US1863566A - Negative resistance - Google Patents

Description

June 21, 1932. M, DOLMAGE 1,863,566

NEGATIVE RESISTANCE Filed Jan. 13, 1927 2 Sheets-Sheet l INVENTOR June 21, 1932- M. M. DOLMAGE NEGATIVE RESISTANCE 2 Sheets-Sheet 2 OOOW QODN Filed Jan. 15 1927 INVENTO Mai,

Patented June 21, 1932 Tumreo stares MIHRAN M. YDOLMAGE, OF WASHINGTON, DISTRICT OF COLUMBIA NEGATIVE RESISTANCE Application filed January 13, 1927. Serial No. 160,978.

This invention covers a new method for obtaining a negative resistance which is automatically variable within predetermined limits, increasing or decreasing in a definite man: nor, as may be required, with an increase in the outside electromotive force impressed upon the electrical network having the above defined negative resistance characteristics. I have shown in U. S. A. Patent No; 1,606,850, how an arrangement of apparatus can be realized which will have a constant negative resistance characteristic, independent of frequency within the practical limits of a few hundred cycles to about a million cycles per second. There a negative resistance is used for the purpose of obtaining a constant load upon an electric generator with an actually varying line impedance (for a specific application'reference may be made to U. S. A. Patent No. 1,606,350 granted to me, relative to impedance variation compensator circuits) I find greatly improved results can be obtained if the negative resistance instead of being constant should vary with the e1ectromotive forced impressed upon it in a definitely predetermined manner, as fully explained hereunder in these specifications.

The novel features of my invention I have pointed out in my claims. The invention itself will be best understood as to its construction and method of operation, together with further objects and advantages by ref erence to the following description taken in connection with the drawings.

Figure 1 shows one form of the invention. Figure 2 is a modification thereof. Figure 3 is a diagram for illustrating certain features of the operation of the invention.

In Fig. 1 of the drawings the combination of apparatus connected to terminals 1 and 2 to the right of the drawings represents a negative resistance having the particular characteristics fully described in these specifications and shown graphically in Fig. 3 of the drawings. Connected to terminals 1 and 2 by dotted lines, are the terminals of an alternating source 20 in series with impedance 19. Impedance 19 includes the internal impedance of the generatOrand such additional impedance as service requirements might demand. As an illustration, the alternating current source 20 may be working .over a given transmission'line, in which case impedance 19 would represent the impedance of the transmission line as seen from terminals '1 in turn wired across thegrid and filament circuit of the amplifier through the grid biassing battery 13. The plate circuit'ofamplifier bulb. 1a is connected through a suitable plate'battery to the secondary windings 4t and 6, which have as primary windings 3 and 5, respectively.

The system shown on Fig. 1, wired as just described, consists broadly of a networkin which a given electromotive force (that'of alternating current'source 20) is amplified and reimpressed back upon the same network so that, in generahtherewill result an increased flow of current in the circuit of alternating current source 20 and impedance19. The arrangement will be made clearer by the following description of the method of operation of the system. Electromotive force E of alternator (20) forces a current through the transmission line (represented schematically by impedance 19) and winding 3 of transformer A. put circuit, as measured from terminals 8 and 9, is made equal to the impedance of the plate circuit as seen through primary winding 3 when resistance 7 is equal to infinity ohms, then there will he no initial current flow through resistance 7, even when the circuit of resistance 7 is closed. This is a result of the balanced VS- heatstone bridge arrangement of the circuit shown on Fig. 1

of the drawings. Under the above mentioned assumed conditions'this initial current is amplified by bulb 14; andreimpressed back through primary windings3 and 5 upon the outside circuit represented by generator 20 If the impedance of the inc and impedance 19. Impedance 7 can be so chosen with reference to impedance 19 that no amplified current will flow through the input circuit and the negative resistance network represented by the combination of apparatus connected to terminals 1 and 2, as shown in full line, will then be working under conditions of maximum stability. This con dition of maximum stability is obtained when, for instance, windings 3 and 5 are of equal number of turns and of equal impedance, the corresponding secondary windings 4 and 6 bear the same equality relations to each other, and impedance 7 is exactly equal to impedance 19.

The numerical value of the negative resistance maybe changed within very wide limits by a change in the adjustment of the potentiometer without it being necessaryto make any. other'changes in the network. The wide range of adjustment possible with the system, described in these specifications will also. be clear if it is realized that, by a proper shifting of the. potentiometer, the amplified electromot-ive force may be so regulated, that there will be nov current flow whatever through source 20 or winding 3. Conversely, if the wires betweenbulb 14 and secondary windings a. and 6 are reversed, the potentiometer of the. amplifier canagain be so adjusted that theimpedance of the combination of apparatus connected to terminals 1 and 2 will. be numerically. equal to zero. These rather two interesting and striking applications Will be made clear by reference. to the formulze, giving the values of the negative resistance of thenetwork-as indicated hereunder.

