JP3940743B2 - Semiconductor integrated circuit device and level conversion circuit - Google Patents

Description

  The present invention relates to a semiconductor integrated circuit device and a level conversion circuit, and particularly to a semiconductor device suitable for application to a case where a plurality of circuit units driven by a plurality of types of power supplies are provided on one semiconductor substrate. The present invention relates to a level conversion circuit used in a semiconductor device.

  In a semiconductor integrated circuit device such as a large-scale LSI, the power supply voltage for lowering power consumption is being lowered particularly in an integrated circuit using CMOS. Such a semiconductor integrated circuit device is driven by a power supply having a low voltage such as 1.2 volts, for example, but the I / O part that is an interface part with a circuit driven by an external 3.3 volt power supply is driven by a 3.3 volt power supply. Circuit.

  In addition to the I / O unit, there may be a plurality of circuit units driven by power supplies having different voltages in one semiconductor chip.

  In such a semiconductor integrated circuit device, a level conversion circuit for level down or level up is required. Here, FIG. 2 (a) shows a conventional level down circuit (a large amplitude signal is converted into a small amplitude signal). FIG. 2 shows a circuit diagram of a circuit for converting into a signal) and an operation waveform diagram thereof. Hereinafter, VDDQ is 3.3V, VDD is 1.2V, VSS is ground potential, a signal with a large amplitude is an amplitude of VDDQ potential, and a signal with a small amplitude is an amplitude of VDD potential.

  Reference numeral 200 denotes a P-type MOS transistor (hereinafter referred to as PMOS), and 201 denotes an N-type MOS transistor (hereinafter referred to as NMOS). As shown in the waveform diagram, IN0 is an input signal with 3.3V amplitude, and out0 with 1.2V amplitude is obtained as an output.

  Since there is a possibility that a potential of 3.3 V at the maximum is applied between the gate and the source, the transistors 200 and 201 are composed of thick MOS transistors.

  FIG. 2B shows a circuit diagram and an operation waveform diagram of a level-up circuit (a circuit for converting a small amplitude signal into a large amplitude signal) conventionally used. 202 and 203 are PMOS, and 204 and 205 are NMOS.

  in0 and in0b are 1.2V small amplitude input signals, and in0 and in0b are dual rail signals in a complementary relationship. OUT0 is a 3.3V large amplitude output signal. The MOS transistors 202 to 205 are MOS transistors formed of a thick oxide film similar to 200 and 201 in FIG.

Japanese Patent Laid-Open No. 06-209256

  In the conventional level-down circuit as shown in FIG. 2 (a), the logic threshold is about VDD / 2, that is, around 0.6V. In general, a large amplitude signal has a large amplitude, so that the type of noise in which the ground level fluctuates easily occurs. If the ground level of the large amplitude signal fluctuates by 0.6 V or more, the circuit of FIG. 2 is judged to be “H” level, and “L” level is output to out0.

  Therefore, the conventional level-down circuit has a drawback that the logical threshold value is lowered as VDD is lowered, and an erroneous logical value is output to out0 due to a little noise.

  In addition, in the level-up circuit having the configuration as shown in FIG. 2B, in0 and in0b become undefined values when VDDQ power is turned on and VDD power is not turned on, and a through current flows between VDDQ and VSS. There is. As a result, in a system in which the VDD power supply is made from the VDDQ power supply by a DC-DC converter, a phenomenon occurs in which a large load is applied to the VDDQ power supply and the VDD power supply cannot be turned on. If VDD power is not turned on, the previous in0 and in0b remain indefinite, and this system will not start normally forever.

  Further, not only when the power is turned on, but also when the VDDQ power is turned on, the VDD power cannot be shut off. This is because the values of in0 and in0b become indefinite values by shutting off the VDD power supply. This causes a through current to flow through VDDQ, which significantly increases the power consumption of the system.

  Also, in the input / output circuit section including the output buffer circuit section as well as the level conversion circuit section, in the conventional input / output circuit, when the VDDQ power supply is turned on and the VDD power supply is not turned on, the input of the output buffer There was a problem that the signal value became indefinite and a through current flowed between VDDQ and VSS of the output buffer circuit.

  An object of the present invention is to provide a level-down circuit that hardly generates noise even when the ground level of a large amplitude signal fluctuates, and a semiconductor integrated circuit device using the level-down circuit.

  Another object of the present invention is to provide an integrated circuit device using a plurality of high voltage and low voltage power supplies even when the high voltage power supply is turned on and the low voltage power supply is not turned on. It is an object to provide a level conversion circuit and a semiconductor integrated circuit device in which no through current flows between ground power supplies.

In order to achieve the above object, the following means are adopted in the present invention.
(1) The input of the level down circuit is a differential input.
(2) A thin oxide film MOS transistor was used for a MOS transistor in which a 3.3 V potential was not applied between the gate and drain and between the gate and source in the level down circuit.
(3) A logic operation function is provided in the level-up circuit.
(4) An element for preventing a through current from flowing through the output buffer by turning on only one of the MOSs of the output buffer is provided on the input side of the output buffer circuit.

  According to the present invention, it is possible to realize a level down circuit that hardly generates noise even when the ground level of a large amplitude signal fluctuates.

  Further, according to the present invention, in an integrated circuit device using a plurality of high voltage and low voltage power supplies, the high voltage power supply and the ground power supply can be used even when the high voltage power supply is turned on and the low voltage power supply is not turned on. It is difficult for through current to flow between them.

  The present invention and other objects and features of the present invention will become apparent from the following description with reference to the drawings.

  In the embodiment described below, the logic “1” indicates the “positive” side of the two values of the logic signal, and the logic “0” indicates the “negative” side (described in a so-called positive logic system). .

  An insulated gate field effect transistor (Insulated-Gate FET or Metal-Insulator-Semiconductor FET) typified by a MOSFET (Metal-Oxide-Semiconductor Field-Effect-Transistor) is referred to as a MOS in the following description. An N channel MOS in which majority carriers are electrons is referred to as NMOS, and a P channel MOS in which majority carriers are holes is referred to as PMOS.

  The threshold voltage (Vth) qualitatively indicates the potential difference between the gate and the source when the drain current begins to flow, but quantitatively the drain current is the square of the difference between the gate-source voltage and the threshold voltage. Plotting some so-called MOS saturation regions represented by curves yields measured values. The threshold depends on several parameters such as the concentration of the surface of the semiconductor substrate where the inversion channel is induced and the thickness of the gate insulating film. In the following embodiments, the threshold values are compared, but it should be understood that both PMOS and NMOS operate in the enhancement mode and are compared in absolute value. When the process parameters for determining the channel conductance β are the same, the threshold Vth is higher when the drain current flows more at the same gate-source voltage, assuming that the device design parameters of the channel width W and the channel length L are the same. You may think that it is low.

