458,585

PATENT SPECIFICATION

Application Date: March 20, 1935. No. 8739/35.

Complete Specification Left: March 9, 1936.

Complete Specification Accepted: Dec. 21, 1936.

-------------------------

PROVISIONAL SPECIFICATION

Improvements in and relating to the Transmission of Electrical Signals having a Direct Current Component

I, ALAN DOWER BLUMLEIN, a British Subject, of 32, Audley Road, Ealing, London, W.5, do hereby declare the nature of this invention to be as follows:

The present invention relates to the transmission of electrical signals having a direct current component. The term "direct current component" is intended to include not only the actual direct current component but also components of very low frequency which may be regarded as slow variations in the direct current component. In television for example, the direct current component represents the average brightness of the picture transmitted and slow changes in the average brightness.

Where such signals have to be transmitted by modulated carrier wave it is not possible to make use at a receiver of the average amplitude of the received carrier in order to correct for varying attenuation, due for example to fading, as is commonly done in so-called A.V.C. systems for sound broadcast receivers. This is because the average carrier amplitude changes at the transmitter with changes in the value of the direct current component.

Some difficulty therefore arises in providing means for compensating for varying attenuation. The difficulty does not, of course, arise where the direct current component is suppressed at the transmitter but, to obtain the original signal at the receiver, it is then necessary to provide some means for re-inserting the missing direct current component. The use of a "stabilised" carrier, that is one bearing the appropriate direct current modulation, is also of advantage in that for a given depth of modulation the average carrier power can be made lower than when the carrier is unstabilised.

It is accordingly one of the principal objects of the present invention to provide improved means whereby compensation for varying attenuation in a carrier transmission can be obtained irrespective of whether the signal modulations contain a direct current component or not.

It is a further object of the present invention to provide improved means whereby the direct current component may be established in signals which have been wholly or partly deprived of this component or in which this component is incorrectly represented.

The present invention accordingly provides a method of correcting for variations in effective amplitude of signals, such for example as may arise due to varying attenuation or changes in a direct current signal component which is wholly or partly suppressed at some point in a transmission channel, the method comprising transmitting at spaced intervals, along the same transmission channel as said signals, a check signal having a portion of fixed predetermined amplitude and, at a desired point in said channel, causing said portions to influence a correcting device to develop a corrective signal dependent upon the amplitude of said portions, applying said corrective signal to compensate at least partially for the variations in effective amplitude thereof and utilising a part of each of said check signals, occurring before said fixed amplitude portion thereof, to change said correcting device from its normally insensitive condition to a condition in which it is responsive to said fixed amplitude portion.

Signals are regarded as transmitted along the same channel either when they are used to modulate the same carrier wave or, if a carrier is not used, when they are transmitted along the same transmission line. The check signals may be a part of the signals themselves or they may be interposed between trains of the signals.

The present invention further provides television and like apparatus for use in connection with signals in the form of trains of picture signals interspersed with auxiliary signals which may include synchronising signals, and apparatus having a correcting device for controlling the effective gain of said apparatus in response to said auxiliary signals and means responsive to a part of said auxiliary signals for converting said correcting device from its normally insensitive condition into a condition in which it is capable of developing its gain controlling effect.

As applied to a carrier wave transmission system for transmitting electrical signal variations containing a direct current component, the present invention provides a transmitter having means for modulation a carrier with the signal variations deprived of the whole or part of their direct current component, the signal variations containing a recurrent check signal comprising two different amplitude values which are fixed values when the direct current signal components are present, and a receiver co-operating with said transmitter and having means for adjusting the effective gain of an amplifier thereof, in dependence upon the difference between the two different amplitude valves of said check signals.

A receiver for use with the system according to the preceding paragraph therefore has means responsive to the difference between two recurrent amplitude values for adjusting the gain of an amplifier thereof. The receiver may be provided with means for reinserting the direct current component with reference to one of said amplitude values (for example in the manner set forth in Patent Specification No. 422,906) and means for thereafter deriving a gain controlling voltage from the amplitude of the other of said amplitude values.

The invention will be described with reference to the accompanying drawings, in which

Figs. 1 to 3 are explanatory diagrams,

Figs. 4, 5 and 6 are circuit diagrams illustrating parts of apparatus according to this invention,

Fig. 7 is a circuit diagram illustrating a further form of the present invention,

Figs. 8 and 9 are further explanatory diagrams, and

Fig. 10 shows a further circuit arrangement illustrative of a feature of the present invention.

Fig. 1 shows a form of television signal which may be used in carrying out the present invention. Picture signals 1 are interspersed with line synchronising impulses 2 occurring between the scanning of adjacent lines of the object to be transmitted. The line marked W indicates the picture signal amplitude representative of the brightest element of the object, the line marked B corresponds to the amplitude level of a "black" signal and the synchronising impulses are seen to be in the "blacker than black" direction.

The dotted line 3 indicates the end of the last line of one traverse or frame of the object and the dotted line 4 indicates the commencement of the next frame. In this interval here is transmitted a frame signal comprising, in the present example, three pulses 5. The leading edges L indicate the commencements of the line impulses and the leading edge F that of the frame signal. The line impulses and the frame signal serve to control the generation of saw-tooth oscillations oat the line and frame frequencies respectively in known manner. The particular form of frame signal shown in Fig. 1 is intended for use in interlaced scanning where the lines traced in one traverse of the image interlace with the lines traced in the next traverse; in the present example the image is completely scanned in two traverses thereof and in such a case it is desirable that the leading edges of the line impulses should occur at double frequency during the frame signal. This is n order that the frame signal should always have the same energy content in spite of the fact that the leading edge of one frame impulse occurs at point F in Fig. 1 whilst the next frame impulse will be displaced by half a line interval relatively to this. For this reason an additional leading edge L1 is provided.

The signal of Fig. 1 will be assumed to have been used to modulate a carrier in such a way that the direct current component is present at the modulating point, so that any given value of ordinate in Fig. 1 will correspond to a fixed carrier amplitude. Further the modulation is such that the carrier amplitude is reduced to zero on the peaks P of the synchronising signals.

Referring now to Fig. 4, it will be assumed that there is applied from a suitable radio receiver to the tuned circuit 14 the modulated carrier referred to in the preceding paragraph. It will further be assumed that the carrier is subject to varying attenuation and that it is desired to correct for this at the receiver.

The signals developed across the circuit 14 are rectified by the diode rectifier 6 giving a rectified signal across a condenser 7 and resistance 8. The time constant of these elements 7 and 8 is so short that the voltage across 7 follows the modulation signals; in fact a wave as shown in Fig. 1 will be obtained. The rectified signals so obtained are taken through a radio frequency choke 9 and high resistance 10 to the grid of a valve 11. With negligible radio frequency signal (on the peak of a synchronising pulse) this valve 1 is biased to cut off by the battery 12. For any signal of about "black" amplitude, or greater, the valve 11 is conducting and keeps a condenser 13 discharged. During the synchronising pulses 2 and 5 the condenser charges through a resistance 15 and during a long frame pulse 5 sufficient charge is accumulated on the condenser 13 to take the grid of a valve 16 above the cut off as fixed by the bias applied from battery 17. The valve 11 served to provide a positive pulse on the grid of valve 16 at the occurrence of the first frame pulse 5. The anode of valve 16 is fed through a resistance 18 which also feeds the screen grid of a valve 19. Valves 19 and 20 are connected as a "multi-vibrator," their grids being cross-connected to their anodes through condenser 21 and 22 and being connected to earth through leak resistances 23 and 24. The condenser 21 is of comparatively large value so that, with resistance 23, it has a time constant which is long compared with the frame period, that is the time interval between successive occurrences of the leading edge F in fig. 1. The leak 24 is made adjustable. The multi-vibrator tends to lock in the condition in which valve 19 is conducting and valve 20 insulating. The first series of frame pulses 5 which arrives makes valve 11 insulate for long enough to allow condenser 13 to charge up sufficiently to make valve 16 conductive. This lowers the potential of the screen grid of valve 19 and makes this valve insulate. When valve 19 insulates, its anode becomes more positive and the resulting positive pulse on the grid of valve 20 makes this valve conduct and drive the grid of valve 19 negative. After a short period, the charge on condenser 22 leaks away through resistance 24 and valve 19 becomes conductive driving valve 20 to a non-conductive state in which, owing to the long time constant of elements 21 and 23, it remains until the next series of frame impulses arrives.

By suitable adjusting the value of resistance 24 it can be arranged that the valve 19 remains insulating for nearly as long as the interval between frames, that is the interval between line 3 and 4 in Fig. 1. The interval during which the valve 19 is insulating may thus commence during or just after the arrival of the first pulse 5 and may end just before the picture signals commence again, for example during the last line impulse of the interval between successive frames. A suitable insulating interval is that indicated between lines 25 and 26 in Fig. 1.

The circuit of Fig. 4 is illustrative of circuits which my be used to develop a control signal which can in turn be used to render sensitive a correcting device which a t other times is insensitive. This correcting device serves to develop a corrective current or voltage for correcting any variations which may take place in the attenuation of the received signal with reference to the part of the received signals occurring during this "sensitive" interval. Many other forms of circuit can of course be used in place of that shown in Fig. 4.

In Fig. 4 a control signal is developed at points 27 and 28, the former being in the positive sense and the latter in the negative sense. The form of the control signal developed at 27 is shown in Fig. 2.

A part of a correcting device according to the invention is shown in Fig. 5. The received carrier is applied as in Fig. 4 to a tune circuit 14 and is demodulated by a diode detector 6 having a condenser 7 with a parallel resistance 8 connected as shown. The elements 14, 6, 7 and 8 need not be duplicated in the circuit of Fig. 5 because, if desired, the required voltages may be taken from the terminals of condenser 7 in Fig. 4. Voltages may also be taken, if desired, from the same terminals to supply the picture reconstituting device. Alternatively some or all of the elements referred to may be separate for the different parts of the receiver circuit.

