US2299944A - Direct current reinserting circuit - Google Patents

Description

Oct. 27, 1942. K. R. wEND'r DIRECT CURRENT REINSERTING CIRCUIT Filed Oct. 23, 1940 2 Sheets-Sheet 1 7 a w W m w c a 0 m w a m a m m 4 n z 1 1 1 flw v v.\ a M 1 6 a Q m o a Karl E. Wen

Oct. 27, 1942. wEN 2,299,944

DIRECT CURRENT REINSERTING CIRCUIT Filed Oct. 25, 1940 2 Sheets-Sheet 2 Fire. 3.

zaommf l LEVEL GETTING I Vol; W765 CL/PvS OF'F' PICTURE Jnnentor Karl E. Wendi (Ittomeg as synchronizing pulses.

Patented on. 27, 1942 UNITED STATES PATENT OFFICE 2,299,944 DIRECT CURRENT REINSERTING CIRCUIT Karl E. Wendi, Audubon, N. 1., assignor to Radio Corporation of America, a corporation of Dela- Application October 23, 1940, Serial No. 362,311

8 Claims. (01. 178-73) My .invention relates to the reinsertion of direct current and/or low frequency components of electrical signals, and particularly to the reinsertion of such components in a television system.

Various circuits have been proposed for the reinsertion of the above-mentioned components where the components have been suppressed during transmission, as, for example, where they are suppressed by transformers or by alternating current amplifiers. These circuits depend for their operation upon 'the transmission of periodically recurring electrical pulses which are caused to go to a fixed voltage level, such as black," in a picture, or a few volts beyond "black before the said components have been suppressed. In television systems, these recurring pulses usually are the synchronizing pulses. A D. C. reinsertion circuit of this character is described and claimed in application Serial No. 658,894, filed February 28, 1933, in the name of Ralph S. Holmes, entitled Television systems, and assigned to the Radio Corporation of America.

Among the various proposed circuits are the so-calledkeyed circuits of the general type described in Brown, et al., Patent .No. 2,190,753, issued February 20, 1940, and in application Serial No. 276,082, died May 2'7, 1939, in the name of Henry R. Rhea, both assigned to the Radio Corporation of America. In the circuits as described in these two cases, there is in-' cluded a triode which is made conducting periodically (that is, it is "keyed) by means of synchronizing pulses or the like to change the charge on a condenser in the grid circuit of the vacuum tube which is to-have the reinserted D. C. in its output.

It is. an object of my invention to provide im- 4 proved direct current and/or low frequency component reinserting apparatus and, especially, to provide improved reinse'rting apparatus of the keyed" type.

A further object of my invention is to provide an improved D. C. reinserting circuit which is both simple in design and eflective in operation. In practicing the invention, the charge of acondenser in the grid circuit of a vacuum tube is controlled by means of one or more diodeswhich are keyed by periodically recurring pulses, such In one embodiment of the invention, there is a single diode which is automatically biased at a voltage sufficient to prevent the condenser from discharging through the diode between successive pulses. Upon the occurrence of a synchronizing pulse, the diode is rendered conducting and the condenser will discharge therethrough a. certain amount if its charge should be reduced to provide the correct D.C. reinsertion, or it will charge 7 a certain amount through the grid-cathode portion of the tube following the condenser if its charge should be increased.-- 'In this case, the

controlpuises are ofipositive polarity at the grid synchronizing between successive control pulses.

of the said tube, and the reference level is the cathode voltage.

In the above-described embodiment, the au tomatic bias which is applied to the diode may be obtained by a D. C. reinserting circuit whereby the D. C. and low frequency components are inserted at the diode. By thus reinsertlng the D. C. and low frequency components at the keyed diode, any difficulty due to a varying portion 013 the signal feeding from the keying circuit into the picture signal amplifier channel is avoided. If the picture signal channel received picture signal from the keying circuit which did not contain the D. C. and low-frequency components, there obviously would be an error in the output signal, the error depending upon how much signal'fed through the keying circuit.

In another embodiment of the invention, two diodes are provided in order that the desired reinsertion may be obtained when the synchronizing or other control pulses are of negative polarity at the point of D. C. reinsertion, that is, at the grid of the tube following the condenser. In this circuit, also, a diode is biased automatically to prevent the condenser from discharging The reference level in this case is determined by a bias applied to the other diode.