  Although the source and drain of the MOS are originally determined by the bias of the circuit, an arrow is attached to the source in the drawing of this application, and the PMOS is indicated inward and the NMOS is indicated outward. In addition, arrows that change the bias direction during operation, such as transmission gates, are marked with arrows. Further, when the source and drain are collectively referred to without being distinguished, they are expressed as source / drain.

  In an actual integrated circuit, a MOS that requires a large conductance is connected to a plurality of MOS gates, sources and drains in common (current paths between the sources and drains are connected in parallel) or equivalently. There are many cases in which such distribution is made, but unless otherwise noted, the present application will be described as one MOS. Similarly, even when the current paths between the source and drain of a plurality of MOSs are connected in series and substantially the same signal is applied to the gates, the present application will be described as one MOS unless otherwise noted.

  FIG. 1 shows a circuit diagram of the level down circuit of the present invention and an operation waveform diagram thereof. IN0 and IN0B are 3.3V large-amplitude input signals, and IN0 and IN0B are dual rail signals in a complementary relationship. out0 is a small amplitude output signal of 1.2V.

  Hereinafter, in the description from FIG. 1 to FIG. 13, the signal indicated by the capital letter signal name (IN, OUT) is a signal of 3.3 V amplitude, and the signal indicated by the lower case signal name (in, out) is 1.2 V. An amplitude signal is indicated.

  Reference numerals 102 and 103 denote thick oxide NMOSs similar to 201 in FIG. 100 and 101 are PMOS. Since the potential difference applied between the gate and drain of 100 and 101 and between the gate and source is at most VDD (1.2 V), the oxide film thickness withstand voltage is not as high as 102 and 103. For this reason, 100 and 101 are thin oxide film PMOSs having a smaller oxide film thickness than 102 and 103, and although not particularly limited, they are constituted by MOS transistors having threshold values lower than those of 102 and 103.

  According to the above embodiment, since the input is a differential input of IN0 and IN0B, an erroneous logic level is not output to out0 even if noise of a type whose ground level fluctuates occurs. In addition, even if the VDD voltage is lowered, it is less susceptible to noise.

  Further, according to the above embodiment, since 100 and 101 are constituted by thin oxide film MOS transistors, high speed operation is possible.

  100 and 101 are set to the same oxide film thickness and threshold voltage as the MOS transistors constituting the circuit to which the output out0 is connected, and 102 and 103 are the MOS transistors of the circuit that outputs the inputs IN0 and IN0B. If the same oxide film thickness and threshold voltage are set, the man-hours for the process can be reduced. For example, 102 and 103 may be the same as the output stage MOS transistor of the I / O circuit or the MOS transistor used in the protection circuit.

  FIG. 3 is a circuit diagram of the level-up circuit and its operation waveform diagram. In0 and in0b are small amplitude input signals with VDD (1.2V). in0 and in0b are dual rail signals in a complementary relationship. OUT0 is a 3.3V large amplitude output signal. Reference numerals 300, 301, 302, and 303 denote thick oxide film PMOSs similar to 200 in FIG. Reference numerals 304 and 305 denote thick oxide NMOSs similar to 201 in FIG. As shown in the waveform diagram in the figure, the logic level of in0 is increased in amplitude and output to OUT0. Since the input is differential, it has the feature of being strong against noise.

  FIG. 4 is a circuit diagram and an operation waveform diagram of the level-up circuit as in FIG.

The circuit in FIG. 3 is a circuit that converts a 1.2 V amplitude signal from 1.2 V (VDD) to 0 V (VSS) into a 3.3 V amplitude signal from 3.3 V (VDDQ 0 to 0 V (VSS). This circuit converts a 1.2V amplitude signal from 1.2V (VDD) to 0V (VSS) into a 3.3V amplitude signal from 1.2V (VDD) to -2.1V (VSSQ). In0 and in0b are low-amplitude input signals with VDD (1.2V) and in0 and in0b are complementary rail signals.OUT0 is a large 3.3V from 1.2V to -2.1V 2, 400, 401, 402, and 403 are thick oxide film PMOSs similar to 200 in FIG. 2A, and 404 and 405 are thick oxide film NMOSs similar to 201 in FIG. is there.
As shown in the waveform diagram in the figure, the logic level of in0 is increased in amplitude and output to OUT0.
Similar to the circuit of FIG. 3, since the input is differential, it has a feature of being resistant to noise.

  Since FIG. 3 and FIG. 4 are in a complementary relationship, the level conversion circuit of the present invention will be shown below based on the circuit of FIG. It is obvious that the following invention can also be extended in the negative voltage direction as shown in FIG.

  FIG. 5 shows the circuit of FIG. 3 that can be used even at a low VDD voltage.

  Compared with FIG. 3, a PMOS 306 used as a current source is added. When the VDDQ potential is fixed and the VDD potential decreases, the NMOS 304, 305 on-current (current when the source-gate potential difference is VDD) is the PMOS 302, 303 off-current (source-gate potential difference is VDD Current). Then, the cross-coupled PMOS composed of the PMOSs 300 and 301 does not reverse. In order to prevent this, it is necessary to reduce the gate width of the PMOS 300, 301, 302, 303 and increase the gate width of the NMOS 304, 405. However, this leads to an increase in area and increases the input capacitance of in0 and in0b. In FIG. 5, the PMOS 306 is connected to the power supply VDDQ. By doing so, it is not necessary to reduce the gate width of the PMOS 300, 301, 302, 303 and increase the gate width of the NMOS 304, 405. The area increase is only 306 minutes for PMOS, and the input capacitance of in0 and in0b does not increase.

  The PMOS 306 may be anything that can limit the current. NMOS may be used. Also, the insertion location may be inserted between the PMOS 300 and 302 and between the PMOS 301 and 303. Although the insertion position is not particularly limited, in short, not only the gate width from 300 to 305 is adjusted, but a current limiting element may be inserted in the circuit.

  FIG. 6A is a circuit in which an inverter circuit 331 is further connected to the output stage in the circuit of FIG. Since the output OUT0 in FIG. 3 is also an internal node of the level conversion circuit, the behavior of the internal node potential of the level conversion circuit changes depending on the circuit connected to OUT0. This affects the delay time of the level conversion cell, which in turn causes a malfunction. By inserting an inverter in the output stage as shown in FIG. 6 (a), the circuit connected to OUT0 does not adversely affect the node in the level conversion cell. In addition, since the output impedance of OUT0 can be reduced compared with FIG. 3, the total delay time can be reduced when many circuits are connected to OUT0.

