US2227906A - Envelope current device - Google Patents

Description

Jan. 7, 1941. E w, KELLQGG 2,227,906

u ENVELOPE CURRENTDEVICE Filed oct. 29, 19:58 v s'sneets-sheet 1 nverttor Jan. 7, 1941. E w KLLOGG y 2,227,906

ENVELOPE CURRENT DEVICE Filed oct. 29', i938 s sheets-sheet 2 Enventor Lamm/Zaza Bg /v/ Jan. 7, 1941. E. mq-11.0@ 2,227,906

ENVELOPE CURRENT DEVICE Filed Oct. 29., 1938 3 SheetS-Sheel 3 nvenfor v f Gttorne'g l Patented Jan. 7, 1941 PATENT orf-Flor..A

ENVELOPE CURRENT DEVICE Edward W. Kellogg, MoorestowmN. J., assignor to Radio Corporation of America, a corporation of Delaware Application october 29, 193s, serial No. 237,603

13 Claims.

This invention relates to devices for producing a current having a wave form similar to the envelope of the peaks of an audio` frequency current, or the like, which contains components of many different frequencies and varies rapidly in amplitude.

The principal object of the invention is to provide an improved apparatus and method of operation whereby the envelope current potential is made to follow the alternating voltage peak values more closely than has been heretofore possible.

The invention will be better understood from thefollowing description considered in connection with the accompanying drawings, and its scope is indicated by the appended claims'. Referring to the drawings,

Figure l is a wiring diagram of a prior art device on which the improvement of the present invention is based,

Figures 2 and 3 are explanatory curves relating to the operation of the apparatus of Fig. 1,

Figure 4 ls a wiring diagram of one form of the invention,

Figures 5 and 7 are explanatory diagrams relating to the operation of the apparatus of Fig. 4.l

Figures 6 and 8 are wiring diagrams of modifications wherein discharge of an envelope voltage delivery capacitor is delayed by interposing in its discharge path a separately charged auxiliary capacitor,

Figures 9, 10, 11, 12 and 13 are wiring diagrams of various modifications wherein an artificial transmission line or delay network is arranged to apply successive charges to an envelope voltage delivery capacitor delaying the discharge of this capacitor and Figures 14 and 15 illustrate further modifications wherein means including a grid controlled electron discharge device is provided in the capacitor discharge path for delaying discharge of the capacitor.

Fig. 1 shows, in elementary form, a circuit arrangement which is in wide use for providing a current proportional tothe envelope of an audio frequency current. After suitable amplification, the audio frequency current is applied to terminals I--I of transformer 2. A rectifier 3 permits current to pass in one direction only and ccndenser 4 becomes charged by the rectified current. Since the resistance of the rectifier'3 to reversed current or to the discharge of condenser 41s very high, condenser 4 would remain for a long peri fd of time charged to a voltagecorre- (Cl. Mii-100.3)

sponding to the highest voltage impressed across the rectifier. In order that the charge on condenser 4 may be. reduced when the audio frequency amplitude becomes less, a discharge rey sistance 5 must be provided. The magnitude of 5 this resistance is so chosen relative to the capacity of condenser 4 that the decrease in voltage E1 across condenser 4, when the audio frequency voltage falls, willbe at the desired rate. Obviously if condenser 4 is permitted to discharge 10 very rapidly through resistance 5, the voltage E1 which should represent an envelope, will attempt to follow individual waves of the audio frequency and this is not desirable. On the other hand, if resistance 5 is made very high, E1`will follow very slowly when the audio frequency amplitude drops and this may in part defeat the purpose of the device for which the envelope wave is wanted. A desirable value of resistance 5 is therefore chosen which will cause E1 to follow 20 decreases at a suitable rate. It is, in general, desirable to make the resistance of the rectifier 3 and of the input circuit as rcected through the transformer 2, so low in relation to the capacity of condenser 4 that an audio frequency peak of 25 very short duration will suffice to build up the voltage Ei to a value substantially equal to this peak voltage. The impedance of the rectier andinput circuit must also below compared with resistance 5 so that the latter will not constitute an appreciable drain on the rectifier during charging.