The system: herein described differs. physically in no essential parts from the system disclosed in U. S. A. Patent 1,606,350, granted to me. It differs, however, in this outstanding feature, viz. that the plate battery and the. grid battery are so chosen with respect to each. other, and further the load impedance of thevacuum tube is also sochosen, that this-vacuum tube operates as a detectoramplifier combination instead of being arranged to operate as a pure amplifier which isthe arrangement utilized in U. S. A. Patent- 1,606 ,35O granted to me above referred to. The valuable features obtained: by means of the above essential difference will be made very clear hereunder. Be it. noted first that two mainpoints must be properly cared for before a detector-amplifier arrangement, such as referred to above, can be realized.

1. The grid-volts and plate current characteristic' must remain curved and not straightened. out into linear relationship with respect to each other. characteristic becomes linear with the inclusionofan impedance in the plate circuit at least equal tothe impedance of the plate-filament vacuum path. Theclroice in the rela- As is well known, this tive number of turns of primary and secondary windings of transformer A included in the plate circuit must, therefore, be so made that the impedance of the outside circuit connected to the plate-filament path will be small compared with the resistance 1 of the platefilament circuit itself.

2. The value of the equivalent voltage must be numerically smaller than E where E is the filament voltage and E =Eflective voltage available between filament and plate. =Amplification constant of the vacuum tube. E =Voltage between grid and filament.

=Contact potential, usually not exceeding 0.50 volts between the materials forming the plate and the filament respectively.

To show very clearlyv the effect of the use of a detector-amplifier combination, instead of a pure. amplifier, we will first assume for the circuit arrangement of Fig. 1 of the present drawings that the dynamic characteristic has been straightened out and that the varying portion of the current in the plate circuit is given by the first term only in Equa tion (12). r

Let us call V=Electromotive force of an alternating current generator having no internal impedance connected to terminals 1 and 2 of the negative resistance, as shown on Fig. 1 of the drawings.

i,=Current through primary winding 10 of transformer A.

2 Current through primarywinding 3-.and

element 7.

A=Series resistanceofgrid. circuit of amplifier 14 as measured through primary winding 10 and of the resistance 15 in series withsaid winding.

B Series resistance of plate circuit as measured through either winding 3 or winding 5, withthe other winding and its corresponding secondary open- Y=Resistance of element 7.

Y =Resistance of element 19. N =R-atio of terms of winding 11 to wind- N =Rat1o of turns. of winding 3 to winding 4, or similarly,winding 5 to winding 6. R =Resistance of grid circuit as measured through winding 10'. K=Coefficient having a value less than, or equal to unity, depending upon the position of the potentiometer.-

The equations governing the performance of the circuit are as follows The derivation of V will be clear if it is borne in mind that the potential across the terminals of winding is equal to R 21, the corresponding secondary potential is equal to N lil z' which is reduced by the potentiometer to the value KN R Q' The equivalent alternating current potential in the plate circuit is obviously given by V ,aK N R i Introducing the value of V, obtained from Equation into Equations (1) and (2),

the following two relations are obtained between V, 2' and i (4) V= (Al+B+4a2l)i +2Bi (5) V=(213r8aA)i +(4B+Y)i The solution of the above two simple equations gives for (2' and (2' V A (2B+ Y) 11: {B+A(1+4a)}{4B+Y}-2B{2B+8aA} The common denominator in the above two fractions can be further simplified into 12 iA+Y +AY(1+4a) The total current (i l-i into the negative resistance network is given by A(1+4a)+B+Y The total resistance offered by the circuit arrangement shown on Fig. 1 of the drawings is evidently equal to In Fig. 1 of the drawings a single amplifying-detecting unit is shown. It is clear that a number of such units could be used in series and secure thereby as large a value for (a), as we please. It will be noted, first, that when (a) is sufiiciently small so that the denominator of Equation (9) is positive in value, the resistance offered by the network to alternating currents is also positive. In order that the network should act as a negative resistance as intended, we must make It will be also noted that when (a) is given its critical value A+B Y then the resistance oiiered to the passage of of (a) greater than the critical value indicated above, we have anegative resistance. With a further increase in the value of (a) a decrease in the absolute value of the nega- The lowest critical value of (a) which will provide a negative resistance characteristic, is given by a: B(4A+ Y) +AY One of the veryi-nteresting and important applications of the present invention may be briefly referred to here. If (a) is given its critical value as indicated by Formula (11), the network shown in Fig. 1 of the drawings is a short circuit for a very large range of alternating current frequencies, while olfering a definite resistance to direct currents.

