USRE23366E - Automatic control of variable - Google Patents

Description

May 22, 1951 A. CALLENDER ETAL I R 23366 AUTOMATIC CONTROL OF VARIABLE PHYSICAL -CHARACTERISTICS Original Filed Feb. 17, 1956 5 Sheets-Sheet 1 INVENTORS ALBERT CALLENDER ALLAN BROWN STEVENSON ATTORNEYE.

5 Sheets-Sheet 2 A. CALLENDER ET AL AUTOMATIC CONTROL OF VARIABLE PHYSICAL CHARACTERISTICS May 22, 1951 Original Filed Feb. 17, I956 INVENTORS ALBERT CALLENDER ALLAN BROWN STEVENSON 'BY ATTORNEYS.

May 22, 1951 A. CALLEYNDER ET AL Re. 23,366

I AUTOMATIC CONTROL OF VARIABLE PHYSICAL CHARACTERISTICS Original Filed Feb 1'7, 1936 5 Sheets$heet 5 r fi r0 U. INVENTORS ALBERT CALLENDER Ll- ALL-AN BROWN STEVENSON ATTORNEYS.

May 22, 1951 A. CALLENPER ETAL 23,366

AUTOMATIC CONTROL OF VARIABLE PHYSICAL CHARACTERISTICS Original Filed Feb. 17, 1936 5 Sheets-Sheet 4 INVENTORS ALBERT GALLENDER ALLAN BROWN STEVENSON ATTORNEYS.

May 22, 1951 A. CALLENDER ETAL Re. 23,

AUTOMATIC CONTROL OF VARIABLE PHYSICAL CHARACTERISTICS Original Filed Feb. 17, 1956 5 Sheets-Sheet 5 INV EN I 0R5 ALBERT CALLENDER ALLAN BROWN STEVENSON ATTORNEYS.

Reiuued May 22, 1951 AUTOMATIC CONTROL OF VARIABLE PHYSICAL CHARACTERISTICS Albert Callender, Northwich, England, and Allan Brown Stevenson, deceased, late of Northwich,

England, by Leeds and Northrup Company, assignee, Philadelphia, Pa., assignors, by mesne assignments, to Leeds and Northrup Company, Philadelphia, Pa., a corporation of Pennsylvania Original No. 2,175,985, dated October 10, 1939,

Serial No. 64,336, February 17, 1936. Application for reissue July 18, 1950, Serial No. 174,529. In Great Britain February 13, 1935 23 Claims. 1

invention relates to the automatic con trol of variable physical characteristics, for example temperature or pressure, and more particularly to the achievement of this aim by a system which is entirely or substantially entirely electrical in operation.

The essential desideratum in the type of automatic control system here proposed is that any devatlon of the characteristic, as indicated, from its desired or set standard value, shall operate compensating means in such a way that the compensating effect (V), resulting from .the operation of the compensating means,,sha1l be dependent upon the behaviour of 0, in accordance with come particular chosen law.

If the compensating effect V is applied in direct proportion to the magnitude of the deviation 0, over-compensation will inevitably result. To eliminate the consequent hunting and instability of the system, the compensating effect is additionally regulated in accordance with other characteristics of the deviation in order to bring the system back to the desired balanced "condition as rapidly as possible. These characteristics include in particular the rate of deviation (which ,may be indicated mathematically by the timederivative of the deviation) and also the summation or quantitative total change of the deviation over a given time (which may be indicated mathematically by the time-integral of the deviation). pensating efiect may be based upon a mathematical formula incorporating the indicated characteristics. A formula of this type is set forth hereafter, together with certain variations therein.

An object of the invention is to provide a control which will apply a compensating effect in accordance with the principles indicated, and preferably which may be arranged to follow with accuracy a mathematical law incorporating said principles. A further object is to provide an electrical system operating in said manner; and more specifically, to provide the simplest possible .elements in such an electrical system such as resistances and condensers, which may be readily interchanged or adjusted. The system likewise is adapted to be designed with a minimum of moving parts.

An important feature of the invention is the arrangement of a system which will operate in the manner indicated through changes in potential, the various rate and quantitative factors being provided by suitable adjustment of leakage rates through appropriately selected and ar- 2 ranged parts of the circuits, this type of system being extremely simple to construct and adjust, and beingexceptionally free from mechanical and electrical difliculties.

A specific object of the invention is to provide a system which will produce a compensating effect governed by factors proportional to the total extent of the deviation, the rate of the deviation and the summation of the deviation during a given period, as indicated above. A further purpose of the invention is to provide an electrical system which will apply a compensating effect in-accordance with factors corresponding to the extent of the deviation and the summation of the deviation.

The effectiveness of the system may be enhanced for certain purposes by the addition of an exponential factor which is sometimes desirable inasmuch as it produces a temporarily exaggerated compensation whenever the characteristic to be'controlled changes suddenly; and a fea- This principle of control of the comture of the invention is the introduction of this factor together with suitable apparatus, and, in a preferred embodiment, suitable electrical apparatus, for superposing this factor upon the other factors already indicated.

It has been found that control apparatus functioning in accordance with the above indicated principles should include the following elements:

(a) An indicator ofthe quantity under control, which provides a physical change in position or intensity proportional to the deviation 0.

(b) A compensating means, (for example, a steam valve where the physical characteristic to be controlled is temperature) for making compensatory changes in the quantity under control and operated, for instance, by a reversible motor, whose operation is directed by an electrical control system.