The circuit so far described willprovide a voltage which comes as near to following the true peak envelope as it is practically possible to provide. Such an envelope voltage, however, is found .to contain in greater or lessmagnitude, numerous fluctuations which are in the audio frequency range, and it is practically almost always 40 necessary to subject the envelope-voltage or current to low pass filtering in order that it shall not introduce sounds in the system. Thisgives a voltage En which rises and falls withEi but does not have the small rapid fluctuations which 45 are present in Ei. The simplest type of low pass lter employed for this purpose consists' in a resistance 6 and condenser 1.

The RC product of this stage of filtering must be sumcient to practically eliminate the audio 50 components. In general, condenser 1 is of much smaller capacity than condenser 4 (for example, one-tenth) and resistance 6 is correspondingly high. With this arrangement, the current through resistance 6 is too small to greatly af- 55 fect the voltage E1. 'I'he filtered voltage En may be applied to the grid of a thermionic amplifier tube 8, from the plate circuit of which is derived the necessary power for control or for ground noise reduction purposes. InA some arrangements, a low pass filter involving inductance and capacity is employed instead of resistance and capacity. This gives a sharper cut-oil.' characteristic which may be desirable under certain conditions.

An arrangement similar to that shown in Fig. 1 for obtaining an envelope voltage is already in wide use and is satisfactory, providing a sufficiently slow discharge of condenser 4 is not objectionable. When attempt is made, by reducing the value of resistance 5 in Fig. 1, to produce an envelope voltage or current, which follows the decreases in the audio frequency voltage more rapidly, it is found that the system no longer gives an envelope voltage equal Ato that of peak voltages of short duration in the audio frequency system, but tends to give a voltage more nearly corresponding to the average or to the r. m. s. audio frequency voltage. This is true regardless of the fact that resistances 5` and 6 do not appreciably load the rectifier 3 during the instant of charging.

Fig. 2 illustrates what happens when there is a brief audio frequency peak followed by an interval of inverse or low audio voltage, comparable in length to the time constant of resistance 6 and condenser 1. Curve 9 represents an assumed audio frequency voltage wave. Curve E1 represents the voltage E1 across condenser 4 of Fig. 1, assuming that the input impedance is very low, permitting development of the substantially full peak voltage across condenser 4. Curve Earepresents the filtered voltage Ez.- It is obvious that this will never reach the peak value of E1 because condenser-4 will have discharged to a lower voltage before condenser 1 has had time to become charged. If some means may be provided which will maintain condenser 4 at its maximum voltage for a period of time comparable with the time constant of resistance 6 and condenser 1 and then permit condenser 4 to discharge, we should have the conditions in.

dicated in Fig. 3, where it is seen that the voltage E2 rises to a value much more closely approaching that of the peaks of' the audio waves.

It is the purpose of my invention to provide means for maintaining the peak voltage on condenser 4 for a longer period of time than the actual duration of the maximum voltage in the audio frequency wave. 'I'his may be accomplished either by delaying the discharge of condenser 4 or by supplying additional charging current which reaches condenser 4 subsequent to the initial charging wave. It is well recognized that the employment of a full wave rectifier will in some cases assist in maintaining the voltage on such a condenser, but only a small and doubtful advantage is obtained in this manner. In many cases, it is desired to obtain an envelope of the peaks of one polarity only. and not of the other. For this purpose, the full wave rectifier is obviously not appropriate. In the second place, in spite of all of the filtering applied, there is a residuum of components of audible frequency. If this residuum is of the same frequency as the audio waves, it is not objectionable in quality. If it is oi' double frequency, such as is produced by a full wave rectifier, it may be much more objectionable in sound, even though smaller in magnitilde.

One method of providing for a retarded discharge of condenser 4 would be to substitute an inductance or an inductance in combination with a resistance for the simple resistance 5 of Fig. l. This arrangement is shown in Fig. 4. Fig. 5 shows the effect of such an inductance on the discharge curve for a condenser. the voltage as a function of time when a condenser is discharging through a resistance. Curve I6 shows the discharge curve through an inductance. It is well known that such an inductance will have the effect, not simply of making the discharge begin more slowly, but it also helps to maintain the discharge at a high rate after the condenser voltage has fallen, or even against -a reversed voltage. In order that this second effect, which is not desired, may be avoided, I have shown in Fig. 4 a short-circuiting rectifier element l1 which comes into action whenever the inductance tends to assist the discharge of the condenser, and prevents this from happening. Thus the combination can delay discharge, but never accelerate it or cause it to go too far. The effect of the rectifier is indicated by curve II in Fig.` 5, the first part of which coincides with curve I6, while the latter part approaches the curve that would be obtained with resistance alone.

