JPH03183210A - High dielectric voltage semiconductor switch - Google Patents

Abstract

PURPOSE:To realize the reduction of a chip occupying area and the reduction of power loss by constituting the switch so that a p-type and an n-type MOS transistors(TRs) provided with main circuits connected in parallel are provided and a common drive signal is inputted to the gates of these p-type and n-type MOS TRs. CONSTITUTION:When an ON signal from a drive circuit connected to a gate terminal G becomes large, and the inter-source-drain current of an MOS-FET increases, and voltage between both ends of a resistor 4 becomes over about 0.6V, a thyristor 1 is turned on. On the other hand, when the anode potential of the thyristor is lower than ON signal potential, an n-channel MOS-FET 5 is turned on, and the current having linear voltage dependency flows through the resistor 3. When the voltage of both ends of the resistor 4 becomes over about 0.6V by these currents, since the thyristor is turned on, ON-driving can be executed surely. Thus, an ON-resistor which has less number of fundamental constituent elements, and in addition, is small in a large current area because a main output part is the thyristor can be realized within the area smaller than the MOS-FET.

Description

[Detailed Description of the Invention] [Field of Industrial Application] The present invention relates to a high-voltage semiconductor switch, and in particular, the input and output are electrically isolated, and when the output section is on, linear voltage dependence is achieved in the low current range. This article relates to high-voltage semiconductor switches with high voltage characteristics.

[Conventional technology]

Conventionally, this type of semiconductor device has been strongly requested in the communication field as a substitute for electromagnetic relays.
A capacitively coupled integrated circuit composed of T is known. (ISSCCDigest of Technica
lPapers, P, 238-239. Feb, 198
1 or U.S. Pat. No. 4,170,740).

This circuit configuration is as shown in FIG.

In this circuit, the input and output are coupled by capacitance, so they are isolated in terms of direct current.As a result, even if the output section is in a potential floating state with respect to the input open circuit section, the output section is reliably driven. becomes possible. Further, since the output section is constituted by a MOS-FET, when the output section is turned on, the current flowing through the output section exhibits linear voltage dependence. As a result, it is possible to save power on the output side compared to a semiconductor device in which the output section is composed of bipolar elements. In other words, the current flowing through a bipolar element exhibits nonlinear voltage dependence, and the nonlinearity is particularly significant in the low current range, so a constant DC bias current must be passed in order to obtain a nearly linear voltage dependence. This is because this current is not necessary in the case of a FET.

Because the circuit described above has such characteristics, the present device can be widely applied as a substitute device for electromagnetic relays that can be integrated into a single chip.

[Problem to be solved by the invention]

However, (1) since the output section is a MOS-FET, the on-resistance is large and the power loss when large current is applied is large.

(2) Since the number of circuit components is large, the chip occupies a large area when converted to RC. In particular, the input-output coupling capacitance is about 15 pF, which requires a large area. Furthermore, in order to reduce the on-resistance of the output section, it is necessary to increase the channel width and channel length of the FET, which increases the chip area.

(3) Since it is capacitively coupled, it cannot be driven unless it is an AC signal.

It had the following drawbacks.

The purpose of the present invention is to insulate the input and output to an electrical value, to be able to drive even when the output section is in a floating state, and to make the current flowing when the output section is on exhibit linear voltage dependence in a low current range. However, it is an object of the present invention to provide a high voltage semiconductor switch that has a low on-resistance in a large current range and occupies a small chip area.

[Means to solve the problem]

In order to achieve such an object, the present invention provides that the gate terminal and main terminal of an insulated gate field effect element such as a MOS-FET are electrically insulated by a gate oxide film, and that a bipolar element is replaced with a MOS-FET. N-channel MOS-
Focusing on the fact that when driving an FET and an n-channel MOS-FET, the potential of the gate terminal is opposite to the potential of the main terminal, we combined a p'n-channel MO3-FET, a bipolar element, and a resistor. be.

