US2440553A - Amplifier system - Google Patents

Description

April 27, 1948.- I

Mall I M. 'MoR rsoN 2,440,553 AMPLIFIER SYSTEM v Filed Aug. 5, 1943 2 Sheets-Sheet 1 i -IIIIMIIIIF:

IL ll IN VEN TOR. MON7'F,'0RDMORR/SUN ATTK I Patented Apr. 27, 1948 Montford Morrison, Upper Montclair, N. J.

Application AugustS, 1943, Serial No. 497,497

This invention relates to amplifier and detector systems with regenerative feed-back, and relates in particular to the stabilization of such systems.

Among the objects of the invention are; to provide a system of regenerative amplification and detection for certain types of signal-currents which system does not produce heterodyne effect in the output circuit; to provide a receiving circuit filter-system for certain types of signal-currents which effects a sharpness of tuning corresponding to a higher Q than present in the structural design of the filter per se; to provide an amplifier-detector system for certain types of signal-currents which produces a relatively high output level and which level is, within wide limits, not sensitive to the voltage level of the input signal; and to provide positive stabilization in a regenerative, non-heterodyning amplifier-detector circuit for certain types of signal currents.

These and other objects will be obvious upon reading the specification in connection with the accompanying drawings, in which Fig. 1 shows an embodiment of the invention in a voice-frequency carrier telegraph receiving circuit; and Fig. 2 shows a series of oscillograms illustrating some of the characteristics of the circuits shown in Fig. 1, operating under the present invention,

scams. (or. 179-171) lator does not respond because of the oscillating frequency being above the response frequencyof the translator. If a signal frequency is injected into the input circuit and which signal frequency is sufficiently different from the oscillator fre- 1 quency, and at the same time within the band of frequency transmission of the tuned circuit, this input signal frequency may be superimposed upon the oscillator frequency, heterodyning with it and producing a resultant frequency to which the output translator will respond. As: the injected signal input frequency approaches very In order to properly define the scope of this invention, it will be advantageous to review and define the pertinent prior art.

In this specification, degenerative amplifiers are considered as not coming within the category of regenerative amplifiers.

This invention has particular application to embodiment in tuned circuit amplifiers in which the output current does not necessarily bear a close qualitative relation to that of the input sigclosely to that of the oscillator frequency, the oscillator is forced to oscillate at the signal fre- JIuency, producing the so-called silent region. In

other Words, this amplifier- -detector becomes inoperative for the purpose for which it is, intended 'at the point where heterodyning ceases.

' In this class of amplifier-detectors, the feedback energy is limited by theuse of a grid-leak condenser structure, that is, by a self-adjusting bias applied to the grid circuit, and in some cases,

by the saturation characteristic of the'plate curi rent of the tube employed. With an increase in signal current, either perse, or with amplification and by regeneration (if the tube is working below plate-saturation), the plate current is increased also by an appreciably larger amount. These operative characteristics are pointed out as they will be hereinafter referred to.

The regenerative detector, while resembling the oscillating detector in circuit, and in fact only lstructu rally differing from it in the matter of nal. Since in the literature relating to thepertinent prior art, there has been considerable confusion in the nomenclature as well as in the 01 erative theory, I shall not only distinguish and define the two pertinent groups, but also shall review the precise operative theory, in order that guished fromathem.

' These two pertinent groups of devices are cor? rectly referred to as oscillating and regenerative detectors, although the literature quite often confuses the two. It may be well to recall at this well as detectors. In the case of the oscillating detector, the feedback is sufficient to set up continuous oscillations in the output circuit to which the signal transthe present invention may be clearly distiri lfi feed-back adjustment, operates on a widely different principle.

g The feed-back is adjusted to a pointwhere "oscillations in the local circuit do not occur, or at least are reduced to a very low amplitude.

