Active solid-state devices (e.g. – transistors – solid-state diode – Regenerative type switching device
Reexamination Certificate
2002-07-25
2004-11-16
Nelms, David (Department: 2818)
Active solid-state devices (e.g., transistors, solid-state diode
Regenerative type switching device
C257S115000, C257S123000, C257S601000
Reexamination Certificate
active
06818927
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to the manufacturing in monolithic form of medium power bidirectional switches.
2. Discussion of the Related Art
The most current static bidirectional switches are triacs. A triac corresponds to the antiparallel connection of two thyristors. It can thus be directly connected in an A.C. network, for example, the mains. The gate of a conventional triac corresponds to the cathode gate of one at least of the two thyristors forming it and is referenced to the electrode located on the front surface side of this triac, that is, the surface including the gate terminal. As a result, the other surface or rear surface of the triac, which is currently connected to a radiator is at the high voltage, which poses isolation problems.
Bidirectional switches of the type described in U.S. Pat. No. 6,034,381 (B3073), the triggering of which is ensured by application of a voltage between a control electrode located on the front surface of the component and a main electrode located on the opposite surface of the component will more specifically be considered hereafter. U.S. Pat. No. 6,034,381 is hereby incorporated by reference.
FIG. 1
shows an equivalent electric diagram of such a bidirectional switch. A control electrode G of the bidirectional switch is connected to the emitter of a bipolar transistor T having its collector connected to the anode gates of first and second thyristors Th
1
and Th
2
arranged in antiparallel between two terminals A
1
and A
2
. Terminal A
1
corresponds to the anode of thyristor Th
1
and to the cathode of thyristor Th
2
. Terminal A
1
is also connected to the base of transistor T. Terminal A
2
corresponds to the anode of thyristor Th
2
and to the cathode of thyristor Th
1
.
FIG. 2
is a simplified cross-section view of an example of monolithic forming of the bidirectional switch described in relation with FIG.
1
. Transistor T is formed in the left-hand portion of the drawing, thyristor Th
1
is formed at the center, and thyristor Th
2
is formed to the right.
The structure of
FIG. 2
is formed from a lightly-doped N-type semiconductor substrate
1
. The anode of thyristor Th
1
corresponds to a P-type layer
2
, which is formed on the rear surface side of substrate
1
. Its cathode corresponds to an N-type region
3
formed on the front surface side in a P-type well
4
. The anode of thyristor Th
2
corresponds to a P-type well
5
formed on the front surface side and its cathode corresponds to an N-type region
6
formed on the rear surface side in layer
2
. The structure periphery is formed of a heavily-doped P-type region
7
extending from the front surface to P-type layer
2
. Conventionally, region
7
is obtained by drive-in from the two substrate surfaces. The rear surface is coated with a metallization M
1
corresponding to first terminal A
1
of the bidirectional switch. The upper surfaces of regions
3
and
5
are coated with a second metallization M
2
corresponding to the second terminal A
2
of the bidirectional switch. An N-type region
8
is formed, on the front surface side, in a P-type well
9
in contact with peripheral region
7
. The surface of region
8
forms one piece with a metallization M
3
connected to control terminal G of the bidirectional switch. A metallization M
4
may be formed on the upper surface of peripheral region
7
. Metallization M
4
is not connected to an external terminal. As an alternative, well
9
may be separated from peripheral region
7
and electrically connected thereto via metallization M
4
.
The operation of the bidirectional switch is the following.
When terminal A
2
is negative with respect to terminal A
1
, thyristor Th
1
is likely to be on. If a sufficiently negative voltage with respect to metallization M
1
is applied on gate G, the base-emitter junction of transistor T is forward biased and this transistor turns on. A vertical current i
c
shown in dotted lines in
FIG. 2
thus flows from metallization M
1
, through the forward junction between layer
2
and substrate
1
, then into regions
1
,
9
, and
8
corresponding to transistor T. There thus is a generation of carriers at the level of the junction between substrate
1
and well
9
close to the junction between substrate
1
and well
4
, and thyristor Th
1
is turned on. It can also be considered that the triggering of an auxiliary vertical NPNP thyristor including regions
8
-
9
-
1
-
2
, region
9
of which forms the cathode gate region, has been caused.
Similarly, in the case where terminal A
2
is positive with respect to terminal A
1
, the applying of a negative voltage on terminal G turns transistor T on. The carriers present in the vicinity of the junction between substrate
1
and layer
2
turn thyristor Th
2
on, as will be better understood by referring to the simplified top view of
FIG. 4
in which it can be seen that the region corresponding to transistor T is close to a portion of each of thyristors Th
1
and Th
2
.
Practice reveals that this type of bidirectional switch has a non-optimal control sensitivity, that is, especially, the current required to trigger thyristor Th
1
is of several hundreds of milliamperes.
The applicant has provided in unpublished French patent application 99/10412 (B4341) of Aug. 9, 1999, which is incorporated herein by reference, another embodiment in monolithic form of a bidirectional switch of the above-mentioned type, in which thryistor Th
1
has a greater control sensitivity.
FIG. 3
is a simplified cross-section view of an embodiment of such a monolithic bidirectional switch. The structure of the various areas formed in semiconductor substrate
1
is identical to that illustrated in FIG.
2
. The difference between the two drawings is that a region
10
having an isolation function, substantially under the above-mentioned auxiliary vertical thyristor, is provided on the rear surface side, between layer
2
and metallization M
1
. This also appears from
FIG. 4
in which the contour of region
10
is designated by a dotted line in the bottom left-hand portion of the drawing. Layer
6
, not shown in
FIG. 4
, occupies the entire lower surface except for the area located under P-type well
4
and the surface occupied by region
10
.
Region
10
is formed of a semiconductor N-type doped material, preferably silicon oxide (SiO
2
).
The operation of the bidirectional switch remains substantially similar to what has been described in relation with FIG.
2
. However, base current i
b
of transistor T, running from metallization M
1
to region
8
, is now deviated by the presence of region
10
, according to path i
b
of FIG.
3
.
The main current of the auxiliary vertical thyristor is also deviated, as shown by arrows i
c
. It can be seen that by modifying the dimensions of region
10
, the passing of current i
c
is favored in the vicinity of the areas where it is most efficient to turn thyristor Th
1
on, that is, close to the limit of well
4
.
Tests performed by the applicant have shown that the triggering current of thyristor Th
1
is minimized when region
10
extends to face P-type well
4
in which N-type region
3
forming the cathode of thyristor Th
1
is formed.
The thickness of region
10
must be sufficiently small to initially enable the starting of transistor T by the conduction of current i
b
from layer
2
to region
8
via peripheral region
7
. Indeed, if region
10
is too thick, the remaining thickness of layer
2
between region
10
and substrate
1
causes the existence of too high a resistance that opposes to the flowing of base current i
b
.
In practice, the thickness of region
10
will be smaller than that of layer
6
. Indeed, layer
6
forms the cathode of thyristor Th
2
and its thickness is determined by the characteristics, especially relating to the turn-on current, of this sole thyristor. The thickness of layer
6
will for example be on the order of 10 to 15 &mgr;m, while the thickness of region
10
will be as small as possible.
SUMMARY OF THE INVENTION
An object of the pr
Jorgenson Lisa K.
McClellan William R.
Nelms David
Nguyen Dao H.
STMicroelectronics S.A.
LandOfFree
Sensitive high bidirectional static switch does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Sensitive high bidirectional static switch, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Sensitive high bidirectional static switch will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3277126