515,361

PATENT SPECIFICATION

Application Date: May 30, 1938. No. 16081/38

(Patent of Addition to No. 458,585: dated March 20, 1935.)

Complete Specification Left: May 26, 1939.

Complete Specification Accepted: Dec. 4, 1939.

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

PROVISIONAL SPECIFICATION

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

We, ELECTRIC & MUSICAL INDUSTRIES LIMITED, a British Company, of Blyth Road, Hayes, Middlesex, and 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 apparatus for handling electrical signals having a direct current component, and is concerned with improvements or modifications of the invention described in the Parent Specification No. 458,585.

In the Parent Specification there is described 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 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 uppon 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.

There is described in said Parent Specification, on page 15, line 7 with reference to Fig. 7 of that Specification, a method of developing corrective signals wherein the corrective signals are applied to a point prior to the observing point in order to correct for any errors in the direct current component, but not to re-insert the direct current component, the latter being carried out by other means.

Briefly, the D.C. component is re-inserted in a television signal, by means of the method described in Patent Specification No. 422,906, at the grid of a modulator valve. The resultant modulated carrier wave is rectified and the resultant D.C. datum portion is observed and the corrective signal is passed from the observing device back to the re-insertion device, thus correcting for any variation in the datum level in the transmitter output.

It is the object of the present invention to provide improvements in or modifications of the invention described in Parent Specification No. 458,585.

According to the invention 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 signal according to Claim 1 of the Specification of the Parent Patent No. 458,585, wherein said signals, after the D.C. component is lost, are applied to a D.C. coupled amplifier and said switching portions, or switching signals derived therefrom, are fed to said observing device from a point prior to or subsequent to said D.C. coupled amplifier to switch the observing device from the inoperative to the operative condition to observe the datum portions after the latter have passed through said D.C. coupled amplifier and developing a correcting signal from the datum levels observed by said device, which correcting signal is fed back to said D.C. coupled amplifier in such a manner as to cause said datum portions to have a substantially constant value at the output of said D.C. coupled amplifier.

According to one arrangement of the invention, television signals are applied to the input of a D.C. coupled amplifier and the amplified output signals are applied to one grid of a hexode valve and delayed synchronising pulses are applied to the other grid of the hexode valve so that the hexode valve is only switched on during the datum potions of the television signal which correspond to black. The resultant pulses of anode current in the hexode valve produce potential pulses which are rectified and the rectified signal is fed back to the input of the D.C. coupled amplifier.

For the purpose of describing the invention more in detail, reference will now be made to the accompanying drawings, which illustrate particular embodiments according to the invention.

In order to simplify the description, reference will also be made to the drawings accompanying the Parent Specification No. 458,585, and for this reason the drawings accompanying the present Specification have been numbered Figs. 13 and 14. Any reference in the following description to Figures other than Figs. 13 and 14 refer to the Parent Specification. Identical reference numerals have been used in the drawings to represent parts which have similar functions.

Referring to Fig. 13 which shows one embodiment of the invention, television signals which may be of the form shown in Fig. 8 of the Parent Specification are applied at terminal 201 with or without the D.C. component present and are passed through the A.C. coupling consisting of condenser 202 and resistance 203. The signals are then amplified by the D.C. coupled amplifier 204 and pass out of line 205. It is assumed in the arrangement shown that the amplifier 204 gives an output signal in the same sense as the input signal. In addition, it is assumed that the television signals are applied at the terminal 201 with the picture signals in the positive sense and therefore the signals in the output at 205 have also picture amplitudes in the positive sense. The amplified signals at line 205, which may be of quite large amplitude, are applied through a potentiometer to the grid of a phase reversing valve 206, in the anode circuit of which is a delay network 207 from which the signals are applied through a condenser and leak resistance to the inner grid of the hexode valve 208. The valve 206 is a pentode and the screening grid is supplied with a suitable positive potential. The signals are applied through a potentiometer to valve 206 in order to prevent this valve from overloading. The delay network 207 comprises the usual series inductances and shunt capacitances and the anode resistance of valve 206 serves to terminate the delay network.