In other words, it has at least one junction that primarily blocks forward voltage when a forward voltage is applied and at least one junction that mainly blocks reverse voltage when a reverse voltage is applied, and these junctions are connected in series between the main terminals. an element, a resistor connected in parallel to one of the junctions for blocking the reverse voltage, and the entirety of the junction for blocking the reverse voltage and the junction for blocking the forward voltage, excluding the junction to which this resistor is connected in parallel; A p-type MOS transistor with a main circuit connected in parallel and an n
type MOS transistor, and this p-type MOS transistor
This is a high-voltage semiconductor switch configured to input a common drive signal to the gates of the transistor and the n-type MOS transistor.

[Function] In this way, the bipolar element and the gate terminal can be electrically isolated, and when the potential of the bipolar element is lower than the gate potential, it can be driven by the n-channel MOS-FET, and when it is higher than the gate potential, it can be driven by the n-channel MOS-FET. -FE
Since it can be driven by T, it can be driven reliably when the bipolar potential is in a floating state and the alligator potential is fixed. In addition, since current flows through the MOS-FET and the resistor until the bipolar element is turned on, the voltage dependence of the conducting current can be made linear, and the potential across the resistor is approximately 0, at which the forward bias junction of the bipolar element operates. Since the bipolar element turns on when the voltage exceeds .6V, the on-resistance in a large current range can be reduced despite the small area.

In addition, since the coupling is not by capacitive coupling but by the gate oxide film of the MOS-FET, it can also be driven by a DC gate bias signal.

〔Example〕

Hereinafter, the present invention will be explained based on illustrated embodiments.

FIG. 1 is a conceptual diagram of a high voltage semiconductor switch according to an embodiment of the present invention. As shown in the figure, a p'n double channel MOS-FET 3.5 is connected in parallel to a thyristor 1, which is a main switch connected between the main circuit terminal A, and a PEnBPC transistor part of this thyristor 1,
This configuration has a resistor 4 connected in parallel to the NG junction. Further, a common drive signal is inputted from the confidence terminal G to the gates of both the p- and n-channel MO8-FETs 3 and 5.

The operation of the embodiment configured as described above will be explained next. This device is driven by supplying an on signal from an opening circuit connected to the gate terminal G.

When the cathode potential of the thyristor is higher than the on-signal potential, the p-channel MO8-FET3 is turned on, so that a current flows between A- and A- via the resistor 4.

When the p-channel MO8-FET3 is turned on, the current flowing between the source and drain depends linearly on the voltage between the source and drain in a relatively low current region, so the current flowing between A- and A- also depends on the voltage between A- and linearly dependent. A in this case
The resistance between - and the resistance is approximately equal to the sum of the channel resistance of the MOS-FET and the resistance 4. The on signal increases and MO
The current between the source and drain of the S-FET increases, and the voltage across the resistor 4 increases to the build-up voltage of the thyristor 1 (approximately 0
．． 6V) or more, thyristor 1 turns on. For example, if resistor 4 is 100Ω, a current of about 6mA will flow through the MOS-
When supplied from FET3, the thyristor turns on. Generally, the on-resistance of a thyristor is equal to that of a MOS-
Since it is significantly smaller than a FET, in a large current range, most of the current flows through the thyristor side, and the on-resistance between A- and A- becomes small. For example, the on-resistance when energizing at 4 mA is as large as at least 4 or more, that is, 100 Ω or more, but at 6 mA or more, the on-resistance can easily be reduced to 10 Ω or less up to about 200 mA.

On the other hand, when the anode potential of the thyristor is lower than the on signal potential, the n-channel MOS-FET 5 is turned on and the resistor 3
A tIl current with linear voltage dependence flows through. In an intermediate state where the on signal potential is lower than the anode potential of the thyristor and higher than the cathode potential, both Pan MO5-FET
may be turned on, but a current with linear voltage dependence still flows. In any case, when the voltage of the resistor 4 becomes 0.6 V or more due to these currents, the thyristor is turned on, so that even if the potential of the thyristor is in a floating state, it can be turned on reliably.

Since the gate oxide films of MOS-FETs 3 and 5 are insulated between the thyristor and the G terminal, a high and good insulation state corresponding to the gate dielectric breakdown voltage can be achieved. -For example, if the threshold voltage of a MOS-FET is set to about 3V, its gate dielectric breakdown voltage, that is, the dielectric strength voltage between the thyristor and the G terminal, can easily be made to about 500V, and the leakage between the thyristor and the G terminal can be easily increased to about 500V. The current can also be easily reduced to 10-''A or less.

