US2047729A - Transmission control in two-way signaling systems - Google Patents

Description

H. J. FlS HER July 14, 1936.

TRANSMISSION CONTROL IN TWO-WAY SIGNALING SYSTEMS Filed Sept. 29, 1933 FIG/ El? REt'T/FIER w W l P M A RECTIFIER PAD VALUES- DEC/EELS A T TORNE Y Patented July 14, 1936 UNITED STATES TRANSMISSION CONTROL IN TWO-WAY SIGNALING SYSTEMS Harold J. Fisher, Port Washington, N. Y., assignor to Bell Telephone Laboratories, Incorporated, New York, N. Y., a corporation of New York Application September 29; 1933, Serial No. 691,465

3 Claims.

This invention relates to two-way signal transmission systems and particularly to two-way signal repeating systems.

An object of the invention is to improve the operation of two-way repeaters used in signal transmission systems subject to noise.

It is frequently found desirable to insert repeating or amplifying systems between two-way transmission lines in a two-way signaling system, for example, a two-way telephone system, to compensate for attenuation in the signals transmitted in opposite directions thereover. In the prior art one form of repeating system comprises two oppositely-directed, one-way repeating paths which, to enable amplification of the signals for both directions of transmission, are coupled to each other and to the two-Way transmission line by three winding transformers or hybrid coils in such manner as to form a Wheatstone bridge circuit'. As is well known, the balanced condition of such a bridge circuit is maintained by providing a balancing network in one arm of each bridge circuit which closely simulates the impedance of the associated transmission line in another arm. For any degree of balance maintained by the balancing networks, there is a critical amount of amplification that can not be exceeded without setting up a local circulation of energy known as singing. In another form of circuit in the prior art, repeating of the signals in opposite directions is obtained without necessitating balancing of the impedances of the transmission lines to avoid. singing due to local circulation of energy, by utilizing switching means responsive to the transmission of voice currents or other signals in either direction to disable the repeating path for the other direction and the associated switching means, so that signal transmission can take place only in one direction at a time. Design of the latter type of repeating system is complicated by problems such as break-in difliculties, clipping of the signals and lock-up of the switching de-' vices by line noise. 7

The disadvantages of the above systems are obviated or minimized in the repeating system of the invention, one embodiment of which utilizes in addition to means such as hybrid coils and line balancing networks for obtaining partial conjugacy of the oppositely directed repeating paths, means responsive to initiation of signal transmission in either direction to switch sufficient loss out of the repeating path transmitting the signals and sufficient loss into the oppositely directed. path so as to maintain the necessary singing margin.

The objects and advantages of the invention will .be better understood from the following detailed description thereof when read in connection with the accompanying drawing in which Fig. 1 shows schematically a two-way telephone repeater embodying the invention and Fig. 2 shows curves illustrating the invention.

The two-way telephone repeater circuit of Fig. 1 comprises a transmission path EA including the loss networks I and 2 and the one way amplifying device 3, for repeating telephonic signals in the direction from west to east between the west twoway line section Lw and the east two-way line section LE, and the transmission path WA including the loss networks 4 and 5 and the one way amplifying device 6, for repeating signals in the direction from east to west between the east two-way line section LE and the west two-way line section Lw. The oppositely directed one-way repeating paths EA and WA are connected in conjugate relation with each other and in energy transmitting relation with the west line section Lw and the east line section LE by the hybrid coil transformers H1 and H2 and associated balancing networks N1 and N2 in well known manner.

Connected across the path EA between the hybrid coil transformer H1 and. the loss networks I and 2 is the input of a control circuit comprising the signal-controlled, amplifier-rectifier device l, and the mechanical relays 8, 9 and I0 having their windings connected in parallel to the output of the device 1. Similarly, connected across the repeating path WA between the hybrid coil transformer H2 and the loss networks 4 and 5 is the input of another control circuit comprising the signal-controlled, amplifier rectifier device II and the mechanical relays l2, l3 and I4 having their windings connected in parallel to the output of the device H. The amplifier-rectifier devices may be of any appropriate type for converting the alternating current signals applied to their inputs into direct current suitable for operating the mechanical relays connected to their outputs.