The delayed synchronising signals applied to the inner grid of the hexode are in a positive sense, and are of such an amplitude as to exceed the grid base for the inner grid of the hexode 208. As explained in the Patent Specification No. 422,906, the D.C. component of the signal will be re-established at this point and the valve will only pass current during the periods when these (delayed) synchronising pulses are applied to the inner grid of the valve. The synchronising signals are delayed so that when they arrive at valve 208 they coincide with the arrival of the datum potion B representing black (see Fig. 8) on the outer grid of the hexode 208 which grid is connected to line 205. If the duration of the synchronising pulse P (see Fig. 8) is shorter than the duration of the datum portion B, then P can be delayed to lie totally within B. If P is longer than B, the delay should be slightly less than the duration of B. The fact that the hexode 208 will be switched on before the occurrence of the datum portion B does not matter, since during this time the signals on the outer grid will correspond to P, that is, they will be very negative and so will prevent the valve passing any anode current. During the occurrence of the datum portion B the inner grid of the hexode valve will be switched on. The absolute potential of the outer grid during this period will be absolute value of B, and so the anode current of the valve will be determined by the difference of the absolute value of the grid potential and the potential of the cathode battery 209, which in effect provides a reference potential may be zero if the desired level of the datum portion B at lead 205 is close to zero. Since during periods other than the datum period B the valve 208 will be switched off, the anode current will consist of a series of pulses, the absolute value of which depends on the difference between the absolute value of the potential of the datum potion B at 205 and the potential of the battery 209. The signal amplitude at lead 205 may very much exceed the grid base of valve 208, since the outer grid is inactive except during the datum periods B. The current pulses flowing through the anode resistance of the valve 208 produce corresponding potential pulses, which are applied to the rectifiers 210 (shown as diode valves operating in a voltage doubler type of circuit) with a condenser and leak resistance circuit 211, which generates a negative bias at the input of the amplifier 204, tending to make the output signal negative. Any positive drift of potential of the datum portion B at lead 205 will increase the output of the hexode 208 and hence the negative input bias of the amplifier 204 is increased, thus counteracting the drift at the output.

As an example of the operation of the circuit, suppose that the mutual conductance of the outer grid of hexode 208 to the anode is 1.0 milliampere per volt and the anode resistance of valve 208 has a value of 2000 ohms: suppose also, that the rectifier valves 210 are 80% efficient and that the amplifier 204 has a gain of 100. Now, if the D.C. level, that is, the absolute level of the datum portion B at the input of the amplifier 204 tends to drift, due to the coupling 202 and 203, by 0.8 volts, which requires 0.8 volts from the rectifier 210 to compensate for the drift, then a change of pulse amplitude of 1 volt from the valve 208, that is, 0.5 milliampere change of current or 0.5 volt change of potential on the outer grid of the hexode is required. The datum level of B would therefore only change at line 205 by about 0.5 volt in place of 0.8 x 100, that is, 80 volts had there been no corrective device. The battery 212 is inserted between the rectifiers 210 and the leak resistance 203 to ensure that the bias at the input of amplifier 204, in the absence of any output from rectifiers 210, is always too positive so that it can be corrected by the rectifiers.

The above arrangement has been described with reference to an amplifier which produces an output signal in the same sense as the signal applied to the input. The circuit arrangement of Fig. 13 may be modified to suit other conditions, for example, if the picture signals are applied to the input of amplifier 204 in a negative sense and the amplifier is such that a positive picture output is obtained, then the rectifier diodes 210 must be reversed so that they generate a positive correcting potential. Similarly, for a negative output the phase reversing valve 206 may be omitted, although if the duration of the synchronising pulses P is longer than the datum portion B, it will be necessary to generate short pulses by such a circuit as shown in Figs. 12 of the Parent Specification. Similarly, as described with reference to Fig. 7 of the Parent Specification, the amplifier 204 may include a carrier frequency modulator and in this case a rectifier is required between line 205 and the connections to valve 206 and valve 208. Referring to Fig. 7 of the Parent Specification, if the diode valve 55 is omitted and the leak resistance 54 replaces it between the grid of valve 52 and the lead 56, the circuit of Fig. 13, together with the other apparatus described in connection with it, provides one example of this method of D.C. re-insertion applied to the output circuit of a radio transmitter.

