US2284085A - Electric signal transmission system - Google Patents

Description

ELECTRIC SIGNAL TRANSMISSION SYSTEM Filed Oct. 26. 1939 4 Sheets-Sheet 1 T 20 we lNVENTOiQ Jomv 001.1. RD

BY g

ATTORNEY y 1942- J. C'O LLARD 2,284,085

ELECTRIC SIGNAL TRANSMISSION SYSTEM Filed Oct. 26, 1939 4 Sheets-Sheet 2 /5 I W I3 g T T T417 INVENTOR JOHN COLLARD y 2 1.942- COLLARD 2,284,085

ELECTRIC vSIGNAL TRANSMISSION SYSTEM Filed Oct. 26, 1939 4 Sheets-Sheet 4 16' I MP lk 7 1 71% 73 V r 70 72 I 1' Q r81 l az V sa, 514 v I INVENTOR JOHN COL LA-RD ATTORNEY Patented May 26, 1942 ELECTRIC SIGNAL TRANSMISSION SYSTEM John Collard, Hammer-smith, London, England,

assignor to Electric & Musical Industries Limited, Hayes, Middlesex, England, a company of Great Britain Application October 26, 1939, Serial No. 301,324 In Great Britain October 19, 1938 15 Claims.

This invention relates to systems for the transmission of electric signals, such as television signals. It provides methods and means whereby low-frequency components difierentially attenuated as compared with higher frequency components in the course of transmission may be re-established, or wherein interference components of low frequency arising in such systems may be substantially removed. It may also be employed for compensating for varying attenuation such as fading in signals possessing two sets of recurrent fixed-level or datum portions such as television signals.

It is well known that for the transmission and reception of television signals a particularly wide frequency band is required which includes very low frequencies and even a direct current component. It is, however, usually uneconomical to construct amplifiers to deal with the direct-current component and even the components up to two or three hundred periods per second often present difiiculties. The signals at the output of an amplifier are, therefore, very often lacking in low frequency components, and some method for re-establishing them is needed in order to obtain satisfactory reproduction.

A problem very closely allied to that of such re-establishment is the one of removing unwanted lowfrequency components due to induction from neighbouring power circuits. the addition of a certain frequency component in this way is equivalent to subtracting an equal component of opposite sign, the solutions to D. C. re-establishment also serve to eliminate effects of similar low frequency that may arise by induction, for example, in neighbouring power circuits or such low-frequency effects as tilt in which case the present invention provides an. improvement inthe method of tilt elimination as described in British patent specification 490,205. 4

Several methods have been suggested for reestablishing direct and low-frequency compoments and have been described in British patent specifications 422,906, 449,242, 464,828 and 507,- 239, for example, but these have only been successful up to comparatively low frequencies such as 50 cycles per second. In cases, however, where several transformers may have to be included in the circuit even components of several hundred periods per second may be so far attenuated as to require some form of re-establishment. Furthermore, unwanted frequencies up to one or two thousand periods per second may be produced in the circuit by induction from the Since higher harmonics of power systems. It is evident, therefore, that some form of re-establishment is often required which is capable of operating up to frequencies of the order of 1000 cycles per. second. This form of re-establishment is provided by the invention, and besides being applicable to television signals is capable of application to any signals possessing recurrent datum level periods.

According to the invention there is provided a 1 system for the transmission of electric signals having recurrent fixed-level or datum portions in which in the course of transmission undesired changes occur in said signals comprising means for providing a corrective signal which is a substantially continuous function of time and is dependent on the simultaneously observed levels of a plurality of different datum portions and for utilising said signal to correct substantially for said undesired changes in said signals.

The invention may be carried into effect according to one aspect which may be referred to as a "delayed or interpolated method, or according to another aspect which may be referred to as an extrapolated method.

Before, however, proceeding to give a detailed description of ways in which the invention may be carried into effect according to either of these. aspects the broad basis of the invention will be explained first more fully.

Thus, as is well known, television signals consist of alternate line and synchronising signals, the line signals varying according to the picture being scanned and lasting for about microseconds, the synchronising pulses lasting for about 10 microseconds and having a definite magnitude and being followed by a black-level period of about 5 microseconds duration. If upon such a waveform a single-frequency induced component is superimposed, it will be impossible during the line signal period to decide how much of the received signal is due to the induced component and how much to the actual line signal. During synchronising pulse or blacklevei periods, however, the matter is different and by taking the diflerence between the received magnitude of the pulse or height of the black-level and the magnitude that should exist it is possible to determine the amount of the induced component during that brief interval.