The invention. will be better understood from the following description taken in connection with the accompanying drawings in which' Figure 1 is a circuit diagram ,showing one embodiment of my invention applied to television apparatus,

Figure 2 is a group of curves that are referred to in explaining the operation of the circuit shown in Fig. 1,

Figure 3 is a circuit diagram of another embodiment of my invention, and

Figure 4 illustrates a modification of the circuit shown in Fig. 3. A

Referring to Fig. 1, the invention is shown applied to a television system in which there is produced a composite signal consisting of picture signals, periodically recurring line synchronizing pulses and periodically recurring framing pulses. As previously described, the peaks of the synchronizing pulses have been made to go to a fixed voltage level such as a fixed number oi. volts beyond "blac in the picture before the direct current and low fre quency have been suppressed. This may be accomplished in various ways, as by varying the clipping level of a vacuum tube in the way described in Patent No. 2,192,121, issued February 27, 1940,'in the name of Alda V. Bedford, or by utilizing a suitable mask with a scanning disc so that the signal goes to "black at the end of each scanning line.

In Fig.1, there is shown a conventional alternating current amplifier Ill which has the composite signal applied thereto with such polarity that it appears in the output circuit with the synchronizing pulses of positive polarity as indicated at I l. The pulses H are line synchronizing pulses set on a black level or pedestal indicated at l2. Picture signal is shown at t3.

It is desired to reinsert at the control grid of the next amplifier tube l6 any direct current and/ or low frequency components that have been lost by transmission of the signal through the preceding portion of the signal channel.

Stated diiferently, it is desired to have the tube l6 supply an output signal in which the peaks of the synchronizing pulses all go to a fixed voltage level. If this is done, correction is made for signal components that have been lost (i. e., they i are reinserted) and correction is also made for undesirable effects which may have been introduced, such, for example, as 60-cycle hum.

The amplifier tubes In and I6, which may be conventional tetrodes or pentodes, are coupled by a coupling condenser C which, as will be explained hereinafter, has a suitable capacity value to permit, first, the necessary initial charging and, later, the necessary additional charging or discharging in cooperation with the complete cir= cuit for the desired reinsertion of lost components.

In place of the usualgrid leak resistor for the tube 16, there is a diode 2| having a cathode 22 connected through a resistor 23 to the control grid I4, and having a plate 24 connected to ground through a resistor 28.

In operation, the positive pulses H produce a fiow of grid current in the tube l6 which charges the condenser C. It will be seen that, since the cathode of the diode 2| is connected to the negative side of the condenser C, it is connected in the proper direction to discharge condenser C and will do so unless a suitable opposing bias is provided. Such a bias is maintained across the resistor 26 by the action of the positive synchronizing pulses or keying pulses which are applied to the plate of diode 2!, these pulses being taken from the output of the tube IS in the specific example illustrated andxapplied to the diode plate through an amplifier 3| and a coupling condenser Cl.

It may be noted that the keying signal applied to the diode to be keyed should have a peak to peak voltage at least as great as the peak to'peak voltage of the signal applied to the amplifier tube I6, and, preferably, should be substantially greater. 'I'his is true, also, of the circuit of Fig. 3 which is described in detail hereinafter.

The positive pulses fed through condenser Cl cause a fiow of current through the diode 2|, the resistor 23, and the grid-cathode impedance of amplifier tube l6. Between successive pulses, the condenser Cl discharges a slight amount through a path that may be traced through the plate resistor 30 of amplifier 3| to ground, through ground and the resistor 26 back to condenser Cl. This current flow through resistor 26 holds the diode plate 24 negative with respect to ground during the time of the picture signal l3, 1. e., between succeeding pulses ll, whereby the diode is non-conducting and the charge on condenser C remains unchanged during the transmission of the picture signal itself.

The above-described charging of condenser Cl through the diode and partial discharging through resistor 26 functions to reinsert the D. C. and low frequency components at the diode plate 24 whereby any signal that is fed back through the keying circuit and into the picture signal channel does not vary with picture D. C. content and thus partially spoil the effectiveness of the D. C. insertion in the said. signal channel.

