Circuit responsive to a voltage change for detecting erroneous writing into a memory

FIELD OF THE INVENTION
This invention relates to a circuit for detecting a voltage in a memory apparatus, and more particularly to, a circuit for detecting a decrease of a power supply voltage and an input high voltage in an electrically erasable and programable read-only memory (defined "EEPROM" hereinafter) including mainly insulated-gate field-effect transistors (defined "IGFETs" hereinafter).
BACKGROUND OF THE INVENTION
An EEPROM is provided with external input terminals such as a chip enable terminal CE, a write enable terminal WE, an output enable terminal OE, etc., to be mounted in, for instance, an information processing system, so that a writing or an erasing is carried out in a writing mode which is set by aplying high and low signals to the above terminals.
One example of modes to be set is explained in the below table.
______________________________________ TERMINAL MODE ##STR1## ##STR2## ##STR3## I/O______________________________________READ L H L DATA OUTPUTWRITE L L H DATA INPUTSTAND-BY H NON- NON- HIGH- SPECI- SPECI- IMPEDANCE FIED FIEDINHIBIT OF NON- NON- HWRITE SPECI- SPECI- FIED FIED NON- H NON- SPECI- SPECI- FIED FIED______________________________________
In this table, the letter "H" is a high signal, and the letter "L" is a low signal. Furthermore, the letters "I/O" indicate an input and output terminal.
Even if one of the modes is correctly selected by applying the signals specified in the table to the respective terminals, an erroneous writing or an erroneous erasing is liable to occur in memory cells in the EEPROM, where a power supply is shut off or a voltage thereof is decreased outside an operation voltage range due to a failure, etc. of the system in which the EEPROM is mounted. For instance, where the stand-by mode is set in the state that a high signal is applied to the CE terminal, a low signal is applied to the WE terminal, and a high signal is applied to the OE terminal, it is assumed that a power supply voltage is abruptly decreased, so that the high signals of the CE and OE terminals are changed to be low. In this situation, it is further assumed that a capacitive load is connected to the OE terminal ten times faster than a capacitive load is connected to the CE terminal, so that a rate by which a voltage is changed from the high signal to a low signal at the CE terminal is faster than a rate of the change at the OE terminal. This results in a signal state of a writing mode in the above table momentarily. As a result, indefinite data on the I/O terminal having a state of a high impedance for the stand-by mode are erroneously written into an address of a memory cell.
For the purpose of avoiding such an erroneous writing occuring due to the decrease of the power supply voltage, a circuit for detecting the voltage decrease is generally built in an EEPROM system. The power supply voltage detecting circuit supplies a high signal at its output terminal, when a power supply voltage Vcc is larger than a predetermined voltage Vcc.sub.(INV), and a low signal at its output terminal, when the voltage Vcc is equal to or smaller than the voltage Vcc.sub.(INV). Consequently, the EEPROM is controlled to be in an inhibit state of a data writing, where the low signal is supplied from the output terminal of the detecting circuit. As a matter of course, the voltage Vcc.sub.(INV) is determined to be lower than an operating range of the power supply voltage Vcc, and is desired to be suppressed from the change thereof due to the fluctuation of an environmental temperature around the EEPROM, device parameters of transistors in fabricating the EEPROM, etc.
If it is assumed that the voltage Vcc.sub.(INV) is increased to be inside the operating range of the power supply voltage, a data writing becomes impossible even in an ordinary writing mode. On the contrary, if the voltage Vcc.sub.(INV) is set to be too low, it becomes impossible to detect the decrease of the power supply voltage Vcc precisely, so that the aforementioned erroneous writing occurs in the EEPROM.
A conventional circuit for detecting the decrease of a power supply voltage includes three Nchannel enhancement mode IGFETs (defined "NE-IGFETs" hereinafter), two N-channel depletion mode IGFET (defined "ND-IGFET" hereinafter), and an inverter, such that a high or low signal is supplied from an output terminal of the inverter dependent on a detecting result of a power supply voltage Vcc, although the structure and operation of the detecting circuit will be explained in detail later. Accordingly, an erroneous writing can be avoided in accordance with the output signal of the inverter in the detecting circuit.
However, the conventional circuit for detecting a decrease of a power supply voltage has disadvantages as follows, the cause of which will be also explained in detail later.
(1) The predetermined voltage Vcc.sub.(INV) is set to be three times a threshold voltage of the NE-IGFET. Therefore, where the threshold voltage is changed due to a temperature change by .DELTA.V.sub.THN, the predetermined voltage Vcc.sub.(INV) is changed as much as 3. .DELTA.V.sub.THN. As a result, the predetermined voltage Vcc.sub.(INV) is increased at a low temperature to be near the operating voltage range (for example, 4.5 V to 5.5 V), while the predetermined voltage Vcc.sub.(INV) is decreased at a high temperature, thereby making it difficult to detect a decrease of a power supply voltage precisely.
(2) In order that the predetermined voltage Vcc.sub.(INV) is not inside the operating voltage range at a low temperature, the predetermined voltage Vcc.sub.(INV) must be designed by considering a temperature characteristic of the threshold voltage of the NEIGFET. For this reason, the design of the circuit for detecting the decrease of a power supply voltage becomes difficult.
(3) The predetermined voltage Vcc.sub.(INV) is liable to be too low at a low temperature, thereby making it difficult to avoid an erroneous writing of the EEPROM occuring due to the decrease of the power supply voltage in a wide temperature range. Consequently, the usage temperature range of the EEPROM becomes narrow.
SUMMARY OF THE INVENTION
Accordingly, it is an object of this invention to provide a circuit for detecting a voltage in a memory apparatus, the design of which is easy.
It is a further object of this invention to provide a circuit for detecting a voltage in a memory apparatus, in which the detection of a power supply voltage is carried out in a wide temperature range.
It is a still further object of this invention to provide a circuit for detecting a voltage in a memory apparatus, by which the inhibit of an ordinary operation of the EEPROM is avoided at a high temperature, and an erroneous writing is avoided in the EEPROM at a low temperature.
According to this invention, a circuit for detecting a voltage in a memory apparatus, comprises:
a first p-channel field-effect transistor having a source connected to a power supply, and a gate and a drain commonly connected;
an n-channel depletion mode field-effect transistor having a drain connected to the drain of the first p-channel field effect transistor, and a gate and a source commonly connected to the ground;
a field-effect transistor having a gate connected to the power supply to be turned on;
a first n-channel enhancement mode field-effect transistor having a drain and a gate commonly connected to a drain of the field-effect transistor and a source connected to the ground;
a second p-channel field-effect transistor having a source connected to the power supply, a gate connected to a common node between the first p-channel field-effect transistor and the n-channel depletion mode field-effect transistor, and a drain connected to a voltage detecting node;
a second n-channel enhancement mode field-effect transistor having a drain connected through the voltage detecting node to the drain of the second p-channel field-effect transistor, a gate connected to a common node between the field-effect transistor and the first n-channel mode enhancement mode field-effect transistor, and a source connected to the ground; and
an inverter connected between the power supply and the ground, and having an input terminal connected to the voltage detecting node and an output terminal for providing a voltage detected result.