The loss networks I, 2, 4 and 5 in one form, as illustrated, are identical resistance pads of the balanced ladder type each comprising two resistance elements in series in each side of the path EA or WA and a shunt resistance arm connected or adapted to be connected across the junction points between the series resistance elements in each side of the path. In the normal condition pad I is ineffective to insert loss in the path EA because the incoming signals are by-passed around the pad through the circuit formed by the normally closed switch contacts l5 and the shunt resistance arm of the pad is open-circuited by the normally open switch contacts l6 therein. Operation of relay l3 to open the switch contacts l5 in the by-pass circuit and to close the switch contacts H5 in the shunt resistance arm will render the pad I effective to insert its maximum loss in the path EA. Similarly in the normal condition, the incoming signals are by-passed around the pad 4 in the path WA, and the pad 4 is ren- 6 dered efiective to insert its maximum amount of loss in the path WA by operation of relay 9 to open the normally closed switch contacts I1 in the by-pass circuit and to close the normally open switch contacts I8 in the shunt resistance arm of the pad.

In the normal condition, pad 2 is effective to insert its maximum loss in the path EA, and is rendered inefiective to insert any loss therein by operation of relay 8 to open the normally closed switch contacts I9 in the shunt resistance arm of the pad, and to close the normally open switch contacts 29 to provide a by-pass circuit for signals around the pad. Similarly, in the normal condition pad 5 is effective to insert its maximum loss into the path WA and is rendered ineffective to insert any loss therein, by operation of relay I2 to open the normally closed switch contacts 2| in the shunt resistance arm of. the pad, and to close the normally open switch contacts 22 to provide a by-pass circuit for signals around the pad.

The normally closed switch contacts 23 in the output of the amplifier-rectifier device I are arranged to be opened by operation of relay I4 to disable the inputs of relays 8, 9 and I0, and the normally closed switch contacts 24 in the output of amplifier-rectifier device II are arranged to be opened by operation of relay ID to disable the inputs of relays I3, I4 and I5.

The system of Fig. l operates as follows:

Let it be assumed that signals are applied to the four-wire circuit from the west line section Lw and that no signals are being simultaneously applied thereto from the east line section LE. The incoming signals are impressed by hybrid coil transformer H1 on the path EA and are transmitted thereover. A portion of, the signals will be diverted into the control circuit connected to the input of the path EA and will cause operation of the amplifier-rectifier 'l causing relays 8, 9, and It] to be operated. In the manner explained above, operation of relay 8 eiT-ectively removes loss network 2 from the transmission path EA, and operation of relay 9 introduces loss network 4 into the transmission path WA. The main portion of the signals impressed on EA will be transmitted around the loss network I through the by-pass circuit including closed switch contacts l5 and around the loss network 2 through the by-pass circuit including closed switch contacts 2f to amplifier 3, and the amplified signals in the output of the amplifier 3 will be impressed on the line section LE by hybrid coil transformer H2.

Since the loss values of the two networks 2 and 4 are equal, the singing margin of the circuit is not changed but the gain of the path EA through which the signals are passing is increased by an amount equal to the loss value of the network 2 effectively removed from that path. At the same time, the gain of the path WA through which no useful signal is passing is eifectively reduced by an amount equal to the loss value of the network 4 inserted therein. Relay It! operates, substantially at the same time as relays 8 and 9, to open switch contacts'24 to eifectively disconnect relays l2, I3 and I 4 from the amplifier-rectifier II. Thus any unbalance signal energy which may be applied to the path WA through the hybrid coil transformer H2 due to imperfect balancing of the line section LE by the balancing network N2, can not cause false operation of the relays I2, i3 and I4.

Similarly, signals applied from the east line section LE when no signals are being applied from the west line section Lw will cause operation of the amplifier-rectifier II to energize the windings of relays I2, I3 and I4. These relays will operate respectively to efiectively remove the loss network 5 from the transmission path WA, to introduce loss network I into the transmission path EA, and to open switch contacts 23 in the output of amplifier-rectifier I thus preventing subsequent operation of the latter by any unbalance signal energy which may be transmitted to the path EA through hybrid coil transformer Hi from falsely operating relays 8 to ID.