The charge on condenser C determines the negative bias on the grid ll of amplifier Hi, this a being another way of saying that it determines the amount of D. C. insertion. Variations in this negative bias result in the desired reinsertion of low frequency components and in the removal of hum efiects or the like.

During the intervals of transmission of picture signal l3, the bias voltage on the control grid H is the voltage across condenser C minus the D. C. voltage on the plate of tube l0. During these intervals, the circuit 23, 2|, 26, which corresponds to the usual grid leak, is open-circuited as previously stated. Therefore, the bias on grid It cannot be determined by figuring 1R drop through a grid leak resistor as is commonly done in the case of a grid-leak biased tube.

The operation of the circuit will-now be considered more in detail. First, it should be noted that, just as in the Holmes invention previously referred to, the basis oi the circuit operation is the fact that the height of a synchronizing pulse as measured from the A. C. axis of the composite signal is a measure of the amount of D. C. or low frequency component to be reinserted. Therefore, it is desired to have the present circuit so operate that the charge on condenser C increases if the height of the line synchronizing pulses increases, and that the charge on condenser C decreases if the height of the line synchronizing pulses decreases.

- Next, let us assume that the apparatus is in operation and let us consider what happens upon the occurrence of a synchronizing pulse.

This pulse is applied with positive polarity both to the grid ll of tube l6 and to the diode plate 24.

Also, condenser Cl is charged through the diode 2| and through the grid-cathode impedance of tube l8. At the end of this pulse, condenser Cl is partially discharging through resistor 26-to hold the diode plate negative, whereby condenser 0 cannot discharge during the picture signal interval.

Meanwhile, thecharge on condenser C holds the grid M of. tube' i6'p'roperly biased and no undesirable saw-tooth components are introduced as may be the case with circuits in which the condenser C may discharge a. slight amount during the picture signal interval.

It is essential that -the circuit operate to decrease the charge of condenser. C in the event that the recurring pulses decrease inamplitude with respect to the signal A. C. axis. The operation is as follows: Upon the occurrence of a synchronizing pulse or less amplitude'than a preceding pulse, the condenser C will not be charged any further because its voltage is greater than the applied voltage. On the other hand,

condenser C will discharge enough to bring the condenser voltage to a value of the applied voltage.

This discharge path for condenser C may be traced through the plate resistor I of amplifier Hi to ground, through the plate resistor 30 of amplifier ll, the condenser Cl and the diode 2i and resistor 23 to the other side of condenser C. All resistors in this discharge path have such low resistance that the condenser C may discharge sufiiciently in a given time to permit the condenser voltage (and, therefore, the voltage on grid II) to follow the variations in height of the synchronizing pulses.

That the condenser C is discharged to the proper voltage level will be seen considering the discharge path of the condenser .C as follows: At the beginning of the. synchronizing pulse of reduced amplitude, the signal voltage across the plate resistor 8 is of less amplitude than the voltage across condenser-C, the condenser voltage having been determined by the. preceding synchronizing pulse. It is desired to discharge the condenser C to a point where its voltage is equal to the signal voltage now appearing across output resistor 8.v Obviously, this may be accomplished by connecting the grid side of the condenser C.

to the ground end of the plate resistor 8. In effect, this is done when the pulse through condenser Ci makes the diode 2| conducting, since the impedance of diode 2| and the impedances of resistors 23 and 38 and condenser Cl are comparatively low. The condenser Ci has a fairly large capacity, usually at least: ten times the capacity of the condenser C Y The capacity of condenser C! is so large that any discharge current from the condenser C can change the voltage across condenser Cl only a negligible amount. As to the D. C. voltage be-- tween the diode plate and the ground end of plate-resistor 8, the D. C. voltage from the cathode of amplifier II to the plate of the diode is zero, since the D. C. charge on condenser Cl is equal and opposite to the voltage applied from the plate battery of the amplifier 3|.

The operation or the D. C. inserter may be described from a different viewpoint as follows:

First, it should be noted that the positive pulse on the plate of diode H which is causing current fiow through diode 2| and through the gridcathode impedance of tube It, is holding the plate of diode 2l close to ground potential, the diode and grid-cathode impedances being low and the resistance of resistor 23 being low. More particularly, the grid side of condenser C has been driven substantially to ground potential by the pulse through condenser Cl. Since the plate side of condenser C is held ata potential determined by the signal, 1. e., determined by the height of the synchronizing pulse from the A. C. axis, it follows that the voltage across .0 is brought to the proper value either by charging or by discharging the condenser. C slightly.

