US 2856533 A
Abstract available in
Claims available in
Description (OCR text may contain errors)
Oct. 14, 1958 J. ROSENTHAL 2,
MOVING WIRE CORONA Filed Jan. s, 1956 2 Sheets-Sheet 1 Q CORONA L 5' POTENTIAL 5H\ELD POTENTiAL CORONA POTENTIAL \SHIELD POTEN'HAL CORONA I POTENTIAL POTENTIAL CORONA POTENTIAL -2'5 SHIELD POTENTIAL CORONA POTENTIAL SHIELD E5 POTENTlAL IN VEN TOR.
122g: 5 JOSEPH F. ROSENTHAL ww T Arm/ 21w? Oct. 14, 19558 J. F. ROSENTHAL 2,
MOVING WIRE CORONA Filed Jan. 3, 1956 2 Sheets-Sheet 2 24' g CORONA -zs POTENTIAL SHIELD POTENTIAL CORONA POTENTIAL SHIELD POTENTIAL A C -50 SUPPLY w 4 35 CORONA i 1 POTENTIAL I sHlELD POTENTIAL 5'3 .6'5' CORONA T154 M POTENTIAL 56 $HIELD POTENTIAL INVENTOR.
JOSEPH E ROSENTHAL United States Patent MOVING WIRE CORONA Joseph F. Rosenthal, Rochester, N. Y., assignor to Haloid Xerox Inc., Rochester, N. Y., a corporation of New York Application January 3, 1956, Serial No. 556,870 Claims. (Cl. 250-495) This invention pertains to a xerographic process and apparatus therefor, and more particularly to the negative charging of insulating surfaces.
In Xerography it is usual to form an electrostatic latent image by applying an electric charge to the surface of a photoconductive insulating layer and selectively dissipate charge from the surface by exposing the charged surface, optionally through a suitable projection system, to a document, scene or other image to be reproduced. The resulting electrostatic latent image is thereupon developed by dusting it with oppositely charged powder which adheres to charged portions of the photoconductive insulating layer. In this manner there is formed a reproduction of the original being photographed or copied. In order to produce a Xerographic picture it is usual to employ a dark colored powder and to transfer it ultimately to a white support base such as a sheet of white paper.
By the operation just described, it is apparent that there is formed a dust image corresponding to a direct positive reproduction of the original. Thus, the portions of the original that are white or light in tone are reproduced as an absence of deposited powder, whereas the dark portions of the original are reproduced by a heavy pattern of deposited powder.
One particular method of Xerographic development is known as continuous tone development and is carried out by placing the electrostatic latent image in a closely spaced parallel relationship with a conductive surface and dusting the latent image by blowing a gaseous suspension of finely divided charged particles through the narrow space between the photoconductive insulating layer and the conductive surface. This method deposits on the insulating layer a dust image which is particularly responsive to slight variations in the electrostatic potential on the insulating layer. This method of development is therefore particularly adapted to the reproduction of continuous tone subjects such as photographs or original scenes. The deposition of the dust or powder is likewise sensitive to small variations in the initial potential to which the photoconductive insulating layer is charged and it is therefore necessary to place a particularly uniform potential on the surface in order to avoid defects in the developed image. Charging devices for use with continuous tone xerography must therefore charge the photoconductive layer evenly and it is also desirable to charge to a predetermined potential, avoiding potential variations such as might be caused by changes in power line voltage or changes in rate of motion or other valiables which may enter into the charging process. These needs have been met in the past by moving the photoconductive insulating layer under one or more corona generating wires which are surrounded by a form of corona control element.
Continuous tone Xerography has commonly been carried out with photoconductive insulating layers of vitreous selenium and because of the electrical characteristics of this particular material it has been expedient to charge the layer positively. It may be noted that xerography,
Patented Oct. 14, 1958 ice as it has been described to this point, forms positive reproductions in that black areas of an original subject are reproduced as black in the xerographic copy. Occasionally, however, it is desired to employ xerography for the reproduction of reversal or negative images in which the original copy to be reproduced in itself corresponds to a tone reversal of an original scene or document. For example, if it is desired to make a Xerographic print from a conventional photographic negative it is necessary to reverse the blacks and whites. A method for accomplishing this result is disclosed in co-pending application Serial No. 556,869, filed January 3, 1956, by Richard E. i-Iayford and Alfred C. Haacke. In this method an additional operation is performed on the electrostatic latent image prior to the development of the image. This operation consists in the deposition of a uniform amount of negative charge per unit area of the previously positively charged photoconductive insulating layer. When properly carried out this has the effect of neutralizing those areas which previously had the greatest positive charge and leaving a net positive charge on those areas which previously had less negative charge. This results in a new electrostatic latent image which is related to the old in that the polarity of charge is reversed and the amount of charge on any area of the plate is inversely proportional to its former charge. Development of such an electrostatic latent image will produce an image in which the blacks and whites are reversed from the original copy.