The wave form of he kind shown in Fig. 1, developed across the condenser 7 is fed to the outer control grid of a mixer valve 29. Positive pulses, such as those shown at 32 in Fig. 2 which may be derived from point 27 in Fig. 4, are applied at 30 through a condenser 31 to the inner control grid of the valve 29. The inner control grid is suitably biased through a leak 33, and the bias value may be so adjusted that the inner grid comes to cathode potential during each pulse 32 but falls during the remainder of the cycle t such a negative value that the valve 29 is then entirely inoperative. By giving the bias voltage a value slightly less negative than required for the above purpose, it can be arranged that the grid automatically fixes itself during the pulse t about cathode potential, owing to grid current charging the feed condenser 31. The valve 29 is biased by a battery 34 so that with negligible incoming radio signals (peaks P of synchronising pulses) the outer grid will cut off current from the anode, even if the inner grid is at cathode potential. When however a signal representing black occur, the outer grid will allow current to flow to the anode whenever the inner grid allows the valve to conduct. The anode current of the valve 29 will therefore take the form of the wave shown in Fig. 1, only during periods in which the inner grid allows the valve to conduct and for the remaining time the anode current will be zero.

Fig. 3 shows a curve of the anode current of the valve 29. The dotted parts of the curve show the current that would be obtained were the inner grid switched on all the time and the full line part of the curve showing the actual anode current. In order that the valve 29 may operate efficiently, it can be arranged that the outer grid is arranged to be at approximately cathode potential for a signal representing black amplitude. Although the vision signals will drive the outer grid ore positive than the cathode, no grid current will flow, owing to the valve being cut off by the negative signal on the inner grid during these periods. The anode current so obtained will be dependent upon the amplitude of the black signals during the frame interval and this amplitude is dependent on the amplification or attenuation of the transmission path supplying the signals, and is independent of the average brightness of the picture. The anode current may therefor be used to provide automatic gain control.

The average value of the D.C. component of the anode current will vary with the black amplitude and the voltage across the anode resistance 34 may therefore be used, after smoothing, to provide a D.C. potential for operating controlling devices, e.g., variable Mu valves for automatic gain control alternatively, the A.C. component of the anode current may be passed out through a condenser 36 and may be amplified to provide an A.C. controlling wave.

Fig. 6 shows a circuit, employing the latter method, by which the A.C. signals may be utilised. A.C. signals from condenser 36 of Fig. 5, after amplification if desired, are fed in at 41 through condenser 42 to a load resistance 43. The signals should be fed in at 41 in such a way that the pulses in the direction P to B in Fig. 3 during the frame interval are in the negative direction. The A.C. voltage across resistance 43 operates a diode rectifier 44, which, owing to the condenser 45 and leak 46, operates as a peak rectifier. Any alteration in the black amplitude of the incoming signal (that is of the level B in Fig. 3) will alter the amplitude of the A.C. wave fed in at 41, and will thus alter the received voltage across condenser 45. The receiver is shown so connected as to produce a negative voltage across condenser 45 to earth. By providing suitable amplification between the valve 29 of Fig. 5 and the rectifier 44 of Fig. 6, it can be arranged that a small change in the black amplitude produces a very large change in rectified voltage at 45. If it is required to hold the resultant signal to fairly close limits, a battery 47 may be inserted in series with the receiver, so as to prevent any rectification, unless the black signal reaches a predetermined value. Any small increase in the black signal will then produce a comparatively large negative voltage across the condenser 45 and this voltage may be used for controlling the bias of amplifier units in the radio receiver, so that an increase of black amplitude automatically reduces the gain of the receiver. If it is arranged that a small increase of black amplitude produces a very large change in voltage at 45, the condenser 45 and leak 46 should be given a sufficiently long time constant to prevent an unsteady state arising which would allow the whole device to "hunt." In any case, the time constant of condenser 45 and leak 46 should preferably exceed by several times the intervals between successive frames. S the leak 46 may be of rather high value, it is preferable to pass the rectified voltages through a further valve 48 before utilisation, in order to ensure that any leakage of the utilisation circuits do not affect the voltage across 45. This valve 48 has its anode connected to a suitable source of high tension, and its cathode fed from a negative voltage source through a comparatively high resistance 49. The voltage of the cathode of the valve 48 will then follow almost exactly the voltage of the grid and the voltage on the cathode can be utilised at 50 for any required purpose. For example, the connected 50 may be taken to provide the bias for the grids of variable Mu valves in the radio frequency amplifier of the wireless receiver. Alternatively, if desired, instead of controlling the amplification of a circuit preceding the device of Fig. 5, the amplification of a later amplifier can be controlled.

The invention has so far been described with reference to the correction of a variable attenuation which equally affects the whole transmitted wave. This invention can however also be utilised for reinserting the direct current component into the signals assuming that the signals are not subject to varying attenuation. For example, suppose that television signals are fed through an amplifier channel of steady amplification which does not however transmit the direct current component. Such signals may be fed in to a circuit such as shown in Fig. 5 at the point 37 (the radio frequency rectifier being omitted). As before, signals are fed in at 32 to render the valve 29 responsive for a period during the interval between frame pulses. The D.C. current obtained in the anode circuit is then proportional to the absolute amplitude of the black pulses fed into this valve, and the D.C. potential drop across the resistance 34 in the anode circuit may be utilised to reinsert the D.C. component. The signals are preferably fed in at 37 through a condenser, the grid to which they are applied being connected through a leak resistance to the cathode. The anode circuit resistance 34 shunted by a condenser is then arranged between the cathode of the valve 29 and the negative terminal of the anode current source and the cathode is connected through a grid leak, to a subsequent valve to provide a bias therefor. This subsequent valve may be fed through a condenser with the purely A.C. signals and, as the bias of the valve is proportional to the average amplitude of "black" in the A.C. wave, it can be arranged that the resultant anode current of this subsequent valve represents the television wave with the D.C. component inserted.

In a further example of the use of this invention, the television wave such as that in Fig. 1, is transmitted through a channel wherein, although the absolute gain for vision signals is substantially constant, the D.C. component and the amplification of the synchronising pulses vary owing to the variants say of H.T. voltage and the effect o curved amplification characteristics. An example of such a case is partly shown in Fig. 7. This represents a typical input to the modulator of a transmitter for television signals. The amplified television signal wave such as shown in Fig. 1 is fed I at 51 though a condenser 53 to the grid of a modulator 52. The sense of this wave is such that the vision signals are negative and the synchronising signals positive. The grid of the valve 52 is connected to the cathode through the leak 54 and this leak, together with a diode 55, re-establishes the direct current component f the wave on the grid of valve 52 with reference to the peak values P of the signals. The absolute value of the voltage of grid 52 representing the synchronising pulse is fixed by the negative bias supplied to the cathode of 55 through a connection 56 as will hereinafter be explained. This stabilisation circuit operates in the manner described in Patent Specification No. 422,906. The anode of the valve 52 is fed through a floating anode battery 57 and modulator resistance 58. The television signals will then appear across 58 complete with their direct current component. The upper end of resistance 58 is connected through any necessary bias 59 to the grids of two push-pull modulator valve 60, whose grids are fed with radio frequency oscillations or carrier frequency through the coils 61 and 62. The radio frequency output from the anodes of the valve 60 is then pass through further radio frequency amplifiers if required, to the aerial. The wave fed in at 51 is given a higher ratio of synchronising signal to picture signal amplitude than is required in the final modulated output of the transmitter. This relative increase of synchronising signal input allows for the curvature of the characteristic of the modulator and further amplifier stages. It is arranged that the peaks P of the synchronising signals cause the radio frequency output to fall to substantially zero, which means that the synchronising signals must modulate the transmitter over the curved and less efficient lower part of the modulation and amplification characteristics. Since the synchronising signals are square topped, there will be no shape distortion other than a relative attenuation due to their operating on the curved part of the characteristic, so that this part of the modulation characteristic can be used for the synchronising signals, leaving the upper straighter part for the vision signals. The resultant radio frequency output at the aerial will then have particular values corresponding to particular light intensities and a value, substantially equal to zero, corresponding to the synchronising signal. Although this latter value is normally zero, it is often inconvenient to provide sufficient synchronising signal input to modulate the transmitter right down to zero anode current around the bottom bend of the characteristic.

At the receiver the correct absolute brightness is obtained either by providing a direct current coupling between the demodulating detector and the light modulating means, such as the control electrode of a cathode ray tube, or by allowing the D.C. to be suppressed and then re-establishing the D.C. with reference to the peak amplitude P of the synchronising pulses previous to application of the signals to the light modulating device. If now the carrier amplitude representing black varies, or if the received amplitude representing the difference between the peaks P of synchronising pulses and black varies, the direct current or average picture brightness component will effectively vary at the receiver. For example, a slight fall of the carrier value representing black, or the amplitude of the synchronising pulses relative to black, will cause the screen of the cathode ray tube to be darkened. Such darkening will be relatively unimportant in the high lights, but may be sufficient to obscure detail in the dark parts of the picture, due to the very dark greys and blacks becoming simultaneously black.

Variation in the carrier amplitude representing black or of the received amplitude representing the difference (P to B Fig. 3) between the peaks of the synchronising pulses and black may be caused for example by variation in the voltage of the source supplying the anode circuits of the modulator or radio frequency amplifiers at the transmitter of by variation in the voltage of battery 57 in Fig. 7. The effect of the variation at the receiver will be the same whichever of the two methods, above referred to, of obtaining the D.C. component are used. Unless the carrier amplitude representing black remains constant it is impossible to keep constant the amplitude of the synchronising peaks P from black B because these signals extend over the curved part of the characteristic. It is therefore difficult to maintain a correct representation of the average picture brightness at the receiver and, when the variations in question are considerable, it may prove difficult to separate the synchronising signals correctly from the picture signals.

In order to correct for variations of the kind above described, therefore, it may be arranged according to a feature of the present invention, to feed in at point 56 in Fig. 7 a voltage which varies in such a manner as to keep the "black" level of the carrier output at a constant value in spite of the variations referred to.

One way in which this may be done is to feed modulated carrier frequency energy from the transmitter of Fig. 7, or from the transmitting aerial, to the circuit 14 of an arrangement such as that of Fig. 5. Signals of the form shown in Fig. 2 are applied at point 32. These signals may for example be generated with the aid of a circuit such as that of Fig. 4 or from a part of the mechanism which serves to generate the synchronising signals. The output of the valve 29 will be proportional to the carrier amplitude corresponding to black, less any effect due to slight variation in the carrier amplitude representative of the peaks P of the synchronising signals, if this is not quite zero. The A.C. output of valve 29 may therefore be fed, after any necessary amplification, at 41 into the device of Fig. 6 and there will then, as already explained, by developed cross resistance 49 voltages which can be arranged to vary substantially with relatively small variations in the level corresponding to black in the signals fed in at 37 in Fig. 5.

