The invention relates to electronic contacts enabling to establish a low or high impedance between a first and a second terminal under the control of a circuit providing a control signal between a third and a fourth terminal.

Such electronic contacts are for instance used in the Belgian patent No 896 388 particularly relating to a controlled capacitive charge circuit enabling to positively or negatively charge a capacitance which, in accordance with the sign of this charge, opens or closes an electronic contact constituted by two DMOS transistors in series opposition, in such a way that their drains respectively constitute the two terminals of the electronic contact while their sources are each tied to the same terminal of the capacitance and their gates both connected to the other terminal of the capacitance, with the possibility for this last to be constituted by the parasitic capacitance between these paired terminals. In such a way, by using transistors able to withstand relatively high voltages, one obtains an electronic contact which can be inserted in a circuit where one or the other polarity may appear at the contact termi¬ nals. Indeed, when the polarity of the charge on the control capacitance for the contact is such that it does not offer a low resistance path, i.e. that the two transis-

tors are blocked, the parasitic diodes which appear for this transistor state between the source and the drain are thus connected also in series opposition which maintains a high impedance whatever the polarity applied by the circuit in which the contact is inserted.

One of the objects of the present invention is to enable the use of a more advantageous type of electronic contact able also to be controlled by the polarity of the charge of a capacitance, and particularly thyristor type device able to work with high breakdown voltages (300 Volts for instance) as envisaged in the above mentioned patent but which can only pass current in one direction while they can block voltages of one or the other polarity, the transistors of the above mentioned patent having inverse properties, i.e. they can conduct the current in one or the other direction but block only one voltage polarity.

The general object of the present invention is to enable the use of such electronic contacts while avoiding a complication of the control circuit. In accordance with a first characteristic of the invention, the electronic contact defined above is charac¬ terized in that two auxiliary electronic contacts are foreseen and enable to establish a low or high impedance between the first and the third terminals and between the second and the third terminals, the impedance conditions of these two auxiliary contacts being opposed.

Such an arrangement offers the advantage that two electronic contacts of the thyristor type can be connected top against bottom as a triac and controlled with the help of the same control circuit and particularly that of the above mentioned patent using a positive or negative charge for a capacitance to close or open the electronic contact. Indeed, with the help of the auxiliary electronic contacts, in accordance with the polarity of the voltage applied to the terminals of the electronic contact constituted by the

two biassed contacts connected in shunt opposition, it will automatically be possible to obtain a connection between a terminal of the control capacitance and the terminal of the main electronic contact having a given polarity. In this manner, the same capacitance charge circuit, i.e. the voltage doubler AC/DC converter described in the above mentioned patent, shall always be used to close or open that of the two polarized contacts which is effectively inserted in a load circuit and depending upon the polarity of the voltage appearing at the terminals of these contacts connected in shunt opposition.

On the other hand, the advantage of thyristor electronic contacts able to be controlled in the indicated manner, and with respect to the DMOS transistors connected in series opposition as in the above mentioned Belgian patent, instead of the top against bottom connection proposed, is that the resistance for the closed condition of the contact is distinctly lower, i.e. below 10 ohms instead of 25 + 25 = 50 ohms. Moreover, for the thyristor solution, the necessary surface in an integrated circuit is reduced to one quarter.

The present invention also relates to an elec¬ tronic device forming part of a circuit including also a power source and a load, said device including means to limit the power dissipation therein.

Such a device is already known, e.g. from the PCT patent application WO 82/03733. Therein the power limiting means are designed to produce a current versus voltage or I/V characteristic which, starting from the origin, rises to a maximum current at a predetermined voltage, remains on that current until a maximum voltage is reached and then suddenly drops to a current substantially equal to zero. At the last breakpoint of this characteristic the power dissipated in the device is maximum (maximum current and voltage) and in some circumstances this may be inadmis-

sible, e.g. in case the device is to be integrated in an electronic chip.

An object of the present invention is to provide an electronic device of the above type, but having a reduced power dissipation.

This object is achieved due to the fact that said power limiting means are designed to produce a current versus voltage characteristic for the device which, star¬ ting from the origin, crosses the load line of the device and thereafter tends to follow said load line towards the voltage axis without again crossing the load line defined by the short-circuit current through the device and the « open-circuit voltage across it.

The minimum of power dissipated in the device occurs in the working point thereof, i.e. the point where the I/V characteristic crosses the load line. Unwanted abnormal signals having various origins like lightning striking such a telephone circuit or main power supply- accidentally connected to it could affect the characteristics of the circuit. Indeed, such signals are added to the normal signals generated by the power source so that the position of the load line is modified. The working point then moves along the part of the I/V characteristic crossing the load line. In case these unwanted abnormal signals become very large, the load line could be displaced in such a way that the working point reaches the upper end of that part of the I/V characteristic. This working point then becomes unstable and moves to higher voltages. Since the I/V characteristic then follows the load line the power dissipated in the device during this transition of the working point is reduced to a minimum. When the unwanted abnormal signals disappear the load line returns to the position initially mentioned and due to the fact that the part of the I/V characteristic following the load line does not cross the latter, the working point moves from the

CPI fay, IPO

higher voltages to its initial position. This would not be the same if there was a crossing between that part of the

I/V characteristic and the load line. Indeed, such crossing would create a working point distinct from the one mentioned above and affect the normal working of the electronic device.

Another object of the present invention is to allow, for relatively small voltages, currents which are much larger than the short-circuit current to flow through the electronic device while keeping the advantages men¬ tioned above for higher voltages.

This object is achieved due to the fact that the part of said current versus voltage characteristic crossing said load line has a first portion extending, for voltages which are relatively much smaller than said open circuit voltage, to a current relatively much larger than said short-circuit current and a second portion joining said first portion to the part of the characteristic which follows the load line. The working point of the device may thus move along the first portion of the I/V characteristic so that the current in this device may reach said relatively large value for small voltages without activating the power limiting means. For higher voltage values the device operates as described above.

Another object of the present invention is also to use such electronic contacts in telecommunication systems and particularly in telephone line circuits, in order to enable the accomplishment of various supervision and control operations, including the provision of a ringing current, functions which previously were generally accomplished by means of relay contacts even in central exchanges where the rest of the equipment was electronic. Thus, the invention is also related to a tele- communication system line circuit comprising a series

OMPI

^«S<_

impedance in each of the two line conductors and contacts on each side of these two impedances enabling to selecti¬ vely connect their terminals towards the exchange or the line respectively or alternatively, towards auxiliary circuits.