Existing methods of D. C. re-establishment in effect measure the discrepancy in this way and then apply a substantially constant and opposite effect during the following line signal. If the 2 7 variations in synchronising-pulse level, for ex ample, are of such low frequency that there is practically no change in magnitude over the duration of the line signal, then the restoring eil'ect applied in this way will successfully annul the departure'thathas arisen. At frequencies.

greater than about 50 cycles'per second, however, it is inaccurate to assume that there is subable to provide a restoringefiect which varies over the line interval so as to supply a more exact correction.

In its simplest form the invention consists basically in eflfecting a delay corresponding to the interval between two synchronising pulses. The

some time t' after the time ti when m is measured may be expressed by the series:

A(A I 2!" where- This series is well known for'use in interpolating between values given in tables. It illustrates very clearly the various approximations obtained by existing and by the new methods of D. C. reestablishment. In previous methods where only 121 is obtained, the first term only arises; thus:

This equationcorresponds to so-called clamp" method of D. C. re-insertion described in the specification of British Patent No. 449,242.

received signals are therefore passed through a delay network which is so proportioned that when one synchronising pulse has just reached the end of the network the previous pulse is just entering the network. j It is thus possible to measure simultaneously the difference between the actual and correct level of these two pulses. It may be assumed that the changes in this difference are relatively low in frequency so that the diiference can be considered as varying linearly over the line interval. The correction necessary in order to restore D. C. or eliminate induction over the line interval thus'has a value which starts at in. say, and changes linearly to 0:,where in is the value determined by observation of the synchronising pulse before the line interval and v: is the value determined by observation of the pulse after the line interval. This restoring signal is added to the signals as they leave the delay system.

When the whole of a line has emerged from the delay system, the synchronising pulse originally at the input of the delay has reached the output and a new pulse is just entering the input. The whole process can thus be repeated. It will be realized that this method enables higher frequencies to be handled than with earlier methods, because with the earlier system the frequency needed to be so low that there was no appreciable change in the magnitude of the required correction over the line interval with the new method the frequency need only be low enough for the required correcting signal during the 'line interval to be regarded as varying linearly.

If the frequency of the induced signal exceeds about 1000 cycles per second, however, an error is introduced by assuming that its magnitude varies linearly over the line interval; by an extension of the methods of this invention, however, this error can be reduced as far as may be desired. Thus, if the signals are delayed for a time corresponding to a number of line intervals, 0

it is possible to determine the magnitude of the required correction at each of the synchronising signals occurring within, the total delayed period. In this way a series of values 221, v2, v3, etc., for the correcting potential are determined at uniform intervals T, where T is the time between successive synchronising pulses is thereby obtained. Now if a series of values m, m. m, etc., of a variable are given the value of the variable at In the first application of the new method described here v: and m are obtained, so-that the second term may also be. used. Thus a second approximation is made according to- In the further application, of this new method in which v1, 02, 12:, etc., are obtained any desired number of terms of the series may be employed giving closer and closer approximations to the true value of v.

In order that the invention may be more fully understood the same. will now be described more particularly by way of examplel'with reference to the accompanying drawings in which Figures 1, 2 and 3show-circuit arrangements whereby methods of the delayed" kind according to the invention may be carried into effect; while a Figures 4, 5, 6 and 'I'illustrate circuit arrangements wherein "extrapolated" methods are applied.

In a number of these arrangements it is necessary to provide some form of delay means so that the signals may be delayed by the requisite amount.

V Two methods are available for producing this delay. One is to use any of the well known electrical delay networks. If this is adopted it may be necessary to correct for a large amount of distortion and attenuation by means of suitable equalisers and amplifiers, since the requirements demanded of such a network, namely, a delay time of theorder of several hundred microseconds over a frequency range up to about 2 megacycles per second, are exceptional. The delay means may conveniently comprise a suitable length of. television cable and its associated equipmen The other methodof obtaining the delay is to use a device allied to the electron camera in which two beams travel continually over'a circular track. The first beam lays down the signals and the second picks them up again after the required interval as described in the specification of British'Patent No. 471,913. a 1

