Search Images Maps Play YouTube News Gmail Drive More »
Sign in
Screen reader users: click this link for accessible mode. Accessible mode has the same essential features but works better with your reader.


  1. Advanced Patent Search
Publication numberUS3754154 A
Publication typeGrant
Publication dateAug 21, 1973
Filing dateFeb 8, 1971
Priority dateFeb 8, 1971
Publication numberUS 3754154 A, US 3754154A, US-A-3754154, US3754154 A, US3754154A
InventorsMassie P
Original AssigneeMassie P
Export CitationBiBTeX, EndNote, RefMan
External Links: USPTO, USPTO Assignment, Espacenet
Sealed pump and drive therefor
US 3754154 A
A completely sealed magnetically driven pump havng a piston-armature driven by electrical windings. Unique electrical driving circuits are provided for the pump embodying feed-back windings magnetically coupled with the driving windings of the pump for controlling the reciprocation. Mechanical variations are provided for variable displacement, high pressure and valve shearing action characteristics. The device is also reversible in that the piston-armature can be reciprocated mechanically whereby to generate pulsating electromotive force in the electrical windings.
Previous page
Next page
Description  (OCR text may contain errors)

United States Patent 1 Massie Aug. 21, 1973 SEALED PUMP AND DRIVE THEREFOR [76] inventor: Philip E. Massie, 4220 Irving P1.,

Culver City, Calif. 90230 [52] US. Cl. 310/30, 310/34, 310/17, 417/418 [51] Int. Cl. 110211 333/00 [58] Field 01 Search 318/128, 138, 130; 310/34, 14, 22, 30, 154; 417/418 [56] References Cited UNITED STATES PATENTS 3,022,450 2/1962 Chase 317/123 2,579,723 12/1951 Best 251/65 3,178,151 4/1965 Caldwell 335/229 X 2,749,453 6/1956 Cassell et 31.... 335/229 2,832,919 4/1958 Reutter 417/418 2,988,264 6/1961 Reutter.... 417/418 2,721,453 10/1955 Reutter.... 417/418 2,701,331 2/1955 Holst 318/128 1,912,167 5/1933 Anderson 318/123 2,892,140 6/1959 Praeg 318/128 2,447,230 8/1948 Brown 310/31 2,686,280 8/1954 318/128 3,328,659 6/1967 318/138 3,293,516 12/1966 318/128 2,690,128 9/1954 Basilewsky 318/122 FOREIGN PATENTS OR APPLICATIONS 728,771 2/1966 Canada 318/130 Primary Examiner-G. R. Simmons Attorney-Herzig & Walsh [5 7] ABSTRACT A completely sealed magnetically driven pump havng a piston-armature driven by electrical windings. Unique electrical driving circuits are provided for the pump embodying feed-back windings magnetically coupled with the driving windings of the pump for controlling the reciprocation. Mechanical variations are provided for variable displacement, high pressure and valve shearing action characteristics. The device is also reversible in that the piston-armature can be reciprocated mechanically whereby to generate pulsating electromotive force in the electrical windings.

1 5 Claims, 29 Drawing Figures Patehted Aug. 21, 1973 5 Sheets-Sheet 2 INVENTOR. pfl/Z/I E. MASS E drroe/ves' Patented Aug. 21, 1973 3,754,154

5 Sheets-Sheet 4 INVENTOR. pH/z/P E. MASS/E 1 SEALED PUMP AND DRIVE THEREFOR SUMMARY OF THE INVENTION The need for a sealed pump is emphasized by the many forms of flexible tube roller pumps and vibrating diaphram pumps on the market. The limitations of crank driven piston pumps are well known and include: seal problems around rotating shafts, bearings, mass balance problems, belt or coupling means and the necessity of including a driving motor of some form with the attendant efficiency reduction and maintenance requirements. The threee step conversion of energy in a motor driven piston pump, electricity to rotating pulley to belt to reciprocating piston, shows the large number of opportunities for energy losses, maintenance requiements and production cost. Associated with the device is the possibility of escape of noxious, corrosive or lethal materials through the fenestrations of the pump house for shafts and controls.

The flexible tube and roller pump and the diaphram pump minimize the leakage problem up to the point where the flexible material fails from fatigue or overpressure.

The herein described pump minirnizes or eliminates most of these problems. The objects of the invention reside in eliminating these problems and in realizing certain advantages.

The herein pum'p offers the following salient and advantageous features.

There is no penetration or fenestration of the pump body.

There is no rotating shaft, crank or roller to drive the pump.

There is no external mechanical drive of any type.

The pump has no elastomeric materials as pumping elements.

The pump is a positive displacement piston pump, of doubleacting type. a

The pump is capable of being assembled, welded shut and helium leak tested for integrity, if this is desirable.

The pump can be made in any reasonable size, from sub-fractional horse-power equivalent to multikilowatt equivalent.

The basic pump will operate to a pressure of about 300 psi and is self limiting at some. modest overpressure above that.

The high pressure version of the pump can be operated at any pressure within the limits of the materials from which the pump is fabricated.

The pump can be made pressure limiting by design of the input power circuit and the pressure limit can be varied by change of the input voltage to a given pump.

The pump can be made of stainless steel, 300 series non-magnetic and 400 series ferro-magnetic, thus lending its use to highlyv corrosive materials. The basic pump cylinder (tube) can be made of glass so that the pumping action or pumped material is visible.

The pump can be made to operate on a single pulse" basis with no power consumed between pulses. Thus, two or more pumps, operated by suitable proportioning pulse logic circuits can be made into a proportioning assembly.

The basic pump is direct current operated but suitable operating controls, described below, make the pump usable on alternating current sources. The pump, in one specific mechanical size and configuration, can be designed for a broad variety of operating voltages by a change in the pump coil design. Since these coils are bobbin wound" and separate parts, it is entirely practical to dis-assemble the pump, SLIP off the coils and replace them with coils designed for a different operating voltage. Control circuitry can'be modified for the new operating voltage.

The pump in its basic form has many applications and adaptations, the range of which can be very greatly increased by way of providing for the creation of higher pressures and for variable displacement. In one form of the invention, the piston and the cylinder are constructed to have portions of two diameters, the arrangement of magnetic poles relative to the piston being such as to produce desired high pressures, the realization of this purpose being an object of the invention.

In modifications of the basic form of the pump, constructions are provided by way of an adjustable head, adjustable relative to the pumping cylinder to provide for variable displacement. The realization of this variable displacement capability is another object of the invention.

Another object is to realize the aforesaid variable displacement capability by way of the use of an additional conventional fluid operated cylinder operating bypass valves to thereby control displacement of the pump.

Another object is to provide a pump as referred to provided with valves which incorporate shearing action for handling certain types of fluids or slurries.

A basic characteristic and advantage of the pump as referred to is that operation requires merely control of the windings. Another object of the invention is to provide automatic magnetically actuated switching or contact means associated with the magnetic fields of the windings for operating the pump. The windings of the pump are adapted for control from different types of electrical power or feed circuitry and embraced within the invention are concepts of controlling and/or driving the pump from particular electrical circuits which com bine in them certain windings associated with the pump itself which form feedback windings magnetically associated with the drive windings of the pump. in one such form of the invention, the driving circuit is a modified Eccles-Jordan circuit embodying feedback windings which are in the form of windings magnetically coupled to the drive windings of the pump. A further object of this invention is to realize this type of electrical drive capability of the pump.

Another object of the invention is to take advantage of the peculiar capabilities of that type of circuitry known as SSRs in the drive of the pump. In one form of the invention, the pump is driven by circuitry constituted by dual SSRs embodying feedback windings which are again windings coupled to drive windings of the pump. The object is to gain the advantages of this type of circuitry in an unique way by coupling the feedback windings to the drive windings of the pump.

From the foregoing, it will be understood and appreciated that a primary adaptation of the invention is in a sealed pump as described. However, it is reversible and in such adaptation, the device constitutes an electrical generator rather than a pump, the electromotive force being generated in the windings by virtue of linear movement of the armature (piston). The armature may be reciprocated by any type of mechanical means. In the form of the invention as disclosed in detail herein, the armature may be driven by an internal combustion engine type of drive means which reciprocates the armature by means of pressure. Therefore, the operation is the reverse of the device as an electrically driven pump.

A further object of the invention is to realize a device of this type having the reversible capability as described, that is, it can be designed as an electrically driven pump or on the other hand, as a mechanically driven electrical generator.

BRIEF DESCRIPTION OF THE DRAWINGS Further objects and additional advantages of the invention will become apparent from the following detailed description and annexed drawings wherein:

FIG. 1 is a cross-sectional view of a basic form of the FIG. 2 is a schematic view illustrating magnetic switching control of the windings of the pumps;

FIG. 2a is a schematic view of the circuitry involved in FIG. 2;

FIGS. 2b and 2c are views of the arrangement of FIG.