The curves in Fig. 2 illustrate, in general, the operation of the circuit. The characteristic curve for the amplifier tube It is indicated at 34, while.

the composite signal which is applied to the control grid of this tube is indicated at, l2,-il.

Itwill be seen that the synchronizing pulses il produce the charging current for the condenser C by driving the grid of tube ll into the positive region whereby the resulting voltage across condenser C produces the correct biasing volt age to hold the A. C. axis of the composite signal the proper distance away from zero grid volts, as indicated. It will be noted that between successive pulses ii, the picture signal it is am- 'pliiled by the amplifier l0 operating as a linear amplifier. No attempt has been made to illustrate in Fig.2 the portion of the cycle of operation involving discharging of the condenser C.

From the foregoing, it will be evident that the charge on the condenser C, and therefore the bias on the control grid i4, is set at the proper value, since the positive pulse fed through condenser Cl drives the grid, side of condenser C substantially to ground whereby condenser C may either acquire some additional charge through the grid-cathode impedance of tube It or lose some of its charge through the diode 2!.

When the circuit connections are as shown in Fig. 1, the resistor 23 generally is included in the circuit to prevent oscillations resulting from the feedback connection through the. keying circuit. When the keying pulses are supplied from a feedforward connection, as when the switch I! is closed and the switch 20 opened, the resistor 23 is preferably removed from the circuit, in effect, by closing a short-circuiting switch 25.

The resistor 23 may be omitted also in the case where the keying pulses are supplied over a line 35 from a separate source which supplies keying pulses occurring in synchronlsmwith the synchronizing pulses. Keying pulses may be so supplied by connecting the switch arm 20 to the line 35, the switch It remaining open.

Instead of .feeding both the picture signal and thesynchronizing pulses through the keying circuit of Fig. 1, it may be preferable to include clipping tubes in the keying circuit whereby synchronizing pulses only of positive polarity are applied to the diode plate 24. In this case, there is a negative bias voltage applied to the diode plate 24, but the voltage does not represent the picture D. C. component. Since the clipped synchronizing pulses are of constant amplitude, they do not introduce a variation in the signal at the input of amplifier it when added thereto and the lack of D. C. component on the diode plate is unimportant.

It has been found that the circuit has less tendency to oscillate due to the feedback of keying signal if the above-mentioned clipping tubes are employed to prevent picture signal from appearing on the diode plate 24.

Referring to Fig. 3, there is shown an embodiment of the invention in which synchronizing since the pulses are of the wrong polarity to cause grid current flow in the tube It. Another difference is that, in Fig. 3, the voltage level is not ground potential but, instead, is a certain bias or level setting voltage which is obtained as a result of current flow through the diodes.

A discharge path for the condenser C is provided by a diode 31 which has its plate connected to the cathodeoi the diode 36, both or these electrodes being connected through a resistor 33 to the grid side oi condenser C. The resistor 33 is not essential to the circuit operation and, for many applications of the invention, it may be omitted, or it may be short-circuited by means of a switch 33. When included in a circuit, the resistor 38 functions primarily to prevent circuit oscillations which otherwise might occur because of the feedback connection through the amplifier tube 48.

The cathode of the diode 31 and the plate of the diode 36 are connected together through a resistor 4|. The plate of diode 36 is connected to ground through a condenser 42. The resistor 4| is provided. with a variable tap through which any desired point on the resistor 4| may be connected to ground through a resistor 43.

Keying pulses are applied from the output of amplifier l6 through clipping circuits 46 and." and through a coupling condenser 48 to the junction point of the resistor 4| and the cathode of diode 31.

These keying pulses are for the purpose of making the diode 31 conducting during the occurrence of synchronizing pulses whereby the condenser.C may discharge some it its voltage should be reduced, and also for the purpose of producing .a fixed voltage which may be em ployed for the voltage level setting.

