US2944207A

US2944207A - Current loss compensating device

Info

Publication number: US2944207A
Application number: US593889A
Authority: US
Inventors: Rogers Stanley; Toivo M Dannback
Original assignee: General Dynamics Corp
Current assignee: General Dynamics Corp
Priority date: 1956-06-26
Filing date: 1956-06-26
Publication date: 1960-07-05
Anticipated expiration: 1977-07-05

Description

y 1960 5. ROGERS EIAL 2,944,207

CURRENT LOSS COMPENSATING DEVICE Filed June 26, 1956 IN VEN TORS' STANLEY ROGERS 6" 70/ V0 M. 0A NNBAGK United States Patent Tire LOSS COMPENSATING DEVICE Stanley Rogers and Toivo M. Dannback, San Diego, 'Califi assignors to General Dynamics Corporation, San

Diego," Calif., a corporation of Delaware AIFiled June 26, 1956, Ser. No. 593,889

4 Claims. (Cl. 32344) -;-'Aetual etransfo-rme'rs riiffer from ideal transformers in ainumber.s..off:re'spects. All of'these 'ditferences protaduce terrorswhen the' transformer is' b'eing used for com- .z'putationpurposes, as for'example in 'statie stress' c alculations. In the idealstran'sformer itheproduct of the voltwagei: across "the primary times the current through the prii rn'ary is exactly equal to the product of the voltagekacross the secondary times the" current th'rough the secondary. .tThe. t primary: curre'nt and -voltageare in -phase and the sccoiidary current and woltagexare ini phase. In actual wtransformersethewindingsthave resista-nce; th'ereare eddy c'currents rand core: :l-osses, zthel primary inductance and mutualminductancec aret l'ess thani infinite; 'the're is "finite impedance between the primary i'and secondary' windings,

there is distributed capacitance among ithe turns of eaeh winding, and there are" 'impedancesr between leach wind- -'ing and. ground. Further. the effective "transformation natio-eof the transformer varies with the lead: applied' to s-the secondary and Withthe voltage across the primary. 'All of these-things. upsetv the simple relationships of the ideal transformer.

Corrections heretofore have been made on current transformers, voltage transformers and 'instruinenttransformers. :These compriseextra turns" in. the windings, artificially exciting-the core to increase 't-hev'effective per- .meability, and sensing and regulatingdevices forwcorrecting-errors. Since errors normally increase :with increase in input,?one such device adds a correction factor. pro- ;portional: to the input. -tAnother devicez'oomparesi the --output-with.1oad-to" the output of another transformer -.-.w ithout: load, the difference volt-age controlling -aroorerection feedback to the transformer output. Still 'an- ..other device uses-thesecondarywurrent'E-of ithe' trans- -former to 'be corrected .tooppose the ,primarywurrent, the diiference being added to the' secondary' as .a' 'correction current. zvicesapproach the accuracy required in computer work. The transformation ratio varies with--input--voltage-. or r load ;and-- there isno means for controllingi 1thetrans formation ratio as desired. "Theydo not; provide "for current measurements without afie'cting the character- .istics 'of the transformer itself.

.In. the present invention a current-generator type of correction isused \to correct" for the above-mentioned transformer losses and errorsandnforulosses involved in making the transformation ratio. adjustable to any desired value and is independentrof variations in input voltage .or secondary .load. Currentgenerators of adjustable phase. andmagnitude are placed -in error correcting cir- ,cuits to. provide current exactly equal; to andoutofphase i with Lthecurrent flowing: through; those elements: causing the "error. Stepless control of the transformation ratio However, none-of these. correcting de- 2,944,207 Patented J y 12960 "is achieved by the co'mbination of error cancelling circ'uits'and"ajpotentiometer or variable resistance across fp'artof oneof thetraiisformer windings. Additional elernents'areprovided whereby current measurements may 5 be made in which the error correction circuits nullify the additional errors introduced thereby, permitting elements of larger 'magnitude'to'be used for greater accuracy in current measurements.

An' objectofthisn invention is the provision of compensating'means' for making a"-transformer approach the performance of theoretically perfect transformers in operation.

Another'object is the provision'of a transformer circuit wherein all losses and phase shift due to transformer action" arecorrected.

Anoth'ef object 'is the provision of a corrected transformer" wherein the transformation. ratio does not "vary with'fehan'ge in voltage "input 'or "change in transformer 1 load.

Still '1 another object is the provision" of transformer circuitry wherein"current"'measure1nerits may be made without affecting. theid'eal"characteristics er the corrected "transformer. Still 'anothenobjecfis theip'rovis'ion'of stepless control of the transformation ratio ofatransformerin combination with transformererror cancelling circuits.