By requiring a negative charging step, this new method raises additional problems in continuous tone xerography since positive corona discharge from a wire generally occurs in a volume in the form of a continuous uniform sheath surrounding the wire, while negative corona has a tendency to concentrate itself at discrete points along the wire with the result that negative charge deposits on the photoconductive insulating layer in a non-uniform pattern corresponding to the non-uniformity of the corona discharge itself. The design features required of a negative charging corona device adapted to deposit uniform charge per unit area tend to emphasize rather than diminish the charging non-uniformities that can result from a non-continuous corona discharge at a corona wire. I have discovered that such non-uniformity in negative charging can be eliminated by imparting suitable motions to the corona generating wires and this is one important basis of my invention.
It is, therefore, an object of my invention to provide improved means and methods of negatively charging insulating surfaces.
It is a further object to provide improved methods and apparatus for uniformly charging xerographic plates to a negative potential.
It is still a further object to provide improved methods and apparatus for the polarity reversal of xerographic latent images.
The nature and scope of my invention will now be set forth in greater detail in the following specification and drawings in which:
Fig. 1 is a digrammatic representation of an electrostatic charging apparatus;
Fig. 2 is a side view of the corona unit shown in Fig. 1;
Fig. 3 is a side view of one form of corona unit according to my invention; 7
Fig. 4 is a side view of an improved corona unit according to another embodiment of my invention;
Fig. 5 is a side view of another form of corona unit according to my invention; 7
Fig. 6. is a vertical section taken through the deviceof Fig. 4 taken on the line 55;
ofwhich is a narrow slit 3. The assembly composed of parts 1, 2, 3, and any auxiliary parts is commonly referred to as a corona unit and is indicated as assembly 4. A power supply 5 supplies the required voltages to the corona unit. Beneath the slot 3 is a xerographic plate 6 composed of a photoconductive insulating layer 7 as, for example, of vitreous selenium, on a grounded rigid conductive support member 8 such as aluminum, brass, conductive glass or the like.
The xerographic plate is moved past the corona unit by rollers 9 which are driven through a chain of spur gears 10 by electric motor 11. Additional rollers 12 are incorporated to support the plate. The corona wire 1 is commonly a length of .0035 inch stainless steel wire and corona shield 2 commonly has dimensions of about 1 inch by 1 inch. Spacing between slit 3 and the photoconductive insulating surface 7 is commonly inch and the rate of travel of layer 7 past slit 3 is commonly about 2 /2 inches per second.
Fig. 2 is a side view of the corona unit 4 of Fig. 1.
Corona wire 1 is supported between insulator 13 and feed-through insulator 14.
Fig. 3 illustrates one form of corona unit according to my invention. It may be incorporated into apparatus of the type shown in Fig. 1 in place of the corona unit 4 comprised of elements 1, 2, and 3 shown in Fig. 1. Referring again to Fig. 2 the corona wire 1 is in the form of a continuous loop supported between pulleys 15 and 16. The corona shield 2 is of the same form as shown 'in Fig. 1 except that there is a hole at either end to permit the corona wire 1 to pass through. Slit 3 is the same as in Fig. 1. Those portions of corona wire 1 that lie outside of shield 2 are enclosed in a conductive shield'17 which is electrically connected to the corona wire 1 by a contact finger 18. This shield suppresses the emission of corona from those parts of the corona wire which are not in position to participate in charging the photoconductive insulating surface. Pulley 15 is rotated by electric motor 19 acting through belt 20 and corona wire 1 is accordingly kept in motion. A corona reducing rounded lip 21 is shown at the aperture through which the belt passes.
