US3532976A - Fault detecting and correcting circuitry for crosspoint networks - Google Patents

Description

Oct. 6, 1970 H ADELAAR ETAL 3,532,976

FAULT DETECTING AND COBRECTING CIR CUI'IRY FOR CROSSPOINT NETWORKS 2 Sheets-Sheet 1 Filed Jan. 17, 1968 I II. II I II II I llllll I I I I I I I I I I i I I I II J QQNHNF I switch come from each of the four A switches. Each of the 16 A-switch outlets is coupled via a crosspoint to one a-link in each of the four pairs associated with the four B switches. This can be made in various combinations and for instance each of the 2 :16 A-switch outlets can be connected to a different combination of 4 a-links, one in each of the four pairs.

In the manner described, there is full accessibility between any of the 1024 external outlets and any of the 128 external inlets via a unique path since in each concentration plane each of the external outlets may reach any of the 16 b-links going out of that plane, each of these 16 b-links leading in turn to one of the 16 mixing planes involving the C and D switching stages while conversely each of the 16 b-links out of a mixing plane leads to one of the 16 concentration planes involving the A and B switching stages. In the mixing planes (C/ D) as in the concentration planes (A/B), there is full accessibility between the two switching stages since each of the 4 inlets of a C-switch is connected via a c-link to the 4 different D switches in the same mixing plane while each of the 4 outlets of a D-switch is connected via a c-link to the 4 dilferent C switches in the same mixing plane.

In order to establish a connection between a predetermined external inlet out of the 12 8 and a predetermined external outlet out of the 1024, positive battery will be applied to one of the 128 D-switch inlets, there being altogether 64 D-switches each with two inlets. Thus, due to the closure of a particular In contact, this positive battery potential will be applied to four crosspoint relays Dr and in order to operate one of these four marked Dr relays, a marking ground will be applied to the multiple marking wire corresponding only to the desired relay. Thus a particular md marking contact will be closed to provide ground at the cathode of a series of diodes GD, only one being connected to a relay Dr marked by a positive potential at its terminal away from the diode and this relay only will energize.

The operation of a particular Dr relay will cause the positive battery potential to be propagated to one particular c-link through the close make contact dr and since this is connected to 4 Cr relays in the C switching stage, a similar marking operation by a ground applied to the cathodes of such rectifiers as GC will enable the subsequent operation of only one relay Cr. As long as marking contact mc is closed together with marking contact ma, the operated relay Dr only finds a locking circuit through the C switching stage but upon reopening of the marking contact md the selectedrelay Cr will then operate in series with the already operated selected Dr relay. In a similar fashion, due to the closure of a marking contact mb and the subsequent reopening of marking contact mc, a selected Br relay will be operated in series with the previous two and in turn, upon the closure of a selected marking contact ma and the reopening of the contact mb, a selected Ar relay will operate in series with the previous three. Closure of make contact ar will then lead to the energization of the desired cut-off relay Cor which is biased to negative battery.

Then, in the junctor circuit, make contact can be closed so that ground will be applied to the operated Dr relay through the decoupling diode GI and thereafter the marking contacts m' and ma. can also be reopened so that the operated connection is now held under the control of the closed jcontact.

Further marking operations can occur throughout the network since the lowering of the potentials of an estab lished holding connection, i.e. they are all lower than ground means that new marking connections cannot interfere with the established ones since these involve potentials high than ground so that the marking diodes GA, GB, GC, GD afford adequate decoupling.

So far, the fault detection network FD represented in FIG. 1 as enclosed between a dotted line rectangle has not been discussed and before examining the effect of this network coupled to the various multiple marking wires connected to such marking contacts as ma, mb, me and md the effect of a faulty marking diode such as GB and more particularly a short-circuited diode will be considered, using also the circuits represented in FIGS. 2, 3 and 4.

FIG. 2 shows one of the 64 coordinate switches forming the B stage and it will be recognized to include 32 crosspoints of which only the four extreme ones involving relays Brll, Br14, Br81 and E184 are represented together with their associated make contacts and marking diodes. There are four b-link inlets, labelled from b1 to 124, for such a switch and eight a-link outlets, labelled from all to a8, to which access is secured when a crosspoint relay such as Brll operates and closes its make contact br11 thus interconnecting the latter in series with the relay winding between b1 and al. This is secured by first providing a suitable ground potential at the marking lead ml which is multipled to the crosspoints in the same way as the corresponding outlet al, this being true for all outlets and marking leads.