FIG. 3 is a view of a manually actuable switch for controlling the pump;

FIG. 4 is a graph of magnetization curves associated with the pump of FIG. 1;

FIG. 4a is a graph of a demagnetization curve of a permanent magnet;

FIG. 5 is a circuit diagram of a modified multivibrator circuit with feedback windings;

FIG. 5a is a circuit diagram of a component of the circuit of FIG. 6;

FIG. 6 is a circuit diagram of the dual SSR control circuitry for the pump windings;

FIG. 7 is a cross-sectional view of the pump modified to produce high pressures;

FIG. 8. is a binary logic driving circuit;

FIG. 9 is a cross-sectional view of a modified form of the pump utilizing valves having a shearing type action;

FIG. 10 is a sectional view taken along the line 10-10 of FIG. 9;

FIG. 10a is a similar view showing the valves of FIGS. 9 and 10 in another position;

FIGS. ll, 12 and 14 show views of a modified form of the pump having variable displacement;

FIG. 13 is a view of a preferred sealing arrangement;

FIG. 15 is a sectional view of a modified form of pump having variable displacement;

FIG. 16 is a sectional view of another modified form of pump having variable displacement;

FIGS. 17, 18, I9 and 20 are cross-sectional views of three-way valves associated with the variable displacement pumps;

FIG. 21 is a sectional view of the device constructed to operate in reverse and to operate as an electrical generator;

FIGS. 22 and 22a are views of current rectifiers associated wih the device of FIG. 21.

The pump, in the basic form, is shown in FIG. I. The mechanical parts of the pump consist of: a piston/armature l of ferro-magnetic material, a cylinder 2 of mm magnetic material, two cylinder head/pole pieces 3 and 4 of ferro magnetic material, a magnetic core (back iron) material 6 that magnetically connects the two pole pieces, a permanent magnet 7 properly interposed to couple the armature/piston to the magnetic core, suitable seals between the cylinder and the cylinder head/pole pieces 5. Assembly means may take any desirable configuration, such as tension rods to cross bars on the outside of the pole pieces to pull the assembly together or screw threads between the cylinder and the cylinder head, including tapered pipe threads, screwed together with plumbers pipe dope. The pole pieces and the back iron are not necessarily one part, preferably not, for reasons to be explained below.

The electrical portion of the basic pump consists of the coils 11 and 12 that drive the pump. These coils are actually wound around the cylinder and preferably cover the air gaps l5 and 16 that constitute the pump displacement. The coils l3 and 14 are not power coils and their function will be explained later.

The cylinder head/pole pieces of the pump contain inlet and outlet ports 8 and 9 to admit and release the pumped material. Suitable valves of any type; reed, plug, flapper or ball valves are applicable. The valves are spring loaded pressure operated in response to pressure differentials across the valves. The valves are of non-magnetic material to minimize sticking due to magnetic forces. (There may be a state of operation where ferro-magnetic valves are desirable, particularly the inlet valve 9).

The pump operates in the following manner. The permanent magnet supplies a magnetic flux to the pistonlarmature 1 through the non-magnetic cylinder wall 2. Two paths exist for the magnetic flux through the two pole pieces 3 and 4. The magnetic flux divides between the two paths 17 and 18. The division of the flux between 17 and 18 is a function of the relative permeance of each circuit. Permeance is that function that relates to the ease with which magnetic flux passes through a circuit under the influence of a given magnetomotive force. P /F. (Units used in magnetic design are strange and wonderful to behold. There are three different systems of units, two of which are in common use in the United States. An explanation of these two is in order here.

A. Mixed English units are based on the inch system.

I. Flux is expressed in lines per square inch or maxwells.

2. Magnetomotive force is expressed in ampereturns per inch.

3. The premeability of a vacuum is 3.192 in this system.

B. C.G.S. units are metric based.

1. Flux is expressed in gausses.

2. Magnetomotive force is expressed in oersteds.

3. The permeability of a vacuum is 1. The ratio between the two systems is conversion from inches to centimeters, with exceptions. Mixed English units will be used herein.)

FIG. 4 shows the magnetization curve. The ordinate is flux in lines (maxwells) for the particular circuit. Kilo-lines is the generally used term because of the large numerical values. The abscissa is magnetomotive force in ampere-tums. The magnetization curve for a specific magnetic circuit is developed in parts. One half the ferro-magnetic material (one side of the pump) is represented by the curve 30 showing the high permeability of the ferro-magnetic material up to the knee of the curve, called saturation. The magnetization curve of an air gap is a straight line. The slope of the line is proportional to the area divided by the length of the air gap. Line 31 represents a short air gap (closed gap) such as 16 in FIG. 1, with a relatively high permeance. Line 32 represent a long air gap (open gap) such as 15 in FIG. 1 with a relatively low penneance. The total magnetization curve for the two sides of the pump are represented by line 33 for the closed gap plus the iron and line 34 for the open gap plus the iron. These curves are developed by summing the magnetomotive force for a given flux quantity. 1

Line 39 represents the external magnetomotive force of the permanent magnet under a specific condition. FIG. 4a shows the internal demagnetization curves of a permanent magnet of a specific size. The permeance of the total external magnetic circuit is represented by the line 65. The intersection of this line with the demagnetization curve 275 defines the MMF 276 and flux 84 the permanent magnet will develop in the external circuit. The MMF 276 of FIG. 4a is the inverse of the MMF 39 of FIG. 4. The flux 38 in the closed gap circuit is found at the intersection 82 of lines 9 and 33 (FIG. 4.). The flux 37 in the open air gap circuit is found at the intersection 83 of lines 39 and 34 of FIG. 4. The sum of the two fluxes 37 and 38 of FIG. 4 is equal to the flux 84 in FIG. 4a. It can be seen that there is a large difference between the flux in the closed air gap gap circuit of FIG. 1 and the open air gap circuit 17. The

mechanical force in pounds in a air gap is equal to */72A, where d is in kilo-lines and A is the effective gap in square inches, allowing for air gap fringing, which increases the area. The net force on the piston is the difierence between the two force values. Thus, it is apparent that a large differential force exists to hold the piston in the closed gap position, since most of the flux flows in that gap. No external power is required to hold the piston in this position. This covers the static or starting position of the pump.

With respect to the operation of the pump, if a voltage is applied to coil 1 1 in such a direction as to aid (increase) the flux in the open gap, flux will be diverted to that gap 15 of FIG. 1. This will tend to pass more flux through the permanent magnet and reduce the magnetomotive force of the magnet along the line 275 of FIG. 4a. The flux of the permanentmagnet moves from line 84 to line 86 as the magnetomotive force moves from line 276 to line 85. The effect of the decrease in pennanent magnet MMF is to decrease the flux in air gap 16, curve 33 of FIG. 4. The addition of an electric MMF is to impress a high MMF on, the open air gap 15. This produces an increase in flux along line 34 of FIG. 4 while the flux in the closed gap decreases along the line 33 of FIG. 4 to a level where the greater amount of flux is in the open air gap. At this point, the balance of force is shifted in accordance with the force equation and the piston starts to move to close the gap 15.

With the constant value of electric MMF the flux in the "open gap 15 increases with movement of the piston 1. At the same time, the flux in the closed gap decreases. This produces an increasing force to move the piston to close gap 15, as the gap closes. The movement of the piston reduces the volume in gap 15 and displaces any material out through outlet port 8. The

pump has completedone pumping stroke. As gap 15 decreases in volume, gap 16 increases in volume, drawing material in for the next pumping stroke.

If power is now removed from coil 11 and applied to coil 12, the same action takes place and the piston displaces the material drawn into gap 16 and draws more material into gap 15, ready for the next pump stroke. The pump has now -completed one cycle, two pumping strokes.

Alternate application of a voltage to coils l1 and 12 will cause the piston to move from one pole piece 3 to the other 4, in the process, alternately increasing and decreasing the volume of the air gaps, 15 and 16, thus drawing material in through inlet port 9 and expelling material through outlet port 8, controlled by the logic of normalspring loaded valves. To establish means of generating the voltage on alternate coils, 11 and 12, a 60 hertz, constant volume pump may be used; it is possible to design the magnetic and electric circuits to operate on the positive and negative half cycles of a 60 cycle power source. Coils 11 and 12 are connected so that current through the corresponding diodes will generate a flux to aid the flux, 17 or 18, in the corresponding coil. This is a highly limited application. The utility of this type of operation can be extended by using a variable frequency AC power source, such as a transistor inverter, for driving the unit.