In the low power and relatively high impedance circuits which are usual in sound recording for pictures, it is not always feasible to obtain a sufllciently high inductance for the purpose just described. A similar eect may be obtained in another manner, as illustrated in Fig. 6. In this arrangement, two rectifiers are provided, each of which charges a condenser at substantially the same rate that condenser 4 was charged in Fig. 1. The additional rectifier'is shown at I9 and the additional condenser at 26. Condenser 4 -in Fig. 6 corresponds in function to the similarly numbered condenser of Fig. 4', but it will be noted now that condenser 4 discharges through resistance 6 into condenser 20, which itself must discharge through resistance 2 I. The time constant or RC value of 26 and 2| is of the same order of magnitude as that of filter stage 6 and Condenser 4 cannot begin to discharge rapidly until the voltage across condenser 26 has dropped con siderably. The effect on the discharge of condenser 4 islllustrated in Fig. 7, in which curve 22 represents the discharge as it would be in Fig. 1. The mere substitution of a higher resistance for 5 in Fig. l wouldlcause the discharge to take place as shown in curve 23, but the fall is still too rapid at first. The effect of the circuit shown in Fig. 6 would be to cause the discharge to take place as indicated in curve 24. A discharge characteristic of this type. for condenser 4, would give condenser 1 time to become charged up to practically peak value.

Some further delay in the dischargeo! condenser 4 may be obtained by employing a full Wave rectifier for charging condenser 2l instead of the single wave rectifier as indicated at Il in Fig. 6. This arrangement is shown in Fig. 8. The rectifier I functions in the manner described above. The full wave rectifier is indicated by the transformer 22 and the rectifying elements 2l and 24. The objection to the full wave rectifier that it tends to introduce double frequency components into the envelope wave, which are themselves objectionable because not completely filtered out, does not apply to the rectifier used to Curve II shows charge condenser 20 in Fig. 8, for the reason that the voltage across 20 goes through an extra stage of filtering before it reaches the condenser 1, whose voltage represents E2, the final ltered envelope wave. Some help may therefore be obtained from the full wave rectifier. If the'negative half wave which follows the positive peak is `of equal magnitude, it will help to maintain the chargefor a slightly longer period in condenser '20. I have also shown in the discharge circuit for condenser 20 in Fig. 8, an inductance 25 in series with the discharge resistance 26 and a short-circuiting rectifier element 21 as previously described. The employment of this discharge circuit is not an essential feature of the full Wave rectifier system of Fig. 8,-but is merely an expedient which may be added-to further slow down the discharge.

In the arrangements already described, I have special discharge circuit features, it being understood that these may be employed.

Fig. 9 shows an arrangement for repeating the audio frequency wave, which in my discussion is ,supposed to have provided only a single peak. Across a high impedance audio circuit which I have represented by the output of a pentode 28 is shunted a short length of audio frequency artificial line 29. 'I'he voltage produced between the conductors 30 and 3| is applied to the input or grid of an amplifier tube 32 which drives the rectifier 3 as in the previous diagrams. The input of amplifier 32 is of high impedance compared with the characteristic impedance of the line 29, so that little energy is absorbed. When a peak voltage is developed in the output of the pentode 28,' a wave is transmitted along line 29 and reflected from the end, surging back and forth through lthis line a number of times and with each return the peak voltage is again impressed on the input of amplifier 32. The line thus provides an effect akin to reverberation. Each time the reflected voltage returns it will be of smaller amplitude, but it will assist in maintaining the charge on condenser 4. The losses in the line are proportionedto give the most desirable duration of the series of refiected voltages..

An improvement over Fig. 9 is the circuit arrangement shown in Fig. 10. In this arrangement, a much longer line 38 is required, since the waves traverse it only once.' Reflections at the far end of the lineare prevented by a resistance 34 having a value equal to in which L and C are the inductance and capacity per unit of line length. A series of rectifier elements 35, spaced at uniform intervals along the line, cause conductor 36 to become charged to the maximum voltage developed at I any point in the line. Conductor 36 transmits its voltage to condenser 4, with the result that the voltage of a single wave is repeatedly applied to condenser 4. A subsequent wave of the same or opposite polarity does not interfere with the action, owing to the rectifylng properties of the elements 35.