Such a system is to be found for instance in the article published on pages 316 to 324 of the IEEE Journal of Solid State Circuits of June 1983, and more particularly on page 317. It is seen there that the two series resistan- ces serve to feed a telephone subscriber line and also to measure the voltage appearing across these resistances and this for supervision and control operations. On the exchan¬ ge side of these resistors, a ringing current can be injected by means of the corresponding contacts and by measuring the voltages across the resistances one may thus supervise the ringing operation. On the other hand, on the other side of these resistors, the contacts on the side of the subscriber line enable to have access to busses to perform tests, either internal (towards the exchange and through the series resistors) or external towards the subscriber line. Until now, these contacts were generally realized by means of change over contacts of three relays which automatically implied that when the make part of the contact was closed in shunt towards one of the control circuits, the break part in series with one of the resis¬ tors was automatically open and vice versa.

In accordance with another characteristic of the invention, these contacts are constituted by four pairs of electronic contacts, the first connecting the line to the impedances, the second connecting these to the exchange, the third connecting the impedances on the line side to a first auxiliary circuit and the fourth connecting them on the exchange side to a second auxiliary circuit.

In accordance with yet another characteristic of the invention, only the first pair of electronic contacts

QMP1_ R T\4

is provided with power limiting means such as described above.

In accordance with an additional characteristic of the invention, the eight electronic contacts which are always operated in pairs are additionally controlled in such a way that only eight combinations among the sixteen possible for the four pairs are permitted.

In accordance with yet another supplementary characteristic of the invention, the control device of the four pairs of electronic contacts comprises a decoder able to be fed by three binary signals in parallel and providing four binary output signals to control the four pairs of electronic contacts, a binary selection circuit being additionally foreseen to enable or inhibit the decoder outputs and in this last case enable corinections permitting to the four input binary signals to respectively control the four pairs of electronic contacts.

In this manner, it becomes possible in particular to realize in the form of a single integrated circuit, not only a series of eight electronic contacts able to with¬ stand relatively high tensions and operating in pairs, but also to control the operation of these electronic contacts either with the help of a code comprising only three binary elements, or directly by a signal corresponding to the pairs of electronic contacts. This versatility can yet be increased by the incorporation in such an electronic circuit of a clock enabling to operate the control circuits of the electronic contacts in a manner described in the above mentioned patent and thus avoiding to rely on a separate clock circuit.

The invention will be best understood and other characteristics thereof appearing in the claims will be better outlined from the following detailed description of preferred embodiments to be read in conjunction with the drawings accompanying the description and which represent :

Fig. 1, the circuit of an electronic contact in accordance with the invention;

Fig. 2, the electronic contact control circuit of the above mentioned patent and modified in accordance with the invention;

Fig. 3, that part of a telephone line circuit incorporating eight electronic contacts in accordance with the invention;

Fig. 4, the whole of the circuits enabling to control the eight electronic contacts and shown solely in the form of a single block in Fig. 3;

Fig. 5, an input protection circuit shown as a block in Fig. 4;

Fig. 6, an electronic gate shown as a block in Fig. 4;

Fig. 7, a double electronic gate controlled by clock pulses and shown as a block in Fig. 4;

Fig. 8, the circuit producing the clock pulses and shown as a block in Fig. 4; Fig. 9, a first logic circuit used to realize the decoder shown as a block in Fig. 4; and

Fig. 10, a second logic circuit used in this decoder;

Fig. 11, another embodiment of the circuit of an electronic contact of Fig. 1 including power protection circuits according to the invention;

Figs. 12 and 13, current versus voltage charac¬ teristics of the power protection circuits of Fig. 11, the characteristics being not drawn at scale; Fig. 14, current versus voltage characteristics of the electronic contact shown in Fig. 11, the charac¬ teristic being not drawn at scale;

Fig. 15, a fault indication circuit FC associated to the power protection circuits represented in Fig. 11 and also shown in Fig. 3.

The electronic contact able to withstand relati¬ vely high tensions and shown in Fig. 1 can be part of an assembly of eight identical electronic contacts (Fig. 3) arranged as four contact pairs, the two contacts of a pair being always simultaneously open or closed, this combina¬ tion being able to be used in a telephone line circuit and particularly as described in Belgian patent No 896 468. Apart from the eight electronic contacts corresponding to that of Fig. 1 and the eight circuits for such contacts appearing in Fig. 2 which essentially corresponds to the controlled capacitive charge circuit of Belgian patent No 896 388, Fig. 4 represents a decoder able to be activa¬ ted either by three or by four binary signals. In the first case, the eight possible combinations of the three binary signals are decoded on four output terminals respectively used to control the four pairs of electronic contacts. In the second case, the enabling signal allows this time to the four input binary signals to be respectively applied to the four electronic gates while the same signal inhibits the decoder operation. Additionally, the circuit of Fig. 4 comprises at the decoder output a converter destined to produce appropriate signals for the capacitive charge circuit of Fig. 2 and this with the help of an oscillator producing complementary clock pulses. The five parts identified above, i.e. the electronic contacts, the control circuit, the decoder, the converter and the clock oscillator can be associated in a same integrated circuit combining a DCMOS low voltage logic and TRIMOS high voltage contacts. The manufacturing technique used can particularly employ the process described in the Belgian patent 897 139. The whole then provides four pairs of electronic contacts able to block in the two directions voltages of 300 Volts and having a dynamic resistance of 10 Ohms when they are conductive, the two terminals of each electronic contact floating with

OMPI

respect to the control circuit. The four contact pairs can be operated in accordance with the 16 possible combinations with the help of four binary signals or in accordance with eight predetermined conditions with the help of three binary signals.

By returning to Fig. 1, it is seen that the electronic contact comprises two identical parts S and S* in such a way that only the first has been represented in detail. In function of the control signal, circuit S can present either a low or a high impedance between its two output terminals S. and S ₂ to which are respectively connected the corresponding terminals S' ₂ and S', of S', the two circuits being thus connected in anti-shunt. This enables to operate them under three different conditions : both S and S* present a high impedance between their terminals, S presents a low impedance for one voltage polarity at the terminals of the contact while S ¹ can also present this ⁴ low impedance but -for the other polarity. Circuit S is of the TRIMOS type constituted essentially by a transistor T.. of the PNP type associated to a transistor T_ of the NPN type so as to form a thyris¬ tor between terminals S, and S-. Manufacture of such a device generally entails the appearance of a parasitic transistor T_ of the PNP type which is connected in paral- lei with the first two. This thyristor combination is controlled by transistor N of the DMOS type associated to transistor P of the PMOS type and whose gates interconnec¬ ted at the same terminal S ₄ present a capacitance C towards terminal S ₂ of the contact to which the drain of transistor P and the source of transistor N are connected.

