE 3 have been built. The devices were operated at 500 amperes per microsecond at 1000 volts, 2500 amperes peak, and C. case temperature with a repetition rate of 60 Hz.

It was found that the devices are completely insensitive to the quality of the initiating gate signal and are safe in high di/dt circuits with a wide variety of gate signals. The regenerative gate signal is a function of the external load circuit and the particular device geometry but can be an extremely fast pulse rising to volts in a fraction of a microsecond. Thus, as the circuit di/dt increases, the regenerative gate signal increases which provides ideal device protection. The power delivered by the regenerative gates, which can be thousands of watts for a very brief time, is a redistribution of an internal device loss for a useful purpose. In prior devices, this power was dissipated without serving any useful purpose.

The regenerative gates will introduce much less stress where they turn-on remote areas, since these areas are comparatively distant from the current initiation hot spot and because they turn on under reduced anode voltage. The total switching losses in the device are much lower than in the conventional device.

When a conventional thyristor is turned on by exceeding the forward blocking voltage, it must operate at a much lower di/dt to prevent failure than if the device is turned on by a powerful gate signal. The regenerative device just described also has a much lower di/dt ability when switched on in the forward direction with excess voltage. Typically, turn-on by excess voltage will be at a random location caused by some defect in the structure. However, if breakdown can be forced to occur first on the regenerative source emitter lip, the device will generate its own regenerative gate signal and be able to withstand high di/dt.

FIGURES 5, 6 and 7 show three ways to force a device to fire at a desired location by building a weakness into the device at that spot.

In FIGURE 5, the device defines a weakened area comprising a dip 74 formed at the end of the emitter lip 20. The dip is provided by forming a depression in the crystal prior to difi'using the emitter to thereby shorten the distance between the emitter n region and the base 11 region. This arrangement provides higher turn-on gain and will insure that breakdown will occur first on the emitter lip so that a regenerative gate signal will become available at the point 32. In the arrangement shown in FIGURE 6, a depression is formed before diffusion of the original silicon crystal whereby this same weakened condition is carried on to the adjacent p-n junction as shown at 76.

FIGURE 7 illustrates an alternative manner in which a weakness can be built into a thyristor. In this instance, the base n portion is tapered so that the thin section 78 is located in the area beneath the regenerative gate takeoff 30.

The dip or weakened emitter arrangement can be provided by any means which will insure that device turnon gain is greatest along a line extending approximately from the regenerative gate source and parallel with the thyristor axis. The weakness could be formed in the anode emitter and still provide the desired result. Selective diffusion of impurities in the area of the regenerative gate source provides one alternative means for achieving the result.

As indicated, the devices of FIGURES 5, 6 and 7 can be broken over safely by excess voltage from the forward locking condition with no loss in di/dt ability.

Interdigitation of conventional power thyristors to obtain high speed turn'on has not worked because of basic limitations, specifically, current concentration at the initiating point, and reduced gate sensitivity. Using the regenerative gate concept, these limitations are overcome.

FIGURE 8 illustrates one form of an interdigitated thyristor. The device includes emitter contact 80 defining three fingers 82. The periphery of the emitter extends beyond the cathode contact fingers as shown at 84. The regenerative tab or emitter lip is shown at 86. The initiating gate is connected at 88.

The element comprises the regenerative gate contact. A central portion 92 of this contact is connected to the emitter lip while the four fingers 94 extend in opposed relationship relative to the periphery 84 of the emitter. In the operation of this device, the additional initiation provided by the fingers 94 in response to the powerful gate regenerative signal present, starts additional conduction spots and forces spreading to occur at a speed much faster than if the signal were not present.

Devices of the type shown in FIGURE 9 are also characterized by the interdigitated concept. In this device, the triggering signal is provided through gate 100, and the regenerative gate signal is supplied through emitter 102 which is connected through line 104 to cathode 106. When the initiating gate signal is applied, the regenera tion of positive bias to line 108 results in application of the signal to the interdigitated gate 110 resulting in turnon of the emitters 112. In the design shown in FIGURE 9, the area of the emitter 102 in the line of current flow from cathode contact 104 through the device corresponds in function with the resistor 52 described in FIGURE 10. Thus, the regenerative gate operating principle turns on the interdigitated device and overcomes the limitation of reduced gate sensitivity.

