US1487691A - Telegraph synchronizing system - Google Patents

Description

March 18 l924` ,487,691

D. MURRAY TELEGRAPH SYNCHRONI Z I NG SYSTEM Filed Feb. 15 1924 2 Sheets-Sheet 2 DONALD MURRAY,

Patented Mar. 18, 1924.

UNITED STATES- PATENT ori-uca.

DONAL) MURRAY, OF LONDON ENGLAND, ASSIGNOR A'.lIO THE WIESTER' UNION TELE- GRAPH COMPANY 0F NEW YORK, N. Y., ACORPORATION OF NEW YORK.

TELEGRAIH SYNCHRONIZIN G SYSTEM.

Application led February 15, 1924. Serial No. 693,112.

To all whom t may conce/mf.

Be it known that I, DONALD MURRAY, a subject ofthe King of Great Britain, residing at Goswell Road, London, E. C.. 1, in the county of London, England, have invented certain new and useful Improvements in Telegraph Synchronizing Systems (for which an application for patent has-been filed in Great Britain, Dec. 6, 1922), of which the following is a specification. v

lThis invention relates to multiplex printing telegraph systems, in which the correction of synchi'onism is secured from the message signals instead of from a special correcting si nal.

Correction from the message signals is old, but it has hitherto been achieved in multiplex systems in a rather complicated way, requiring considerable alteration 1n the ordinary pattern of multiplex apparatus.

The object of the present invention is to apply the method of correction from the message signals in a lmultiplex system 1n a very simple way without involving any material departure from standard multiplex apparatus such as the Baudot.

The invention is illustrated inthe accompanying drawings.

Fig. 1 is a eneral diagram ofthe 'connections of a multiplex printing-telegraph system, including stationA and station B arranged in accordance with the present invention.

Fig. 2 illustrates an electrical method of securing synchronism automatically between the two stations A and B.

Fig. illustrates a. mechanical method of securing synchronism automatically between the two stations A and B.

Correction of synchronism from the message signals appeared lirst in the Hughes printing telegra h, and I applied this method in the lVlhriay automatic printing telegraph in the year 1899 and I proposed its application to the multiplex in 1903. Picard afterwards applied it to the Baudet multiplex across the Mediterranean. The Western Union Telegraph Company in America subsequently adopted the Picard plan with modifications.

The main feature of the present invention is the combination of ,the plan used in the Murray automatic system with part of the Picard plan. In the Murray automatic system, the speed of the vibrator V2, Fig. 1, is

made sensitive to current variations in the magnet 9 driving the reed of the vibrator. This is achieved bythe use of buffer springs 7y and 8 on `each side of the reed.V The arriving message signals operate a local correcting relay l0, the contacts of which are inthe circuit'of the vibrator magnet 9. The vibrator magnet contact 13 on the reed and the local correcting relay contact 11 consequently interfere with each other and Vary the strength of the current impulses in the vibrator magnet in proportion to any increase in speed of the reed over 'the speed ofthe arriving signals. I use this plan also in the invention now being described.'

In the Picard system, alternations of current in the line are used while no signals are being sent, each channel of the multiemployed, and correct phase is securedy rectly registered, and the synchronizing mechanism is then allowed to operate again.

In4 the Western Union multiplex system, the Picard plan of reversals and rectifying by use of the back as well as the front contact of the line relay is followed with this difference, that the channels themselves are alternately positive and negative. {Each channel therefore operates with one polar-- ity of current, thus --I- -----insteadof+++-"'as in Picards plan. In other respects, the Picard plan is more or less closely followed, with the exceptionfthat the correcting inagnet revolves,thus getting more than one correction per revolution if necessary. In'the' Western Union system, like the Picard,

manual operation of the switches is required to secure correct phase.

After these explanations, the novel features of my invention will be more readily understood. I arrange the distributors 1 and .2, Fig. l, so that only one channel of the multiplex, that shown in Fig. l, has reversed polarity and operates on the back contact of the line relay. It will be seen that in the case of channel A at. A, Fi 1, station A, the five contact levers Z, in tieir position of rest, press against the marking contact bar M instead of against thev spacing contact bar S. For the other three channels at station A this arrangement is reversed. At station B printer A on channel A operates normally on the back contact 5 of the recording relay 6. In the quadruple installation therefore the sequence of polarity in the four channels when no signals are being transmitted, is as follows:-

There is therefore at the beginning and end of the channel A, a change in the condition of the line when no signals are being transmitted, and this change in the condition of the line from'negative to positive and from positive to negative effects correction of synchronism, and as it takes place on one particular channel only, it also secures correct phase automatically.

