Semiconductor Memory Device

(1) Field of the Invention

The present invention relates to a semiconductor memory device, and more specifically to a decoder circuit of a semiconductor device which writes information by causing selected memory cells such as PROM (programmable read only memory) cells, to be conductive (short circuited) or non-conductive (opened).

(2) Description of the Prior Art

The programmable read only memory usually includes address inverters, decoders, memory cells, multiplexers, output buffers and a programming circuit. The memory cells are either of the junction type, in which the emitter-base junctions are connected to intersecting points of word lines and bit lines so that the junctions are short-circuited by the heat generated by a writing current, or of the blown-fuse type, in which a region of evaporated metal or polycrystalline silicon is melted and cut by the writing current.

With such a semiconductor memory device, the address inverters, memory cells, decoder circuits, multiplexers, and output buffers must be simplified and reduced in size with the increase in the memory capacity.

The semiconductor memory device which employs programmable read only memory cells operates on such greatly different writing current levels and reading current as, for example, 200 mA and 0.5 mA. In such a programmable read only memory, if a current which is enough to drive a writing current is fed to the base of a transistor which drives a reading current when the information is being read out, the transistor is saturated and electric charges accumulate in its base. Therefore, an extended period of time is required to render the transistor non-conductive; i.e., the circuit is not suited for high-speed operation. In order to ameliorate the above problem, therefore, a conventional decoder circuit has been employed which uses a feedback circuit having a diode to supply a suitable base current to the above-mentioned transistor when information is to be written in and read out (refer to ELECTRONICS DIGEST, Practical Application of IC Memory PROM, pp. 2-7, August, 1974, that was published in Japan).

Provision of the feedback circuit, however, makes wiring complicated, and is not desirable from the standpoint of highly integrating the circuits. Furthermore, the feedback circuit is not adequate for high-speed operation. Instead of providing the feedback circuit, one might consider switching over the power supply between reading and writing. However, since the power supply must feed a voltage not only to the decoder circuit but also to other circuits, it is difficult to switch over the power supply in view of design requirements for other circuits.

The object of the present invention is to provide a semiconductor memory device which eliminates the above-mentioned defects inherent in the conventional art, which features a simply constructed circuit setup, and which makes it possible to highly integrate the circuits.

Another object of the present invention is to provide a semiconductor memory device which is capable of being operated at high speeds.

In order to achieve the above objects, the present invention discloses a semiconductor memory device which comprises an address buffer which receives address signals, address signal lines which will be set to levels corresponding to the address signals by the output of the address buffer, a decoder circuit which decodes outputs of the address signals, and memory cells which will be rendered conductive or non-conductive in accordance with write information, wherein when a memory device is in the state, the current which feeds a base current to a transistor in a final stage of the decoder circuit to drive a selected word line is fed at power supply voltage, and when the device is in write state, the current is fed through a source whose voltage is higher than power supply voltage.

Further features and advantages of the present invention will become apparent from the ensuing description with reference to the accompanying drawings to which, however, the scope of the invention is in no way limited.

FIG. 1 is a block diagram illustrating a conventionally employed semiconductor memory device;

FIG. 2 is a circuit diagram of a conventional semiconductor memory device;

FIG. 3 is a circuit diagram of a semiconductor memory device according to the present invention, and especially a decoder circuit according to an embodiment of the present invention; and

FIG. 4 is a circuit diagram of a circuit for providing a current which feeds driving current to transistors of decoder driver in the semiconductor memory device according to the present invention.

In general, a programmable read only memory consists of, as illustrated in FIG. 1, an address buffer 1 which receives address signals A.sub.0 to A.sub.8, a decoder driver 2 which decodes the output of the address buffer 1, memory cells 3, a programming circuit 4 which introduces information in the direction Y of the address buffer 1 and which feeds an output to the memory cells 3, a multiplexer 5 which transmits information in the direction Y of the memory cells 3, a CE (chip enable) circuit 6 which introduces chip enable signals, and an output buffer 7 which introduces the output of the CE circuit as well as the content of the multiplexer 5.

FIG. 2 is a partial circuit diagram of a conventional semiconductor memory device, and particularly, a partial circuit diagram of a junction-type programmable read only memory.

In FIG. 2, symbols INV.sub.00 to INV.sub.11 denote address inverters, AL.sub.0 to AL.sub.3 denote address signal lines, 2 denotes a decoder driver circuit, Q.sub.1 to Q.sub.4 denote transistors, R.sub.1 to R.sub.5 denote resistors, D.sub.1 denotes a diode, 4 denotes a programming circuit, 3 denotes a memory cell unit, Q.sub.00 to Q.sub.1n denote memory cells, WL.sub.0, WL.sub.1 --denote word lines, and BL.sub.0 to BL.sub.n denote bit lines.

The writing operation of the circuit is described below with reference to the case where information is to be written down by short-circuiting the emitter-base junction of a memory cell Q.sub.00 in the memory cell unit 3.