Test results show the additional areas 112 can be turned on quickly and that fast switching occurs. Current rises to 1000 amperes in 8 10 microsecond have been obtained with this structure. Applications requiring high current devices, and megacycle switching speeds are contemplated.

FIGURES 12 and 13 illustrate two additional configurations embodying the principles of this invention. The construction of FIGURE 12 is similar to that of FIGURE 3 in that the thyristor comprises a cathode contact 16, gate 18 and emitter lips 44 and 50. In this design, however, the take-off 132 from the emitter lip 44 comprises a conductive material diffused into the surface of the device. This material extends in the form of a ring 134 all around the device. The regenerated signal will fire the device at a plurality of emitter lips 50.

In FIGURE 13, the device comprises a cathode contact 142 which defines a cut out portion 144 having an emitter lip 146 located therein. The main emitter body is positioned entirely beneath the contact with the exception of this lip and two additional projections 148. A gate is provided for triggering the device.

Because of the proximity of the projections 148 to the emitter lip, regenerative firing will occur at the projections. Transfer and multiple turn-on can thus be effected without the need for a wire lead or other conductor extending from the emitter lip. It will be understood that one or more projections of the type shown at 148 may be employed in lieu of the specific configuration shown.

The dip emitter arrangements of FIGURES 57 can be utilized in conjunction with the structures described in FIGURES 2, 3, 4, 8, 9, 1.2 and 13. Thus, in each instance, a weakness can be formed beneath the emitter lip to provide a point of preferential firing.

The principle of the regenerative device can be used to advantage when connecting conventional thyristors in parallel as shown in FIGURE 11. In this case, thyristor 117 is provided with initiating gate 118 and gate return lead 123, and a resistor 119 is connected between this thyristor and cathode 121. The signal generated by the gated thyristor is fed back through line 120 to a multiplicity of thyristors 122 which is analogous to the original regenerative gate concept. Arrays have been made which demonstrate that this is a useful technique to employ particularly when working with short pulses and high repetition rates. This method has the advantage of allowing higher circuit di/dt and heat removal in proportion to the number of devices used. Good load sharing is achieved with parallel devices because the very powerful regenerative signal minimizes turn-on variations of devices. An array of six 35 ampere 1000 volt conventional side fire gate devices with a 3 ohm resistor at 119 was built to verify the utility of this approach. Rise time of A microsecond to 1000 amperes are readily obtained with this array.

In the arrangement of FIGURE 11, the current through thyristor 117 is rapidly reduced as soon as a signal is produced which provides for turning on of thyristors 122 through line 120. The element 119 may be any impedance.

The circuit of FIGURE 11 may also be built to provide advantages analogous to the dip emitters described herein, Specifically, the thyristor 117 may be selected to have a lower breakover voltage than any of the thyristors 122. In such event, application of excess voltage will preferentially turn on thyristor 117. This will create a signal which will fire each of the other thyristors in the normal manner so that they will not burn out. In addition, the thyristor 117 is safe from burnout since its current is rapidly reduced when thyristors 122 are turned on.

A parallel array of devices including at least one regenerative device provides advantages since the regenerative voltage signal can be brought out of the initially triggered device and used to trigger the other devices in the array. Where a regenerative device is the one triggered in such an array, the resistor 119 is not necessary. For example, if a device such as shown in FIGURE 2 is employed as the initially triggered device, the emitter lip of this device will serve as a take-off point with the lead 20 being connected to the gates of the other devices in the array.

In a parallel arrangement such as shown in FIGURE 11, the end thyristor 117 may advantageously be a thyristor of the dip emitter type, for example as shown in FIG- URES 5, 6 and 7. The thyristor 117 must also be characterized by the lowest ibreakover voltage relative to the other thyristors in the assembly in which case the application of excess voltage to the assembly will trigger the device 117 thereby providing a signal which will fire each of the other thyristors. This arrangement thus substantially reduces the possibility that the devices will breakover destructively.

It is significant to note that the instant invention utilizes an available signal which is extremely valuable for regeneration purposes. This signal is, however, also available for other purposes since it may be used as a timing signal, as a source to excite oscillator circuits or for a variety of other applications. Thyristors constructed in accordance with FIGURE 2 may provide an exposed terminal connected to the lead so that use of this signal can be readily accomplished.