Fig. 4 is a modified version of the corona unit shown in Fig. 2. Corona wire 1 is in the form of a continuous loop but lies entirely within the corona shield 2. It passes over pulleys 15 and 16 and further passes under idler pulleys 22 which serve to keep the two straight parallel sections of the wire in closer proximity. The corona wire 1 is kept in continuous motion by electric motor 19 acting through belt 20 on pulley 15. A contact finger 23 mounted on feed-through insulator 24 contacts the corona wire 1 and provides means for making electrical connection to corona wire 1. Pulleys 15, 16, and 22 are either constructed of insulating materials, as shown, or may be mounted on insulating supports.
Fig. 5 is a side view of another form of corona unit according to my invention. A continuous loop of corona wire 1 passes around pulleys 25, 26, 27 and 28 and is kept in motion by an electric motor 19 acting through pulley 25. A length of insulating shaft material 29 serves to electrically insulate motor 19 from the high potential applied to wire 1. Insulating blocks 30 insulate the corona shield 2 from the high potential of wire 1. Electrical connection to the corona wire is made through feedthrough insulator 24 and contact finger 23.
Fig. 6 isa section view of Fig. 5. It more clearly depicts the relation of corona wire 1 to slit 3.
Fig. 7 shows another form of corona unit in which corona wire 1 is rotated. Corona wire 1 is attached at either end to gears 31 and 32 which are supported by bearings 33 and 34. Bearing 33 is mounted to shield 2 on an insulating block 35 but bearing 34 is attached to a metal terminal rod 36 which passes through shield 2 but is electrically insulated from it by insulator 37. Electric motor 19 drives an insulating shaft 38 which is supported in bearing 39 and carries gears 40 and 41 which mesh with gears 31 and 32 respectively, thus rotating the co rona wire 1. The corona wire is shown as being rotated simultaneously from both ends because it is commonly a very fine wire and lacks the mechanical strength required if it were to be rotated from one end only.
Fig. 8 represents another form of corona unit according to my invention in which the corona wire is caused to execute a reciprocating motion. Corona wire 1 is attached at one end through spring 42 to feed-through insulator 43 which serves as an electric terminal for wire 1. At its other end corona wire 1 is connected by means of a knot 44 to a length of insulating string 45 which passes through a guide bushing 46 set in the wall of shield 2. String 41 is attached to an iron rod 47 the opposite end of which is connected to a spring 48. Iron rod 47 lies partially within an electrical solenoid coil 49 which is connected to a source of alternated current 50. The alternating current acting through solenoid 49 intermittently attractsrod 47 thus causing corona wire 1 to move rapidly back and forth at a rate determined by the frequency of alternating current and with amplitude determined by the strength of the current and the characteristics of springs 42 and 48.
Fig. 9 is a schematic diagram of a power supply adapted to furnish the required corona and shield potentials to the devices of Figs. 1 through 7. Step up transformer 51 multiplies the input voltage many times and this multiplied voltage is converted to pulsating direct current by thermionic rectifier tube 52. Inductance 53 and capacitors 54 comprise an electric filter to smooth out the fluctuations in the pulsating current furnished by the rectifier. A reversing switch 55 makes it possible to supply bleeder resistor 51 with voltage of either positive or negative polarity. The corona potential and shield potentials are taken from suitable taps on bleeder resistor 56.
To charge a xerographic plate according to my invention, the plate 6, composed of a photoconductive insulating layer 7 on metal plate 8, as shown in Fig. l, is laid on rollers 9 and 12 and motor 11 energized to move the plate past slit 3 while power supply 5 is simultaneously energized to supply necessary potentials to the corona unit. The corona unit can be of the form shown in Fig. 3, where motor 19 is energized at the same time as motor 11 and power supply 5 in Fig. 1. Motor 19, acting through belt 20, rotates pulley 15 which causes continuous motion of corona wire 1 which is stretched in a continuous loop between pulley 15 and pulley 16. Shield potential is fed directly to shield 2 while corona potential is fed to shield 17 and then through contact finger 18 to corona wire '1, causing corona emission from corona wire 1.