If single shot, manual operation is desired, a manual, spring loaded, single pole, double throw switch will suffice, as in FIG. 3. The switch is spring loaded in the center position with neither position connected. The

' switch is manually pushed from one contact to the other and the pump will respond with one stroke per switch contact. This might be a desirable method for laboratory test application where precise pressure or air quantities are desired.

A method of automatic lever switch operation is shown in FIG. 2. Spring 20 is a ferro-magnetic and con ductive leaf spring carrying back side contact 27 and front side contact 21. A similar spring is installed in relation to air gap 16. Referring to the magnetization curves of FIG. 4, curve 30 is the magnetization curve of the iron only. Curve 32 is the magnetization curve of the open air gap. The two combine into curve 34. At the flux level, 37, it can be seen that most of the magnetomotive force is across the air gap, 16. Now with the reed switch 20 close to piston 1 (separated only by the non-magnetic cylinder 2) as compared to the spacing between armature l and pole face 3 (air gap 16) and with a similar magneti material 19 supporting a contact 22, in close proximity to pole face 3, it is seen that a high MMF will be impressd across the magnetic circuit 1, 20', 19', 3; and this will produce a corresponding high flux in the air gap between 19 and 20, diverting flux from air gap 16. An equation can be used to design a sufficient force to move reed spring 20', open contacts 27' and 28', and close contacts 21' and 22'. From FIG. 2a, it is seen that this will place power. on coil 11 through leads 24 and 29 and contacts 28 and 27, moving the piston to close gap 16. As gap 15 opens, the magnetization curve shifts along line 39 (with deviations) to the intersection with line 34, which is now the curve for gap 15, the open gap. At this point, contacts 27 and 28 are opened, removing power from lead 23a and coil 12. Contacts 21 and 22 are closed by the large MMF across gap 15. As gap 16 is closed, the MMF moves to the intersection of line 39, the permanent magnet MMF, and line 33, closed gap. Since the MMF across gap 16 is now small and with appreciable air gaps in the reed switch magnetic circuit at the two penetrations of the tube wall 2, reed switch 20' will move and close contacts 27' and 28' and complete the power circuit to lead 23 and coil 11. This will move the piston back to close gap 16 and the cycle will continue.

During the interval that contacts 21 and 22 and 21 and 22 are both closed with gap 16 closing, the change of flux in coil 12 due to movement of the piston will be such as to produce a current in coil 11 that reduces the flux in gap 15. This will keep the MMF across gap 15 reduced but not low until the piston stops moving and there is no further change of flux and corresponding induced voltage in coil 12.

The mounting rings 19b, 19b, 20b and 20b support the various contacts and increase the area of the air gap 19c, 20c between the magnetic portions. of the reed switch parts and the pole pieces and the piston/armature. This reduces the reluctance (increases the penneance) of the air gaps and allows more magnetic flux to flow through the reed switch magnetic circuit'at any point in the operation. This is typical and not the only method of support of the reed switch contacts. The coils 11, 12, 13 and 14 can be placed at any point on the iron circuit (around the loop) as long as the coils enclose all the circuit cross section. The preferred location is along some portion of the pole piece/piston region to minimize leakage flux.

Permanent magnet material is hard and difficult to cut or machine. Since cast bars are more efficient than similar material pressed and sintered from powder, it is desirable to use flat-sided magnet sections. This does not adapt well to the round configuration of the center of the pump.

Further, the energy obtainable from a magnet is a function of cross-section area and length in relation to the external magnetic circuit. The external magnetic circuit has a permeance characteristic of the crosssection and the length of the various elements. It is desired to operate the external magnetic circuit at a given flux (flux density times area at the critical point, in this case, the working air gaps l and 16). The required cross-section area of the permanent magnet is selected to provide the desired flux when operated at the optimum point. The length of the magnet is selected to provide the required MMF to force the flux through the external circuit, again operating at the optimum point on the curve 275, FIG. 4a. It is now apparent that the dimensions of the permanent magnet may be entirely independent of all of the other parameters of the magnetic system, i.e., coil cross-section, central tube diameter, etc.

For the above reasons, it is often desirable to provide coupling" of the permanent magnet to the iron circuit by a shoe 270 as shown in FIGS. 2b and 2c, the permanent magnet being designated 7. This is usually a material more adapted to machining operations and much cheaper than permanent magnet material. The designer is now free to design his permanent magnet in any desired configuration. It may be one or more magnets in parallel, as shown in FIG. 2b, or it may be one magnet only. The pump design may be such as to have only one external leg and thus a position for only one permanent magnet. The core is designated at 6 and 6'. The permanent magnet should be close to the iron in all cases. The use or omission of a shoe is optional with the design.

DRIVE CIRCUITS FIG. 5 shows the operation of the pump on a modi- I and 54/56 are designed to turn on switches 51 and 52 alternately. The load on the switches consists of the pump coils, 11 and 12. The rest of the resistors 57, 57, 58, 58. and 59 in the circuit are typical of this configuration of multivibrator (flip-flop). The circuit of FIG. 5 will be considered in some detail in relation to its feedback implications.

When switch 51 is ON resistance 55 and capacitor 53 are designed so that sufficient base current is supplied to switch 51 through the capacitor but not through the resistor. Switch 52 is OFF and the collector is at a high potential since no current flows through the circuit. Capacitor 53 charges from the high potential at a time constant determined by resistor 55, the base bias resistors of switch 51 and the resistance of the base to collector junction of switch 51. When capacitor 53 is charged to the potential of the collector of switch 52, no further current flows through the capacitor. This reduces the current through the base of switch 51 and correspondingly reduces the current through switch 51. This reduction in current in switch 51 causes a rise in potential of the collector of switch 51. This rise in potential causes the capacitor 54 to start charging and supplying current to the base of switch 52. This in turn causes current to start flowing in switch 52 and reduces the collector voltage, further reducing the base current to switch 51. This feedback continues until switch 51 turns off and switch 52 is on. Since capacitor 53 was charged to a high potential prior to the switch and cannot rapidly discharge, when the collector of switch 52 drops in potential, the base of switch 51 is driven hard off (positive as shown) and fully cuts off switch 51. The cycle now repeats in the opposite direction.

It will be noted that the loads on the two switches are the operating coils, 11 and 12, of the pump. Thus, the coils are powered alternately and cause the pump to operate.

This circuit is designed to operate from direct current. It can be designed to operate over a wide range of voltage levels. The limit on power capability is the current carrying capacity of the switches 51 and 52. Transistors (as shown) now are available with very significant voltage ratings and power capacities. Operation in a saturated switching mode, such as this, increases the current carrying capability of the switches. NPN transistors are shown but PNP transistors are equally applicable. If greater power is required or higher voltages are available, the circuit can be converted to use silicon controlled rectifiers (SCR). Additional components are required to turn the SCR switches off, but circuits of this nature are available in most handbooks on SCR operation.

A typical Eccles-Jordan astable multi-vibrator circuit operates at a fixed frequency established by the time constants of the passive components of the circuit. It is possible to add feedback coils to the pump that will speed up operation of the multi-vibrator as a function of pump air gap closure, and to modify such circuit to include the feedback coils. In FIG. 5 are shown the pump and astable multivibrator with the addition of feedback coils l3 and 14 to the pump. These coils are magnetically coupled to the power coils and sense the flux (not current) changes therein. See FIG. 1. When switch 51 is ON and supplying power to move the piston to close gap 16, capacitor 53 is charging at some rate toward a value that will turn off switch 51. Now, it is desirable to switch the feedback multivibrator 60 when the piston reaches the end of travel. Coil 14 is so connected that the large change of flux with respect to time at the end of travel (d/dt) will induce a voltage in coil 14 such as to charge capacitor 53 at a higher rate than the normal power circuit. Now coil 14 tends to receive a voltage spike near the end of piston travel due to the increasing rate of flux change with respect to piston movement and the increasing acceleration on the piston with the increased flux and resultant force. This voltage spike can be adjusted to charge the capacitor 53 very rapidly near the end of piston travel, thus decreasing the effective time constant of the circuit and hastening the switch time of switch 51. This feedback circuit speeds up operation of the pump by making the multivibrator switch as the piston reaches the end of travel instead of waiting for the full time constant of the MV circuit.

The lower part of the network of FIG. is simply that part of an Eccles-Jordan circuit.

The semiconductor operated circuit is highly desirable for power requirements from miniscule to moderatefWhere alternating current power only is available,

a simple rectifier or transformer/rectifier circuit with moderate filtering will supply the required direct current.