The local circuit, which is" subject to oscillation "under operation, is tuned'to a frequency which will heterodyne with that of the incoming signal. "When the incoming signal is-injected into the "input circuit and amplified 'with the additional efi'ect of regeneration due to th'efeed-back circuit,

this amplified'signal n'er'gy 'ih the plate circuit becomes a source of energy for the local tuned circuit, providing the power required for its oscilflation and'heterodynin in the case of the "oscillating detector. The regenerative detector point that these devices are, in fact, amplifiers as thus operates like two differently-tuned circuits in parallel, which are coupled together and which has energy fed to one directly, the other acquiring its energy by coupling effect. The regenerative I detector, like the oscillating detector, does not respond operatively in its output circuit, except istics, among which are; there is no heterodyning in the output circuit; the tubeor tubes never work under'conditions of plate current saturation; the grid-cathode impedance of the tube used in the output circuit is never relatively any appreciable value of that of the total grid circuit; the feed-back circuit always contains a very high impedance; the positive voltageleveliof the output tube grid never rises substantially above. the zero value; and many other differences 'will be pointed out and obvious in the descriptionhereinafter given.

In Fig. l, I is the output transformer of a receiving amplifier of a voice frequency carrier telegraph system; '2 is a channel comprising a singlestage inductively coupled filter 3 feeding into a class A1. amplifier ,4, which is resistancecapacitance coupled to a detector tube 5 having anelectroemagnetic relay 6 inserted as a load in theplate circuit, with the plate circuitshunted bya condenser I. The resistors 8 and 9 have a very high-value in comparison to the grid-cathode impedance of tube 5; the plate circuit of tube 5 .supplies-feed-back current to filter 3 through .a

high. resistance [0 having a conventional blocking condenser H inserted. I

|2 is a second channel identical with channel 2, withthe exception ofhaving an additional stage inserted into the filter circuit I3, by employment of the reactor l land the condenser l5. In thisparticular channel, the feed-back from the output tube through the resistor Ibis made between the reactor l4 and the condenser of the The connection of the feed-back circuit into the filter .systemdepends upon the phase. position ofthe harmonic in the feed-back current that is used forfeed-back purposes,,with relation to the phase position of the corresponding frequency in-the filter system and which phase position in the filter system is afiected by filter design as well as reflections within the filter itself.

Since the filter used in the embodiment of this invention shown in Fig. 1 operates into a. class A1 amplifier, the ohmic resistance of the circuit .comprising the end section l1.of filter 3, channel 2, as well as the corresponding circuit in channel 1 l2, may beincluded in the reactor of that circuit andthereb-y decreasing the physical size ofthis reactor over what would be required by the use of an external resistor.

The feed-back energy received by this filter circuit serves to neutralize'a large portion of the filter insertion losses, thereby greatly increasing .the'Q of the filter reactors, and thus not only decreasing the over-all size of the filter circuit,

but alsogreatly increasing the sharpness of tun- .ing.

Further, the resistors! and I6 have such a high .value in comparisonwith the impedance of the filters in series with these resistors, that the impedance of the filters themselves contribute no appreciable veffect to the impedance 'of the feed-back circuit as a whole. Therefore,

the current in this feed-back circuit is rather .-:strictly directly proportional to the plate current of the output tube, :and the voltage developedin 4 the filter by the feed-back current is determined by the selectivity of the filter itself. In other words, irrespective of the shape of the wave form of the plate voltage of the output tube and the identical shape of the feed-back current, the filter voltage only takes from that wave form the harmonic component in a very pure form, corresponding to the carrier frequency of the particular channel involved. Therefore, the particular channel under consideration picks out of the feed-back current the correct harmonic component (when its phase position is correct) and utilizes this component in the regenerative amplification employed in the system.

'The'impedance of filter 3 to feed-back current of the'proper'carrier frequency at the point of injection of the feed-back current in the said filter,

'has a very low value looking east and a very high value looking west. In actual practice, under op erating conditions, the presence of feed-back currentfrom resistor H) in outputtransformerl is not detectable in comparison with theamplitude of the corresponding harmonic in output transformer I that is received as signal current.

Referring to Fig. 2,. abscissa Aillustrates in a general-way the amplitude of the voltage .across the coil l8 in Fig. 1, due to signal current alone. Abscissa B shows a relative amplitude of thevoltage across the same coil under feed-shack operation. Abscissa C shows the wave form of the voltage applied to the grid of amplifier tubetlfi. Abscissa D shows .typicalvoltage wave forms developedacross resistor 9 included. in the gridv cir- .cuitof the output tube. 5.

