JPS61144065A - Reverse conductive gate turn-off thyristor - Google Patents

Abstract

PURPOSE:To increase the substantial areas of a GTO section and a RCD section, and to obtain sufficient current capacity and low ON voltage by disposing a gate electrode etc. formed at sections except for an isolation region onto the isolation region. CONSTITUTION:When a gate electrode section extracting a gate lead is shaped into an isolation region, an insulating film 21 is formed so that a gate electrode in a section up to a section where the gate electrode 17 for a GTO section is brought into contact with a second base layer 13 from the end of a RCD section is not brought into contact with the second base layer 13 and an N<+> layer 20 because the section functions as the isolation region. An On gate electrode 22 is shaped onto the isolation region on an application for an amplifying gate type reverse conduction GTO. That is, the insulating film 21 is formed onto the second base layer 13 in the isolation region and the N<+> layer 20, and the ON gate electrode 22 is shaped onto the insulating film and the gate lead is extracted. A second emitter layer 24 in an auxiliary GTO is formed to the GTO section adjacent to the ON gate electrode 22, and an electrode 23 connecting the second base layer 13 and the second emitter layer 24 in the auxiliary GTO is shaped onto the layer 34.

Description

DETAILED DESCRIPTION OF THE INVENTION [Technical Field of the Invention] The present invention relates to an electrode structure of a reverse conduction GTO in which a gate turn-off thyristor (GTO) and a reverse conduction diode (RCD) are integrally formed on the same semiconductor wafer.

[Technical background of the invention and its problems] A reverse conduction GTO device is one in which a GTO and an RCD that conducts a current in the opposite direction to the current flowing therein are integrally formed, and an example thereof is as shown in FIG. It is something. GTO Department a is
p+ type first layer, transmitter layer 11, n type first base layer 1
2. Consists of a four-layer structure including a p-type second base 8113 and an n+-type second emitter @14 (7). The second emitter layer 14 is divided into a plurality of parts. The RCD part includes an anode layer 1 made of a p-type layer common to the second base layer 13 of GTOISa.
3-1 Consists of an n-type layer 12' common to the first base layer 12 and an n1-type power source layer 15. 18 is an anode electrode provided in common to the first emitter layer 11 and the N9 type power source layer 15; 16 is a cathode electrode provided to each divided second emitter layer 14; 17 is a gate electrode;
is the anode electrode of ROD. The seven-node electrode 19 and the cathode electrode 16 are electrically connected and placed at equal potential.

There is an isolation region C between the GTO section a and the RCDfiSb, in which an n+ layer 20 is formed to substantially isolate the second base layer 13 and the seven-node layer 13'. That is, this n+ layer 20 prevents short-circuiting via the anode layer 13- of the RCD when applying a negative bias between the gate electrode 17 and the cathode electrode 16 of the GTo.

Figure 13 shows that after the forward current In flows through the RCD, the GTO
These are the voltage and current waveforms when a positive voltage ■^ is applied to.

If the diode current ID flows as shown in the figure while the GTO is off, after time t1, a positive voltage is applied to the GTO again and the voltage ＾ recovers as shown by the solid line, and the GTO continues to flow.
It is required that the TO remains off. However, R
In the case shown by the broken line where the rate of decrease in the diode voltage lip of the OD is large, the GTO loses its blocking ability after time t1, resulting in erroneous firing. This is because excess carriers in the ROD act as a trigger current for the GTO.

That is, during the period when the diode current 1D is flowing, holes flow from the anode layer 13- of the RCD to the cathode layer 15, and electrons flow from the cathode layer 15 to the anode layer 13'. Then, at time t1 in FIG. 13, the voltage between the seven nodes and the cathode of the GTO becomes opposite to that before time t1, with the anode becoming positive and the cathode becoming negative. At this time, RCDI
Excess electrons existing in Sb are discharged from the cathode layer 15 of the RCD, and excess holes are discharged from the seven-node layer 13'. However, the excess carriers that have protruded to the vicinity of the n+ layer 20 formed in the isolation region C and to the GTO section a do not return to the RCD section, and the excess electrons pass through the first emitter layer 11 and are extracted from the 7-node electrode 18. , the excess holes pass through the second base layer 13, pass through the gate electrode 17 near the isolation region C, and are normally injected into the GTO d.
In order to improve the v/dt withstand capability and the forward breakdown voltage, it is discharged to the cathode electrode 16 through a resistor Ra (not shown) connected between the gate and cathode outside the device. in the end,
A displacement current accompanying voltage recovery of the GTO and a current due to discharge of excess holes flow in RGK in a superimposed manner. When the voltage drop due to the current flowing through this RGK exceeds the minimum gate trigger voltage corresponding to the built-in potential of the junction between the Wi2 base layer 1 layer and the 32 emitter layer 14, the holes are transferred from the second base layer 13 to the second emitter layer 14. 114 to the cathode electrode 16, and corresponding electrons are injected from the second emitter layer 14 into the second base layer 13. Such an operation causes the GTO to fire incorrectly. This erroneous firing is caused by the rate of decrease d of the diode current ID.
As ln/dt increases, RCDISb and isolation area C
This is likely to occur due to an increase in the amount of excess carriers remaining, especially holes, which have a lower mobility than electrons.