Within moderately broad limits the dimensions of the corona unit according to my invention are not critical. The minimum diameter of corona wire 1 is set by considerations of mechanical strength and the maximum diameter by the fact that the voltage required for corona discharge increases with increasing wire diameter and approaches that required for sparking. The corrosive nature of corona imposes the further requirement that the corona wire be corrosion resistant. A .0035 inch stainless steel wire operated at a potential of 5000 volts with respect to shield 2 was found satisfactory. The corona shield 2 may be constructed of any rigid conducting material such 'as steel or aluminum sheet and is typically, although not necessarily, of square cross section. In the apparatus of Fig. 3 the shield dimensions were 1 inch by 1 inch. The shield potential is of the same polarity as the corona potential and is preferably as large as can be maintained without risk of causing spark or other discharge between the shield and the plate surface. For a shield to plate spacing of inch, a voltage of 2500 volts between the shield and the plate support member 8 was found satisfactory. Slit 3 should have a uniform width great enough to permit only sufficient ions or electrons to escape from the shield 2 to charge the plate in a reasonable length of time. A slit width of inch was found satisfactory. The length of slit 3, and therefore of shield 2, is determined solely by the width of the surface to be charged.
In order to more fully understand my invention it is necessary to consider the characteristics of positive and negative corona discharges as they affect the charging of insulating surfaces. It is believed that there is a certain small amount of free electrons and positive ions normally present in air and that when a sufliciently high positive potential is applied to a wire, the surrounding free electrons move towards the wire with suficient velocity to ionize gas molecules which they strike while traveling towards the positive wire, thus creating additional positive ions and electrons. The newly created electrons themselves are accelerated towards the corona wire, collide with gas molecules, create still more ions and electrons, and so forth in like manner. As a result of this process the Wire becomes surrounded by a sheath of electrons and positive ions and some of the latter, repelled by the positive potential on the corona wire, diffuse away from the wire and can be made to strike a nearby insulating surface, thereby giving the surface a positive electrostatic charge. The corona wire itself plays essentially no part in the corona generating process other than providing the necessary electric field. Variations in wire diameter will, according to the laws of electrostatics, vary the surrounding electric field strength and therefore the rate of corona generation, and isolated points or other surface imperfections in the wire will create locally high electric fields near the wire but these points produced field anomalies exist primarily close to the wire surface, while the corona generating process occurs in a sheath extending some distance from the wire. Because positive corona production is relatively independent of the exact nature of the corona wire by which it is generated, it is possible to get relatively very uniform positive corona emission along the surface of wire of only commercial grade of surface finish. Y
The situation with negative corona is entirely different. Negative corona sets in at a lower voltage than positive corona, when pre-existing positive ions around the corona wire are accelerated towards the wire and in striking it release electrons which are accelerated away from the wire, strike gas molecules, and create more electrons and more positive ions, which are accelerated towards the corona wire and continue the process. The rate at which the wire releases electrons when struck by positive ions is very much a characteristic of the wire material and the exact state of the wire surface. This relation between impinging ions and released electrons is probably dependent on such factors as dirt spots on the wire, areas of oxidation, variations in the crystal structure of the wire and the like, but whatever the true causes may be, or even whether the theoretical explanation given above is correct, it is an experimental fact that as the negative voltage on a small Wire is increased corona discharge commences at discrete points along the wire. As the voltage is further increased corona discharge generally spreads along the surface of the wire but never becomes adequately uniform. This non-uniformity may be reduced by operating the wire at a potential well above that at which negative corona commences and by using a clean, smooth wire, but it is virtually impossible to eliminate it sutficiently for uniform charging of a xerographic plate.
When negative corona discharge is used to negatively charge an insulating surface, such as that of a xerographic plate, the non-uniformity of the corona discharge is reflected in non-uniform charging of the plate surface. This is particularly true if the apparatus used is of the type shown in Figures 1 and 2. In the corona unit shown there the corona wire is almost entirely surrounded by the shield 2 with the result that the rate of corona generation is almost entirely independent of such factors external to the corona unit itself as the pre-existing potential on the surface being charged and that electrons travel in substantially straight lines from the wire to the surrounding shield. Those electrons which emerge through slot 3 are urged toward the surface of the plate by the strong electric field which is maintained between shield 2 and the plate and it has been found that these electrons travel in substantially straight and direct paths from the slit 3 to the plate. Because the electric field maintained between the shield 2 and the plate surface by the voltage applied to the shield is much greater than the fields which may be produced by electrostatic charges on the plate surface, electrons travelling from slit 3 to the plate surface are substantially unaffected by the presence of prior charge on the plate surface. It follows from this that the density of charge deposited on the plate by the corona unit depends only on the rate of corona production and the speed at which the plate moves past the corona unit and such, in fact, is one of the essential purposes and objects of the hereinbefore mentioned Hayford and Haacke invention. It also follows from the straight line paths of the electrons that each element of area on the plate surface receives charge from the corona produced at a small area of the corona wire facing slit 3. Conversely, each element of area on the corona wire facing the slit causes charge to deposit on a small area only of the plate surface at any given time. Thus a point on the wire which emits an unusually strong corona will cause a line to be formed on the plate, parallel to the direction of relative motion between the plate and the corona unit, which is more strongly charged than adjacent areas. This effect does in fact occur and it had been found that xerographic plates charged negatively by the apparatus of Figure I tend when developed to exhibit undesirable parallel streaks corresponding to streaks of non-uniform charge deposition.