The intervals between the positions of rectifier elements 35 should be such that the voltage ripple due to the successive charges applied to condenser 4, will be readily filtered out by the low pass filter 6, 1. The total length of the line is such that the time of wave propagation from one end to the other is equal to the period during which it is desired to sustain the voltage across condenser 4. It is obvious that if design considerations make is desirable, amplification may be introduced between conductor 36 and rectifier element 3, in order that it may be possible to operate without supplying as much power to the line as would be necessary without such amplification. The rectifiers l35 must be of sufficiently high impedance in comparison with the characteristic line impedances so that they will not cause serious attenuation of the waves as they travel along the line. The arrangement of Fig. 10 has the advantage that a given wave can continue tosupply charge to condenser 4 for a definite period of time, after which its effect ceases altogether, whereas with Fig. 9 there is a gradually decreasing effect.

Since the avoidance of too much drain on the line, or in other words bleeding the same wave too many'times, may present some design difiiculties, it may be desirable to employ another arrangement, shown in Fig. l1. In Fig. 11, rectifier element 3 supplies the initial charge to condenser 4 in exactly the same manner as in Fig. 1. The subsequent impulses are provided through separate lines 31, 38, 39 and 40 of successively increasing length. Each line is damped at the far end with a resistance which prevents reflections. Line 40 would be of the same length (in terms of propagation time) as line 33 in Fig. l0; With the arrangement shown in Fig. ll, each of the five successive waves applied through rectifiers to condenser 4 has the same amplitude.

The artificial audio frequency lines called for in Figs. 10 and 11 must be capable of transmitting waves of fairly high frequency,kand this requires that these lines be built with a very large number of small units of capacity and inductance, thereby rendering them expensive. Fig 12 shows an arrangement which avoids this objection. If the peak voltage appearing in the audio frequency circuit is of very 'short duration, and therefore represented by high frequency components, it Would be attenuated on a transmission line unless this were made up of a large number of elements. If, however, this brief peak voltage can be caused to produce a more sustained voltage, it may be transmitted over a line having larger units of induction and capacity. The expedient of causing a short peak voltage to operate through a rectifier and charge a condenser is an expedient for making a pulse of short duration produce one of greater effective duration. It is, of course, assumed. in this as in previous arrangements Vdescribed, that the rectifier is of such low impedance that it can charge up its condenser in a very brief interval. The arrangement of Fig. 12 is such that condenser 4 will receive a succession of charges at definite intervals. The initial charge is delivered through rectifier 3 as in Fig. 1. A second and larger condenser 42 is simultaneously charged through rectifier 4|. Condenser 42 discharges through inductance:43 and resistance 44, the elements being given such values that the voltage is sustained vlong enough to give the equivalent of a moderately low frequency wave, which will be in the transmission band of the artificial line 45. This line-is damped by a Fig. 13 shows an arrangement similar in principle to that of Fig. 12, but employing a full wave rectifier 50 to build up a voltage for supplying the repeated charges. Condenser 5i, which is larger than the condenser units of the transmission line 52,- receives the charge as first delivered by the rectifier 50. 'Ihis voltage is propagated along the line, delivering a succession of charges to condenser 20 through rectiers 53, instead of delivering the extra charges direct to condenser 4 (as in Fig. 12) which corresponds to condenser 4 in Fig. l. The condenser .20, into which condenser 4 must discharge through resistance 5, is maintained charged by the successive action of rectifiers 53. In preventing condenser 4 from discharging for a brief interval by maintaining a voltage across condenser 20, the purpose of the invention is accomplished just as well as though the additional charges had been delivered direct to condenser 4, While an additional filtering advantage is obtained. The discharge circuit 43-44 for condenser 5I and also for condenser '-42 in Fig. l2, is not an essential feature, but results in a slightly better wave (with more rounded top) than would result if condenser 5I or 42 were drained entirely by the transmission lines which have a more nearly pure resistance characteristic. With the extra drain circuit, the condenser must be larger than with the line alone acting as discharge circuit. y