The voltage pulses which are applied to the cathode of diode 31 are referred to as keying pulses and are of negative polarity and of fixed amplitude, these pulses being indicated at 5|. The pulses 5| of fixed amplitude are obtained by removing the picture signal in the clipping circuits 46 and '41 whereby the amplitude of the resulting pulses is fixed by the vacuum tube characteristics and by the associated circuit elements.

The clipping tube 46 may be of the well-known type operating on grid leak biasing, as described in Patent 2,132,655, issued October 11, 1938, to J. P. Smith. The next clipping circuit 41 is also of the grid leak biased type and may be omitted if sufllcient clipping'is obtained in the first clip ping circuit. However, the second clipping circut usually is desirable and, in the example illustrated, it is a circuit of the cathode-follower type which does not reverse the signal polarity and which, therefore, has negative polarity pulses appearing across the cathode resistor 52.

Considering the operation of the circuit of Fig. 3 more in detail, upon the occurrence of a negative keying pulse 5|, the diodes 31 and 33 are made conducting whereby'the grid side of condenser C is connected through the resistor 38 and the diode 36 to the plate end of resistor 4|, this point'on resistor 4| being at a predetermined level setting voltage, such as minus four volts. Therefore, if the synchronizing pulse applied through amplifier I is of greater amplitude with respect to the signal A. C. axis than was the preceding synchronizing pulse, the condenser re ceives a slight additional charge through the diode 36.

This charge path may be traced from the lower end of the plate resistor 8 of the amplifier l0,

through ground, and through the parallel cir cuit comprising condenser 42 shunted across resetting value.

In the event that the next synchronizing pulse supplied from amplifier I0 is of less amplitude than the preceding pulse the condenser C will discharge slightly, the discharge path being from 10 the grid side of the condenser C through the resistor 33, the diode 31 and through the comparatively low impedance elements 48 and 52 to ground, and through ground and the plate resistor 3 back to the plate side of condenser C.

It will be apparent that the condenser C cannot discharge below the level setting value because if it should start to do so it would instantly begin to receive a charge through the diode 36 as previously described. This is also apparent from the fact that, when the diodes 33 and 31 are conducting, the cathode of the diode 31 is at substantially the same potential as the plate of the diode 36, that is, it is at the level setting voltage.

-Between successive synchronizing pulses, that 5 is, during the picture signal interval, the diode 31 is being held in non-conducting condition by a D. C. voltage across resistor 4|, this voltage holding the cathode of diode31 positive with respect to ground. 'l he-signa1 on grid |4 holds the 'sbhathodebfdiode as positive with respect to its plate 4, just as in Fig. 1, it prevents grid current in tube l6 by holding its grid negative. As previously explained, this D. C. voltage across resistor 4| appears as a result of the condenser 42 35 and 48 (which previously were charged through diodes 36 and 31) discharging through resistor 4|, the discharge path being from the condenser 48 through resistor 4| and the condenser 42 to ground, and through ground and the cathode resistor 52 and back to the condenser 48. Thus,

it will be seen that, just as in thecircuit of Fig. l, the condenser C cannot discharge during the interval between successive synchronizing pulses.

In Figs. land 3 the values of certain circuit 45 elements have been indicated in ohms, megohms,

microfarads and micro-microfarads. These values are given merely for the purpose of illustration and may be varied within wide limits.

In Fig. 3, just as in Fig. l, the keying pulses may be obtained from a point other than one following the D. C. insertion. For example, the keying pulses may be fed over a line 60 and through the switch 6| to the diodes 36 and 31, the pulses being obtained from any suitable source of pulses which are synchronized with the line synchronizing pulses.

It may also be mentioned that the circuit of Fig. 3 will operate for the case where positive pulses are applied to the control grid of tube I6 providing the connections of diodes 36 and 31 are reversed, and also providing the keying pulses are fed through the condenser 43 with positive polarity instead of with negative polarity.

As illustrated in Fig. 4, the level setting volt age for the circuit of Fig. 3 may be obtained by means of a biasing battery indicated at 65 instead of by means of the resistor 43 and condenser 42 which, in Fig. 3, utilized the flow of diode current for biasing. The circuit of Fig. 3

7 is the preferred circuit since it is usually desirable to avoid the use of batteries which eventual ly must be replaced by new ones.