Hence, when the circuit is in synchronism and in phase, the four, printers A, B, C, D, are silent and operate only when message signals are transmitted. For synchronizing at station B `(the corrected station), various plans may `be followed. Picards method may be employed, or the lVestern Union method. Both of these involve extra contact rings on the distributors and extra brushes. This entails considerable alteration of the commutators on the distributors and also requires various relays and switches.

These methods maybe adopted without departing from the scope of this invention; but a far simpler plan is to employ the synchronizing method used i'n the Murray automatic system already referred to, in which the s eed of the vibrator V2, Fig. 1, is contro led by interference from the line signals. For thisv purpose bu'ersprings 7 and 8, Fig. 1, may be employed on each side ofy the reed of the vibrator -V2 as in the Murray automatic system, or the Delany method of a magnet or magnets at the end of the reed may be employed; or I may utilize the variation in speed of the reed due to the reaction of the reed on the driving magnet 9. In this case I may employ a soft iron armature attached to they reed with the magnet very close to the armature to secure maximum electromagnetic interference and therefore maximum variation of the speed of the reed with variation of current. With any of these devices I employ the Murray automatic system arrangement of a local correcting relay 10, Fig. 1, operated by the line relay, the baci; and 1front contacts niemeer of the correcting relay being joined together electrically as shown at 11, and the armature of the relay being allowed a fair amount of play, adjustment of play being made to suit the amount of' correction required. In Fig. 1 the recording relay 6 and the correcting relay 10 are shown operating direct in main telegraph line 15. but they can of course be arranged to operate locally from the main line relay. The armature 12 of this correcting relay 10 crosses over between. its contacts at the beginnifng and end of 'every change of polarity in the line. AS the contacts of this local correcting relay are in the sameloeal circuit as the contact 13 on the reedfthat sends vibratory impulses' into' the reed magnet 9, there is interference between the two contacts. iffthey do not open and close at the same instants. lVhen they open and close exactly at the same time, maximum current impulses flow through the vibrator magnet y9 and the reed 14 vibrat'es with maximum energy and speed, owing tothe buffer springs 7 andy 8. (In the case of utiiizing the reaction of the reed 14 on the vibrator magnet 9 above mentioned, the speed and energy of the reed are reduced with increased current.) If the relay and vibrator contacts get out of step with each other, there is interference, and smaller current impulses flow through the vibrator magnet, resulting in the speed of the reed being brought into harmony again with the speed of the line signals.

When no signals are being transmitted over the line, there are two points, one at the beginning and the other at the end of the channel A as already described atl which correction of the speed of the vibrator takes place. Experience shows that this is suicient. When signals are being transmitted, the corrections of the vibrator speed become numerous and irregular, corresponding to the position of the brush arms all round the distributor commutator, but the correcting action is the same, and correct speed and phase Iare preserved.

It is possible to operate without connecting the back and front contacts of the 4correcting relay 10, but in this case the correcting action is more 'violent and irregular and electric connection vof the back and front contacts as shown in Fig. 1 is preferable.

The usual multiplex method is employed of running the distributor at station B at a speed of about .2 per cent :faster than the speed at station A. That is to say, the vibrator at station B runs about .2 per cent 'faster than the arriving line signals. Interference between the contacts 11 of the correcting relays 10 and 13 of the vibrator magnet 9 then occurs, and the speed of the v1- brator is then modified by the alteration in the amount of current. until itis in harmony with the line signals.

The distributors 1 and 2 at stations A and B are driven by 'phonic or other motors 16 and 17 controlled by the speed of the vibrators V1 and V2.

Finally, provision must be made for auto matic phase-finding. If all the transmitters are stopped, transmitter A -at station A, Will send out the five-unit marking signal and the other'transmitters B, C, D, (not shown) will send out spacing current, and if the action of the correcting relay 10, at the receiving station is cut out by moving over a trigger catch by hand so as to lshort circuit the contacts of the relay or short circuit the relay itself, the trigger catch also reversing the connections of printer A so that it temporarilyl operates on the front' contact of the line relaylike printeis B,y C, D, then the live-marking-unit signal Will gradually travel round the distributor commutator at the receiving station. Consequently it will affect each printer A, B, C, l), in succession; but no operative action will take place on any of the printers except printer A immediately after printer D.