In this way, assuming that capacitance C has been charged positively at its terminal connected to the two gates of transistors P and N with respect to terminal S~, and that on the other hand the voltage on terminal S. is more positive than that on S_, transistor N becomes conduc-

OMPI IPO

tive which enables a current to flow from terminal S- towards terminal S ₂ through transistor T., due to transistor

N short-circuiting by its drain/source path the base of transistor T, to which this drain is connected, the emitter of T» being connected to S.. The effect of this conductibi- lity of T ₁ is to pump current into the base of transistor

T ₂ which is directly connected to the collector of T., in such a way that T ₂ which is of the NPN type begins to pump current into the base of T. which is directly connected to the collector of T-, whose emitter is directly connected to S_. In this way, by this cumulative action, the two transistors T, and T ₂ are placed into a saturation mode offering a low impedance between S, and S ₂ . Transistor T, which is also of the PNP type as T,, the basis and the emitter of these two transistors being respectively interconnected while the collector of T ₃ is at the potential of S ₂ , becomes also conductive but as indicated, this concerns a parasistic element without influence on the main operation of the circuit. The biassed contact S offering a low resistance between S. and S ₂ can now be put back into its high impe¬ dance condition with the help of a negative charge on capacitance C, this negative potential at the gates of transistors P and N with respect to S ₂ now entailing conduction of transistor P which is of the PMOS type. While transistor N of the NMOS type has its drain connected to the base of T_, the source of P is connected to the collec¬ tor of T» is such a way that P draws current from the T_ collector so that the ₂ base current become insufficient to ma'intain conductibility of this NPN transistor which, by cumulative effect, entails its blocking and that of T.. and T-, thyristor T.. ,_ becoming non conductive. It will still be noted from Fig. 1 that the P and N substrates are connected to the N drain and to the P source respectively. The other S' contact shown solely in the form of

a block in Fig. 1 operates exactly in the way described but this time under the control of a positive or negative charge on capacitance C and more particularly at its terminal S. with respect to its terminal S^- ^Bu t these operations of half contact S ¹ will this time be produced when the polarity of the circuit in which the switches are inserted in anti-shunt is positive in S' with respect to S' ₂ . To be noted that the realization of S/S ¹ in a single integrated circuit entails a connection between the common base of T, and T- for the two contacts S and S' .

As already indicated in Belgian patent No 896 388, capacitances such as C and C can be constituted by parasitic capacitances, particularly those appearing at the gates of transistors P and N in the case of capacitance C. Both C and C* can be charged with a desired polarity by way of the control circuit shown in Fig. 2 and corresponding essentially to a version already described in the Belgian patent last mentioned.

Indeed, transistors-NA and NB which are both of the NMOS type are shown in Fig. 1 as having their sources directly connected to terminal S, while the drains of NA and NB are respectively connected to S, and S-,. On the other hand, the gate of NA is connected to S ₂ while that of NB is connected to S.. Such a circuit arrangement has as a consequence that if the potential of S, for instance is higher than that of S ₂ , that of S ₃ cannot be outside this range and transistors NB and NA are conductive and blocked respectively which implies in fact that terminal S- is practically (0.7 Volt) connected to terminal S ₂ and by referring to Fig. 2 it is seen that it is in fact capacitance C which is effectively connected between terminals S^ and S-. of the charging device shown in Fig. 2. The parasitic diodes between the source and the drain of transistors NA and NB, i.e. DA and DB such as shown in Fig. 1, are biassed in such a way that they play an analo-

gous role by enabling terminal S-, to align itself on the potential at terminal S ₂ when the latter is less positive than that at terminal S..

Of course, in view of the symmetry of the circuit constituted by transistors NA and NB, when the potential of

S ₂ is higher than that of S,, the conditions are inversed and following the conductibility of NA or DA, terminal S- is this time practically connected to terminal S* ₂ in such a way that in these circumstances it is capacitance C which is effectively connected to output terminals S. and S- of the control circuit of Fig. 2.

In this manner, a single control circuit can automatically close or open the half electronic contact S or S' depending on the polarity of the voltage applied between terminals S./S^ on the one hand and S _j /S'. on the other.

As already indicated, the circuit of Fig. 2 is essentially described in Belgian patent No 896 388 and particularly in relation to Fig. 6 of this patent which is very similar to Fig. 2. The latter constitues an AC/DC converter in the form of a full wave cascade voltage doubler and fed in puch-pull by clock pulses of complemen¬ tary polarity. The polarity control of the circuit of Fig. 2 is effected by signal DC ^* applied to series capacitance C- while, the complementary clock signals CL and CL are perma¬ nently applied to the two other series input capacitances C, and C ₂ respectively. In the same way as for the output capacitance C/C (Fig. 1) present between terminals S. and S ₂ /S' , the three input capacitances are not necessarily constituted by discrete physical elements. The first voltage doubler rectifier is essentially constituted by a series capacitance C, followed by the series diode D.. , transistor P. of the PMOS type to reach shunt capacitance C/C* between terminals S. and S-,, the shunt diode of this voltage doubler being D ₂ connected as indicated between the

-In _¬ junction of C ₁ and D, on the one hand and that of C- and of the gate of P. on the other. When the control potential DC- applied to series capacitance C- corresponds to clock pulses CL applied to series capacitance C ₂ , it is the described charging circuit which is effective in ensuring a charge of capacitance C/C in such a manner that the potential at terminal S. will be more positive than that at terminal S_.

In the inverse case, when control signal DC applied to series capacitance C- corresponds to clock pulses CL perminently applied to series capacitance C,, the output capacitance C/C will this time be charged with S. more negative than S^, the elements of this voltage doubler now effective to negatively charge the shunt output capaci- tance being C ₂ , D~, N. and D., corresponding to C, D., P. and D ₂ respectively as shown in Fig. 2, transistor N. being of the NMOS type.

As in Belgian patent No 896 388, the positive charging circuit using the conductibility of transistor P ₁ is completed by transistor N ₂ of the NMOS type whose drain is connected to S- and its source to capacitance C~ by means of the series diode D_, the gate of N ₂ being connec¬ ted to S.. This connection thus permits to complete the return circuit for the positive charge by offering a path between the "ground" output terminal S-. and the input

"ground" terminal constituted by the right-hand electrode of series capacitance C ₂ . Similarly, upon a negative charge of the output capacitance between S, and S-,, the return path is this time effected by way of transistor P_ of the PMOS type in series with diode D _g " these two elements corresponding to ₂ and D ₅ respectively as indicated by the circuit which is practically identical to that of Fig. 6 of Belgian patent No 896 388 to the exception of diodes D. and D, which are this time to be found on the source side of transistors P, and N, respectively instead of being located

on the drain side as in the prior patent. Another version of this circuit shown in Fig. 4 of this prior patent already put diodes D _χ and D ₃ on the side of the sources o transistors P. and N,, but in this circuit the gates of transistors P ₂ and N ₂ were interconnected in another circuit and not connected to the drains of transistors P. and N, (S.) in such a way that diodes D_ and D _g were this time on the drain side of transistors ₂ and P ₂ . In the version of Fig. 2 on the other hand, the four diodes D,, Do-, DD _c and Db, are all located on the source side of the transistors to which they are associated in such a way tha with the diodes D ₂ , D. and D _1Q , they are all arranged on the side of the three input capacitances C. ₂ / ₃ « This last diode D _1Q directly connects capacitances C, and C ₃ in the same manner as in the prior patent and the Zener diodes D_ D„ and D_ respectively in parallel on the output terminals S. _/3 and the source/drain path of transistors N ₂ and P ₂ ar also connected in the same manner as previously.