  When registering a level conversion cell as a cell to be used by the automatic placement and routing tool, a high-speed level conversion cell having excellent noise resistance can be configured with the configuration shown in FIG. 6 (a). In addition, since the delay dependency on the output load is the same as that of the CMOS inverter, it can be applied to the timing analysis as it is.

  Fig. 6 (b) is the waveform diagram of Fig. 6 (a). By inserting the inverter 331, the slew rate of the internal node 333 is slow, whereas the slew rate of the output OUT0 is fast.

  The same effect can be obtained by adding an inverter circuit to the output of the circuit of FIG. Hereinafter, although the inverter of FIG. 6 (A) is not added to the level conversion circuit, it is obvious that it can be added.

  FIG. 7 is obtained by adding a logic operation function to the level-up circuit of FIG. in0 and in1 are small amplitude input signals having an amplitude of 1.2 V, and in0b and in1b are complementary signals thereof. OUT0 is a large amplitude output signal with 3.3V amplitude. Compared with FIG. 3, the inverter composed of 302 and 304 and the inverter composed of 303 and 305 are replaced with NOR composed of 502, 504, 506 and 508 and NAND composed of 03, 505, 507 and 509, respectively. With this configuration, logical operation of OUT0 = in0 or in1 becomes possible.

  Furthermore, if the NOR consisting of 502, 504, 506, 508 is replaced with a logic circuit that performs LOG1 operation, and the circuit complementary to that LOG1 circuit is replaced with NAND consisting of 503, 505, 507, 509, OUT0 = ~ A level-up circuit with a logical operation function of LOG1 (~ means invert) can be configured. Although FIG. 7 shows a 2-input (4 inputs including complementary signals) signal, a circuit configuration with more inputs than that can be used.

  FIG. 8 is an application of the level-up circuit of FIG. 7 to a level-up circuit with an output fixing function. Reference numeral 513 denotes a level-up circuit with an output fixing function. Compared with FIG. 7, an inverter 512 is added, in1b is changed to a large amplitude signal IN1 of 3.3V, and in1 is made from IN1 by the inverter 512.

  510 is a circuit block that operates at a power supply voltage of 1.2V, and 511 is a circuit block that operates at a power supply voltage of 3.3V. By setting IN1 = 0V, OUT0 = 3.3V is output regardless of the potentials of in0 and in0b. In this state, there is no through current flowing from VDDQ to VSS in the level-up circuit 513 with the output fixing function between the power supplies.

  By setting IN1 = 0V, the power supply of the circuit block 510 can be turned off. Although the potentials of in0 and in0b become indefinite, no through current flows through 513 and its output OUT0 is also determined, so that the circuit block 511 does not malfunction.

  When the circuit block 510 is composed of low threshold MOS transistors, a subthreshold leakage current flows even during standby when the circuit block is not operated, and power is consumed. With the configuration of FIG. 8, the power supply of the circuit block 510 can be turned off during standby, and power consumption due to the subthreshold leakage current can be suppressed.

  Further, although circuit constants such as the gate width of the MOS transistor are not described in FIG. 8, since a large amplitude signal is input to IN1, the gate length of the MOS 503, 509, 504, 508 is set to the MOS 505, What is necessary is just to make it smaller than the gate length of 507,502,506. Although the circuit constants are not described in the following level conversion diagrams as well, generally, when a CMOS circuit is composed of a MOS transistor to which a large amplitude is input and a MOS transistor to which a small amplitude is input, a large amplitude is input. Symmetry can be maintained by making the gate length of the MOS transistor smaller than the gate length of the MOS transistor to which a small amplitude is input.

  A level-up circuit 514 with an output fixing function in FIG. 9 is a level-up circuit with an output fixing function that fixes OUT0 = 0V when IN1 = 3.3V. To fix at OUT0 = 3.3V, use 513 in FIG. 8, and to fix at OUT0 = 0V, use 514 in FIG.

  FIGS. 10 and 11 respectively implement the functions of FIGS. 8 and 9 by different methods, and 515 in FIG. 10 and 516 in FIG. 11 are level-up circuits with an output fixing function. Even if the power supply of the circuit block 510 is turned off, by setting IN1 to an appropriate value, a through current flowing between the power supplies 515 to 516 can be prevented, and the output OUT0 can be stabilized.

  8 to 11 show a level-up circuit with an output fixing function that fixes the output OUT0 to a certain level. If these circuits and a latch circuit are used, the level of OUT0 is maintained when IN1 becomes a certain value. Complete the circuit. An example is shown in FIG. 513 is a level-up circuit with an output fixing function shown in FIG. 8, and 522 is a latch circuit. When IN1 changes from 3.3V to 0V, the signal level of 521 at that time is latched to OUT0. As described above, when IN1 is 0V, the power supply of the circuit block 510 can be turned off. Although the potentials of in0 and in0b become indefinite, no through current flows through 513 and its output OUT0 is also determined, so that the circuit block 511 does not malfunction.

  This embodiment is not limited to 513, and it goes without saying that the same can be achieved by using the level-up circuit with an output fixing function shown in FIGS.

  FIG. 13 shows an embodiment of the circuit system 600 using the level-up circuit with the output fixing function and the level-down circuit described above.

  Reference numeral 601 denotes a low voltage circuit block supplied with VDD = 1.2 V, which is composed of a low threshold MOS transistor. On the other hand, 602 is a high voltage circuit block to which VDDQ = 3.3 is supplied, and is constituted by a MOS transistor having a threshold higher than that of the MOS transistor constituting 601. Therefore, the subthreshold leakage current flowing between the power supplies of the circuit block 602 is negligible compared to that of the circuit block 601. Reference numerals 6031 to 603n denote level-up circuits with an output fixing function shown in FIGS. Reference numerals 6041 to 604n denote level down circuits as shown in FIG.

  Since the circuit block 601 is composed of low threshold MOS transistors, a subthreshold leakage current flows even during standby when the circuit block 601 is not operated, and power is consumed. By inputting appropriate values to IN1 of the level-up circuit group 603 during standby, the circuit block 601 can be powered off, and power consumption due to the subthreshold leakage current can be suppressed. Since the output OUT0 of the level-up circuit group 603 is also fixed, the circuit block 602 does not malfunction.

  Although the circuit function and the like mounted on the circuit block 602 are not limited, the power supply of the circuit block 601 can be frequently turned off by storing a circuit group in which the power supply such as a clock function and a memory cannot be turned off.

  A means for turning off the power supply of the circuit block 601 is not particularly limited, but a PMOS may be inserted between the circuit block 601 and the power supply VDD. If these circuit systems 600 are integrated on one chip, there is no need to provide a switch for turning off the power of the circuit block 601 outside the chip.