While the signals from the west line Lw maintain the relay I operated, it is apparent that signals subsequently applied to the repeating circuit from the east line section LE can not gain control of the switching circuits due to the open condition of the contacts 24 in the output of amplifier-rectifier II. During this period the maximum losses of both networks 4 and are effective in the path WA. Since transmission over the path WA is not completely blocked, but only reduced in amplification by the loss networks, the signals from the east line section LE will pass through to the west line section Lw. Similarly while signals from the east line section LE maintain relay I 4 operated, the path EA includes both loss networks I and 2 which act to reduce the amplification in path EA but not to completely block that path, so that signals may still pass from line section Lw to line section LE. Thus, if the circuit is used as a telephone repeater neither talker is ever completely blocked from the other; one talker can always interrupt the other. This improves the ease of carrying on a two-way conversation. When speech is first applied to the circuit, for example, from the west line section Lw, a certain portion of the speech currents may pass through to the amplifier 3 in the path EA before the relay 8 has operated under control of the amplifier-rectifier I to effectively remove the network 2 from the path, and thus, this portion of the speech will be transmitted to the east line section LE with reduced amplification as compared to the following portion. The amount of the speech thus reduced in relative amplifications or clipped depends largely on the voltage required to cause amplifier-rectifier I to operate relay 8 if the pad 2 introduces a loss of more than -15 decibels, being reduced in proportion to the reduction in the operating voltage. The minimum values of the operating voltage of the amplifier-rectifiers are limited by the noise present on the lines Lw and LE, since, if noise operates the amplifier-rectifier for either direction steadily, the talker from the other direction can not obtain the desired amplification. If, as is often the case, the amplitude of the noise present is higher than that of the weaker but still important parts of speech, and the sensitivity of the amplifier-rectifiers is set so as not to operate on noise, operation on speech may occur at a point where significant parts of words do not receive full amplification.

The importance of this clipping depends largely on the sensitivity of the amplifier-rectifiers and the loss values of the pads switched. Fig. 2 illustrates graphically the advantage from the standpoint of allowable increase in sensitivity of the amplifier-rectifiers and usable gain obtainable with the circuit of the invention as shown in Fig. 1 for different pad values effectively Switched from the path transmitting signals to the oppositely directed repeating path to maintain the required singing margin. Curve A shows the sensitivity advantage, AS, and curve B the gain advantage, AG. Sensitivity advantage may be defined as the difierence in sensitivity in decibels required to give equally noticeable clipping with'an amplifier-rectifier switching a pad of infinite loss and an identical amplifier-rectifier switching a pad of any selected loss from the talking path into the non-talking path. As indicated by the curve A, the sensitivity advantage increases for pad values below 16 decibels and a large sensitivity advantage is obtained when the pad values are in the order of 5 decibels or less. For example, if the loss of the pad switched is 5 decibels, the amplifier-rectifier can be made 8 decibels less sensitive than if the loss of the switched pads were infinite (which is the case where the non-talking path is completely suppressed in response to the signals in the talking path) to give equally noticeable clipping of speech. Thus, if each of the pads switched in the circuit of Fig. 1 are made of value 5 decibels or less, operating difliculties due to noise on the lines are minimized. As shown by curve B, where the switched pads have values of 5 decibels the gain advantage is 5 decibels. The invention is especially adapted, therefore, for use in systems where there is a substantial amount of noise and requiring only a moderate increase in amplification. For example, the system is particularly applicable to a suburban switched trunk telephone system for use in a population center, such as New York city or Chicago.

The losses of the pads required may be computed in the following manner. Let

GE=Gain (in decibels) from line Lvv to line LE with pads I and 2 cut out of the circuit.

Gw=Gain (in decibels) from line LE to line Lw with pads t and 5 cut out of the circuit.

Bw=Balance (in decibels) of balancing network N1 against line Lw.

BE=Balance (in decibels) of balancing network N2 against line LE.