The integrated circuit IC which may incorporate eight electronic contacts as shown in Fig. 1 as well as eight control circuits as represented in Fig. 2 , appears i the form of a block in Fig. 3 which essentially correspond to a part of Fig. 1 of Belgian patent No 896 468 bearing o a line circuit for an electronic telephone system. As show by Fig. 3, a subscriber line (not shown) can terminate on terminals LT.. and LT_, preferably through an overvoltage protection circuit such as that which is the object of Belgian patent No 896 468. By means of the first electroni contact S-. part of integrated circuit IC, terminal LT. ca be connected to series resistance R. and then by means of second series electronic contact, i.e. S ₂ 1, to the SLIC circuit containing other elements of the electronic line circuit. The circuit between the second input terminal LT_ and the SLIC circuit is exactly similar, S ₁₂ , R ₂ , S ₂ _ corresponding to S^, R ₁ , S ₂₁ respectively. In addition to

the series contacts enabling to connect the resistances between LT. ,_ and the SLIC circuit, these resistances can also be connected via four shunt contacts towards the test circuits TC (S ₃₁ for R. and S ₃₂ for ₂ ) on the subscriber side (LT- _/2 ) on the one hand and towards the ringing circuit RC (S ₄₁ for R. and S ₄₂ for R ₂ ) on the exchange side (SLIC) on the other. Connections (not shown) going from the terminals of the feed resistors R ₁ , ₂ to SLIC enable the latter to supervise the potentials appearing across these resistors.

The operation of the eight contacts is controlled from the SLIC through four conductors terminating on terminals ^ιc _ι 7 _2/3 /4' ^tle contacts of a pair such as S.. ,.., being controlled by the same signals so that the two wires of the connection are switched simultaneously. The fourth conductor reaching IC. is however indicated in interrupted ^• lines since this control can be effected in accordance with a control mode using only three binary signals, a mode selection signal applied to terminal IC ₅ determining if three or four binary signals are used to control the four contact pairs S _n/12 _ S _21/22 , S _31/32 and S _4χ/42 _

This versatility of the IC circuit is based on the fact that the location of these four contact pairs, directly on each side of resistors R. and ₂ enables to ensure an adequate control with a number of connecting states which does not go beyond eight. Accordingly, when the four contact pairs of circuit IC are used for any application, line circuit or other, necessitating between 9 and 16 possible conditions for the combination of these 4 pairs in their open or closed conditions, each of the four binary signals at terminals ^IC I/2/- _? / _Λ ^can directly control the state of a pair of contacts. On the other hand particu¬ larly in the case of the telecommunication line circuit which will be described further, one may be content with a maximum of eight conditions and the selection signal at

terminal IC will this time indicate that only the three binary signals at terminals IC. _/2/3 must be taken into consideration and the eight possible combinations of these signals will be transformed with the help of a decoder DEC into four binary signals each of which can control a pair of contacts.

This appears in the form of a block in Fig. 4 which represents the constituting elements of circuit IC of Fig. 3, except the electronic contacts and their capacitive charge control circuits already described in relation to Figs. 1 and 2 respectively.

In Fig. 4, each of the inputs ^C , _2/3 , ₄ , _ζ ^ ^s coupled to the input of a corresponding inverter

^Iv _l/2/3/4/5 ^anc ^ ^{eχ e} Pt i ⁿ the case of IC where the connection is direct, through identical protection circuit ^PC _l /2/ ₃ /4' ^{tιe P} circuit being detailed in Fig. 5.

The latter shows that input terminal I •C_.. is directly connected to output terminal- A, the input binary code for IC being identified by ABCD, the output terminals B, C, D corresponding to terminals IC _J /Ά /A ^res P ^ec tively.

Input terminal IC. is connected to poles V., and V ₂ of a DC source through diodes D.. and D ₁₂ respectively, these limiting the potential on IC./A between those applied in V. and V ₂ , this last potential, 0 Volt for instance, being more negative than that of V-, 15 Volts for instance. On the other hand, transistor P ₃ of the PMOS type has its source connected to V.. , its drain to A and its gate to V ₂ in such a way that it is continuously conductive. Transistor T ₄ shown in interrupted lines as being of the • NPN type and having its connector connected to IC. and its emitter to V ₂ , can be used to bring a binary control signal to IC. If its base is at the potential of V., it is conductive and enables the current to pass from V. to V_ through P ₃ and T. in series. The impedance of the latter transistor being lower than that of P ₃ , terminal A stands

- l X^

_ QMPI __ Ϊ?O _Λ*.

«& ^■ NATIO

at the potential of V _2> On the other hand, if the base of

T is at the potential of V ₂ so as to block T ₄ , terminal A is at the potential of V. transmitted through P ₃ «

Fig. 4 indicates that the four potentials (ABCD) at the outputs of PC1 2/ /4 ^{are a} PP ^lied - ^to terminals such as ^" δ ^" of a gate GD through inverters such as IV., in such a way that a complementary binary code B C D appears at the inputs of these gates (gates identical to GT being foreseen for the signals at terminals A, B and C) with respect to the binary code ABCD at the outputs of ^pc _ι /2/3/4* Three of these binary signals A ^" ~B C ^" are on the other hand applied to decoder DEC, as well as the two signals A and B complemen¬ tary to A ^" and IT and obtained by inverters IV- and IV_ in cascade with IV.. and IV_ respectively. Before explaining with the help of Figs. 9 and 10 how decoder DEC can advantageously transform the eight combinations of three binary signals ABC into particular combinations of four binary signals at its outputs E F G H, the description of the other elements of Fig. 4 will be completed, beginning with the above gate GD the output of which is connected.to that of an identical gate GH fed by the ^" if output of decoder DEC, three identical gates (not shown) being used and connected in like manner to outputs ^" E F G. Gates such as GD and GH are controlled from terminal IC determining the operation mode of IC, with or without decoding by DEC, the binary signal in IC being applied to all the gates such as GD/BH as well as the complementary binary signal obtained by inverter I _ζ . Fig. 6 shows the circuit of 'a transmission gate such as GD and GH which connects input terminal ^* D or HT to the output terminal DH by the source/drain path of transistors N ₃ and P ₄ connected in anti-shunt and which are of the NMOS and PMOS type respectively. Their gates are connected to IV ₅ and IC respectively for GD and vice versa

for GH in such a way that one of these gates is conductive and the other blocked in function of the selection signal of IC which does permit to choose for terminals such as

DH, either signal D ^" and part of a four binary element code each identifying a pair of contacts such as S.. ,.-

(Fig. 3) , or the signal H, i.e. one of the four binary elements decoded by DEC from the three binary elements

^" A B C.