  In FIG. 14, the circuit block 601 is further separated into two systems of circuit blocks 601A and 601B.

  The circuit block 601 in FIG. 13 has a drawback that when the power is turned off, the potential of the node in the circuit block 601 becomes indefinite, and if the circuit block 601 has a memory circuit such as SRAM or DRAM, the information cannot be held.

  In FIG. 14, a circuit such as a memory that cannot be turned off is mounted on the circuit block 601a, and a circuit that may be turned off is mounted on the circuit block 601b. PSC is a power switch control circuit that turns on / off the power switches PMOS 702a and 702b by signals 701a and 701b from the PSC. 603a, 603b is a level-up circuit group with an output fixing function. 604a and 604b are level down circuit groups. A fixed circuit is also required between 601a and 601b to prevent the circuit block 601a from malfunctioning when the power to the circuit block 601b is turned off, but it can be easily realized by using a CMOS circuit such as NAND or NOR. It is omitted here.

  In the system configured as shown in FIG. 14, two states are provided for standby. One is a state in which the power supply of the circuit block 601b is turned off by turning off the power switch PMOS 702b (hereinafter referred to as standby 1). The other is a state of turning off the power of the circuit block 601a by turning off the power switch PMOS 702a (hereinafter referred to as standby 2) in addition to the state of standby 1. In standby 1, the subthreshold leakage current of the circuit block 601b can be reduced. The circuit block 601b does not malfunction even when the power is switched from off to on because no circuit such as a memory is mounted. Therefore, recovery from standby 1 can be performed at high speed. On the other hand, when transition is made to standby 2 in which the power of the circuit block 601a is turned off, the contents of the memory and the like in the circuit block 601a are erased, and it takes time to recover from standby 2. However, in standby 2, in addition to the state of standby 1, the subthreshold leakage current of the circuit block 601a can be further reduced, resulting in lower power. If the operation of the circuit blocks 601a and 601b is stopped for a relatively short time, it is in the standby 1 state, and if it is stopped for a long time, it is in the standby 2 state.

FIG. 15 is obtained by adding substrate bias control circuits VBCa and VBCb to FIG. As described above, a subthreshold leakage current flows in the circuit block 601a in the standby 1 state. In the invention of FIG. 14, in the standby 1 state, the substrate potential of the MOS transistor in the circuit block 601a is controlled by the substrate bias control circuit VBCa as follows.
(1) The PMOS is set to a potential higher than the power supply potential.
(2) For NMOS, use a potential lower than the power supply potential.
As a result, the threshold voltage of the MOS transistor in the circuit block 601a increases, and the subthreshold leakage current can be reduced. Since the power remains on, the contents of the memory and the like in the circuit block 601a are held.

  The substrate bias control circuit VBCb connected to the circuit block 601b can be used during the IDDQ test. Since the circuit under test is cut off from the power supply line during the IDDQ test, the power switches PMOS 702a and 702b cannot be turned off. The IDDQ test can be executed by raising the threshold voltage of the MOS transistors constituting the circuit blocks 601a and 601b using the substrate bias control circuits VBCa and VBCb and reducing the subthreshold leakage current.

  The configuration of the present invention is not particularly limited to that shown in FIG. 15. The configuration is composed of a high threshold MOS transistor, which is composed of a circuit block 1 to which a high voltage is applied and a low threshold MOS transistor, and has a low In a system comprising a circuit block 2 to which a voltage is applied, the circuit block 1 and the circuit block 2 are interfaced by a level-up circuit group with an output fixing function and a level-down circuit group. The circuit block 1 is equipped with a device that requires high-speed operation, and the circuit block 2 may be equipped with a circuit such as an RTC that does not consume much power. Further, the circuit block 1 is divided into a circuit block 1A including a circuit that takes time to recover when the power of the memory or the like is turned off, and the other circuit block 1B, and the power supply of each circuit block 1 is controlled. Further, a substrate bias control circuit may be added to each circuit block 1.

  FIG. 16 (a) shows an embodiment of a method for controlling the power switch PMOS 702a used in FIG. 14 and FIG. In FIG. 16 (a), a high threshold PMOS is used for the power switch. When active, the gate terminal potential 701a is controlled as negative as possible as long as the gate oxide breakdown voltage of the power switch PMOS 702a permits. This allows a large amount of current to flow through the PMOS. As the negative potential to be applied, for example, a negative voltage used for substrate bias control can be used. On the other hand, during standby, the gate potential 701a is controlled to the VDD potential of 1.2V. Since the power switch PMOS 702a is a high threshold MOS transistor as described above, the power switch PMOS 702a can be sufficiently turned off at this potential.

  FIG. 16B shows an embodiment of a method for controlling the power switch PMOS 702a when the power switch PMOS 702a of FIG. 16A is configured with a low threshold PMOS. When active, the gate voltage 702a of the power switch PMOS 702a is controlled to 0V. Since the power switch PMOS 702a is composed of a low-threshold MOS transistor, a large amount of current can flow. On the other hand, during standby, the gate potential 701a is controlled to be as positive as possible as long as the gate oxide breakdown voltage allows. Here, it is controlled to 3.3V. By controlling in this way, the power switch PMOS 702a is a low threshold MOS transistor, but sufficient on / off characteristics can be obtained.

  The control method in FIGS. 16 (a) and 16 (b) is not limited to the PMOS. Obviously, when using an NMOS power switch, the same thing can be done with only different polarities.

  FIG. 17 shows an embodiment of a method for generating the gate voltage 701a shown in FIG. 710 is a negative voltage generation circuit that generates -2.1V from 3.3V and outputs it to 712. Reference numeral 711 denotes a power switch control circuit for controlling the gate voltage 701a, and VDD (1.2 V) and 712 (−2.1 V) are supplied as power. Reference numeral 712 denotes a substrate bias, and the substrate potential of the MOS transistor constituting 601a is controlled by this voltage.

  Thus, by sharing the negative power supply voltage 712 used for substrate bias control and the negative power supply voltage necessary for controlling the power switch 702a, a circuit necessary for realizing the control of FIG. 16 (a). Can be greatly reduced.

  Next, an example in which an input / output circuit connected to an external terminal (pin) of an IC (semiconductor integrated circuit) using the level conversion circuit described so far will be described.

  FIG. 18 shows an example of an input / output circuit connected to an external terminal (pin) of an IC (semiconductor integrated circuit) according to the present invention.

