JP3489912B2 - Semiconductor booster circuit - Google Patents

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION The present invention relates to, for example, an EEPROM.
(Electrically Erasable and Programable Read Only M
emory) and a semiconductor booster circuit such as a charge pump circuit used in a flash memory.

[0002]

2. Description of the Related Art In recent years, as semiconductor integrated circuits such as EEPROM and flash memory have been changed to a single 5V power source or a single 3V power source, the voltage has been boosted inside the integrated circuit. For this purpose, a semiconductor booster circuit such as a charge pump circuit is used.

FIG. 7 shows the configuration of a conventional semiconductor booster circuit.

As shown in the figure, N-channel MOS transistors Q _{120 to} Q ₁₂₄ are connected in cascade to form an n-stage booster circuit. The gate terminals of the respective transistors Q _{120 to} Q ₁₂₄ are connected to the source terminals, and the capacitances C _{120 to} C are connected to the respective source terminals N _{120 to} N _124.
The clock signal φ _A or φ _B is input via ₁₂₄ .

As shown in FIG. 8, clock signals φ _A and φ
_B is a signal having a mutually opposite phase, having a period of 1 / f and an amplitude of Vφ. The clock signals φ _A and φ _B are obtained by passing the clock signal CK through the NAND circuits ND ₁ and ND ₂ and the inverter circuits IV _{1 to} IV ₃ of FIG. 7, and the amplitude Vφ of the clock signals φ _A and φ _B is obtained. Is equal to the power supply voltage V _dd .
In addition, in FIG. 7, G is a ground terminal.

As shown in FIG. 7, in this semiconductor booster circuit, the power supply voltage V _dd is input to the transistor Q ₁₂₅ as an input signal.
Is input from the source terminal N _127, and the output voltage V _POUT is output from the output terminal N ₁₂₆ as an output signal.

The output voltage V of such a semiconductor booster circuit
_POUT is, for example, "Analysis and Modeling of On-Chip H
igh-voltage Generator Circuits for Use in EEPROM C
ircuits "(IEEE JOURNAL OF SOLID-STATE CIRCUITS, vo
(l.24, No.5, OCTOBER 1989),
It is expressed by the following formula. V _POUT = V _in −V _t + n [Vφ · C / (C + C _s ) −V _t −I _OUT / f (C + C _s )] (1) V _t = V _tO + K ₂ · [(V _bs + 2φ _f ) ^1/2 − (2φ _f ) ^1/2 ] (2) where, V _in : input voltage of booster circuit Vφ: clock amplitude voltage f: clock frequency C: coupling capacity to clock signal C _s : boosting Parasitic capacitance n at each stage of the circuit: number of stages of the booster circuit V _POUT : output voltage at the final stage of the booster circuit I _OUT : load current at the output stage V _tO : threshold voltage V _bs when there is no substrate bias : Substrate bias voltage (potential difference between source and substrate or well) φ _f : Fermi potential K ₂ : substrate bias coefficient

From the equation (1), the output voltage V _POUT is (Vφ when the load current I _OUT is 0 and C / (C + C _s ) ≈1.
It can be seen that it increases in proportion to −V _t ) and the number of stages n of the booster circuit. In the conventional booster circuit shown in FIG. 7, since the amplitude voltage Vφ of the clock is equal to the supply voltage V _dd, the number of stages n of values and the step-up circuit of the output voltage V _POUT is (V _dd -V _t)
Increases in proportion to and.

[0009]

However, in the conventional booster circuit, as the output voltage V _POUT becomes larger, the transistors Q _{120 to} Q _{124 are} caused by the substrate effect.
There is a phenomenon that the threshold voltage V _t of V becomes larger as shown in the equation (2).

[0010] Therefore, when the substrate effect constitutes a boost circuit discrete was prevented from being generated, the output voltage V _POUT is become larger in proportion to the number of stages n of the step-up circuit, the transistors Q ₁₂₀ when formed on the same substrate to Q ₁₂₄ are integrated, since the substrate effect occurs, the value of (V _dd -V _t) becomes smaller as the number of stages n of the booster circuit increases.