It may be noted here that for any given value of (a), a constant negative resistance is obtained regardless of the value. of the resistance of the outside circuit represented by element 19. To secure proper operation, it is sufficient that the unbalance between elements 19 and 7 should not exceed a definite attenuating currents is infinity.- For values tive res? stance results. The asymptotic value limit' If this limit is exceeded the stability of the network would be destroyed. This point has been carefully reviewed in my U. S. A. Patent No. 1,606,350 and will not be referred to further here. In the derivation of all the preceding formulae the assumption is that the amplifier bulb 14 is operating as a pure amplifier. As already noted hereinabove in this specification, the essential characteristics. of the present invention are mainly two+ 1 (1) The amplifier bulb (14) is arranged to function as an amplifier-detector combination and not as a pure amplifier.

(2) The load upon the plate circuit is sufficiently low in impedance so that the characteristic of the grid-volts and plate-current curve will be represented by a curve rathe than by a straight line. 7

lVith regard to the first point, reference may be made here to the following relation which gives the value of the variable portion of the current in the plate circuit, for the general case, which is y I l 1 z 1 1 where y i e =Alternating current electrornotive force impressed upon the grid.

1', =Resistance of plate-filament path to alternating current.

1",,=The first derivative of the alternating current resistance with respect to E (For proof of above relation-see Van der Bijl-Thermionic Vacuum Tubes, page 2517,

Formula 8.)

The first term is the "alternating current component of frequency p in the plate circuit. The second term is proportional to c If e=e sin pt, then 6 =6 sin jot. Replacing sin pt by its equivalent I sin pt=' 2 the second term may be writt i th f 2 lpjil 2 1 2 p p J2 7 p 7 3 +710 7 under.

- Since. as shown in'Formu'la (.13), the increase in the direct current through the plate circuit is inversely proportional to it is evident that to obtain the best results possible for the specific purpose we have in mind, we should-make r, as small as possible with reference to 1",. One of the essential re quirements to secure proper. operation of bulb (14) as an amplifier-detector combination consists in a choice for the value of E for the plate battery such that It must be also clear that Equation (12) is valid, provided (6) the alternating potential imposed upon the grid of bulb (14), satisfies the equation When vacuum tube (14) is working under the conditions which have been specified above, the resistance of the plate-filament circuit will vary in a definite manner, depending upon the value of the electromotive force (0) instead of remaining a constant or very nearly a constant. Referring, therefore, to Equation (9) which can still be used to determine the numerical value of the negative resistanceof the network as measured between terminals (1) and (2), we note that B in said equation instead of being a definite constant varies with (e) as just stated. It follows neces sarily that the negative resistance of the elec' trical network itself as described in these specifications will also vary. It will vary in a predetermined manner, as a definite f1inction of (6). Such a variable negative resistancevarying in a definite manner as a function of the electromotive force imposed upon it, has certain very definite and valuable uses in the electrical art. To show in a more concrete manner how and to what extent a variation in the negative resistance may be secured, for a given fixed setting of the potentiometer of the amplifier bulb (1'4) we will consider an amplifier having the grid volts-platecurrent characteristic, as indicated in the following table:

E 130 volts (plate battery) E 9 volts (grid battery) I =8 mils (plate current) E =4.4 volts (filament volts) We will analyze hereunder the performance characteristics of the above tube and show the effect of the use of such a tube in the circuit of Fig. 1 of the drawings as element (14). We are concerned here with the variation of 17 Since, by definition, 13, is given by the relation It also follows that r which expresses the variation in the plate-filament resistance for a small change in the efiective potential E between plate and filament, is given by 1 i 12 llel ifi dE n I In the case of most vacuum tubes, operated in the ranges now considered most effective in practice, n is approximately 2. The value of r' is therefore equal to, approximately T P: a 1,, n1, Since, further i4 2 I11 21 it follows that In the case of the tube chosen above r =6000 ohms 40 1 =1.6 approximately Hence 48 ohms approximately is equal to 0 volts, we find that the instantaneous current in the plate circuit'is equal to 28 mils. For this value of grid potential the value of E5 130 54 7I ,,".1.6 .002f The next problem consists in the deter- 6000 ohms mination of the value of the detected direct current J which is the direct current component of J Thiscurrent depends upon 1",, and r' which in turn depend upon the same component, so that derivationof l is some-1 what complicated. To simplify the problem, we will neglect as a first approximation, the efiiect of the potential drop 1%, J G in the expression for E,,.