As indicated by Fig. 4, the signal at the terminal such as DH must still be synchronized by the gate such as GC with the clock pulses provided by oscillator CO to be applied, as well as the complementary clock signals CL and CL, to the three input capacitances C. , _2/3 of the AC/DC push-pull converter (Fig. 2) serving to positively or negatively charge capacitances C/C* controlling the biased electronic contacts S/S' connected in anti-shunt (Fig. 1). Fig. 7 represents the circuit of the clock gate. The binary signal at the terminal such as DH and determining the open or closed condition of the corresponding contact is this time applied to control gates GCA and GCB identical to those of Fig. 6 but to the inputs of which are applied the complementary clock pulses CL and CL used by oscillator CO. Gates GCA and GCB are controlled in complementary fashion by signal DH and its complement produced by inverter IV _g in such a way that the output signal DC from gate GC is either a CL pulse or a complementary CL pulse in accordance with the value of the binary signal in DH.

Fig. 8 represents the clock oscillator CO comprising the three inverters ^ιv ι _n /i i _/ -, connected in cascade in a loop comprising also the series resistances R- and R ₄ on each side of inverter ^ιv _ιι . the output of IV ₁₂ feeding a second series of three inverters IV.-,.. ,._ in

13/14/15 cascade the last of which provides the clock pulses CL and ^ιv i4 ^the complementary pulses ^" CL. Oscillator CO is also fed

- UR£ ^> OMPI NAT\4 >

by voltages V ₂ (not shown) and V^ this last being connected to inputs of ^ιv ι1 12/13 through capacitances

^C _4/ς/fi respectively. These can be of 6 picofarads and _3/ - ₄ of 20 kohms to produce oscillations at a frequency of the order of 1.2 MHz.

As indicated by multipling arrows in Fig. 4, oscillator CO feeds the four gates such as GC whose output terminal controls two charging circuits such as those of

Fig. 2 to control a pair of contacts such as that of Fig. 1. This last connection is performed by inverter IV _g providing the signal 1C is such a way that the three signals arrive at the capacitances C. , _j ι- _t ^of Fig. 2 through the output impedance of an inverter.

Decoder DEC of Fig. 4 will. finally be described by referring also to Figs. 9 and 10 representing the type of logic circuits advantageously used for its realization.

To this end one will first of all define the eight conditions of a telephone line circuit which can be characterized by a combination of the input signals ABC of the decoder DEC. These eight conditions are identified by the truth table which follows :

^S ll/12 ^S 21/22 ^S 31/32 ^S 41/42

A B C Y E F G H

Isolated 0 0 0 1 1 1 1 1

Ringing test 0 0 1 0 1 1 0 0

Ringing supervision 0 1 0 0 0 1 0 0

Ringing 0 1 1 0 0 1 1 0

External test 1 0 0 1 0 1 0 1

Internal test 1 0 1 0 1 0 0 1

Supervision 1 1 0 0 0 0 0 1

Switching through 1 1 1 0 0 0 1 1

, wr?o

The table comprises three columns corresponding to the input signals ABC, a fourth column for an intermediate Y signal whose usefulness will appear later, and four other columns E F G H defining the signals at the output of decoder DEC, each of these last columns corresponding, as indicated, to the state of a pair of contacts, e.g. If for S.. ₁₂ -

The complement bar on top of the references identifying these contacts corresponds to the complementary form of E F G H in such a way that the indication 0 identifies a closed contact for the columns at stake, with 1 for an open contact. The eight conditions for the line circuit appear in the successive rows in increasing order for the binary code from 000 to 111 for ABC, this last code corresponding to the switching through of the line circuit of Fig. 3, i.e. the series contact S-. ,. ₂ and S-.. _/22 are closed and the shunt contacts S... _/32 and S ₄ . , . _j ^are open. The fourth line gives the code 011 for ABC as ringing condition enabling to connect RC (Fig. 3) towards terminals ^L - ^r ι _/ °^ ^tιe subscriber's line through resistors R.. < ₂ in such a way that the voltages on the latter can also be used for the supervision of the ringing operation of the called subscriber. The sixth line corresponds to code 101 for ABC and to an internal test (towards the exchange) enabling this time to connect the test bus of TC to the SLIC through resistors R. ,_. On the other hand, the external test (towards the subscriber) of the fifth line (100 for ABC) produces a connection between TC and terminal LT.. ,_ without passing through the resistors while the ringing test (001 for ABC) interconnects this time RC and TC through the resistors.

Apart from these five conditions, the line circuit still enables complete isolation of the resistors (000 for ABC) , ringing supervision (010 for ABC) where TC is additionally branched by S... _/32 on the above output

-£υl £ ^»

_ OMPI fa^ IPO -

connection and finally, a supervision (110 for ABC) where

TC _. is also branched but this time on the normal connection.

The eight ABC codes enabling these various conditions have been assigned to the eight contact combinations ^s n/ 2' ^S 21/22' ^S ₃ 1/32 ^{7 S} 41/42 ^as i ^nα -i ^cat e<3. by the table and this to enable as simple a realization as possible for DEC. Indeed, the correspondence table indicates that ^* E = ^~ B and that ^" = ~k~, except for ABC = 100, that G = 0 except for ABC = 000 and BC = 11, and finally that IT = A except for ABC = 000. The realization of decoder DEC is facilitated by these simple correspondence relations and by the introduction of Y = 0 except for BC = 00, Y being an intermediate binary signal appearing at the fourth column of the table. Accordingly, one may write the corresponding

Boolean relations :

^" E = B ^" (A + C) = A B ^" + B ^" C ^" = X + ¥ = + B C

G ^" » A Ϋ ^" + BC = A~¥ C ^" + BC ^" H = A + Y = A + B ^* C ^"

T = B C or Y = B + C where the second expression for ^" E~ (Fig. 9) will facilitate a comparison with G (Fig. 10) and those for F, G and H are obtained by replacing ^" ? by the indicated value, that for Y corresponding more directly to the logic circuits used and more precisely a derivation of that of Fig. 10 for ^" G ^" .