The output obtained at point 50 in Fig. 6 may then be applied to control the potential of point 56 in Fig. 7. In the particular arrangement shown in Fig. 6, an increase in "black" amplitude produces a negative voltage at 50, whereas a positive voltage is required at point 56 of Fig. 7 I order to correct such an increase in amplitude. To overcome this difficulty, the sense of the input to the diode 41 in Fig. 6 may be made opposite to that previously described and the diode 41 and battery 47 may also be reversed. The point 50 in Fig. 6 may then be connected through any necessary biasing potential source directly to point 56 in Fig. 7 to give the desired result.

Alternatively there may be applied to point 56 of Fig. 7 a voltage proportional to the direct current flowing in the valve 29 of Fig. 5, for example by inserting a resistance in the cathode circuit of the valve 29, as explained above in connection with the re-establishment of D.C., and the cathode of the valve 29 may then be connected to point 56 of Fig. 7.

Instead of injecting the control voltage at point 56 of Fig. 7, it can be arranged that the potential at point 50 in Fig. 6 is effective in controlling the amplitude of the synchronising signals applied at the point 51. For example before these synchronising signals are mixed with the picture signals, they may be passed through an amplifier having its amplification controlled suitably in accordance with the voltage at point 50 of Fig. 6. If the amplification is controlled by variation in the bias of one or more variable Mu valves, the point 50 may be connected to control the bias of the valves. The connections in Fig. 6 would then be as shown and the sense in which the signals are applied at 41 would be the same as that originally described for that Figure. Thus an increase in black amplitude would produce a negative voltage at point 50, tending to reduce the amplification of the synchronising signal amplifier and thus tending to reduce the black amplitude.

In a system in which the carrier amplitude is not educed to the neighbourhood of zero on the peaks of the synchronising signals, it can be arranged that the black signals alone are effective in producing a voltage for controlling the black level. Such an arrangement would not take count of changes in the amplitude P to B in Fig. 3 unless this were accompanied by a change in the black level. This result can be obtained with the circuit of Fig. 5 by biasing the valve 29 so that the synchronising signals carry the outer control grid well below cut-off so that variation in the amplitude of the synchronising signals is not effective in changing the anode current and so that the amplitude of the black signals (B Fig. 3) is effective in controlling the anode current.

Where it is not desired to apply a correction to the synchronising signals before they are mixed with the picture signals, the synchronising signal amplitude may be increased or decreased relatively to the picture signal amplitude by passing the composite signal through a suitable circuit. Such a circuit may comprise a thermionic valve having the curvature of its characteristic variable under the control of a voltage such as that obtained from point 50 in Fig. 6. An example of a circuit of this kind is shown in Fig. 10. The combined signal, having the form shown in fig. 1 for example, is fed through a condenser 71 on to the grids of two valves 74 and 75. The sense of the signals is such that the synchronising signals are positive. A diode 72 and leak 73 serve to reinsert the D.C. component into the signals with reference to the peaks P of the synchronising signals. The cathode of valve 74 is connected to earth through a resistance 78 which reduces the mutual conductance of the valve and lengthens and straightens its characteristic. The valve 75 has its cathode biased very positively so that the valve is inoperative for the negative picture signals on its grid but is operative to give increased amplification for the synchronising signals. The combined output of the valves 74 and 75 is passed out at 76.

A control voltage, such as that from point 50 in Fig. 6, is applied at point 7 and is effective in altering the bias on the valves 74 and 75. Thus an increase in the "black" level of the outgoing signals will produce a negative voltage at 77 which will reduce the amplitude of the synchronising signals, so tending to neutralise the change in "black" level.

In the systems so far described, the corrective effect is derived from an "observation" of the "black" level during the intervals between frames.

Where the line synchronising signals do not occupy the whole of the interval between successive trains of picture signals representative of successive lines of the object, however, use may be made of and "observation" in this interval (which may be called the line interval) to derive a corrective effect. A line synchronising impulse usually occupies about one tenth of a line period and the other nine tenths is usually occupied by picture signals.

In Fig. 8 is shown a wave form in which the line synchronising impulse 2 occupy only a fraction of the line interval D, for example a quarter of this interval. Thus the impulse lasts for one fortieth of a line period and during three fortieths of a line period the signal is at "black," indicated by reference B. with such a signal it is possible to derive a corrective effected once every line and a more rapid control is therefore possible than with the signal of Fig. 1. Signals so the kind shown in fig. 8 may be useful for example for providing automatic gain control for the receiver of a relay station receiving signals from a moving van or over some other channel subject to rapidly varying attenuation, with a view to re-transmitting the signals after correction.

If the maximum possible rapidity of action is desired with signals of the kind shown in Fig. 8, the signals in the frame interval should be modified from those shown in fig. 1. Thus each of the pulses 5 may be constituted by an excursion (at F, L, and L1) to level P to form a pulse lasting one fortieth of a line interval, a return to level B for a period of three fortieths of a line interval, a further excursion to level P for a period of three tenths of a line interval followed by a return to level B.

The signals of Fig. 8 may be utilised in a circuit which is a modification of that shown in Fig. 4. The valve 11 is omitted and the output of the rectifier from the choke 9 is taken (without resistance 10) to the grid of the valve 16. The rectifier 6 is inverted so that picture signals produce a negative rectified voltage. The valve 16 is so adjusted or biased that it does not pass current for the negative voltages representing picture signals. At the occurrence of a line synchronising signal the valve 16 conducts, thus triggering the multi-vibrator valves 19 and 20. The leak 24 is adjusted so that the valve 19 remains insulating for the moment of triggering to just before the beginning of the next train of picture signals, that is for most of the one-tenth of a line interval. Instead of making the condenser 21 so large that the multi-vibrator is quasi-aperiodic it may be found advantageous to make the natural period of the device just longer than a line period, so that the device tends to run at the required frequency. This tends to reduce the amplitude of the pulses required from valve 16 and thus false triggering due to interference may be reduced. The output at 27 then takes the form of pulses occurring once a line and having a length slightly less than one tenth of a line period.

The pulses form 20 may be fed in at 30 in Fig. 5 and may control the valve 29 to give output signals at 36 dependent on the black amplitude occurring in the short interval between a line synchronising pulse and the beginning of the picture signals of the next line. The signals at 36 may be amplified and passed to a circuit such as that of Fig. 6 for producing the control voltage at 50. In this case the time constant of condenser 45 and leak 46 may be much smaller than that employed for "observation" once per frame, thus giving a more rapid control. The time constant of elements 45 and 46 should however be made sufficiently greater than the line period to prevent any instability of the control.

The examples given above are of systems in which the black level is "observed" and a corrective signal dependent on this observation is injected at an earlier point in the system so as to correct as far as possible any variation of black level at the point of observation. The invention is equally applicable to systems where the controlling signal is utilised to correct the black level at some point after the point of observation. For example in the automatic gain control systems described above the control signal developed at 50 in Fig. 6 may be used to control the gain of amplifiers following the point of observation. Alternatively where large variations of transmission efficiency are to be corrected it may be arranged that the corrective signal serves to vary the amplification both before and after the point of observation. For example the control voltage at 50 in Fig. 6 may be utilised to control the radio frequency amplification ahead of the rectifier in Fig. 5. Such a control may for example reduce the variation of the black level at the rectifier in Fig. 5 to 10% for a 40 decibel change I incoming signal strength. The control voltage at 50 in Fig. 6 may also serve to produce a slight variation in the amplification following the observation point (e.g. the modulation frequency amplification following the rectifier) so that a 10% change of black level at the observation point produces a 10% compensating change of gain following it, thus giving a final output signal substantially free from any variation.

Examples previously given have been based on an observation of the "black amplitude." The invention can also be applied to the observation of any definite picture amplitude or amplitude related to the picture amplitude. For example, instead of the wave shown on Fig. 8, the wave shown in Fig. 9 may be transmitted. In this case during the interval between the synchronising pulse and the beginning of the next line, the signal amplitude assumes a value E between the synchronising peaks P and the black amplitude B. such a signal level may be fixed as being a certain fraction of the distance between P and B, or alternatively, if the amplitude of the synchronising impulses from the black level varies, may be defined as being a certain fraction of the amplitude of the maximum white signal W relatively to the level B, below the level B. signals of the type shown in Fig. 9 may be useful for direct reception on a cathode ray tube, the slightly "blacker than black" signals E serving to black out the cathode ray during the return stroke. This signal could be "observed" in circuits (with slightly modified adjustment) similar to those employed for the wave of Fig. 8. Similarly a wave form may be employed where during the interval between the synchronising signal and the picture signals of the succeeding line, the amplitude corresponds to a value within the picture range, say half way between black B and white ". such a wave provides a strong controlling signal but special means will probably be found necessary to "black out" this signal before the picture is reproduced.

Alternatively, a wave of the kind last referred to may be used for replacing a signal to a final transmitter and after the corrective signal has been derived and used to correct the signal, the excursion into the picture range occurring in the line intervals may be suppressed before final transmission of the signals.

Similarly, wave forms of the kind shown in Figs. 8 and 9 may be converted before final transmission to the form shown in Fig. 1, for example by superimposing upon the signals pulses such as those developed by the multi-vibrator of Fig. 4 when used with a wave of the kind shown in Fig. 6, and then limiting the resultant pulses to the required amplitude.

A radio relay station for a television transmission system may employ the corrective means according to the present invention more than once. For example the wave received at the relay station may be of the general form shown in Fig. 8 and may be observed once per line. The radio frequency gain may then be adjusted automatically as already described in dependence upon the observation. A further automatic adjustment of gain may the be carried out in order to correct for the error introduced due to the fact that slight changes in the black level at the observation point are necessary to develop a corrective signal to correct the gain of a preceding amplifier. The corrected modulation frequency output may be passed through a circuit to increase the relative amplitude of the synchronising signals so as to compensate for the subsequent reduction of the amplitude of these signals by the curved characteristic of the transmitter. Simultaneously the wave form may be converted to the form shown in Fig. 1 or Fig. 7 so as to be more suitable for final reception. The wave may then be applied to the transmitter and the radiated signal level corresponding to black may be controlled by one of the methods already described. For this latter control, observation may be made once per line for waves of the general form of Figs. 8 and 9 or once per frame for waves of the general form of Fig. 1.

