DE2537075C3 - Coupling matrix produced using integrated technology - Google Patents

Description

The invention relates to a coupling matrix with bistable, a control electrode, produced using integrated technology and two crosspoints having different main line connections, their Main routes both by control on their control electrode from their second, conductive in their primary

2nd , non-conductive state as well as by control on their same control electrode from their first to their second Ztitand can be controlled and their main lines are used to switch through and disconnect connection paths, the control electrodes of the crosspoints each with the output of an individually assigned to them, in turn from the line control signal and column control signal controlled logic circuit produced in integrated technology are connected and thus controlled by the row control signal and column control signal, for telecommunications, in particular for telephone switching systems.

Such a coupling matrix is already described in DE-OS 2348414, with dor * the coupling points each contain a non-bistable transmission gate made of two complementary field effect transistors and wherein the control electrodes of the crosspoint transmission gates for generating the bistability are each connected to a flip-flop.

The linking function of the link is selected here so that at the logic element output only when a line control signal is applied at the same time and a column control signal, a third signal, the so-called adjustment signal, for coupling point is forwarded, whereby depending on the level of the setting signal, the crosspoint in question either is controlled in its first, non-conductive state or in its second, conductive state. Both of these cases are therefore coincident

', ■; the same row control signal and the same column control signal at. If, on the other hand, there is only this row control signal or only this column control signal or neither of the two control signals is present. then the relevant crosspoint remains in its previous first or second state, regardless of what the setting signal is. Regarding of the row control signal and the column control signal, this logic element has a pure AND function on. A special type of separation called “cleaning” in switching technology, that is a simultaneous control of all connected to a row line and / or all connected to a column line Crosspoints that are not at the intersection of the

matching column line and line line are in | :, their first, non-conductive state, is with a so

γ is not possible - on the

ε »cleaning«, which is necessary in case you have finished

transmission of information does not immediately interrupt the connected connection path and that

• z. B. may be necessary to address an urgent,

' Establishing a legitimate connection with the simultaneous interruption of disruptive, less authorized, existing connections will be discussed further below

"f received.

In the case of non-electronic, namely non-integrable bistable relay crosspoints, it is known that they can intentionally or unintentionally retain their second, conductive state, although the connection path originally established via these crosspoints is no longer required. If you set up a new connection path, then not only the> o coupling point located at the selected crossing point of the column line and row line of the matrix has to be controlled into its second, conductive state. In addition, every further coupling point that is directly connected to the same selected row line and to the same selected column line is to be controlled in its first, non-conductive state, because otherwise undesired double connections to other, already established connection paths or to others, only later connection paths to be established are possible. In order to avoid such double connections, it is therefore necessary to "clean" them, ie all of them are connected to the same selected row line and column line. Crosspoints that are not at the intersection of the selected row lines and column lines to control prior to building or during the construction r> of the compound in its non-conducting state. This is dealt with in DE-ASn 1 213006 and 1267 725, according to which relays must have special, non-integrable coils in order to achieve this task.

You can clean each individual, in the first, non-conductive state to be controlled coupling point on the z. B. attached parallel to the matrix column line, associated column control line and via the corresponding, z. B. v, parallel to the matrix row line brought in, control associated Zeüensteuerleitung for itself and thereby steer this crosspoint into the first, non-conductive state. However, such a procedure would be time consuming.

Numerous publications are bistable

Koppelp'.mkte known in coupling matrices through Thyristors are formed, in turn by an AND logic element and thus by coincident Row control signals and column control signals are controlled in their second, conductive state.

see e.g. DE-AS 1298188. But even with this coupling matrix, cleaning is not yet planned and not possible according to the link function there. In addition, thyristors are generally e> o not at all via their control electrode in their

] ■: 'most non-conductive state controllable. Accordingly

Accordingly, DE-AS 1198188, column 5, lines 1

to 3 assumed crosspoints that do not have

j their control electrode, but solely via the main

; Streckenanschlüssc controllable in their first state

I, are. Take the measures according to the invention

but presuppose that the coupling points have their tax * Electrodes in their first, and via the control electrodes but also in their second Zuütand controllable Are. By the way, you can't either, z. B. by interrupting the anode-cathode current in the DE-AS 1298188 specified crosspoints, all at the selected column line and line line connected crosspoints, d. H. Cleaning thyristors, so control in their first, non-conductive state, and at the same time during the same control process the am Connect through the crossing point of these two lines lying crosspoint, d. H. in his second, Control the conductive state, but this is possible with the invention.