BRIEF DESCRIPTION OF THE DRAWINGS
This invention will be explained in more detail in conjunction with appended drawings, wherein:
FIG. 1 is a circuitry diagram showing a conventional circuit for detecting a decrease of a power supply voltage in a memory apparatus.
FIG. 2 is an explanatory diagram showing voltages changing at nodes C.sub.1, C.sub.2, C.sub.3 and C.sub.4 in acordance with the change of a power supply voltage,
FIG. 3 is an explanatory diagram showing the change of a predetermined voltage to be detected in the circuit in FIG. 1,
FIG. 4 is a circuitry diagram showing a circuit for detecting a decrease of a power supply voltage in a memory system in a first preferred embodiment according to the invention,
FIG. 5 is an explanatory diagram showing the change of currents IL.sub.1 and IL.sub.2 in accordance with the change of a power supply voltage,
FIG. 6 is an explanatory diagram showing voltages at nodes A.sub.1, A.sub.2 and A.sub.4 changing in accordance with the change of a power supply voltage,
FIG. 7 is an explanatory diagram showing a predetermined voltage to be detected which is not changed dependent on a temperature,
FIG. 8 is a circuitry diagram showing an input high voltage detecting circuit in a second preferred embodiment according to the invention,
FIG. 9 is an explanatory diagram showing the change of currents IL.sub.3 and IL.sub.4 in accordance with the change of an external applied voltage in FIG. 8, and
FIG. 10 is an explanatory diagram showing the change of voltage at nodes B.sub.1, B.sub.2, and B.sub.4 in accordance with the change of the external applied voltage