In an alternative arrangement to that shown in Fig. 4, a synchronising pulse is caused to generate the required switching signal to turn on the observing valve. For example, using the waveform shown in Figs. 8 or 9, the synchronising pulses P can be separated from the vision signals in the manner described in Patent Specification No. 422,824. These pulses may then be delayed by a delay network and inverted by subjecting them to one stage of valve amplification, after which they can serve to make sensitive the observing valve such as 29 in Fig. 5 during the period B or E in Figs. 8 and 9. Alternatively the observing valve may be turned on by a pulse obtained not from the leading edge of the synchronising pulse but from the trailing edge. In such an arrangement the synchronising pulse is separated from the picture signal, passed through a simple high pass filter (e.g. a small series condenser), which coverts the synchronising pulse into two sharper pulses, one corresponding to the leading edge, and one in the reverse direction corresponding to the trailing edge. This second pulse can then be separated from the leading edge pulse by a suitably biased valve, which at the same time amplifies it. The output of this valve passes to a low pass filter (e.g., a shunt condenser) which broadens the pulse again, producing a pulse of length approximately that required to sensitise the observing device during the period B in Fig. 8.

In the above description there have been described methods of correcting for varying attenuation or for complete or partial absence of the direct current component of signals. For this it has been shown to be sufficient to derive a corrective signal dependent upon the received amplitude of a signal which at the transmitter is a fixed value.

Where, however, it is desired to correct both for varying attenuation and direct current component, it is necessary to derive a corrective signal dependent upon the difference between two different received amplitudes both of which have fixed values at the transmitter. Thus in the case of a signal of the form shown in Fig. 8, it my be assumed that at the transmitter the level B corresponds to picture black and that level P has a fixed difference from value B. now if such a signal be transmitted through a channel which is incapable of transmitting the D.C. component and which subjects the signals to varying attenuation, the procedure may be as follows:-

The signals are given a datum co-incident with the peaks P with the aid of a D.C. reinserting device of the kind set forth in Specification No. 422,904 and at the same time they are used to derive a corrective signal dependent upon the amplitude of level B, in the manner already described. As the D.C. reinserting device ensures that the datum remains on the peaks P and as the corrective signal makes the amplitude P to B substantially correct, the desired corrections will have been applied.

Although the invention has been described in some detail with reference to its application to television systems it is also applicable to other systems in which signals of the requisite character are either present for the purpose of enabling a corrective effect to be derived according to the present invention.

Dated this 20th day of March, 1935.

REDDIE & GROSE,

Agents for the Applicant,

6, Breamís Buildings, London, E.C.4.

COMPLETE SPECIFICATION

Improvements in and relating to the Transmission of Electrical Signals having a Direct Current Component

I, ALAN DOWER BLUMLEIN, a British Subject, of 32, Audley Road, Ealing, London, W.5, do hereby declare the nature of this invention and in what manner the same is to be performed, to be particularly described and ascertained in and by the following statement:-

The present invention relates to apparatus for handling electrical signals having a direct current component. The term "direct current component" is intended to include not only the actual direct current component but also signal components of very low frequency, which may be regarded as slow variations in the actual direct current component. In television, for example, the direct current component represents the average brightness of the picture transmitted and slow changes in the average brightness.

It is already known, for example, from Patent Specification No. 422,906, that when a signal having a D.C. component is passed through a channel, such as an A.C. amplifier, which is incapable of passing the D.C. component, that component is wholly or partly lost.

If the wave form of the signal without its D.C. component were plotted, and a straight line drawn through the curve in such a manner that the parts of the curve above and below the line enclosed equal areas, it would be found that parts of the curve which correspond to black (or any other fixed brightness) in the object of which an image is to be transmitted would not all have the same amplitude with resect to the above mentioned straight line. This line is the electrical zero of the signal, and is the datum line which the signal assumes when it loses its D.C. component. If, however, a corresponding curve were plotted for the signal with its D.C. component, it would be found that a straight line could be drawn through all parts of the curve corresponding to any fixed brightness value.

Now suppose that two television signals which are identical except that one contains its D.C. component while the other does not are to be reproduced by means of two identical cathode ray tube reproducers. To reproduce the signal containing its D.C. component, the cathode ray tube bias is given a fixed value such that the ray is just extinguished when parts of the signal representing black are operative; other parts of the signal will then be correctly reproduced relative to black, whatever the average picture brightness may be. Now clearly it is not possible when reproducing the signal which has lost its D.C. component to choose a bias value such that all parts of the signal corresponding to black just extinguish the ray, and a compromise must be adopted. The result of this is usually that when the average brightness of the transmitted picture increases there is a serious loss of detail in the darker parts of the reproduced picture, all parts of the signal representing the darker greys being reproduced as black.

From the above explanation it will be clear that if a signal loses its D.C. component, a given picture brightness is not always represented by the same signal amplitude and it may therefore be said that variations in the effective amplitude of the signal are introduced. In order to restore the signal to its original form, it is necessary to re-insert the lost D.C. component.

Means have already been proposed for effecting this re-insertion, and it is an object of this invention to provide novel or improved means for this purpose. The invention also aims to provide means for correcting for the incorrect representation of the D.C. component in electrical signals.

Furthermore, where signals having a direct current component are transmitted by modulated carrier wave, it is not possible to make use at a receiver of the average amplitude of the received carrier in order to correct for varying attenuation, due for example t fading, as is commonly done in automatic gain control systems for sound broadcast receivers. This is because the average carrier amplitude changes at the transmitter not only with changes in attenuation but also with changes in the value of the direct current component.

Some difficulty therefore arises in providing means for compensating for varying attenuation. The difficulty does not of course, arise where the direct current component is lost or suppressed at the transmitter but, to obtain the original signal at the receiver, it is then necessary to provide some means for re-inserting the missing direct current component. The use of a "stabilised" carrier, that is one bearing the appropriate direct current modulation, is however of advantage in that for a given depth of modulation the average carrier power can be made lower than when the carrier is unstabilised.

It is accordingly a further object of the present invention to provide improved means whereby compensation for varying attenuation in a system of signal transmission be carrier wave can be obtained, irrespective of whether the direct current component of the signals is represented in the transmitted carrier or not.

The present invention accordingly provides a method of correcting for variations in the effective amplitude of electrical signals representative of intelligence, such as may arise in the transmission of said signals as a result of the complete or partial loss of the D.C. component of said signals, the incorrect representation of that component, or varying attenuation of the signals, the said method comprising transmitting at spaced time intervals along the channel through which the intelligence signals are passed, check signals each of which has a switching portion and a datum portion, said datum portion having, at the input of said channel, either a predetermined fixed amplitude value of a predetermined wave form comprising fixe amplitude values and the said method being characterised in that the datum portions are applied through one path to an observing device, while the switching portions, or switching signals derived therefrom, are fed to said observing device through another path, and serve to change said observing device from the inoperative condition into the operative condition, said observing device, when in the operative condition, serving to develop a corrective signal dependent upon the amplitude of amplitude and wave form of the datum portions applied thereto, and said corrective signal being applied at a point either before or after the observing point to compensate wholly or in part for said variations in effective amplitude.

Two sets of signals are regarded as transmitted along the same channel either when they are used, for example, to modulate the same carrier wave, or, if a carrier is not used, when they are transmitted along the same transmission line. The check signals may be a part of the intelligence signals themselves, or they may be interposed between trains of the intelligence signals.

The invention further provides a receiver for carrying into effect the method of the statement of invention in the last paragraph but one above, and so designed as to be suitable for use in a transmission system for transmitting electrical intelligence signals containing a recurrent check signal, each check signal comprising two datum portions which, in the transmitted signal, have a fixed amplitude difference, and a switching portion either separate from said datum portions or constituted wholly or in part by one of said datum portions, said receiver comprising an observing device, a path for conveying the datum portions of said check signals to said observing device, and another path for conveying the switching portions of said check signals, or switching pulses derived from said switching portions, to said observing device and for enabling said switching portions, or the switching pulses derived therefrom, to change said observing device from the inoperative condition into the operative condition, said observing device being capable, when in the operative condition, of development a corrective signal dependent upon variations in said amplitude difference, and means being provided for utilising said corrective signal to vary the gain of an amplifier of said receiver to compensate at least partially for said variations.

A receiver according to the preceding paragraph may be provided with means for re-inserting the direct current component with reference to one of the datum portions (for example in the manner set forth in Patent Specification No. 422,906) and means for thereafter deriving a corrective signal for controlling the gain of an amplifier of the receiver.

The invention will be described, as applied by way of example to television, with reference to the drawings accompanying the Provisional Specification, in which

Figs. 1 to 3 are explanatory diagrams,

Figs. 4, 5 and 6 are circuit diagrams illustrating parts of apparatus according to this invention,

Fig. 7 is a circuit diagram illustrating a further form of the present invention,

Figs. 8 and 9 are further explanatory diagrams, and

Fig. 10 shows a further circuit arrangement illustrative of a feature of the present invention, and with reference to the accompanying drawing, in which

Fig. 11 illustrates a part of apparatus for carrying the invention into effect, and

Fig. 12 shows a further form of apparatus according to the invention.

Fig. 1 shows a form of television signal which may be used in carrying out the present invention. Picture signals 1 are interspersed with line synchronising impulses 2 occurring between the scanning of adjacent lines of the object to be transmitted. The line marked W indicates the picture signal amplitude representative of the brightness element of the object, the line marked B corresponds to the amplitude level of a "black" signal and the synchronising impulses are seen to be in the "blacker-than-black" direction.