The object of the present invention is without the use of coils, namely with integrable coils Means to enable quick cleaning by all with a selected line or Column line of the switching matrix connected crosspoints by control with a row control signal or the column control signal is initially controlled in the first, non-conductive state.

The invention is based on the introduction in DE-OS 2348414 and in the preamble of the »claim 1 specified coupling matrix.

The object of the invention is given in this coupling matrix by the specified in claim 1 Measures resolved.

By the coincidence of the row control signal fed to the row control line and that of the column control line The column control signal supplied to the column control line may be the one at the intersection of the selected line line and the selected column line crosspoint are controlled in its second conductive state, all other crosspoints connected to the line and column lines at the same time cleaned, i.e. controlled in their first, non-conductive state. The invention thus allows how will be described later, simultaneously in a single control process both on the Zeüenleitung as ach to clean the column line and at the same time also the one at the intersection of these lines To switch through the crosspoint. The crosspoints are in integrated technology together with the links can be produced on a single integrated module.

The invention is described in more detail with reference to FIGS. 1 and 2, which show special embodiments of the invention for two-wire connection paths. Fig. 1 largely the embodiment shown in Fig. 9 of DE-OS 2348414 the switching matrix with the Zeüen'.eitungen a, b ... and the corresponding SpaltfMiltitungen al, fei ... Incidentally, in Figs. 1 JNCI i the same reference symbols and the same circuit symbols promoting clarity are used that are already used in DE-OS 2 3 ^ 8414.

The two coupling points KP , formed here by an s, ich non-bistable, integrable transmission gate of complementary field effect transistors, are, as described in DE-OS 2348414, each steered by deft fiip-flops k / k ' via their control electrodes, whereby the coupling points The coupling points KP thus assume their first, non-conductive state or their second, conductive state depending on the respective state of the flip-flop k / k ' , with each coupling point or each coupling point pair being individually assigned to a flip-flop.

is ordered.

Fig. 1 differs from Fig. 9 of DE-OS 2348414 essentially only by the logic circuit G1 / G2, which does not have the switch ■ Se the steering of the two connected to the flip ^ FIops k / k ' coupling points KP of the two-wire Allow connecting path stage, but also the cleaning of this coupling point pair KP connected to the associated matrix column lines a, b and matrix row lines al, bl.

In each of the flip-flops of the coupling matrix, individually assigned logic circuits, here Gl / G2, are attached, which in turn are controlled by the row control signal Z and the column control signal S. Such a flip-flop k / k 'is inserted between each logic circuit and the assigned input. The inputs of the logic circuits are connected to the assigned row control line Z and the assigned column control line 5, see FIG. 1. The logic element G7> and the input EN of the logic circuit are provisionally neglected. The logic circuits G1 / G2 are each designed so that their two outputs control the assigned flip-flop kk ' via the switch 5c in three different ways:

If the row control signal Z fed via the row control line and, at the same time, the column control signal 5 fed via the column control line each have their "1" state, i.e. when controlled by both the row control signal and the column control signal, the NOR gate Gl of the logic circuit Eq
G2 the assigned flip-flop k / k ' in its second state, in which this flip-flop directs the coupling point KP into its second, conductive state. In this case, the coupling point KP is switched through.