As a result, as shown in FIG. 9, as the number of stages n of the booster circuit increases, the output voltage V _POUT decreases from the value obtained when there is no substrate effect, and (V _dd −
When the value of V _t ) becomes 0, the output voltage V _POUT becomes saturated. This shows that the obtained output voltage V _POUT has a limit no matter how large the number n of the booster circuits is. FIG. 10 shows the relationship between the power supply voltage V _dd and the maximum output voltage when the number of stages n of the booster circuit is infinite. When the number of stages n of the booster circuit is infinite, the output voltage V _POUT obtained is theoretically infinite when there is no substrate effect, but when there is a substrate effect, it is determined by the power supply voltage V _dd . You can only get up to a certain value. That is, in the conventional booster circuit, when the power supply voltage V _dd is low, the desired output voltage V is set regardless of the number of stages n of the booster circuit.
There was a problem that I could not get _POUT .

For example, in the conventional booster circuit shown in FIG. 7, the power supply voltage V _dd is 2.5 V, and the threshold voltage V _tO when there is no substrate effect is 0.6 V (substrate bias is 0 V).
In the case of, the output voltage V _POUT of 20V could be obtained when the number n of the booster circuits was set to 20, but the number n of the booster circuits was set to 100 when the power supply voltage V _dd was 2.0V. However, only 12 V could be obtained as the output voltage V _POUT .

On the other hand, in Japanese Patent Laid-Open No. 61-254078, the threshold voltage V _t of the MOS transistor on the rear side where the substrate effect is remarkable is made lower than the threshold voltage V _t of the MOS transistor on the front side. Discloses a Cockloft-type booster circuit in which the reduction of the output voltage due to the substrate effect is improved.

However, even in this configuration, the rise in the threshold voltage V _t itself due to the substrate effect cannot be suppressed, and for example, when the power supply voltage V _dd becomes about half (V
_dd = 1 to 1.5 V), the desired output voltage V _POUT cannot be obtained no matter what value the number of stages n of the booster circuit is set to. Also, the threshold voltage V of the MOS transistor
To set a plurality of _t , for example, it is necessary to add an extra photomask and ion implantation process, which also has a drawback that the manufacturing process becomes complicated.

Therefore, an object of the present invention is to provide a semiconductor booster circuit which can obtain a desired output voltage even when the power supply voltage is low, without requiring a particularly complicated manufacturing process.

[0016]

In order to solve the above-mentioned problems, in the semiconductor booster circuit of the present invention, each stage has one end connected to the first MOS transistor and the drain terminal of the first MOS transistor. And a second capacitance whose one end is connected to the gate terminal of the first MOS transistor, wherein the first MOS transistors are connected in cascade to connect each stage. And the first MOS in each stage
The source terminal of the transistor and the base portion are electrically connected to each other, and the substrate portion is the first M of another stage.
It is electrically insulated from the base of the OS transistor,
A drain terminal and a source terminal of the first MOS transistor are connected to each other via a second MOS transistor, and a gate terminal of the second MOS transistor is connected to a source terminal of the first MOS transistor. In each stage, the gate terminal and the source terminal of the first MOS transistor are connected to each other through a third MOS transistor and a fourth MOS transistor arranged in parallel, The gate terminal of the MOS transistor is connected to the one end of the first capacitance, the gate terminal of the fourth MOS transistor is connected to the source terminal of the first MOS transistor, and the two consecutive stages are connected. A pair of first clock signals having opposite phases are input to the other end of the first capacitance, respectively. Both the second clock signal of a pair of pulse timing from each other to the other end of said second capacitance of the two-stage continuous different is characterized in that it is respectively input.

In one aspect of the present invention, the first MOS
The transistor is a P-channel MOS transistor formed in the N-type well region, and the N-type well region is electrically insulated and separated for each stage.

[0018]

[0019]

[0020]

[0021]

[0022]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The present invention will be described below with reference to embodiments with reference to FIGS.

FIG. 1 shows the configuration of a semiconductor booster circuit according to an embodiment of the present invention.