One class of magnetic core amplifying devices is called by a variety of names. The term self-saturating reactor (SSR) will be used. The "direct current form of SSR unit operates on alternating current and providefsa plusating direct current to the load. Two common types are the center tap SSR and the bridge SSR. FIG. 5a shows a center tap SSR. The basic circuit consists of two magnetic cores (which may be interconnected magnetically so as to look like one core) 101 and 102. Alternating current power is supplied through a center tapped transformer 116. Each SSR core has a load carrying winding, commonly called the gate winding 103 and 104, and a control winding 105 and 106. AnTalternating current supply is connected across the gate windings 103 and 104 in series with a load 108. Two rectifiers 110 cause the SSR to pass pulsating directjcurrent through the load 108. The inductance of a coil wound on a ferromagnetic core is well known as an impedance to the flow of alternating current. Thus, within the limits of the core and coil design, the two gate windings act to limit the flow of current in the load to thatsmall amount, known as exciting current, necessary to supply the core excitation. Referring to FIG. 4, if the core is operating on line 30 for ferro-magnetic material, and operating below the knee of the curve (saturation), the flux in the core (instantaneous and varying value) is proportional to the voltage across the winding. The current through the coilis proportional to thetmagnetizing' force,expressed in ampere turns. It

" can seen that the current will be very small for core operation below saturation. With the core windings connected as shown in FIG. 5a, each gate winding limits the current on one half cycle of the alternating voltage, one gate limiting the positive half cycle, the other limiting the negative half cycle.

Referring to the control windings on the two cores 105 and 106. These windings are connected in series and are powered by a direct current power source 107. The DC source can be varied in magnitude, varying the current through the control windings. (All devices in this class of magnetic core amplifiers are current control and limiting devices as compared to voltage control of vacuum tubes). If a small amountof direct current is passed through the control windings 105, 106 of the two cores, the ampere turns of the control winding will cause the magnetic flux of the cores to move up on the magnetization curve 30, FIG. 4, to some level, such as 87. If the alternating voltage is now impressed on the gate windings, the total flux in the core will be the sum of that produced by the control winding and the peak value of one half cycle of alternating current. This flux will reach a value such that the core passes over the knee of the magnetization curve. At that point, the alternating voltage supply can push a large amount of current through the gate windings since a small change in flux is caused by a large change in ampere turns. The same effect will be repeated in the other core on the next half cycle. Thus, impressing a small amount of DC power on the control windings causes the gate windings to pass some significant amount of alternating current during a portion of each half cycle. The portion of the half cycle during which the gate winding conducts is a function of the amount of control winding ampere turns impressed on the core. This is a simplified explanation of the operation. The basic law is that of equal ampere turns in control and gate windings, neglecting the exciting current required for the core.

Series connected and parallel connected saturable reactors operate in the same manner. The final results are slightly different. The load has an alternating currentpassed through it that is a function of the amount of control ampere turns impressed on the two magnetic cores. The control windings, and 106, and the gate windings, 103 and 104, and the magnetic cores, 101 and 102, and the control direct current power supply 107 are similar to the saturatable reactor. There are several significant differences.

In the typical operation curve of an SSR, AC or DC type, if no control current is supplied, a load current will flow at a particular point on the characteristic curve. To bring the load circuit to minimum output, a reverse bias current must be supplied. The ampere turns for this bias are normally supplied by a separate winding(s) similar to the control windings 105 and 106 (FIG. 5a). A constant direct current supply to this bias winding will place the operating point of the SSR at the minimum current level. This bias winding is not illustrated. It should be noted that there are many core and coil mechanical configurations for both saturable reactors and SSRs. One form has the cores 101 and 102 so placed that the control windings 105 and 106 can be replaced by a single winding. Two gate windings 103 and 104 are required. Another version uses the typical .transformer E-I core in the form called the three-leg reactor. v

A modification of the SSR has an additional windings(s) placed on the cores 101 and 102. This is called an extrinsic positive feedback winding (not shown). It can be connected in series with the load 108 or in parallel with the load. This winding is connected so that when power is applied to the load, ampere turns in the positive feedback winding aid the ampere turns of the control windings 105 and 106. This added control increases the power supplied to the load and increases the ampere turns in the winding. The winding can be designed to be of the snap-action type, this is, when power is applied to the control windings 105 and 106, the positive feedback winding supplies the increasing ampere turns to move the SSR to full output. The snapaction SSR is a two-state device.

The center tap SSR 115 with direct current output shown in FIG. a with positive extrinsic feedback will produce a snap-action SSR. Bias ampere turns are required from some source. The unique capability of this type of SSR lies in the ability of the transformer 116 to change the operating voltage to any desired level. Thus, devices designated to operate from automotive battery voltages can be readily operated from commercial AC line voltages.

FIG. 6 shows the application of the SSR to operation of the pump, i.e., two snap-action SSRs 117a and 117k to the drive of a pump. The bridge type SSR's shown have four rectifiers 110. Positive extrinsic feedback windings 111 provide the snap action. Bridge type SSRs are shown but center tap SSRs are equally applicable. For detailed description of the SSR, see Magnetic Amplifiers by H. F. Storm, John Wiley & Sons, 1955.) The load 108 of each SSR is the operating coil 1 l or 12 of the pump. A negative feedback winding 120 is added to each SSR. The control windings 105 n and 106 and the direct current supply 107 of each SSR are used as a bias winding and adjusted so that each SSR is ON, that is operating at a point on the upper limb of the control characteristic curve of the SSR. This method of operation is chosen to assure that the SSR multivibrator will start when power is applied. Other methods of operation are possible.

Since the SSRs are snap action, negative feedback ampere turns on either SSR will cause it to snap from the upper limb of its characteristic curve to the lower limb. When power is applied, current will immediately flow in both power coils of the pump 11 and 12.

Now assume that SSR 1 17B and coil 12 are operating on a closed gap 16. The high permeance of the circuit will produce high magnetic coupling between the load coil 12 and the feedback coil 14. The in-rush current in coil 12 will produce a large transient voltage on coil 14. SSR 117A and coil 11 are operating on an open gap 15. The coupling between coil 11 and feedback coil 13 is low and a minimum pulse will be induced in coil 13. The large pulse in coil 14 will pass readily through capacitor 122 and negative feedback coil 120 on SSR 117A, causing this SSR to snap off. SSR 117A will remain off until capacitor 122 is charged and the current is limited by the resistor 121. If this current is properly adjusted, SSR 117A will snap on. In the absence of a negative feedback pulse from coil 14, coil 13 will produce sufficient negative feedback ampere turns in coil 120 on SSR 11713 through capacitor 122, to snap off SSR 1 178. The current through load coil 108 will move the piston l and producea continuing large feedback pulse due to the flux switching described above. This large pulse will rapidly charge the capacitor 122 on SSR'117B so that as soon as the flux change stops due to end of piston travel, no added current flows in the negative feedback coil 120 on SSR 1 17B. This will permit SSR 1178 to snap on, since that is the normal quiescent state. Further, in the absence of further voltage 4 input from coil 13, the capacitor 122 on SSR 1178 will discharge in the reverse direction through coil 12 on SSR 1178, providing positive feedback to further speed the tum-on of SSR 1178. The process becomes fully oscillatory in a repetition of the above cycle.

The normal ultra-high gain SSR uses wound tape ferro-magnetic cores of high nickel content and special production, cost competitive techniques used in the manufacture of commercial transformers. The final appearance of each SSR is that of a single unit with a single core and requiring a single mounting means.

HIGH PRESSURE PUMP The pump described so far has had a piston of simple cylindrical form. This type of pump is limited to a maximum pressure of about 300 psi, to be understood. Higher pressure pumps can be built by suitable modification of the piston, cylinder and pole pieces. Such a pump is shown in FIG. 7. The operating volume of the pump is contained in the air gap 15 with a pressure proportional to the force applied on the piston ends 91 and 92, divided by the area of those piston ends. The piston is made in stepped form having a total crosssectional area of piston faces 91 and 93. The force generated by this area is directly proportional to the area of the face. If the force equation is stated in terms of flux density rather than total flux, the force becomes equal to BA/72; where B is the flux density in kilo-lines per square inch. B, the flux density, is limited by the saturation of the iron, as shown in FIG. 4. Thus, the total force is proportional to area. Since the area of the piston increases as the square of the diameter, a piston of twice the diameter will have four times the force. There are limitations on how far the stepped piston can be extended without excessive losses due to leakage flux that produces no force.