Fig. 9 represents the CMOS logic circuit enabling to realize E by using three PMOS transistors connected as indicated between potential V. and the output terminal giving function ^~ E~, as well as three NMOS transistors connected as indicated between the output terminal and the potential V ₂ « The three transistors are identified by the signals A, ^" B, C applied to their gates, as well for the PMOS as for the NMOS transistors. Their source/drain paths are connected in such a manner that B is in series with a

-23- parallel combination AC for the PMOS transistors while the duality of the circuit comprising the NMOS transistors implies that B should be in parallel with the series combination C Hence, signals X, I, C being respectively the complement of those intervening in the equation giving

E, it is seen in particular that if B is at the low potential, the transistor B among the PMOS transistors is conductive while the corresponding NMOS transistor is blocked. If one forgets transistors and C among the PMOS transistors by replacing them by a short circuit, while by duality, the NMOS transistors ^" X and C are replaced by an open circuit, ~& would then be at high potential (V ₁ ) which corresponds to E = B~, this latter being the first factor of the expression defining E and a condition which, as previously indicated, is true for all the combinations of

ABC except 100. But for this last combination, with A and C ^" at the high potential and B at the low potential, the PMOS transistors controlled by A and C are thus both blocked while the corresponding NMOS transistors are both conductive. For this particular combination, ^" E ^" is thus at the low potential (V ₂ ) which corresponds to ^" 1 ^" = B. For the seven other combinations, these four transistors A and C~ (PMOS and NMOS) are irrelevant because they cannot either short-circuit the blocked transistor B (NMOS) nor put the conductive B transistor (PMOS) in an open circuit in such a way that only B is relevant and that one has ~Ε - B~.

Fig. 10 represents the circuit enabling to realize G ^* following principles identical to what has been outlined for Fig. 9, the second form given above for ^" IT enabling a direct comparison with the first for G ^" . Indeed, it is readily seen that there are four independent variables X, B, C and Y for G instead of only three (A, B and ^" C) for ^" E ^" . In this way, four pairs of PMOS and NMOS transistors connected as indicated are this time necessary and one calls on the intermediate variable Y.

To obtain the latter from B and C, as well as F and IT from X and ^~ Ϋ~, and from A and Y respectively, it suffices each time to take half the circuit of Fig. 10, e.g. transistors T and ^" C, both for the series PMOS transistors and for the shunt NMOS transistors by controlling them by appropriate signals, e.g. B instead of IT and C instead of C~to provide Y.

In this manner, the whole of decoder DEC which only calls on the five signals A, A, B, B and C by economizing an inverter for C (Fig. 4) , uses only 13 PMOS and 13 NMOS transistors.

Reference is now made to Fig. 11 which shows the electronic contact S ₁₁ of Fig. 3 in detail, this contact being a modification of the one represented in Fig. 1. While being still of the TRIMOS type, the thyristor TRX formed between the terminals S. and S ₂ differs slightly from the one schematically represented in Fig. 1 in that the PNP transistor T. is now replaced by a PNP transistor Q. having two distinct collector electrodes respectively connected to the gate electrodes of two NPN transistors Q ₂ and Q ₃ substituting the NPN transistor T ₂ , transistor T ₃ being not represented. Furthermore, individual power protection circuits which are described below in more detail are associated to TRX. To be noted that contact S. ₂ is identical to S.. butthati±e other six contacts do not include power protection circuits. Electronic contact S-. of Fig. 11 includes two identical switching circuits S and S' which are coupled in anti-shunt. More particularly, terminal S. of S is connec- ted to terminal S ¹ - of S ¹ , whilst terminal S ₂ of S is connected to terminal S'. of S'. The control circuit mentioned above is connected to both S and S' via terminal S.. The switching circuits S and S ¹ are also each provided with a detection output terminal DT., , DT ₂ of which only DT. is connected to a fault indication circuit FC via detection

terminal DET.. FC is also included in the integrated circuit IC and will be described later. Because S and S' are identical only one of them, e.g. S is now considered. As already mentioned above, transistor Q. of the thyristor TRX has two distinct collector electrodes connec¬ ted to the base electrodes of the transistors Q ₂ and Q ₃ respectively. The collector electrodes of the transistors Q ₂ and Q ₃ are both connected to the base electrode of Q.. Terminal S. is connected to the emitter electrode of Q ₁ and terminal S ₂ is connected to the emitter electrode of Q ₂ directly and to the emitter electrode of Q- __. Vi.a el S_._c_H„S_._c_ resistor R, -.. The base electrode of Q ₂ is also connected to the collector electrode of an NPN transistor Q ₄ whose emitter electrode is connected to the terminal S-,. Terminal

S. is connected to the base electrode of Q ₄ via the cascade connection of a diode D ₂₁ , the collector-to-emitter path of an NPN transistor Q ₅ and a resistor R ₁₂ * ^τ ^ ^e cathode of diode D ₂ . is also connected to the emitter electrode of Q ₃ via the series connection of a resistor R-, ₃ . the drain- to-sourσe path of an NMOS transistor N.. and a resistor R... The junction point of resistor R. ₄ and the source electrode of N... is connected to the base electrode of an NPN transistor Q _fi whose collector electrode is connected to ^• the base electrode of Q. and whose emitter electrode is connected to the terminal S_. The cathode of diode D-. is also connected to the drain electrode of an NMOS transistor N. ₂ whose source electrode is connected to the base elec¬ trode of Q _g together with the detection output terminal DT _; .. The gate electrodes of N.. and N ₁₂ are both connected to the drain electrode of a DMOS transistor N ₁₃ having its source electrode connected to the terminal S ₄ and its gate electrode connected to the terminal S ₂ . To be noted that the DMOS transistor N. ₃ has a parasitic diode (not shown) whose anode is connected to the source electrode of N.. a whose cathode is connected to the drain electrode of this

transistor, and that the NMOS transistors N^ and ^ ₂ have high gate capacitances (not shown) .

The thyristor TRX is turned ON and OFF by means of MOS transistors (not shown) corresponding to transistors P and N of Fig. 1, controlled via terminal S4. As already mentioned above, the switching circuit Sincludes power protection circuits which are also able to control TRX and the functioning of which will be described in detail hereinafter. The thyristor TRX is associated to two distinct power protection circuits, called primary and secondary power protection circuits respectively. The primary power protection circuit comprises the components D ₂ ., N ₁₂ , Q_, R ₁₂ and Q-, and more particularly limits the current through TRX when the voltage across the circuit S exceeds a predetermined value. The secondary power protection circuit comprises the components D ₂₁ , R ₁₃ -/ ^N n/ Q _g ' ^R i ₄ ^anσ ^ ^R n« o be noted that in the following description of the operation of the protection circuits it is assumed that the voltage at terminal S. is positive with respect to the one at terminal S ₂ so that diode D ₂₁ is forwardly biased. The same functioning is valid for S' in case the voltage at S* ₂ is positive with respect to the one at S'..