A'fur'th'er'object is theprovision of a current gene'rator type of transformenerror correction whereby "an i external current source provides an additionatcurr'ent to a "transformer through a negative impedance in an amount equal to the transformer losses.

Other objects and' features of the present invention will be readily apparent 'to those skilled :in' the'iart. from the'fol'lowingspecification and appended drawings wherein-isillust-rated a preferred form of'theinv'ention;'andin which: p

Figure lshows' circuitry symbolic of transformer error in combination-with a theoretically perfecttransfor'rner;

Figure 2 shows compensating circuitry for correcting the errors shown in Figure 1;

Figure B shoWs an alternate transformer showingstepless change in transformation "ratio; and

Figure' -4-shows 'one' preferred"embodirnent'jof correction circuitry.

Refe'rring" now; to"Figure" lglthe" elements equivalent 'to those which'produceerror or-loss 'in' a transformer are shown externally of an ideal transformer for the p'urpose of emphasizing 'thoseerrors for 'which'corrections "must bemade, to thereby obtain a theoretically perfect transformer for use in'analog computer Work, for'ex- =amp1e. The excitation generator 11 energ'izes"e1ements -12 of a computing networkto providean outputsignal acrossthe primary windings 13 of an idealstep-uptrans- 55 former 14. -In the 'ide'al transfor rner 14the"product"of the voltage acrossthe -prim'ary 13 times the current through the primary is-exactly equal to the product or "the voltage across" the secondary 4 16"timesfthe current through the secondaryand these currents 'and voltages 'havephaseangles of 0"--or 180. Howeventhisflsitiration is ideal-and not a'ctually the'casein'conventional transformers "because of the effects *of error elements '17, 18, 19,:21 and 22.

Consider first"errorelements-17 m:Fi-gure 1. They consist of the effectiveleakageinductance:and winding resistances of both the primary and secondary' andsare represented by their sum referred tothe primary. rsError elements 18 'consist ofthe finite inductance offithe primary 13, here shown fin-parallel with ithe infiniteszin- 0 ductance of theideal transformer flythe distributed capacitance of the primary; and the power-absorbing (-resistive) elements not included in 17. Error elements 19 represent the leakage resistance and capacitance between the primary winding 13 and ground. Error elements 21 consist of the same elements between the secondary winding 16 and ground. Error elements 22 represent the interwinding capacitance and leakage resistance. Impedance 23 between the low side of the primary winding and ground consists of circuit elements connected between these points. Impedance 23 does not represent an error term.

In addition to the errors that appear in the conventional transformer, it is generally necessary to add small series resistances such as 24 (t, u and v) and 26 in Figure 2 in order to be able to measure the currents in the primary and secondary windings. The correction circuits to be described will cancel the errors introduced by these metering resistors. Uses of resistors 24 and 26 will be described later. Resistor 27 in Figure 2 is the sum of the copper resistance in both primary and secondary windings plus the value of resistors 26, 32A, 32B and the effective resistance of 33 as seen from the primary.

The correction of error 17 is done by correction circuit 171 in Figure 2. This consists of a negative impedance inserted in series with the primary 13 of the transformer 141. The negative impedance is made up of a current generator 28 whose output is of controllable phase and amplitude, and resistor 29 through which the current from the generator flows. This negative impedance can be adjusted so that the current through the primary of the transformer is in phase with the voltage across the primary. The output of the current generator 28, is, of course, at the same frequency as the excitation to the transformer from 11. For reasons which will be developed hereafter, the correction impedance applied at this point usually will have a net resistive component. There is an alternative method for correcting error elements 17 that can be used, provided the resistance 29 is to be positive. In this method, 29 is chosen of the desired value and is shunted by an inductance or capacitance to obtain the desired phase angle to bring the primary voltage and current into phase with each other and with the exciting voltage. A further alternative is to put the inductance or capacitance in series with the resistance 29. The current-generator type of correction as shown, however, is the most general, since it can produce impedance with either positive or negative resistive components.

The error elements 18 in Figure ,1 can be corrected by a current generator 31 in correction circuit 181 in Figure 2, either alone or in combination with correction circuit 171. By connecting current generator 31 as shown in Figure 2, its phase may be with respect to the exciting voltage for the transformer. Other corrections for this current generator are possible, for example, its high side might be connected just below correction circuit 171 instead of above it. With this connection, however, the phase angle of current generator 31 would have to be adjustable since it would have to correct for the reactive part of error elements 18. Current generator 31 operates on the resistive elements it sees looking into the transformer and causes current to flow through them. It is adjusted so that it feeds in just enough current to offset the losses caused by the resistive element in error elements 18 in Figure 1. If error correction circuit 171 has a positive resistive component, it will be necessary for correction circuit 181 to cancel this resistance as well as the resistive element of error elements 17. Correction circuit 171 can be adjusted to take care of the phase angle errors caused by the finite inductance and distributed capacitance of error elements 18.