DESCRIPTION OF THE PREFERRED EMBODIMENTS
Before explaining a circuit for detecting a voltage in a memory apparatus in a preferred embodiment according to the invention, the aforementioned conventional circuit for detecting a power supply voltage in an EEPROM will be explained again.
FIG. 1 shows the conventional circuit for detecting a power supply voltage which comprises NEIGFETs Q.sub.21, Q.sub.22 and Q.sub.25, ND-IGFETs Q.sub.23 and Q.sub.24, and an inverter I connected between a power supply CC and the ground, wherein the NE-IGFETs and ND-IGFETs Q.sub.21 to Q.sub.25 are connected via nodes C.sub.1, C.sub.2 and C.sub.3 to each other, and the inverter I is connected at an input terminal to the mode C.sub.3 and at an output terminal to the node C.sub.4. In this circuit, each IGFET is designed in regard to a ratio of a gate width to a gate length (defined "W/L" hereinafter), such that an equivalent resistance of Q.sub.23 is higher than those of Q.sub.21 and Q.sub.22, and an equivalent resistance of Q.sub.24 is higher than that of Q.sub.25. For instance, the ratio of W/L is designed to be 60/3 for Q.sub.21 and Q.sub.22, 6/2000 for Q.sub.23 and Q.sub.24, and 60/3 for Q.sub.25. Therefore, currents flowing through Q.sub.21 to Q.sub.23, and through Q.sub.24 and Q.sub.25 are as small as approximately 0.5 .mu.A. Furthermore, a logic threshold value of an inverter composed of Q.sub.24 and Q.sub.25 is set to be V.sub.THN which is a threshold value common to the NE-IGFETs.
In operation, a voltage Vcc of a power supply CC is applied to the circuit, so that voltages VC.sub.1, VC.sub.2, VC.sub.3 and VC.sub.4 of the nodes C.sub.1 to C.sub.4 are changed, where the voltage Vcc is changed, as shown in FIG. 2.
Vcc>3.multidot.V.sub.THN (1)
The voltage VC.sub.2 is higher at the node C.sub.2 than the threshold voltage V.sub.THN, so that the voltage VC.sub.3 is O V (the ground level) at the node C.sub.3, and the voltage VC.sub.4 is Vcc at the node C.sub.4.
Vcc.ltoreq.3.multidot.V.sub.THN (2)
As the power supply voltage Vcc is decreased, the voltages VC.sub.1 and VC.sub.2 of the nodes C.sub.1 and C.sub.2 are also decreased. Where the voltage VC.sub.2 becomes equal to or less than the threshold voltage V.sub.THN, the voltage VC.sub.3 of the node C.sub.3 is Vcc, and the voltage VC.sub.4 of the node C.sub.4 is O V.
Accordingly, the predetermined voltage Vcc.sub.(INV) which is a reference voltage in detecting the decrease of a power supply voltage is expressed by the equation (1).
Vcc.sub.(INV) .apprxeq.3.multidot.V.sub.THN (1)
The threshold voltage V.sub.THN is expressed, as described in "the Physics and Technology of Semiconductor Devices", by the equation (2). ##EQU1## where V.sub.FBP is a flat-band potential of a p-semiconductor substrate, .phi. is a Fermi level of a p-semiconductor substrate (.phi..sub.FP >0),
Cox is a gate capacitance per a unit area,
Ks is a dielectric constant of a substrate,
.epsilon.o is a permittivity of vacuum, and
N.sub.A is an acceptor concentration of a p-semiconductor substrate.
In the following explanation, it is presumed that,
Ks=11.8,
.epsilon.o=8.854.times.10.sup.-12 (F/m),
Cox=.epsilon.o.multidot.K.sub.ox /t.sub.ox,
and
Kox=3.8.
In the equation (2), V.sub.FBP is expressed by the equation (3). ##EQU2## where .phi..mu. is a work function of an aluminum electrode, .phi.sp is a work function of a p-semiconductor,
X is an electron affinity of silicon,
Eg is an energy gap of silicon,
k is Boltzmann constant,
T is absolute temperature, and
Ni is an intrinsic carrier concentration.
In the following explanation, it is presumed that, k=8.62.times.10.sup.-5 eV/K, and Eg=1.1 eV.
In the equation (2), .phi..sub.FP is expressed by the equation (4).
.phi..sub.FP =kT/q ln N.sub.A /Ni (4)