Assuming now that a crosspoint connected to link b1, e.g. Br81, is operated but that after the closure of make contact br81 and the establishment of a complete holding connection as discussed in relation to FIG. 1, due to marking diode associated with this link, e.g. GB11, being short-circuited, the potential present on link b1 and which, as is clear from FIG. 1 is a negative potential, will be transmitted to the marking lead m-l through Brll and the short-circuited diode GB11. This lead is not only coupled to the four crosspoints Brll 13114 of the switch represented in FIG. 2 but in fact to the analogous crosspoints in all the other 64 switches forming the whole of the B switching stage. This negative potential on marking lead m1 supplied through a relatively low resistance formed by the crosspoints thus simulates a permanent marking condition and this is applied to all the remaining 256 b-links such as link [24 through diode GB14 and relay winding Br14 if it is assumed that the links of the switching network of FIG. 1 including the b-links are normally biased by a positive potential through a relatively high resistance to indicate their availability.

FIG. 3 represents such a positive biasing arrangement. It is seen that each link, e.g. Z2, is connected to positive battery via two biasing resistors R1 and R2 in series. Thus, as long as the b-link is not involved in a connection in the switching network the only potential which it can receive is the positive biasing potential which is an indication that the link is available to establish a connection. Thus, a positive probe pulse may be applied through capacitor C to the junction of resistors R1 and R2 and this will pass through the decoupling rectifier G poled as shown to reach a common detector (not shown) and indicating that the link is available. If it is engaged in some connection, the potential at the junction of resistors R1 and R2 will necessarily be lowered and it can be arranged that in this case the reading pulse is insufficient to trigger the detector, thereby indicating that the link is not available. The positive biasing potential can be chosen with a much lower value egg. 6 volts, than the positive marking potential, e.g. 48 volts supplied through marking contact m, and R1/2 can be relatively high valued resistors as compared to the crosspoint resistances so that a negative potential or ground on a marking lead such as mil cannot send a significant current through a crosspoint relay.

Thus, with such a biasing arrangement, while a shortcircuited diode such as GBll will not cause undesired crosspoints to be operated due to the faulty appearance of a negative potential on a marking leaddue to this single faulty diode, all the free b-links in the exchange will assume negative potentials indicating that they are busy whereby no new connection can be established throughout the exchange until that using the b1 link coupled to the faulty Br11 crosspoint is released. This is obviously a very undesirable situation, since the network will be repeatedly blocked and it is not particularly easy to trace the faulty crosspoint.

One may however use a negative potential biasing the links to indicate their availability. This is represented in FIG. 4 which will be recognized as being entirely similar to FIG. 3 except that the polarities of the battery, the reading pulse and the decoupling rectifier leading to the link-available detector (not shown) have all been reversed. With the biasing arrangement of FIG. 4, the short-circuited diode GB11 will thus be unable to block the whole exchange because the negative potential present at b1 on the holding connection and which it transmits to the marking lead m1 will not create a busy condition on all the other b-links since such diodes as GB14 will now remain blocked.

Nevertheless, it still creates a permanent marking condition on this lead ml which mean-s that as soon as any other connection is set up towards an a-link not associated with m1, a double connection will arise. Indeed, in any of the B switches the positive marking potential coming from the C switching stage on the b-link will lead to the operation of two crosspoint relays in the switch considered instead of one. Assuming for instance that the positive marking potential arrives on b4 and that ground has been provided through a contact mb (FIG. 1) at marking lead m8 because it is desired to operate crosspoint relay Br84, not only will this relay operate but also relay Br14 due to the negative marking potential on m1. Thus, a connection will be established from b4 not only to a8 through br84 as desired but also to al through 11114 and a double connection will be established and held.

If the marking and holding switching arrangements described form part of a telephone exchange so that the crosspoint relays shown have their windings inserted in the test wire of a telephone connection and in addition to their make contacts shown, to hold the connection, are provided with two additional make contacts to establish the speech wire connections, it is clear that such double connections will play havoc with the telephone traffic.

Assuming a negative biasing of the links such as shown in FIG. 4, the fault detection circuit FD of FIG. 1 will prevent such undesirable situations and with simple means limit the inconvenience to subscribers on the whole, to a minim-um.