In this embodiment, since the bipolar element is a thyristor, it has a self-holding function to maintain the on state even if the on signal disappears once it is turned on.

In addition, this embodiment has fewer basic components than the conventional example, and the main output section is a thyristor, so it is possible to achieve a small on-resistance in a large current range with a smaller area than a MOS-FET, and it is I do not need a capacity that requires area.
In the disclosed example in the case of C conversion, a chip occupying area of Lonm'' or more was required to realize a 400 V, 100 mA class semiconductor device, but in this example, it can be realized with a chip occupying area of 2 m2 or less.Furthermore, capacitive coupling Dena (MOS-FE
Since they are connected through the T gate oxide film, a DC bias signal can be driven.

FIG. 2 shows a conceptual configuration diagram of a high voltage semiconductor switch according to another embodiment of the present invention. In this embodiment, an npn transistor 2 is used as a bipolar element, and the collector junction is p”n
Both channel MO8-FETs 3.5 are connected in parallel, and a resistor 4 is connected to the emitter junction. The basic operating mechanism of this embodiment is the same as that of the previous embodiment, so its explanation will be omitted. In this case, since the bipolar element is a transistor, it has the characteristic that it can be turned on and off between A- and A- depending on the presence or absence of an on signal at the G terminal.

Furthermore, although the on-resistance in a large current range is larger than that of a thyristor, it is possible to achieve an on-resistance as low as that of a MOS-FET with a smaller chip area than a MOS-FET.

As explained above, the details of the present invention and its effects have been explained based on each embodiment, but the present invention is not limited to these embodiments, and bidirectional switching can be achieved by connecting the embodiments in antiparallel. In the example shown in Fig.
It will be obvious to those skilled in the art that various modifications and applications are possible, such as connecting an n-channel MO3-FET in parallel with the npn transistor section and connecting a resistor in parallel with the Pp nn junction.

〔Effect of the invention〕

As is clear from the above explanation, according to the high voltage semiconductor switch according to the present invention, since the bipolar element is triggered by the p"n double channel MO3-FET whose gates are connected, the input and output can be electrically isolated. In addition, it has the effect of being able to drive reliably even when the output section is in a floating state.Furthermore, in the low current range, current can be passed through a MOS-FET and a resistor, so linear voltage dependence can be achieved, and in the large current range, it can be driven via a bipolar element. Since current can be passed, low on-resistance can be achieved with a small chip area, and power loss can be reduced.

[Brief explanation of drawings]

FIG. 1 is a block diagram showing one embodiment of a high voltage semiconductor switch according to the present invention, FIG. 2 is a block diagram showing another embodiment of a high voltage semiconductor switch according to the present invention, and FIG. 3 is a block diagram showing a conventional high voltage semiconductor switch. FIG. 3 is a circuit diagram of a switch. DESCRIPTION OF SYMBOLS 1... Thyristor, 2... Bipolar transistor, 3... P channel MO5-FET, 4... Resistor,
5- n-channel MOS-FET.

Claims (1)

[Claims] 1. There is at least one junction that mainly blocks forward voltage when forward voltage is applied and at least one junction that mainly blocks reverse voltage when reverse voltage is applied, and these junctions are connected in series between the main terminals. a bipolar element connected to a bipolar element, a resistor connected in parallel to one of the junctions for blocking the reverse voltage, and a junction for blocking the reverse voltage and the junction for blocking the forward voltage other than the junction to which this resistor is connected in parallel; A configuration including a p-type MOS transistor and an n-type MOS transistor having a main circuit connected in parallel to the junction, and a common drive signal is input to the gates of the p-type MOS transistor and the n-type MOS transistor. high voltage semiconductor switch. 2. In claim 1, the bipolar element is a thyristor, the resistor is connected in parallel between the p-base and n-emitter of the thyristor, or the n-base and p-emitter of the thyristor, and the p-type MOS transistor and an n-type MOS transistor connected in parallel between the base of the thyristor to which the resistor is connected and the emitter of the same conductivity type as the base. 3. In claim 1, the bipolar element is a transistor, the resistor is connected in parallel between the base of the transistor and one main electrode, and the p-type MOS transistor and the n-type MOS transistor is connected in parallel between the base of the transistor and the other main electrode.