The pump of FIG. 7 has coils 11 and 12 to carry load current and coils l3 and 14 (FIG. 1) for feedback. Piston face 91, pole face 3 and cylinder 63 form one pumping chamber 15. Piston face 92, pole face 4 and cylinder 64 form the other pumping chamber 16. Piston faces 93 and 94, pole faces 97 and 98 and cylinder 2 are non-pumping volumes. Bleed holes 62 are provided in the piston between volumes 95 and 96 to allow any leakage material to flow from one volume to the other and prevent hydraulic lock before the end of the stroke. Valve ports 8 and 9 serve as inlet and outlet from the pump volumes. Cylinder 63 is non-magnetic to reduce the side force on the piston and resulting parasitic drag. The thickness of this cylinder wall should be nearly equal to the length of the pump stroke to minimize leakage flux from the pole face 97 to the cylindrical surface of the pumping piston 91. A short stroke pump of this type, such as might be used for hydraulic fluid (or other incompressible fluid) can be designed for pressures in thousands of pounds per square inch. The size of the pump is limited only by the daring of the designer.

A comment is in order on the design of a pump of this type, stepped or cylindrical piston. Since the force in the air gap builds up with decreasing gap length, the general design philosophy is to use a short stroke pump with incompressible fluids and along stroke pump with compressible fluids. This can be modified radically depending on the'desired pumping pressure.

It should be noted that the pump is self pressure limiting. If there is insufficient magnetic force in the air gap to overcome the back pressure on the pump, the piston will not move. It is thus possible to design pressure limiting into the basic pump.

It should also be noted that this pump basically has zero head space, which makes it look like a highly efficient pump. The pump piston is not allowed to hit the cylinder head each stroke to avoid noise. Small trapped volumes are designed into the piston and cylinder head to minimize noise.

Though it has been stated that pump design should be different for compressible and incompressible fluids, this does not prohibit using one pump for both purposes. The viscosity of the fluid in relation to the valve size is the limiting factor. Thus, in a small installation, such as a farm, one pump might be used for air, water or oil, as required. Conversely, the pump can be tailored for most efficient operation with a single fluid. The key features of the pump remain; simplicity, minimum parts and no holes in the pumping body.

BINARY LOGIC DRIVE CIRCUIT Two state circuits of the type commonly known as binary logic can be used to operate the pump, as shown in FIG. 8. A bistable device known as a flip-flop is shown at 130, 131, and 132. The R-S, flip-flop is shown, but the type D, type T, or type J-K flip flop can be used. The R-S F-F has the characteristic that if a pulse is applied to, the S (set) terminal, a continuous output is developed at the X terminal, even after the pulse is removed from S. If a pulse is now applied to the R (reset) terminal, a continuous output is developed at the X (not X) terminal. Terminals X and X are mutually exclusive. If a driving source alternatively applies pulses to terminals R and S of F-F 130, pulses will be applied to the S terminals of F-F 131 and 132 alternately. The use of capacitors 133 and 134 (with suitable dicharging and loading resistors 138 and 138') assure that only pulses reach the set terminal of F -Fs 131 and 132. When a pulse reaches the set terminal of F-F 131, an output (continuous) is developed on X. This output (of low power capability) is used to operate a switch 135 to control power gto the load coil 11 of the pump. The piston 1 will move to close the gap 15. Coil 13 will have an induced pulse generated by the change of flux in the gap (as explained above). This pulse will reach the reset terminal of F-F 131 and cause the X terminal to be active and the X terminal to be turned off. X is not used. When the reset terminal of F-F 130 is pulsed, the X terminal is activated which pulses the set terminal of F-F 132. This causes switch 136 to be turned on and the coil 12 operates to move the piston to close the gap 16. Coil 14 pulses the reset terminal of F-F 132 and turns off power as soon as the gap is closed. Power for the coils 11 and 12 is from lines 139 and 139' controlled by switches 135 and 136.

Pulses from some clock source 137 may be used to drive the F-F 130. The unit may be made free running by coupling the leads from coils l3 and 14 back to 137. This may require some external pulse to generate the first movement of the piston. The prime advantage here is theready availability and low price of monolithic integrated circuit (IC) flip flops in small sizes and standard packages. The power switch can be electromechanical relay or semiconductor or other type that is compatible with the output of IC flip flops.

Computer control of mixing processes is becoming widely used. The logic drive form of the pump lends itself to this system. Since the pump has a constant displacement for each stroke (half cycle), the quantity of a mix component can be varied in a continuous mix process by computer control of the pump rate. type T F-F has a single input and changes the X and X state with each pulse on the input. If a type T F-F is used at F-F 130, the computer can command the required component flow by timing the pulses to the input. Thus, with a pump for each component that may be used in a variety of mixtures, the computer can accurately and continuously control the mix ratios by proper pulse rate on the logic input of each pump. If the mix ratio of a component varies over a wide range, two parallel pumps of widely different stroke volumes can be used.

Other logic fonns are possible to generate the required operating pulses. Variable rate is possible with most of such circuits. Clocked operation is possible with the type D and type J-K flip-flops. Type J-K is a universal type that will perform the functions of all other generally classified flip-flops.

SHEARING ACTION VALVES In the case where it is undesirable to put outlet and inlet ports in the cylinder head 3 and 4, or where a very coarse material is being pumped and poppet or reed valves may stick open due to material collecting on the valve seat, it will be of advantage in some of these cases to have a valve with a shearing action to clear the port. A rotating sleeve valve will fulfill the needs of both cases. This valve sleeve can be made to operate by the magnetic forces of the piston.

Considering the explanation above, the same MMF distribution explanation applies to the sleeve valve with magnetic actuation. (Other means of actuation are possible and not prohibited by this explanation.)

FIGS. 9 through 10a show a sleeve valve surrounding the cylinder 2 with inlet port 154 and outlet port disposed in a common plane through the pump and disposed at some radial angle around the axis of the pump cylinder 2. Piston l is in the open gap postion with the chamber 16 filled. A high MMF now exists across the open air gap 16 as explained above.

Two magnetic' shunts are placed in the wall of cylinder 2 on each side of the pump axis. (This is for force balance.) Shunt 152 and 1524 will be designated the piston shunts; shunts 153 and 153a will be designated the head shunts. On the outside of the valve sleeve 150 are placed two torque bars 151, 151a of magnetic material. Further, an operating spring 158 is placed on the valve sleeve (of any reasonable configuration). A coil spring is shown but a leaf or lever spring will work equally well. The spring operates with a torque 159 to rotate the valve in such a direction as to place the valve sleeve intake port 154 and the cylinder intake port 156 opposite each other, as shown in FIG. 10 (port open).

The high MMF across the open gap now passes through the magnetic shunts, 152, 152a, 153, and 1530 to couple the magnetic flux into the torque bars 151 and 151a. The flux in the air gap (and the nonmagnetic valve sleeve 150) exert a force on the torque bars such as to reduce the permeance of the circuit. This force causes the torque bars to rotate the valve sleeves 150 until the torque bars are directly opposite the magnetic shunts. This rotation closes the intake ports 154 and 156 and opens the outlet ports 155 and 157 (FIG. 10a).

As the pump piston is moved (by a current through coil 12), the material is forced out of the air gap 16. With the piston close to the head, the MMF across the sleeve valve magnetic circuit 152-151-153 is reduced and the flux is diverted through the piston 1 and head 4. The rotator spring 158 turns the valve sleeve 150 until the inlet ports 154, 156 are open. The pump is now ready to draw material into the air gap 16 on the next intake stroke. Because the permeance of the shunt magnetic circuit 1-152-151-153-4 is decreased in the outlet position, the valve ports will remain in the outlet position, even .though the MMF across the circuit is greatly lowered. The air gaps in the circuit assure that the permeance of the circuit is sufficiently low in the outlet" position that the flux will be diverted through gap 16 and let the spring rotate the valve.

The discrete magnetic shunt plugs are required in the wall of the cylinder to provide a specific flux path of low permeance in relation to all other radial positions around the axis of the pump such that a force in the torque bar 151, 151a will exist in relation to this specific point.

It is feasible to put a separate valve sleeve operating coil in the system and use the same electrical power source for valve operation as is used for piston operation.

VARIABLE DISPLACEMENT PUMP The requirement for a variable, positive displacement pump is obvious. Such a pump might be used as the fuel throttling device for fuel injection of internal combustion engines. Another use will be in mixing of two or more components in variable ratio. Two pumps can be operated in synchronism from the same source of power with one pump operating at a fixed displacement and the other operating with displacement varied to suit the mix ratio required. A third application is for those fluid or gas moving requirements that require a variable flow rate for any reasons, such as variable air supply for operating air powered tools at varying rates or where demand will vary with the number of tools used and the utilization rate of the tools. This use stands in comparison to present practice of using constant displacement pumps on an intermittent duty cycle or running continuously and dumping excess air.

FIGS. 11, 12, 13, and 14 show various means of making the pump have a variable displacement by varying the length of the pump stroke. FIGS. 15 and 16 show means of using a constant displacement form of the pump with diverter valves to return excess fluid or gas to the inlet side of the pump.