The circuit FD provides for each of the l6+8+4+4 multiple marking wires to be connected to a common detector DET. As shown the multiple marking wires on the side of the marking contacts away from the marking ground are connected through a decoupling diode, e.g. diode SA for the multiple marking wire shown connected to a marking contact ma, to the terminal of the detector DET away from ground through a break contact detO pertaining to a relay (not shown) part of this detector.

Assuming now that relay Br11 (FIG. 2) being operated and involved in a holding connection (FIG. 1) between ground and negative battery at the cut-off relay Cor, marking diode GB11 becomes short circuited, ground through the detector impedance, break contact detO, diode SB poled in opposition to the short-circuited marking diode GB11 will thus reach the closed make contact br11. Since a holding ground is applied through make contact j to relay Dr at the right-hand side of the holding connection, if the impedance of the detector has a sufiiciently low value, all three Brll, Cr and Dr crosspoint relays will be short-circuited and accordingly all three will release, involving in turn the release of the remaining Ar and Cor relays involved in the holding connection.

Thus, from that moment, the short-circuited marking diode GB11 cannot transmit an interfering negative potential to its multiple marking conductor since the connection involving this faulty crosspoint is forcedly released. After this release, the gro rid provided by the detector DET means that the potential on link [11 will be sufficiently raised, i.e. close to ground, with respect to the negative biasing potential (FIG. 4) to cause an artificial busying of this b1 link whereby the latter can no longer be selected until the fault is cleared.

A more complex situation however arises if the faulty crosspoint is not itself involved in a holding connection but if this is the case for another crosspoint connected to the same inlet, such as relay Br81 which is also connected to link b1. Thus, if there is an established holding connection between b1 and as through the winding of relay Br81 in series with closed make contact [1181, the current supplied from ground through detector DET will reach link b1 through dett), SB, ml, the shorted diode GB11 and the winding of relay Brll. This additional current path will cause an increase of the holding current flowing through the operated Br81 relay and the operated Ar and Cor relays (FIG. 1). On the other hand, it will reduce the holding current through the Cr and Dr relays. Thus, the condition of the relays on the left-hand side of the fault, i.e. Br81, Ar and Cor cannot in principle be affected but as to the relays in the right-hand part of the holding connection, i.e. Cr and Dr there are two possibilities:

(X) the holding current in the right-hand part of the holding connection is still sufficient to maintain it;

(Y) the holding current in the right-hand path of the holding connection is no longer sufficient to maintain it. As to the current flowing through the faulty crosspoint, i.e. winding of relay Brll, and which is equal to the sum of the current increase through the left-hand part plus the current decrease in the right-hand part, there are again two possibilities:

(X) the fault current is insufficient to operate the faulty crosspoint;

(Y) the faulty current is insufficient to operate the faulty make contact brll towards link a1.

Examining now the four combinations of these two pairs of possibilities, by definition possibility Xx creates no effective disturbance upon affected connection which will be normally released at the end of the conversation.

Automatically then, case Yx will arise in a transient manner, since the reopening of make contact j will be followed by the subsequent release of relays Dr and Cr.

At that moment, condition Yy will automatically arise because the current supplied from ground at the detector DET and passing through the short-circuited marking diode GB11 will flow through the winding of the faulty cross point relay Brl l in series with the operated crosspoint relay Br81 which will thus remain held in that connection in series with the other two relays Ar and Car.

Thus, following the operation of the faulty crosspoint relay Brll, upon the closure of its make contact brll, a potential higher than negative battery will appear, i.e. practically ground, on link a1. If this a-link is not engaged in another connection it will be made artifically busy and cannot thus be seized again. If link a1 is already engaged in another connection, e.g. connected to link b4 through make contact br14 and the winding of crosspoint relay Br14 the appearance of this ground potential on link a1 will as already explained cause the release of the relays on the right hand side of the connection, i.e. crosspoint relay B114 and the Cr and Dr relays involved in that connection through links a1 and b4.

Moreover, if the faulty crosspoint happens to be in the A switching stage, when it operates in the manner described, irrespective of whether the cut-off relay Cor was operated or not, i.e. irrespective of whether the line circuit was free or busy, the operation of a faulty Ar crosspoint by closing make contact ar will apply ground to the cut-off relay and always operate the latter.

