US1730411A - Herbert grove dorset - Google Patents

Description

OctS, 1929. H. G. DORSEY FILTER Filed Nov. 11,3192? Patented Oct. 8, 1929 UNITED STATES4 PATENT OFFICE HERBERT GROVE DORSEY, OF GLOUCESTER, MASSACHUSETTS, ASSIGNOB TO SUB- MARINE SIGNAL COMPANY, OF PORTLAND, MAINE, A .CORPORATION OF MAINE rILTER Application tiled November 11, 1922. Serial No. 600,350.

The object of this invention is to eliminate noises in submarine signaling. In the art of submarine signaling it is usual to employ microphones or magneto telephones, now commonly known as hydrophones, which are mounted upon orA towed from a ship or similar vessel at a point which will be at all times Well below the surface of the Water in which such vessel is floating. It

has always been found in the past that hydrophones mounted upon a moving vessel were subject to excessive noises created in the hydrophones not only by the machinery upon' rangement or association with each other,

these disturbing noises are liable to be so loud as to completely overpower the sounds ofthe submarine signals produced by a distant source. In order to reduce these disturbing ship and water noises it has been necessary in the past to mount the hydrophones in tanks Within the vessel in the manner described in U. S. Patent No. 768,570 although this reduces somewhat the sensitiveness of the hydrophone. Another method has been to use a large number of hydrophones so arranged and connected that these disturbing noises are suiiiciently out of phase with each other at the several hydrophones to produce an interference which reduces their intensity and thus permits the submarine signals to be more easily heard.

Submarine signals have in the past been most commonly created by bells struck under water or by electromagnetic devices such as the Fessendenoscillator (U. S. Patent No. 1,167,366) Both of these types of submarine sound producers create submarine sound waves of a Well dened dominating frequency, their difference-being that the sound Waves created by the submarine bell are highly damped While those from the oscillator are undamped when heard at any deinite distance from the oscillator. Innumerable attempts have been made to take advantage of this uniformity in the frequency of the sound Waves created by submarine signaling devices by using mechanical means which it Was hoped would be resonant at the frequency of the source of the signals and consequently non-resonant at other frequencies. Water and ship noises are composed of tones of various frequencies and it was hoped that by some such means such noises would be reduced in intensity at. the receiver While the submarine sig'nals would be increased in intensity. None of these devices has been commercially successful, partly owing to the -fact that it is difficult to maintain the desired pitch in a resonant system of which one element is a microphone, and more particularly for the reason that there were often tones in such noises of the pitch of the-submarine signal source to cause the water and ships noises to energize the resonant system and cause tones of considerable energy of the same pitch as that of the submarine signal source. This resonant system therefore failed as false signals were liable to be created. Another method which has been tried has been the use ofelectrical filters between the hydrophones andthe telephone receivers. These electric filters consisted of condensers and inductance coils so arranged and of such proportions that the circuit was resonant only to currents of the frequency of the 'source of submarine signals. Filters of this kind have never been found practicable partly for the reason that the apparatus added to the circuit tended to reduce the intensity of the signal but more particularlyfor the same reason that mechanically resonant systems failed, that is to say, for the reason that the noises, produced by the ship or by the 'Water inthe vicinity of the hydrophones create -by impulse action on the hydrophone some tones of the same pitch as that of the submarine signals and, as a consequence, make it impossible to distinguish the submarine signal.

I have invented a filter which in service under most adverse conditions entirely removes the objections inherent in resonant systems whether mechanical or electrical or in the filters hitherto employed. This filter comprises two thermionic tubes each con! nected with 'one of two interconnected re-Y generative circuits. It has been possible by these means to suppress water and other disturbing-noises and permit the tones of submarine signals to be heardwith relatively great power or intensity. It has been found possible to give such proportions to the electrical devices as far as inductance, resistance and capacity are concerned that all disturbing sounds other than low audible frequencies are filtered out.

In efforts to malie a filter of this character I have tried a very considerable number of circuits and differently proportioned circuits without getting satisfactory results, and I. have described below a circuit which I have found very effective, though my invention may be otherwise embodied as Will be apparent to those skilled in the art.

The effectiveness of the present type of circuit for detecting signals in a medium agitated by disturbing noises and other compressional impacts due to the motion of the water or otherwise is due it is believed to the peculiar characteristics of the circuit. Tt is well known that circuits which are highly tuned and have only slight damping' are very liable to be affected by so-called impact excitation which simply acts in such a manner as a bell would act when struck with a hammer; in other Words, the impact whether it has a characteristic frequency or not sets the circuit in oscillation at the natural frequency of the circuit.

In the present system this is avoided by providing a circuit which is highly timed and slightly damped., notat all times but only periodically at such intervals as controlled by some external means.