X=Loss in decibels of pads l, 2, 4 or 5 (for purposes of illustration all pads are assumed to have the same loss value).

Without loss of generality, it may be assumed for simplicity in illustration that GE=GW=G and that BE=BW=B.

Also, let singing margin be defined as the amount in decibels that the gain around the loop comprising paths EA and WA, or 2G, needs to be increased to just cause singing. Then, if it be assumed that the minimum singing margin required is 10 decibels, it can be seen that the limiting condition occurs when 2B2G+2X=10 or when G=B+X5. For a twoway repeater not employing voice operated switching to maintain the 10 decibel singing margin, 2B2G=l0, or G=B5. Thus the difference in usable gain due to the pad switching is or AG=X. The difference in usable gain is equal to the loss of the pad switched from one direction of transmission to the other. This relationship is shown graphically by curve B in Fig. 2.

In the preferred form of the invention which has been illustrated and described, two loss networks have been employed in connection with each one-way portion of the circuit. It is to be understood that it is within the scope of the invention to utilize only a single loss network in each one-way path which is normally effective but under the control of signals is removed from the path for the direction of transmission. The circuit of Fig. 1 would represent the latter form of the invention if pads l and i, and relays Q and it are eliminated.

In the circuit of the invention as illustrated and described, a particular form of resistance pad is utilized for each loss network and switching contacts controlled by mechanical relays are utilized for switching the pads in and out of the repeating paths to control the transmission equivalent of said paths. It is within the scope of the invention to utilize other means for producing and removing the required loss in the transmission paths. For example, the loss device may be a. tapped transformer or potentiometer connected in the transmission path, the values of which may be changed by switching the grid lead of the line amplifier from one tap or contact point to another. Also, the loss in the transmission path might be controlled by changing the grid bias of one vacuum tube stage of the lin amplifier to control the gain thereof in proper manner, under control of the amplifierrectifier. Any other method known in the art for automatically varying the losses in the oneway paths may be used.

The invention has been illustrated and described in connection with a voice operated twoway telephone repeater. However, the advantages of the invention may be utilized in any two-way voice or signal-operated device in which it would be desirable to switch small rather than large or infinite losses to directionally control transmission and to maintain a, margin against singing.

The invention is to be limited only within the scope of the appended claims.

What is claimed is:

1. In a telephone system comprising a fourwire repeating circuit coupling two two-wire lines subject to noise, means for minimizing the effects of noise while preventing singing in said circuit comprising voice-operated switching devices for removing loss from that side of said circuit carrying voice signals and simultaneously inserting loss in the opposite side of said circuit to maintain a desired margin against singing, the net amount of loss effectively switched from the first side to said opposite side of said circuit being within the range of values over which the amount of protection against false operation of said devices by noise fer a given reduction in speech clipping increases rapidly as the value of said switched loss decreases.

2. In a two-way telephone system comprising a four-wire repeating circuit coupling two twowire lines subject to noise, means for improving the operation of said circuit comprising switching means connected to each side of said circuit and responsive to voice signal transmission thereover to remove loss. from that side of the circuit and simultaneously to insert an equivalent loss into the opposite side of the circuit to maintain the transmission equivalent of said paths such as to function of the value of said switched loss and varies rapidly with small changes in the value of said loss.

3. In a two-way telephone transmission system, two two-way lines, a four-wire repeater therebetween consisting of two oppositely directed one-way repeating paths coupled in partial conjugate relation with each other and in energy transmitting relation with said two-way lines, and means for improving the operation of said circuit under line noise conditions comprising means normally inserting in said paths an amount of loss which is just sufiicient to prevent singing for a given amount of repeater gain, and switching means connected to each path and responsive to signal transmission therein to remove loss from that path and to simultaneously insert loss in the other one-way path, the net amount of loss thus effectively switched between the paths being within that range of values for which the amount of protection against false operation of the switching devices by noise for a given reduction in signal clipping is a function of the value of the switched loss and increases rapidly as the value of said switched loss is decreased, and the sensitivities of the switching means being set at the optimum values permitted by the selection of the proper value for the loss switched.

HAROLD J. FISHER.

Images