Consequently, while the fault detection circuit FD of FIG. 1 will avoid a complete interruption of the telephone traflic, nevertheless the above case Yy leading to the operation of the faulty crosspoint could lead to one or two partial paths being held, extending from the faulty crosspoint to the left and each including an operated cut-off relay in a line circuit. In such a condition,

United States Patent 3,532,976 FAULT DETECTING AND CORRECTING CIR- CUITRY FOR CROSSPOINT NETWORKS Hans Helmut Adelaar, Kapellenbos, Jan Van Goethem, Antwerp, and Marcel Arthur Van Brussel, Hoboken, Belgium, assignors to International Standard Electric Corporation, New York, N.Y., a corporation of Delaware Filed Jan. 17, 1968, Ser. No. 698,541 Claims priority, application Netherlands, Jan. 23, 1967, 6701055 Int. Cl. G01r 31/02 US. Cl. 324-73 8 Claims ABSTRACT OF THE DISCLOSURE Fault detecting and correcting circuitry for use in detecting and correcting troubles such as double connections or false busy conditions caused by shorted marking diodes in glass reed switching networks. For example, the invention provides an artificial busy potential on half of the links of a stage including the faulty diode to thereby assure that the remaining half of the stage will remain operating until the shorted diode is located.

The invention relates to crosspoint network formed by coordinate arrangements of crosspoints each of which includes asymmetrically conducting impedances used to establish an operating path for the crosspoint, and more particularly to fault detecting and clearing circuitry for such networks.

Networks are known wherein coordinate arrangements of crosspoints are such that each crosspoint comprises a relay winding, a make contact for this relay and a markirrg diode, these three elements having a common point. By applying a marking potential source to the winding of the relay in series with the diode, the relay operates and may thus apply a marking potential to a subsequent switching stage so as to operate another relay in that stage whose diode is also marked by an adequate potential. If the marking potentials applied to the diodes have the same value for all the switching stages, before a relay can be operated in a given stage, it is evident that the marking potential applied to the diode in the preceding switching stage must be disconnected, so as to remove the short circuit on the relay which then operates in series with the previously operated relay in the preceding stage. This sequential withdrawal of the marking potential may in fact be avoided by using different marking potentials in each switching stage, the potential difference between successive stages corresponding substantially to the voltage needed to operate a relay plus a blocking voltage for the marking diode.

Silicon diodes can be used for the marking diodes needed for each crosspoint but although these are generally very reliable, one may not absolutely exclude one of said diodes becoming short circuited. With the present reliability figures, such an occurrence will ve very rare, but it is clear that in a relatively large exchange, the correspondingly large number of crosspoints implies that over the years, such a fault cannot be dismissed entirely and may have unfortunate results.

Indeed, in such a crosspoint network using a step-by step marking procedure, especially one in which there is a unique path between every external inlet and every external outlet, it is possible to associate the marking diodes in large groups because in each switching stage there may be a relatively large number of coordinate crosspoint arrangements and because as a marking potential reaches a selected inlet of a selected coordinate arrangement, it is merely necessary to direct it to one 'ice selected outlet out of a relatively small number which can be reached from this particular inlet. Thus, homologous outlets of all of the coordinate switches in a stage may have their crosspoint marking diodes all multipled together on a common marking lead. This means that if a marking diode becomes short-circuited and if the crosspoint is involved in a holding connection the potential present at the junction of the relay winding with the make contact and the short circuited diode and which may for instance vary between ground and negative battery depending on the switching stage concerned, will be prop- Thus, in this manner, clamping a marking lead connected to a short-circuited diode can be achieved. If a holding connection is established between ground and negative battery, going through a crosspoint relay winding in series with a make contact thereof in each stage, and if ground is used as a test potential, this clamping to ground when a marking diode is short-circuited will cause at least one of the crosspoint relays involved in the holding connection to have its winding short-circuited with the result that such relay(s) releasetjs) and this interrupts the whole connection whose remaining operated crosspoint relays release also. In such a case, the test impedance must be relatively low valued to enable the desired shortcircuit but this is not the only way to realize the test detector which may also have a relatively high impedance in accordance with another aspect of the invention.