Error elements 19 in Figure 1 may be cancelled by current generator 191 of adjustable phase and magnitude which is set to be the negative of leakage impedance 19. Error elements 21 are corrected by current generator 211 in exactly the same way. In practice, it is probable that these error elements can be reduced to negligible significance by reasonable care in the design of a computer using transformers. In this case, of course,

correction generators

191 and 211 would not be necessary. Note that correction generator 221 is connected to the secondary winding 161 through a pair of resistors 32A and 323. These resistors should normally be chosen very large compared to any of the resistors driven by the secondary of the transformer when it is in use, to prevent leakage from the secondary 161 to primary 131 or vice versa. The errors caused by these

resistors

32A and 328 can be cancelled by the correction circuits.

Error elements 22 are corrected by current generator 221 of adjustable phase and magnitude. It is set so that current produced by it is exactly equal to and exactly out of phase with the current flowing through error element 22.

Since in a computing transformer, it is necessary to be able to adjust the transformation ratio to any desired value, a series of taps (or equivalent arrangement) is necessary on the secondary side 161. In order to obtain nearly stepless variation in the effective transformation ratio, a low resistance potentiometer 33 may be placed across a small portion of the secondary winding as shown in Figure 2. The resistance which is added to the circuit by this and the power loss which is entailed in the current circulating through the potentiometer can be oflset by the error correcting circuits previously described, principally by correction circuit 181.

Another method for stepless adjustment of the transformation 'ratio of the corrected transformer is shown in Figure 3. In this case, the secondary 161 of the computing transformer 141 is connected to the tap and one side of an adjustable autotransformer 34. While a'd justable autotransformers have larger losses than some other types of transformers, the correction circuits are capable of effecting these losses so that an autotransformer can be made to appear like an ideal transformer of adjustable ratio. 7

One more correction element is necessary before the characteristics of an ideal transformer is achieved. This is a controllable reactance 35 across the secondary. In all probability it will always be a capacitive reactance and may most simply be a variable capacitor as shown in Figure 2. However, a current generator with the proper phase angle could be used. It would also be possible, if desired, to correct the power losses in the secondary 161 by adding an in-phase current generator across the secondary winding. In practice, however, it is simpler and better to make all power loss corrections by correction circuits in the primary 131 instead of using the

additional correction circuits

171, 191, 221 and 211.

Before turning to the correction circuits themselves, the current metering system should first be considered. In Figure 2, there are three resistors marked 24. These are further identified by t, u, and v, and all have the same resistance value. Assume that correction circuit 171 is adjusted so that the primary voltage and current are in phase. Correction circuit 181 is then making up for the power lost in the resistance 36 which is equal to the equivalent shunt lossm, and the current flowing through resistor 24t is in phase with the current in the primary winding 131. The volt drop acros 24: is, then, in phase with the volt drop across 2414. In the ideal transformer, the current flowing into one terminal of the primary winding equals the current flowing out of the other. With the correction circuits properly adjusted, the current flovn'ng in at 37 equals the current flowing out at 38. This is the corrected primary current Ip.

The primary current may be measured without disturbing the ideal characteristics of the corrected transformer. Since 24! and 2414 have the same resistance, the currents through them are proportional to the voltages across them. From .Kirehofis Law we know that the current flowing through 2414, Iu is the sum of the current flowing out through point 38, Ip, and the correction current, Ic, flowing through 24tin correction circuit 181. Thus 'Ei. Eu Eu;"E t IUI=II7+IL lp-i -Rf z, 17+? .M

also Rt=Ru.' Thus? the primary current Ip. is proportional to the voltage dropacross resistor 24a minus thevoltage drop across resistor 24:.

Eu- Et correction circuits completely'n'ullify errors introduced by metering-resistors 24-and26, these resistors need not be "chosen-very-small (around l ohm is customary) 'tokeep the errors small. Using resistances of lto'100 ohms t greatly faci'litates making current" measurements.