  PB1 and NB1 are PMOS and NMOS having conductances high enough to drive the load of the external circuit to be connected to the external terminal I / O, respectively, and both constitute an output buffer circuit. The inverter INV7, NAND gate NAND1 and NOR gate NOR1 at the left end of the figure are configured such that when the output control signal / OE is “0”, the information of the output signal Out is guided to the external terminal I / O through the output buffer (output). Either one of the MOS transistors becomes conductive and the output impedance is low). When / OE is “1”, both MOSs of the output buffer are made non-conductive regardless of the state of the output signal Out. Is a circuit that performs a tri-state logic operation to make the output impedance high.

  The external terminal I / O is also connected to the input side to the NOR gate NOR2 and is used as an input / output shared terminal. When the input control signal / IE is logic “0”, the NOR gate NOR2 transmits information supplied from the outside of the IC to the external terminal I / O to the terminal / IN at the left end of the figure (the / In terminal is an external terminal) When the input control signal / IE is logic “1”, the transmission is blocked (the / In terminal is forcibly set to logic “0”).

  P3 is a PMOS for pull-up, and is used when an external input is supplied to the I / O terminal in one of two states of logic “0” and open (high impedance). P3 conducts when the pull-up control signal / PU is logic "0", and outputs a signal that becomes logic "0" when the external input is logic "0" and logic "1" when the external input is open. It is transmitted to the gate NOR2. The channel length of P3 is made larger than the channel width W so that the impedance when P3 is conductive becomes sufficiently larger than the impedance of the external input logic “0”.

  The block surrounded by the dotted line on the left side of the figure is a circuit of a low voltage power supply. Within the range shown, all PMOS N-type substrates (N-type wells) N-SUB are connected to the PMOS well power supply Vbp. All NMOS P-type substrates (P-type wells) P-SUB are connected to the NMOS well power supply Vbn. The power supply uses Vss of 0V and Vdd of 1.2V. Most MOSs have a lower threshold voltage than a high-voltage power supply circuit described later, and the gate insulating film is also formed thin. As an example, the minimum channel length is 0.2 μm, which is shorter than the minimum channel length of 0.32 μm of the high voltage power supply circuit.

  21 (c) is used as the inverter circuits INV4 to INV9, and (b) and (a) are used as the NAND circuit NAND1 and the NOR circuit NOR1, respectively.

  The block surrounded by the dotted line on the right side of the figure is a high voltage power supply circuit, and Vssq of 0V and Vddq of 3.3V are used for the power supply. Within the range shown, all PMOS N-type substrates (N-type wells) N-SUB are connected to the power supply Vddq, and all NMOS P-type substrates (P-type wells) P-SUB are connected to the power supply Vssq. All the MOSs are set to a high threshold voltage, and the gate insulating film is also formed thick. The power sources Vss and Vssq are connected to the outside of the IC, for example, by a printed wiring board on which the IC is mounted, but the external terminals (pins), bonding pads, and IC internal wiring are separated. As described above, the power sources Vss and Vssq are separated in the IC in order to prevent fluctuations in the load current from becoming operational noise on the power source wiring.

  The LSD in the low power supply voltage circuit block converts the level of the 3.3V high amplitude signal supplied via the high power supply voltage circuit into a 1.2V low amplitude signal so that it can be processed by the low power supply voltage circuit. 1 is used, and the circuit shown in FIG. 1 is used. In the MOSs 102 and 103, the gate insulating film is formed thick by the same gate oxide film forming process as the MOS used in the high power supply voltage circuit. The channel lengths of the MOSs 102 and 103 are not the minimum channel length of 0.2 μm of the low power supply voltage circuit, but the minimum channel length of 0.32 μm of the high voltage power supply circuit.

  In the low voltage block on the left side of FIG. 18, the gate insulating film is formed thin except for the above-described level down circuit LSD, and the channel length is 0.2 μm which is the minimum channel length of the low power supply voltage circuit.

  LSU1 to LSU4 in the high power supply voltage circuit are level shift circuits for leveling up the 1.2V low amplitude signal supplied from the low power supply voltage circuit to the 3.3V high amplitude signal, as shown in FIG. A circuit is used.

  INV1 and INV2 are pre-buffer circuits for driving the output buffers PB1 and NB1, and the inversion circuit shown in FIG. 21C is used. Since the output buffers PB1 and NB1 are formed with a large area so as to have a low output impedance, the input capacitance (gate) becomes large. This pre-buffer has the following roles and configurations.

  (1) The load capacity of the level shift circuits LSU1 and LSU2 can be reduced, and the setting of the design parameters of the level shift circuit is not restricted by the large input capacity of the output buffer.

  (2) The level shift circuits LSU1 and LSU2 have a larger impedance at the time of conduction of the PMOS 300 to 303 on the cross couple side than the input side NMOSs 304 and 305 so that the previous output state can be inverted by the input signals I and / I. To do. If the output buffer is directly driven by reducing the impedance on the cross-couple side, the impedance of the input MOS must be further reduced, which is not a good measure in terms of occupied area and power consumption. Therefore, the level conversion function is a level shift circuit, and the drive of the output buffer is shared with the pre-buffer. As shown in FIG. 3, when the input side is NMOS, the output impedance of each circuit when outputting logic “1” is the order of the output buffer, pre-buffer, and level shift circuit in ascending order. The output impedance of each circuit when outputting logic “0” is almost always set in the same order as described above, but considering the switching characteristics of the output buffer described later, the output buffer, level shift circuit, pre- Sometimes in buffer order. Similarly, when the input side is a PMOS as shown in FIG. 4, the output impedance of each circuit when outputting logic “0” is the order of the output buffer, pre-buffer, and level shift circuit in ascending order. In most cases, the output impedance of each circuit when outputting a logic “1” is set in the same order as described above, but considering the switching characteristics of the output buffer described later, the output buffer, level shift circuit, pre- Sometimes in buffer order.

  (3) When the output buffer changes from the previous output state to the inverted state, it is desirable to avoid both MOSs from conducting at the same time, and at least to reduce the period during which they are conducted at the same time. In other words, both MOSs want to turn off early and turn on late. Also, if the waveform of the signal output to the output terminal I / O is too steep or rising, the differential noise can easily be applied to the surrounding external pins or the wiring around the printed circuit board, so that it should be smoothed to some extent. The output impedance of the pre-buffer is set in consideration of the above points.