In accordance with another characteristic of the invenagated through the short-circuited diode to all the other diodes multipled thereto. The switch inlets (i.e. the links between switching stages) are permanently biased to a positive potential to readily indicate when these links are not involved in a connection. They are then idle and available for the establishment of a new connection. However all the diodes multipled to a faulty short-circuited one are effectively conductive. Thus, all links lead ing to the preceding stage will now assume a false potential indicating that they are busy, i.e. a negative potential, if the impedance through which the positive bias is provided is substantially larger than that of the holding connection seen from the link. This means that no new connection can be established until the faulty crosspoint is released.

By modifying the link biasing potential, i.e. using a negative potential, one may avoid rendering all the links leading to the preceding switching stage busy when a marking diode is short-circuited. But this time, when this faulty element is connected to a holding connection, the marking lead corresponding to a particular outlet for all the switches in a stage will be at such a potential that this will simulate a marking condition with the result that upon a new connection being established towards any outlet marked by any marking lead other than that to which the defective diode is coupled, two diodes will be made conductive to operate two crosspoints and a double connection will inevitably result.

A general object of the invention is to remedy these disadvantages in a simple way and produce an arrangement in which short-circuited diodes can be readily detected and the faulty conditions cleared in a simple and effective manner.

In accordance with a characteristic of the invention, in a crosspoint network of the type initially defined, one end of the crosspoint impedances is coupled to a test potential via test impedance, associated with a plurality of crosspoints, in such a way that upon one of the asymmetrically conducting impedances becoming conductive in both directions, a common detector is operated.

In accordance with a further characteristic of the invention, said couplings to the test potential are provided through asymmetrically conducing test impedances oppo-- 3 sitely poled with respect to the crosspoint asymmetrically conducting impedances.

Thus, in this manner, clamping a marking lead connected to a short-circuited diode can be achieved. If a holding connection is established between ground and negative battery, going through a crosspoint relay winding in series with a make contact thereof in each stage, and if ground is used as a test potential, this clamping to ground when a marking diode is short-circuited will cause at least one of the crosspoint relays involved in the holding connection to have its winding short-circuited with the result that such relay(s) release(s) and this interrupts the whole connection whose remaining operated crosspoint relays release also. In such a case, the test impedance must be relatively low valued to enable the desired shortcircuit but this is not the only way to realize the test detector which may also have a relatively high impedance in accordance with another aspect of the invention.

In accordance with another characteristic of the invention, in each switching stage of the network, said asymmetrically conducting impedances include a relay winding in series with a diode, a holding connection may be provided through a plurality of crosspoints in series with a hold potential source, each crosspoint holding circuit comprising the relay winding in series with a make contact of the relay connected to the junction of the winding with the diode, and said test impedance being chosen high enough to allow a short-circuited diode coupled to a holding connection to communicate a potential different from said test potential to a marking lead connected to the end of said short-circuited diode away from the relay winding.

In this alternative, free links can again be made artificially busy due to a fault by biasing them with a positive potential, so that the negative potential passed by the short-circuited diode may cause the links connected thereto through a cross-point winding in series with a diode to assume a potential near that communicated from the holding connection whereby they are no longer treated as free. But in accordance with yet a further characteristic of the invention, in each switching stage the cross-points giving access, when operated, to a predetermined switch outlet, are divided in at least two groups whose diodes are multipled to separate marking conductors, whereby a short-circuited diode renders only part of the links unavailable.

Thus, in this alternative service remains ensured but on a reduced trafiic basis until, following the reaction of the detector, the fault is cleared.

This division of the marking leads is not necessary in accordance with the first aspect of the invention using a low impedance detector. In brief, in a preferred embodiment of this type, the low impedance fault current detector is coupled via diodes to the various marking leads, the diodes from the detector being oppositely poled with respect to the crosspoint marking diodes multipled to the marking leads. If a short-circuited marking diode is involved in a holding connection, this will immediately cause the detector to react and the latter may in turn apply short-circuiting potential to all the marking diodes in one stage of the network causing thereby the release of all established connections. Thereafter, this shortcircuit is suppressed and new connections may be established with the sole exception of those using the inlet directly coupled to the faulty marking diode, which inlet is now marked as artificially busy due to the test potential supplied through the detector.

Though the breaking down of all connections existing at a given moment in the exchange is a serious disturbance, this unlikely event is however a minor evil when compared to unchecked double connections extended throughout the network due to a single faulty component.

