Programmable Logic Array for Carrying Out Logic Operations of Binary Input Signals

This invention relates generally to programmable circuits for performing logic operations and, in particular, to a programmable circuit having a matrix with crossed first and second sets of lines connected through a respective resistor to an operating voltage source and also having logic units connected at intersections between the first and second lines.

Various types of programmable circuits for implementing logic operations are known in the prior art (see, for example, Weiss, Horninger, Integrierte MOS Schaultungen, Springer Verlag 1982, pp. 295 through 298). These programmable circuits typically contain an AND level of circuitry and an OR level of circuitry. Input signals are received by the programmable circuit in the AND level and are stored in the AND level in accordance with a function table. The operation results of the AND level, also referred to as product terms, are supplied to the OR level and are operated thereon in accord with the function table contained in the OR level to form what are referred to as sum terms. The sum terms are the output of the OR level and also the output of the programmable circuit. Therefore, such a prior art programmable circuit has input lines which lead to the AND level and has output lines which lead from the OR level.

The present invention overcomes the restrictions of the prior art that certain lines are always input lines and other lines are always output lines.

It is therefore an object of the present invention to provide a programmable circuit such that it can be bidirectionally operated, that is, the lines connected to the programmable circuit can be either input lines or output lines.

The novel programmable circuit performs logic operations on binary input signals. The programmable circuit has a matrix with crossed first and second sets of lines connected through a respective resistor to an operating voltage source and also has logic units connected at intersections between the first and second lines. The first set of lines are paired to form line pairs. The logic units are bidirectionally operable and are connected between each of the second lines and a selected one of the lines of the line pair at each of the intersections. A bidirectional operable driver circuit is connected to an end of one of the first lines of the line pair and opposed to the connection of that first line to the voltage source. Similarly, a driver stage is connected to an end of the other of the first lines of the line pair and opposed to the connection of that first line to the voltage source.

The programmable circuit is therefore no longer composed of a separately executed AND level and an OR level as found in the prior art. On the contrary, it is composed of a matrix of crossed first and second lines and has logic units arranged at the intersections between the first and second lines in accordance with a function table. These logic units are capable of being bidirectionally operated. The first lines can be employed as inputs or as outputs in accordance with the setting of the logic units and of the driver circuit. When first lines are used as inputs, then the input signals supplied to these inputs are transmitted via logic units to the second lines and are subjected to an AND operation on the second lines. When the first lines are employed as outputs, as a result of the setting of the logic units and of the driver circuit, they carry out a NOR operation on the signals which had been transmitted via the logic units from the second lines onto the first lines.

The programmable circuit thus has the advantages that logic operations on input signals can be carried out on a single matrix and that it is possible to utilize the first lines both as inputs for the input signals, as well as, outputs of the operational results. A further advantage of the programmable circuit is the determination of whether the input bit patterns applied to the programmable circuit are in accordance with the realized function table and whether the outputs as a result of the setting of the logic units are a correct function result in accordance with that function table. These determinations are made possible by utilizing a testing network.

The features of the present invention which are believed to be novel are set forth with particularity in the appended claims. The invention, together with further objects and advantages, may best be understood by reference to the following description taken in conjunction with accompanying drawings, in the several figures of which like reference numeral identify like elements, and in which:

FIG. 1 is a general block diagram of the programmable circuit;

FIG. 2 is a general circuit diagram of a matrix of the programmable circuit;

FIG. 3 is a circuit diagram depicting a logic unit used in the FIG. 2 circuit;

FIG. 4 is a circuit diagram of an alternative embodiment of the logic unit in the FIG. 2 circuit diagram;

FIG. 5 is a preferred embodiment of a first switching element of the logic unit;

FIG. 6 is a circuit diagram of a second switching element of the logic unit;

FIG. 7 is a schematic illustration of a portion of the matrix depicting first lines used as inputs;

FIG. 8 is a schematic illustration of a portion of the matrix depicting first lines used as outputs;

FIG. 9 is a schematic illustration showing an example of programming the matrix;

FIG. 10 is a schematic illustration depicting the circuit arrangement for determining whether the input bit pattern applied to the matrix is contained in a function table stored in the matrix;

FIG. 11 is a schematic illustration of the matrix depicting the matrix utilized as an adder;

FIGS. 12 and 13 are schematic illustrations depicting two examples of programming the matrix as an adder and operating the matrix by setting logic units according to the function table; and

FIG. 14 is a general block diagram depicting a testing network for use with the matrix.