Referring to Figure 1 there is shown a functional circuit for applying the invention in its simplest form. The signals are applied at the terminals III, II. From there they pass through the delay network II which is assumed to be distortionless and free from loss, and so reach the output terminals l3, It by way of the condenser l5. Consider a state of aifairs where one synchronising pulse is just leaving the delay network I! while the following one is Just entering it. At this moment switches I6 and I1, which connect the terminal III to the terminal H through the condenser 13 and the battery l3 as shown in the figure, are closed. The battery I! has an E. M. F. equal to the potential to which the terminal l should be raised, that is, the potential to which the terminal I! would be raised if the synchronising pulse had its correct magnitude. If the pulse which has just reached the terminal III has not its correct amplitude the condenser l3 will be charged to an extent equal to the difference between the amplitude the pulse should have and that which it actually has. In the same way condenser' l5, by virtue of its connection through the lead 20 and the switch H to the battery l3, will be charged to a corresponding extent by the pulse which has just arrived at the output of the delay network. Since the signals are taken through condenser Hi the signal amplitude at the terminals 13, I4 will consist of the signal amplitude at terminals III, II together with the potential diflerence across the condenser l5. If the condensers l8, 15 are connected together and if it is assumed that the former is large compared with the latter then the former will eventually charge the latter to its own potential. If the connection is effected through the resistance 2| it is possible by suitably proportioning the resistance 2| in relation to condenser l8 and condenser l that the latter reaches substantially the potential of the former in a time equal to one line interval. Thus condenser l5 will add to the line signals a potential which starts at some value 111' and gradually changes to some different value '02.

If, however, only a resistance is used to connect the condensers l8 and Hi the potential difference across the latter will change exponentially from m to on so that the requirement of a linear variation will not be fulfilled. This difilculty may be met by connecting an inductance '22 in series with the resistance 2i, By suitably proportioning this inductance it is possible to arrange that the change in the potential difierence across the condenser I5 is substantially linear.

The condensers I5 and I3 constitute in eilect observing devices and enable a correcting signal.

taneously observed levels of a plurality of diflerent datum portions.

In practice the switches 16 and I5 may be replaced by valve circuits arranged to perform the same operations. Various forms of valve switches have been proposed in connection with so called clamp" D. C. re-establishment, and in particular the six-diode switch described in British Pat ent No. 512,109 may be applied here.

Figure 2 shows an alternative arrangement for obtaining a linear restoring. potential. As before there are the input and output terminals III, II and I3, l4 respectively connected through the delay network l2, the condenser l5 being connected in the lead from the delay network to the terminal 13. The terminal Ill and the terminal of condenser connected to the output of the delay network are connected respectively through the switches 23, 24 and the condensers 25, 26 to ground, these condensers forming the observing devices in this example of the invention. The plates of these condensers not con nected to earth are taken to the control grids 21, 28 of the valves 29an'd 30 respectively. These valves have a large resistance 3| connected in their common-cathode circuit, so that the potential of either anode 32, 33 is proportional to the difference of the potentials on the two grids.

The anode 33 is connected through the condenser 34 and the resistance 35 to the output terminal l3. the condenser 34 being of large capacity compared with that of condenser IS. The output terminal I3 is connected through the switch 33 to the battery 31 and through the battery to ground. At the moment one synchronising pulse reaches the end of the delay network l2 and the following one is entering the beginning, the switches 23, 24 and 35 are closed. Condenser thus becomes charged to a value corresponding to 0:, say, and. condenser 26 some other value corresponding to m. The potential change arising at the anode 33 of valve 33 is proportional .to mvi.

Accordingly, the charging current flowing from valve 33 through the condenser 34 and the resistance 35 to condenser I5 is also proportional to 122-111. At the beginning oi the line the condenser I! has a potential difference between its terminals equal to '01 due to the charge obtained from the battery 31 through the switch 36. By suitably adjusting the magnitude or the resistance 35 it is possible to arrange that the steady charging current from valve shall increase the potential diflerence across condenser l5 to 02 at the end of the line. Thus condenser l5 supplies the correct restoring potential starting at m and increasing linearly to m. In order that the current from valve 30 shall be independent of the potential changes of condenser I5 during the line interval, it is necessary to arrange that the potential changes on the anode of valve 30, which produce the charging current, shall be large compared to the changes occurring in condenser 15. The gainof valves 29 and 30 is therefore made large and the resistance is also made large.

It was pointed out that a better approximation to the correct restoring potential could be obtained by observing a. larger number or synchronising pulses. If three pulses are observed the expression for the restoring potential is The circuit corresponding to this equation is shown in Figure 3. The arrangement is similar to that 01 Figure 2 except that two delay networks l2 and I2 are used, there being connected to these delay networks in manner precisely sim ilar to that of Figure 2, the pairs of valves 29,30 and 29', 30 respectively. In similar manner also the anodes 33 and 33' of the valves 30 and 33' are taken'through series resistance- capacity circuits 34, 35 and 34, 35' respectively to the terminal l3 with which is associated as before the switch 36 and battery 31.