The primary and secondary protection circuits are brought into and out of service by the respective NMOS transistors $- . and N. ₂ which are themselves controlled by DMOS transistor N ₁₃ . When the thyristor TRX is in the ON-state, a positive control voltage of about + 20 Volts applied to terminal S. is transmitted to the gate electrodes of N.. and . ₂ via the parasitic diode of transistor N.^. As a result, the transistors N.. and N ₁₂ are conductive and the protection circuits are in service. To turn OFF the thyristor TRX, the control voltage at S ₄ is decreased ^' from its positive value of about + 20 Volts to a negative value of about - 20 Volts. During this voltage

transition, TRX turns OFF when the voltage at S ₄ reaches about - 3 Volts, whilst the protection circuits remain in service even when the parasitic diode of the DMOS transistor N., is blocked. Indeed, the NMOS transistors are then still conductive due to the positive voltage latched by their gate capacitance. When the voltage at S ₄ reaches about - 8 Volts transistor N_ ₃ becomes conductive so that this negative control voltage is applied to the gate electrodes of N.. and N. ₂ and blocks them. The power protection circuits are then out of service. The transistor N ₁₃ coupled to the gate capacitances of the transistors N.. and N. ₂ thus constitutes a delay circuit which brings the protection circuits out of service a time interval after the blocking of TRX. Thus the protection of this device remains operative as long as TRX is in the ON-state.

When TRX is in the ON-state, no current flows through the primary power protection circuit as long as the voltage across the switching circuit S does not exceed about three diode voltage drops. These three diodes are the diode D ₂₁ , the base-to-emitter path of transistor Q_. and the base-to-emitter path of transistor Q . . To be noted that the current flowing through the primary power protection circuit is so small that the voltage drop across resistor R. ₂ is negligible and that the voltage drop across the cirain-to-source paths of conductive transistor N. ₂ is also negligible since this latter voltage drop is proportional to the base current of transistor Q ₅ which is still blocked. When the voltage across S increases, transistor Q ₄ becomes conductive and the collector current of Q. is drained from the base electrode of Q ₂ to the terminal S_. As the base current of transistor Q ₂ is reduced, its collector current and, consequently, the base current of transistor Q. are also reduced. As a result, the portion Q./Q ₂ of the thyristor TRX turns OFF whilst its portion °-ι^°-2 ^a -^ ^rema ^- ⁿ turned ON as will be described later.

During the above operation the main current in the tran¬ sistor Q ₄ is limited by resistor R. ₂ and transistor Q ₅ . which is itself controlled by transistor N.-.

Considering only the portion Qι/Q ₂ °f the thyris- tor TRX, the current versus voltage characteristic of the switching circuit S would be the one represented in Fig. 12 where part 1 is the normal I/V characteristic of the forwardly biased thyristor TRX. As shown, the voltage V rises to a maximum voltage V _D equal to the above mentioned three diode voltage drops (+ 2,1 Volts) and corresponding to a maximum current I. of 320 milli-amperes. From the above it follows that, transistor Q ₄ becomes active at the maximum voltage V _D corresponding to the current I. so that TRX turns OFF following part 2 of the I/V characteristic shown in Fig. 12. The current in the thyristor TRX and therefore in the switching circuit S is then substantially equal to zero whatever the voltage across this circuit may be so that I/V characteristic then nearly coincides with the voltage axis for voltages exceeding V_. To be noted that this I/V characteristic is valid for the thyristor TRX as well as for the switching circuit S.

The DC load line 3 of the switching device S is also represented in the diagram of Fig. 12. It is defined by two points corresponding respectively to the maximum current I (70 milli-amperes) in the telecommunication line when the latter is short-circuited and to the maximum voltage V (70 Volts) when this line is open. This DC load line 3 crosses part 1 of the I/V characteristic of the switching circuit S at a stable working point 4. When unwanted abnormal signals are applied to the telecommunication line they are added to the normal signals generated by the telecommunication exchange so that the load line moves in the I/V diagram of Fig. 12. Such abnor¬ mal signals may have various origins like lightning striking the telecommunication line or main power supply

accidentally connected to these lines. The working point then moves along part 1 of the I/V characteristic. When these unwanted abnormal signals become very large, the load line could be displaced in such a way that the working point reaches the upper end of part 1 of the I/V charac¬ teristic. This working point then becomes unstable and moves to higher voltages whilst the TRX turns OFF (part 2) . However, the maximum voltage V„ across the switching circuit S is limited to about 250 Volts by the overvoltage protection circuit (not shown) mentioned above so that the working point is then located at point V _M on the voltage axis.

When the abnormal signals disappear the DC load line shifts back into the position drawn in Fig. 12 and the working point moves from V Α„ (250 Volts) to V_ LI (70 Volts) where the part of the I/V characteristic of TRX which coincides with the voltage axis crosses the DC load line 3. The working point thus becomes stable at that voltage V_ and since the primary power protection circuit is then still active, it is impossible to turn ON the thyristor

TRX. To allow TRX to be turned ON again, the part 2 and the part of an I/V characteristic of the switching circuit S coinciding with the voltage axis should not cross the DC load line 3 so that no stable working point such as V_ should exist between V _M and the normal working point 4. A solution is to use the secondary power protection circuit described hereinafter.

Considering only this secondary power protection circuit, when the switching circuit S is in the ON state a current flows from S. to S ₂ (Fig. 11) not only via TRX but also via diode D ₂₁ , resistor Ri , drain-to-source path of NMOS transistor N.. and resistors . ₄ and R... in series. As long as the voltage between the terminals S. and S ₂ is relatively so small that the voltage drop produced by the above currents across R... and R.. in series is smaller than

the base-to-emitter saturation voltage V _B „ of Q, the latter remains blocked. The current I flowing through TRX then varies in function of the voltage V measured across the switching circuit S according to part 5 of the I/V σharac- teristic shown in Fig. 13. To be noted that due to,the values of the resistors which will be given later the current I (Fig. 11) flowing through TRX is much larger than the current flowing through the secondary protection circuit. Hence, the current I may be considered to be the current flowing through the switching circuit S and, as for Fig. 12, the I/V characteristic of Fig. 13 is valid for the thyristor TRX as well as for the switching circuit S. When the voltage between the terminals S. and S ₂ is so large that the voltage drop produced across R.. and R- ₄ in series by the above mentioned currents becomes larger than V_ _E of Q _fi , the latter becomes conductive and thereby forms a shunt path to S. for the collector current of Q.. Thus the base current of Q ₃ is reduced as a result of which the impedance of TRX increases so that the current I flowing through it varies in function of V in the way represented by part 6 of the I/V characteristic of Fig. 13. This variation is function of the power dissipated in TRX because the voltage drop developped across the switching circuit S not only depends on I because R.. is connected in series with TRX but also on V since an additional current which is function of V flows through R.. via R ₁₃ and R-, ₄ - Without . ₃ and R ₁₄ the current I would remain constant and be equal to the maximum current I ₂ as shown by part 7 of the I/V charac¬ teristic of Fig. 13. In that case the power dissipated in the switching circuit S may become excessive since part 7 crosses the line of maximum power dissipation 8 of the circuit S. For the reasons mentioned above, part 6 of the I/V characteristic should not cross the DC load line 3. On the other hand, because the minimum power dissipated in the switching circuit occurs in the working point of this

