MOS Type Circuit Device

The present invention relates to a semiconductor integrated circuit device and, more particularly, to improvements in a circuit for applying a potential to a back gate of a MOS transistor. As is well known in the art, such as in NAND, NOR and EXOR gate, a P-channel MOS transistor in an N-well formed in a P-type substrate and an N-channel MOS transistor in a P-well formed in an N-type substrate can have source potentials different from the supply potential. Such an arrangement is also used with the MOS transistors which construct the differential amplifiers in an operational amplifier. In the conventional circuits, these MOS transistors have had their back gates (i.e., the wells insulated from the substrate) set at the same potential as the source, that is, at the highest potential in the case of P-channel MOS transistors, or at the lowest potential in the case of N-channel MOS transistors.

In a circuit in which the source potential is applied to the back gate of the MOS transistor by connecting the back gate directly to the source electrode, however, the large capacitance between the well forming the back gate and the substrate is applied to the source electrode. As a result, if the source potential fluctuates, the time period for charging or discharging this capacitance is elongated to invite a reduction in the switching speed.

In the circuit in which the back gate is connected to the source electrode, moreover, there arises another defect in that the substrate bias effect causes the threshold voltage to increase with an increase in the reverse bias voltage between the source electrode and the back gate.

Therefore, an object of the present invention is to provide a MOS type circuit having a high switching speed.

Another object of the present invention is to provide a MOS type circuit which is free from the substrate bias effect.

Yet another object of the present invention is to provide a MOS type circuit having a steady threshold potential.

A further object of the present invention is to provide a current summing type D/A converter.

According to the present invention, there is provided a MOS type circuit comprising a MOS transistor having source, drain and gate electrodes; and an amplifier having a gain of about "1", the input and the output of the amplifier being coupled to the source electrode and the back gate of the MOS transistor, respectively.

Other objects and features of the invention will be clarified by the following description, with reference to the accompanying drawings.

FIGS. 1 through 3 are circuit diagrams showing first through third embodiments of the present invention;

FIG. 4 is a circuit diagram representing a fourth embodiment of the present invention; and

FIG. 5 is a circuit diagram indicating the current cell circuit shown in FIG. 4.

FIG. 1 is a circuit diagram showing a first embodiment of the present invention. A MOS transistor Q.sub.1 is connected between a constant source I.sub.1 and a load resistor R.sub.OUT1 for switching the application of the constant current to the load resistor R.sub.OUT1 in accordance with a clock input from an input terminal C.sub.1. An amplifier A, having a gain of about "1" is provided with the source potential of the MOS transistor Q.sub.1 and is connected to the back gate of the MOS transistor Q.sub.1 at its output. The output of the amplifier is fed back to the inverting input thereof as shown in FIG. 1.

The large capacitance due to the PN junction between the back gate and substrate of the MOS transistor Q.sub.1 is connected to the low-impedance output of the amplifier A.sub.1 so that the capacitance at the source of the MOS transistor Q.sub.1 is limited to the low input gate capacitance of the amplifier A.sub.1. Since the source and the back gate of the MOS transistor Q.sub.1 are connected through an amplifier A.sub.1 having unity gain, moreover, their potentials are always kept at the same level. This eliminates any increase in the threshold voltage that might otherwise be caused by the substrate bias effect.

FIG. 2 illustrates a second embodiment of the present invention. MOS transistors Q.sub.11 and Q.sub.12 have their respective sources connected to a constant-current power supply I.sub.11 and their respective drains connected to load resistors R.sub.12 and R.sub.13. The MOS transistors Q.sub.11 and Q.sub.12 switch the constant current from the constant-current power supply I.sub.11 to the load resistors R.sub.12, R.sub.13 in accordance with the clock input or inputs from input terminals C.sub.11 and C.sub.12. A MOS transistor Q.sub.13 has its gate connected to the sources of the MOS transistors Q.sub.11 and Q.sub.12 and its source connected to the back gates of the MOS transistors Q.sub.11 and Q.sub.12 and the resistor R.sub.11. In this case, the MOS transistor Q.sub.13 acts as a source follower amplifier having a gain of about "1". As a result, the operations of the MOS transistor Q.sub.13 are the same as those of the amplifier A.sub.1 of FIG. 1.

FIG. 3 shows a third embodiment of the present invention which is obtained by replacing the P-channel transistor Q.sub.1 in the first embodiment with an N-channel MOS transistor Q.sub.21. More specifically, the N-channel MOS transistor Q.sub.21 has its gate connected to receive the clock input applied via input terminal C.sub.21 ; has its drain connected with a load resistor R.sub.OUT2 and its source connected to both a constant-current power supply I.sub.21 and the input of an amplifier A.sub.2 having a gain of about "1". The output of this amplifier A.sub.2 is applied to the back gate of the N-channel MOS transistor Q.sub.21. According to this circuit, similar operations and effects as in the first embodiment are obtainable.

FIG. 4 illustrates the fourth embodiment of the present invention, which is a current summing type D/A converter.

Digital data in the form of 8 bit words DB.sub.0 -DB.sub.7 from input terminals 10-17 are fed to an input buffer 23. Of the digital data, the least significant bit (LSB) DB.sub.7 and the 4th through 7th bits, DB.sub.3 through DB.sub.6 are supplied to a first latch circuit 27 via a first decoder 25. The first decoder 25 generates and applies first control signals C.sub.1 -C.sub.16 to the first latch circuit 27 on the basis of the bits DB.sub.3 -DB.sub.6. The most significant digit (MSB) DB.sub.0, the 2nd bit DB.sub.1 and the 3rd bit DB.sub.2 are supplied to a second decoder 24. The second decoder 24 generates and applies second control signals A.sub.1 -A.sub.8 and third control signals B.sub.1 -B.sub.8 to a second latch circuit 26. The latch circuits 26 and 27 latch the first through third control signals and the bit DB.sub.7 in response to a sampling pulse SP supplied from a terminal 18.