FIGURE 1 illustrates a further improvement related to the use of a signal which becomes available when conduction is initiated through a device. Specifically, the lead 26 can be attached to a conventional thyristor as shown and the terminal of the lead will then be available for utilization of the signal generated.

When conventional thyristors are series connected into high di/dt circuits, it is necessary to not only balance the array with condenser resistor networks to maintain uniform voltage division but to ensure that all devices are gated simultaneously with powerful and complex gate supplies. It is also necessary to select devices for identical turn-on characteristics so that they all turn on at the same instant in time. Failure to achieve exact synchronism during the instant of turn-on causes high voltages to appear across the last device to turn on. The result is that excess voltage triggering will occur which will cause catastrophic failure of the last device to fire. In fact, series operation at high di/dt is such a demanding application that it is an inappropriate application for conventional thyristors unless protected by suitable reactors.

Using devices which have the dip emitter design, for example, as shown in FIGURES 5, 6 and 7, results in a series combination in which it is no longer necessary to balance the turn-on characteristics since either gate or voltage breakover will trigger them safely in high (Ii/alt 8 circuits. Another advantage is that only enough gates need to be fired to break down the remainder of the series string. In fact, this series string does not require gating to initiate breakdown and can be triggered by applying an excess voltage to it. As shown in FIGURE 14, a pair of devices 70 may utilize a single triggering lead 71 and still provide a completely safe arrangement.

It should also be noted that the invention is not limited to devices which have a gate physically attached. Other triggering means, for example, as in light activated devices, are contemplated.

There is one fundamental characteristic of all thyristor designs falling within the scope of this invention. In all instances, a voltage generated when a device first turns on is applied to a separate emitter portion so that at least two firing points are provided. In the preferred forms of the invention, there is a base region (a p region in the case of cathode gated devices) interposed between the take-off on the emitter lip and the additional firing points on the emitter.

Where additional emitter portions are referred to in the claims, it will be understood that these are portions of the emitter from which breakdown will initiate upon application of the firing signal. These additional emitter portions may extend physically beyond the associated electrode, however, emitter portions underneath the electrode may also be fired if the electrode does not completely short out the firing signal, and such emitter portions may be considered additional emitter portions in accordance with this invention.

In the case of construction as shown in FIGURE 3. the p region extends between the contacts points 48 and the adjacetn emitter edge. In the case of the construction shown in FIGURE 13, the p region extends from the sides of the emitter lip to the firing points on the projections 148. Where the claims refer to a base region extending between the take-01f and the additional emitter portion. it will be understood that this refers to the presence of some base region in the electrical path between the takeoif and additional emitter portion. Thus, other material or an opening may be interposed in this area but this will not be totally disruptive and may even be advantageous unless there is complete insulation betwen the takeoff and the additional emitter potrion.

Where a regenerative gate signal is applied to an emitter lip, voltage is built-up on this lip when the lip starts to turn on. As shown in dotted lines in FIGURE 3, this voltage built up on a lip 50 can be utilized for firing at additional points instead of firing all points from the originally gated lip. This arrangement is especially contemplated for use where a large number of points on the emitter are to be fired. In this way, a regenerative gate is used to fire a regenerative gate in a cascade arrangement. The cascading arrangement may involve several steps rather than just two where a large number of firing points is involved.

Another embodiment of this invention consists of using an emitter area that is isolated from the main emitter area but that is connected to the main emitter area by a resistive element. In such a device, the gate signal would be used to trigger the isolated emitter area. When this area turns on, the current through the resistive element will create a voltage across the resistive element. A takeoff lead, such as previously described, can be connected to the isolated emitter area or to a portion of the resistive element. This take-off lead can be used to trigger the main emitter by being positioned in close proximity to an adjacent emitter portion or portions. The aforementioned resistive element can be formed as part of the device or can be made external to the active area of the device. In the latter case, the element may be either inside or outside the device housing 21. In addition to triggering the main emitter, the take-oft lead can be used for generating a voltage signal that can be taken out of the device and used for other purposes.

FIGURE 2 serves to illustrate this alternative form of the invention. Thus, the resistance illustrated may be formed by resistive element rather than being emitter-lip resistance. The end portion of the emitter lip will be the only actual emitter portion, and this end portion will be isolated from the main emitter.