In the present invention a three-electrode vacuum tube is used, having plate and grid circuits wherein some of the energy of the plate circuit is fed back to the grid circuit for the purpose of regenerative amplification. By feeding back sufficient energy from the plate to the grid circuit it is well known that such a system can be made to continue its oscillations indefinitely without the aid of any signal. At the point'where such a system just begins to oscillate the resistance and therefore the damping of the system may be considered as negligible. When the system is made to oscillate, the resistance or damping is thought of as being of a negative valine and therefore energy is supplied to the system. Under the latter conditions a tuned circuit such as described in the present specification will, of course, have no damping at all and will tend to maintain itself in oscillation indefinitely. When an impact therefore is applied to a. circu`it under these conditions this impact may easily `start the system oscillating in 'such a manner as to be maintained indefinitely. However, in thel present invention such continuous oscillations by impact excitation is avoided by increasing periodically the amount of energy absorbed by the grid circuit so that the circuit becomes highly damped. Therefore any impact which might have a tendency to maintain the circuit in oscillationwould periodically be damped out by increasing the energy consumed in the grid circuit since in effect the system then becomes a highly damped system.

In the accompanying drawings is shown a diagram of connections for an improved circuit and in which a receiving microphone 10 is connected in series with a battery 11 and a winding 12 which forms a primary coil of a transformer T1. This winding 12 might be lconnected to a magnetophone instead of the microphone 10 and in general is connected to any instrument which may receive the compressional waves or other sound signals which are to be used in the work and generate an electrical current or cause variations of current corresponding to the compres sional wave impulses received. Transform er T1 is made with a laminated iron core and has also a secondary coil 13, the outside end of which is connected by wire 130 to the plate 15 of the thermionic valve B and the inside end by wire 131 to the positive side of battery 20. The tertiary coil 14 of the transforirter Tl has the inside end of its winding connected by wire 140 to thc grid 16 of the thermionic valve B and the outside end of its winding by wire 141 to the filament 17, this side of the filament being also connectedto the battery .18, whose positive side is connected to the negative side of the plate battery 20. The secondary winding 13 has one end connected to the'plate 15 of the thermionic tube and the other end to the positive side of the battery 20. The primary winding 12 is connected across the microphone 10 in series with the battery 11. The other side of the filament 17 is connected tothe rheostat 19, which is connected to the master rheostat 21, which is connected through the ammeter 22 to the negative side of 'the battery 18.

Across the terminals of the tertiary winding 14 is arranged a switch arm 35 so that it may be connected to any one of the condensers 32, 33,` 34, and there is also connected across those terminals a variable condenser 31.

The grid 16 of the thermionic valve B is also connected by wire 23 to the secondary wind ing 36 of the laminated iron core transformer T2, and this winding is shunted with a variable condenser 37. The inside end of the winding 36 is connected by wire 24 to the grid 38 of thermionie valve C which has one side of its filament 39 connected by wire 25 to the positive side of the battery 18, the other side of the filament being connected through the variable resistance 51 and the master rheostat 21 and ammeter 22 to the negative side ofbattery 18. The plate 40 of the thermionic valve C is connected to the `grid through the variable condenser 53 and the variable inductance 52 and the plate is also connected bv wire 26 to the outside end of the winding 54 of the transformer T2. The inside end of said winding 54 being connected by wire 27 to one side of the telephone 55, the other side of which is connected by wire 28 to the battery 20. A condenser 56 is bridged across the telephone 55 and plate battery 20. It should be understood that I refer to inside and outside ends of the windings of transformer T1 and T2 to show that there is to be a regenerative action in these 'two transformers and that they must be properly connected or regeneration'will not take place.

To those versed in the art it will be readilyV noticed that the connections might be reversed by pairs and the regenerative action still maintained.

The operation of this circuit will now be described. In terms of telephone coil coupling there is very close coupling between the coils 13 and 14 and very loose coupling between these two and the coil 12. In transformers built for actual work the coil 14 is wound with about 1200 turns of #24 enameled. wire. having an inductance of 95 millihenries and coil 13 is Wound with about 1250 turns of #24 enameled Wire having an inductance of about 133 millihenries, the two coils having a coefficient coupling of .99. The very loose coupling between the coils 12 and 13 may be obtained in any suitable manner known in the art.

Considering the circuit composed of the thermionic tube B, the two batteries 18 and 20, the transformer T1 and the condensers across its terminals, the regenerative action between the coils 13 and 14 will cause violent singing of the system at a frequency depending upon which of thecondensers 32, 33 and 34 are connected across the terminals of coil 14 by the switch 35. These condensers are so adjusted that they will tune the system to any three standard fixed frequencies, such as 540 cycles, 1050 cycles and 1215 cycles, being for the coil values above mentioned about .73 mic'ofarads, .20l m.f. and .10 m.f. respectively, these being common frequencies used with different submarine signaling oscillators and bells. The variable condenser 31 is used to produce slight changes in tuning to meet the conditions of any individual submarine signal sender.