(c) A coupling means by which changes in position of the compensating means are translated by interchanges in the electrical system by an amount sensibly proportional to the changes in compensating efiect into changes of potential or current flow at a tapping point.

(d) An electrical control system including the tapping point X and a terminal point P, together with suitable electrical means varying the potential at P in accordance with the indicated principles covering the desired rate of application of the compensating effect.

(e) Means for converting the resultant change of potential at P into a physical change in posivided by .(a) above.

- cases A further purpose of the invention is to provide an arrangement of the type indicated in which the compensating means may be controlled in accordance with any reasonable desired behavior of the characteristic. For example, the system may be applied to the automatic control of a temperature which is required to change with time.

In a further extension of the system the action of the coupling means may be modified or varied in accordance with changes in the magnitude of a further characteristic, hereinafter referred to as a causal characteristic, the variations of which may be expected to lead to variations of the characteristic to be controlled. For instance, where the characteristic under control is temperature, the causal characteristic may be pressure.

A further feature includes the incorporation of means for compensating for backlash or lost motion between the actuation of the compensating means and its controlling eifect. This is accomplished in general by providing a delay between the commencement of motion of the compensating means (when oppoflte to the preceding movement) and the corresponding change in the electrical control system.

The invention likewise includes the incorporation of meanssuch as a suitably calibrated resistance for applying to the electrical control system a change, such as a. change in potential, which is sensibly proportional to the change in the compensating effect rather than to the movement of the compensating means.

Other objects and advantages will appear and be considered in connection with the accompanying drawings, illustrating certain preferred embodiments of the invention, in which- Fig. l is a diagram illustrating the fundamental principles of the invention and an arrangement of electrical apparatus suitable for its operation; I

Fig. 2 is an elaborated and somewhat modified system incorporating the principles and arrangement indicated in Fig. 1;

Fig. 3 is a similardiagramoi' a modified system;

Fig. 3a shows a modified form of resistance for use in the circuit illustrated in Fig. 3;

Fig. 4 is a diagram indicating a method of conmeeting a thermocouple to the system shown in Fig. 3;

Fig. 5 is a view of a portion of the system shown 7 in Fig. 3 indicating an arrangement substituting an electrochemical cell for the initial condenser; and

Fig. 6 is a diagram illustrating the addition of an exponential factor to the arrangement illustrated in Fig. 1.

Referring to Fig. 1, the suppl of heating medium, e. g., steam, to the system the temperature of which is to be controlled, is via a pipe I having a regulating valve 2, this valve being operated by a motor, and reduction gearing not shown. Such a motor is indicated diagrammatically at VM. The indicator of the valve carries a wheel 3. A second wheel 4, coaxial with 3, is connected to the latter by an adjustable pin and slot device so as to permit of a small pre- 4 y and terminals. A potential dependent upon the setting of the valve is thus impressed on the fixed contact arm 6, corresponding to thepoint X previously referred to.

Contact arms Ii is connected to One plate of a condenser 1, the other plate of which is connected via resistances 8 and 9 to a suitable indicator ID of the potential at point P which provides in member M: a displacement of its end sensibly proportional in this example to the deviation 0 in the potential at P and corresponds to the device before described in paragraph (e). Point P is itself connected to one plate of a condenser l I, the other plate of which is connected to a point Y, the potential of which is kept constant, e. g., earthed.

At the junction point of resistances 8 and 9, a further resistance [2 branches ofi to point Q, the potential at which is maintained by means of potentiometer l3 at a value corresponding to the desired value of the temperature.

The deviation 0 of temperature from its desired value is indicated by the deviation of the member M1, which is sensibly proportional in this example to 0. By providing a thimble Illa containing mercury at the extremity of member Ml and a 4-point contact system of the kind described in British patent specification No. 440,115 at the extremity of M2 and thus actuating the motor operating the compensating means the maintenance of the desired continuingregistry between M1 and M2 is obtained.

This arrangement is illustrated diagrammatically in Fig. 1, in which M1 is provided with a mercury cup la. The member M2 carries insulated points Ila and Ho connected through the windings of relays 49 and 52 (corresponding to the windings s1 and se of said British patent) to the other side of the supply. Armatures 2| and 22' are suitably connected to the motor VM and operate said motor in the appropriate direction as indicated in said patent, the operation corresponding to that described in detail in connection with Fig. 2. Assume that when the temperature rises the indicator arm M1 rises and that when the temperature falls said arm falls. When the temperature as shown by the indicator is correct the system is at rest, that is, point I 7c is submerged in the mercury in cup Ilia while [1a is located directly above the surface of the mercury. If the temperature rises, contact is made between point Ila and the mercury, thereby energizing relay 49 and permitting current to flow to the servomotor VM in such a direction that the valve is moved in the direction of closing, thus, bringing about a compensating effect by restricting the flow of steam through pipe I. The movement of the valve rotates potentiometer 5, say clockwise, and changes the potential applied thereby at X. I

Assume that this change of potential is an increase. The increase of potential at X produces a corresponding increase of potential on the right determined amount of free relative movement of the two wheels, indicated by the angle a; this free movement is adjusted to correspond to the backlash between the point of attachment of the valve indicator and the valve itself. The wheel 4 carries a circuit adjusting element shown as the resistance 5 of a potentiometer system, connected to a source of supply indicated by the hand side of the condenser! and at the point P. The increase of potential at P moves the current; responsive member M: upwardly, that is, following the'direc tion of the previous movement of M1. Point l'la thus tends to break contact with the mercury in cup Illa: and if the temperature has meanwhile ceased to rise, such contact will be broken, relay 49 will be de-energizedand the servomotor will stop.