As shown in FIG. 1, n first P-channel MOS transistors Q ₁ , Q ₅ , Q ₉ , Q ₁₃ , ..., Q are provided.
₁₇ are connected in cascade, and in parallel with them, n second P-channel MOS transistors Q
₃ , Q ₇ , Q ₁₁ , Q ₁₅ , ..., Q ₁₉ are cascade-connected to form an n-stage booster circuit. The first transistor Q ₁ ,
The substrate parts of Q ₅ , Q ₉ , Q ₁₃ , ..., Q ₁₇ are electrically isolated from each other, and these substrate parts are respectively connected to the first transistors Q ₁ , Q ₅ , Q ₉ , Q ₁₃ ,. The source terminal of Q ₁₇ and the second transistors Q ₃ , Q ₇ , Q ₁₁ , Q ₁₅ ,
..., connected to the source terminal of Q ₁₉ . And the first
, Q ₁₇ , the drain terminals of the transistors Q ₁ , Q ₅ , Q ₉ , Q ₁₃ , ..., and the second transistors Q ₃ , Q ₇ , Q ₁₁ , Q
_15, ..., and the drain terminal of Q ₁₉ are each electrically connected (node _{N 1, N 3, N 5} , N 7, ..., represented by N _9.), The capacitance C to the drain terminal of the respective
_The clock signal φ _1A or φ _1B shown in FIG. 3 is input via ₁ , C ₃ , C ₅ , C ₇ , ..., C ₉ .

Further, the first transistors Q ₁ , Q ₅ , Q
₉ , Q ₁₃ , ..., Q ₁₇ gate terminals (nodes N ₂ , N ₄ ,
N _6, N _8, ..., represented by N _10. 3) are input to the clock signals φ _2A or φ _2B shown in FIG. 3 via capacitances C ₂ , C ₄ , C ₆ , C ₈ , ..., C ₁₀ , respectively.

Further, the second transistors Q ₃ , Q ₇ , Q
The gate terminals of ₁₁ , Q ₁₅ , ..., Q ₁₉ are connected to their respective source terminals.

Further, the first transistor Q₁ , Q_Five, Q
₉, Q₁₃, ..., Q₁₇Gate terminal N₂ , N_Four , N₆ , N
₈ , ..., N_TenAnd source terminal (node N₃ , N_Five , N₇ ,
N₁₁, ..., N₁₂Indicated by. ) And the third P
Channel MOS transistor Q ₂ , Q₆, Q_Ten, Q₁₄,
…, Q₁₈And a fourth P-channel MOS transistor
Q_Four, Q₈, Q₁₂, Q₁₆, ..., Q₂₀Are connected in parallel
Has been. And the third transistor Q₂ , Q₆, Q
_Ten, Q₁₄, ..., Q₁₈The gate terminal of the first transistor
Q₁ , Q_Five, Q₉, Q₁₃, ..., Q₁₇Drain terminal N
₁ , N₃ , N_Five , N₇ , ..., N₉ Respectively connected to the 4th
Transistor Q_Four, Q₈, Q₁₂, Q₁₆, ..., Q₂₀Ge of
The terminal is the first transistor Q₁ , Q_Five, Q₉,
Q₁₃, ..., Q₁₇Source terminal N₃ , N_Five , N₇ , N₁₁,
…, N₁₂Connected to each.

In the booster circuit of this embodiment, the power supply voltage V _dd is the N-channel MOS transistor Q as the input signal.
₂₂ and Q ₂₃ source terminals (indicated by a node N ₀ ) are input to the drain terminals N ₁ and N ₃ of the transistors Q ₁ , Q ₃ and Q ₅ , Q ₇ , respectively, and output as an output signal V _POUT. Are output from the output terminal (shown by the node N ₁₃ ) via the P-channel MOS transistor Q ₂₁ . As shown, the gate terminals of the transistors Q ₂₂ and Q ₂₃ are connected to the source terminal N ₀ , respectively. Further, the drain terminal of the transistor Q ₂₁ (indicated by the node N ₁₂ )
The clock signal φ _1A shown in FIG. 3 is input to the capacitor via the capacitance C ₁₁ . Furthermore, the gate terminal of the transistor Q ₂₁ is connected to the source terminal (. Represented by the node N _13).

As shown in FIG. 3, clock signals φ _1A , φ
_1B are signals having opposite phases, and clock signals φ _2A and φ _2A
2B is a pulsed signal that is turned off while the clock signals φ _1A and φ _1B are on.

Next, the operation of the semiconductor booster circuit according to this embodiment will be described with reference to FIGS.

FIG. 2 is a circuit diagram showing two consecutive stages (first stage and second stage) of the semiconductor booster circuit of FIG. Also,
FIG. 4 is a diagram of FIG. 2 in the periods (I) to (VI) shown in FIG.
3 shows voltage waveforms at nodes N _{A to} N _D of the circuit of FIG. Further, FIG. 5 is a diagram of FIG. 2 in each period (I) to (VI).
3 is a circuit diagram for explaining the conduction states of the transistors M _{1 to} M ₈ of FIG.