The anodes 33 and 33' are, however, also connected through series resistance- capacity circuits 33, 33 and 33', 39' to the grid 40 of the valve whose cathodeis earthed. Between the grid 40 and earth there is connected the condenser 42 and across this condenser is arranged the switch 43 and battery 44 in series. The anode 45 of valve 4| is connected through the condenser 46 and resistance 41 to the terminal l3.

In order to relate the above expression to this circuit arrangement-it may be written in the re-arrangedform:

proportional to (tn-ml-(m-m).

.',I'he"term 121 is obtained by closing switch a operating in .the same way as in Figure 2.

-' *Th tworesistances 3i and are adjusted in the roportion of three to one and the gain of all four valves' is arranged to be substantially the same so that-the char ing currents from valves 30' and 3. are in the correct ratio. The anodes of these valves are also connected to condenser 42 through large condensers and resistances. Hence the total charging current into condenser 42 is The potential diilerence across the condenser 42 is thus equal to {(112-121) -(va-v'z) }A. Since the grid of valve 4| is connected to condenser 42 the anode voltage of valve 4| is therefore proportional to '{(vzv1)-(v:-vz)}A. Also as the anode of valve 4| 'is connected to condenser I! through the large condenser 46 andresistance 41., condenser AIS is therefore supplied with a charging current proportional to {(v:v1)(v:m) }A. The potential. difference built up across condenser l5 .due to this currentis accordingly proportional topotential -{(vz-v1)(v:.vz)}A. The total diiference across condenser I5 is thus made to vary in accordance with the full expression given above for v and is thus correct to a secondorder of approximation throughout the line interval.

' Since the process is required to siart fresh at the .end of each line the switch 43 is provided to .short' circuit condenser and'so prepare it for the 'new charge. The battery 44 is used to avoid short circuiting the gridof valve 4ltoearth.

The steps involved in the extension of this ment has the disadvantage that it is not so accurate as the delayed method. This follows naturally from the fact that the "delayed" method depends uponinterpolating between certain values while the other method operates by extrapolating beyond the last observed point.

The first stage of'the extrapolated"- method is the same as'the original -clamp" D. C. re-establishment. (see British'latent No. 449,29). It consists ofrneasuring the restoring potential in for the synchronising pulse immediately preceding the. line interval. and then applying a constant potentialofthis magnitude throughout the following lineinterval. This assumes, of course.

. per cent, where T is the time interval of one'line' two successive synchronizing ply a restoring potential of:

that the restoring potential remains constant during the line interval, which is only true for very low frequencies. If the restoring potential is given by v=A sin out, then the maximum error produced by applying the constant voltage in is:

The second stage of approximation is to observe the restoring potentials in and 0: during pulses and then ap- =v1) where t is the time interval after the last observed synchronising pulse and varies from zero to T. This is equivalent to assuming that the mean slope of the restoring potential during the scheme to include still further terms will be obvious from the above.

It will be appreciated from the foregoing illustrations that the chief feature of the invention as carried into eiiect according ,to delayed methods is that the restoring potential is determined from observations made on synchronising pulses before and after the line interval for which the restoring potential is to be used. This neces-' sitates the delaying of the line signals, which is a disadvantage as it is comparatively expensive to construct a delay system to give a uniform delay up to 2 or 3 megacycles. To avoid this difllculty "extrapolated methods of re-establishment which will now be illustrated, maybe used. In these a number of successive synchronising pulses are observed and the observations are used to set up the restoring potential'for the line interval following immediately after. the last observed synchronising pulse. Thus no delay of the line signals is required. In order to be able to observe simultaneously'a number of synchro nising pulses, these may be passed through a delay system tapped at suitable points as in the case of the delay networks used for delaying the line signals in delayedire-establishment. Since the delay network has now only to deal with synchronising pulses and not line signals it may possess characteristicsof much less perfect form with a resultant economy. Furthermore, as will be shown later, there is a methodof dispensing with the delay network altogether. Hence extrapolated" methods of re-establishment are considerably more simple to effect than delayed" methods. However, extrap9lated" .re-establishline when the correction is being applied, is the same as the mean slope in the previous line interval. This assumption is valid up to somewhat higher frequencies. than the previous assumption and is therefore an improvement upon clamp" D. C. rte-establishment. The maximum error per cent is:

e,=400 Sill '5' A' yet closer approximation is obtained by observing successive synchronising pulses,

thus deriving three consecutive correcting potentials v1, or and m. The restoring potential that can then .be applied is:

The corresponding maximum error per cent by:

showing thev extent to which the approximation is better than the previous one.