1 Figure 4'jshows one "form of error correction circuit "in -which*the error resistance is connected in' parallel with 'the load resistor. By increasingthe voltage drop: across the loadythe voltage across the error resistance also increases"; Using a constant current source,'this operates to inc-reasethe "efiective error resistancewhich, in parallel, becomes-so great' thatall the current goes'throngh the load resistor. ""Through the use of positive fee'dbackthe circuit acts as a negative impedance device which supplies power tdthe load in an amount controlled by the Voltage drop acrossithe load. Because the circuit in Figure 4 may be usedin any of the correction circuits in Figure 2, letters instead of numerals are used to identify the parts to avoid confusion. As shown, the'circuit consists of two cathode .followers, M'andN. In each cath'odeis a potentiometer. The potentiometer arms are connected to the ends of the primary coil of step-up transformer T. To keep'direct current, from flowing throughthe primary 'of the trans- .former, a blocking capacitor is placed in series with'the primary. 1 Each side of the secondary of thetransformer is connected, to one of the gridsof the cathode followers. "The terminals of thesecondary winding (H and K) are the .outputoterminals of the correctioncircuit. For correction .lcircuits which require phase correction (and most of them probably. will),'-the secondary windingcan beconnected to a ya'riablereactance. j Load resistanceR shunted by an error resistance Re, is connected between the two terminals of the correction device and is in series with other resistances R andtR ;Resistor.R .is grounded and R, is driven by a sinusoidal current generator G. Re

diverts someof'the current that'should flow through R; I ;and causes the voltage across R to be too small. i Since Re I6: Re

3 When the voltage drop acrossresistors.R .and R (to-.be measured) are. proportional to -their..known; resistive values, the compensating current-.10 equals'the currentle through the error resistance. Whenthe cur-rent Icfv has been adjusted so that 1: compensation for the current iloss' through Rhas been Ry-need bemade'inadj-u'st-ing 'Ic. 'Fhis-correctiOncw-rent is derived by the proper positioning of the otentiometers C and D.

The arms of the twopotentiometersCQD nitric-cathodes s of the cathode followers are placed at the B- end "of. the potentiometers. Even-thoughavoltageappears on the grids of tubes' M and N, there is -no"signal 'across 'the l1 primary of transformer T. with' switch S "openfithe voltage across -R g R -"in parallel-"with 'Re, and Ro will divide in proportion to their resistances. I-f-wewill new run the potentiometers up until i such time as *the voltage =-across-secondary winding terminals" H and -K'-'of"trans former T is -the=same as the '=voltageacross resistor R we can close'switchS without'affecting'the circuit. With 0 switch S c1osed-,-=we move' the potentiometer arms farther vup. The transformer will a now deliver powerto resistor R This will' increase thevoltage drop across R which amounts to exactlythe same thing 1 as increasing the resistanceofRe.

--Suppose-that the potentiometer arms are'not at the "bottoms of the potentiometers. Since an AC: voltage appears across resistor R unequal A.C. signals with respect to ground will appear on the grids of the cathode followers M and N. Thevoltage on the grid ofcath'ode 45= follower M' will be greater than that on "thegrid'i'of cathode follower N. The potentiometers; if they are set for: the same gain from each o'f'the'cathode followers, will deliver a signal'across'the primary of'transformeiiT, which is normally a step-up transformer. Itv'vill'tryio 'increase the-'voltage"across"R and so"tryto"drivethe grids more strongly, thus increasing the signalacross -thep'rim-ary. 'Its ability todo? this is names by the fact that 'jit' 'must deliven power to resistor'Ri in order to rincreasethevoltage across it. If'the' potentiometer arms --are set too lowg 'the'tra'nsformer' wilhabsdrbfpowef from the external circuit fand decrease the volta'ge across R If the potentiometers" are 'set so"'-tha.t-the voltage acr'oss points H, K of'the transformer secondary is greater" than the-voltage across R when switch 'S isopen,then when switch S is closed; there will' be anincreasein voltage across R and an increase in drive onthe cathode'followers,'*'-fu-rther increasing the voltage acrossR """Ihisus positive feedback and accounts for the'beha'vior'ofmhe correctioncircuit as a negative impedance device. The "correction device delivers-power to the load R in'an "amount-controlledby voltage drop across R Itshould be n'oted that jsince the feedback ratio is less than unity,

it has a stable operation.

By setting potentiometer arm'C enough nearer to B- 'than"D,"the flow of' currentthrough the transformerprimary is reversed. This reverses the secondary voltage. Current Is flowing through R and Re acts to reduce the volt drop across them. This reduces the excitation on M and increases the excitation on N, producing a further increase in secondary voltage and consequently in Is which madeu =In' thiscaseonlythe voltage drop reading across 7 now flows in the opposite direction. Thus, while action 'on Is is regenerative, it diminishes the total current through R By adjusting C and D, the magnitude of Is can be controlled and hence, Re can be made to appear infinitely large or even negative, as desired.