The dotted line 3 indicates the end of the last line of one traverse or frame of the object and the dotted line 4 indicates the commencement of the next frame. In this interval there is transmitted a frame signal comprising, in the present example, three pulses 5. The leading edges L indicate the commencements of the line impulses and the leading edge F that of the frame signal. The line impulse and the frame signal serve to control the generation of saw-tooth oscillations at the line and frame frequencies respectively in known manner. The particularly form of frame signal shown in Fig. 1 is intended for use in interlaced scanning where the lines traced in one traverse of the image interlace with the lines traced in the next traverse; in the present example the image is completely scanned in two traverses thereof and in such a case it is desirable that the broad pulses 5 constituting the frame signal should occur at double the line impulse frequency during the frame interval, so that the frame signal should always have the same energy content in spite of the fact that the leading edge of one frame impulse occurs at point F in Fig. 1 whilst the next frame impulse will be displaced by half a line interval relatively to this. For this reason, the broad pulse having the leading edge L1 is provided in the present example.

The signal of Fig. 1 will be assumed to have been used to modulate a carrier in such a way that the direct current component is present at the modulating point, so that any given value of ordinate in Fig. 1 will correspond to a fixed carrier amplitude. Further, the modulation will be assumed to have been so effected that the carrier amplitude is reduced substantially to zero on the peaks P of the synchronising pulses.

Referring now to Fig. 4, it will be assumed that there is applied from a suitable radio receiver to the tuned circuit 14 the modulated carrier referred to in the preceding paragraph. It will further be assumed that the carrier is subject to varying attenuation and that it is desired to correct for this at the receiver.

The signals developed across the circuit 14 are rectified by the diode rectifier 6 giving a rectified signal across a condenser 7 and resistance 8. The time constant of these elements 7 and 8 is so short that the voltage across condenser 7 follows the envelope of the modulating signals; in fact, a voltage wave of the form shown in Fig. 1 will be set up across condenser 7. The rectified signals so obtained re taken through a radio frequency choke 9 and high resistance 10 to the control grid of valve 11. With negligible radio frequency signal (on the peak of a synchronising pulse) this valve 11 is biased to anode-current cut-off by the battery 12. For any signal of about "black" amplitude, or greater, the valve 11 is conducting and keeps a condenser 13 discharged. During the synchronising pulses 2 and 5, the condenser 13 charges through a resistance 15, and during a long frame pulse 5, sufficient charge is accumulated on the condenser 13 to take the potential of the control grid of a valve 16 to a value above that corresponding to anode-current cut-off; battery 17 serves to bias the control grid of valve 16, and the valve 11 serves to provide a positive pulse on the grid of valve 16 at the occurrence of the first frame pulse 5. The anode of valve 16 and the screen grid of a valve 19 are fed through a common resistance 18. Valves 19 and 20 are connected as a "multi-vibrator," their grids being cross-connected to their anodes through condensers 21 and 22 and being connected to earth through leak resistances 23 and 24. The condenser 21 is of comparatively large capacity so that, with resistance 23, it forms a circuit having a time constant which is long compared with the frame period, that is, the time interval between successive occurrences of the leading edge F of Fig. 1. The leak 24 is made adjustable. The multi-vibrator tends to seek its relatively more stable condition in which valve 19 is conducting and valve 20 insulating. The first series of frame pulses 5 (Fig. 1) which arrives makes valve 11 insulate for long enough to allow condenser 13 to charge up sufficiently to make valve 16 conductive. This lowers the potential of the screen grid of valve 19 and makes this valve insulate. When valve 19 insulates, its anode becomes more positive and the resulting positive pulse on the control grid of valve 20 makes this valve conduct, and the control grid of valve 19 is driven negative. After a short period, the charge on condenser 22 leaks away through resistance 24 and valve 19 becomes conductive once more, driving valve 20 to a non-conductive state in which, owing to the long time constant of elements 21 and 23, it remains until the next series of frame impulses arrives.

By suitably adjusting the value of resistance 24 it can be arranged that the valve 19 remains insulating for nearly as long as the interval between frames, that is, the interval between lines 3 and 4 in Fig. 1. The interval during which the valve 19 is insulating may thus commence during or just after the arrival of the first pulse 5 and may end just before the picture signals commence again, for example during the last line impulse of the interval between successive frames. A suitable insulating interval is indicated between lines 25 and 26 in Fig. 1.

The circuit of Fig. 4 is illustrative of circuits which may be used to develop a control signal which can in turn be used to render sensitive an observing device which at other times is insensitive. This observing device serves to develop a corrective current or voltage for correcting for any variations which may take place in the attenuation undergone by the received signal with reference to the part of the received signals occurring during this "sensitive" interval. Many other forms of circuit can, of course, be used in place of that shown in Fig. 4.

In Fig. 4 a control signal is developed at points 27 and 28, the signal at the former point being in the positive sense and that at the latter in the negative sense. The form of the control signal developed at 27 is shown in Fig. 2.

A part of an observing device according to the invention which is suitable for use with the arrangement of Fig. 4 is shown in Fig. 5. The received carrier is applied as in Fig. 4 to a tune circuit 14 and is demodulated by a diode detector 6 having a condenser 7 with a parallel load resistance 8 connected as shown. The elements 14, 6, 7 and 8 need not be duplicated in the circuit of Fig. 5 because, if desired, the required voltages may be taken from the terminals of condenser 7 of the apparatus of Fig. 4, in which case it is to be noted that the cathode of valve 29 of Fig. 5 is conductively connected to earth through batteries 35, 12 and 17. Voltages may also be taken, if desired, from the same terminals to supply the picture re-constituting device. Alternatively, some or all of the elements referred to may be separate for the different parts of the receiver circuit, in which case the negative end of battery 35 of Fig. 5 should be earthed.

The wave form of the kind shown in Fig. 1, developed across the condenser 7 of Fig. 5, is fed to the outer control grid of a hexode mixer valve 29. Positive pulses, such as those shown at 32 in Fig. 2 which my be derived from point 27 in Fig. 4, are applied at 30 through a condenser 31 to the inner control grid of the valve 29. The inner control grid is suitably biased through leak 33, and the bias value may be so adjusted that the inner grid comes to cathode potential during each pulse 32 but falls in the intervals between pulses 32 o such a negative value that the valve 29 is then entirely inoperative. By giving the bias voltage a value slightly less negative than required for the above purpose, it can be arranged that the grid automatically fixed itself during the pulse at about cathode potential, owing to grid current charging the feed condenser 31. The outer control grid of valve 29 is biased by a battery 35 so that with negligible incoming radio signals (peaks P of the synchronising pulses) the outer control grid will cut off or at least very considerably reduce the anode current, even if the inner control grid is at cathode potential. When, however, a signal representing black occurs, the outer grid will allow current to flow to the anode whenever the inner grid allows the valve to conduct. The anode current of the valve 29, during periods in which the inner grid allows the value to conduct, will have a wave form which is the same as that of the part of the signal of Fig. 1 between dotted lines 25, 26, while at all other times the anode current will be zero or substantially zero.

Fig. 3 shows the wave form of the anode current of the valve 29. The dotted parts of the curve show the current that would be obtained were the inner grid switched on all the time, and the full line part of the curve shows the actual anode current. in order that the valve 29 may operate efficiently, the outer grid is arranged, by adjustment of bias battery 35 or otherwise, to be at approximately cathode potential for a signal representing black amplitude. Although the picture signals will drive the outer grid more positive than the cathode, no grid current will flow, since the valve is biased to cut-off by the potential on the inner grid thereof at all times except during the frame intervals. The anode current of valve 29 is dependent upon the amplitude of the signals in the blacker-than-black sense occurring during the frame interval and this amplitude is dependent on the amplification of attenuation of the transmission path supplying the signals, and is independent of the average brightness of the picture. The anode current may therefore be used to actuate automatic gain control means.

The magnitude of the D.C. component of the anode current of valve 29 will vary in accordance with variations in the amplitude representing black, and the voltage across the anode resistance 34 may therefore be used, after smoothing, to provide a D.C. potential for operating controlling deices, e.g. variable-mu valves for automatic gain control. Alternatively, the A.C. component of the anode current may be passed out through a condenser 36 and may be amplified to provide an A.C. controlling wave.

Fig. 6 shows a circuit, employing the latter method, by which the A.C. controlling wave referred to may be utilised. The signals taken off through the condenser 36 of Fig. 5, after amplification if desired, are fed from terminal 41, through condenser 42, to a resistance 43. The signals should be fed in at 41 in such a way that the pulses in the direction P to B in Fig. 3 during the frame interval are in the negative direction. The A.C. voltage set up across resistance 43 is fed to a diode rectifier 44, which, owing to the values chosen for the condenser 45 and the load resistance 46, operates as a peak rectifier. Any alteration in the black amplitude of the incoming signal (that is, of the level B in Fig. 3) will introduce a change in the amplitude of the A.C. wave fed in at 41, and a corresponding change in the rectified voltage set up across condenser 45. The rectifier 44 is shown so connected as to produce a negative potential at the upper terminal of condenser 45 with respect to earth. By providing suitable amplification between the valve 29 of Fig. 5 and the rectifier 44 of Fig. 6, it can be arranged that a small change in the black amplitude produces a very large change in rectified voltage set up across condenser 45.

If it is required to hold the resultant signal to fairly close limits, a battery 47 may be inserted in series with the rectifier, so as to prevent rectification taking place until the black signal reaches a predetermined value. Any small increase in the black signal will then produce a comparatively large negative voltage across the condenser 45 and this voltage may be used for controlling the grid bias of radio-frequency amplifier valves in the receiver, so that an increase of black amplitude automatically reduces the gain of the receiver. If it is arranged that a small increase of black amplitude produces a very large change in voltage at condenser 45, the condenser 45 and leak resistance 46 should be given a sufficiently long time constant to prevent an unsteady state arising in which the apparatus as a whole might tend to "hunt." In any case, the time constant of condenser 45 and leak resistance 46 should preferably exceed by several times the time interval between successive frames. As the leak resistance 46 may have a relatively high value, it may be desirable to pass the rectified voltages through a further valve (48, Fig. 6) before utilisation, in order to ensure that leakages in the utilisation circuits do not affect the voltage across condenser 45. The valve 48 has its anode connected to a suitable source of high tension (not shown), and its cathode fed from a negative voltage source (not shown) through a comparatively high resistance 49. The voltage of the cathode of the valve 48 will then follow almost exactly the voltage of the grid of that valve, and the voltage on the cathode can be utilised at 50 for any required purpose. For example, the connection 50 may be taken to provide the bias for the grids of variable-mu valves in the radio-frequency amplifier of the wireless receiver. Alternatively, if desired, instead of controlling the amplification of a circuit preceding the device of Fig. 5, the amplification of a later amplifier can be controlled.