If the temperature remains steady in the incorrect (high) value the next thing that happens is that point I'Ia again makes contact with 1 the mercury owing to the loss of potential at P due to the leak to Q, which causes M2 to move downwardly.

Thus after the first make and break at M1, M2, and if M1 is still high, a second make occurs, followed by a break as before. This sequence is repeated so long as M1 is above the value corresponding to the predetermined setting of M: (controlled by potential at Q). Thus the servomotor runs discontinuously to displace the compensating means (valve 2) step by step so as to apply the necessary correction to cause M1 to be restored to the desired value.

A similar sequence takes place if the temperature falls, when M1 falls and contact is broken between the mercury and point IIc. This deenergizes relay 5! and permits armature 22' to fall, causing the servomotor to run in the opposite direction to move the valve 2 in the direction 01 opening. The potentiometer 5 is now rotated counter-clockwise and the potential at X falls, leading to a fall in potential at P. This causes the member m to move downwardly.

, tendin to make contact between point I and th mercury. As the potential at P is now below the standard value, current passes to it from point Q until it is restored. This brings about repeated operation of the relay 52 so long as the temperature is below the desired value, in a similar manner to that described for the case where the temperature is too high.

As the potential at P constantly tends to that at Q, member M2 also constantly tends to return to a standard position. The system can only be in equilibrium when the level of the mercury in cup Illa isbetween points Ho and He; and therefore after every disturbance the compensating means is varied until the correct temperature is restored. Th compensating means camtake up any equilibrium position within its working range without the necessity for adjustment of any part of the apparatus.

Over a given period of time the applied change of compensating eifect is dependent upon the number and duration of the contacts made between M1 and M2.

It can be shown that the apparatus of Fig. l obeys a control law expressed by the equation where -V is the compensating effect 0 is the deviation oi the temperature t is time k1, k: and k: are constants.

If resistances B, 9, condenser II and its earth connections are omitted, k3 is equal to zero and the control reduces to Although the three portions of the electrical system do not behave independently, it is neverthelesssubstantially true (1) that the condenser 1 provides a term in the law proportional to 0; (2) the addition of resistance l2 provides a term proportional to 1 and (3) the further addition of-condenser II and resistances 8 and provides a term proportional to Th choice or the values of condenser I, resistances 8, 9 and I2, and condenser II, determines the values or the three constants of proportionality k1, k: and k: which have been previously selected in accordance with the particular nature or the system under control. The sum of these three terms of potential (klradwawkgg is equated at X to a value of potential which is proportional to the compensating eifect in contra fashion, 1. e., to V. V

The function of condenser I I is to impart to the control system a correction corresponding to the term The apparatus described in connection with Fig. 2 functions in essentially the same way as the arrangement indicated in Fig. 1, and employs the same basic circuit. The principal changes include the addition of means for straight amplification or the movements of the indicator M1; the elaboration of the device indicated diagrammatically at III; and replacement of the simple relay system 49, 52 by a thermionic relay and mercury switch system.

Referring to Fig. 2, the arrangement for amplification of the movement of the indicator M1 includes the parts numbered 2II39. In this figure the indicator illustrated by the thermometer arm 20 carries at its insulated extremity a light aluminium extension II and to this is connected a platinum wire 22, which floats in a small gap between two silver contacts 23, 24. These are carried on a synthetic resin block 25 with a tapped thread of say, /1" Whitworth, into which there fits a similarly screwed spindle 26 having, say, 20 threads per inch. As the spindle rotates, the contact block 25 moves vertically, being constrained hy a pin 21 which fits in a vertical slot in the guide 28. This guide terminates in a horizontal plate at the top, carryin the bearing 29 to the upper end of the spindle 25. At the bottom the spindle fits in a coupling 30, arranged to be driven'by the reduction gear I at a speed of, say, 5 R. P. M. from the servometer 32, when the latter is put in action.

This servomotor is caused to drive the contact block 25 so as to. follow the free movement of the thermometer arm III by the action of two small relays whose coils are indicated at 33 and 34, which coils are included in the respective branches of a relay circuit arranged to be energised by contact of II with either of the contacts 2 3 or 24, the respective relay then closing the power circuit of servomotor 32, at the respective switch 35 or 35, thereby setting the motor in operation to'raise or lower, respectively, the

block 25.

.3 inches, a set of contacm which is composed of two light silver contacts is, a, connected by means of light flexible connections 40, 4| (which may be carried downthe arm Mrto the hub) andv inclusion of resistances i and 54 in the grid 'cir-' ammeter needle, designated M2, the movement 43 of which is of the usual pattern.

The apparatus of which Hlis the diagrammatic representation in Fig. 1 is mounted on a separate panel, and consists of a pentode thermionic valve 44 and milliammeter 43, which transform the potential at the grid of valve 44 into a positive displacement of member M2. Direct current mains suppy the heater and grid bias circuits; across the heater there are two largeecapacity 2-volt accumulators 45 in series, and across the grid bias resistance 46 is a high tension battery 46a of any suitable voltage, say 20 volts. This latter also supplies (1) the graded potentiometer 5, which is attached to the wheel 4, and from which the point X derives its potential; and (2) the linear potentiometer l3 which is fixed, and from which the point Q derives its potential. The condenser I, the resistances 8 and 9, the condenser II and the resistance I2 are substantially the same as in Fig. 1.