Different connections within the correction device also can be used to substitute negative feedback for positive feedback, if desired. Terminals H and K may be re versed and arm D set nearer B than arm C so that current will flow in the windings of P as shown. A decrease in driving current from I now results in a decrease of Is, which is likely to fit real-life situations better than the positive feedback case.

Note that if R of Figure 4 had been replaced by a general impedance, the phase angle of the current from the correction circuit of Figure 4 would have to have been adjusted by capacitor Q to the appropriate value to cancel the reactive portion of the impedance.

It should be noted that this correction circuit shown 1 in Figure 4 would not operate unless R is greater than 0. This is the reason that it is necessary to have an im- 7 in the voltage applied across the primary can be automatically compensated for.

If the correction circuit is to be applied to a transformer one side of which is grounded, it can be simplified by omitting one of the cathode followers (replacing it by a connection to ground) and grounding one side of the secondary and one side of the primary.

It should be understood that the particular form of the correction circuit in Figure 4 is not intended to limit the scope of the invention to this particular kind of amplifier. In general, any type of amplifier could be used in place of the cathode followers and isolation circuits other than transformers could be used in place of transformer T.

It is not necessary to use amplifiers and vacuum tube circuits at all for these corrections, though it is believed that they are the most convenient form of correction work. It would be entirely possible to pick up power from the main source of exciting voltage for the computer and feed it through resistors, transformers, and phaseshifters into the various points of the computing transformer networks where it is needed for correction purposes.

While certain preferred embodiments of the invention have been specifically disclosed, it is understood that the invention is not limited thereto as many variations will be readily apparent to those skilled in the art and the invention is to be given its broadest possible interpretation within the terms of the following claims:

What we claim is:

1. Means for correcting transformer error comprising in combination, a transformer having current losses between a first and second point having a current flow therebetween, said first point receiving current from an energy source, means connected between said first point and its energy source for measuring the amount of input current flowing into said first point from said source, means connected to said second point for measuring the amount of output current flowing out of said second point, and circuit means connected across said points for producing a correction current flowing into said first point to equalize said input and output currents, said circuit means comprising means comparing said input and output currents for controlling the amount of correction current flowing into said first point and including generating means producing a current corresponding to the resultant of said input and output currents, said circuit means also including a transformer having its primary winding connected to said generating means and its secondary winding connected across said first and second points.

2. Means for correcting transformer error comprising in combination, a transformer having current losses between two points, said means comprising cathode follower circuits having potentiometer resistors therein, said cathode follower circuits generating a current in said potentiometer resistors responsive to grid bias from a current source connected to said transformer, said points being connected to said potentiometer resistors to receive said current in predetermined amounts therefrom.

3. Means for correcting transformer error comprising in combination, a transformer having current losses between a first and second point having a current flow therebetween, means connected between said first point and its energy source for measuring the amount of input current flowing into said first point, means connected to said second point for measuring the amount of output current flowing out of said second point, and circuit means connected across said points for producing a correction current flowing into said first point to increase the output current to the value of the input current to said first point from its energy source, said circuit means comprising a power supply, a resistor and vacuum tube connected to said power supply, the grid of said tube being connected to said first point for biasing thereby to regulate current flow through said resistor, an isolation step-up transformer, means connecting said resistor to said step-up transformer to induce a current flow therein, the secondary of said step-up transformer being connected across said points to provide said correction current thereto.

4. Means for correcting transformer error comprising in combination, a transformer having current losses between two points, a first known resistance connecting one of said two points to a current source, a second known resistance connected across said two points, current means connected to said one of said points for supplying a correction current thereto, and means for adjusting said current means until the voltage drops across said resistances are proportional to their resistive values, said current means comprising cathode follower circuits having potentiometer resistors therein, said cathode follower circuits generating a current in said potentiometer re- 'sistors responsve to grid bias from said current source,

an isolation transformer having the primary thereof connected to said potentiometers to receive said current in predetermined amounts therefrom, the secondary of said isolation transformer being connected to said one of said points for supplying said correction current thereto.

References Cited in the file of this patent UNITED STATES PATENTS 1,399,968 Knopp Dec. 13, 1921 1,994,279 Higgins Mar. 12, 1935 2,071,832 Harder Feb. 23, 1937 2,179,333 Horsley Nov. 7, 1939 2,243,162 Lee May 27, 1941 2,428,613 Boyajian Oct. 7, 1947 2,602,838 De Boisblanc et al July 8, 1952 OTHER REFERENCES Transformers Their Principles and Design, F. C. Connelly 1950, Isaac Pitman & Sons Ltd, London, pp 39-65.