The level setting illustrated in Fig. 3 is claimed broadly in my copending application Serial No.

7 367,323, filed Nov. 27, 1940, entitled Direct cur- 2,200,944 a r g 5 rentreinsertingcircuit,andassignedto the Ratween said picture signal intervals and nondio Corporation oijmerica.

From the roregoing, it willbe apparent that various other-modifications y be made in my invention without departing from the spirit and scope thereof.

I claim as my invention:

1.'In combination, a source of composite slg nal consisting of the principal intelligence conveying signals and of control pulses each recurring at intervals, said pulses having an amplitude at least as great as said principal signals like polarity, the combination of an amplifier tube having a control grid and a cathode, a condenser connected in series with said source and the control grid of said tube whereby said composite signal may charge said condenser, and a dis-' charging circuit for said condenser which com prises a diode so connected that said condenser may discharge therethrough when the diode is conducting, means torso biasingsaid diode that said condenser cannot discharge therethrough during the intervals of said principal signal, and

keying means for making said diode conducting between said last intervals.

2. The invention according to claim 1 wherein said control pulses are applied with positive polarity to said control grid whereby said condenser is charged through said grid-cathode irnpedance.

3. The invention according to claim 1 wherein aseconddiodeisconnectedinserleswithsaid' condenser and the source of said composite signal, and connected in-the direction to supplya charging current to said condenser when control 35 pulses or negative polarity are applied to said control grid.

4. The invention according to claim liwherein' saidrkeying means includes means for applying tosaiddiodeasignalconsistingoivoltagepulses of substantially constant amplitude and containing no picture signal.

5. In a circuit for direct current reinsertion and low frequency correction, terminals across which there appears a composite signal consisting or picture signals and control pulses each recurring at regular intervals, said pulses having an amplitude at least as greatas picture signals oi like polarity, the combination or an amplifier tube having a control grid and a cathode. a condenser connected in series with the grid-cathode impedance'oi saidtuhe. means for impressing saidcomnositesignalacrosssaidserie'seircmt' and for charging said condenser, and a dischargingcircuitwhichcunprisesadicdesocon throughwhenthediodeisecnducting ndhsyingmeansiormakingsaiddiodecmductingbe- 5 composite signal.

conductingduring said picture signal intervals, said keying means also comprising means for reinserting the direct current component or said 6. In a circuit for direct current and low irequency correction 01 a composite signal consisting oi picture signals and control pulses each recurring at regular intervals said pulses having 7 an amplitude at least as great as picture signals of like polarity, the combination oi an amplifier tube having a control grid anda cathode, a coupling condenser connected to said control grid, means for feeding said composite signal through said coupling condenser to said control grid, and a charging and discharging circuit for said condenser which comprises a pair of diodes so ccnnected that said condenser may charge through one or said 'diodes and discharge through the other or said diodes when said diodes are made conducting, means for so biasing said-diodes that said condenser cannot discharge therethrough during picture signal intervals, and keying means for making said diodes conducting between said picture signal intervals.

'1. In a circuit for correction oi'.a composite signal consisting of picture signals and control pulses each recurring at regular intervals said pulses having an amplitude at least as great as picture signals of like polarity, the combination 01 an amplifier tube having a control grid and acathode, a coupling condenser connected to said control grid, means fori'eeding said composite signal through said coupling condenser to said control grid, and a charging and discharging circuit ior said condenser which comprises-a pair of diodes each having aplate and acathode, the

plate or one diode andthecathcde oitheother diodebeingconnectedtosaidcontrolgrimthe 40 other plate and cathode of said diodes being connected to each other through a resistor, means rorholding one end of said resistor substantially at ground potential for alternating currents, means for applying keying pulses to theotherendoisaidresisterrormaklngsaiddiodes conducting'between said picture signal inmeans for maintainingsaid one end tervalaand oitheresistorataflxedD.C.potentialwithrespe'ciitogroundandiorholdingthe'iiiodes-nonconductingduringthepicturesignalintervals.

8,. 'lheinvention accordingtoclaim'lwherein saidlastmeans comprisuacondenser connected betweensaidoneendoitheresistorandground and further comprising a second resistor connectedbetweengroundandapointontheiirst xsnnawaun'r.

Images