E,, E,,+,i ,-r.Jd

Where p l r Resistance of plate circuit load J =Direct current component of J When the above assumption is made and We take for 1 the value equal to Then (by Formula 13)- r lp We have also shown Hence, substituting these values of 7",, and 1",, into the value of J as given above, we

obtain Jd: 900n(1 m1, 132E Obviously I w I I b J 4 Therefore i v i J 1 milli ampere approx,

The values of r, and 1",, can now be deter-:

mined more accurately by substitution of the above value of J into the formulae relating together these quantities. Such a substitu tionwhen carried out, results in a value for W -44 ohms and for 73, 5237 ohms.

The comparison is therefore as follows: When the alternating current electromotive force (6) is absent, the vacuum tube resistance is equal to 6000 ohms. When the electromotive force (6) is equal to ten volts, the tube resistance drops down to 5237 ohms. It is this variation which it is proposed-to utilize. A specific design was worked out in full detail to indicate the extent of the variation obtainable in the negative resistance network with a specific type of vacuum tube. A similar method maybe used with any other type of vacuum tube. The law relating together the negative resistance' of the electrical network shown on" Figure 1 of the drawings may now be derived in its'general'form, as follows. 7 Theplate resistance is given by, as shown which is'approximately correct within a few per cent, and is quite general in type. In the case of the tube for which a detailed design was given above G is foundequal to .00127. It will naturally vary with the particular type of vacuum tube used.

In order to simplify the expression for 7 as given in Formula (20), we can write it in the form a relation which indicates definitely and clearly the law of variation of 7 with If we now refer to Formula (9) which gives a value of the negative resistance network, we note in that formula the symbol B which is equal to (21) B=N1 r =N (0 0 6 The operation of the detecting-amplifying system as described in some detail hereinabove was outlined on the basis that the amplifying constant of vacuum tube (14) was non-varying, It is known, however, that this Constant does vary to some extent with both t e plate and, grid batteries. ncreasing with increase in the absolute value of both the plate battery and the grid battery. It has lately been found that this varying value for the amplification constant results in the production of the larger amount of second harmonic current than previously derived on the basis of the assumption of a constant value for-theamplification constant. Since the direct current produced in the plate circuit and the second harmonic current are numerically equal to a first degree of approxiniation (see F orniula l3 of the present application) it" follows that the corresponding change in the tube resistance would also be somewhat in excess of the value obtained, using Formula. (18) of the present specification.

It may be of interest, here to also note that the effect of a change in the'value ofthe tube resistance upon the value obtained for the greater than negative resistance network terminating at terminals 1' and 2 of Fig. 1- of the drawings, depends upon the mode of connection between the plate circuit and secondary windings i and 6. WVithout goin into all of the details in the mathematical evelopment it may be shown that for a small increment in the value of tube resistance, the incrementinz the cur rent flowing through the negative resistance network, as wired to. terminals 1 and 2 of Fig. 1 of the drawings, is proportional to {8aA 2A Y} {2A+ Y} It is therefore clear that if constant (a) is 8A then an increase in the value of tubere" sistance (B) will result in an increase in the value'ozt the current flowing through the network. On the contrary it the wire connections between the plate circuit and transformer windings 4a and are reversed, which is equivalent to changing the sign of (a) in the above formula,then any increase in the value of the tube resistance (B) will result in a decrease in the value of the current flowing into the network wired at terminals 1 and 2 of Fig. 1 of the drawings.

As already discussed in detail, we see that a number of interesting and important applications are possible where a plate resistance varying as shown in Formula ('21) is available. For instance, suppose wev place the negative resistance in series with a load receiving its power through impedance (19). from generator (20), and so chose (a) that for a normal voltage (V) upon said load the negative resistance represented, by the "net:- work shown on Figure 1. of, the drawings, is. zero in value. Should a change now occur in the load, the negative resistance will tend to. maintain constant the potential across the load. For proof, it the" load decreases in resistance, the potential (V), available across the terminals of the grid increases; this increase will. result in a decrease of the resistance offered by the network fromzero ohms tosome negative value. The total drop in. potential, between the terminals of the generator source and the load will therefore tend to. decrease, andconsequently, the potential across the load will tend to remain constant. In this case, therefore, the, combination, oi apparatus designated as a variable negative resistance will act asan automatic potential regulator. It is sufiicient and necessary for this purpose that the circuit between the plate and transformer windings i; and 6 should be as exactly shown on the. drawings. If the circuit were to be reversed, potential regulation cannot be obtained.