When address signals are fed to input terminals A.sub.0 and A.sub.1 of the inverters INV.sub.00 and INV.sub.10, the address signal lines AL.sub.0 to AL.sub.3 are set to levels that correspond to the signal levels fed to the input terminals A.sub.0 and A.sub.1.

Namely, when the terminal A.sub.0 assumes a low level and the terminal A.sub.1 assumes a low level, the address signal line AL.sub.0 assumes a high level, the address signal line AL.sub.1 assumes a low level, the address signal line AL.sub.2 assumes a high level, and the address signal line AL.sub.3 assumes a low level. The emitters of a multi-emitter transistor Q.sub.1 of a decoder driver which is connected to word line WL.sub.0 are connected to the address signal lines AL.sub.0 and AL.sub.2. Therefore, the transistor Q.sub.1 is rendered non-conductive, and a base current is supplied from a power supply V.sub.CC to the transistor Q.sub.2 via resistor R.sub.1. Accordingly, the transistor Q.sub.2 is rendered conductive, and the transistors Q.sub.3 and Q.sub.4 are also rendered conductive to select the word line WL.sub.0. In this case, a writing current of about 200 mA is supplied to the memory cell Q.sub.00, via a path BL.sub.0 -Q.sub.00 -Q.sub.4 -V.sub.GND, to short-circuit the emitter-base junction of memory cell Q.sub.00. In the writing state the base current of transistor Q.sub.4, which must sink a writing current of about 200 mA, is supplied through the transistors Q.sub.2 and Q.sub.3 of the decoder circuit.

Next, the reading operation of the circuit is described below with reference to the memory cell Q.sub.00, in which the emitter-base junction is assumed to be short-circuited as shown in FIG. 2.

Information can be read out by detecting whether or not a current flows into the transistor Q.sub.4 through the memory cell Q.sub.00. When a low level voltage is applied to the terminal A.sub.0 and terminal A.sub.1, the word line WL.sub.0 is selected, whereby the transistor Q.sub.1 is rendered non-conductive and the transistors Q.sub.2, Q.sub.3 and Q.sub.4 are rendered conductive, in the same manner as mentioned in discussing the writing state. In the reading state, there is no need for applying such a strong current as is applied in the writing state. A reading current of about 0.5 mA, which is two to three orders smaller than the writing current, flows into the bit line BL.sub.0 and word line WL.sub.0 through a path BL.sub.0 -Q.sub.00 -Q.sub.4 -V.sub.GND. In the writing state, a current of about 200 mA is sinked in transistor Q.sub.4, the collector potential of transistor Q.sub.4 is sustained high, and the diode D.sub.1 is rendered non-conductive, while, in the reading state, a small current of about 0.5 mA is sinked in transistor Q.sub.4. Therefore, the collector potential of the transistor Q.sub.4 is sustained low, and, as a result, the diode D.sub.1 is rendered conductive, whereby the base current of the transistor Q.sub.2, which is supplied from the power supply V.sub.CC via the resistor R.sub.1, is divided into a current of the emitter side and a current of the collector side. Accordingly, the base current supplied to the transistor Q.sub.3 is smaller than when information is being written down. Hence, the base current of the transistor Q.sub.4 becomes so small as to be sufficient to draw a current of about 0.5 mA from the word line WL.sub.0. In general, in the programmable read only memory in which the writing current and the reading current have such greatly different values as 200 mA and 0.5 mA, if the base current of the transistor Q.sub.4 is so great as to adequately drive the writing current of 200 mA, then in the reading state the transistor Q.sub.4 will be saturated due to excessive base current, and excessive electric charge will accumulate in the base. Therefore, the transistor Q.sub.4 takes an extended period of turn off time, and is unsuited for high-speed operation. In order to prevent the above-mentioned inconvenience, therefore, the conventional decoder circuit has been equipped with a feedback circuit which switches the base current of the transistor Q.sub.4, so that suitable base currents can be supplied to the transistor Q.sub.4 both in writing and reading states. Provision of the feedback circuit, however, makes the circuit setup complicated, and is not desirable from the standpoint of highly integrating the circuits. For example, as will be obvious from FIG. 2, the collector of the transistor Q.sub.2 and the collector of the transistor Q.sub.4 in the feedback circuit must be connected via the diode D.sub.1 over the wiring of the transistor Q.sub.3, thus making the wiring complicated. It is of course allowable to switch over the power supply V.sub.CC to control the base current of transistor Q.sub.4 during writing or reading states, instead of providing the above-mentioned feedback circuit in decoder driver circuit. However, the power supply voltage V.sub.CC must feed a voltage not only to the decoder driver circuit but also to other circuits, and thus it is undesirable to switch over the power supply voltage V.sub.CC in view of the design requirements of other circuits.

An embodiment of the present invention is described below with reference to FIG. 3.

FIG. 3 illustrates a decoder driver circuit, address signal line and a word line according to an embodiment of the present invention. The circuit of FIG. 3 is connected to address inverters and a programming circuit as is illustrated in FIG. 2.