In order to avoid such problems, the width of isolation region C is made wide so that the influence of excess carriers of RCDISb is reduced to GTOIS.
Generally, it is done to avoid reaching a.

However, since the isolation region C becomes a completely dead space, if the width of the isolation region C is made large, the actual area of the 070 section and the ROD output will become smaller, leading to problems such as not being able to obtain sufficient current capacity and having a large on-voltage. was.

[Purpose of the invention]

In view of the above points, the present invention increases the actual area of the 070 part and the RCD part by effectively utilizing the isolation area,
Reverse conducting GTO with sufficient current capacity and low on-voltage
The purpose is to provide

[Summary of the invention]

In the present invention, gate electrodes, etc., which were conventionally formed in parts other than the isolation region, are disposed on the isolation region.
It is characterized by increasing the actual area of the 0 part and the RCD part.

(Effect of the invention) In the case of GTO, in order to improve the turn-off balance and increase the peak turn-off current, it is essential to reduce the resistance of the gate electrode and draw out the gate current efficiently. The cathode area, which is the path for the 7-node current, accounts for only 25 to 35% of the area of the
remains. Therefore, 60 to 7 of the area of 070 parts
By providing the gate electrode portion, which occupies 0% of the gate electrode area, and the portion that contacts the gate lead and press-contact type gate, which require a particularly large area, in the isolated region, the portion conventionally used as the gate electrode can be used as the cathode portion. . As a result, it is possible to increase the current capacity of the GTO and reduce the on-state voltage. Calculating this for a 100OA class conductor GTO with a diameter of 60xxφ, the cathode area increases by 20% compared to the conventional one, so if the present invention is used, a 120OA class conductor GTO can be manufactured with the same pellet size. Become.

Further, according to the present invention, since the gate electrode and the like are provided on the isolation region which requires a relatively large area, the contact with the gate lead line from the outside can be made in a significantly larger space than in the conventional GTO. Therefore, when the GTO of the present invention is housed in the envelope, the area of the contact part of the gate lead becomes larger, which has the advantage of allowing more leeway in the design of the cathode side electrode, making it easier to assemble it roughly. This is an advantage.

[Embodiments of the invention]

The present invention will be explained in detail below with reference to Examples. In each embodiment, parts corresponding to those in FIG. 12 are denoted by the same reference numerals, and detailed description thereof will be omitted.

FIG. 11 shows an embodiment in which a gate electrode portion from which a gate lead is taken out is provided in the isolation region. The area from the end of the RCD part to the place where the gate electrode 17 in the 070 part contacts the second base layer 13 becomes an isolation region, so the gate electrode in this part contacts the second base layer 13 and the n+ layer 20. The insulation 1121 is formed to prevent this from occurring.

FIG. 2 shows an embodiment applied to an amplified gate type reverse conduction GTO, in which an on-gate electrode 22 is provided on an isolated region. i.e. the second base layer 13 and the n+ layer 20 in the isolation region.
An insulating film 21 is formed on top, and an on-gate electrode 2 is formed on it.
For setting 2, the gate lead is taken out. A second emitter layer 24 of the auxiliary GTO is formed in the 070 portion adjacent to the on-gate electrode 22, and an electrode 23 connecting the second base layer 13 and the second emitter layer 24 of the auxiliary GTO is provided on the second emitter layer 24. . This electrode is preferably electrically connected to the gate electrode 17, but may be separated. In this embodiment, the off-gate is taken out directly from the gate electrode 17, but it may be taken out from an electrode formed on a monolithically formed diode (described later).