A further modification of the invention is shown in Fig. 14. In this Figure the D.C. coupled amplifier 204 has two input circuits; for example, the input circuit of the amplifier may comprise two valves, the anodes of which are connected in parallel and the input grids of these two valves provide the input terminals for the amplifier. It is also assumed in connection with this Figure that the amplifier 204 produces a phase reversal between the input and output circuits. In addition, it is assumed that the amplifier 204 produces a potential of substantially zero volts at the output 205 when the bias applied to both inputs of the amplifier is zero. Due to the presence of the positive bias battery 212, the output at line 205 tends to be very negative unless corrected by the negative bias produced by rectifier 210, which, in this case, is represented in a half-wave rectifier circuit. The two valves 214 and 215 are provided with a large common cathode resistance which is taken to a source of negative potential. The anodes of both these valves are taken to suitable positive potentials, the grid of valve 215 being connected directly to the line 205 and the grid of valve 214 being connected via a coupling condenser and grid resistance to the device 213 which represents an arrangement for producing negative pulses on the grid of valve 214 during only those periods when line 205 is at datum potential, that is, during the datum periods. Such pulse producing devices are described in the Parent Specification and are illustrated, for example, in Fig. 12 of that Specification. The cathode of valve 214 is normally biased positively with respect to the potential of line 205 so that the cathode of valve 215 is held positive irrespective of the potential on its grid, thus keeping diode 210 from conducting. The negative pulses on the grid of valve 214 cause the cathode potential of this valve to become negative until valve 215 conducts. The cathode potential and thus the output potential of valve 210 then depends on the potential of the grid of valve 215 during the datum period. The potential stored in condenser 216 is passed to the input of the amplifier 204 so as to correct the output datum level to a value substantially equal to zero volts (or slightly negative of zero due to the bias required by valve 215). The time constant of condenser 216 and its leak resistance must be long compared with the intervals between successive datum portions but must be short enough having regard to the fact that the potential across it is, in effect, amplified by amplifier 204, to enable correction to be made for the errors in the D.C. component of the input signal. Suitable values for the components for the diode load impedance can be found experimentally, although, of course, correction by means already described can only be made for changes in the D.C. level of the input signals which are slow compared with the interval between successive datum periods. This also applies to the condenser and leak resistance of the rectifiers 210 in Fig. 13.

The arrangement described above can be used omitting condenser 202 and resistance 203 for an input signal which contains the D.C. component to correct for any errors in the input or to correct for any errors in the D.C. amplification of amplifier 204.

The arrangements of Figs. 13 and 14 can also be operated with the switching pulse derived from the input, or from an intermediate point in the amplifier. For example, in Fig. 14 the device 213 may receive its input from the input to amplifier 204. In this case an amplifier may be required in the device 213 in order to generate sufficiently large pulses. This arrangement has the advantage that any faults in the amplifier 204 which might overload it and reduce the amplitude of the switching pulses cannot affect the D.C. re-insertion. Such reduction in the amplitude of the synchronising pulses might be a disadvantage with the arrangement of Fig. 13 when first switching on. This can always be corrected, however, by a manually controlled bias applied to a resistance 203.

Dated this 28th day of May, 1938

F. W. Cackett

Chartered Patent Agent

COMPLETE SPECIFICATION

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

We, ELECTRIC & MUSICAL INDUSTRIES LIMITED, a British Company, of Blyth Road, Hayes, Middlesex, and 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, and is concerned with improvements or modifications of the invention described in the Parent Specification No. 458,585.

In the Parent Specification there is described 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 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 uppon 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.

There is described in said Parent Specification, on page 15, line 7 with reference to Fig. 7 of that Specification, a method of developing corrective signals wherein the corrective signals are applied to a point prior to the observing point in order to correct for any errors in the direct current component, but not to re-insert the direct current component, the latter being carried out by other means.

Briefly, the D.C. component is re-inserted in a television signal, by means of the method described in Patent Specification No. 422,906, at the grid of a modulator valve. The resultant modulated carrier wave is rectified and the resultant D.C. datum portion is observed and the corrective signal is passed from the observing device back to the re-insertion device, thus correcting for any variation in the datum level in the transmitter output.

It is the object of the present invention to provide improvements in or modifications of the invention described in Parent Specification No. 458,585.