It will be understood that the apparatus of Figure 1 is particularly susceptible to this form of image streaking. Prior art corona units such as those now commercially in use in Xerography do not have the narrow slit 3 nor the high shield to plate potential of the unit shown in Figure l and therefore permit more mixing and diffusing of the electrons as they travel from the corona wire to the plate surface. In these prior art corona units there is also a tendency to charge the plate to a preselected potential and then stop the flow of electrons to the plate, making the final plate potential quite independent of local or general variations of corona emission from the corona wire. Nevertheless, as was pointed out earlier, the new method of negative to positive xerographic reproduction disclosed in the hereinbefore mentioned application of Richard E. Hayford and Alfred C. Haacke, requires the step of negatively charging a xerographic plate with apparatus of the type shown in Figure 1. The problem of non-uniformity in negative charging is thus seen to have been a real one, requiring a novel solution.
I have found that it is possible to average out the nonuniformity of negative corona emission from a wire and eliminate the corresponding non-uniformity of charge deposition by using a moving corona wire. By moving the wire at a sufficient speed every point on the plate will receive an increment of charge proportional to the average corona emission over a reasonable length of corona wire. The critical quantity in determining the necessary wire speed is the length of time during which a point on the moving plate receives charge from the corona unit or, equivalently, the width of the charge pattern deposited on the plate at any instant. If the width of the charge pattern is 1 a typical figure, and the wire moves with respect to the shield at 16 times the rate at which the plate moves with respect to the corona unit then the wire will have moved one inch during the time that any element of the plate surface was exposed to the corona discharge. As the average corona emission over anyone inch length of the corona wire is substantially the same as the corona emission averaged over any other one inch length, all points on the plate surface will receive substantially the same amount of charge. In more general terms it is necessary forthe corona wire to move a reasonable distance in the length of time required for an element of plate surface to cross the width of the charge pattern from the corona unit. The width of this pattern can be readily determined by momentarily energizing the corona unit with a stationary plate beneath it, developing the plate, and measuring the width of the pattern of powder deposition on the plate, which is the same as the width of the charging pattern from the .corona unit. Insulficiently rapid motion of the corona wire may be recognized by the appearance of diagonal streaks in a developed xerographic image.
The corona unit shown in Figure 3 is a straightforward application of the principles outlined above to the corona unit of Figure 1. Figures 4, 5, and 6 differ somewhat in that the active corona generating element consists of two parallel lengths of wire traveling in opposite directions. In Figure 8 a reciprocating motion of the wire is used rather than a continuous unidirectional motion. The wire can attain a high instantaneous velocity, but the maxi mum wire travel is restricted. In Figure 7 the wire is rotated rather than set in linear motion. Either form of motion is effective in eliminating non-uniformity of charging because, as was previously explained, only the corona coming from that portion of the wire facing the slit in the corona shield is efiective in charging the plate. If the wire rotates at a speed such that it makes at least one revolution in the time that it takes any element of plate surface to cross the width of the charge pattern from the corona unit then the charge deposited on the element of surface is proportional to the average intensity of corona discharge around a circumference of the wire. While the figures show either linear or rotary motion of the corona wire it is to be understood that both types of motion can be combined.
I have hitherto described my invention in terms of a particular type of corona unit because that type presents most clearly the difiiculties which my invention overcomes, but it Will be understood that my invention is also applicable to other forms of corona units.