If, however, the logic circuit G1 / G2 neither the row control signal Z nor the column control signal 5 is supplied, so if these two signals have the "O" state at the same time, then influenced

rtt.IlIt-I VJC-I UClUCIl nUStiUllKC VJC JV Ci IVH ULJl Uiltludl.! IuI ~ tung Gl, G2 the state of the assigned flip-flop k / k '. This flip-flop is then left in its first or in its second state, in which it was previously. The connected coupling points KP are thus also left in their state. If a crosspoint is controlled neither via the row control line nor via the column control line, then this crosspoint is not influenced, but rather is left in its first or second state. Connection paths that are established via this coupling point, which is then left in its second, conductive state, are therefore not disturbed as long as a control signal S or Z is not supplied via the assigned row control line or the assigned column control line.

If the logic circuit Gl, Gl is supplied exclusively either via the row control line or via the column control line in a "1" state corresponding control signal, but no such control signal is supplied via the other control line, but only a "signal" corresponding to an "0" state then this logic circuit Gl, G2 located at the intersection of this row control line and column control line controls the associated flip-flop k / k ' into its first state, in which this flip-flop directs each of the two connected coupling points KP into their first, non-conductive state . If a row control signal is fed to a single selected row control line and a column control signal is not fed to a single column control line of the same coupling matrix at the same time, then all of the flip-flops assigned to this row control line are controlled in their first state and thus

i "simultaneously steered all crosspoints KP connected to the respective row line parallel to the first, non-conducting state. There are thus in this case, all steerable via the respective row control line coupling points KP this case at the same

c · cleaned early.

In a corresponding manner, all crosspoints that can be steered via a column control line can be cleaned. To do this, a “!” Control signal is applied to the relevant column line and, at the same time, a “1” control signal is not applied to any of the row control lines. In this case, all of the flip-flops that are connected to the relevant column control line are controlled into their first state, and thus at the same time all of the crosspoints KP connected to the relevant, parallel column control line are directed to their first, conductive state. In this case, the crosspoints of the relevant splitter line are cleaned.

To connect a connection via the

ίο Switching network could initially only be supplied with a "1" control signal S or Z to the selected column control line (or only the selected row control line) in a first process, but only a "0" control signal could be supplied to the remaining control lines of the switching matrix. As a result, the crosspoints of the relevant parallel row line or parallel column line of the coupling matrix are cleaned. Following this, the crosspoints of the selected column line (or the selected row line) of the coupling matrix could be cleaned in a second process by just adding a "1" control to the relevant control line : 1 "7 / 1r \ £ 'itt: j : i /

l. Ill UlllV / tll UlllLt ^ ll

after brushing both to the selected column line and the devices connected to the selected row line coupling points of the mounted at the intersection of the selected column line and the selected row line of the switching matrix cross-point CP could be directed in its second conducting state to the communication path via the switching matrix to establish by applying coincident "1" control signals Z and S via the relevant control lines.

Remarkably, however, you can use the one shown Invention to build a connection path much faster. You don't need to be first two cleaning processes and the third then the actual switching process. You can with the Invention synchronously the selected crosspoint

so switch through immediately without needing two cleaning processes beforehand. By simultaneously supplying a "1" control signal S and Z, the selected crosspoint in question is immediately switched to its second, conductive state - at the same time, all other crosspoints KP connected to the same selected column line al, bl and to the same selected row line a , b are connected to the coupling matrix, cleaned, i.e. in their first state

steered, because the these cleaned crosspoints each individually assigned logic circuits each exclusively either a row control signal Z = "1" or a column control signal 5 = "1", but not both control signals are supplied coincidentally. Remarkably, therefore, they are superfluous special cleaning processes in addition to the peeling process.

The structure of the coupling matrix is made possible by the creation of the logic circuits in integrated technology easier to manufacture than if you had special coils according to DT-AS 1 213006 and 1267 725 would attach. The logic circuits can also be used together with the crosspoints can be accommodated on a common chip.