The present invention has general applicability, but is most advantageeously utilized as a programmable circuit for performing logic operations. In general terms, the programmable circuit has a matrix which has a plurality of first lines crossing a plurality of second lines to define a plurality of intersections. The first lines are grouped by twos to form a plurality of line pairs and each line pair has an A first line and a B first line. Logic units are bidirectionally operable and connect each of the first lines to each of the second lines and include in series a switch and a disconnecting mechanism. A plurality of bidirectional driver circuits are connected to first ends of the A first lines of the line pairs. A plurality of driver stages are connected to first ends of the B first lines of the line pairs. A voltage supply is connected to the second ends of the first and second lines via a respective resistor. For performing a logic operation, selected bidirectional logic units are deactivated by their associated disconnecting mechanisms such that each of the second lines is connected by an operational logic unit to either the A first line or the B first line of each of the line pairs.

Referring now to FIG. 1, a general block diagram of the programmable circuit BPLA, the programmable circuit BPLA is composed of a matrix MA and may also including a testing network PRN. The matrix MA is connected to data lines D.sub.0 through D.sub.n which can be bidirectionally operated. A first set of control signals EN.sub.0 through EN.sub.n determine whether the data lines D.sub.0 through D.sub.n are utilized as input lines. A second set of control signals AS.sub.0 through AS.sub.n determine whether the data lines D.sub.0 through D.sub.n are utilized as output lines. The matrix MA outputs test signals P.sub.0 through P.sub.n which, together with the first control signals AS.sub.0 through AS.sub.n and the signals on the data lines D.sub.0 through D.sub.n are supplied to the testing network PRN. By comparison of the test signals P.sub.0 through P.sub.n to the signals appearing on the data lines D.sub.0 through D.sub.n, when these data lines are used as output lines, the testing network PRN can determine whether the matrix MA is being utilized in accordance with a function table stored in the matrix. The significance of the signals F1, F2 and NB.sub.0 through NB.sub.n will be described below.

Referring now to FIG. 2, a complete matrix is depicted which has not yet been programmed in accordance with a function table. It contains first lines LH, second lines LV, logic units G and GN, arranged at the intersections between the first lines LH and the second lines LV, driver circuits EA, and driver stages E. All first lines LH and all second lines LV are connected to an operating voltage source VCC via a respective resistor R.

The two first lines LH are combined to form a line pair LP. Succeeding pairs of lines LH are combined to form other line pairs. The matrix MA contains n+1 line pairs LP.sub.n. A first driver circuit EA is connected between a data line D and one first line of the allocated line pair LP. Driver stage E is connected between the data line D and the other first line of the line pair LP. For example, the data line D.sub.0 is connected to the first line LH.sub.01 via the driver circuit EA.sub.0 and is connected to the first line LH.sub.02 via the driver stage E.sub.0.

In the depiction of FIG. 2, m+1 second lines LV.sub.0 through LV.sub.m are provided, logic units G, GN, which are each composed of two switch elements SC1 and SC2 in the preferred embodiment, lie between the first lines LH and the second lines LV. Utilizing the first control signals EN, the first switch elements SC1 can be switched off; and utilizing the second control signals AS, the second switch elements SC2 can be switched off. In the FIG. 2 embodiment, the first switch elements SC1 are shown as non-inverting switch elements and the switch elements SC2 are shown as inverting switch elements. The arrangement of the switch elements SC1 and SC2 is referred to as a connection in antiparallel fashion.

All logic units are identically operated, however, the logic units which are connected to the first line are referenced G and the logic units which are connected to the other first line of a line pair are referenced GN. The difference in designation will be explained in reference to FIG. 9.