The anode current through the valve 44 is. under steady conditions, of a constant value, determined by the value of grid bias voltage at Q, and flows through the milliammeter 43, shunted by a variable resistance 41, to maintain a constant deflection of suitable magnitude. When, however, the temperature changes so as to close contacts 42 and 38, the grid potential of the valve 48 is increased from the negative value due to the grid bias battery 45:: to that of the filament, the valve is rendered more conducting and the contactor coil winding 49 of the mercury reversing switch 50 is energised suiliciently to close the armature circuit of the steam valve motor VM in a certain direction from the terminals marked and which may be connected directly or through a limiting resistance to any suitable source of electrical power, the field circuit of the motor VM being continuously supplied in one sense only from the same power source, which -may be direct current or alternating current supply. The positive supply main is connected to the pentode heater via two resistances in series, BI, 62, 50 rated to pass the heater current con tinuously, of which 62 is adjusted to give a voltage drop adequate to supply the heaters of valves 48 and 5|. The direction of closing of the armature circuit of the motor VM is such as to cause the compensating means (i. e. the steam valve) to move in a direction to tend to restore the value of the temperature, and as the steam valve moves, the change in potential of the point X applies'a changing potential to the right-hand plate of condenser I, and causes a change in the potential of the point P which is the control grid potential of valve 44, through the intermediary of resistances 8 and 9. This latter change is of the correct polarity to change the value of anode current through the valve 44 in such a manner as to make the milliammeter needle M: follow the movement of the amplified temperature indicator M1, i. e., to maintain the desired condition of correspondence, and is continuous until the pla 'tinum wire 42 is once more clear of contact 38. The converse is equally applicable when platinum wire 42 touches contact 39 whereupon valve ii is rendered conductive to energise contactor winding 52 to close the direction.

Momentary contact, due to mains fluctuationsor;other cause is, howevenof itself incapable of rendering .valve 48 or 5| conductive. Due to the cults of the valves, of condensers 55 and 56 in the grid-cathode circuits, and of'by-pass resistances 51 and 58, contact must be maintained between wire 42 and contacts 38- or 39 for arr-appreciable time, say 5 seconds, before the gridattains its final potential; similarly, momentary breaking of contact does not interrupt the contactor circuit and cause chattering, since thegrid condenser 55 or 56 does not immediately lose the grid potential but maintains it to a sumcient degree, for, say, 5 seconds.

The resistances 8, 9, and I2 however, any one of which may be variable, allow the charge on the control grid of valve 44 to leak away slowly until the potential at point P is once more restored to that of point Q, when the grid bias and anode current are once more at their normal values. Hence the milliammeter needle Ma slowly restores itself to the mean position of constant deflection, intermittently causing wire 42 to touch contact 38 or 39 and hence altering the setting of the compensating means, and since this can only occur concurrently with the restoration of the indicated temperature to its normal value, the valve is further compelled to stabilise itself in a position suited to the changed conditicns.

It is well known, of course, that the leakage rates through the respective resistances 8, 9 and l2'are governed entirely by the potentials across them. The mathematical discussion which explains their actions has been set forth at the end of this specification.

The potential at the point P may be modified by insertion or alteration of the resistances 59 and ill, which are controlled by variations in the magnitude of any causal characteristic, or of a number of pairs of these resistances, corresponding to the number of causal characteristics it is necessary to consider, as has been previously described.

For example, let it be supposed that the pressure of the heating medium is being regarded as a causal characteristic. If the pressure changes, which would lead eventually, i. e., after an interval of time peculiar to the system under control, to 'a change in-temperature, the resistances 59 and ill are directly increased and/or decreased by an appropriate amount, causing the potential at point P to change, whereupon the control system operates as though the temperature had already changed, although it has not yet been aifected by the change in pressure of the heating medium. Thus, the control system anticipates the eilect of variations in the causal characteristic and the control is smoother than would otherwise be obtained. In some cases it may be rents operating through a double contact'reversing relay. The manner 'of functioning of the arrangement shown in Fig. 3 is fundamentally mercury switch St in a reverse ases 9 the same as that described in connection with Fig. 2.

Referring to Fig. 3, the thermometer, or other indicator, I20, carries at its insulated extremity a light contact I2I made of platinum or other noble metal, which bears on the close meshed resistance I22 of a potentiometer system, across the opposite ends I22a, I221), 01 which a potential of, say, 20 volts, is applied. The metallic resistance I22 may alternatively be replaced by a liquid resistance I23, as shown in Fig.- 3:1, for example, in. which .the moving electrode I24 moves between the two fixed electrodes I25, I26, the electrolyte being water or any other inorganic .or organic liquid with a high specific resistance.

Resistance 5 is connected in parallel with resistance I22 (or I23) and'resistance I3, i. e.,.

substantially across the sanie source of E. M. F. I21, but the voltage applied across each resistance need not be the same. Tapping point 6 is led via its network of condenser 1, resistances 8, 9. and I2, and condenser II to point .P which is the grid I3I of a triodeor pentode amplifying valve I30 of large handling capacity, the anode I32 of which is connected to a positive potential of, say, 200 volts or over with respect to earth, via a resistance I33 of approximately 10 times the value of the resistance of relay I50, of which one'terminal of the moving coil is connected to the anode I32. The cathode I34, which may be directly or indirectly heated, is connected, in common with cathode I44, to earth.