In the embodiment shown in FIG. 1, the logic circuit consists of three logic elements, all three of which can be integrated together on the same chip. The HXKLU-SIV-OR element Gl emits an output signal when a row control signal Z or a column control signal S is exclusively fed to this logic element G2. Accordingly, in this example, the output of this logic element G2 is connected to a switch Sc so that when this exclusive-or condition is met, the flip-flop k / k 'is controlled in its first state, whereby the coupling points KP in their first, non-conductive state State can be controlled. In addition, an AND gate or NOR gate Gl is attached, which then controls the flip-flop k / k ' in its second state and thus directs the crosspoints in their second state if both a row control signal Z and a column control signal S at the same time are fed. If neither a row control signal nor a column control signal is supplied, neither the logic element Gl nor the logic element Gl affects the associated flip-flop k / k ', in that the switches 5c are controlled in their non-conductive state.

Instead of CMCS transmission gates with flip-flop

control points KP by means of the integrable logic circuits. The flip-flops KZ, KS attached between the control line inputs and the logic element inputs serve, as described, to supply a defined potential to the logic circuits Gl, G2, G2> and to protect the module.

In and of itself, it is not necessary to additionally attach the third link G3 in order to fulfill the functions described above. However, the logic element G3 allows the additional advantage that the potentials fed to the inputs Z, S can only control the two logic elements G1 / G2 if a "1" release signal EN is also fed to the input EN. The selection signal EN is generally also called “enable signal”, “chipenable signal (CE)” and “chip select signal (CS)” in the specialist field, cf. B. Becker / Mäder, highly integrated MOS circuits, publishers Berliner Union and Kohlhammer, 1972, p. 85. This signal EN = "1" causes the logic elements Gl, G2 to work in the manner described above. However, if no signal EN is supplied, ie if EN = "0", then the two logic elements Gl, G2 are blocked, so that neither the line control signal Z nor the tray control signal S can influence the flip-flop k / k ' . Because the logic circuit only then emits output signals that control the flip-flop k / k ', i.e. that control the coupling point,

^r > if the logic circuit is additionally supplied with a signal EN , the coupling points KP can be protected, namely prevent individual interference potentials temporarily supplied to the terminals Z, 5 from having undesirable consequences on the state of the coupled coupling points KP .

This protection of the crosspoints is particularly important if the row control signal Z or the column control signal S has a certain transit time or is overlaid by oscillating distortions.

H In this case, undesired cleaning of these crosspoints would very easily be effected, if not because of the lack of the signal EN, ie when EN = "0". the VerknüngsglipHer Eq. G2 are blocked until the then clearly correct potential is supplied to the inputs Z, 5 with certainty. Even if the row control lines Z and column control lines S are temporarily disturbed - z. B. by transient effects of a decoder that generates a specific control signal S or Z on a specific control line of the matrix from a coded signal - so reliable operation of the coupling matrix is possible due to the signal EN. The state of the flip-flops and the crosspoints connected to them is not influenced by the logic circuits of the crosspoints until the signal EN = "1" via the row control lines and column control lines.

The signal EN normally represents a DC potential. Instead, however, a clock pulse CL can also be fed to the input EN without changing the circuit. The clock pulse has basically the same effects as the signal EN in FIG. 1. The clock pulse CL therefore has the effect that the flip-flops k / k 'of the coupling matrix are only controlled during the clock pulse. If no Λ aktimpuls CL is supplied, all link gücuci GI, G2 of the Küppeimairix are blocked, so that the flip-flops k / k 'of the coupling matrix cannot be influenced. If a clock pulse CL is fed to the input EN shown in FIG. 1, the switching matrix can be operated clock-controlled, which is often very convenient, especially when using the switching matrix in large, computer-controlled switching systems because of the uniformity of the operating technology. At the same time, disturbances due to transient processes or runtimes in the control signals are rendered harmless,

FIG. 2 shows such a development in which the flip-flop k / k ' shown in FIG. 1, here forming an auxiliary part SK , is preceded by an additional flip-flop MK , here forming a main part. The auxiliary part SK controls the coupling point KP from one to the other state only at the beginning of the clock trailing edge, if a corresponding command was received in the main part MK during the clock pulse CL that has just ended. This flip-flop MK / SK caused by its internal delay between dtm reception of the command in the main part MK and the output of the command in the auxiliary part of SK that transient distortions in the logic gates G, G2, or in the control signals Z and / or S, while they are still occur during the beginning of the clock pulse CL , have no influence on the associated auxiliary part SK . Of the