The invention has so far been described with reference to the correction of a variable attenuation which equally affects the whole transmitted wave. This invention can however also be utilised for re-inserting the direct current component into the signals, assuming that the signals are not subject to varying attenuation. For example, suppose that television signals are fed through an amplifier channel of steady amplification which does not however transmit this direct current component. Such signals may be fed in to a circuit such a s shown in Fig. 5 at the point 37 (the radio frequency rectifier 6 being omitted). As before, signals are fed in at 30 to render the valve 29 responsive for a period during the interval between frames. The direct component of the anode current of valve 29 is then proportional to the absolute amplitude of the black pulses fed into this valve, and the potential drop across the resistance 34 in the anode circuit due to this direct component may be utilised to re-insert the D.. component of the signals fed in at point 37. These signals are preferably fed in at 37 through a condenser, the grid to which they are applied being connected through a leak resistance to the cathode of valve 29. The anode circuit resistance 34, shunted by a condenser, is then conveniently arranged between the cathode of the valve 29 and the negative terminal of the anode current source (not shown) and the potential of the cathode is applied through a grid leak to the control grid of a subsequent valve to provide a bias therefor; in this arrangement, it should be arranged that only the anode current, and not the screen grid current of valve 29 passes through the resistance 34, and this can be achieved by supplying the anode and screen grid from different current sources. Black level is then represented by a potential at the cathode of valve 29 which is positive with respect to earth, and which can be employed to bias the control grid of a further valve to which the alternating component of the signal is fed in a sense opposite to that of signals fed to valve 29. In another arrangement, the cathode of valve 29 is earthed, and the anode resistance 34 is inserted between the cathode and the negative terminal of a source of current which is arranged to supply the anode but not the screen grid; grid bias for the further valve can then be taken from the junction point of resistance 34 and the anode current source.

The signals fed to the further valve mentioned in the preceding paragraph may be obtained from lead 37 of Fig. 5, and another valve may be inserted if desired to reverse the phase of the signals. The magnitude of the direct component of the potential difference set up across resistance 34 may be adjusted so as to ensure correct re-insertion of the D.C. component by adjustment of the value of resistance 34.

In a further example of the use of this invention, a television wave such as that shown in Fig. 1, is transmitted through a channel wherein, although the absolute gain for the picture signals is substantially constant, the D.C. component and the amplification of the synchronising pulses vary owing to the variations say of H.T. voltage and the effect of curved amplification characteristics.

An example of such a use will be described with reference to Fig. 7. This Figure represents a typical form which the input circuit to the modulator of a transmitter for television signals may take. The amplified television signal wave such as is shown in Fig. 1 is fed in at 51 through a condenser 53 to the grid of a modulator 52. The sense of this wave is such that the picture signals are negative and the synchronising signals positive. The grid of the valve 52 is connected to the cathode through the leak 54, and this leak, together with a diode 55, serves to re-establish the direct current component of the applied signals on the grid of valve 52 with reference to the peak values P of the signals. The absolute value of the voltage of the grid 52 which represents the synchronising pulse is fixed by the negative bias supplied to the cathode of diode 55 through a connection 56, as will hereinafter be explained. The stabilisation (or re-inserting) circuit comprising elements 53, 54, 55 operates in the manner described in Patent Specification No. 422,906.

The anode circuit of the modulator valve 52 comprise a floating anode battery 57 and a modulator resistance 58, and the television signals appearing across resistance 58 thus contain their direct current component. The upper end of resistance 58 is connected through a source 59 for providing the desired bias voltage to the grids of two modulated radio-frequency amplifier valves 60 connected in push-pull; the grids of these valves are fed with radio frequency oscillations of carrier frequency through the coils 61 and 62. The radio frequency output from the anodes of the valves 60 is then passed through further radio frequency amplifiers (not shown) if required, to the aerial.

The wave form of the signals fed in at 51 is such that the ratio of synchronising signal amplitude to picture signal amplitude is larger than is required in the final modulated output of the transmitter. This relative increase of synchronising signal input allows for the curvature of the characteristic of the modulator and further amplifier stages.

It is arranged that the peaks P of the synchronising signals cause the radio frequency output to fall to substantially zero, which means that the synchronising signals must modulate the transmitter over the curved and less efficient lower part of the modulation and amplification characteristics. Since the synchronising signals are of a substantially square-topped wave form, there will be no wave-shape distortion other than a relative attenuation due to the operation over the curved part of the modulator characteristic, so that this pat of the characteristic can be used for the synchronising signals, leaving the upper straighter part for the picture signals. The resultant radio-frequency output at the aerial of the transmitter has then a range of values corresponding to particular light intensities in the object of which an image is to be transmitted, and a value, substantially equal to zero, corresponding to the peaks of the synchronising signals. Although this latter value is normally zero, it is often inconvenient to provide sufficient synchronising signal input to modulate the transmitted carrier right down to zero amplitude by swinging the modulator around the bottom bend of its characteristic.

At the receiver, correct representation of absolute brightness is obtained either by providing a direct current coupling between the demodulating detector and the light-modulating means, such as the control electrode of a cathode ray tube, or by allowing the D.C. component to be suppressed and then re-establishing the D.C. component with reference to the peak amplitude P of the synchronising pulses previous to application of the signals to the light-modulating device. If the first of these methods is employed and the received carrier amplitude representing black varies, the direct current or average picture brightness component will effectively vary at the receiver. A similar effect will occur in cases in which the D.C. component is lost at the receiver and subsequently re-inserted if the amplitude difference between the peaks P of the synchronising pulses and black in the received signals varies. For example, a slight fall of the carrier value representing black, or the amplitude of the synchronising pulses relative to black, will cause the screen of the cathode ray tube to be darkened. Such darkening will be relatively unimportant in the high lights, but may be sufficient to obscure detail in the dark parts of the picture, due to the very dark greys and blacks becoming simultaneously black.

Variation in the carrier amplitude representing black, or of the received amplitude representing the difference (P to B, Fig. 3) between the peaks of the synchronising pulses and black may be caused for example by variation in the voltage of the source supplying the anode circuits of the radio frequency amplifiers at the transmitter, or by variation in the voltage of battery 57 in Fig. 7. The effect at the receiver of such variations will be that changes in the average brightness of the transmitted picture will not be correctly represented in the reproduced picture: if the receiver is one in which the D.C. component is lost and subsequently reinserted, variations of the amplitude P to B will render the correct re-insertion of the D.C. component impossible. Further, in a D.C. coupled receiver, it is clear that if the synchronising pulses extend over the curved parts of the characteristics of the receiver valves, it is impossible (if the carrier amplitude representing black varies) to keep constant the amplitude of the peaks P of the synchronising pulses with respect to black (B) because different peaks will extend to different extents along the said curved parts and, when the variations in question are considerable, it may prove difficult to separate the synchronising signals correctly from the picture signals.

In order to correct for variations of the kind above described, therefore, it may be arranged, according to a feature of the present invention, to feed in at point 56 in Fig. 7 a voltage which varies in such a manner as to keep the "black" level of the carrier output at a constant value in spite of the variations referred to.

One way in which this may be done is to feed modulated carrier frequency energy from the transmitter of Fig. 7, or from the transmitting aerial, to the circuit 14 of an arrangement such a that of Fig. 5. Signals of the form shown by 32 in Fig. 2 are applied at point 30 in Fig. 5. These signals may, for example, be generated with the aid of a circuit such as that of Fig. 4, or may be derived from part of the mechanism which serve to generate the synchronising signals. The output of the valve 29 will be proportional to the carrier amplitude corresponding to black, less any effect due to slight variation in the carrier amplitude representative of the peaks P of the synchronising signals, if this is not quite zero. The A.C. component of the output of valve 29 may therefore be fed, after any necessary amplification, to terminal 41, and thence into the device of Fig. 6; here, as already explained, there are developed across resistance 49 voltages which can be arranged to vary substantially with relatively small variations in the level corresponding to black in the signals fed in at 37 in Fig. 5.

The output obtained at point 50 in Fig. 6 may then be applied to control the potential of point 56 in Fig. 7. In the particular arrangement shown in Fig. 6, an increase in "black" amplitude produces a negative voltage at 50, whereas a positive voltage is required at point 56 of Fig. 7 in order to correct such an increase in amplitude. To overcome this difficulty, the sense of the input to the diode 44 in Fig. 6 may be made opposite to that previously described; the diode 44 and battery 47 are then also reversed. The point 50 in Fig. 6 may then be connected, by means of a direct-current connection which may if necessary include a source of biasing potential, to point 56 in Fig. 7 to give the desired result.

Alternatively, there may be applied to point 56 of Fig. 7 a voltage proportional to the direct component of the current flowing in the valve 29 of Fig. 5; this may be achieved, for example, by inserting a resistance in the cathode circuit of the valve 29, as described above in connection with the re-establishment of D.C., and the cathode of the valve 29 may then be connected to point 56 of Fig. 7.

Instead of injecting the control voltage to point 56 of Fig. 7, it can be arranged that the potential at point 50 in Fig. 6 is effective in controlling the amplitude of the synchronising signals applied at the point 51. For example, before the synchronising signals are mixed with the picture signals, they may be passed through an amplifier having its amplification controlled suitably in accordance with the voltage at point 50 of Fig. 6. If the amplification is controlled by variation in the bias of one or more variable-mu valves, a lead may be taken from the point 50 to the control grids of these valves in order to control the bias thereof. In such an arrangement, the apparatus of Fig. 6 is connected up as shown, and signals are applied at point 41 in the same sense as that mentioned in the original description of that Figure. An increase in black amplitude accordingly produces a negative voltage at point 50, tending to reduce the amplification of the synchronising signal amplifier and thus tending to reduce the black amplitude.

In a system in which the carrier amplitude is not reduced to the neighbourhood of zero on the peaks of the synchronising signals, it can be arranged that the black signals alone are effective in producing a voltage for controlling the black level. Such an arrangement would not take count of changes in the amplitude P to B in Fig. 3 unless such changes were accompanied by changes in the black level. This result can be obtained with the circuit of Fig. 5 by biasing the valve 29 so that the synchronising signals carry the outer control grid well below cut-off, so that variation in the amplitude of the synchronising signals is not effective in changing the anode current. the anode current is then controlled in accordance with changes in the amplitude of the black signals (B Fig. 3).