FIG. 11 shows a pump having one head 4 fixed in position and one head 171 being movable axially into and out of the cylinder tube 2. Inlet ports 9 and outlet ports 8 are contained in each head. The movable portion of the head 171 is ferromagnetic. It is threaded and screws into a mating threaded portion of the fixed portion of the head/pole piece 170. Rotation of the moving portion 171 causes it to move from the retracted position 172 to the extended position 173. Since the displacement of piston 1 is fixed by the length of the stroke, reducing the distance between heads 171 and 4 reduces the stroke length and thus reduces the piston displacement whether the piston moves left or right. By suitable design of the thread pitch, the relation between rotation of the head 171 and the displacement can be made to have any reasonable value.

To establish a specific utilization of this variable displacement pump, the fuel injection system for an internal combustion engine may be considered. Present systems use a pump to draw fluid from the storage system or tank and present it under pressure to a series of valves. The length of time a specific valve is held open determines the amount of fuel injected. Systems exist that have valves assigned one to each cylinder. Other systems have one valve serving multiple numbers of cylinders. The variable displacement version of this pump will have one fuel injection for each stroke of the pump, whether left or right. Thus, this pump version will serve to displace the pump on existing systems and two valves. Each "side of the pump will serve the function of one valve.

Diesel engines require fuel injection under pressure and at a variable rate. A high pressure version of this pump, as described above, will serve as the injection means for'a diesel engine.

The pump configuration described above requires rotation of the head 171 for operation. This implies that the inlet and outlet of the moving head 171 be connected to flexible tube means to permit rotation of the head. FIG. 12 shows a movable head 171 and fixed portion of the pole piece having the same relation as above. A threaded nut 174 is interposed between a portion of the moving 171 and fixed 170 portions of the head/pole piece. The nut 174 may be single or double threaded. The movable piston is fixed in a nonrotatable position with provision-to move axially along the cylinder. The threaded nut is configured in one of several forms so that rotation of the threaded nut 174 causes the head 171 to move axially. The amount of rotation of the threaded nut 174 determines the amount of movement of the piston 171 and thus the stroke and displacement of the pump. The threaded nut may be of ferro-magnetic material or not. FIG. 12 shows the nut 174 with a left-hand thread on the inner surface in contact with and mating with a similar left-hand thread on the moving piston 171. The nut has a right-hand thread on the outer surface in contact with and mating with a similar right-hand thread on the pole piece portion 170. Rotation of the nut causes the nut to move with respect to the fixed portion 170 and causes the movable portion of the head 171 to move with respect to the nut and in the same direction. By this means the axial movement of the movable head 171 is greater than the movement of the nut 174. For equal pitch on left-hand and right-hand thread, the head moves twice as far as the nut. The movement of the head 171 still changes the displacement of both strokes of the pump since the stroke length is the same in both directions.

The double threaded nut 174 of FIG. 12 may have the threads cut in the nut 174, movable pole 171, and fixed pole 170 so the threads mate, i.e., male threads on the nut 174 and female threads on the movable pole 171 and fixed pole 170. Sealing of a thread of this type against fluid leakage may be significant. FIG. 13 shows a means where a female" thread is cut in all parts. A flexible sealing means 175 and 176 is so placed as to tightly fill the combined recesses of the threads. This sealing means has the advantage of a large sealing area, i.e., multiple sealing rings in the axial direction and a long and high resistance flow path along the axis of the seal 175. Any form of sealing means maybe supplemented or replaced by the circumferential seals 5. Design of the air gaps between the various elements of the movable head is well known to those skilled in the art.

The double threaded nut .01 FIG. 12 may be replaced by a single threaded nut, as shown in FIG. 14. The nut 174 is shownwith a thread mating with a similar thread on the movable head 171. The thread may take the fonn shown: in FIG. 13 or may be directly mating. The nut 174 is suitably constrained so as not to move axially with respect to the fixed pole 170. The movable pole 171 is constrained against rotation by suitable means.

The location of the threads may be reversed with the threads between the nut 174 and the fixed pole 170. The nutis. constrained to movewith the movable head 171.

Other means may be used to displace the movable pole 171 such as a cam, an external drive rod from any source (such as a hydraulic piston) or other mechanical linkage;

External by-pass valves may be used to control the displacement of a constant volume pump, as shown in F 1G. 15. Piston 1, cylinder 2, pole piece/cylinder heads 3 and 4, and permanent magnet 7 constitute parts of a constant displacement pump such as described in paragraphs above with inlet and outlet valves in the pole piece/cylinder heads .(PP/Cl-ls) 3 and 4. Three-way valves 178 and 180 areconnected to the outlet valves 8 and-8a and to the two outlets of the system. A typical three-way valve is shown in the two positions of operation in FIGS. 17 and 18. Valve 178 is controlled by a ferromagneticarmature 177 placed in conjunctionto the displacementair gap with an air gap 186 and coupled to valve 178 by-a mechanical linkage 183. The armatures 177 and 179 are not circles but single sided and more perpendicular'to the pump as is due to flux flow in gaps 185 and 186. Similarly, valve 180 is coupled to a'ferromagnetic armature 179 through a linkage 184 with thearmature disposed in relation to displacement air gap 16 with an air gap 185. The armatures 177 and 179 will'be attracted toward the air gaps 15 and 16 in a manner as described relative to the magnetic reed switch.- The armatures will be restrained away from the displacementair gaps 15 and 16 by suitable springs, thus keeping the air gaps 185 and-186 at some specified distance. With the armatures away from the magnetic circuit, valves 178 and .180 will be in the position to pass fluid to the outlet of the system. With the armature pulled in to reduce the length of air gaps 185 and 186, the valves-180 and 178 will be in a position to return fluid to the intake side of the system through lines 181 and 182.

Referring to piston 1 in the position shown, most of the MMF of the right-hand side of the magnetic circuit will exist across the air gap 15. This will produce a high MMF across the air gap 186 with a resulting flux that will produce a force closing the air gap 186 and moving the armature 177 in the position to put valve 178 in the positionto bypass fluid to the retum'line 182 to the inlet of the system. With the air gap 16 closed, the MMF across the gap will be low, and there will be little flux in the air gap 185. The low force will allow annature 179 to move away from the magnetic circuit and movevalve 180 through linkage 184 to the outlet position. The nonnal one-way valves are installed in 8, 8a, 9 and 9a. Q

If a power pulse is applied to coil 11 around air gap 15, the'piston 1 will'move to close air gap 15 and expel fluid to the outlet through the three-way valve 180.

Fluid will flow to the outlet until the MMF builds up across air gap 16 to a level sufficient to move the armature 179 and switch the valve to divert flow from air gap 15 to the by-pass line 181 and back to the inlet 4 side of the system. The permeance of the air gap 185 can be varied by making the restrained position of armature 179 closer to or farther away from the magnetic circuit of 1 and 4. The value of this permeance will control the required MMF necessary to pass sufficient flux through the air gap 185 to operate the armature. The MMF across air gap 16 will increase with the displacement of the piston 1 and will likewise increase the MMF across air gap 185. Thus, by adjusting the length of air gap 185 the position of piston 1 that will actuate armature 179 and valve 180 can be varied. Since the movement of valve 180 diverts flow from the outlet, the amount of fluid flowing to that outlet during a stroke of the piston can be controlled and varied by varying the length and permeance of air gap 185. The fluid passed on the opposite stroke by air gap 16 can be varied by adjusting the length and permeance of air gap 186 to vary the point at which valve 178 diverts flow from the outlet to by-pass line 182 and back to the system inlet supply.

This system makes use of a standard form of the pump and diverts a portion of the flow of each stroke by three-way diverter valves and magnetic armatures operating on the MMF of the basic pump to control the amount of fluid pumped through the outlets on each stroke. The point of actuation of the diverter valves is controlled by varying the length and permeance of the air gaps between the armatures and the main operating magnetic circuit.

If it is desired to use a fully standard pump of the type described above, an external floating piston can be used, as shown in FIG. 16. A pump as described is represented by elements 1, 2, 3, 4 and 7. A common inlet line is established by line 199. (This will be true of all pumps but is significant here). A diverter piston and valve assembly 187 is coupled to the outlets 8 and 8a by lines 200 and 201. The pump outlets are also coupled to the inlet manifold 199 by diverter valves 203 and 202, each being a three-way valve like FIGS. 19 and 20. The diverter assembly 187 contains a free floating piston 189 in a cylinder body 188 with operating rods 190 and 191 at each end of the cylinder. A variable length means 192 is mounted on operating rod 190. The fluid outlet lines 200 and 201 are connected to the two ends of the cylinder 188 and to outlets and 196 through valves 193 and 194. Operating rods 191 and 192 are mechanically connected to valves 193, 194, 202 and 203 through linkage 197, including the variable length means 194. Consider the pump with piston 1 as shown. Valve 202 is set to pump with fluid flowing up through line 200. Valve 193 is closed. Valve 203 is set to by-pass to manifold 199 with line 201 closed and valve 194 open.