In the equations (2) to (4), the below described values are substituted.
.phi..mu.=4.252V,
X=4.1V,
Eg=1.1eV,
Cox=0.067.times.10.sup.-2 F/m (500 .ANG.),
Ks=11.8,
N.sub.A =2.5.times.10.sup.22 /m.sup.3,
and
Ni=1.45.times.10.sup.16 /m.sup.3 (at a temperature of 25.degree. C.).
Thus, .phi..sub.FP is expressed at a temperature of 25.degree. C. by the equation (5).
.phi..sub.FP =0.37V (5)
The threshold voltage V.sub.THN is expressed at a temperature of 25.degree. C. by the equation (6), as a result of the substitution of the above described values into the equation (2). ##EQU3##
Accordingly, the predetermined value Vcc.sub.(INV) is expressed at a temperature of 25.degree. C. by the equation (7). ##EQU4##
On the other hand, .phi..sub.FP is expressed at a temperature of 100.degree. C. by the equation (8), where Ni is presumed to be 2.times.10.sup.18 /m.sup.3.
.phi..sub.FP =0.30V (8)
Furthermore, the threshold voltage V.sub.THN is expressed at a temperature of 100.degree. C. by the equation (9), as explained in the equation (6). Here, Eg is not changed dependent on a temperature. ##EQU5##
Accordingly, the predetermined voltage Vcc.sub.(INV) is expressed at a temperature of 100.degree. C. by the equation (10). ##EQU6##
As understood from the above calculations, the predetermined voltage Vcc.sub.(INV) which is a reference voltage in detecting a decrease of a power supply voltage is changed between 2.85 V and 3.42 V in accordance with a temperature change between 100.degree. C. and 25.degree. C. In other words, the voltage Vcc.sub.(INV) is changed by a voltage range of 0.57 V, as a temperature is changed by a range of 75.degree. C. This phenomenon of the conventional circuit is shown in FIG. 3. As apparent therefrom, the predetermined voltage Vcc.sub.(INV) which is three times the threshold voltage V.sub.THN is decreased in an accordance with a negative temperature coefficient, along with the decrease of the threshold voltage V.sub.THN.
Next, a circuit for detecting a voltage in a memory apparatus in the first preferred embodiment according to the invention will be explained in FIG. 4. The detecting circuit comprises P-channel enhancement mode IGFETs (defined "PE-IGFET" hereinafter) Q.sub.1 and Q.sub.5, ND-IGFETs Q.sub.2 and Q.sub.3, NE-IGFETs Q.sub.4 and Q.sub.6, and an inverter connected between a power supply CC and the ground. In this detecting circuit, a ratio of "W/L" for each IGFET is designed, such that an equivalent resistance of Q.sub.2 is larger than that of Q.sub.1, an equivalent resistance of Q.sub.4 is larger than that of Q.sub.3, a ratio of "W/L" for Q.sub.5 is larger than that for Q.sub.1 by n times, and a ratio of "W/L" for Q.sub.6 is larger than that for Q.sub.4 by n times. Therefore, current IL.sub.1 flowing through Q.sub.1 and Q.sub.2 is determined by a current derived power of Q.sub.2, and current IL.sub.2 flowing through Q.sub.3 and Q.sub.4 is determined by a current derived power of Q.sub.4. For instance, the ratio of "W/L" is designed to be 10/3 for Q.sub.1 and Q.sub.5, and 6/2000 for Q.sub.2, Q.sub.4 and Q.sub.6, and the numbers of "m" and "n" are presumed to be "1" (m=n=1).
In such an instance, the currents of IL.sub.1 and IL.sub.2 are very small to be several 100 nA, respectively, where the power supply voltage Vcc is 5 V.