In a device constructed along the lines of FIGURE 2, the resistive element forms a part of the active area of the device. This arrangement can be accomplished by difiusing or otherwise providing a resistive material on the face of the device between the isolated emitter area of the main emitter area. An external connection can be provided by attaching leads to the isolated emitter area, and to the main emitter area or contact 16 with the resistive element attached between the leads. In this case, the resistive element may be located within the device housing, or the leads may extend outside the housing for attachment of the resistive element.

Conventional practices employed in manufacturing thyristors can be employed in producing devices in accordance with this invention. Conductors other than wire leads can be formed by diffusion, epitaxial growth, vapor deposition, plating techniques, printing techniques, alloying techniques, or other means. Resistance values can be altered to achieve desired values by employing acid etching techniques. The depressions provided in the case of FIGURES 5 and 6 can be formed by chemical or physical removal of material.

In discussing gated devices, reference has been made to cathode gated devices which are the more common type. The invention is applicable, however, to anode gated devices, and to devices which may have gates at both the anode and cathode for alternative use. The invention is also applicable to non-gated devices, for example, devices switched by over voltage.

It will be appreciated that the structural characteristics of thyristors designed in accordance with this invention are not limited by the illustrations presented herein. On the contrary, a wide variety of structural variations and configurations are possible while oeprating within the principles of this invention. This applies to the structures illustrated in FIGURES 2, 3, 4, 12 and 13, the weak erred structure of FIGURES 5, 6 and 7, and the inter digitated structures of FIGURES 8 and 9. In addition, the concepts of this invention are applicable to circuits beyond those specifically referred to herein.

It will be understood that various other changes and modifications may be made in the constructions described which provide the characteristics of this invention.

That which is claimed is:

1. In a thyristor construction having cathode and anode electrodes with associated emitters of the thyristor in contact with said electrodes, and a triggering gate in contact with a base region adjacent at least one of said emitters, the improvement comprising a first emitter portion free of said electrodes adjacent and contiguous to said one emitter, and intermediate said gate and said one emitter so as to be adjacent to the line of current flow therebetween, at least one additional emitter portion adjacent and contiguous said one emitter, and regenerative gate means comprising a conductor electrically connecting said first emitter portion to a point on said base region opposite said additional emitter portion.

2. A construction in accordance with claim 1 including an emitter lip extending outwardly from said emitter, said emitter lip providing said first mentioned emitter portion.

3. A construction in accordance with claim 1 including an emitter lip extending outwardly from said emitter, said emitter lip providing said additional emitter portion.

4. A construction in accordance with claim 1 wherein a plurality of emitter lips extend outwardly from said emitter around the periphery thereof, one of said emitter lips providing said first mentioned emitter portion and the remaining emitter lips providing additional emitter portions, said conductor being connected to said first mentioned emitter lip and extending into contact with base regions near the additional emitter lips.

5. A construction in accordance with claim 4 wherein the electrode in contact with said emitter extends at least to the periphery thereof whereby said emitter lips comprise the only portions of the emitter extending beyond said electrode.

6. A construction in accordance with claim 1 wherein the resistance in said line between the take-0E means and said electrode exceeds the resistance of the additional emitter portion between said junction and said electrode.

7. A construction in accordance with claim 1 wherein a plurality of emitter portions isolated from each other are formed on a face of the thyristor, said triggering gate being positioned adjacent one of said emitter portions, said conductor extending between said first mentioned emitter portion and a contact piece, said contact piece providing an additional conductive connecting point located near edge portions of each additional emitter portion, and means connecting each of said emitter portions to an electrode.

8. A construction in accordance with claim 1 including a lead secured to said first mentioned emitter portion and extending outwardly of the construction with a connecting terminal on said lead.

9. A construction in accordance with claim 1 wherein the first emitter portion is isolated from its associated electrode by a resistive element.

10. A construction in accordance wtih claim 1 includ ing an additional conductor connected to an additional emitter portion, said additional conductor extending to at least one further connecting point on the face of said thyristor, said further connecting point being located adjacent a further emitter portion with base region disposed therebetween whereby firing of said further emitter portion is adapted to take place at the junction defined between said base region and said further emitter portion.