CC.sub.01 through CC.sub.716 are current cells arranged in matrix form. Each current cell has first through third control input terminals CI.sub.1 through CI.sub.3, first and second output terminals O and O and a bias input terminal BI connected to a bias circuit 28. The details of the current cell circuit construction will be explained referring to FIG. 5.

The first control terminals CI.sub.1 of those current cells arranged in the same column are connected in common and the first control signals C.sub.1 -C.sub.16 from the first latch circuit 27 are supplied to the cells in respective columns, as shown. The second and third control terminals CI.sub.2 and CI.sub.3 of cells arranged in the same row or line are respectively connected in common, and the second and third control signals A (A.sub.1 -A.sub.8) and B(B.sub.1 -B.sub.8) from the second latch circuit 26 are fed thereto, respectively, as shown. The first output terminals O, the second output terminals O and the bias input terminals BI of the current cells CC.sub.01 through CC.sub.716 are respectively commonly connected with a first current output terminal 21, a second current output terminal 22 and an output terminal of the bias circuit 28. It is noted here that the current capacities of the current cells CC.sub.01 through CC.sub.716 are substantially the same. Each of the current cells CC.sub.01 through CC.sub.716 generates a current from its first output terminal O when its first and second control signals C.sub.(n), A.sub.(n) are at the "HIGH" level or when the third control signal B is at the "HIGH" level, and generates a current from its second output terminal O otherwise.

A current cell CC.sub.00, which has half the current capacity of the other current cells, generates a current from its first output terminal O and from its second output terminal O when the bit DB.sub.7 takes the values "1" and "0", respectively. The first and second output terminals O and O of the cell CC.sub.00 are connected in common with outputs 21, 22, respectively.

The first decoder 25 generates a decoded signal, as shown in table 1, which determines the levels of the first control signals C.sub.1 through C.sub.16, responsive to the bits DB.sub.3 through DB.sub.6.

where "H" and "L" are "HIGH" and "LOW" levels, respectively.

The second decoder 24 generates a decoded signal, as shown in Table 2, determining the levels of the second control signals A.sub.1 -A.sub.8 and the third control signals B.sub.1 -B.sub.8.

As a result, the output current I at the first current output terminal 21 increases stepwise by the current capacity of the current cell CC.sub.00 with each increase by "1" of the input digital data DB.sub.0 through DB.sub.7. On the other hand, the current I. from the second current output terminal incrementally decreases at the same rate.

FIG. 5 represents a detailed circuit diagram of each of the current cells CC.sub.01 through CC.sub.716. Transistors Q.sub.33 through Q.sub.35 are sequentially connected between the power source V.sub.DD and the ground level GND. The third control signal B is applied to the gates of the transistors Q.sub.33 and Q.sub.35 through the third input terminal CI.sub.3. Between the drain of transistor Q.sub.33 and ground are connected transistors Q.sub.30 through Q.sub.32. The first control signal C is applied to the gates of the transistors Q.sub.30, Q.sub.31 through the first input terminal CI.sub.1. To the gates of the transistors Q.sub.32, Q.sub.34 the second control signal A is fed through the second input terminal CI.sub.2. With this circuit configuration, a "LOW" level signal appears at the connection point of the transistors Q.sub.34, Q.sub.35 when both the first and second control signals C.sub.(n), A.sub.(n) are "HIGH", or when the second control signal B is "HIGH".

The signal appearing at this connection point is inverted in an inverter composed of transistors Q.sub.36, Q.sub.37, and the inverted signal is applied to the gate of transistors Q.sub.38, Q.sub.41. A transistor Q.sub.40 is inserted between the transistors Q.sub.38 and Q.sub.41, and a transistor Q.sub.39 is connected to the transistor Q.sub.38 in parallel. A transistor Q.sub.42, having a gate supplied with a bias voltage through the bias terminal B.sub.I, operates as a constant current source. Condensors C.sub.1 and C.sub.2 are used as filters.

Transistors Q.sub.43, Q.sub.47 have their sources connected to the drain of the transistor Q.sub.42 and their back gates supplied with the source voltage through a source follower amplifier of unity gain comprised of transistors Q.sub.44 and Q.sub.45 in accordance with the present invention, to prevent the threshold increase of the transistors Q.sub.43 and Q.sub.47. The gates of the transistors Q.sub.43 and Q.sub.47 are connected to the connection points between the transistors Q.sub.39 and Q.sub.40 and between the transistors Q.sub.46 and Q.sub.48. To the gates of the transistors Q.sub.39 and Q.sub.40, the source voltage of the transistors Q.sub.43 and Q.sub.47 is applied. Accordingly, if a "LOW" signal appears at the connection point of the transistors Q.sub.34 and Q.sub.35, the gate voltage of the transistor Q.sub.43 is lower than that of the transistor Q.sub.47, generating a constant current from the transistor Q.sub.42 at the first output terminal O through the transistor Q.sub.43. In contrast, the current output appears at the second output terminal O through the transistor Q.sub.47, if the signal at the connection point of the transistors Q.sub.34 and Q.sub.35 is at the "HIGH" level.

As will be clearly understood, since the source electrode of the MOS transistor is not connected to the back gate, the large capacitance at the junction between the back gate region and the substrate is not applied to the source electrode. This makes it possible to obtain high speed operation. The input capacitance of the amplifier is applied to the source electrode, but this capacitance is much smaller than that at the junction.

Since the source electrode and the back gate region of the MOS transistor are kept at the same potential through an amplifier having a gain of about "1", moreover, the so-called "substrate bias effect" is not caused and the threshold voltage is not varied.