-31- . circuit, i.e. in the crosspoint of part 5 of the i/V characteristic and the DC load line 3, part 6 of the I/V characteristic should be chosen as close as possible to the

DC load line 3 in order to obtain a minimum of power dissipation in the switching circuit S. Therefore, the slope of part 6 of the I/V characteristic is chosen similar to the slope of the DC load line 3. This slope is function of the ratio R ₃ /R.. Indeed, when Q _g starts conducting, its base-to-emitter saturation voltage V„ _E may be defined by the following expression :

Where V and I are the voltage across and the current flowing through the switching circuit S respectively. This expression immediately leads to

I.R _ι R ₁₃ = V _BE (Rn+Ri4+R ₁₃ ) - V(R _χι + R ₁₄ ) according to the values of the resistances which are R.. = 7.6 ohms

^R 12 ^{= 500 ohms}

R. ₃ = 145 Kilo-ohms.

R. ₄ = 1 Kilo-ohms the following assumptions can be made

R ₁₃ » ₁₄ » R the final expression is

,R _n .R ₁₃ = V. BE R 13 - V.R 14 so that

^{1 =} ST1T1 ^{(VβE "} R R ₁ ¹ ₃ ^{' V)}

It appears clearly from this expression that the current I is dependent on the voltage h14.V.

R,

¹³ Since part 6 of the i/V characteristic is chosen

as close as possible to the DC load line 3 in order to limit the power dissipated in the switching circuit S the maximum current I ₂ must be chosen slightly above I _L and the maximum voltage V ₂ must be chosen slightly above V_. In the present example and with the values of the resistors given above I ₂ = 100 milli-amperes and V ₂ = 100 Volts approxi- matively. However, according to requirements which are conventional for a telecommunication system the protection circuit should only be activated for a current exceeding 300 milli-amperes. If, for this reason I ₂ is chosen higher than the required 300 milli-amperes part 6 of the charac¬ teristic should be shifted up and a portion of it may be located above the line of maximum power dissipation 8. In that case, when the power protection circuit becomes active, the power dissipated in S may be so large that the latter is destroyed.

The drawbacks of the two power protection circuits taken separately may be eliminated by combining these two circuits, this combination providing the overall I/V characteristic of the switching circuit S represented in Fig. 14. This characteristic has part 1 and partially part 2 of the I/V characteristic related to the primary power protection circuit, and part 6 of the I/V charac¬ teristic related to the secondary power protection circuit. From this figure it appears clearly that the I/V charac¬ teristic crosses the above mentioned DC load line 3 at a unique stable working point 4 and that the power dissipated in the switching circuit S is reduced to its minimum since part 6 is very close to the DC load line 3. A fault indication circuit FC is shown in

Fig. 15, FC has input terminals DET ₁ and DET ₂ , terminals LT^ and LT ₂ , output terminal F _Q and power supply terminals

V _DD (-33 Volts) and Vg _S (-48 Volts) . The fault indication circuit FC is only associated to the protection circuit S of the switching unit S.. (Fig. 11) and to the correspond-

ing protection circuit of the switching unit S. ₂ (not shown) . This is sufficient to detect abnormal signals of any polarity on the telecommunication line loop connected between LT. and LT ₂

Input terminal DE^ of FC is connected to the like named detection output terminal of the protection circuit S of the switching unit S.. (Fig. 11) , whilst input terminal DET ₂ of FC is connected to the detection output terminal of the protection circuit corresponding to S of the switching unit S. ~ (not shown) . The terminals LT.. and L ₂ of the fault indication circuit FC are respectively connected to the like named line terminals of the subscriber line. The output terminal F _Q of the fault indication circuit FC is connected to a digital signal processor or DSP circuit (not shown) which forms also part of the telecommunication line circuit.

The fault indication circuit FC includes an NPN transistor Q_, whose base electrode is connected to input terminal DET.. via a resistor R. ₅ and whose emitter elec¬ trode is connected to terminal LT.. The supply terminal V__ is connected to the collector electrode of Q_ via resistor R ₁₇ and diode D ₂₂ connected in series. Another NPN tran¬ sistor Q ₈ has its base electrode connected to input terminal DET ₂ via a resistor R. _g and its emitter electrode connected to terminal T ₂ , whilst the junction point of resistor R._ and diode D ₂₂ is connected to the collector electrode of Q _a via diode D ₂₃ . This junction point is also connected to the gate electrodes of an NMOS transistor N ₁ . and of a PMOS transistor P.,, the source electrode of P. being connected to V _ and the source electrode of N.. being connected to V S.S _c . The drain electrodes of P.11. and N.1.4„ are both connected to the gate electrode of an NMOS tran¬ sistor N.c whose source electrode is connected to V _ςς via a resistor R- _>8 - Output terminal F _Q is directly connected to the drain electrode of transistor N_._.

/

\

The fault indication circuit FC works as follows.

When no abnormal signal is detected by the power protection circuits of the switching circuits S of S.. and S. ₂ or when these power protection circuits are out of service, the voltage at the input terminals DET. and DET-, is not sufficiently positive, with regard to the respective terminals LT., and LT ₂ , to make their associated transistors

Q_ and Q _fi conductive. No current then flows through the diodes D ₂₂ and D ₂₃ and, consequently, through the resistor R. _η so that the voltage (V_ _D ) applied to the gate electrode of . ₄ is more positive than the voltage (V _qs ) at its source electrode. Transistor N ₁₄ is thus conductive, whilst transistor P.,, is blocked since it has the same voltage (V _nn ) at its source and gate electrodes. As a result, transistor N. ₅ is also blocked and no signal is transmitted to the output terminal F _Q . Alternatively, when an abnormal signal is detected by the power protection circuits of a switching circuit S a voltage, positive with respect to the one at terminal LT. (LT ₂ ) , appears at the input terminal DET- (DET ₂ ) of FC. To be noted that the voltage at terminal LT. (LT- is more negative than the voltage at terminal V . The transistor ₇ (Q _g ) then becomes conductive and a current can flow from- V _ to LT. (LT ₂ ) via resistor R ₁₇ , diode D ₂₂ ( ₂ _) and the collector-to-emitter path of transistor Q_ (Q_) . As a consequence, transistor - ₄ blocks and transistor P.. becomes conductive so that the voltage V__ appears at the gate electrode of transistor N _lς which becomes also conductive. The circuit N. ₅ /R _lg then generates a current which is transmitted via terminal F _Q to the DSP circuit, the latter being able to take the appropriate actions.

While the principles of the invention have been described above in connection with specific apparatus, it is to be clearly understood that this description is made only by way of example and not as a limitation on the scope of the invention.