By applying power to the coil 11 around air gap 15, the piston 1 will move to the right and close air gap 15 causing it to discharge. Air gap 16 will open and charge. This will force fluid through the valve 202 into manifold 200 and into the right side of cylinder 188 moving piston 189 to the left. This will force fluid out of the left side of cylinder 188 into the manifold 201. Since valve 203 is closed to line 201 and valve 194 is open, fluid will be forced out of outlet 195 to the desired location. After a volume Q has been discharged,

the piston 189 will push rod 191. This causes all valves to change position, i.e., valve 202 will switch to by-pass fluid from the air gap 15 to the inlet manifold 199 and valve 195 will close stopping flow from outlet 195. Since line 200 is closed to valve 202, valve 193 can open without flow through it. Piston 189 is stationary. Valve 203 is open to air gap 16 but cannot pass fluid to its because of the one-way valve in outlet 8a. Piston 1 will complete travel to the end of the cylinder 2 and discharge any excess fluid from air gap 15 into the inlet manifold 199. Now by applying an electrical power pulse to coil 12 around air gap 16, piston 1 moves to the left, discharging fluid from air gap 16 through valve 203, which is open to pump, to manifold 201. Valve 194 is closed so the fluid is forced into the left side of cylinder 188 moving the piston 189 to the right. This discharges fluid into manifold 200. Since valve 202 is still in by-pass position, it is closed to the manifold 200 and fluid passes through the open valve 193 to the outlet 196. When the piston 189 contacts operating rod 190, the valves are all re-set. Excess fluid in air gap 16 is by-passed to intake manifold 199. The system has completed one discharge each from outlets 195 and 196.

To vary the amount of each discharge pulse, the variable length means 192 is changed. This increases or decreases the stroke length of piston 189 and increases or decreases the amount discharged on each stroke. Any excess fluid in air gaps l and 16 is by-passed to intake manifold 199.

Electrical means of changing the positions of the valves are possible and practical.

ELECTRICAL GENERATOR It is practical to make an electrical generator from the pump device described herein. As pointed out above, the magnetic flux switches to the side of the magnetic circuit having the highest permeance. Thus, if the armature 1 of FIG. 1 is mechanically moved from one side to the other, the flux will switch from one side of the magnetic circuit to the other. In so doing, the coils will experience a change in the amount of flux linking each coil. As pointed out above, in connection with the operation of the feedback coils 13 and 14, the change in flux will produce an electromotive force. The relation drb/dt, time rate of change of flux, determines the magnitude of the voltage developed in the circuit. Thus, if some means is employed to mechanically move the armature from side to side, an alternating voltage will be developed in the coils surrounding the magnetic circuit.

Various means are available for moving the armature mechanically. Such means may comprise the application of air, steam, gasses or fluids under pressure to each side of the piston 1 alternately. Mechanical movement through a push rod from any source is also possible. One unique configuration is the application of a modified form of internal combustion engine, discussed below. Another possibility is the application of the external gas generator combustion engine known as the Stirling Cycle engine. This engine employs a free floating piston with energy exchange taking place in an external circuit. It has been extensively discussed in the literature. The moving piston of this Stirling cycle engine suggests itself to use in this device, used as a generator. The application of the Stirling cycle engine will be understood by those skilled in the art from the discussion of another form of power generation described below.

FIG. 21 shows a modified form of the abovedescribed pump. A piston 21 1 moves in a sealed cylinder 212 with cylinder head/pole pieces 213 and 214. A central permanent magnet 217 completes the magnetic circuit, identical to that described above. Coils 250 and 251 surround the air gaps l5 and 16 or the magnetic material associated with those air gaps. A single orifice penetrates the Cl-I/PPs, orifices 215 and 216. The various configurations of the coils 250 and 251 will be explained below.

This device is presumed to operate on a combustible material in mixture with air, similar to but not identical with, the four-stroke Otto cycle. To complete the four strokes of the Otto cycle with only two strokes of the piston 211 requires an auxiliary device. A two-stroke compressor 209 is used to perform the intake and compression strokes of the four-stroke cycle. The piston 211 and cylinder 212 of the generator perform the other two strokes, expansion (or firing) and exhaust. A pre-compression chamber takes the place of the head space" of a conventional internal combustion engine, 226 and 227.

The compressor 209 will be described in terms of the pump mentioned above, but other forms can be used, such as a conventional crank and piston two-stroke compressor. Compressor 209 takes a form similar to the pump shown in FIG. 1. A piston 1 moves in a cylinder 2 between PP/CI-I's 3 and 4, each having a single orifice, 218 and 219. A permanent magnet 7 and coils (not shown) 11 and 12 complete the compressor. Referring to the right side of the circuit, with the left side identical, the CH/PP orifice 218'is connected by a manifold 238 to a three-way valve 230 with an air inlet 232. A fuel mixer nozzle 220 is placed in the air flow manifold 238 such that air flowing into the cylinder will entrain fuel in the air mix. Fuel is supplied from a source (tank) 224 through a one-way (check) valve 222 to prevent blow back to the fuel source during the compression stroke. The valve 230 can selectively couple the manifold 238 to the air inlet 232 for the intake stroke and to the pre-compression chamber 226 during the compression stroke. The pre-compression chamber 226 is connected to a three-way valve 234 by a manifold 240. The precompression chamber contains an electrical ignition means 228 such as a spark gap or glow plug. The three-way valve 234 can selectively couple the exhaust orifice 236 to either or both of manifolds 240 and 242 or couple the two manifolds without coupling the exhaust orifice, after the fashion of FIGS. 17 and 18 (with the third position not shown). Manifold 242 couples the valve 234 to the orifice 215 in the CPI/PP 213 of the generator. The opposite side contains: fuel jet 221 in manifold 239 with fuel check valve 223 and fuel source 225, three-way valve 231 connecting intake inlet 233 and the manifold to the precompression 227, ignition means 229, manifold 241 connecting to three-way valve 235, exhaust port 237, connecting manifold 243.

The three-way valves 230 and 231 can be replaced by two check valves, i.e., one in line 232 will admit air but block flow out of 232; a valve between 230 and 226 will pass fuel/air mix to 226 but block flow in the opposite direction (during the intake stroke).

Referring to the operation of the system. With piston 1 as shown, an electrical power pulse applied on the left part of the circuit will draw the piston to PP/CH 4. With valve 230 connecting manifold 238 to the intake inlet 232, air will be drawn into the cylinder 2 through the manifold 238. In passing the fuel inlet 220, fuel will be drawn into and mixed with the air in proper proportions as controlled by the design of the fuel inlet 220 and associated parts. Valve 230 decouples manifold 238 from the precombustion chamber 226 during the intake stroke. An electrical power pulse applied to the right side of the circuit of compressor 209 will expel the combustible mixture. If valve 230 is set to couple manifold 238 to the pre-combustion chamber 226, the chamber will be charged with the combustible mixture under high pressure. If, during this interval, valve 234 is set to couple the precombustion chamber to the air gap 15, manfiold 242 and the air gap orifice 215 and de-couple from the exhaust inlet 236, the manifold 242 and orifice 215 will also be charged with the combustible mixture. At the end of the stroke of piston 1, valve 230 is closed, the compression stroke of the system has been completed. The ignition means 228 is fired and the combustible mixture ignited. The release of the energy of the combustion will drive the piston 211 to the left end of cylinder 212. This effects the flux change in the coils 250 and 251. If the charging and firing cycle is repeated on the left portion of' the cycle and valve 234 is positioned to vent the air gap and precombustion chamber 226 to the exhaust valve, the return motion of piston 211 will accomplish the exhaust stroke. Suitable timing of the valves and electrical power pulses to the compressor 209 will produce continuous reciprocating motion of the piston 211 and repeated flux switching in the coils 250 and 251, with resulting output of electrical power.

Since the flux is alternately switched into and out of the coils 250 and 251, the electromotive force generated is alternating. Because of the highly non-linear motion of the piston 211 and the rate of change of flux, the output E.M.F. will not be sinusoidal. It may be desirable to convert the E.M.F. to direct current. Various coil and auxiliary component configurations can accomplish this. The conventional bridge rectifier is one form. FIG. 22 and 22a illustrate another method, similar to a center-tapped transformer and rectifier direct current source. The coils 250 and 251 are divided equally with the center tap brought out. Connection of rectifiers 256 as shown to the four half coils 252, 253, 254 and 255 will produce a pulsating direct current. The outputs can be connected in series for higher voltage or in parallel for higher current capacity. The interconnection can be on the AC or DC side of the rectifiers. This configuration is familiar to persons skilled in the art of power supplies and is shown as a typical example of direct current output from the flux switching generator.