First, in the period (I), as shown in FIG. 3, the clock signal φ _1A changes from the ground potential 0 V to the power supply voltage V _dd , and the potential of the drain terminal N _A of the transistor M ₁ shown in FIG. , As shown in FIG. 4 (a), the power supply voltage V
_It rises by the voltage of _dd .

Further, the clock signal φ _1B changes from the power supply voltage V _dd to the ground potential 0 V, and the potential of the source terminal N _B of the transistor M ₁ becomes the power supply voltage V _{dd as} shown in FIG. 4B.
The voltage drops by.

At this time, the source terminal N of the transistor M _1
The electric charge carried from the previous stage is accumulated in the capacitance C _A2 connected to _B , and the transistor M ₁
The potential of the source terminal N _B of the above is boosted by the voltage of the charge accumulated in the capacitance C _A2 .

Further, the gate terminal N _{A of the} transistor M ₂
Potential becomes higher than the potential of the source terminal N _B , and the transistor M ₂ is turned on → off as shown in FIG. 5 (I).

The transistor M ₅ is turned off when the potential difference between the drain terminal N _A and the source terminal N _B of the transistor M ₁ is larger than the threshold voltage of the transistor M ₅ , and the source terminal N ₅ is turned on. Transistors M ₁ and M ₅ connected to the potential of _B and the source terminal N _B
The potential of the substrate portion of is held at the potential of the drain terminal N _A minus the threshold voltage of the transistor M ₅ .

Further, as shown in FIG. 4C, the potential of the gate terminal N _C of the transistor M ₁ is maintained at the potential of the source terminal N _B plus the threshold voltage of the transistor M ₆ , The transistor M ₁ remains off, as shown in FIG.

Further, as the clock signal φ _1A changes from the ground potential 0V to the power supply voltage V _dd , the potential of the source terminal N _D of the transistor M ₃ becomes, as shown in FIG. _It rises by the voltage of _dd .

At this time, the charge carried from the previous stage is accumulated in the capacitance C _A3 , and the transistor M
The source potential of the terminal N _{D 3} is boosted by the voltage of the charge stored in the capacitance C _A3.

Further, when the clock signal phi _1B becomes the power supply voltage V _dd to the ground potential 0V, the potential of the gate terminal N _B of the transistor M ₄ is lowered, the transistor M ₄ is turned OFF → ON state , the potential of the gate terminal N _E of the transistor M ₃ are, the source terminal N _D of the transistor M ₃
It becomes the same potential as. At this time, the transistor M ₃ remains off, as shown in FIG.

Further, at this time, the potential of the source terminal N _B of the transistor M ₁ is equal to that of the source terminal N _{D of the} transistor M _3.
Since the potential of the transistor M ₇ is off, the potential of the gate terminal N _{E and} the potential of the source terminal N _D of the transistor M ₃ are the same.
_{8 is} also off.

Next, in the period (II), the clock signal φ _2A changes from the power supply voltage V _dd to the ground potential 0 V, and the potential of the gate terminal N _C of the transistor M ₁ becomes as shown in FIG. 4 (c). The voltage drops by the power supply voltage V _dd .

Therefore, as shown in FIG. 5 (II), the transistor M ₁ is turned on, and the drain terminal N _A to the source terminal N _B of the transistor M _{1 is changed to} the drain terminal N _B.
The current flows until the potentials of _A and the source terminal N _B become equal.

That is, the charge is transferred from the capacitance C _A1 to the capacitance C _A2, and the potential of the drain terminal N _A of the transistor M ₁ drops as shown in FIG. As shown, the potential of the source terminal N _B of the transistor M ₁ rises.

Further, the source terminal N _{D of the} transistor M ₃
The same applies to the case of the drain terminal N _A of the transistor M ₁ , and the potential of the source terminal N _D drops as shown in FIG. 4 (d).

[0046] At this time, the clock signal phi _2A for the transistors M ₁ to the ON state is supplied from the outside via the capacitance C _B1, the drain terminal N _A and the source terminal when the transistor M ₁ in an on state Since no voltage drop occurs with N _B , the boosting capability is improved as compared with the conventional case. That is, this state corresponds to the state where V _t = 0 V in the parentheses in the above-mentioned formula (1) is satisfied, and the voltage can be boosted significantly efficiently.