Where it is undesirable or inconvenient to apply a correction to the synchronising signals before they are mixed with the picture signals, the synchronising signal amplitude may be increased or decreased relatively to the picture signal amplitude by passing the composite signal comprising mixed synchronising and picture signals through a suitable circuit. Such a circuit may comprise a thermionic valve, the curvature of the characteristic of which can be varied under the control of a voltage such as that obtained from point 50 in Fig. 6. An example of a circuit of this kind is shown in Fig. 10. The composite signal, having the form shown in Fig. 1 for example, is fed through a condenser 71 on to the grids of two valves 74 and 75; the signal fed in loses its D.C. component, if that component is present, in its passage through condenser 71. The sense of the signals is such that the synchronising signals are positive. A diode 72 and leak 73 serve to re-insert the D.C. component into the signals with reference to the peaks P of the synchronising signals. The cathode of valve 74 is connected to earth through a resistance 78 which reduces the mutual conductance of the valve and lengthens and straightens its characteristic. The valve 75 has its cathode biased very positively so that the valve is inoperative for the negative picture signals on its grid but is operative to give increased amplification for the synchronising signals. The combined output of the valves 74 and 75 is passed out at 76.

A control voltage, such as that from point 50 in Fig. 6, is applied at point 7 and is effective in altering the bias on the valves 74 and 75. Thus an increase in the "black" level of the outgoing signals will produce a negative voltage at 77 which will reduce the amplitude of the synchronising signals, so tending to neutralise the change in "black" level.

In the systems so far described, the corrective effect is derived from an "observation" of the "black" level during the intervals between frames.

Where the line synchronising signals do not occupy the whole of the interval between successive trains of picture signals representative of successive lines of the object, however, use may be made of an "observation" in this interval (which may be called the line interval) to derive a corrective effect. A line synchronising impulse usually occupies about one tenth of a line period and the other nine tenths is usually occupied by picture signals.

In Fig. 8 is shown a wave form in which the line synchronising impulses 2 occupy only a fraction of the line interval D, for example a quarter of this interval. Thus the impulse lasts for one fortieth of a line period and during three fortieths of a line period the signal is at "black," indicated by reference B. with such a signal it is possible to derive a corrective effect once every line and a more rapid control is therefore possible than with the signal of Fig. 1. Signals of the kind shown in Fig. 8 may be useful for example for providing automatic gain control for the receiver of a relay station receiving signals from a moving van or over some other channel subject to rapidly varying attenuation, with a view to re-transmitting the signals after correction.

If the maximum possible rapidity of action is desired with signals of the kind shown in Fig. 8, he signals in the frame interval should preferably differ from those shown in Fig. 1. Thus each of the pulses 5, Fig. 1, may be constituted by an excursion (at F, L and L1) o level P to form a pulse lasting one fortieth of a line period, a return to level B for an interval of three fortieths of a line period, a further excursion to level P for a time equivalent to three tenths of a line period followed by a return to level B.

The signals of Fig. 8 may be utilised in a circuit which is a modification of that shown in Fig. 4. In the modification, the valve 11 and resistance 10 are omitted, and the output of the receiver 6 is taken from the choke 9 to the grid of the valve 16. Further, the rectifier 6 is inverted, so that picture signals produce a rectified voltage in the negative sense. The valve 16 is so adjusted or biased that is does not pass current for the negative voltages representing picture signals. At the occurrence of a line synchronising signal however, the valve 16 conducts, thus triggering the multi-vibrator comprising valves 19 and 20. The leak 24 is adjusted so that the valve 19 remains insulating from the moment of triggering to just before the beginning of the next train of picture signals, that is for most of the one-tenth of line period. Instead of making the condenser 21 so large that the multi-vibrator is quasi-aperiodic, it may be found advantageous to make the natural period of the multi-vibrator just longer than a line period, so that the device tends to run at the required frequency. The amplitude of the pulses required from valve 16 is thus reduced. The output at points 27 and 28 then takes the form of pulses occurring at the line frequency, each pulse having a length slightly less than one tenth of a line period.

The pulses from point 28 may be fed in at point 30 to the apparatus of Fig. 5, and may control the valve 29 to give output signals at 36 dependent on the black amplitude occurring in the short interval between a line synchronising pulse and the beginning of the picture signals f the next line. The signals set up at point 36 may be amplified and passed to a circuit such as that of Fig. 6 for producing a control voltage at 50. In this case, the time constant of condenser 45 and leak 46 may be made much smaller than that employed for "observation" once per frame, so that a more rapid control is obtained. The time constant of elements 45 and 46 should, however, be made sufficiently greater than the line period to prevent any instability of the control.

The examples given above are of systems in which the black level is "observed" and a corrective signal dependent on this observation is injected at a point earlier in the system than that at which the observation is made so as to correct as far as possible any variation of black level at the point of observation. The invention is also applicable to systems in which the controlling signal is utilised to correct the black level at some point after the point of observation. For example, in the automatic gain control systems described above, the control signal developed at point 50 in Fig. 6 may be used to control the gain of amplifiers following the point of observation. Alternatively, where large variations of transmission efficiency are to be corrected, it may be arranged that the corrective signal serves to vary the amplification both before and after the point of observation. For example, the control voltage at point 50 in Fig. 6 my be utilised to control the radio frequency amplification ahead of the rectifier 6 in Fig. 5; such control my, for example, reduce the variation of the black level at the rectifier 6 in Fig. 5 to 10% for a 40 decibel change in incoming signal strength; the control voltage at point 50 in Fig. 6 may also serve to produce a slight variation in the amplification following the observation point (e.g. the modulation frequency amplification following the rectifier) so that a 10%change of black level at the observation point produces a 10% compensating change of gain following it, thus ensuring that the final output signal is substantially free from any variation.

The arrangements so far discussed are based on an observation of the black amplitude. The invention can also be carried into effect by an observation of any definite picture amplitude, or of an amplitude related to a definite picture amplitude. For example, instead of the wave shown on Fig. 8, the wave shown in Fig. 9 may be transmitted. In this case, during the interval between the synchronising pulse and the beginning of the next line, the signal amplitude assumes a value E which lies between the synchronising peaks P and the black amplitude B. such a signal level may be fixed as being a certain fraction of the distance between P and B or, alternatively, if the amplitude of the synchronising impulses from the black level varies, may be defined as being below the level B by a certain fraction of the amplitude relatively to the level B of the maximum white signal W.

A signal of the type shown in Fig. 9 may be useful for direct reception by the use of a cathode ray tube, the slightly "blacker than black" signals E serving to black out the cathode ray during the scanning return strokes. Such a signal may be "observed" by the use of circuits similar to those employed for the wave of Fig. 8, but with suitable modification and adjustment. Similarly, a wave form may be employed in which, during the interval between a line synchronising signal and the picture signals of the next succeeding line, the signal amplitude corresponds to a value within the picture range, say half way between black B and white W. Such a wave form provides a strong controlling signal, but it may be found necessary to provide means to "black out" this signal so as to prevent its appearance in the reproduced picture.

Further, a wave of the kind last referred to may with advantage be used in a system which involves relaying a signal to a final transmitter; after the corrective signal has been derived and used to correct the signal, the excursion into the picture range occurring in the line intervals may then be suppressed before final transmission of the signals.

Similarly, wave forms of the kind shown in Figs. 8 and 9 may be converted before final transmission to the form shown in Fig. 1, for example by super-imposing upon the signals pulses such as those developed by the multi-vibrator 19, 20 of Fig. 4 when used with a wave of the kind shown in Fig. 1 and the limiting the resultant pulses to the required amplitude.

A radio relay station for a television transmission system may employ the corrective means according to the present invention more than once. For example, the wave received at the relay station may be of the general form shown in Fig. 8 and may be observed once per line. The radio-frequency gain may then be adjusted automatically as already described in dependence upon the observation. A further automatic adjustment of gain may then be carried out in order to correct for the error remaining as a result of the fact that slight changes in the black level at the observation point are necessary to develop a corrective signal for controlling the gain of a preceding amplifier. The corrected modulation frequency output may be passed through a circuit adapted to increase the relative amplitude of the synchronising signals so as to compensate for the subsequent reduction of the amplitude of these signals by the curved characteristic of the transmitter. Simultaneously, the wave form my, if necessary be converted to the form shown in Fig. 1 or Fig. 9 so as to be more suitable for final reception. The wave may then be applied to the transmitter and the radiated signal level corresponding to black may be controlled by one of the methods already described. For this latter control, observation may be made once per line for waves of the general form of Figs. 8 and 9 or once per frame for waves of the general form of Fig. 1.

In a further modified arrangement based upon that described with reference to Fig. 4, the synchronising pulses are again caused to generate the required switching signals, which in turn are employed to turn on the observing valve. In the modified arrangement, using the waveform shown in Figs. 8 or 9, the synchronising pulses P are separated from the vision signals in the manner described in Patent Specification No. 422,824. The separated pulses are then suitably delayed by a delay network and inverted by subjecting them o one stage of valve amplification, after which they can serve to make sensitive an observing valve such as valve 29 of Fig. 5 during a part of the period B or E in Figs. 8 and 9.

In a practical case, however, a wave form is employed which is similar to that shown in Fig. 8, but which differs in that the synchronising pulse lasts for one tenth of the line period while the black interval B lasts for one twentieth of the line period. Signals of this wave-form are illustrated on page 373 of the issue of "The Wireless World" of the 4th October 1935. The method last described is applicable here, but since the synchronising pulse lasts longer than the black interval the observing device, each time it is turned on, observes a part of the synchronising pulse as well as the black interval. This may be a disadvantage, and there is therefore preferably employed an alternative method in which the observing valve is turned on by a pulse obtained not from the leading edge of the synchronising pulse but from the trailing edge.