According to the invention there is provided 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 signal, or the incorrect representation of that component, the said method comprising transmitting at spaced time intervals along the channel through which the intelligence signals are passed, check signals each of which are a switching portion and a datum portion, said datum having, at the input of said channel, either a predetermined fixed amplitude value or a predetermined waveform comprising fixed amplitude values and the said method being characterised in that the datum portions are applied through one path which includes a direct current amplifier to an observing device and said switching portions or switching signals derived therefrom, are fed to said observing device through another path to switch the observing device from the inoperative to the operative condition to observe the datum portions after the latter have passed through said direct current amplifier, and observing device, when in the operative condition, serving to develop a correcting signal from the datum portions observed by said device which correcting signal is fed back to said direct current amplifier in such a manner as to cause said datum portions to have a substantially constant value at the output of said direct current amplifier.

According to one arrangement of the invention, television signals are applied to the input of a D.C. coupled amplifier and the amplified output signals are applied to one grid of a hexode valve forming the observing device and delayed synchronising pulses are applied to the other grid of the hexode valve so that the hexode valve is only switched on during the datum potions of the television signal which correspond to black. The resultant pulses of anode current in the hexode valve produce potential pulses which are rectified and the rectified signal is fed back to the input of the D.C. coupled amplifier.

In order that the said invention may be clearly understood and readily carried into effect it will now be more fully described with reference to the drawings accompanying the Provisional Specification which illustrate particular embodiments according to the invention.

In order to simplify the description, reference will also be made to the drawings accompanying the Parent Specification No. 458,585, and for this reason the drawings accompanying the present Specification have been numbered Figs. 13 and 14. Any reference in the following description to Figures other than Figs. 13 and 14 refer to the Parent Specification. Identical reference numerals have been used in the drawings to represent parts which have similar functions.

Referring to Fig. 13 which shows one embodiment of the invention, television signals which may be of the form shown in Fig. 8 of the Parent Specification are applied at terminal 201 with or without the D.C. component present and are passed through the A.C. coupling consisting of condenser 202 and resistance 203. The signals are then amplified by the D.C. amplifier 204 and pass out of line 205. It is assumed in the arrangement shown that the amplifier 204 gives an output signal in the same sense as the input signal. In addition, it is assumed that the television signals are applied at the terminal 201 with the picture signals in the positive sense and therefore the signals in the output at 205 have also picture amplitudes in the positive sense. The amplified signals at line 205, which may be of quite large amplitude, are applied through a potentiometer to the grid of a phase reversing valve 206, in the anode circuit of which is a delay network 207 from which the signals are applied through a condenser and leak resistance to the inner grid of the hexode valve 208 which forms the observing device. The valve 206 is a pentode and the screening grid is supplied with a suitable positive potential. The signals are applied through a potentiometer to valve 206 in order to prevent this valve from overloading. The delay network 207 comprises the usual series inductances and shunt capacitances and the anode resistance of valve 206 serves to terminate the delay network.

The delayed synchronising signals applied to the inner grid of the hexode are in a positive sense, and are of such an amplitude as to exceed the grid base for the inner grid of the hexode 208. The valve will only pass current during the periods when these (delayed) synchronising pulses are applied to the inner grid of the valve. The synchronising signals are delayed so that when they arrive at valve 208 they coincide with the arrival of the datum potion B representing black (see Fig. 8) on the outer grid of the hexode 208 which grid is connected to line 205. If the duration of the synchronising pulse P (see Fig. 8) is shorter than the duration of the datum portion B, then P can be delayed to lie totally within B. If P is longer than B, the delay should be slightly less than the duration of B. The fact that the hexode 208 will be switched on before the occurrence of the datum portion B does not matter, since during this time the signals on the outer grid will correspond to P, that is, they will be very negative and so will prevent the valve passing any anode current. During the occurrence of the datum portion B the inner grid of the hexode valve will be switched on. The absolute potential of the outer grid during this period will be absolute value of B, and so the anode current of the valve will be determined by the difference of the absolute value of the grid potential and the potential of the cathode battery 209, which in effect provides a reference potential. Of course, this reference potential may be zero if the desired level of the datum portion B at lead 205 is close to zero. Since during periods other than the datum period B the valve 208 will be switched off, the anode current will consist of a series of pulses, the absolute value of which depends on the difference between the absolute value of the potential of the datum potion B at 205 and the potential of the battery 209. The signal amplitude at lead 205 may very much exceed the grid base of valve 208, since the outer grid is inactive except during the datum periods B. The current pulses flowing through the anode resistance of the valve 208 produce corresponding potential pulses, which are applied to the rectifiers 210 (shown as diode valves operating in a voltage doubler type of circuit) with a condenser and leak resistance circuit 211, which generates a negative bias at the input of the amplifier 204, tending to make the output signal negative. Any positive drift of potential of the datum portion B at lead 205 will increase the output of the hexode 208 and hence the negative input bias of the amplifier 204 is increased, thus counteracting the drift at the output.