Fig. 14: shows another method of retarding the discharge of condenser 4 after this has received its initial charge through rectifier 3. Instead of a simple resistance, such as 5 in Fig: l, the disp charge path includes a thermionic triode tube 55 in series with which the resistance 56 may be employed if desired. In the absence of any audio frequency voltage, the bias on the grid 51 of tube 55 is such that the tube is of low resistance as compared with fixed resistor 56, so that the latter largely determines the rate of discharge of condenser 4. As soon as condenser 4 receives any charge through rectifier 3, the voltage E1 is impressed on the grid 58 of an amplifier tube 59 which is employed as a D. C. amplifier and polarity reverser in a well known manner. A plate voltage bucking battery 50 is required for this purpose. A voltage is impressed at terminal 6I which is proportional to but opposite in sign to voltage E1 across condenser 4. A fraction of this voltage appears at the grid 51 of the discharge tube 55 and serves to bias tube 55 to cut-off so that condenser 4 cannot discharge therethrough. After a short interval deter-mined by the capacity 1 and resistance 6, the voltage E; will reach a value substantially equal to E1. lThis voltage, which is positive, is impressed on terminal 62`and a portion of this voltage reaches the grid 51, neutralizing the negative bias which was applied through terminal Si. Tube 55 then becomes conducting and permits discharge. Tube 55 is preferably designed to pass quickly from conducting to non-conducting condition (or, in other words, is a sharp cut-ofi tube), and the network 63 and bias voltage 64 are adjusted so that discharge can take place when Eris substantially equal to or greater than E1, but not so long as En is less than E1.

Fig. l5 shows a modication giving results substantially equivalent to those obtained wth Fig. 4, 5 while avoiding the use of a D. C. amplifier and bucking battery. The negative voltage to be applied to terminal 5I of network 53, in order to bias the tube 55 to cut-ofi, should be proportional to and opposite in sign to the voltage E1 10 across condenser 4. An approximation to this voltage maybe obtained by means of a second rectifier represented -by the transformer 66 and the rectifier element 61 which works on the same peaks as rectifier ibut is oppositely poled so that 15 it gives a reversed voltage. The condenser 68 is charged to this reversed voltage and a discharge circuit 69 may be provided, or this may be omitted, since the network 53 provides a discharge path. The time constant or RC product of condenser 20 58 and its discharge path is designed to maintain the negative voltage across 5I long enough for condenser 1 to become charged. In other words, the RC product is somewhat greater than that of condenser 1 and resistance l6, in a ratio of the 25 order of 2:1. This relatively slower discharge of condenser 58 will not prevent a reasonably rapid dischargejof condenser 4, once the voltage Ez has reached its full magnitude. The resistance of network 63 must be high enough so that such 30 currents as may flow therethrough will not appreciably affect the voltage En.

The delayed discharge of condenser 4 in all of the arrangements I have shown will Anot only serve to insure a system providing a voltage more nearly proportional to peak audio frequency voltage (rather than to an average), but will also very materially improve the filtering and perhaps make it unnecessary to employ as large an RC value for the filter stage 8, 1, as would be necessary without this delayed discharge feature. Referring to Fig. 2, in which curve l0 shows the value of E1 resulting from an audio frequency wave 9 having sharp peaks spaced at relatively large intervals, it will be evident that the quick 45 drop in the voltage E1 following the audio peak, means that there will be a much larger fluctuation or ripple in E1 than would be the vcase with a delayed discharge such as shown in curve E1 of Fig. 3. It is this ripple which must be substan- :ial'ly filtered out by the low pass filter stage In my figures I have represented rectier units in a conventional way, and it is to be understood that the symbols employed may stand for any of the well known types of rectifier, whether of the hot cathode type of non-linear resistance such as copper oxide rectiflers. It is also to be understood that the relative positions of rectifier and transformer winding may be interchanged if desired for design purposes, in the manner well known to engineers familiar with vacuum tube applications.

I claim as my invention:

1. In a system for deriving an envelope voltage 35 proportional to the peak values of an alternating voltage, the combination of means for rectifying said alternating voltage, a capacitor arranged to be charged substantially to said peak value voltage by said rectified voltage, and means for delaying the discharge of said capacitor.