Next, in the period (III), the clock signal φ _2A changes from the ground potential 0V to the power supply voltage V _dd , and the potential of the gate terminal N _C of the transistor M ₁ becomes as shown in FIG. 4 (c). The voltage rises by the power supply voltage V _dd .

Therefore, as shown in FIG. 5 (III), the transistor M ₁ is turned off.

Further, as shown in FIG. 4 (a) (b) ( d), the drain terminal N _A of the transistor M _1, a source terminal N _B, the potential of the source terminal N _D of the transistor M ₃ unchanged.

Next, in the period (IV), the clock signal φ _1A changes from the power supply voltage V _dd to the ground potential 0 V, and the potential of the drain terminal N _A of the transistor M ₁ tries to drop by the voltage of the power supply voltage V _dd. However, in the first stage, since the transistor Q _{22 of} FIG. 1 is turned on, the potential becomes (V _dd −V _t ) as shown in FIG. 4A.

[0051] Further, since the clock signal phi _1B from the ground potential 0V to the power supply voltage V _dd, the potential of the source terminal N _B of the transistor M _1, as shown in FIG. 4 (b), the power supply voltage V _dd
Voltage rises.

[0052] At this time, the capacitance C _A2 is the charge that has been carried from the previous stage is accumulated, the potential of the source terminal N _B of the transistor M ₁ is the capacitance C _A2
The voltage is boosted by the voltage of the electric charge stored in.

Further, the gate terminal N _{A of the} transistor M ₂
Potential becomes lower than the potential of the source terminal N _B , and the transistor M ₂ is turned off → on as shown in FIG. 5 (IV).

[0054] Therefore, the potential of the gate terminal N _C of the transistor M _1, as shown in FIG. 4 (c), rises to the same potential as the source terminal N _B of the transistor M _1.

Further, as the clock signal φ _1A changes from the power supply voltage V _dd to the ground potential 0 V, the potential of the source terminal N _D of the transistor M ₃ changes to the power supply voltage V d as shown in FIG. 4 (d). _It drops by the voltage of _dd .

At this time, the charge carried from the previous stage is accumulated in the capacitance C _A3 , and the source terminal N _D
The potential of is boosted by the voltage of the charge stored in the capacitance C _A3 .

Therefore, the potential of the drain terminal N _B of the transistor M ₄ becomes higher than the potential of the source terminal N _D ,
The transistor M ₄ goes from the on state to the off state as shown in FIG. 5 (IV).

As in the case of the transistor M ₅ described above, the transistor M ₇ operates when the potential difference between the drain terminal N _B and the source terminal N _D of the transistor M ₃ becomes larger than the threshold voltage of the transistor M _7. Transistor M connected from the off state to the on state and connected to the node N _D
The potentials of the substrate portions of ₃ and M ₇ are held at the potential of the drain terminal N _B of the transistor M ₃ minus the threshold voltage of the transistor M ₇ . Also, the transistor M ₃
The potential of the gate terminal N _E of the transistor M ₃ is held at the potential of the source terminal N _D of the transistor M ₃ plus the threshold voltage of the transistor M ₈ .

Next, in the period (V), the clock signal φ _2B changes from the power supply voltage V _dd to the ground potential 0 V, and the potential of the gate terminal N _E of the transistor M ₃ drops by the power supply voltage V _dd .

Therefore, as shown in FIG. 5 (V), the transistor M ₃ is turned on and the drain terminal N _B of the transistor M _{3 is changed to} the drain terminal N _D.
The current flows until the potentials of _B and the source terminal N _D become equal.

That is, the charge is transferred from the capacitance C _A2 to the capacitance C _A3 , the potential of the drain terminal N _B of the transistor M ₃ drops as shown in FIG. 4 (b), and FIG. 4 (d). As shown, the potential of the source terminal N _D of the transistor M ₃ rises.

Further, since the transistor M ₂ remains in the ON state and the gate terminal N _C of the transistor M ₁ and the drain terminal N _B of the transistor M ₃ are at the same potential, FIG.
As shown in (c), the gate terminal N of the transistor M ₁
The potential of _C drops.

[0063] At this time, the clock signal phi _2B for the transistor M ₃ and the ON state is supplied from the outside via the capacitance C _B2, the drain terminal N _B and the source terminal when the transistor M ₃ to the ON state Since no voltage drop occurs with N _D , the boosting capability is improved as compared with the conventional case.