Apparatus for carrying this method into effect is illustrated in Fig. 11; signals of the form shown in Fig. 8, but with synchronising pulses lasting longer than the black intervals B, are fed in at 80 through a condenser 81 to a valve 82, it being arranged that the synchronising impulses are in the positive sense on the control grid of valve 82. The valve 82 serves in the manner described in Patent Specification No. 422,824 to separate the synchronising pulses from the picture signals, that is, its control grid tends to assume zero potential at the peaks P of the synchronising pulses, the picture signals lying beyond anode-current cut-off.

In the anode circuit of valve 82 is a small condenser 83 and a resistance 84 in series, the arrangement being such that differentiated synchronising pulses are set up across resistance 84; at the beginning of each pulse, sharp negative pulse appears at the top of resistance 84, while at the trailing edge of the pulse, a positive pulse appears at that point. The positive pulse causes the anode current of a valve 85 to increase; the anode circuit of valve 85 contains a resistance 86, and the potential of the anode of valve 85 consequently falls. A negative pulse is thus passed to the screen grid of valve 87, which, together with valve 88, forms a multivibrator, and the multivibrator is triggered off.

The multivibrator is so adjusted that a positive pulse is set up at the anode of valve 87 during each black interval B, and this pulse may be taken off at 89 and employed to actuate an observing valve, such as valve 29 in Fig. 5. It is to be noted that, in the practical waveform referred to above, the frame signals each comprise a plurality of pulses of longer duration than the line pulses, successive frame pulses being separated by black intervals; the multivibrator is thus triggered off by the trailing edges of both the line and frame pulses, and the observing valve is thus switched on during the black intervals following the frame pulses as well as during those following the line pulses.

If desired, the multivibrator 87, 88 may be omitted, the pulse from valve 85 being broadened by means of a low-pass filter, reversed in sense in an amplifying valve, and fed directly to the observing valve.

A further form of apparatus suitable for use in generating a signal for actuating an observing valve, which is particularly suitable for use with a signal of the form shown in Fig. 8, but with synchronising pulses of longer duration than the black intervals B, is illustrated in Fig. 12. Signals of the form mentioned are fed in at point 90, through condenser 91, to a valve 92, which serves the same function, and operates in the same manner as valve 82 of Fig. 11. Synchronising pulses, freed from picture signals, are fed from the anode of valve 92 to the inner grid of a hexode 93, where they appear in the negative sense, the arrangement being such that no anode current flows when a synchronising pulse is present on the inner grid.

The synchronising pulses are also fed, through a delay network comprising series inductances 94 and shunt condenser 95, to a reversing valve 96 having a resistance 97 in its anode circuit. Each synchronising pulse causes a positive pulse to appear at the anode of valve 96, and these positive pulses are fed through condenser 98 to the outer control grid of hexode 93; the outer control grid is connected through a leak resistance 99 to a source of grid bias 100, which serve to hold the outer grid at such a potential than, normally, no current flows to the anode.

The hexode 93 acts as a switch, the delayed positive pulses from valve 96 tending to open the switch, and the undelayed pulses applied to the inner grid tending to hold it closed. The delay positive pulses can thus only "open" valve 93, and allow anode current to flow therein, in the absence of negative pulses on the inner grid, and the delay introduced by network 94, 95 is made such that the valve 93 is opened for a part of the black interval of the signal immediately after each synchronising pulse. The anode circuit of valve 93 contains a resistance 101, and the pulses set up at the anode of valve 93 are taken off at 102, through condenser 103, and fed, after being reversed in sign, to an observing valve such as valve 29 of Fig. 5.

It is to be noted that phase error between the beginning of the switching pulse fed to the observing device and the occurrence of the black level or other datum signal to be observed may be compensated for by the use of suitable delay networks inserted in that part of the apparatus between the point where the switching portions of the check signals are feed from the datum portions, and the observing device. In cases in which the datum portions of the check signals precede the switching portions, in order to ensure that the datum portions and the switching portions are operative upon the observing device in the correct phase relation, a suitable delay network is provided in the channel along which the datum portions pass to the observing device.

In the above description, there have been described methods of correcting for varying attenuation or for complete or partial absence of the direct current component of signals. For this it has been shown to be sufficient to derive a corrective signal dependent upon the received amplitude of a signal which, at the transmitter, is a fixed value.

Where, however, it is desired to correct for varying attenuation of signals which have no direct current component, for example signals which have lost their D.C. component, it is necessary to derive a corrective signal dependent upon the difference between two different received amplitudes both of which have fixed values at the transmitter. Thus in the case of a signal of the form shown in Fig. 8, it may be assumed that at the transmitter the level B corresponds to picture black and that level P has a fixed difference from value B. now if such a signal be transmitted through a channel which is incapable of transmitting the D.C. component and which subjects the signals to varying attenuation, the procedure may be as follows:

The signals are given a datum co-incident with the peaks P with the aid of a D.C. reinserting device of the set forth in Specification No. 422,904 for example, and at the same time they are used to derive a corrective signal dependent upon the amplitude of level B, in the manner already described. As the D.C. reinserting device ensures that the datum remains on the peaks P and as the corrective signal makes the amplitude P to B substantially correct, the desired corrections will have been applied.

Although the invention has been described in some detail with reference to its application to television systems it is also applicable to other systems in which signals of a fixed amplitude value or a predetermined waveform comprising fixed amplitude values are either present inherently or are arranged to be present for the purpose of enabling a corrective effect to be derived according to the present invention.

Having now particularly described and ascertained the nature of my said invention and in what manner the same is to be preformed, I declare that what I claim is:-

  1. A method of correcting for variations in the effective amplitude of electrical signals representative of intelligence, such as may arise in the transmission o said signals as a result of the complete or partial loss of the D.C. component of said signals, the incorrect representation of that component, or varying attenuation of the signals, the said method comprising transmitting at spaced time intervals along the channel through which the intelligence signals are passed, check signals each of which has a switching portion and a datum portion, said datum portion having, at the input of said channel, either a predetermined fixed amplitude value or a predetermined wave form comprising fixed amplitude values and the said method being characterised in that the datum portions are applied through one path to an observing device, while the switching portions, or switching signals derived therefrom, are fed to said observing device through another path, and serve to change said observing device from the inoperative condition into the operative condition, said observing device, when in the operative condition, serving to develop a corrective signal dependent upon the amplitude or amplitude and waveform of the datum portions applied thereto, and said corrective signal being applied at a point either before or after the observing point to compensate wholly or in part for said variations in effective amplitude.
  2. Apparatus for carrying into effect the method according to claim 1, for use with composite signals comprising trains of intelligence signals interspersed with check signals, said apparatus comprising an observing device, a path for conveying the datum portions of said check signals to said observing device, and another path for conveying the switching portions of said check signals, or switching pulses derived from said switching portions, to said observing device and for enabling said switching portions, or the switching pulses derived therefrom, to change said observing device from the inoperative condition into the operative condition.
  3. Apparatus according to claim 2, wherein said observing device comprises a thermionic valve, and wherein the second-named path includes apparatus for producing electrical switching pulses under the control of the switching portions of the check signals and for applying said switching pulses to said valve to permit the flow of anode current therein under the control of the datum portions of the check signals.
  4. Apparatus according to claim 3, particularly for use in television systems in which said check signals are constituted by or include frame synchronising impulses, wherein there is provided a circuit adapted to generate said switching pulses under the control of said frame impulses.
  5. Apparatus according to claim 3, particularly for use in television systems, in which each of said check signals comprises a synchronising impulse of one amplitude followed by a datum portion of a different amplitude, the amplitude of the datum portion being fixedly related to the amplitude corresponding to picture black at least in the absence of variations in the effective amplitude of the composite signals of which said check signals form a part, wherein there a re provided means for utilising the synchronising impulses as switching pulses or to generate switching pulses, and delay means arranged to ensure that the switching pulses change the observing device into the operative condition during the occurrence of the datum portions at the observing device.
  6. Apparatus according to claim 3, particularly for use in television systems in which each of said check signals comprises a synchronising impulse of one amplitude followed by a datum portion of a different amplitude, the amplitude of the datum portion being fixedly related to the amplitude corresponding to picture black at least in the absence of variations in the effective amplitude of the composite signals of which said check signals form a part, wherein there are provided means for utilising the trailing edges of said impulses to control the generation of switching pulses.
  7. A receiver for carrying into effect the method of claim 1, and so designed as to be suitable for use in a transmission system for transmitted electrical intelligence signals containing a recurrent check signal, each check signal comprising two datum portions which, in the transmitted signal, have a fixed amplitude difference, and a switching portion either separate from said datum portions or constituted wholly or in part by one of said datum portions, said receiver comprising an observing device, a path for conveying the datum portions of said check signals to said observing device, and another path for conveying the switching portions of said check signals, or switching pulses derived from said switching portions, to said observing device and for enabling said switching portions, or the switching pulses derived therefrom, to change said observing device from the inoperative condition into the operative condition, said observing device being capable, when in the operative condition, of developing a corrective signal dependent upon variations in said amplitude difference, and means being provided for utilising said corrective signal to vary the gain of an amplifier of said receiver to compensate at least partially for said variations.
  8. A receiver according to claim 7, so designed as to be suitable for use in the reception of transmitted signals in which said datum portions are of substantially fixed amplitude, said receiver comprising means for re-inserting the direct current component of said intelligence signals with reference to one of said datum portions.
  9. The method of correcting for variations in the effective amplitude of electrical signals representative of intelligence, such as may arise in the transmission of said signals as a result of the complete or partial loss of the D.C. component of said signals, the incorrect representation of that component, or varying attenuation of the signals, substantially as herein described with reference to any of Figs. 4, 5, 6, 7 or 10 of the drawings filed with the Provisional Specification, or to the accompanying drawing.
  10. Apparatus for correcting for variations in the effective amplitude of electrical signals representative of intelligence, such as may arise in the transmission of said signals as a result of the complete or partial loss of the D.C. component of said signals, the incorrect representation of that component, or varying attenuation of the signals, substantially as herein described with reference to any of Figs. 4, 5, 6, 7 or 10 of the drawings filed with the Provisional Specification, or to the accompanying drawing.

Dated this 9th day of March, 1936.

REDDIE & GROSE,

Agents for the Applicant,

6, Breamís Buildings, London, E.C.4.

---------------------------

Leamington Spa: Printed for His Majestyís Stationery Office, by the Courier Press. Ė 1937.