  The N1 and P1 MOS transistors whose drains are connected to the input side of the pre-buffer have the power supply Vddq raised when the application system is turned on, but the power supply Vdd is not yet turned on (the power-on sequence is Vddq). Therefore, it is possible to prevent both PB1 and NB1 buffer MOSs from conducting simultaneously and causing a large through current to flow because the signal from the low power supply voltage circuit is indefinite. It is provided for the purpose. P1 conducts when the gate potential of PB1 is low level “L”, and N1 conducts when the gate potential of NB1 is high level “H”. Considering the normal operation time, in the high output impedance mode in which both PB1 and NB1 are OFF, both N1 and P1 are OFF and do not affect the normal operation. In the low output impedance mode in which only one of PB1 and NB1 is turned on, the MOS with N1 and P1 turned on functions to turn off the other of PB1 and NB1 turned off, so that the normal operation is not substantially affected. In normal operation, it is prohibited to turn on both PB1 and NB1, and such an input potential (abnormal state: the gate potential of PB1 is “L” and the gate potential of NB1 is “H”) is not supplied. If the signal from the low power supply voltage circuit described above is indefinite, such an abnormal state may occur. However, when the abnormal state approaches, N1 and P1 begin to conduct, and the gate potentials of PB1 and NB1 are set in the same direction. Since it is going to move, only one of PB1 and NB1 will eventually be turned on.

  The N2 to N5 MOS circuits are circuits for further ensuring the countermeasure against through current when the power is turned on. When the power is turned on, when the outputs Q and / Q of the level shift circuit LSU1 start to rise, N3 starts to conduct and the input / I is drawn to the “L” level side, and the output Q is drawn to the “H” level side. Similarly, N2 starts to conduct and the output / Q is pulled to the “L” level side, and the output Q is pulled to the “H” level side. That is, N2 and N3 act so as to pull the output Q of the level shift circuit LSU1 to "H" level when the power is turned on. During normal operation, when the input I is at the “H” level, the output Q is at the “H” level. At that time, the N2 and N3 can move the output Q to the “H” level, that is, in the same direction. When the input I is at the “L” level, N2 and N3 are non-conductive, and N2 and N3 do not adversely affect the logic operation of the outputs Q and / Q based on the inputs I and / I.

  Since N4 and N5 operate in the same manner, the description thereof is omitted. However, since the connection to the input / output terminals of the level shift circuit LSU is opposite to that in the level shift circuit LSU1, the output Q at power-on is “ The difference is that it is drawn to the L ″ side.

  In this way, N2-N5 pulls the output Q of LSU2 to "L" and the output Q of LSU1 to "H" when power is turned on, so that both output buffers PB1 and NB1 are made non-conductive. Therefore, if N1 and P1 operate earlier when the power is turned on, only one of the output buffers PB1 and NB1 becomes conductive. If N2 to N5 operate earlier, both the output buffers PB1 and NB1 become nonconductive. Therefore, in either case, it is possible to prevent the output buffers PB1 and NB1 from becoming conductive.

  Similarly, N6 pulls the output Q of the level shift circuit LSU3 at power-on to "H" level so that the state of the input / output terminal I / O is not transmitted to the internal circuit / In.

  Similarly, N7 pulls the output Q of the level shift circuit LSU4 at power-on to "H" level so that the pull-up P3 becomes non-conductive.

  Note that one of N4 and N5 connected to LSU2 and one of N2 and N3 connected to LSU1 may be omitted in the same manner as LSU3 and 4.

  ESD1 and ESD2 are electrostatic breakdown protection circuits shown in FIG. 21 (d), and prevent the gate insulating films of the output buffers PB1 and NB1 from being destroyed when a surge voltage jumps into the input / output terminal I / O. Circuit.

  Returning to FIG. 18, the resistors R1, P2, and N8 MOS are circuits for preventing the gate insulating film of the MOS of the NOR gate NOR2 from being destroyed when the surge voltage jumps into the input / output terminal I / O. The resistors R1 and N9 are circuits for preventing the gate insulating film of the pull-up MOS of P3 from being destroyed when a surge voltage jumps into the input / output terminal I / O.

  ESD3 to ESD10 are electrostatic breakdown protection circuits shown in FIG. 21 (e), and are between different power sources Vdd and Vddq, between Vdd and Vssq, between Vss and Vddq, or between Vss and Vssq (Vss and Vssq are in the mounted state). 18 when the surge voltage jumps into the IC itself, which is connected on the printed circuit board, but is open in the handling state of the IC alone that requires countermeasures against surges, and the low voltage power load circuit on the left side of FIG. This is a circuit for preventing the MOS gate insulating films of the level shift circuits LSU1 to LSU4 from being destroyed through the high voltage power supply load circuit. The resistor R3 smooths the waveform of the surge voltage I together with the strange capacitance, or generates a voltage drop when a bypass current flows through the protective element N16 or P16, and outputs connected to the MOS gates of the level shift circuits LSU1 to LSU4 It has a function of suppressing the application of a surge voltage to the terminal O. In P16, when node I becomes more positive than power supply Vddq due to surge, the source junction (PN junction) connected to node I side of P16 is forward biased, and N connected to the junction and power supply Vddq A surge bypass path is formed between the node I and the power supply Vddq through the substrate (N well). When node I becomes more negative than the power supply Vddq due to surge, the drain junction (PN junction) connected to the node I side of P16 breaks down in the reverse direction, and N connected to the junction and power supply Vddq A surge bypass is formed between node I and power supply Vddq through the substrate (N-well) (or even through the opposite source junction). Since the gate of P16 is connected to the power supply Vddq, the electric field concentration at the drain junction is large and the breakdown voltage is low (in absolute value).

  When a surge voltage is applied between the node I and the power supply Vssq, N16 forms a bypass path between the node I and the power supply Vssq in the opposite positive / negative relationship.

  In normal operation, the drain junction of P16 and N16 on the node I side described above is not forward-biased and no reverse bias voltage exceeding the breakdown voltage is applied, and both P16 and N16 are gate and source. Are short-circuited because they are shorted, and the protection circuit does not affect normal logic operation.

  The electrostatic breakdown protection element described above is formed in the circuit block of the high voltage power supply surrounded by the right dotted line frame in FIG. 18, and the gate insulating film is formed thick so that the protection element itself is not destroyed.

  The input / output circuit shown in FIG. 18 is arranged around the bonding pads of a large number of chips and is provided as a standard circuit. Depending on each application and each product type, the input / output terminal I / O may be an input-only terminal, an output-only terminal, or an input / output terminal. Circuits that are unnecessary in each input / output circuit can be practically disabled by the method shown in FIG. C1 to C10 indicate “disconnections” for preventing a specific circuit of the high voltage power supply from operating by not providing wiring between the circuit of the low voltage power supply and the circuit of the high voltage power supply. S1 to S10 indicate that when the input path is blocked in such a manner, the input is fixed to a specific logic with a low impedance, and is connected to Vssq or Vddq by an internal wiring of Ic. For example, when the terminal I / O is dedicated to input, the wiring is cut off at C7 to C10 (no wiring pattern is provided), and the inputs I and / I of the level shift circuits LSU1 and 2 are connected to the power supply as shown in the figure. By connecting, the output buffers PB1 and NB1 are both turned off. If the input of the level shift circuit is fixed at a specific logic level, a switching operation is not caused by noise, and malfunction and waste of power can be prevented. Further, by fixing the input of the preceding circuit as much as possible, it is possible to eliminate the need for work in the subsequent circuit.