Considering now the second part of the circuit, namely, the thermionic tube C, transformer T2, the batteries 18 and 20 and the telephone 55, the regenerative action in the transformer T2 will produce an electrical oscillation in the circuit depending upon the constaluits` ofthe transformer and the condenser 37, as. welljasthe variable inductance 52 and the variable condenser 53. Good results have been obtained by having the transformer T2 made of 2400 turns of #34 enameled wire for the primary coil 54 and of 8000 turns of :#:28 enameled wire for the coil 36 and having a capacity of 1 m.f. for the condenser 37. The frequency of the oscillations of the first tube feeding back on thc grid of y the same tube in the well known regenerative method and is adjusted and tuned so that the greatest variations of the grid potential will be symmetrical with a point on the above mentioned characteristic curve,-Which is the middle of the straight portion of the curve. Pure maximum amplification of the signal may then take place without an;7 distortion and without rectification. The oscillations of the incoming signal are passed on to the second tube similar in wave formv to what they were when received but ampli fied. If the self oscillations set up in the first tube, however, by reason of the operation ofthe tube at the Apoint in the characteristic curve mentioned above, are allowed to continue without interruption the ability of the tube to amplify is greatly diminished.` If the tube could remain in a state so that each impressed oscillation due to the incoming signal would force the tube to build up oscillations, a maximum amount of amplification would result. The oscillations due to the signal would in this way be regenerated over and over to the maximum capacity of the tube from practically an inoperative state of the tube, in contradistinction to the regeneration from the self oscillating state to a maximum oscillating state as in thecase of simple regeneration.

By increasing the filament current in the second tube which is tuned to a frequency of about 5 to 12 vibrations per second, these low frequency vibrations are impressed through the feed back system, transformers T2 and T1 upon the grid of the first amplifier varyingl periodically the grid potential of the first amplifier. By varying the grid potential of the first tube in this way, the self oscillations of the tube are not given a chance to build up and sustain themselves because during certain times, though relatively short as co-mpared to the total periodic variation, the point of openation of the tube is at the ends of the straight portion of the characteristic curve its fullest capacity.

Where `the self oscillations of the tube cannot Well exist.

The action of the circuit is then as follows The incoming signals are received by the detector, microphone or any suitable means and act to vary the electrical oscillations in l the coil 12. These electrical oscillations, transformed by the transformer T, act on the Grid 16 to vary its potential, thereby chang ing the plate 15 current-according to'the variation of the grid potential 16.- The action of the coil 13 of T then feeds these oscillations back upon the grid 16 which process continues until the tube has amplified to At the same time the tuned filament grid circuit is partly tuned to the incoming signals, thereby eliminating some of the disturbing noises. The oscillations 'of the first tube are passed on to the second tube through the transformer T2 and in a similar Way acting upon the grid 38 are thereby amplified. The local circuit 40, 53, 52, 24, 38 is tuned to the incoming signal thus eliminating all the disturbing noises. The signals Without other noises may then be heard plainly in the telephone 55. During this process the second tube has the other function of feeding back sub-audible oscillations on the first tube to keep the firsttube at the point Where the incoming signals will cause the tube to commence to oscillate. This last process is sometimes. called superregeneration.

Now, if the circuit for tube B is operated alone, it will sing at say a frequency of 1050 cycles. Ifthe second tube C is then put in circuit and the filament current gradually made, stronger and stronger a point will be reached when the tube B will cease to oscillate,but there Will still be oscillations in the tube C. At this point incoming signals will be amplied and noise suppressed, but will be varied or modulated at the frequency of the oscillations of the. tube C and, 'in some form of work, this is not objectionable and in o\ further adjustments need be made. However, for certain other forms of submarine ,signaling reception, especially with bell signals, it is desirable to have the signals come in as nearly as possible with their own characteristics, and to reach this condition the filament current in tubes B and C. are increased in steps, first causing B to oscillate and then destroying'the oscillations by in'- creasing the filament current in C, and at the same time condenser 53. and inductance 52 are adjusted until a condition is reachedin which incoming signals are received with almost their own characteristics and are greatly amplified, While any noises are entirely suppressed unless theybe of such a nature that they continue to excite the system by impulses sustained at exactly the frequency at .which it is tuned.. However, thislatter condition is almost never reached by the noises one encounters in submarine signaling Where one is dealing with frequencies in the neighborhood of from about 200 to 2000 cycles per second.

Vhat I claim as my invention is 1. A. system for receiving substantially undamped compressional-wave disturbances comprising detector means for converting compressional .wave energy into electrical energy, a normally deenergized tuned circuit adapted to continue tol oscillatc when set int-o oscillation by shock, means for damping said circuit periodically at a non-audible frequency andl means coupled to said tuned circuit for receiving signals of undamped audible frequency.

2. A system for, receiving substantially uudamped eompressional Wave disturbances comprising sound translating means for converting compressional wave energy into eleccircuit periodically at a nonsau'dible fre,

quency vincluding a local oscillatory circuit and -means for impressing potential varia-v tion therefrom on the gridv of said vacuum tube at said non-audible frequency and means coupled to said tunedcircuit forreceivmg signals of undamped audible frequen C Q HERBERT GROVE DORSEYY.

Images