Next, in the period (VI), the clock signal φ
_2B changes from the ground potential 0V to the power supply voltage V _dd , and the potential of the gate terminal N _E of the transistor M ₃ rises by the power supply voltage V _dd .

Therefore, as shown in FIG. 5 (VI), the transistor M ₃ is turned off.

Further, as shown in FIGS. 4 (a) to 4 (d),
The potentials of the nodes N _{A to} N _D do not change.

In the above-described operation, the source terminals of the transistors M ₁ and M ₃ are boosted in the subsequent stages, so that the substrate effect should occur and the equation (2) should be taken. Thus, the threshold voltage V _t of each of the transistors M ₁ and M ₃ tends to rise. However, in this embodiment, as shown in FIG. 2, since the substrate portion of each of the transistors M ₁ and M ₃ is connected to the source terminal,
The substrate effect does not occur, and the charge is efficiently transferred from the former stage to the latter stage.

FIG. 6 is a schematic sectional view showing the element structure of the transistors M ₁ and M ₃ of FIG.

As shown in FIG. 6, N well regions 11 isolated from each other are formed in a P-type semiconductor substrate 10, and in each N well region 11, polycrystalline silicon formed via a gate oxide film 15 is formed. Having the layer 16 as a gate electrode,
MOS having P ⁺ diffusion layer 12 as source / drain
A transistor is formed.

Source side P ⁺ diffusion layer 1 of each transistor
2 is electrically connected to the N well region 11 in which the transistor is formed via the N ⁺ diffusion layer 14, and the source of the transistor at the front stage is connected to the drain of the transistor at the rear stage.

As a result, the N well region 11 serving as the substrate portion of each transistor is fixed to the source potential of each transistor, and the substrate effect is prevented.

Further, P ⁺ on the drain side of each transistor
PN formed between the diffusion layer 12 and the N well region 11
When the junction is in the state of FIG. 5 (I) or (IV), it is forward-biased and passes through this PN junction from the N well region 11 of the substrate portion through the N ⁺ diffusion layer 14 to the node N _A →
It is possible to transfer charges of N _B and N _B → N _D. In this case, the forward junction bias voltage V of the PN junction independent of the threshold voltage V _t of the MOS transistor
The potential difference of _F (usually about 0.7 V) is used for boosting, and V _F is used instead of V _{t in} the above equations (1) and (2). Forward junction bias voltage V of this PN junction
_{Since F} is not affected by the substrate effect, it is possible to realize a booster circuit in which the boosting capability does not decrease due to the substrate effect even if the number of stages of the booster circuit increases.

However, when the booster circuit is constructed using the forward junction bias voltage V _F of the PN junction, the P ⁺ diffusion layer 12, the N well region 11 and the P type substrate 10 are formed.
It is necessary to prevent the PNP-type bipolar transistor parasitically formed between and from turning on.
In this case, the emitter terminal of the bipolar transistor is P
^{The +} diffusion layer 12, the base terminal corresponds to the N well region 11, and the collector terminal corresponds to the P type substrate 10. When the booster circuit operates, the potential of the emitter terminal becomes higher than that of the base terminal, and current is injected from the emitter terminal to the base terminal. In general, for bipolar transistors,
The larger the potential difference between the base and the emitter or the current, the easier the ON state between the emitter and the collector.
In general, the lower the impurity concentration of the base region and the narrower the base width, the more easily the emitter-collector is turned on. Therefore, when the PN junction is sufficiently turned on, the PNP-type bipolar transistor described above is turned on, current flows from the P ⁺ diffusion layer 12 to the P-type substrate 10, and the booster circuit does not operate successfully. Can be.

Therefore, in the booster circuit of this embodiment shown in FIG. 1, the second transistors Q ₃ , Q ₇ , Q ₁₁ and Q are used.
₁₅ , ..., Q ₁₉ and fourth transistors Q ₄ , Q ₈ ,
By providing Q ₁₂ , Q ₁₆ , ..., Q _{20 respectively} , FIG.
In the state of (I) or (II), the current is prevented from flowing through the PN junction.