  FIG. 20 shows another embodiment of the through current prevention circuit when the output buffers PB1 and NB1 are turned on. In the figure, the same parts as those in FIG. OG is a one-shot pulse generation circuit that generates a pulse OSP for a specific period after the power source Vddq is turned on. After power-on, the MOS of N1 and P1 is turned on by this pulse OSP, the outputs of the inverters INV1 and 2 become “L” and “H”, respectively, and the subsequent output buffers PB1 and NB1 are turned off. This one-shot pulse generation circuit OG can be integrated in a compact manner if it is connected in common (through a buffer) to similar parts of other input / output circuits, and the level shift circuits LSU1-4 It can also be used to set the initial state when the power is turned on.

  FIG. 22 is a diagram showing an embodiment of the layout of the input / output circuit shown in FIG.

  As shown in FIG. 22, a plurality of I / O pads 2202 are arranged along the chip end portion 2201 in a direction parallel to the chip end portion side. Then, the circuits described in FIG. 18 are arranged in a direction perpendicular to the chip end side. The NMOS buffer 2203 and the PMOS buffer 2204 are the MOS transistors NB1 and PB1 of the output buffer of FIG. 18, and are arranged near the I / O pad as shown in the figure. Further, toward the inside of the chip, the electrostatic breakdown protection circuits ESD1 and ESD2 (2205), the pull-up circuit (2206), the prebuffer (2207), the level shift circuit (2208), and the tristate logic operation described in FIG. Circuits (2209) are sequentially arranged.

  On each of these circuits, the power supply wiring is laid out in the direction parallel to the chip end side across the adjacent circuit blocks using the third and fourth metal wiring layers. Yes. Vssq and Vddq on 2203, Vssq and Vddq on 2204, Vssq on 2205, Vddq on 2206, Vssq on 2207, Vddq on 2208, Vddq on 2209 Vss and Vdd are respectively wired.

  Next, a configuration of a protection element between power supplies that is suitable for a chip that uses a plurality of power supply voltages like the semiconductor integrated circuit device of the present invention will be described. The semiconductor integrated circuit device according to the present embodiment employs a triple well structure, and the configuration of the most efficient inter-power supply protection element in the triple well structure chip will be described below.

  A chip that uses a plurality of power supply voltages, or a chip that separates the power supplies corresponding to the magnitude of the power supply noise even with the same power supply voltage, has a plurality of types of power supply pins. In such a chip, it is effective to insert an element such as a MOS or a diode between the power source and the GND or between different types of power sources in order to easily release static electricity and improve the electrostatic withstand voltage. In this case, as a matter of course, the connection is made so that current does not flow in the forward direction in the bias state in the normal use state, and the reverse current is applied only when several hundred to several thousand volts of static electricity enters. Flow away the charge.

  Here, in the case of a chip with a triple well structure, for example, when making a diode, between a P-type substrate and an N-type element region, between an N-type element region and a P-type well, and between a P-type well and an N-type diffusion layer. There are four methods between the N-type well and the P-type diffusion layer. Among these, the method of constructing the diode having the smallest area and the least parasitic element effect differs depending on the type of power supply connected thereto.

  In the embodiments of the present invention, the most efficient configuration of such a protection element is shown below.

  FIG. 23 (a) shows the most efficient diode configuration method when a diode is connected as shown in FIG. 23 (b) in a chip in which the Si substrate is P-type and VSS is supplied thereto. This is an example.

  In FIG. 23A, 2301 is a Si substrate (P type), 2302 is an element region (N type), 2303 is an N type well, 2304 is a P type well, 2305 is an N type diffusion layer, and 2306 is a P type diffusion layer. 2307 is a diode between the P-type well and the N-type diffusion layer formed on the P substrate, 2308 is a diode between the N-type well and the P-type diffusion layer formed on the N-type element region (biased by VDDQ), Reference numeral 2308a denotes a diode between the N-type well and the P-type diffusion layer formed on the N-type element region (biased by VDD), and reference numeral 2309 denotes a diode between the P-type well and the N-type diffusion layer formed on the N-type element region. Reference numeral 2310 denotes a diode between the N-type well and the P-type diffusion layer formed on the P substrate.

  If the Si substrate is a P-type chip that feeds VSS, first, the diode connected to the VSS uses the same P-type well as the substrate, and directly without the N-type element region, the P-type substrate It is desirable to form on top. As a result, the diode has a minimum area, no parasitic element operation, and VSS power supply to the P-type substrate can also be used.

  Second, the diode connected to VDDQ is preferably formed on the N-type element region using an N-type well. As a result, this diode has a minimum area, does not operate as a parasitic element, and can also be used for VDDQ feeding to the N-type element region.

  Third, it is desirable to form the diodes other than the above two types directly on the P-type substrate using the N-type well and without the N-type element region. As a result, the diode has a minimum area, and the parasitic element operation can be eliminated.

  FIG. 24 shows another example of the protection element between power sources in the present invention.

  FIG. 24 (a) shows the most efficient MOS configuration method when MOS connection is made as shown in FIG. 24 (b) in a chip in which the Si substrate is P-type and VSS is supplied thereto. This is an example.

  In FIG. 24A, 2401 is a Si substrate (P type), 2402 is an element isolation region (N type), 2403 is an N type well, 2404 is a P type well, 2405 is an N type diffusion layer, and 2406 is a P type diffusion. 2411 is a gate, 2407 is an N channel type MOS on a P type well formed on a P substrate, 2408 is a P channel type MOS on an N type well formed on an N type element region (biased by VDDQ), 2409 Is an N channel type MOS on a P type well formed on an N type element region (biased by VDDQ), and 2410 is a P channel type MOS on an N type well formed on a P substrate.

  If the Si substrate is a P-type chip that feeds VSS, first, the N-channel MOS connected to the VSS is the same P-type well as the substrate, so there is no N-type element region and It is desirable to form on a P-type substrate. As a result, the N-channel MOS has a minimum area, no parasitic element operation, and can also be used for VSS power supply to the P-type substrate.

  Second, the N-channel MOS connected to VSSQ is a P-type well, but is preferably formed on an N-type element region biased by VDDQ. As a result, VSSQ can be supplied to the P-type well of the N-channel MOS, and it can be electrically separated from the P-type substrate to which VSS is supplied, thereby eliminating the parasitic element operation.