The above and other objects and characteristics of the invention as well as the best manner of attaining them and the invention itself will be better understood from the following detailed description of embodiments thereof to be read in conjunction with the accompanying drawings which represent:

FIG. 1, a four stage switching network to which the fault detecting and clearing circuits of the invention are applied;

FIG. 2, a coordinate switching stage part of the network of FIG. 1;

FIG. 3, a way to bias the links of the network of FIG. 1 with a positive potential to indicate their availability; and

FIG. 4, a way to bias the links of the network of FIG. 1 with a negative potential to indicate their availability.

Referring to FIG. 1, the latter represents in diagrammatical form a four-stage switching network enabling connections to be established between any one out of 1024 external outlets and any one out of 128 external inlets, there being a unique path for every possible interconnection. The external outlets may be telephone line circuits while the external inlets may be telephone junctor circuits and particularly an input thereof. When a connection is established, as indicated, the line circuit cut-off relay Cor conected to negative battery will be operated from a ground supplied at make contact j in the junctor circuit and passing through the decoupling diode G] in the junctor circuit and the four switching stages labelled A, B, C and D starting from the cut-off relay Cor side.

The A and B switching stages form a series of 16 socalled concentration planes because they enable to couple the 1024 line circuits to a smaller number of 256 so-called b-links which are the interconnections between the B and C switching stages. The quantities of each element shown in FIG. 1 are each time indicated between brackets next to the various elements so that it is clear that in addition to these b-links, there are 512 a-links between the A and B stages and 256 c-links between the C and D switching stages. The latter constitute so-called mixing planes because they enable the traffic from all the 16 concentration planes (A/B) to be mixed towards any desired mixing plane (C/D) out of a set of also 16 planes.

The multiplying arrows also help to indicate the number of elements and it is seen for instance that the four multipling arrows leading to the marking diodes GA, GB, 60, GD in each of the four stages A, B, C, D are marked with 16 in correspondence With the 16 concentration or 16 mixing planes. In each of such planes, there are four coordinate arrangements or crosspoint switches and this is indicated by the multipling arrows marked with 4 and immediately above those indicating multiplying towards the various concentration and mixing planes. Each of the 16 4= 64 switches in each stage is represented in FIG. 1 as enclosed between a dotted line rectangle and it includes a number of crosspoints each consisting of a relay winding such as Ar for the switch A, a make contact ar of that relay and a marking diode GA all three elements having a common terminal as indicated. The relay winding terminals away from the common point may be considered as the switch inlets whereas the make contact armatures away from the common point can be considered as the switch outlets. The multiplying arrows at the inlet and outlet conductors indicate to how many crosspoints an inlet or an outlet is multipled in a switch. Thus, for a B switch each b-link inlet is multipled to 8 crosspoints while each a-link outlet is multipled to 4 crosspoints which, in a completed rectangular arrangement of crosspoints, means that each B switch comprises altogether 4 8=32 crosspoints, as indicated by the bracketed number. A similar reasoning can be readily made for the C and D switches, but for the A switches this is an incompletely filled rectangular coordinate arrangement of 16 8 and there are only 64 crosspoints. Each A switch has 8 a-link inlets and 16 line circuit outlets leading to the cut-off relay windings Cor. In the same concentration plane, two of the 8 a-links inlets to an A switch come from each of the four B switches while two of the 8 a-links outlets from a B these conductors so that for example the multiple marking conductor mil would only serve the crossp-oints (FIG. 2) leading to the links b1 and b2 only the first of which is shown, while a separate multiple marking lead ml would serve the crosspoints coupled to the remaining links, i.e. b3 and b4- only the last of which is shown. Thus, altogether, there would now be 16 multiple marking leads for the B switching stage and a similar splitting of the multiple marking leads would be made for all the other switching stages such as A, C and D.

In this alternative, one would use the positive link bias ing arrangement of FIG. 3 to indicate the availability of a link. In such a case, if a marking diode GB ll (FIG. 2) becomes short-circuited and if this crosspoint is involved in a holding connection or if the associated b1 link is being used in such a connection, the negative potential necessarily present on link b1 would thus create a negative to the opening of make contact (FIG. 1), again case Yy discussed above will arise.