The-gridbias batterylil of Fig. 1 has a voltage value indicated by the heightfll from the abscis saiD in.Fig.-2. The resistor..9 of Fig.1 may have a value as high as the order of one megohm,and the current produced in such a resistor may be of the order, of .a few micro-amperes. The ability of the class A1 amplifier to reproduce the wave form shownat .C in the resistor depends upon the value .of the resistor being constant and .of every high value, Which operating. conditions obtain so longas the grid. of tube-5 does not draw any current. However, as soon as the instantaneousamplitude of the positive lobe of the alternatingvoltage produced in 9 exceeds the bias voltage 20, the grid, of course, beginsto draw current and the impedance between the grid and the cathode of tube 5 drops to a relatively low value. Because of the relatively small amount -of current available in resistor 9 and relatively considerable amount of carrying capacity available between thegrid and cathode of tube"5, practically all of the energy available in resistor -9 represented at voltage value in excess of the bias 2%] .is-sh-unted through the grid cathodecircuit ,tube 5, as will be-fullyrunderstood by those familiar with theart to which this invention appertains. 7

Referring to Fig. 2, abscissa D, curve .21 represents the wave form of voltage applied to the-grid of'tube 5 when the maximum amplitude of'the said voltage is less than the grid bias, producing .ducing a truncated lobe 26, abscissa E, in the plate current and a corresponding platevoltage -21, abscissa F.

. Again,-.if.the voltage across resistor Bisfurther raised, the wave form corresponding to 28ihaving a wider truncated positive lobe 29 produces a wider truncatedlobe of plate current 30, abscissa E, and a corresponding plate voltage 3 I, abscissa F.

While the truncations of the lobes of the waveforms illustrated along abscissa D are represented as being parallel to that abscissa, they do, in fact, rise some slight amount depending upon the circuit parameters involved, but for all practical purposes in a well-designed circuit, these may be considered as substantially flat. It will be observed that the plate voltages of the output tube, as illustrated along abscissa. F, decrease in magnitude of variation with an increase in grid voltage, thereby producing a positive stabilization in the amplifier circuit.

This highly important stabilizing effect results in a circuit which is highly stable to large variations in operating conditions, and, entirely unlike other regenerative feed-back amplifiers, does not tend to oscillate or operate out of control.

While I'have used two triodes in teaching the use of my invention in the embodiment described in this specification, I do not limit myself to the number of tubes nor'to the type of tube used, as my invention may be embodied in circuits having one triode or one multi-grid tube, as will be fully understood by those skilled in the art to which this invention appertains, upon having learned the teachings of this specification.

It is pointed out that in the claim hereunder, the terms internal and external do not refer to the tubes employed but only to indicate that separate circuits are employed, and of course are coupled in the same sense that in a conventional regenerative or oscillating detector employing a pentode tube, the feed-back is external of the input and output circuits. The expressions input and output circuits are used to mean these circuits per se or, of course, circuits coupled with them.

What I claim is:

1. The method of signal-current amplification by regeneration which comprises receiving signal current, filtering out of said current a form of wave comprised of a narrow band of sinusoidal frequencies, amplifying said wave and truncating said form producing a fiat-top pulsating current,

, discharging said current through means reducing the amplitude of the pulsation of said current and feeding back from last said current into first said current for regeneration, a. current proportional to the amplitude of the reduced pulsations in said means.

2. The method of signal-current amplification by regeneration which comprises receiving signal current, filtering out of said current a form of wave comprised of a narrow band of sinusoidal frequencies, amplifying said wave and truncating said form producing a, fiat-top pulsating current, discharging said current through means reducing the amplitude of the pulsation of said current, conducting feed-back energy from last said current into first said current for regeneration and employing the reducing-of-the-amplitude characteristic to stabilize the amplifying of said WaVe.\

3. The method of signal-current amplification by regeneration which comprises receiving signal current, amplifying said current and converting the form thereof into a wave having a direct current component, discharging said wave into means having a direct-current storage characteristic, conducting feed-back current from said means into first said current for regeneration and employing said direct-current storage characteristic to limit said feed-back current to a stable amount.

. MONTFORD MORRISON.

REFERENCES CITED The following references are of record in the

Images