Tapping point I2I (or I24 in Fig. 3a) is led by connection I2Ia. direct to the grid I of a similar triode (or pentode) amplifying valve I40, the anode I42 being connected to the same potential via a resistance I43 of the same value as resistance I33, the second terminal of relay I50 being connected toanode I42.

In series with the two filament heaters on the negative side is the source resistance I21, which is sufiicient to give a voltage drop of, say, 40 volts, when theheater current is flowing. In parallel with this resistance and each having one end connected to the common connection between cathodes and heater circuit the three parallel potentiometers 5, I3 and I22 (or I23) are arranged, the other ends terminatin in three sliding contacts I285, I201a, I28m. A barretter or loading resistance I23 is included in the positive side.

Relay I50 is a sensitive low-consumption microammeter type and is provided with an adjustable shunt I5I across its moving coil terminals. It is caused to move its light contact I52 to left or to right according as the circuit composed of resistance I33 and valve I30 carries more or less .current than the parallel circuit I43, I40. Since I33 and I43 are invariable, the relative impedances of valves I30 and I40 decide the movements of contact I52, and in practice these impedances are determined by the negative potentials applied to each grid. The object of the relay I50 is to cause as a first consideration the negative potential derived from resistance 5 and applied to valve I30 to follow in stepwise manner the changes in negative potential derived from movements of the thermometer arm I20 as interpreted by contact I2I .(or I24) on resistance I22 (or I23) and applied to valve I40. When the two negative potentials are unequal, and the anode currents ther'efore unequal, the relay I50 I53 orcontact I54.

10 V In the former case valve I is caused to raise its grid potential from earth to that of the cathode I6I, due to charging of the condenser I62 through delay resistance I63, until the contactor coil winding I84 01 the contactor I is sufliciently energised to close the armature circuit of the steam valve motor VM in a certain direction. This direction is such as to cause the steam valve to move in a direction to tend to restore the temperature and as the steam valve moves the change in potential of the point X applies a changing potential to theleft hand plate of condenser I and causes a change in the potential of the point P which is the control grid negative potential of the valve I30. This change is of the same polarity as the.change in potential applied to valve I40 from thermometer arm I20 and when the magnitudes are the same, the anode currents in resistances I33 and I43 are equal and relay I50 is completely de-energised. Before this, however, the contacts I52 and I53 will open circuit; valve I60 does not, however, immediately cease to conduct, but will conduct until the condenser I62 is sufliciently discharged through resistance I66 to lower the anode current in contactor coil winding I 64 below its holdon value; this time interval may, by varying resistance I66, be adjusted to be equal to the interval between parting of contacts I52 and I53, and absolute equality of the anode currents.

In a similar manner, contact on the right hand side between I52 and I54 raises the grid potential of valve I10 to that of its cathode I by charging condenser I12 through delay re-- and I12 and by leakage resistances I66 and I16.

. In this embodiment of the invention, the features which are equivalent to the unit I0 in Fig.

l .are the thermionic valve I30 and the microammeter relay I50, which transforms the posystem previously represented by thermometer arm I 20, contact I2I and potentiometric resistance I22, or alternatively electrodes I24, I 25, I26,

and liquid f resistance I23. The E. M. .F. of thermocouple I is applied direct to the grid of a steep-slope pentode valve I6I, which by means of suitably proportioned values of anode resistance I82 and, coupling condenser I03 amplifies the comparatively small changes in voltage to give a stage gain of, say, 200, and applies its output to a second steep-slope valve I84, with anode resistance I85 and coupling condenser I86, giving the-same stage gain, so that the net efiect is an output of not less than 1 volt per degree centigrade. This output is then finally 11 applied to the grid III of the output triode (or pentode) valve III, with its anode resistance I ll, substantially as hereinbefore described. I

In certain cases where very large time lags are involved in the plant to be controlled, the size of the condenser 1, which for a given decrement is proportional to the time lag, will become unduly large and the cost a deterrent. Since electrolytic condensers are inadmissible owing to the internal losses, an alternative is to be found in an electrochemical polariser, of the type known in the industry as Daniells cell. In this particular application, current is fed at a constant rateto the cell for the duration of time that the valve motor VM is running, in one direction or the other according to whether contactor I55 or I I5 has operated. Passage of current throu h the cell builds up an increasing ing, while subsequent reversal of this current will first gradually depolarise the. cell in the same time that it took to polarise and then build up an increasing E. M. F. in the opposite sense.