One embodiment of the first switch element SC1 is shown in FIG. 5. It is composed, for example, of four npn transistors T1 through T4. The first control signal EN is applied to the base of transistor T3. When it is present, that is, when the switch element is to be conductive for data signal DE, then the transistor T3 is activated and the transistor T4 is inhibited thereby. The switch element is thus not influenced by the first control signal EN. A data signal DE at the transistor T1 is thus transmitted to the output DA. The transistor T2 is realized in open-collector circuitry. Thus, when a binary "0", is applied at the input DE, the transistor T2 is activated and pulls the allocated second line LV to "0" potential. By contrast, when a binary "0" is applied at the input DE, then the transistor T2 is inhibited and the potential on the allocated, second line LV is not influenced via the switch element SC1. Similarily, the potential on the allocated, second line LV is not influenced when the first control signal EN is binary "0". When the first control signal EN is binary "0" the transistor T4 is activated thus inhibiting the transistor T2.

FIG. 6 schematically depicts an embodiment of the second switch element SC2. It is composed of the transistors T5, T6 and T7 and of the resistor R1. The second control signal AS is applied to the base of the transistor T6. In this case, the transistor T6 is activated and the transistor T7 is inhibited. The signal at the input DE can thus act on the transistor T5 via the resistor R1. When the signal at the input DE is binary "1" then the transistor T5 is activated and pulls the potential on the allocated, second line LV to ground or zero potential. When the signal at the input DE is binary "0", then the transistor T5 remains inhibited and the potential on the allocated, second line is not influenced. When the second control signal AS is binary "0", then the transistor T6 is inhibited but the transistor T7 is activated. This results in the transistor T5 being inhibited and, thus, the second control signal AS cannot influence the potential on the allocated, second line LV. Here, too, the output stage of the switch element SC2 is realized as an open-collector circuit.

The driver circuit EA connected at the input of each of the first lines LH is constructed in accordance with the logic units G. It can likewise be switched with the assistance of the control signals EN and AS. When, thus, the first control signal EN is binary "1", then the driver stage ST1 is connected and the driver stage ST2 is inhibited. When, by contrast, the second control signal AS is utilized, then the driver stage ST2 is operational and the driver stage ST1, by contrast, is inhibited.

FIG. 7 shows the electrical operation of the matrix when the first lines LH.sub.z are employed as input lines and, thus, D.sub.z is an input. In this case, all second switch elements SC2 are to be considered as being non-existent and only the first switch elements SC1 are effective. Likewise, the first driver stage ST1 of the driver circuit EA.sub.z is effective but the second driver stage ST2 is not.

FIG. 8 illustrates the opposite situation, wherein D.sub.z is an output. Accordingly, the second control signal AS.sub.z is utilized and the first switch elements SC1 are now not effective but the second switch elements SC2 are. Similarly, the second driver stage ST2 of the driver circuit EA.sub.z is effective but the driver stage ST1 is not. Furthermore, the driver stage E.sub.z is also not effective.

It can be shown from FIG. 9 how the matrix is programmed in accordance with a function table. When a binary "1" is to be programmed at an intersection between the line pair LP.sub.z and a second line, for example LV.sub.1, then the logic unit G.sub.z1 lying between the first line LH.sub.z1 and the second line LV.sub.1 is utilized. By contrast, the logic unit Gn.sub.z1 lying between the other first line LH.sub.z2 and the second line LV.sub.1 remains non-functioning. When, by contrast, a binary "0" is to be programmed, for example between the line pair LP.sub.z and the second line LV.sub.2, then the logic unit G.sub.z2 connected between the first line LH.sub.z1 and the second line LV.sub.2 remains non-functioning but the logic unit Gn.sub.z2 connected between the first line LH.sub.z2 and the second line LV.sub.2 is utilized. When, thus, a binary "1" is to be programmed at an intersection between a line pair LP and a second line LV, then a logic unit G is connected between one of the first lines of the line pair and the second line LV; when, by contrast, a binary "0" is to be programmed, then the logic unit GN is arranged between the other first line of the line pair and the second line LV. The logic units which are provided for programming the binary "1" are referenced G and those employed for programming the binary "0" are referenced GN.

The programming of the logic units can be carried out with the assistance of one or two line levels LT. The two switch elements SC1 and SC2 must then be respectively wired or not wired between the levels. FIG. 3 shows an example of this situation. The locations of the logic unit G or, respectively, GN are indicated with arrows and can be present or not be present.