2. In a system forl deriving an envelope voltage proportional to 'the peak values of an alternating voltage, the combination of means for rectifying said alternating voltage, a capacitor 7s arranged to be charged substantially to said peak value voltage by said rectiiied voltage, means affording a discharge path for said capacitor, low

pass ltering means connected across saiddis- 5 charge path, and means for delaying the discharge of said capacitor for a. period comparable with the transmission time of said filtering means whereby the maximum rate of capacitor discharge is attained subsequent to the instant of maximum applied voltage.

3. In a system for deriving an envelope volt'- age proportional to the peak values of an` alter- Jnating voltage, the combination of means for rectifying said alternating voltage, a capacitor arranged to be charged substantially to said peak value voltage by said rectified voltage, means including an inductor aiording a discharge path for said capacitor, low pass ltering means connected across said discharge path, and means for delaying the discharge of said capacitor for a period comparable with the transmission time of said illtering means whereby the maximum rate of capacitor discharge is attained subsequent to the instant of maximum applied voltage.

4. In a system for deriving an envelope voltage proportional to the peak values oi' an alternating voltage. the combination of means for rectifying said alternating voltage, a capacitor arranged to be charged substantially to said peak value voltage by said rectiiied voltage, means including an inductor shunted by a non-linear resistance device aifording a discharge path for said capacitor, low pass iiltering means connected across said discharge path, and means for delaying the discharge of said capacitor for a period comparable with the transmission time of said nltering means whereby the maximum rate of capacitor discharge is attained subsequent to the instant of maximum applied voltage.

5. In a system for deriving an envelope voltage -proportional to the peak values of an alternating voltage, the combination of means for rectifying said alternating voltage, an envelope voltage delivery capacitor arranged to be charged by said cluding an auxiliary capacitor shunted'by a resistor, and additional rectifying means arranged to charge said auxiliary capacitor for delaying the discharge of said envelope voltage delivery capacitor.

6. In a system for deriving'an envelope voltage l proportional to the peak values of an alternating voltage. the combination of means for rectiiying said alternating voltage, an envelope voltage delivery capacitor arranged to be charged by said rectified voltage, a capacitor discharge circuit includingan impedance device, and means including an additional rectier connected to said discharge circuit'i'or delaying the discharge of said envelope voltage delivery capacitor.

7. In a system for deriving an envelope voltage proportional to the peak values of an alternating voltage, the combination of means for rectifying said alternating voltage, an envelope voltage delivery capacitor arranged to be charged by said4 I rectied voltage, a capacitor discharge path inmeans for varying the impedance of said path in response to change in the value of said rectiiied voltage.

8. In a system for deriving an envelope voltage proportional to the peak values oi an alternating voltage, the combination of means for rectifying said alternating voltage, an envelope voltage delivery capacitor arranged to be charged 'by said rectied voltage, and means including a delay network for applying to said envelope voltage delivery capacitor successive voltage impulses from said alternating voltage whereby discharge of said capacitor is delayed.

9. In a system for deriving an envelope voltage proportional to the peak values of an alternating voltage, the combination of means for rectifyingsaid alternating voltage, an envelope voltage delivery capacitor arranged to be charged by said rectified voltage, and means including a delay network for applying to said envelope voltage delivery capacitor successive rectified voltage impulses from said alternating voltage whereby discharge of said capacitor is delayed.

10. In a system for deriving an envelope voltage proportional to the peak values of an alternating voltage, the combination of means for rectifying said alternating voltage, an envelope voltage delivery capacitor arranged to be charged by said rectified voltage, and means including a delay network for applying to said envelope voltage delivery capacitor a series of successive voltage impulses from said alternating voltage whereby the discharge of said capacitor is delayed.

l1. In a system for deriving an envelope voltage proportional to the peak values oi an alternating voltage, the combination of means for rectifying said alternating voltage, an envelope voltage delivery capacitor arranged to be charged by said rectified voltage, and means including a delay network and a pluralityof rectifying elements connected to different points of said network forv applying to said envelope voltage delivery capacitor a series of successive rectified voltage impulses from said alternating voltage whereby the discharge of said capacitor is delayed.

12. In a system for deriving an envelope voltage proportional to the peak values of an alternating voltage, the combination, means for rectifying said alternating voltage, a capacitor arranged to be charged by said rectified voltage, and transmission line means arranged to apply a single one of said peak voltages repeatedly to said capacitor whereby discharge of said capacitor is delayed.

13. In a system for deriving an envelope volt- EDWARD W. KELLOGG.

Images