9 10

Auxiliary file SK is only used by that generally un- clock pulse from the control line inputs Z, S

distorted command that are separated when the clock is reversed. Through this last mentioned

edge is present at the input of the master part MK , and the disconnector is in the input flip-flops KZ,

so passed on to the auxiliary part SK . This KS each time that signal Z during the clock pulse

Embodiment is particularly reliable ί and 5 blanked and stored that at the beginning of

with regard to transient distortion. Clock pulse these input flip-flops KZ, KS to-

In addition, this implementation guide was provided. In this way, disturbances in the control

play the P ^r lip flop KL at the clock pulse input CL er signals Z and S, which during the clock pulse

To generate the switching matrix controlling, rege CL or /// 'occur without affecting the

Defined clock pulses / and t ' with a defined Po in operation of the switching network. The flip-flops KY, KS

potential exploited. These regenerated clock pulses / then speak together with the downstream

and t ' control the individual components in the coupling flip-flop MK / SK in terms of structure, function and

field. In particular, these regenerative benefits control the Fiip-Flops according to DE-AS2237579;

Clock pulses at the input of the flip-flops KZ, KS lie- caused by delay effects disturbances of the Si

Low circuit breaker (see. Fig. 2) so that this flip \ i signals Z, S and the clock pulse CL are therefore far ^

Flops at the beginning of the signal EN or at the beginning of the going harmless.

^ ii λ Ri ^ tt ueichnuri ^ en

Claims (6)

Patent claims: 1. Coupling matrix produced using integrated technology with bistable coupling points having a control electrode and two different main line connections, the main lines of which are controlled both by controlling their control electrode from their second, conductive to their first, non-conductive state and by controlling their same control electrode from their first, non-conductive can be controlled in their second, conductive state and their main lines are used to switch through and separate connection paths, with all crosspoints being controlled by the output of a logic circuit that is individually assigned to them and is in turn controlled by the row control signal and column control signal and manufactured using integrated technology, for telecommunication and in particular for telephone switching systems, characterized in that all coupling points (KP or K / K ', SK) are controlled by the respective logic circuit (Gl, G2) according to the following logic functions rden: a) with exclusive control (z = 1, s = 0 / s = 1, ζ = 0) from the row control signal (z) or from the column control signal (s) (via Eq first, non-conductive state controlled (»cleaning«), b) with coincident control both (s = 1, Z = I) by the line ... control signal (ζ) and by the column control signal (5) (via Eq «), c) when neither (z = 0, s = 0) is controlled by the row control signal (;) nor by the column control signal (s) , the crosspoint in question is left in its first or second state, in which it was previously. 2. Coupling matrix according to claim 1, characterized in that the respective logic circuit (Gl, G2. G3) only emits an output signal controlling the coupling point if it is additionally supplied with a selection signal (EN). 3. Coupling matrix according to claim 2, characterized in that the selection signal (EN) is identical to a clock pulse (CL) generated by a clock generator. 4. Coupling matrix according to one of claims 2 or 3, characterized in that a flip-flop (KZ, KS) is provided at the input (Z, 5) of the row control line and column control line, which at the beginning of the selection signal ( EN, CL) from the input is separated. 5. Coupling matrix according to one of the preceding claims, characterized in that in each case between the logic circuit (Gl, G2) and the controlled coupling point, a flip-flop (k / k ¹ ) controlled by the logic circuit and causing the bistability of the connected coupling point is inserted, that in his- nem first state controls the coupling point (KP) in its first state and which controls the coupling point in its second state in its second state. 6. Coupling matrix according to one of the preceding claims, characterized in that the coupling points are each formed by a transmission gate from two complementary field effect transistor (CMOS) whose control electrodes are connected to the associated flip-flop and thus controlled by this flip-flop.