Further, a possible realization of the programming with the assistance of the logic units can be achieved in accordance with FIG. 4. Here, fuses SI are inserted between every switch element SC1, SC2 and an allocated first or second line and these fuses are capable of being parted, thus eliminating the switch element from the circuit. When the logic unit is to represent a binary "1" or a binary "0", then the fuses are electrically continuous and connect the switch element in the circuit; otherwise, the fuse element is parted.

The programming of the matrix MA as an adder is shown in FIG. 11. In this depiction, D.sub.0 and D.sub.1 have been selected as inputs and D.sub.2 and D.sub.3 have been selected as outputs. The value for the sum is outputted at the output D.sub.2 and the value for the overflow is outputted at the output D.sub.3. The arrangement of the logic units in the matrix corresponds to a function table that is established given a modulo-2 addition of two input values. For example, the second line LV.sub.0 in combination with the allocated, first lines, shows the programming for D.sub.0 ="0" and D.sub.1 ="0". The sum is then D.sub.2 ="0" and the carry is D.sub.3 ="0". The second line LV.sub.3 shows the relationships when a binary "1" is applied to the inputs D.sub.0 and D.sub.1. The sum at the output D.sub.2 is then D.sub.2 ="0" and the carry at the output D.sub.3 is D.sub.3 ="1". The logic units connected to the second lines LV.sub.1 and LV.sub.2 correspond to the two other combinations possible for input signals at the inputs D.sub.0 and D.sub.1 and to the resulting output signals at the sum output D.sub.2 and at the carry output D.sub.3.

In order to utilize D.sub.0 and D.sub.1 as inputs, the first control signals EN and EN.sub.1 must be applied to the matrix on line pairs LP.sub.0 and LP.sub.1, respectively. In the logic units connected to the line pairs LP.sub.0 and LP.sub.1, the first switch elements SC1 are effective, but the second switch elements SC2 are not effective. The situation is different at the outputs D.sub.2 and D.sub.3. Here, the second control signals AS.sub.2 and AS.sub.3 are applied to the line pairs LP.sub.2 and LP.sub.3, respectively, and, thus, the second switch elements SC2 are respectively effective in the line pair LP.sub.3 and LP.sub.2, but the first switch elements SC1 are not effective. Of the driver circuits EA, the driver stages ST1 are effective for the line pair LP.sub.0 and LP.sub.1 and the driver stages ST2 are effective for the line pairs LP.sub.2 and LP.sub.3.

As an example of the functioning of the adder depicted in FIG. 11, a binary "1" is applied to both inputs D.sub.0 and D.sub.1. A binary "1" then occurs on the first lines LH.sub.01 and LH.sub.11, by contrast, a binary "0" occurs on the other first lines LH.sub.02 and LH.sub.12. Since the first control signals EN.sub.0 and EN.sub.1 are applied to line pairs LP.sub.0 and LP.sub.1, only the first switch elements in the line pairs LP.sub.0 and LP.sub.1 are effective. Accordingly, the logic units GN.sub.00 and GN.sub.01 apply a binary "0" to the second lines LV.sub.0 and LV.sub.1 just as the logic units GN.sub.10 and GN.sub.12 apply a binary "0" to the second line LV.sub.0 and LV.sub.2. By contrast, a binary "1" is applied to the second line LV.sub.3 via the logic units G.sub.03 and G.sub.13. It may be seen that a respective AND operation of the binary values supplied to the inputs D.sub.0 and D.sub.1 is carried out on the second lines LV.

As a result of the second control signals AS.sub.2 and AS.sub.3, the line pairs LP.sub.2 and LP.sub.3 operate as outputs. Accordingly, the second switch elements SC2 are effective, as is the second driver stage ST2. The binary signal "1" on the second line LV.sub.3 places the line LH.sub.22 to binary "0" via the logic unit GN.sub.23 and places the line LH.sub.3 1 to binary "0" via the logic unit G.sub.33. The binary "0" on the second lines LV.sub.0, LV.sub.1, LV.sub.2 act on the lines LH.sub.2 1 and LH.sub.32 via the logic units G.sub.2 1, G.sub.22, GN.sub.30, GN.sub.3 1, GN.sub.32 and generate a binary "1" thereon. It may now be seen that a respective NOR operation of the binary values coupled in from the second lines LV via the logic units appear on the first lines LH. The correct values are then obtained at the outputs D.sub.2 and D.sub.3 via the second, inverting driver stage ST2.