Operation of the circuit as shown in FIG. 4 will be explained in conjunction with FIGS. 5 and 6, wherein the currents of IL.sub.1 and IL.sub.2 are changed dependent on the power supply voltage Vcc in FIG. 5, and voltages VA.sub.1, VA.sub.2 and VA.sub.4 are changed at nodes A.sub.1, A.sub.2 and A.sub.4 dependent on the power supply voltage Vcc in FIG. 6. As shown in FIG. 6, the voltage VA.sub.1 is fixed at the node A.sub.1 to be "Vcc-(.vertline.V.sub.THP .vertline.+.alpha.)" which is a voltage difference necessary for the flowing of the current IL.sub.1 through Q.sub.1, where V.sub.THP is a threshold voltage of the PE-IGFETs. The current IL.sub.1 is determined by the equivalent resistance of Q.sub.2. Therefore, where the power supply voltage Vcc is increased to be larger than the threshold voltage .vertline.V.sub.THP .vertline.(Vcc>.vertline.V.sub.THP .vertline.), Q.sub.1 is turned on, so that the current IL.sub.1 which is determined by a current to voltage characteristic of Q.sub.2 as shown in FIG. 5 flows through Q.sub.2.
On the other hand, the voltage VA.sub.3 is fixed at the node A.sub.3 to be of the power supply voltage Vcc, and the current IL.sub.2 is determined by the equivalent resistance of Q.sub.4. Therefore, where the power supply voltage Vcc is increased to be larger than the threshold voltage V.sub.THN of the NE-IGFET Q.sub.3, Q.sub.4 is turned on, so that the current IL.sub.2 which is determined by a current to voltage characteristic of Q.sub.4 flows through Q.sub.4. As shown in FIG. 4, Q.sub.1 and Q.sub.5, and Q.sub.4 and Q.sub.6 are connected to provide current mirror circuits, so that Q.sub.5 operates to supply the node A.sub.2 with the current IL.sub.1, and Q.sub.6 operates to supply the ground with the current IL.sub.2. The ratio of "W/L" is designed for Q.sub.2 and Q.sub.4, such that the currents of IL.sub.1 and IL.sub.2 are equal to each other, where the power supply voltage Vcc is equal to the predetermined voltage Vcc.sub.(INV), as shown in FIG. 5. Furthermore, the threshold voltage V.sub.THN of Q.sub.2 is designed, such that Q.sub.2 operates in a saturated region, where the power supply voltage Vcc is equal to or larger than the predetermined voltage Vcc.sub.(INV) (Vcc.gtoreq.Vcc.sub.(INV)). Consequently, the current IL.sub.1 is constant without any dependency of the power supply voltage Vcc.
On the other hand, the current IL.sub.2 is proportional to (Vcc-V.sub.THN), because a gate and a drain of Q.sub.4 are commonly connected.
Next, the voltages VA.sub.1, VA.sub.2 and VA.sub.4 of the nodes A.sub.1, A.sub.2 and A.sub.4 will be explained to be changed, where the power supply voltage Vcc is decreased to be lower than a designed value.
Vcc>Vcc.sub.(INV)
As shown in FIG. 5, the current IL.sub.2 is larger than the current IL.sub.1. At this time, the voltage VA.sub.2 of the node A.sub.2 is determined by a ratio of current derived powers of Q.sub.5 and Q.sub.6. Therefore, the voltage VA.sub.2 becomes near the ground potential to be lower than a logic threshold value of the inverter I, because the current derived power of Q.sub.6 is larger than that of Q.sub.5, so that the power supply voltage Vcc is applied to the node A.sub.4 by the inverter I.
(2) Vcc=Vcc.sub.(INV)
Where the power supply voltage Vcc is decreased to be equal to the predetermined voltage Vcc.sub.(INV), the currents of IL.sub.1 and IL.sub.2 become equal to each other. Therefore, the current derived powers of Q.sub.5 and Q.sub.6 become equal to each other, so that the voltage VA.sub.2 is increased at the node A.sub.2. As a result, the currents flowing through Q.sub.5 and Q.sub.6 are equal to the currents of IL.sub.1 and IL.sub.2.
(3) Vcc<Vcc.sub.(INV)
Where the power supply voltage Vcc is more decreased to be lower than the predetermined voltage Vcc.sub.(INV), the current IL.sub.1 becomes larger than the current IL.sub.2. As a result, the current derived power of Q.sub.5 becomes larger than that of Q.sub.6, so that the voltage VA.sub.2 becomes at the node A.sub.2 near the power supply voltage Vcc. At this time, the voltage VA.sub.2 is higher than the logic threshold value of the inverter I, so that the voltage VA.sub.4 of the node A.sub.4 is O V.