As an example of the operation of the circuit, suppose that the mutual conductance of the outer grid of hexode 208 to the anode is 1.0 milliampere per volt and the anode resistance of valve 208 has a value of 2000 ohms: suppose also, that the rectifier valves 210 are 80% efficient and that the amplifier 204 has a gain of 100. Now, if the D.C. level, that is, the absolute level of the datum portion B at the input of the amplifier 204 tends to drift, due to the coupling 202 and 203, by 0.8 volts, which requires 0.8 volts from the rectifier 210 to compensate for the drift, then a change of pulse amplitude of 1 volt from the valve 208, that is, 0.5 milliampere change of current or 0.5 volt change of potential on the outer grid of the hexode is required. The datum level of B would therefore only change at line 205 by about 0.5 volt in place of 0.8 x 100, that is, 80 volts had there been no corrective device. The battery 212 is inserted between the rectifiers 210 and the leak resistance 203 to ensure that the bias at the input of amplifier 204, in the absence of any output from rectifiers 210, is always too positive so that it can be corrected by the rectifiers.

The above arrangement has been described with reference to an amplifier which produces an output signal in the same sense as the signal applied to the input. The circuit arrangement of Fig. 13 may be modified to suit other conditions, for example, if the picture signals are applied to the input of amplifier 204 in a negative sense and the amplifier is such that a positive picture output is obtained, then the rectifier diodes 210 must be reversed so that they generate a positive correcting potential. Similarly, for a negative output the phase reversing valve 206 may be omitted, although if the duration of the synchronising pulses P is longer than the datum portion B, it will be necessary to generate short pulses by such a circuit as shown in Fig. 12 of the Parent Specification. Also, for D.C. reinsertion as applied to a radio transmitter, the amplifier 204 may include a carrier frequency modulator such as described in Figure 7 of the Parent Specification and in this case a rectifier to rectify the modulated carrier wave is required between line 205 and the connections to valve 206 and valve 208. In such a case the diode valve 55 is of said Figure 7 is omitted together with the leak resistance 54.

A further modification of the invention is shown in Fig. 14. In this Figure the D.C. amplifier 204 has two input circuits; for example, the input circuit of the amplifier may comprise two valves, the anodes of which are connected in parallel and provided with a common output impedance, and the input grids of these two valves provide the input terminals for the amplifier. The television signals are fed to one of the valves and the correcting signal is fed to the other of said valves. It is also assumed in connection with this Figure that the amplifier 204 produces a phase reversal between the input and output circuits. In addition, it is assumed that the amplifier 204 produces a potential of substantially zero volts at the output 205 when the bias applied to both inputs of the amplifier is zero. Due to the presence of the positive bias battery 212, the output at line 205 tends to be very negative unless corrected by the negative bias produced by rectifier 210, which, in this case, is represented in a half-wave rectifier circuit. The two valves 214 and 215 form the observing device and are provided with a large common cathode resistance which is taken to a source of negative potential. The anodes of both these valves are taken to suitable positive potentials, the grid of valve 215 being connected directly to the line 205 and the grid of valve 214 being connected via a coupling condenser and grid resistance to the device 213 which represents an arrangement for producing negative pulses on the grid of valve 214 during only those periods when line 205 is at datum potential, that is, during the datum periods. Such pulse producing devices are described in the Parent Specification and are illustrated, for example, in Fig. 12 of that Specification. The cathode of valve 214 is normally biased positively with respect to the potential of line 205 so that the cathode of valve 215 is held positive irrespective of the potential on its grid, thus keeping diode 210 from conducting. The negative pulses on the grid of valve 214 cause the cathode potential of this valve to become negative until valve 215 conducts. The cathode potential and thus the output potential of valve 210 then depends on the potential of the grid of valve 215 during the datum period. The potential stored in condenser 216 is passed to the input of the amplifier 204 so as to correct the output datum level to a value substantially equal to zero volts (or slightly negative of zero due to the bias required by valve 215). The time constant of condenser 216 and its leak resistance must be long compared with the intervals between successive datum portions but must be short enough having regard to the fact that the potential across it is, in effect, amplified by amplifier 204, to enable correction to be made for the errors in the D.C. component of the input signal. Suitable values for the components for the diode load impedance can be found experimentally, although, of course, correction by means already described can only be made for changes in the D.C. level of the input signals which are slow compared with the interval between successive datum periods. This also applies to the condenser and leak resistance of the rectifiers 210 in Fig. 13.