11. A construction in accordance with claim 10 wherein a plurality of emitter lips extend outwardly from said emitter around the periphery thereof, one of said emitter lips providing said first mentioned emitter portion and the remaining emitter lips providing said additional and further emitter portions, and wherein said additional conductor is connected to one of said additional emitter lips and extends into contact with base regions near further emitter lips.

12. A construction in accordance with claim 1 wherein said emitter defines a plurality of fingers, one portion of said conductor engaging one of said fingers, and wherein extensions from said one portion of said conductor comprise a plurality of fingers interdigitated with the fingers of said emitter.

13. An assembly of thyristor constructions of the type described in claim 1, said constructions being connected in parallel with an initiating gate connected to a first construction, and a conductor connected to the take-01f means of said first construction and extending to the triggering gate position of each of the other thyristor constructions to provide for firing of said other constructions.

14. A construction in accordance with claim 1 includ ing means formed in the area of said first emitter portion along said line of current flow, said means providing a weakness which reduces the ability of the device to resist voltage break-over at the point of the weakness when compared with other points of the device whereby said thyristor will preferentially fire in the area of the weakness upon application of excess voltage.

15. An assembly of thyristor constructions of the type described in claim 1 wherein the constructions are connected in series.

16. In an assembly of thyristor constructions, each defining cathode and anode electrodes and a triggering gate, the respective electrodes of the thyristors being connected by means of conductors whereby the thyristors are in parallel, the improvement wherein the thyristor at one end of the assembly comprises a thyristor of the type defined in claim 14, an additional conductor connected to said first emitter portion of said end thyristor, said additional conductor extending to a triggering gate position for the other thyristors in said assembly, and wherein said end thyristor is characterized by the lowest breakover voltage of all the thyristors in the assembly whereby application of excess voltage to the assembly will create a signal which will fire each of the other thyristors through said additional conductor.

17. In a thyristor construction having cathode and anode electrodes with associated emitters of the thyristor in Contact with said electrodes, and means in association with a base region adjacent at least one of said emitters for triggering the construction, the improvement comprising a first emitter portion free of said electrodes adjacent and contiguous to said one emitter adjacent to the line of current flow between said triggering means and said one emitter, at least one additional emitter portion contiguous to said one emitter and positioned closely adjacent said line of current flow, and a section of said base region separating said additional emitter portion from said first emitter portion, the proximity between said first emitter portion and said additional emitter portion providing regenerative gate means whereby firing of said additional emitter portion takes place at the junction defined between said base region and the additional emitter portion substantially instantaneously upon triggering of the construction.

18. A construction in accordance with claim 17 wherein said triggering means comprises a triggering gate in contact with said base region.

19. In a thyristor device having at least four layers of alternating conductivity types including main electrodes connected to the end layers and means for triggering the device into conduction, the improvement comprising a first portion of a predetermined one of said end layers being free from contact with said main electrode whereby a voltage is developed in said first portion, regenerative gate means comprising a conductor connecting said first portion to at least one point on the intermediate base layer adjacent said predetermined end layer and opposite a second portion of said predetermined end layer such that a region of the base layer extends between said point and said second portion whereby firing of said additional emitter portion is adapted to take place at the junction defined between said base region and said second portion.

References Cited UNITED STATES PATENTS 2,993,154 7/1961 Goldey et al. 317235 3,256,470 6/1966 Gerlach 317-235 3,300,694 1/1967 Stehney et al. 317-235 3,403,309 9/1968 Longini 317235 OTHER REFERENCES Proc. of IEEE, Behavior of Thyristors Under Transient Conditions, by Samos et al., August 1967, pp. 1306-1311.

IEEE Trans. on Electron Devices, Probed Determination of Turn-On Spread of Large Area Thyristors," by Dodson et al., May 1966, pp. 478-484.

JERRY D. CRAIG, Primary Examiner US. Cl. X.R. 307252, 305

UNITED STATES PATENT OFFICE CERTIFICATE OF CORRECTION Patent No. 3,486 ,088 December 23, 1969 Donald I. Gray et a1.

It is certified that error appears in the above identified patent and that said Letters Patent are hereby corrected as shown below:

Column 10 line 71 claim reference numeral "1" should read 14 Signed and sealed this 3rd day of November 1970.

(SEAL) Attest:

Edward M. Fletcher, Jr.

Attesting Officer Commissioner of Patents WILLIAM E. SCHUYLER, JR.

Images