This trigger short-circuiting arrangement 'is Shown at station B, Fig. 1, andalso in Figs. 2 and 3. Referring to Fig. 1 station B, pressing over contact lever 18 by hand closes a circuit through the relay magnet 19, which then holds itself closed and attracts armature 20, which throws the common return Wire 27 of printer A over from the back contact 26 on to the front contact 25 of the recording relay 6. It also closes the circuit through the relay magnet 21 energized by the battery 22, which therefore attracts its armature 23 and closes contact 24, which it will be seen short-circuits the correcting relay contact 11, and thus renders the correcting relay inoperative. This condition remains until the contact at 28 is opened. This cle-energizes the relay magnets 19 and 21, and the arrangement is restored to the normal correcting condition With printer A once more operating on the back contact 5 of the recording relay 6. Printer A is provided With a mechanical device so that when the five-unit marking signal reaches the point at which it is correctly recorded as ive marking units on printer A, a lever 29, Fig. 2, is selected by the usual selector plates 35 in the printer which momentarily trips open the Contact 28 and this unlocks themagnet 19 and opens the contact 18 and thus opens the short circuit of the correcting relay 10 so that the correcting relay 10 starts acting at the right phase point. At the same time the trip action as just described reverses the action of rinter A so that' it again operates on the ack contact of the line relay. Hence printer A performsl automatically the action now performed manually on the Picard and the 'Western Union multiplex. `As long as no messages are running, corrections Will be recorded at the beginning and end of channel A during each revolution. While messages are passing corrections Will take place at the beginning and end of each change in the condition of the line.

An appreciableinterval of time elapsesl Abefore this tripping action can take place. Suppose for instance that the trip does not operate in printer A till the distributor brushes have reached channel C, then no. harm is done because the brushes only advance the usual three quarters of a degree per revolution, and that is an amount very easily corrected in the next revolution. There would be no difficulty in correcting eventWo degrees of ladvance of the brushes and this is considerably beyond any practical requirement.

If at any time synchronism fails, all that is necessary is to stop the transmitters, (as at present) and then push over the catch lever 18, Fig. 1 at the receiving station B that short circuits the correcting relay 10 and reverses the action vof printer A. Synchronism Will then re-establish itself automatically, and this will be indicated by the opening oit' the short circuit and the ing of the printers.

In Fig. 2 only magnet 19 is shown, magnet 21 being omitted, but theoperation is the same, Fig. 2 illustrating specially the printer mechanism 28, 29, and 35 for opening the circuit.

In Fig. 3 a mechanical arrangement for the short-circuiting and 'switching arrangement is shown, which avoidsthe use of the tivo magnets 19 and 21 and of the contact 28. Depressing the button 30 by hand closes the contact 24 which short-circuits the contacts 11 of the correcting relay 10. Depressing the button 3() also throws the contact lever 20 over fromthe back contact 26 to the front contact 25, thus putting the common return Wire 27 of printer A on the front or marking contact '25 instead of the back or spacing contact 26. yThe button 30 is connected to a rod 31 having'a hook32 at its end which engages with the pawl 33 pivoted at 34. At the rightmoment, as already described, printer A selects lever 29, Fig. 3, by means of the movement of the selector plates-35. The lever 29 then moves inward and strikes the tail piece 36, thereby trip-` pingthe pawl 33 and restoring all the contacts to the normal operating positions.

The advantage of my method of synchronism is its extreme simplicity. In the modern multiplex it is customary to use the Gulstad relay and a local operating relay is therefore necessary. All that is required additional for the present method of synchronism is a local correcting relay in the same circuitl as the local operating inlay. No correcting magnet 1s required in the dissilenc-` tributor and no epicyclic .correcting gear, a very material simplification. Infact, the distributor reduces to a phonic Wheel and a commutator. No extra receiving rings or brushes are required on the commutator and no extra switches or condensers or relays, as in the Picard systemv and Western Union method. It can be applied to the ordinary Baudet distributors Without changing anything except the number of segments on the commutator. Also the front commutator alone is sufficient for sending andy receiving, leaving the back commutator available for retransmission.

lHaving now particularly described my said invention and in what manner the same is to be performed, I claim 1. In a synchronous multiplex printing telegraph system, a telegraph'line, means for transmitting impulses of different polarities on said line, a multi-channel receiving distributor, printer relays connected to each channel, circuit connections whereby the printer relays in one lchannel are unresponsive to impulses of one polarity and means for changing said circuit connections,`

whereby saidrelays are made responsive to impulses of said polarity.

2. In a synchronous multiplex printing n telegraph system, a telegraph line, a vibrator at the receiving station having a natural period of vibration, a correcting relay controlled by received signals, circuit connections controlled 'by said correcting relay, whereby the vibrator position is corrected to correspond to the received impulses, and means for short-circuiting the clrcuit connections, whereby the main line correcting relay is rendered inoperative.

3. In 'a telegraph system as set forth in claim l, a recording relay, the printer relays ot said one channel being operated from the back contact of said recording relay and the relays of the other printers fromthe front contact, and means for temporarily changing the connection of said back-contact connected printer relays to the front contact of the recording relay to secure correct phase relation when synchronism is being established.

4. In a synchronous telegraph system as set forth in claim 1, a special selecting lever which when operated Vsets the normal s nchronizing mechanism into operation Wiien the correct phase point is reached.

In testimony whereof I aiiix my signatur DONALD MURRAY.

Images