The arrangement described above can be used omitting condenser 202 and resistance 203 for an input signal which contains the D.C. component to correct for any errors in the input or to correct for any errors in the D.C. amplification of amplifier 204.

The arrangements of Figs. 13 and 14 can also be operated with the switching pulse derived from the input, or from an intermediate point in the amplifier. For example, in Fig. 14 the device 213 may receive its input from the input to amplifier 204. In this case an amplifier may be required in the device 213 in order to generate sufficiently large pulses. This arrangement has the advantage that any faults in the amplifier 204 which might overload it and reduce the amplitude of the switching pulses cannot affect the D.C. reinsertion. Such reduction in the amplitude of the synchronising pulses might be a disadvantage with the arrangement of Fig. 13 when first switching on. This can always be corrected, however, by a manually controlled bias applied to a resistance 203.

It has been proposed in the Specification of Patent No. 464,828 to provide for use in a television or like transmission system a method of reinserting the direct current component in a complex signal from which the direct current component has been removed and having a periodic maxima or minima which would have a constant value were the direct current component present, which method consists in amplifying the complex signal obtained at the point of reinsertion by means of a direct current amplifier and feeding back from the output of said amplifier a component which tends to keep the value of the periodic maxima or minima at a predetermined constant value. The present invention, however, distinguishes from this prior proposal in that it utilises the method of D.C. reinsertion referred to in the Specification of the parent Patent No. 458,585.

Having now particularly described and ascertained the nature of our said invention and in what manner the same is to be performed, we declare that what we 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 of said signals as a result of the complete o partial loss of the D.C. component of said signals, or the incorrect representation of that component, 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 waveform comprising fixed amplitude values and the said method being characterised in that the datum portions are applied through one path which includes a direct current amplifier to an observing device and said switching portions or switching signals derived therefrom are fed to said observing device through another path to switch the observing device from the inoperative to the operative condition to observe said datum portions after the latter have passed through said direct current amplifier, said observing device, when in the operative condition, serving to develop a correcting signal from the datum portions observed by said device which correcting signal is fed back to said direct current amplifier in such a manner as to cause said datum portions to have a substantially constant value at the output of said direct current amplifier.
  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 including a direct current amplifier 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 during the datum portions so that said observing device is caused to observe said datum portions and means for developing a corrective signal from said datum portions and for applying said corrective signal to the input of said amplifier.
  3. Apparatus according to claim 2, wherein said observing device is a hexode valve and the output of said amplifier is fed to one grid of said hexode valve and means are provided for applying said switching portions or switching signals to the other grid of te hexode valve so that said hexode valve enables a corrective signal to be developed under the control of said datum periods.
  4. Apparatus according to claim 2, wherein the observing device comprises a pair of valves having a common cathode impedance one of said valves having controlling pulses applied thereto so as to cause the other valve to be rendered conducting during the datum periods, the output of the amplifier being connected to a control electrode of said other valve.
  5. Apparatus according to claim 2, 3 or 4, wherein the output of said observing device is fed to a rectifier with a suitable smoothing circuit which develops said corrective signal.
  6. Apparatus according to claim 5, wherein a pair of said rectifiers is provided arranged to form a voltage doubling circuit.
  7. Apparatus according to any of the preceding claims 2 to 6, wherein the direct current amplifier has a pair of input circuits each comprising a thermionic valve having a common output impedance the electrical signals being fed to one of said valves and the correcting signal being fed to the other of said valves.
  8. A method of correcting for variation in the effective amplitude of electrical signals substantially as described with reference to Fig. 13 or 14 of the drawings accompanying the provisional specification.
  9. Apparatus for correcting for variations in the effective amplitude of electrical signals substantially as described with reference to Figure 13 or 14 of the drawings accompanying the Provisional Specification.

Dated this 25th day of May, 1939

F. W. Cackett

Chartered Patent Agent