The most probable use of a generator of this type is for remote, hazardous or inaccessible locations (underseas) with some form of long-time external heat generator (nuclear reactor) operating an external combustion engine such as the Stirling cycle engine. Another possibility is the requirement for a short-time operation on compressed gas, such as in short-range guided missile applications. Solid fuels, such as gunpowder, are

applicable to this type of generator. Further, the use of this generator for a single shot" use, such as detonation of electrical squibs and blasting caps is possible.

The pump device as shown in FIG. 1 can readily be driven as a generator as shown in FIG. 21. To do so, manifold 242 is connected to both valves in pole 3 (FIG. 1), and manifold 243 is connected to the valves in pole 4 (FIG. 1).

From the foregoing, those skilled in the art will readily understand the nature and construction of the invention and the manner in which it acheives the objects set forth in the foregoing. The device is reversible, operative as either a pump or as a generator using the same windings. As a pump, the device is unique in that it is completely sealed and uniquely adapts itself to drive from particular types of electrical control circuits. The range of applicability of the device as a pump is greatly extended by way of simplified means for variable displacement and for higher pressures and for shearing action of its valves.

The foregoing disclosure is representative of preferred forms of the invention, preferred forms of drive of the pump, and preferred applications and adaptations of it; and it is to be interpreted in an illustrative rather than a limiting sense, the invention to be accorded the full scope of the claims appended hereto.

What is claimed is:

1. A sealed pump comprising: a cylinder of nonmagnetic material;

a free floating piston, of magnetically attractable material, slidable in said cylinder and being free of mechanical connection to the exterior of said cylinder;

means closing the ends of said cylinder and having .inlet and outlet ports therein;

a magnetic core means having pole-defining ends having axially facing pole faces facing inwardly at opposite ends of said cylinder;

permanent magnet means establishing a permanent magnetic field in said core means and through said cylinder andpiston; and 1 electrical windings disposed about said cylinder adjacent the ends thereof to produce, when energized, a magnetic field cooperable with said permanent magnetic field to reciprocate said piston in said cylinder.

2. A pump as defined in claim 1 wherein said poledefining ends of said core means extend into said cylinder and constitute said means closing said ends.

3. A pump as defined in claim 2 including inlet and outlet valves in said pole defining ends and respectively controlling said inlet and outlet ports.

4. A pump as in claim 1, including switch means for energizing the windings in a manner to cause the piston to reciprocate.

5. A pump as in claim 1, constructed to produce relatively high pressures wherein the said cylinder and piston each comprise a portion having a first diameter and a portion having a second diameter, said pole faces being configurated to be of a size comparable to a part of the piston and cylinder of larger diameter.

'0: s s t:

Patent Citations
Cited PatentFiling datePublication dateApplicantTitle
US1912167 *Jul 13, 1931May 30, 1933Leroy J LeahyElectrical reciprocating motor
US2447230 *Jun 11, 1946Aug 17, 1948Mallory & Co Inc P RVibrator
US2579723 *Oct 28, 1947Dec 25, 1951United Aircraft CorpMagnetic device
US2686280 *Oct 25, 1949Aug 10, 1954Klingel John BElectromagnetic piston pump
US2690128 *Mar 24, 1950Sep 28, 1954North American Solvay IncElectromagnetic pumping device
US2701331 *Aug 29, 1949Feb 1, 1955Frank K BenniePumping apparatus with electromagnetically propelled piston
US2721453 *Jan 30, 1953Oct 25, 1955Leon Reutter JeanSynchronous oscillating compressor for alternating current
US2749453 *Apr 30, 1952Jun 5, 1956Reeve Electrical Co IncElectric bell motor
US2832919 *Jun 12, 1953Apr 29, 1958Leon Reutter JeanMovable equipment for electro-magnetically controlled devices
US2892140 *Aug 30, 1956Jun 23, 1959Maxim Electronies LtdCurrent switching device for electromagnetic hammers
US2988264 *Aug 20, 1959Jun 13, 1961Chausson Usines SaAlternating movement synchronous compressor
US3022450 *Sep 15, 1958Feb 20, 1962Bendix CorpDual position latching solenoid
US3178151 *Jan 23, 1963Apr 13, 1965Marquardt CorpLinear displacement electromagnetic actuator
US3293516 *Nov 18, 1964Dec 20, 1966Bosch Gmbh RobertElectromagnetically driven pumps, particularly fuel pumps
US3328659 *Dec 16, 1965Jun 27, 1967Smith Corp A OSwitching apparatus for plural step motors connected by differential gearing
CA728771A *Feb 22, 1966Exxon Research Engineering CoMagnetic pump system
Referenced by
Citing PatentFiling datePublication dateApplicantTitle
US3968387 *May 16, 1975Jul 6, 1976Lawrence Peska Associates, Inc.Linear magnetic generator
US4005346 *Oct 29, 1975Jan 25, 1977Hsia Tsang ChenDirect electromagnetic reciprocating motor
US4314532 *Jun 17, 1980Feb 9, 1982Robertshaw Controls CompanyMethod and apparatus for translating frequency of mechanical movement into a pneumatic signal
US4468177 *Apr 27, 1981Aug 28, 1984Strimling Walter EDiaphragm pump arrangement in which alternately expanded and contracted chambers are used independently
US4664685 *Nov 19, 1985May 12, 1987Helix Technology CorporationSensing the back emf in displacer drive coil
US4671745 *Mar 21, 1986Jun 9, 1987Smith Raymond HMagnetically-activated motorized pump
US4965864 *Dec 7, 1987Oct 23, 1990Roth Paul ELinear motor
US5057724 *Jan 16, 1990Oct 15, 1991Patton James VCeramic magnet motor
US5102306 *May 8, 1990Apr 7, 1992Liu Kuo ShengAC/DC air pump
US5106274 *Jul 23, 1990Apr 21, 1992Mark HoltzappleHermetic compressor
US5166563 *Mar 2, 1990Nov 24, 1992Stuart BassineMagnetically actuated linear displacement compressor
US5203172 *Feb 18, 1992Apr 20, 1993Simpson Alvin BElectromagnetically powered hydraulic engine
US5587615 *Dec 22, 1994Dec 24, 1996Bolt Beranek And Newman Inc.Electromagnetic force generator
US5727388 *Dec 21, 1993Mar 17, 1998Adamides; AlexanderSolar activated positive displacement piston pump-rotor drum turbine
US6208497Jun 26, 1997Mar 27, 2001Venture Scientifics, LlcSystem and method for servo control of nonlinear electromagnetic actuators
US6813885Mar 1, 2001Nov 9, 2004Saab AbElectrohydraulic setting device
US6884040 *Dec 27, 2001Apr 26, 2005Pratt & Whitney Canada Corp.Multi pumping chamber magnetostrictive pump
US6942469Jan 30, 2001Sep 13, 2005Crystal Investments, Inc.Solenoid cassette pump with servo controlled volume detection
US7001158Jan 24, 2003Feb 21, 2006Sturman Industries, Inc.Digital fluid pump
US7040873Feb 14, 2005May 9, 2006Pratt & Whitney Canada Corp.Multi pumping chamber magnetostrictive pump
US7056105 *Nov 28, 2003Jun 6, 2006Ming-Hwa LiuUnit utilizing current to control reciprocation for pushing fluids
US7503756Mar 10, 2006Mar 17, 2009Pratt & Whitney Canada Corp.Multi pumping chamber magnetostrictive pump
US8512011 *Dec 19, 2007Aug 20, 2013Dresser, Inc.Fluid pump and fuel dispenser
US20080164287 *Dec 19, 2007Jul 10, 2008Larsson Bengt IFluid pump and fuel dispenser
CN1068149C *Aug 31, 1995Jul 4, 2001王广生Magnetic energy storage device
CN100420850CNov 20, 2003Sep 24, 2008刘明华;陈定福;简金榜Unit using current to control reciprocating motion to motivate fluid
WO1993016297A1 *Jan 29, 1993Aug 19, 1993Charley W BiselElectromagnetically powered hydraulic engine apparatus and method
WO2001077531A1 *Mar 1, 2001Oct 18, 2001Kjellquist PeterElectrohydraulic setting device
WO2005017357A1 *Aug 5, 2003Feb 24, 2005Chen Chung ChinElectromagnetic pump
U.S. Classification310/30, 310/17, 417/418, 310/34
International ClassificationH02K33/00, F02B75/02, H02K33/14
Cooperative ClassificationF02B2075/025, H02K33/14
European ClassificationH02K33/14