The process can, of course, be carried still further by oburving additional synchronising pulses. The maximum errors for the first three approximations given above are tabulated below for various frequencies.

is expressed re-establishment and it will the closer approximations give considerably better results except at the higher frequencies.

It will be noticed that the variable terms so far considered have all been linear in it, so that the resultant restoring potential varies as a straight line throughout the line interval. It will be obvious that by taking terms proportional plying the second approximation of the extrapolated method, i. e., for producing a restoring potential of form shows diagrammatically how this may be carried out.

As in previous circuit arrangements there are the input and output pairs of terminals l0, H

ing a large cathode impedance 58. At the anode 59 of valve 56 there accordingly appears a potential proportional to vz-m as will be appreciated from previous explanations. This anode is taken through the large condenser and resistance 60 and 6| respectively to the terminal l3, which, as before, is connected through the switch 36 to the battery 31 and thence to earth.

At the instant when a given synchronising pulse reaches the beginning of the delay network 48 the switch 36 is closed, thus charging condenser 15 to an extent corresponding to m. Switches 50 and 5| are also closed so that condenser 52 receives a charge depending on or while condenser 53 which is connected to the far end of the delay network 48, receives a charge dependingon v1. The delay in this case is, of course, equal to the interval between two con- The terminals l6 and I3 are joined through the condenser l5, and terminal I3 is connected to earth through the switch 36 and battery 31 in series as in Figure 4. Terminal id is however, taken to one of three contacts 62, 63, 64 namely 62. Associated with these contacts is a rotating system of threeobserving condensers 65, 66, 61 having a clockwise motion and being connected in a star arrangement with the central point earthed. The free plates are periodically brought into contact with the points 62, 63,,

64 by means of brushes which are indicated by arrows. The contacts 62 and 63 are connected to the device 66 which comprises a pair of valves arranged in exactly the same manner, for exam-' ple, as the valves 56 and 61 in Figure 4 and performing the same function. From the output of this device connection is made by a large condenser and resistance, 16 and 1| respectively, to terminal i3. The contacts 63 and 64 are connected to the terminal I3 in a precisely similar way, namely, through the device 69, similar-to the device 66, and through the condenser 12 and resistance 13.

When the first synchronising pulse reaches the condenser l5 the switch 36-is closed so that consecutive synchronising pulses. a The two valves 56, 51 as in the delayed D. C. re-establishment, supply a charging current to condenser 15 through condenserv 60 and resistance 6 l' proportional to vz-m, and therefore provide the term To apply this method to the third approximation, 1. e., using a restoring voltage of:

merely requires an obvious extension of the circuit of Figure 4 using two delay networks of one line interval delay each and two pairs of valves each like the pair 56 and 51.

Since this method depends on observations taken only on synchronising pulses occurrin before the lineinterval for which the restoring potential is required it is possibleto dispense altogether with the delay system. This provides a. considerable gain in simplicity. Figure 5 15 now becomes charged to a potential 122. Condenser 65 also receives a charge depending on 122. Just before the third synchronising pulse reaches condenser I5, the condenser assembly is rotated again through When the third pulse arrives the switch 36 is closed again so that condenser I5 is now charged to a potential vs. Condenser 66 also receives a charge depending onm, and condenser 65 retains a charge depending on or. The position now is that condenser 61 has a charge depending on m and is connected to contact 64, condenser 66 has a charge depending on m and is connected to contact 62, and condenser 65 has a charge depending on '02 and is connected to contact 63 Since the device 68 is connected to contacts 62 and 63 a charging current is therefore given to condenser 15 proportional to 03-02. By reason of the similar device 69 there is also given'a charging current to condenser '15 proportional to 02-01. Hence by suitably adjusting the gains of the two devices 68 and 69 and the resistances 1|, 13, it can be arranged as previously described that the p tential of condenser l5 has the correct value throughout the line interval. For. subsequent synchronising pulses, the series of operations continues in a similar fashion.