To clear the condition created by this latter case, the current detector DET may be arranged, when it reacts to a fault, i.e. a short-circuited marking diode, to break down all existing connections through the network and completely clear the latter. This may occur manually due to an alarm being given by DET or automatically. As shown in FIG. 1, it suffices that upon such an alarm being given or upon a monostable device operating in the detector, that the ground provided through the latter be temporarily interrupted at break contact detO and on the other hand through four make contacts det1/4 of the same monostable relay device, that negative battery be applied to the four multiple marking wires for the C switching stage. For all established connections this will short-circuit the Cr and Dr relays whereby all existing connections will be interrupted. Admittedly, the final interruption of all existing telephone communications is a serious disturbance but since the event will be very rare it can be considered as a minor evil as compared to the danger of an unchecked number of double connections.

As soon as the one-shot device in detector DET has performed its function as explained, break contact det recloses while make contact dell/4 reopen and the cleared switching network is immediately ready for use. Indeed, only link b1 is artifically marked busy and cannot be seized again so that there is no danger that the same unwanted situation arises again. There will now be ample time to locate and replace the shorted component. Preferably, after the one-shot automatic device has caused the opening and closure of the det contacts, the detector DET remains in an holdover condition until it is reset manually after the fault, i.e. the short-circuiting marking diode GB11, has been cleared. In order to facilitate the location of the faulty component, as is usual in such networks, break jacks may be provided in all or in major branches of each marking wire. Obviously the automatic one-shot detecting device will be particularly desirable in an unattended exchange so that communications may be immediately resumed after clearing down the network and without waiting for maintenance action to be carried out.

An alternative fault detection arrangement may however keep the network operative at all times, though on a reduced grade of service basis. Thus, this alternative solution could at worst cause only congestion during peak traffic hours if a fault occurred at that time. With this this arrangement, a detector DET should still be connected via decoupling diodes such as SA to the various multiple marking conductors, but this time the impedance of the detector should be relatively high, instead of relatively low to supply a clamping ground.

Indeed, with such an alternative, the multiple marking wires like the 8 marking conductors ml to m8 for the B switching stage (FIG. 2) are split in at least two parts as indicated in FIG. 2 by the two short lines crossing the corresponding lines will be blocked, i.e. they can neither make a call nor be called. Moreover, apart from this unwanted condition the close to ground marking potential appearing on lead m1 is a faulty permanent marking condition which may again lead to double connections being set up in the manner previously explained.

Finally, the fourth case Xy has yet to be considered, i.e the holding current in the right-hand path of the holding connection remains sufficient to hold the connection but the current through the faulty crosspoint is also suflicient to cause its operation. Thus, with the established connection through b1 and a8 remaining held operated, a path will be closed from b1 to al through the winding of Br11 in series with make contact br11 and this may result in a double connection if a1 is already involved in an established connection. But in any event, when the connection involving b1 and a8 is releaased due potential on all the b-links coupled via a crosspoint relay winding in series with a marking diode to the multiple marking wire m1. Indeed, this would be allowed due to the relatively high impedance of the detector DET which would not interfere with the transmission of such a negative potential. But only part of the b-links of the whole switching network would be marked as artificially busy due to this negative potential and all the remaining blinks coupled to the multiple marking wires mi'l to m8 would not receive this artificial busying potential.

Accordingly, by splitting the links in two parts, for instance half the junctors, half the c-links, half the b-links and half the a-links would remain available in case of a fault so that service could still be ensured. Evidently, by splitting into more than two parts, the eiIect on the trafiic carrying capacity of the network would be even smaller. On the other hand, the current flowing through the relatively high impedance of the detector DET towards the negative potential through a short-circuited marking diode would be sufiicient to cause the reaction of DET and thereby cause an alarm leading to the identification of the shortcircuited diode and to its replacement.