Furthermore, a level of the predetermined voltage Vcc.sub.(INV) which is detected to inhibit a writing mode of EEPROM by the circuit for detecting the decrease of a power supply voltage in the preferred embodiment will be obtained as follows. As described before, the power supply voltage Vcc, by which values are equal between the currents of IL.sub.1 and IL.sub.2 (IL.sub.1 = IL.sub.2), is the predetermined value Vcc.sub.(INV). The current IL.sub.1 is expressed by the equation (11), because Q.sub.2 operates in the saturated region. ##EQU7## where .beta..sub.D =W/L.multidot..mu..sub.D .multidot.Cox for Q.sub.2, .mu..sub.D is a mobility of the ND-IGFET, and
Cox is a gate capacity of a unit area.
The current IL.sub.2 is expressed by the equation (12), because Q.sub.4 operates in the saturated region in the same manner. ##EQU8## where .beta..sub.N =W/L.multidot..mu..sub.N .multidot.Cox for Q.sub.4, and .mu..sub.N is a mobility of the NE-IGFETs Q.sub.4 and Q.sub.6.
The predetermined voltage Vcc.sub.(INV) is expressed by the equation (13), where the equation (11) is equal to the equation (12), and the power supply voltage Vcc is equal to the predetermined voltage Vcc.sub.(INV). ##EQU9##
If it is assumed that the ratio of "W/L" is designed, such that .beta..sub.N is equal to .beta..sub.D, the equation (13) is expressed by the equation (14).
Vcc.sub.(INV) =V.sub.THN -V.sub.THD (14)
In the equation (14), V.sub.THN is expressed by the aforementioned equation (2), and V.sub.THD is expressed by the equation (15). ##EQU10## where Nc is a donor concentration of the IGFET to be a depletion type, and Tc is a donor depth of a substrate.
Where the equations (2) and (5) are substituted into the equation (14), the equation (16) is obtained. ##EQU11##
In the equation (16), Tc, Nc and Cox are not changed substantially in regard to a temperature, so that the predetermined voltage Vcc.sub.(INV) is not changed substantially in regard to a temperature as shown in FIG. 7.
For instance, if it is assumed that,
Tc=10.sup.-7 m,
Nc=1.38.times.10.sup.23 /m.sup.3,
and
Cox=0.067.times.10.sup.-2 F/m.sup.2
the predetermined voltage Vcc.sub.(INV) is 3.3 V as follows. ##EQU12##
In conclusion, the detected voltage is constant to be 3.3 V in the circuit for detecting a power supply voltage according to the invention, even if a temperature is changed.
As understood from the above, a detected voltage Vcc.sub.(INV) is set to be a difference between threshold values of NE- and ND-IGFETs, so that the three terms of the equation (2), that is, the front three terms of the equation (15) which are changed dependent on a temperature are eliminated therefrom. Therefore, only the term which is not changed substantially dependent on a temperature is left therein, so that a temperature dependency of a voltage which is detected in the circuit for detecting a power supply voltage according to the invention is negligible as compared to the conventional one. Furthermore, a degree of freedom for setting a voltage to be detected is increased as compared to a case where a voltage is set to be detected in the multiplication of the threshold voltage V.sub.THN by n (n:an integer), as described in the conventional circuit, because a voltage is set to be detected in accordance with the conditions of injecting ions such as a kind of substances to be injected, a doping amount, an accelerating energy, etc.
FIG. 8 shows a circuit for detecting a power supply voltage in a memory apparatus in the second embodiment according to the invention which is applied to a circuit for detecting an input high voltage in an electrically programmable read-only memory (EPROM). In FIG. 8, Q.sub.11 and Q.sub.15 are PE-IGFETs, Q.sub.12 and Q.sub.13 are ND-IGFETs, Q.sub.14 and Q.sub.16 are NE-IGFETs, PP is an external input terminal, and I is an inverter, wherein a threshold voltage of Q.sub.14 and Q.sub.16 is set to be V.sub.THN-H which is larger than a threshold voltage V.sub.THN of an NE-IGFET, and a ratio of "W/L" for each IGFET is set in the same manner as in the first preferred embodiment.