A way in which this scheme may be put into practice without the use of moving mechanical parts is shown in Figure 6. This arrangement is identical with that of Figure 5 except that the contacts 62, 63,64 are replaced by leads havin the same reference numbers, the condensers 65, 66 and 61 are now fixed and the brushes are replaced by the array of switches shown. Switches 15, 16, 11 are capable of connecting the condensers as, u, a respectively to the lead "c2.

7 Switches 18, 18, 88 perform asimilar function with respect to the lead 68 as do switches 8|, 82, 88 with respect to lead 64. Switch 14 connects terminals l8 tolead- 62. The various switches, would, of course, consist of suitable valve circuits as already discussed.

When the first synchronising pulse reaches .c9ndenser,l 5-the switches 14, 86' and 15 denser 81 has a charge depending on 01, condenser 88 a charge depending on v: and condenser 65 a charge depending onm. Thus the same state of affairs is arrived at as with the.

previous and more complicated switching scheme.

on m. At the third synchronising pulse the switches 14, 86, 11, 18 and 81 are closed. 14 being opened before the line signal .begins. This leaves condenser 65 with a charge depending on m, condenser 86 with a charge depending on m,

condenser 61 with a charge depending on v: and' condenser IS with a potential equal to 12:. Also device 68 is connected to condensers 61 and 66 and therefore gives a charging current to condenser I 5 proportional to try-vs, while device 88 is connected to condensers 66 and 66 and therefore gives a charging current to condenser l5 proportional to lb-U1. Thus'the correct potential difference across condenser I! may be main-' tained throughout the line interval, and-for subsequent synchronising pulses, the series of operations is continued.

The disadvantage of the above scheme is its obvious cumbrousness and a greatly simplified arrangement is shown in Figure 'l. The condensers 65, 6t and 61, as previously, each have one plate connected to earth but the opposite plates are connected permanently to the leads. 82, 68, 64 respectively. The previous array of switches is replaced by the three switches 84, 85, 86 and the two cathode follower valves 81 and 88 having cathode resistances 88 and 84. Switch 84 connects terminal II to the unearthed plate of condenser 65 and the grid," of valve 81. Swith 85 connects the cathode 8| of valve 81 to the unearthed plate of condenser 88 and the rid 88 of valve 88; Similarly switch 88 connects the cathode 82 of valve 88 to the unearthed plate of condenser 81. 7

At the firstsynchronising signal switch 84 is closed and then opened again; this gives a charge to condenser 85 depending on the potential in. As in previous cases switch 88 is used to give condenser IS the correct potential in, D2, in, etc., as the case may be, and its use, therefore, need not be dealt with again. When the second synchronising pulse arrives switch 85 is first closed. Since valve 81 is a cathode follower this causes condenser 66 to be charged to substantially the same voltage as condenser 65. Switch 85 is then opened, and switch 84 is closed for a short time and opened again, causing condenser 65 to have a charge depending on m. Hence condenser 66 now has a charge depending on m and condenser 65 one depending on in. When the next pulse arrives switch 86 is first closed causing condenser 61 to receive a charge depending on in. Switch 86 is then opened and switch 85 is closed, causing condenser 66 to receive a charge depending on Us. Switch 85 is then opened, and switch 84 is closed and opened again, causing condenser 65 to receive a charge depending on us. Thus con- It will have been appreciated that the series of switching operations requires to be completed before the synchronising pulse is over and this can be eifected by passing the pulse down a delay network. Across this network are-bridged three valves causing the switches, 85, 86 to operate. As the pulse reaches each valve the corresponding switch is caused to operate. By suitably selecting the tapping points the various operations can be made to take place at the correct times. The method of arranging such a delay network and switching device has been already described in connection with clamp D. C. re-establishment for example in British patent specification No. 449,242. v

The arrangements described, it will have been noticed, have all been based on observations on the magnitude of synchronising pulses. As, however, stated earlier an alternative method is to observe the level of the short black interval which follows each synchronising pulse and precedes the line signal. The previous methods all apply equally well to this case except that, of course, the various switches must be closed during the black interval instead of during the syn--' chonlsing pulse interval.