The arrangement shown in Figure 2. of the drawings consist of two-networksof the'type shown in Figure 1, working on the so-called push-pull basis. When the variable negative resistance as described in these specifications is used in a telephone circuit, it tends to produce harmonics. These have a detrimental efiect upon quality, as well known. The arrangement of Figure 2 eliminates the harmonics in the outside circuit. Figure 3 of the drawings show the variation obtained in the value of the negative resistance with a change in amplification.

The electrical network described in these specifications is a negative resistance independent of frequency, since its value is given by Formula (9), in which A and B are either pure resistances (or can be made to represent pure resistances through devices well known to the art). It may therefore be used for high as well as low frequencies. It may be adjusted within extremely wide limits as already shown in the detailed description given hereinabove. The electrical network described herein furnishes a negative resistance which stays fixed in value, even though the resistance of outside circuits changes, provided the amplifying power of the amplifying circuit included in said network remains constant and provided further the initial potential imposed upon its terminals remains constant. It is stable and nonoscillating in character.

The uses to which a negative resistance, as hereinbefore specified, may be put are so many that no attempt will be made to give even a partial list of such uses. An important application consists in its use in the main antenna circuit of a radio receiving system. A second very important application consists in the possibility it affords to compensate line impedance variations in telephone repeater systems, thereby allowing greater amplification. A third important application consists in its use as a line reflection reducing device when used in conjunction with an impedance compensator circuit. A fourth application consists in its use as an automatic potential regulator, substantially as described.

A further important application of the ar rangements described in the present specifications is in conjunction with impedance variation compensator circuits, which are described in my pending application S. N. 358,- 370. lVhen such a combination is used, it is possible to obtain repeater amplifications in two-way telephone systems, under variable line conditions, far in excess of those possible in the art at the present time.

hat I claim:

1. An electrical network, means to connect said network to an outside circuit, a detectingamplifying system in said network, and means to superimpose upon said outside circuit the electromotive force of the final output circuit of said amplifying system in combinect said network to anioutside circuit, a dete'cting-amplifying system in said network, means to vary the amplification of said system, and means to superimpose upon said outside circuit the electromotive force of the final output circuit of said amplifying system, in

combination with means to prevent the said electro-mot1ve force from belng impressed upon the input circuit of said amplifying systern.

'3. An electrical network having a negative resistance independent of frequency, means to connect said network to an outside circuit, a detecting-amplifying system in said net work. and means to superimpose upon said outside circuit the electromotive force of the final output circuit of said amplifying sys tem. I

4. An electrical network having a negative resistance independent of frequency, means to connect said network to an outside circuit, a detecting-amplifying system in said network, means to prevent the singing of said amplifying system, and means to superimpose upon said outside circuit the eleotromotive force of the final output circuit of said amplifying system. i r

5. An electrical network having a negative resistance independent of frequency, means to connect said network to an outside circuit, a detecting-amplifying system in said network, means to vary the amplification of said amplifying system, and means to superimpose upon said outside circuit the electromotive force of the final output circuit of said amplifying system.

6. An electrical network having a negative resistance independent of frequency, means to connect said. network to an outside circuit,

a detecting-amplifying system in said net-- work, means to prevent the singing of said network. means to vary the amplification of said amplifying system. and means to superimpose upon said outside circuit the electromotive force of the final output circuit of said amplifying system.

7. In combination. a bridge circuit containing a detecting-amplifying system with its initial and final output circuits occupying opposite arms. an outside circuit and a balancing network also occupying opposite arms in said bridge, an electromotive force in said outside circuit, and means to superimpose upon said outside circuit the electromotive force of the final output force of said amplifying system.

8. A two-terminal electrical network, including a detecting amplifying system in said network with mutually non-reacting input and output circuits, both connected to said terminals, with the amplification of said amplifying system adjusted to develop zero iresistance for the two terminal network, substantially as described. a

9. A two-terminal electrical network, including a detecting-amplifying system in said network with mutually non-reacting input and output circuits, both connected to said terminals, with the amplification of said amplifying system adjusted to develop infinity resistance for the two-terminal network, substantially as described.

10. In combination, a transmission line, a.

generator, and means to allow for an increase in normal line load consisting of a two-terminal electrical network including a pushpull detecting amplifying system with mutually non-reacting input and output circuits and with a second network, balancing the transmission line in mutually non-reactive position with respect to the transmission line, with means to superimpose the final output circuit of the amplifying system upon said transmission line.

11. A two-terminal electrical network with means to automatically change the resistance of said network in response to the variations of alternating current potentials impressed upon the terminals of said network consisting of a detector-amplifying system in said network with mutually non-reacting input and output circuits.

In testimony whereof I a fiix my signature.

MIHRAN M. DOLMAGE.

Images