In operation, current IL.sub.3 flows through Q.sub.11 and Q.sub.12, and current IL.sub.4 flows through Q.sub.13 and Q.sub.14, where a voltage Vpp is changed to be applied to the external input terminal PP, as shown in FIG. 9. In this circumstance, voltages VB.sub.1, VB.sub.2, and VB.sub.4 are changed at nodes B.sub.1, B.sub.2, and B.sub.4, where the voltage Vpp is changed. In the second preferred embodiment, the voltage V.sub.THN-H is set to be the power supply voltage Vcc (=5 V), and the ratio of "W/L" for Q.sub.12 and Q.sub.14 is designed, such that the current IL.sub.3 is equal to the current IL.sub.4 (IL.sub.3 =IL.sub.4), where the applied voltage Vpp is a predetermined voltage Vpp.sub.(INV). As a matter of course, the current IL.sub.3 is determined by an equivalent current of Q.sub.12, and the current IL.sub.4 is determined by an equivalent current of Q.sub.14 in the same manner as in the first preferred embodiment. Furthermore, Q.sub.11 and Q.sub.15, and Q.sub.14 and Q.sub.16 provide current mirror circuits. The below explanation is for voltage changes of the nodes B.sub.1 to B.sub.4 in operation, where the applied voltage Vpp is increased from 0 V.
(1) Vpp<V.sub.THN-H =5 V
Where the applied voltage Vpp becomes larger than .vertline.V.sub.THP .vertline., Q.sub.11 is turned on, so that current IL.sub.3 flows through Q.sub.11 and Q.sub.12. At this time, a voltage VB.sub.3 is approximately [Vpp] at the node B.sub.3, while Q.sub.14 is non-conductive. Accordingly, Q.sub.15 is conductive, and Q.sub.16 is non-conductive, while a voltage VB.sub.2 is [Vpp] at the node B.sub.2, and a voltage is 0 V at the node B.sub.4.
(2) V.sub.THN-N .ltoreq.Vpp<Vpp.sub.(INV)
Where the applied voltage Vpp becomes a value which is equal to or larger than the voltage V.sub.THN-H, and smaller than Vpp.sub.(INV), the current IL.sub.3 is not changed, even if the applied voltage is changed, because Q.sub.12 operates in the saturated region. On the other hand, Q.sub.14 is turned on, so that the current IL.sub.4 flows through Q.sub.13 and Q.sub.14. As a result, a voltage VB.sub.2 is decreased at the node B.sub.2 from the applied voltage Vpp in accordance with the increase of the applied voltage Vpp, because the voltage VB.sub.2 is determined by current derived powers of Q.sub.15 and Q.sub.16. At this moment, a voltage VB.sub.4 is still 0 V, because the voltage VB.sub.2 is set to be more than a logic threshold value of the inverter I.
(3) Vpp=Vpp.sub.(INV)
Where the applied voltage Vpp is equal to the predetermined voltage Vpp.sub.(INV) in accordance with the increase of the applied voltage Vpp, the currents IL.sub.3 and IL.sub.4 are equal to each other (IL.sub.3 =IL.sub.4). Accordingly, the current derived powers of Q.sub.15 and Q.sub.16 are equal to each other to decrease the voltage VB.sub.2, so that the current IL.sub.3 flowing through Q.sub.15 and the current IL.sub.4 flowing through Q.sub.16 are equal to each other.
(4) Vpp>Vpp.sub.(INV)
Where the applied voltage Vpp is more increased to be larger than Vpp.sub.(INV) (Vpp>Vpp.sub.(INV)), the curent IL.sub.4 is larger than the current IL.sub.3 (IL.sub.4 >IL.sub.3). At this moment, the voltage VB.sub.2 is more decreased to be lower than the logic threshold value of the inverter I, because the current derived power of Q.sub.16 is larger than that of Q.sub.15. Therefore, the voltage VB.sub.4 is [Vcc] at the node B.sub.4.