It is possible in some instances to combine the methods of delayed" and "extrapolated" reestablishment with advantage. For example, Figure '3 shows a method of applying delayed D. C. re-establishment in which three successive synchronising pulses are observed. These values are obtained by two delay networks each giving a delay equal to the interval between synchronising pulses. If. v1, or and in are the three restoring potentials observed then m is obtained by means of one delay network and v: by means of two delay networks. The restoring potential derived from these is applied to correct in the line interval between or and D2. It is possible, however, instead of observing the potential v: to observe and store the synchronising pulse prior to that giving the potential or, thereby deriving a correcting potential say. This can be done by using the methods described for "extrapolated D. C. re-establishment. Then the potentials v0, v1 and or can be used to establish the restoring potential for the line interval between the pulses corresponding to m and or, instead of employing the potentials in, D: and v: as previously. The advantage of this is that the second delay network is now not needed. In "extrapolated methods the restoring potential may be applied in any part of the transmission path along which the signals are transmitted either before or after the point from which the correcting signal is derived.-

The invention may be used in overcoming the eflect of fading in television transmission (see British Patent No. 458,585) by employing the methods of the invention to develop a gain-control potential which depends on a plurality of datum levels simultaneously observed and varies continuously, or substantially so, in like manner to the D. C. restoring potential as described above, this development of a gain-control potential, of course, depending upon the previous reestablishmentin the signalsof the direct and slowly varying components by means of another set of datum levels in the signals. difierent from those to be used in deriving the gain-control potential. The television waveform possesses two full sets of datum levels; namely the peaks of synchronising signals constitutes one, the black level datum portions another. For D. C. re.-in-

serting, of course, and interference elimination either may be used as desired as already stated; this still applies when a gain-control potential is likely to be developed.

I claim:

1. In a television system having video signals combined with interspersed synchronizing impulses and in which direct current and low Irequency components of the video signals are suppressed which combined signals and impulses are subjected to distortion in transmission, the method of correcting signals representative of the suppressed components derived from the synchronizing signals at a receiving point which includes the steps of receiving the combined signals and impulses, comparing the received synchronizing impulses with each other, deriving energy representative of the difference of the compared impulses, and adding progressively varying energy to the received signals during the time interval between the synchronizing impulses, the variation of the energy being between the limits of zero and a maximum value proportional to the difference between the compared impulses.

2. In a television system having video signals combined' with interspersed synchronizing impulses and in which direct current and low frequency components of the video current and low frequency components of the video signals are suppressed which combined signals and-impulses are subjected to distortion in transmission, the method of correcting signals representative of the suppressed components-derived from the synchronizing signals at a receiving point which includes the steps of comparing the received successive synchronizing impulses with each other, deriving energy representative of the difierence of the compared impulses, and adding progressively varying energy to the received signals during the time interval between the synchronizing impulses, the variation' of the energy being between the limits of zero and a maximum value proportional to the diflerence between the compared impulses.

3. In a television system having video signals combined with interspersed synchronizing impulses and in which direct current and low frequency components of the video current and low frequency components of the video signals are suppressed which combined signals and impulses are subjected to distortion in transmission, the method of correcting signals representative of the suppressed components derived from the syn-1 chronizing signals at a receiving point which includes the steps 01 comparing the received successive groups of synchronizing impulses with each other, deriving energy representative of the difference of the compared impulses, and adding progressively varying energy to the received signals during the time interval between the synchronizing impulses, the variation of the energy being between the limits oi! zero and a maximum value proportional to the difference between the compared impulses.

4. In a television system having video signals combined with interspersed synchronizing impulses and in which direct current and low frequency components of the video signals are suppressed which combined signals and impulses are subjected to distortion in transmission, th method of correcting signals representative of the suppressed components derived from the synchronizing signals at a receiving point which includes the steps of receiving the combined signals and impulses, delaying the combined signals and impulses, comparing delayed synchronizing impulses with undelayed synchronizing impulses, deriving energy representative of the difierence of the compared impulses, and adding progressively varying energy to the received signals during the time interval between the synchronizing impulses, the variation of the energy being between the limits of zero and a maximum value proportional to the diiierence between the compared impulses.

5. In a television system having video signals combined with interspersed synchronizing impulses and in which direct current and low frequency components of the video signals are suppressed which combined signals and im'pulses are subjected to distortion in transmission, the method of correcting signals representative of the suppressed components derived from the synchronizing signals at a receiving point, which includes the steps of receiving the combined signals and impulses, comparing the received synchronizing impulses with eachother, deriving potential mcreasing linearly with respect to time between the limits of zero and a maximum value in proportion to the difl'erence of the compared impulses, and adding said derived potential to said received combined signal and impulses.