Referring now to Fig. 5, the electro-chemical cell Ill with its electrodes I92 and I93 porous diaphragm I04 in a cuprous salt solution is connected in the grid circuit of a steep-slope pentode amplifying valve I90 which together with anode resistance I 95 and couplin .condenser I applies an amplified E. M. F. to the output valve I50, which is, as hereinbefore described, coupled to anode resistance I33 and to the relay I50. Contactor I65 or I15, as it closes,

applies a small current in one direction or the other to the circuit of the electro-chemical cell via the resistance I05 mounted on wheel I, which, as before, is loosely coupled to the valve indicator. Resistance I05 differs from the potentiometer 5 in that it is designed to compensate for non-linearity of the compensating means by passing a greater current at the more critical positions of the valve, instead of applying a greater voltage. The mean current may be of the order of say 100 microamps, depending on (a) the size of electrochemical cell, (b) the speed of the valve and (c) the length of the time lagin the plant. The E. M. F. enerated by ionisation orde-ionisation in the'cell is then applied to-the grid of the valve I90. to render the latter more negative or less negative, according plifled voltage on grid I3I governs the output of valve I30. I

As shown in Fig. 5, the only leakage across the cell tending to equalise the potentials on the two electrodes is the internal grid to cathode resistance of the valve I90. This is of a very high order, so that the decrement in grid voltage is very low, being thus suited to long time lags. If it should be desired to use the same size of electro-chemical cell for shorter time lags of, say. one-half of the previous duration, a parallel resistance equal to the internal grid cathode resistance of the valve is inserted, and corresponding values for all other time lags.

A diagrammatic representation of the extra arrangements necessary for the extension of the basic form of the invention, .whereby an exponential factor is introduced, is given in Fig. 6. This law, which can now be conveniently repraented only by use of two equations is;

, to the motion of the steam valve, and the am- 12 where x is defined by:

c, A and s being chosen constants.

In Fig. 6 of the accompanying drawings 00 represents either the temperature indicating arm itself or an arm so driven that it amplifles'the movements of the temperature indicating arm. An example of such amplification has already been given with reference to Fig. 2. At the extremity of 80 a contact 0| bears on the resistance 02 of a potentiometer system, thus imparting to plate R of condenser 03, changes of potential sensibly proportional to changes in 9. Plate L is connected through resistance 84 and condenser 05 to a constant potential and is also connected to an indicator of potential 05, the indicating arm of which may be designated .Ml, the movements of which are utilised in the same manner of those of arm M1, in Figs. 1 and 2. Indicator 86 (which is indicating x by a displacement W) may be of the same form as indicator I0, details of which have already been described with reference to Fig. 2.

In this embodiment of the invention let a: be the potential applied at the end of arm 00 (proportion to 9, the error of the variable) and let w be the potential applied to the indicator 05. The potential at the grounded end may be regarded as zero.

Calling capacities C, currents i, then resistances R and dw 1 1 1 da: 1 l wfinicfinfl -wri in? so that the deflection x of the indicator 86, which is proportional to W, obeys a law of the form dx (l? xc 6+B The physical, as apart from the mathematical,

explanation of this behaviour is that on the oo- '7 by the relative sizes of condenser). In

l3 analogous manner, therefore, a linearly rising characteristic of the arm ll would cause an initially parallel rise of potential at indicator 86, followed by an exponential drift of the potential to some other linear rise of lesser gradient. By suitable choice of the-voltage across 82 and the lengths of arms 80 and M1, the net eiIect of such additional device is to give an exaggerated behaviour or the variable to indicator 86.

The object of exponentially exaggerating the variations of the indicator arm 80 is to produce momentarily a. compensating elfect which is greater than that required to neutralize the disturbance to which the variations of the indicator are due, thus restoring more quickly the desired value or the quantity in cases where sudden variations of the same occur.

In cases where there is no direct current supply available but only alternating current, the systems as shown in Figs. 2, 3, 4 and 5 may be modified to have the filament heater supplies of individual valves connected in parallel to a transformer winding of 4-volt secondary output, while the high tension direct current is supplied from a rectifier of thermionic valve or copper oxide type; substantially as employed in alternating current radio receivers. The valve motor may be single or S-phase.

Itwill be understood that the systems and modifications illustrated in the drawings may be applied to the control of various mechanical, electrical, physical, chemical or other characteristics although for the sake of simplicity they have be p described in connection with the control oi'temperature. The specific systems and modifications shown and described are intended as examples illustrative of the principles of the invention, which is not limited to the specific disclosure.

Various modifications of the systems previously described may be made in carrying out our invention, and all such modifications are intended to come within the scope of the appended claims insofar as they achieve to a useful degree the new results, improvements and advantages hereinbefore described.

To demonstrate mathematically the action which occurs in the electrical systems hereinbefore described, the following is submitted:

As was stated hereinbefore the leakage rates through the resistances 8, 9 and II, as is well known, are governed entirely by the potential across them. How they fit into the system as a whole will now explained with reference to the diagrammatic view, Fig. 1. The values of, the respective resistances are indicated by Ra, Ru, and R12.

Let the protential at point P at any given moment be designated by 0. Then the current into condenser II, will be a function of the capacity Cu and of the voltage 0 and may be indicated as wherefore the potential E0 at the Junction common to the resistances 8, 9 and I2, Fig. 1,-will be The potential E at the point Q is fixed, and for purposes of calculation may be assumed-as oases plate of condenser I, Fig. 1.

zero, whereupon the current through the resistance l2 will be determined. It is: p

ROCHE n Let Er designate the potential at the right hand into the condenser! will be (Eli r)/ 8 n n r (3) From the two Formulae 2 and 3 there can be deduced the B-term equation hereinbefore set forth, namely:

The steps by which the deduction is made will now be indicated.

Equating (2) and (3) where EX is the potential at x.