While my invention has been shown with only one corona wire, more can be used as will be apparent to those skilled in the art. While my invention has been shown with a stationary corona unit positioned above a moving plate it will be understood that the corona unit may also be moved past a stationary plate. My invention is particularly valuable for use with negative corona charging, but it can also be used for positive corona charging. My invention has been shown in use for charging a xerographic plate composed of a vitreous selenium photoconductive insulating layer on a rigid conductive support member, but it is not intended to be limited to such an element. Other photoconductive insulating materials may be used such as, sulphur, anthracene, various alloys of selenium as with tellurium, arsenic and the like, photoconductive phosphors in insulating resin binders, and the like. The support member may be rigid or flexible, conductive or insulating, examples being metal plates and flexible metal foils in sheet or roll form, paper sheets and rolls, rigid plastic sheets, plastic film sheets or rolls, or the like. The support member may be flat,
curved cylindrical, or other shape. If the photoconductive insulating material has sufficient strength to be selfsupporting the support member may be dispensed with entirely.
Xerographic methods have been discovered whereby electrostatic latent images may be formed on non-photoconductive insulating materials such as plastic films and my invention may be used for charging such materials also. While my invention has been described in terms of its usefulness in the art of Xerography, its usefulness is not limited to that art and it may be used for the uniform deposition of electric charge on any insulating surface whether photoconductive or not.
The foregoing specification is intended to be broadly interpreted, limited only by the scope of the appended claims.
1. A corona discharge electrode for depositing on a surface to be charged uniform negative electric charge, said electrode comprising at least one corona discharge wire and a conductive shield substantially completely surrounding said wire on more than three sides, the wire comprising at least one fine uniform electrically conductive strand insulated from said shield and adapted to receive a corona generating potential, an elongated opening in the shield substantially parallel with the corona discharge wire and of uniform width along its length, means to apply a corona generating potential to the corona discharge wire with respect to the surrounding shield, means to cause uniform relative motion between the electrode and the surface being charged in a direction across the longitudinal position of the corona discharge wire, and means to move the corona discharge wire longitudinally along its own length and across the path of motion between the electrode and the surface being charged.
2. A corona discharge electrode for depositing on a surface to be charged uniform increments of charge density, said electrode comprising at least one corona discharge wire and a conductive shield substantially completely surrounding said wire on more than three sides, the wire comprising at least one fine uniform electrically conductive strand insulated from said shield and adapted to receive a corona generating potential, the corona discharge wire being moveable longitudinally within the shield, an elongated opening in the shield substantially parallel with the corona discharge wire and of uniform width along its length, means to apply a corona generating potential to the corona discharge wire with respect to the surrounding shield and means to supply a field potential to the shield with respect to the surface being charged said field potential being at least ten times the desired potential to which said surface should be charged, and means to cause uniform relative motion between the electrode and the surface being charged in a direction across the longitudinal position of the corona discharge wire.
3. A corona wire charging device comprising a corona discharge wire electrode, means to support said wire elec trode adjacent and parallel to an insulating surface of a layer to be charged, means to move said wire electrode relative to the insulating surface and in a direction transverse to the length of said wire electrode, and means during said transverse relative movement to impart additional movement to said wire electrode in a direction differing from said transverse movement while maintaining said wire electrode parallel to and at a constant distance from an insulating surface.
4. Apparatus according to claim 3 in which said means to impart additional movement causes axial wire rotation.
5. Apparatus according to claim 3 in which means to impart additional movement to said wire electrode causes movement parallel to its length.
6. Apparatus according to claim 3 in which said means to impart additional movement to said wire electrode to said wire electrode 9 comprises means to cause reciprocal movemnet of said wire electrode along its own length.
7. Apparatus according to claim 3 including a shield of conductive material substantially completely surrounding said wire electrode on more than three sides and presenting a narrow slit opening toward the surface to be charged, and means to apply a negative corona discharge generating potential to the corona discharge electrode.
8. Apparatus according to claim 7 in which said means to impart additional movement to said wire electrode comprises means to cause said wire electrode to reciprocate along its own length.
9. Apparatus according to claim 7 in which said means to impart additional movement to said wire electrode comprises means to impart Wire axial rotation.
10. Apparatus according to claim 7 in which said means to impart additional movement to said wire electrode comprises means to cause longitudinal movement of said wire electrode along its own length.
References Cited in the file of this patent I UNITED STATES PATENTS 2,551,582 Carlson May 8, 1951 2,576,047 Schattert Nov. 20, 1951 2,684,902 Mayo et a1 July 27, 1954