The novel features of the circuit of the present invention are described below. In the present invention, the feedback circuit of FIG. 2, which comprises the transistor Q.sub.2 and the diode D.sub.1, is eliminated, and, instead, an external terminal EXT is connected to the collector of a pre-final transistor Q.sub.13 to control the base current of the transistor Q.sub.14 in the final stage.

In writing, a voltage which is greater than the power supply voltage V.sub.CC (where the power supply voltage is of, for example, 5 volts) is applied to the terminal EXT, whereby a base current which is great enough to sink a current of about 200 mA from the word line WL.sub.0 is fed to the transistor Q.sub.14 of the final stage via the transistor Q.sub.13. The diode D.sub.13 is provided to prevent the current applied from the terminal EXT from flowing to the voltage supply V.sub.cc.

In reading, on the other hand, an input voltage less than or equal to the supply voltage is applied to the terminal EXT. In this case, however, the transistor Q.sub.14 is driven by a base current consisting of a current which flows from the power supply V.sub.CC via R.sub.1 and Q.sub.1 plus a current which flows from the power supply V.sub.CC through point P.sub.1 via R.sub.12, D.sub.13, and Q.sub.13, so that the base current of the transistor Q.sub.14 is small enough to sink a reading current of about 0.5 mA from the word line WL.sub.0. The diode D.sub.12 is provided to prevent the collector current which is supplied by the supply voltage V.sub.cc from flowing to the terminal EXT.

Thus, by supplying a base current from an external unit to the transistor Q.sub.14 in the final stage, the base current of the transistor can be controlled both in reading and writing states, and thus, the feedback circuit (which requires complicated cross-over wiring) shown in the conventional art can be eliminated. Moreover, by supplying enough base current from the terminal EXT, there is no need of increasing the number of stages to amplify the base current, i.e., the number of stages can be reduced and the circuit can be simplified.

A circuit for supplying the current of the terminal EXT is mentioned below in detail with reference to FIG. 4. As illustrated in FIG. 4, the terminal EXT can be connected to the chip enable input CE through a voltage control circuit which consists of a diode D.sub.15 connected to the terminal CE, a series circuit consisting of a Zener diode ZD.sub.12 and a resistor R.sub.16 connected between the power supply V.sub.CC and the diode ZD.sub.12, a pnp transistor Q.sub.15 of which the emitter is connected to a connection point between the diode D.sub.15 and the Zener diode ZD.sub.12, a series circuit, consisting of a Zener diode ZD.sub.13 and a resistor R.sub.17, connected between the base of the transistor Q.sub.15 and ground, an npn transistor Q.sub.16 of which the base is connected to the collector of the transistor Q.sub.15 and of which the collector is connected to the base of the transistor Q.sub.15, a resistor R.sub.19 connected between the base and the emitter of transistor Q.sub.16, and is served with the collector voltage of the transistor Q.sub.15, a resistor R.sub.18 (preventing the voltage control circuit from malfunctioning by surge current) which is connected between the base of the transistor Q.sub.16 and ground, and a series circuit, consisting of a Zener diode ZD.sub.11 and a diode D.sub.132, connected between the power supply V.sub.CC and a connection point intermediate between the emitter of transistor Q.sub.16, the resistor R.sub.19, and the resistor R.sub.12. In the circuit of FIG. 4, the diode D.sub.15 works to prevent reverse current flow, the Zener diode ZD.sub.12 and the resistor R.sub.16 reduce the input capacitance, the series circuit of the Zener diode ZD.sub.13 and resistor R.sub.17 bias the voltage-current characteristics, and the resistor R.sub.18 prevents transistors Q.sub.15 and Q.sub.16 from being rendered conductive by a surge current.

Referring to the circuit constructed as illustrated in FIG. 4, the transistors Q.sub.15 and Q.sub.16 form an SCR replacing D.sub.12 circuit. The Zener diode ZD.sub.13 has been connected to the base of the transistor Q.sub.15. Therefore, when a voltage greater than the reference voltage of the Zener diode ZD.sub.13 is applied to the terminal CE, the transistor Q.sub.15 is rendered conductive and a collector current flows. As the collector current flows into the transistor Q.sub.15, the base potential of the transistor Q.sub.16 is, and the transistor Q.sub.16 is rendered conductive. Accordingly, the transistor Q.sub.13 supplies a base current to the transistor Q.sub.14 such that Q.sub.14 will be capable of sinking a write current of about 200 mA. In FIG. 4, a series circuit consisting of the Zener diode ZD.sub.11 and the diode D.sub.13a is equivalent to the diode D.sub.13 of FIG. 3.

According to the present invention, described in the foregoing, the decoder circuit does not require a feedback circuit for switching over the base current of the transistor in the final stage from reading to writing state and vice versa. Therefore, the circuit setup and wiring can be simplified and the mask for manufacturing the device can be simplified too, making it possible to form a semiconductor memory device having a high degree of integration.

Although the chip enable terminal was also used as the external terminal EXT in the aforementioned embodiment, it is also allowable to employ an address input terminal upon which can be impressed a voltage (of, for example, 20 volts) which is greater than that of the power supply V.sub.CC.