Next, the predetermined voltage Vpp.sub.(INV) which is detected in the input high voltage detecting circuit in the second preferred embodiment will be calculated. As explained in the first preferred embodiment, a value of the applied voltage Vpp which meets the relation (IL.sub.3 =IL.sub.4) is the predetermined voltage Vpp.sub.(INV). The current IL.sub.3 is expressed by the equation (11) in the first preferred embodiment, and the current IL.sub.4 is expressed by the equation (17). ##EQU13## where .beta..sub.NH =W/L.multidot..mu..sub.NH .multidot.Cox for Q.sub.14, and
.mu..sub.NH is mobilities of the NE-IGFET Q.sub.14 and Q.sub.16.
If it is assumed that the ratio of "W/L" is designed for Q.sub.11 and Q.sub.14, such that .beta..sub.NH is equal to .beta..sub.D, where the equation (11) is equal to the equation (17), the equation (18) is obtained.
Vpp.sub.(INV) =V.sub.THN-H -V.sub.THD (18)
Here, if it is assumed that a concentration and a depth of acceptor which is injected below a gate of a substrate is N.sub.AC and T.sub.A, respectively, the voltage V.sub.THN-H is expressed by the equation (19). ##EQU14##
Thus, the voltage Vpp.sub.(INV) is expressed by the equation (20), because the voltage V.sub.THD is expressed by the equation (15) as explained in the first preferred embodiment. ##EQU15##
For instnce, if it is assumed that,
Tc=10.sup.-7 m,
Nc=1.38.times.10.sup.23 /m.sup.3
Cox=0.067.times.10.sup.-2 F/m.sup.2
T.sub.A =10.sup.-7 m,
and
N.sub.AC =2.1.times.10.sup.23 /m.sup.3
the voltage Vpp.sub.(INV) is calculated below.
Vpp.sub.(INV) =3.3V+5V=8.3V
In the equation (20), the voltage Vpp.sub.(INV) is not changed substantially dependent on a temperature, because Tc, Nc, T.sub.A, N.sub.AC and Cox are not changed substantially dependent on a temperature. Accordingly, the same advantage as explained in the frst preferred embodiment is obtained in the second preferred embodiment.
Although an ND-IGFETm, a gate and a drain of which are conmonly connected, is utilized for Q3 in the first and second preferred embodiments, a circuit structure, in which Q.sub.3 is always turned on, and a power supply voltage is obtained at a node A.sub.3, for instance, an ND-IGFET, a gate and a source of which are commonly connected, a PE-IGFET, a gate of which is connected to the ground, etc. may be utilized.
As understood from the above, following advantages are expected in a circuit for detecting a decrease of a power supply voltage, and a circuit for detecting an input high voltage according to the invention.
(1) A detected voltage is not changed substantialy dependent on a temperature, because the detected voltage is determined by a difference of threshold values of NE-IGFET and ND-IGFET, so that terms which are changed dependent on a temperature are all eliminated.
(2) A detected voltage is set to be an arbitrary value, because the detected voltage is set by the ion injection conditions such as a kind of substanees to be injected, a doping amount, an accelerating energy, etc.
Consequently, an EEPROM which is built in a system operating in a wide temperature range can be provided in a market, because an erroneous writing is precisely avoided, even a power supply voltage is decreased in a wide temperature range.
Although the invention has been described with respect to specific embodiment for complete and clear disclosure, the appended claims are not to be thus limited but are to be construed as embodying all modification and alternative constructions that may occur to one skilled in the art which fairly fall within the basic teaching herein set forth.

Circuit responsive to a voltage change for detecting erroneous writing into a memory

Information

Patent Number

Date Filed

Date Issued

Inventors

Original Assignees

Examiners

Agents

CPC

US Classifications

Field of Search

US

International Classifications

Abstract

Description

Claims

Priority Claims (1)

Foreign Referenced Citations (1)