FIG. 3 shows another embodiment applied to an amplified gate type reverse conduction GTO, in which an off-gate electrode 17 is provided on an isolated region. This off-gate electrode 17 is the gate electrode itself surrounding each GTO element. The layer width gate part is 0
70 copies of other f! The second emitter layer 24 and the auxiliary gate electrode 23 are formed in four regions.

FIG. 4 shows another embodiment applied to an amplified gate type reverse conduction GTO, in which a cathode electrode 25 of a monolithically formed auxiliary diode is provided on an isolation region. In an amplified gate type GTO, in order to turn off the auxiliary GTO reliably, the gate electrode (off gate) 17 of the main GTO and the auxiliary GT
Generally, an auxiliary diode is connected between the gate electrode (on-gate) 22 of O and the cathode layer 2.
6, 00 layers are formed and realized monolithically. The off-gate electrode 25 connected to the cathode layer 26 is
It is electrically connected to the on-gate electrode 22 and exposed to the outside as one gate terminal. In the monolithically formed auxiliary diode part, an n+ layer 27 is formed on the anode side to prevent parasitic thyristor operation, and this part is selectively heavily doped with a lifetime killer to prevent parasitic transistor operation. are doing. Reverse conduction GTO
In this case, it is effective to strongly dope only the isolated region with a lifetime killer so that the excess carriers in the RCD part are recombined and disappear before reaching the 070 part. There is an advantage that doping can be done all at once by placing monolithically formed auxiliary diode sections adjacent to each other.

FIG. 5 shows another embodiment applied to an amplified gate type reverse conduction GTO, in which the cathode electrode 25 of a monolithically formed auxiliary diode is integrated with the on-gate electrode and provided on the isolation region. In this case, it is assumed that the electrode 23 provided on the auxiliary GTO and the cathode electrode 25 of the monolithically formed auxiliary diode are electrically separated by a two-layer wiring (or by planar separation). Further, the electrode 23 provided on the auxiliary GTO is electrically connected to the gate N-pole 17 in order to increase the amplification effect.

In addition, in the embodiments described so far, 070 copies and RCD
The parts may be in any positional relationship in a plane, the structure of each may be another structure such as an anode short, and each element constituting the amplification gate structure may be modified in various ways. Also, the n+ layer 2 formed in the isolation region
0 may be replaced with a groove, a dielectric material, etc., or the second base layer 13 of the GTo and the anode layer 13' of the RCD may be completely separated by selective diffusion or the like. Further, in the figure, the gate lead is taken out from the electrode provided on the isolation region, but it does not necessarily have to be taken out from this part, and other methods such as pressure contact type etc. may be used for taking out the gate lead.

FIG. 6 shows an embodiment in which a diode 29 is provided to bypass excess holes protruding from the RCD section.

In this case, an electrode 28 is provided on the isolated area for connecting a lead to the anode terminal of the diode 29. This electrode 28 makes low resistance contact with the gate layer (second base layer 13) at 070 portion. The gate electrode 17 is a separate electrode. According to this embodiment, the lead wire for connecting to the diode 29 can be taken out from the isolated area, so that the actual area of the GTO section and the R and CD sections can be increased.

FIG. 7 shows an embodiment in which a bypass diode similar to that in FIG. 6 is monolithically formed. In order to efficiently extract excess holes, an electrode 30 forming a Schottky diode is formed between the second base layer 13 and the second base layer 13 . This electrode 30 is also separated from the gate electrode 17. The Schottky electrode 30 and the cathode electrode 16 may be connected by bonding or by a two-electrode connection.

Figure 8 shows the gate upper electrode 1 of the gate turn-off thyristor.
MOS for shorting between 7 and cathode electrode 16
This is an example in which FETs are integrated. A p-type layer 31 is formed in the isolation region, and a source,
N-type layers 32 and 33 serving as a drain are formed, and a gate electrode 35 is formed between these n-type layers 32 and 33 with a gate insulating film 34 interposed therebetween. A gate electrode 17 is provided on the n-type layer 32.
is connected to the n-type layer 33, and an anode electrode 19 of the RCD section that is commonly connected to the cathode electrode 16 is connected. According to this embodiment, MO can be achieved without increasing the area.
A reverse conducting GTO is obtained which can be turned off with an SFET.