Using (2), (4) and (A) 2 E TIUBn-i- Re+ l Ream a.) amongf]:

l 12( s+ R9) Ramayan -2f] Now Ex is proportional to V, the valuej'of the compensating effect, so that Equation 5 generalised in the form (iv (19 (1 9 'iilg+ fai+ i and integrating The current through the resistance I is -(cur- I rent through I 2 current through 0).that is; i t j 'Then the current is where The values of resistances I, 9 and I! in combination with the sizes of condensers I and ll,

govern the time-rate of change of potential at the point P and the values of the constants in the main equation.

In the simple two-term case (e. g., Fig. 5) the rate of leakage of current through the resistance (192193) governs the re-setting rate of the steam (or other) valve movement, and hence in conjunction with the capacity of the cell l9-l fixes the ratio of re-setting to factors hi and 762.

We claim:

1. A system for the automatic control of a variable characteristic comprising responsive proportiona means proportionally responsive to deviations of the characteristic from a desired value, compensating means for adjusting the value of the charand a resistor connected in said circuit between,

said potentiometer and a source of potential to introduce upon change of voltage applied to said condenser'and said resistor as a result of movement of said potentiometer said corrective action in accordance with said sum.

2. A system as set forth in claim 1 in which the electrical operating means comprises a static electrical system including said condenser and said resistor, said potentiometer applying to the sys tem a varying potential proportional to the variations in the compensating efl'ect, potential-responsive means for operating said compensating means, and static electrical means for regulating the operation of said potential-responsive means for transmitting said applied changes in potential thereto in conformity with said extent and said summation of a deviation of said physical characteristic.

3. A system as set forth in claim 1 in which there is included a difi'erentiating circuit comprising a condenser and a resistor whereby said operation of said compensating means is additionally controlled in conformity with the rate of such deviation.

4. A system for the automatic control of a variable physical characteristic comprising an indicator of the value of said physical characteristic; compensating means for restoring the desired value of the physical characteristic following any departure therefrom; a reversible servomotor actuating said compensating means; a relay system controlling operation 01' the servomotor, the

relay tem being responsive to variations of.

and a resistor for the controlled transmission of changes in said electrical characteristic from said point to said responsive means for gradually restoring said-electrical characteristic at said responsive means to a predetermined value.

5. A system asset forth in claim 4 including means providing a correction for backlash in the action of the compensating means, said correction means including a lost motion drive connecducing means.

'6. A system for the automatic control of a variable physical characteristic comprising an indicator of the value of said physical characteristic; compensating meam for restoring the desired value of the physical characteristic following any departure therefrom; a reversible servomotor actuating said compensating means; a relay system controlling the operation of the servomotor. the relay system being responsive to variations of the indicator; means for producing at a given point changes of electrical potential corresponding to changes in the compensating effect 'of the compensating means; potential-responsive means for controlling the relay system; and a static electrical system including a circuit connecting said point and the potential-responsive means provided with a condenser and a resistor interposed tion between the servomotor and the change-probetween said point andthe responsive means for the transmission of changes in potential from said point to said responsive means for gradually restoring the potential at said responsive meansto a predetermined value.

7. A system as set forth in claim 6 in which the static electrical system including said condenser has one-plate thereof directly connected to said point and the other plate connected through a resistance to a point of predetermined potential co to the desired value of the physical characteristic to be controlled, said other plate also being connected to said potential-responsive means.

t 8. A system as set forth in claim 6 in which the static electrical system including said condenser has one plate thereof directly connected to said point and the other plate connected through two in series to a point of predetermined potential corresponding to the desired value of the physical characteristic to be controlled a point between said two resistances being connected through aresistance to said potential-responsive means and thence through a second condenser to a point of constant potential.

9. A system as set forth in claim 6 comprising I for causing variations of the indicator to produce proportional changes -of electrical potential at a selected point in a second static electrical system comprisin a condenser, one plate of which .is directly connected to said selected point and the other plate of which is connected through a resistance and a condenser in series to a oint of constant potential, said other plate. alsobeing connected directly to a potential-responsive memher to control the relay system.

10. A system as set forth inrlaim 6 including 7 17' means for applying to the control grid of the second valve a potential proportional to the variations of the indicator, and current-operated reversing switch means actuated by variations in relative current strength of the output of the valves and controlling the operation of the servomotor.

11. A system as set forth in claim 6 in which the physical characteristic is temperature and the indicator is a thermocouple, said system including thermionic valve means for amplifying the voltage produced by the thermocouple, and means for actuating the relay system in conformity with the proportional relationship between said voltage and the potential applied to the potential-responsive means.

12. A system for the automatic control of a variable physical characteristic comprising an indicator of the value of the characteristic to be controlled, compensating means for varying said characteristic, a reversible servomotor for actuating the compensating means, and a relay system controlling the operation of the servomotor and responsive to variations of the indicator from a normal position corresponding to the desired value of the characteristic; a reversible electrochemical cell; means for passing through said cell a quantity of electricity proportional to variations in compensating efiect caused by actuation of the compensating means; and means for controlling the relay system responsive to the potential of the electrochemical cell.

13. A system as set forth in claim 12 in which one pole of the electrochemical cell is connected to a point of constant potential and the other pole is connected to the control grid of a thermionic valve, the output of which is applied to a circuit associated with the potential-responsive means for controlling the relay system.