FIG. 12 shows the matrix programmed as an adder with D.sub.1, D.sub.2 and D.sub.3 operating as inputs and D.sub.0 operating as an output. This is effected by utilizing the second control signal AS.sub.0 and the first control signals EN.sub.1, EN.sub.2 and EN.sub.3. In order to clearly illustrate the functioning of the matrix, only the respective switch elements which are effective due to the control signals AS.sub.0, EN.sub.1, EN.sub.2 and EN.sub.3 are shown in FIG. 12. The other switch elements are not shown although they are actually present in the circuit although non-conductive. The binary values indicated in FIG. 12 are applied to the inputs D.sub.1, D.sub.2 and D.sub.3. The binary values shown in the FIG. 12 are then derived on the second lines LV.sub.0 through LV.sub.3. They are all binary "0". The first lines of the line pair LP.sub.0, operating as an output, then carry a binary "1". After inversion by the inverting driver stage, a binary "0" is then applied to the output D.sub.0. The lines LV.sub.0 through LV.sub.3 are binary "0". An operation of the matrix which does not correspond to the function table is present. This is determined by means of the circuit shown in FIG. 10 and will be explained in detail below.

In FIG. 13, the matrix is programmed as an adder and is operated with D.sub.1 as input and D.sub.0, D.sub.2, D.sub.3 as outputs. The first control signal EN.sub.1 is now applied at the line pair LP.sub.1 and the second control signals AS.sub.0, AS.sub.2, AS.sub.3 are applied at the line pairs LP.sub.0, LP.sub.2, LP.sub.3. Only the switch elements of the logic units which are effective due to the control signals are again shown. When a binary "0" is applied at the input D.sub.1, then the binary values shown in FIG. 13 derive on the second lines LV.sub.0 through LV.sub.3. The feedback of these binary values onto the line pair LP.sub.0 results in the fact that a binary "0" appears at both first lines. This output D.sub.0 can thus not be determined. The analogous case applies to the line pair LP.sub.2. The allocated output D.sub.2 is likewise not determinable. A binary "1" appears only on one line of the line pair LP.sub.3. A binary "0" appears on the other line of line pair LP.sub.3. This binary value pair corresponds to the function table contained in the matrix. The output D.sub.3 can thus be determined.

The line pairs which operate as outputs and which carry identical binary values can be identified by checking the values at the outputs D and the values of the test signals P with the assistance of the testing network shown in FIG. 14. It was found that the outputs D.sub.0 and D.sub.2 cannot be determined from the FIG. 13 example. A checking of the function of the matrix can be carried out with the assistance of the testing network of FIG. 14. To that end, every line pair LP.sub.0 through LP.sub.n has a respective AND element, UG.sub.0 through UG.sub.n, allocated to it. The signal on the data line D, the test signal P and the second control signal AS are supplied to every AND element UG. The AND element outputs an error signal NB at its output when the signal on the data line D is binary "1" and the test signal P is binary "0". In this case, this output cannot be determined. The output signals of the AND elements UG.sub.0 through UG.sub.n are combined to form an error signal F1, via an OR element OD.sub.2.

Whether a second line exhibits the binary value "1" can be determined with the assistance of the OR circuit OD.sub.1 of FIG. 10 by means of which the second lines LV.sub.0 through LV.sub.m are combined. One second line carries the value binary "1" when the input bit pattern applied to the inputs is present in the function table. A signal F2 can be derived therefrom. When this signal F2 is "0", then the applied bit pattern is not contained in the function table.

Thus, every type of logic function can be realized with the matrix. The requirement is the realization of the function table underlying the logic function in the matrix. It is also possible to utilize a permanently programmed matrix and to select the number of inputs or, respectively, outputs and to check the output values by applying first and second control signals.

Advantageous the testing network can be utilized to identify whether an applied input bit pattern leads to a defineable output value or not.

In the preferred embodiments, the logic units have been depicted as two switch elements. It is also envisioned that the logic units can be bidirectionally operated single switch elements.

The invention is not limited to the particular details of the apparatus depicted and other modifications and applications are contemplated. Certain other changes may be made in the above described apparatus without departing from the true spirit and scope of the invention herein involved. It is intended, therefore, that the subject matter of the above depiction shall be interpreted as illustrative and not in a limiting sense.