6. In a television system having video signals combined with interspersed .synchronizing impulses and in which direct current and low frequency components of the video signals are sup-- ing signals at a receiving point which includes the steps of receiving the combined signals and impulses, comparing the received synchronizing impulses, deriving potential increasing parabolically with respect to time between the limits of zero and a maximum value inproportion to the diil'erence of the compared impulses, and adding said derived potential to said received combined signal and impulses.

7. In a television system having video signals combined with interspersed synchronizing impulses and in which direct current and low frequency components of the video signals are suppressed which combined signals and impulses are subjected to distortion in transmission, the method of correcting signals representative of the suppressed components derived from the synchronizing signals at a receiving'point which. includes the steps of receiving the combined signals and impulses, delaying the received combined signals and impulses, comparing delayed groups of synchronizing impulses with undelayed groups of synchronizing impulses, deriving energy representative of the difierence of the compared impulses, and adding progressively varying energy to the received signals during the time interval between the synchronizing impulses, the variation of the energy being between the limits of zero and a maximum value proportional to the diflerence between the compared impulses.

8. A direct current insertion television system comprising means to receive video signals combined with interspersed synchronizing impulses in which direct current and low frequency components of the video signal are suppressed which combined signals and impulses are subjected to distortion in transmission, means for comparing the received synchronizing impulses with each other, means for deriving energy representative of the diiference of the compared impulses, and means for adding progressively varying energy to the received signals during the time interval between the synchronizing impulses, the variation of the energy being between the limits ofzero and a maximum value proportional to the diflerence between the compared impulses.

9. A direct current insertion television system comprising means to receive video signals compared impulses, and means for adding progressively varying energy to the received 1811 18 during the time interval between the synchronizing impulses, the variation of the energy being between the limits of zero and a maximum value proportional to the difference between the compared impulses.

10. A direct current insertion television system comprising means to receive video signals combined with interspersed synchronizing impulses in which direct current and low frequency components of the video signals are suppressed which combined signals and impulses are subjected to distortion in transmission, means for comparing the received successive groups of synchronizing impulses with each other, means for deriving energy representative of 4.0

the difference of the compared impulses and means for adding progressively varying energy to the received signals during the time interval between the synchronizing impulses, the varia- 12. A direct current insertion television system comprising means toreceive video signals combined with interspersed synchronizing impulses in which direct current and low frequency components of the video signals are suppressed which combined signals and impulses are subjected to distortion in transmission, means for comparing the received synchronizing-impulses with each other, means for deriving potential increasing linearly with respect to time in proportion to the difference of the compared impulses, and means for adding said derived potential to said received combined signal and impulses.

13. A direct current insertion television system comprising means to receive video signals combined with interspersed synchronizing impulses in which direct current and low frequency components of the video signals are suppressed which combined signals and impulses are subjected to distortion in transmission, means for comparing the received synchronizing impulses with each other, means for deriving potential increasing parabolically with respect to time in proportion to the diflerence of the compared impulses and means for adding said derived potential to said received combined signal and impulses.

14. A direct current insertion television system comprisingmeans to receive video signals combined with interspersed synchronizing impulses in which direct current and low frequency components of, the video signals are suppressed which combined signals and impulses are subjected to distortion, intransmission, means for delaying the combined signals and impulses, means for comparing delayed groups of synchronizing impulses with undelayed groups of synchronizing impulses, means for deriving energy representative of the difference of the compared impulses, and means for adding progressively varying energy to the received signals during the time interval between the synchronizing impulses, the variation of the energybeing between the limits of zero and a maximum value proportional to the difference between the comtion of the energy being between the limits of ,45 pared impulses.

zero and a maximum value proportional to the difference between the compared impulses.

11. A direct current insertion television system comprising means to receive video signals combined with interspersed synchronizing impulses in which-direct current and low frequency components of the video signals are suppressed which combined signals and impulses are subjected to distortion in transmission, means for delaying the combined signals and impulses, means for comparing delayed synchronizing impulses with undelayed synchronizing impulses, means for deriving energy representative of the difference of the compared impulses, and means for adding progressively varying energy to the received signals during the time interval between the synchronizing impulses, the variation of theenergy being between the limits of zero and a maximum value .proportional to the difference between the compared impulses.

15. A direct current insertion television system comprising means to receive video signals combined with interspersed synchronizing impulses in which direct current and low frequency components of the video signals are suppressed which combined signals and impulses are subjected to distortion in transmission, a delay network fed by said received signals, said delay network having an output circuit, thermionic means connected across said output circuit, means common to both named thermionic means for deriving a potential proportional to the difference between the received energyand'the ene y Of the output

Images