14. A system as set forth in claim 6 in which the potential-responsive means includes a thermionic valve whose control grid is connected to the static electrical system and a current-responsive member actuated by the output of the valve and associated with the indicator; and means actuated by variations from predetermined proportionality between the movement of the indicator and the current-responsive membar is provided for actuating the relay system to restore said proportionality.

15. A system for the automatic control of a variable characteristic comprising responsive means proportionally responsive to deviations of the characteristic from a desired value, compensating means for adjusting the value of the characteristic, and electrical means having one circuit element associated with and actuated. by responsive variations in said responsive means, and other circuit elements including'a resistor and a condenser for operating the compensating means in conformity with the extent of the deviation, and with the summation of the deviation over the period during whichsaid deviation occurs.

16. A system for the automatic control of a variable characteristic comprising means proportionally responsive to deviations of the characteristic from a desired value, compensating means for adjusting the value of the characteristic, electrical means having one circuit element associated with and actuated by said compensating means and other circuit elements including a resistor and a condenser connected in circuit with said first-named element for modifying the potential output therefrom in conformity with the cases sum .of the extent of summation of the deviation over the period during which said deviation occurs, and actuating means for said compensating means operable under the control of said electrical means.

' 1'7. A system for the automatic control of a variable characteristic comprising compensating means for adjusting the value of said characteristic, actuating means for said compensating means, means comprising an electrical network having an output circuit for controlling said actuating means, said network including a circuit varying element associated with and actuated by said compensating means for varying the output of said output circuit to control said actuating means in conformity with the extent of deviation of said characteristic from a desired value, said network also including an adjustable circuit element associated with said output circuit and adjustable for selection of said desired value of said characteristic, responsive means proportionally responsive to deviations of said characteristic from said desired value for initiating operation of said actuating means, and a resistor and a condenser connected to said output circuit for modifying the output thereof to control said actuating means in conformity with the summation of said deviation over the period during which said deviation occurs.

18. A system for the automatic control of a variable characteristic comprising compensating means for adjusting the value of said characteristic, actuating means for said compensating means, means comprising an electrical network having an output circuit for controlling said actuating means, said network including a first adjustable circuit element-and a second adjustable circuit element associated with and actuated by said compensating means, responsive means proportionally responsive to deviations of said characteristic from a desired value for adjusting said first circuit element, said network including a third adjustable circuit element for selection of said desired value of said variable characteristic, said output circuit including circuit connections leading to said first and second adjustable circuit elements, and a resistor and a condenser connected in one of said last-named connections for modifying the output thereof to control said actuating means in conformity with the summation of said deviation over the period during which said deviation occurs.

19. A system for the automatic control of a variable characteristic comprising a compensating means for adjusting the value ofsaid characteristic, a network having multiple branches, one of which includes an adjustable circuit element for development in the network of a first voltage of magnitude and polarity depending upon the extent and sense of deviation of said variable characteristic from a desired value, a second branch of said network having included therein a second adjustable circuit element for producing in said second branch of said network a voltage of magnitude and polarity dependent upon the position of said compensating means, said network having an output circuit, actuating means connected to said output circuit responsive to the difference between said first voltage and said second voltage for adjusting said compensating means in direction and to extent to maintain equality between said first voltage and said second voltage, and condenser means and resistor means respectively connected in said network for 7 modifying the voltage applied to said output cirthe deviation, and of the cuit for adjustment or said compensating means by said actuating means in accordance with the summation 01' said deviation from said desired value with respect to time.

20. A system for the automatic control of a variable characteristic comprising a compensating means for adjusting the value of said characteristic, an electrical network having multiple branches, one of which includes an adjustable circuit element for development in said network of a first voltage 01 magnitude and polarity depending upon the extent and sense 01' deviation oi said variable characteristic from a desired value, a second branch of said network having included therein a second adjustable circuit element for producing in said second branch of said network a voltage of magnitude and polarity dependent upon the position of said compensating means, said network having an output circuit, actuating means connected to said output circuit responsive to the difi'erence between said first voltage and said second voltage for adjusting said compensating means in direction and to extent to maintain equality between said first voltage and said second voltage, said network including a third branch having resistors therein in parallel with the branch including said second adjustable circuit element, and a condenser and a resistor interconnecting a point on said third branch and said second circuit element, said output circuit being connected to a point between said resistor and said condenser, and also being connected to a branch of said network, the voltage across which branch is under the control of said first adjustable circuit element.

21. The combination set forth in claim 20 in which a second condenser and a second resistor are connected to said output circuit for so modifying the output voltage as to control said actuating means for adjustment of said compensating means in accordance with the rate 01 departure 01' said characteristic from said predetermined value.

22. The combination set forth in claim 21 in which said second resistor is connected between the juncture of said first resistor and said first condenser and one side or said output circuit, and said second condenser is connected from said output circuit to the branch of said network including said first adjustable circuit-controlling element.

23. The combination set forth in claim 20 in which said condenser and said resistor are connected in series in that part of the output circuit from which there is derived the voltage from said second branch of said network, a circuit including a second condenser effectively connected in parallel across said output circuit and including in series with said second condenser a resistor for attenuating said output upon change of voltage to control said actuating means to adjust said compensating means in accordance with vthe rate of departure of said characteristic from said predetermined value. I

ALBERT CAI-LENDER. LEEDS AND NORTHRUP COMPANY, Assignee of Allan Brown Stevenson, Deceased,

By I. MELVILLE STEIN, 1

Vice President.

REFERENCES CITED The following references are of record in the

Images