FIG. 9 shows an embodiment in which an emitter open MO8FET connected in series to a gate turn-off thyristor is integrated. The structure of the MOSFET is similar to the previous embodiment. The Zener diode 36 is for suppressing the source-drain voltage of the MOSFET. In this embodiment, as in the previous embodiment, MOSFETs can be integrated without increasing the area.

FIG. 10 shows an example in which the emitter pattern is devised to further increase the effective area of the GTO. If this pattern is used, the effective area of the GTO increases by 10% compared to the conventional radial pattern.
% effective area increase, an increase of as much as 30% can be achieved.

FIG. 11 shows an embodiment in which the RCD section is divided into a plurality of parts and gate leads are taken out from between them. Using this method, the gate terminal can be easily taken out without changing the shape of the pressure welding boss.

Since the main purpose of the present invention is to effectively utilize the isolated area of the reverse conduction GTO, the present invention is not limited to the above embodiments, and other electrodes may be formed, and other elements may be integrated. You may use this space at any time.

[Brief explanation of drawings]

Figures 1 to 11 are diagrams showing the structure of a reverse conduction GTO according to each embodiment of the present invention, Figure 12 is a diagram showing the structure of a conventional reverse conduction GTO, and Figure 13 is an explanation of the operation of the reverse conduction GTO. This is a diagram for 11... First emitter layer, 12... First base layer,
13... Second base layer, 14... Second emitter layer,
15... RCD cathode layer, 16... Cathode electrode, 17... Gate electrode, 18... Anode electrode, 1
9...RCD anode electrode, 2o...n+ layer, 21
...Insulating film, 22...On-gate electrode, 23...
Auxiliary GTO cathode electrode, 24... Auxiliary GTO second emitter layer, 25... Auxiliary diode cathode electrode, 2
6... Auxiliary diode cathode layer, 27... N'' layer for preventing parasitic thyristor operation of the auxiliary diode,
28... Diode connection electrode for bypass, 29...
Bypass diode, 30... Schottky electrode,
31...p-type layer, 32.33...n-type layer, 34.
...Gate insulating film, 35...Gate electrode, 36...
Zener diode. Applicant's representative Patent attorney Takehiko Suzue Figure 3 Figure 4 Figure 15 Figure 6 (A) Figure 8 + 〉 9

Claims (9)

[Claims] (1) In a reverse conduction gate turn-off thyristor in which a gate turn-off thyristor and a reverse conduction diode are integrally formed on the same semiconductor wafer, the cathode electrode of the gate turn-off thyristor or the reverse conduction A reverse conduction gate turn-off thyristor characterized by having an electrode other than the anode electrode of a conduction diode. (2) The reverse conducting gate turn-off thyristor according to claim 1, wherein the electrode on the isolation region is a gate electrode of the gate turn-off thyristor. (3) The reverse conducting gate turn-off thyristor according to claim 1, wherein the gate turn-off thyristor is of an amplification gate type, and the electrode on the isolation region is an on-gate electrode. (4) The reverse conducting gate turn-off thyristor according to claim 1, wherein the gate turn-off thyristor is of an amplification gate type, and the electrode on the isolation region is an off-gate electrode. (5) The gate turn-off thyristor is of an amplification gate type, and an auxiliary diode connected between the gate electrodes of the main gate turn-off thyristor and the auxiliary gate turn-off thyristor is monolithically formed, and the electrode on the isolation region is connected to the gate electrode of the main gate turn-off thyristor. The reverse conducting gate turn-off thyristor according to claim 1, which is a cathode electrode of an auxiliary diode. (6) The electrode on the isolation region is an electrode separate from the gate electrode that makes low resistance contact with the gate layer of the gate turn-off thyristor, and a bypass diode is connected between this electrode and the cathode electrode of the gate turn-off thyristor. The reverse conduction gate turn-off thyristor according to claim 1, characterized in that the reverse conduction gate turn-off thyristor is connected to (7) The electrode on the isolation region is an electrode separate from the gate electrode that forms a Schottky diode with the gate layer of the gate turn-off thyristor, and between this electrode and the cathode electrode of the gate turn-off thyristor. The reverse conduction gate turn-off thyristor according to claim 1, characterized in that the reverse conduction gate turn-off thyristor is connected to (8) The reverse conducting gate turn-off thyristor according to claim 1, wherein the electrode on the isolation region is a gate electrode of a MOS structure element provided within the isolation region. (9) The reverse conducting gate turn-off thyristor according to claim 1, wherein the gate turn-off thyristor section has a plurality of segments arranged in parallel.