It 'will be appreciated that the network represented in FIG. 1 only illustrates the way in which the fault detection and clearing circuits of the invention can be applied. Various modifications and additions can be made to the network shown. For instance, the dotted line connection included in the positive potential marking branch on the right-hand side of FIG. 1 may incorporate not only a junctor selecting marking contact such as m but also transistor switches to apply the potential after closure of such a contact and to remove it after opening thereof so as to avoid such relay contacts as m from establishing or breaking a current path. A constant current regulating device may also be inserted in this branch. Also, while the c-links have been shown coupled to only two junctor circuits and more particularly to one side of such junctor circuit, the same switching stage D may of course also be used to provide a path between the opposite side of the junctor circuit and the c-links in which case the number of crosspoints in the D switches would be doubled and if the same marking contact m is used for the input or the output of a junctor circuit, distinct marking contacts md being used for such input and output connections, a decoupling diode would be inserted in series with contact m. Further the fault detection measures disclosed herein may also be coupled with other measures of like nature such as a vertification of the correct operation of each crosspoint each time this is required for a new connection.

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

What is claimed is:

1. A fault detecting and correcting circuit for crosspoint networks,

said networks comprising a plurality of stages,

each of said stages having a plurality of switching blocks formed by a plurality of coordinately arranged horizontal and vertical multiples, crosspoints comprising reed relays extending between the horizontal and vertical multiples at the overlapping of the horizontal and vertical multiples,

operating circuit means comprising marking means for supplying a marking potential at one side of said reed relays to prepare an operating path at selected ones of each of said crosspoints,

said marking means including asymmetrically conducting marking impedance means connected to one side of said reed relays,

first potential means applied to selected inputs of the last stage of said plurality of stages for providing sufficient potential difference in cooperation with said marking potential for operating selected crosspoints stage by stage,

hold circuit means operated responsive to a series of crosspoints switching through said stages so that a path is established from one end of said network to the other end of said network, said hold circuit means comprising an output hold potential difference across said established path, and

fault detecting means coupled to one side of the asymmetrically conducting marking impedance means at each of said crosspoints for detecting a short condition in any of said asymmetrically conducting marking impedance means.

2. The fault detecting and clearing circuitry of claim 1 wherein said fault detecting means comprises asymmetrically conducting test impedance means coupled to said crosspoint asymmetrically conducting marking impedance means,

said test impedance means being poled opposite to said marking impedance means, and

test potential means coupled to said test impedance means at the side thereof opposite the coupling to the said crosspoint impedance means.

3. The fault detecting and clearing circuitry of claim 2 wherein said asymmetrically conducting marking impedance means in each of said plurality of switching stages includes a relay winding of said crosspoint reed relays,

marking diode means serially connected with said windeach crosspoint comprising the reed relay winding, a

make contact of the said reed relay connected to the junction of the said winding and the marking diode, and said test potential means having a value relative to said hold potential difference and said test impedance means wherein a short circuit in one of said marking 12 diodes connected to said holding circuit means short circuits at least one of said relay windings thereof to cause the release of said hold circuit means.

4. The fault detecting and clearing circuitry of claim 3 including network clearing means operated responsive to the operation of said detecting means to temporarily prevent all crosspoint reed relay operations so that the temporary operation of said clearing means causes the interruption of all crosspoint connections established through the network.

5. The circuitry of claim 4, wherein said clearing circuitry includes means to connect a release potential to all of the marking diodes to a selected one of said switching stages, so that at least one of the crosspoint reed relay windings in the hold circuit means is short-circuited thereby causing the release of the serial connection.

6. The fault detecting and clearing circuitry of claim 5 wherein each switching stage of said plurality of said switching stages the crosspoints giving access, when operated, to a predetermined switch outlet, are divided in at least two groups, means for multipling the marking diodes to separate marking conductors, whereby a shortcircuited diode renders only part of the links unavailable.

7. The circuitry of claim 3 wherein said test impedance means has a relatively low impedance value and thereby causes a short-circuited marking diode to indicate that a link connected thereto via a crosspoint is busy.

8. The circuitry of claim 3 wherein said test impedance means has a relatively high impedance value and thereby enables the potential of a busy link coupling said switch blocks to be transmitted through a shortcircuited marking diode coupled thereto to all free links connected to said short-circuited diode via a crosspoint impedance whereby the potential of said free links assumes a value indicating that they are busy.

References Cited UNITED STATES PATENTS 2,978,641 4/1961 Voegtlen 324158 3,084,326 4/1963 Mitchell 324158 3,200,392 8/ 1965 Chumakov. 3,206,729 9/1965 Longcroft 340-166 3,363,064 1/1968 Sperlich.

RUDOLPH V. ROLINEC, Primary Examiner E. L. STOLARUN, Assistant Examiner US. Cl. X.R.

Images