As described above, in the semiconductor booster circuit according to this embodiment, the first MOS transistor Q of FIG.
₁ , Q ₅ , Q ₉ , Q ₁₃ , ..., Q ₁₇ are electrically isolated from each other, and the respective substrate parts are separated by source terminals N ₃ , N ₅ , N ₇ , N ₁₁ ,. _By electrically connecting to ₁₂ , the increase of the threshold voltage V _t due to the substrate effect is prevented. Therefore, it is possible to obtain the output voltage V _POUT which increases in proportion to the number of stages n of the booster circuit, and it is possible to provide a semiconductor booster circuit having a higher boosting capability than the conventional one.

Further, in the structure of this embodiment, as shown in FIG. 6, the N well region 11 in which each transistor is formed is formed separately, and the N ⁺ impurity region 14 of each N well region 11 and each N well region 11 are formed separately. It suffices if the source side P ⁺ impurity region 12 of the transistor is electrically connected, and there is no need for a step for differentiating the threshold voltage of each transistor as in the conventional case. Absent.

Further, as shown in FIG. 1, the second MOS transistors Q ₃ , Q ₇ , Q ₁₁ are arranged in parallel with the _first MOS transistors Q ₁ , Q ₅ , Q ₉ , Q ₁₃ , ..., Q _17. ,
Q _15, ..., a Q ₁₉ respectively provided, the first MOS transistor _{Q 1, Q 5, Q 9} , Q 13, ..., great to PN junction formed at the boundary between the drain and the N-well region of Q ₁₇ By preventing the current from flowing, it is possible to prevent the parasitic bipolar transistor from turning on, and to realize stable operation independent of the manufacturing process.

As shown in FIG. 5, the gate signals N _C and N _E of the transistors M ₁ and M ₃ are different from the clock signals φ _1A and φ _1B input to the drain terminals N _A and N _B , respectively. Since independent clock signals φ _2A and φ _2B can be input to turn on the transistors so that no potential difference is generated between the sources and drains of the transistors M ₁ and M ₃ , the transistors in the booster circuit can be turned on. It is possible to send out the charges such that the voltage drop due to the potential difference between the source and the drain does not occur when the charges are sent to the stage. Therefore, in equation (1), since the threshold voltage V _t can be put to zero, can boost efficiency as compared with the conventional circuit, the number of stages n and the power supply voltage V _dd of the booster circuit is the circuit identical to the prior art Even in this case, a higher output voltage V _POUT can be obtained. When the output voltage V _POUT is the same, the booster circuit of this embodiment can take a larger load current I _OUT .

For example, when the power supply voltage V _dd is 2.5 V and the number of step-up circuits n is 20, the capacitance ratio C / (C
Assuming that + C _s ) is 0.9, the absolute value of threshold voltage | V _t | is 0.6 V, and the load current I _OUT in the output stage is 0, only 20 V is obtained as the output voltage V _POUT in the conventional circuit. However, in the circuit according to the present embodiment, it is 47V.
I was able to get a value of the degree.

Further, in the semiconductor booster circuit according to this embodiment, a desired output voltage can be obtained even at a low power supply voltage V _dd which cannot be boosted by the conventional circuit. That is, in the conventional circuit, as shown in FIG. 10, the maximum output voltage is limited to a predetermined value by the power supply voltage V _dd regardless of the value of the number of stages n of the booster circuit. In the semiconductor booster circuit according to the above, there is practically no such limitation.

For example, when the power supply voltage V _dd is 2.0 V, the capacity ratio C / (C + C _s ) is 0.9, the absolute value of the threshold voltage | V _t | is 0.6 V, and the output stage Load current I
_{When OUT is set} to 0, in the conventional circuit, the output voltage V _POUT can obtain only 12V even when the number n of the booster circuits is 50. However, in the circuit according to the present embodiment, the number n of the booster circuits is 20. A value of about 37 V can be obtained, and a value of about 91 V can be obtained when the number n of boosting circuits is 50.

In the semiconductor booster circuit of this embodiment, if the absolute value | V _t | of the threshold voltage is 0.6V, the lower limit of the boostable power supply voltage V _dd is about 0.7V.

[0083]

According to the present invention, the substrate portions of the first MOS transistors forming each stage are electrically isolated from each other, and the substrate portions are electrically connected to the source terminals of the first MOS transistors. Since it is connected to, the substrate effect can be prevented and a high boosting capability can be obtained.