  Third, the N-channel MOS other than the above two types is a P-type well, but is formed on an N-type element region biased by VDD or VDDQ. As a result, VSSQ can be supplied to the P-type well of the N-channel MOS, and it can be electrically separated from the P-type substrate to which VSS is supplied, thereby eliminating the parasitic element operation.

1 is a circuit diagram and an operation waveform diagram of a level-down circuit according to an embodiment of the present invention. FIG. 2A is a circuit diagram and an operation diagram showing a conventional example of a level-down circuit, and FIG. 2B is a circuit diagram and an operation diagram showing a conventional example of a level-up circuit. 1 is a circuit diagram and an operation waveform diagram of a level-up circuit according to an embodiment of the present invention. It is the circuit diagram and operation | movement waveform diagram of the level-up circuit which are the other Examples of this invention. It is a circuit diagram of the level-up circuit which is the other Example of this invention. FIG. 6A is a circuit diagram of a level-up circuit according to another embodiment of the present invention, and FIG. 6B is an operation waveform diagram thereof. FIG. 4 is a circuit diagram of a circuit obtained by adding a logic operation function to the level-up circuit of FIG. 3. It is a figure which shows the example which applied the level conversion circuit with a logic operation function of FIG. 7 to the level-up circuit with an output fixing function. It is a figure which shows the example which applied the level conversion circuit with a logic operation function of FIG. 7 to the level-up circuit with an output fixing function. It is a figure which shows the other implementation example of the level-up circuit with an output fixing function. It is a figure which shows the other implementation example of the level-up circuit with an output fixing function. It is a figure which shows the example of the level-up circuit with an output fixing function of the type holding the output after level conversion. It is a figure which shows the system using the level conversion circuit of this invention. It is a figure which shows the system using the level conversion circuit of this invention when the circuit block comprised by the low threshold value MOS transistor of FIG. 13 is divided | segmented into two. It is a figure which shows the system using the level conversion circuit of this invention which added substrate bias control to FIG. FIG. 16A is a diagram showing an example of a method for controlling the power switch 702a of FIGS. 14 and 15, and FIG. 16B is a diagram of FIG. 14 and FIG. 15 when a low threshold MOS is used for the power switch 702a. It is a figure which shows the example of the control method of the power switch 702a. FIG. 17 is a diagram showing an example of a method for generating the gate voltage 701a shown in FIG. It is a figure which shows the example of the input-output circuit connected to the external terminal (pin) of IC (semiconductor integrated circuit) by this invention. FIG. 19 is a diagram illustrating an example in which an unnecessary circuit unit is not practically operated in the input / output circuit of FIG. 18. It is another Example which shows the through current prevention circuit at the time of power activation of the output buffers PB1 and NB1. FIG. 21A is a diagram showing an example of a specific circuit of NOR in FIG. 18, FIG. 21B is a diagram showing an example of a specific circuit of NAND in FIG. 18, and FIG. FIG. 21 (d) is a diagram showing an example of a specific circuit of ESD1 and ESD2 in FIG. 18, and FIG. 21 (e) is a diagram of ESD3 to ESD10 in FIG. It is a figure which shows the example of a specific circuit. It is a figure which shows the example of the layout of the input-output circuit of FIG. It is a figure which shows the structural example of the protection element between power supplies. It is a figure which shows the other structural example of the protection element between power supplies.

Explanation of symbols

100, 101 Thin oxide low threshold PMOS
102, 103 Thick oxide high threshold NMOS
200,300,301,302,303,400,401,402,403,306,500,501,502,503,504,505,312,314 Thick oxide high threshold PMOS
201, 304, 305, 402, 403, 404, 405, 506, 507, 508, 509, 311 Thick oxide high threshold NMOS
331, 512, 313 Inverter circuit
510, 511, 601, 602 circuit block
522 Latch
603 Level-up circuit group
604 Level down circuit group
PSC power switch control circuit
VBCa, VBCb Substrate bias control circuit
702a, 702b Power switch PMOS

Claims (6)

A first circuit block operating at a first power supply voltage;
A second circuit block operating at a second power supply voltage higher than the first power supply voltage;
A level conversion circuit for level-converting the output of the first circuit block and transmitting it to the input of the second circuit block;
The level conversion circuit includes a first means for cutting off a through current path flowing through the level conversion circuit when the supply of the first power supply voltage to the first circuit block is stopped, and an output of the level conversion circuit only contains a second means for controlling the predetermined potential,
The level conversion circuit includes a first CMOS circuit and a second CMOS circuit respectively responsive to the first output signal and the second output signal having opposite phases output from the first circuit block,
The level conversion circuit includes a first load MOSFET having a source / drain path connected between the second power supply voltage and a power supply voltage supply node of the first CMOS circuit, and a gate driven by an output signal of the second CMOS circuit; And a second load MOSFET having a source / drain path connected between the second power supply voltage and a power supply voltage supply node of the second CMOS circuit, and a gate driven by an output signal of the first CMOS circuit. A semiconductor integrated circuit device. In claim 1,
When the first power supply voltage is supplied to the first circuit block, a semiconductor integrated circuit device consumes power by flowing a subthreshold leakage current even when the first circuit block is not operated. In claim 1,
A semiconductor integrated circuit device in which a switch for stopping the supply of the first power supply voltage to the first circuit block is integrated on one chip. In claim 1,
The level conversion circuit is a semiconductor integrated circuit device having a logical operation function. A first circuit block operating at a first power supply voltage;
A second circuit block operating at a second power supply voltage higher than the first power supply voltage;
A level conversion circuit for level-converting the output of the first circuit block and transmitting it to the input of the second circuit block;
The level conversion circuit includes a first conductivity type MOSFET and a second conductivity type MOSFET in which a control signal output from the second circuit block is input to a gate,
When the supply of the first power supply voltage to the first circuit block is stopped, the first conductivity type MOSFET blocks a through current path flowing through the level conversion circuit, and the second conductivity type MOSFET The output of the level conversion circuit is controlled to a predetermined potential,
The level conversion circuit includes a first CMOS circuit and a second CMOS circuit respectively responsive to the first output signal and the second output signal having opposite phases output from the first circuit block,
The level conversion circuit includes a first load MOSFET having a source / drain path connected between the second power supply voltage and a power supply voltage supply node of the first CMOS circuit, and a gate driven by an output signal of the second CMOS circuit; And a second load MOSFET having a source / drain path connected between the second power supply voltage and a power supply voltage supply node of the second CMOS circuit, and a gate driven by an output signal of the first CMOS circuit. A semiconductor integrated circuit device. In claim 5,
A semiconductor integrated circuit device in which the first CMOS circuit and the second CMOS circuit perform complementary logical operations .

Images