In addition to this, the first M that constitutes each stage
Since the second MOS transistor is provided in parallel with the OS transistor, the PN junction parasitically present between the drain terminal and the source terminal of the first MOS transistor and the substrate section is turned on during the operation of the booster circuit. Is prevented, and unnecessary injection of charges into the substrate portion is suppressed.

Further, in addition to these, a pair of first clock signals having mutually opposite phases are input to the other ends of the continuous two-stage first capacitances, and the continuous two-stage second capacitances are input. Since a pair of second clock signals having mutually different pulse timings are respectively input to the other end of, the number of stages of the booster circuit can be reduced as compared with the conventional case when the same boosting capability as the conventional case is obtained.

According to another feature of the present invention, the first MOS transistor is a P-channel MOS transistor formed in an N-type well region, and the N-type well region is electrically isolated in each stage. I decided to separate them,
A complicated manufacturing process can be eliminated.

[Brief description of drawings]

FIG. 1 is a circuit diagram showing a configuration of a semiconductor booster circuit according to an embodiment of the present invention.

FIG. 2 is a circuit diagram showing a configuration of two consecutive semiconductor booster circuits according to an embodiment of the present invention.

FIG. 3 is a waveform diagram showing clock timing of the semiconductor booster circuit according to the embodiment of the present invention.

FIG. 4 is a waveform diagram showing voltage waveforms at respective nodes of the semiconductor booster circuit according to the embodiment of the present invention.

FIG. 5 is a conceptual diagram for explaining the operation of the semiconductor booster circuit according to the embodiment of the present invention.

FIG. 6 is a schematic sectional view showing an element structure of a semiconductor booster circuit according to an embodiment of the present invention.

FIG. 7 is a circuit diagram showing a configuration of a conventional semiconductor booster circuit.

FIG. 8 is a waveform diagram showing clock timing of a conventional semiconductor booster circuit.

FIG. 9 is a graph showing the relationship between the number of stages of a conventional semiconductor booster circuit and the output voltage.

FIG. 10 is a graph showing the relationship between the power supply voltage and the maximum output voltage of the conventional semiconductor booster circuit.

[Explanation of symbols]

_{Q 1 ~Q 21, M 1 ~M} 8 P -channel MOS transistor Q _22, Q ₂₃ N-channel MOS transistor _{C 1 ~C 11, C A1 ~C} A3, C B1, C B2 capacitance N ₀ ~N _13, N _A ~ N _E node V _dd power supply voltage V _pout output voltage φ _1A , φ _1B , φ _2A , φ _2B clock signal 10 P type semiconductor substrate 11 N well region 12 P ⁺ impurity region 14 N ⁺ impurity region 15 gate oxide film 16 multi Crystalline silicon layer

─────────────────────────────────────────────────── ─── Continuation of front page (58) Fields surveyed (Int.Cl. ⁷ , DB name) H02M 3/07 G11C 17/00 H01L 21/822 H01L 27/04 H01L 29/78

Claims (2)

(57) [Claims] 1. Each stage comprises a first MOS transistor,
A first capacitance having one end connected to a drain terminal of the first MOS transistor;
A second capacitance whose one end is connected to the gate terminal of the transistor, wherein the first MOS transistors are connected in cascade to connect each stage, and the source of the first MOS transistor in each stage is connected. The terminal and the base portion are electrically connected to each other, and the substrate portion is electrically insulated from the base portion of the first MOS transistor in another stage, and the drain terminal of the first MOS transistor is provided. A source terminal is connected to each other via a second MOS transistor, a gate terminal of the second MOS transistor is connected to a source terminal of the first MOS transistor, and the first terminal in each stage is connected to the first terminal. The third MOS transistor and the fourth MO transistor in which the gate terminal and the source terminal of the MOS transistor are arranged in parallel. They are connected to each other through an S transistor, the gate terminal of the third MOS transistor is connected to the one end of the first capacitance, and the gate terminal of the fourth MOS transistor is connected to the first MOS transistor. A pair of first clock signals, which are connected to the source terminal and have opposite phases, are input to the other ends of the first capacitances in two consecutive stages, and the second capacitances in two consecutive stages are connected to each other. A semiconductor booster circuit, wherein a pair of second clock signals having different pulse timings are input to the other end of the semiconductor booster circuit. 2. The first MOS transistor is a P-channel MOS transistor formed in an N-type well region, and the N-type well region is electrically isolated for each stage. The semiconductor booster circuit according to claim 1.