US3264592A - High voltage transformer - Google Patents

Description

Aug. 2, 1966 P. A. PEARSON 3,264,592

HIGH VOLTAGE TRANSFORMER Filed May '7, 1962 3 Sheets-Sheen- 1 PAUL A. PEARSON INVENTOR.

RNEYS Aug. 2, 1966 P. A. PEARSON men VOLTAGE TRANSFORMER 3 Sheets-Sheet 2 Filed May '7, 1962 3 um 1 I 2 3 0OQ\\ I 3 3 ob lite :0

PAUL A. PEARSON INVENTOR,

ATTORNEYS Aug. 2, 1966 Filed May 7, 1962 P. A. PEARSON 3,264,592

HIGH VOLTAGE TRANSFORMER 5 Sheets-Sheet 5 PAUL A. PEARSON INVENTOR.

gffll ATTORNEYS United States Patent 3,264,592 HIGH VOLTAGE TRANSFORMER Paul A. Pearson, 1200 Bryant Sh, Palo Alto, Calif. Filed May 7, 1962, Ser. No. 192,855 1 Claim. (Cl. 336-182) This invention relates generally to transformers and more particularly to high voltage transformers having broad band frequency and good pulse response.

Prior art transformers of this type are rather large, performance is, in many instances, sacrificed for life and the transformers are relatively expensive.

It is a general object of the present invention to provide an improved wide band and/or pulse transformer.

It is another object of the present invention to provide a transformer having at least one tapered winding to reduce leakage inductance and provide protection against high voltage breakdown.

It is a further object of the present invention to provide a transformer in which electric field concentrations are minimized.

It is another object of the present invention to provide a transformer having minimum solid dielectric material in the electric fields.

It is a further object of the present invention to provide -a broad band and pulse transformer which includes multiple windings.

It is a further object of the present invention to provide a transformer suitable for series connection to other transformers for high voltage applications.

It is still a further object of the present invention to provide a high voltage pulse transformer having a biased core to reduce the size of the transformer.

These and other objects of the invention will become more clearly apparent from the following description taken in conjunction with the accompanying drawings.

Referring to the drawings:

FIGURE 1 is a perspective view showing a transformer in accordance with the invention;

FIGURE 2 is a sectional view of the transformer of FIGURE 1;

FIGURE 3 shows parallel connection of the transformer windings shown in FIGURE 1;

FIGURE 4 schematically shows a transformer having two serially connected sections disposed on a common core;

FIGURE 5 shows a plurality of transformers connected in series; and

FIGURE 6 is a schematic diagram of the serially connected trans-formers shown in FIGURE 5.

Conventional high voltage transformers are, in general, provided with primary and secondary windings, each of which include a plurality of winding turns or convolutions. The convolutions are arranged to form cylindrical primary and secondary windings which are nested one within the other. Generally, the spacing between the primary and secondary windings is uniform along the length of the winding and is such as to accommodate the highest voltage which exists between the primary and secondary windings.

As is well known, leakage inductance in a transformer increases with increased spacing of the primary and secondary windings. As a consequence, there is substantial leakage inductance in transformers of the prior art which have uniform spacing between the primary and secondary windings along the same.

In accordance with the present invention, there is provided a transformer in which the high voltage winding is tapered with the winding being relatively close to the other winding at one end and spaced therefrom at the other end of the same. The closely spaced end is connected to points at nearly the same potential; for example, both could be connected to ground. Since, at this point, the voltage between the windings is small, the spacing may be made relatively close without the danger of voltage breakdown between the windings. Away from this point along the windings, the spacing is increased to accommodate the higher voltages which are encountered. The leakage inductance between the primary (generally low voltage) and secondary (generally high voltage) windings is substantially reduced because the average spacing between windings is reduced.

Referring more particularly to FIGURE 1, there is shown a transformer including a base 11 which carries a pair of spaced parallel angle brackets 12 and 13. Secured to the angle brackets there are provided two pairs of spaced winding support members 14. The support member 14a of each pair carries terminal posts 16 which are connected to opposite ends of the low voltage winding turns. The other support member 14b of each of the pairs includes a terminal post 17 which is connected to one end of the high voltage winding. In FIGURE 1, there is only shown one such post, it being understood that the other member 14b (not shown in the figure) includes a similar post connected to one end of the corresponding high voltage winding. A pair of spaced members 18 cooperates with each pair of the spaced members 14 to form an open-ended cylinder which accommodates the core 19. The open-ended cylinder supports the convolutions 21 of the low voltage wind-ing. As illustrated in FIGURE 1, the low voltage winding includes multiple winding turns or convolutions with the opposite ends of each connected to one set of terminals 16. The material forming the winding support may be any suitable insulating non-magnetic material, for example, the material may be paper base phenolic sheet.

The core 19 may be formed in various ways. However, as illustrated, the core includes two U-shaped core sections 22 and 23 which cooperate to form a closed core 19. The core 19 is supported on cross members 24 secured to the spaced angles 12 and 13. Straps 25 encircle the core and the support members 24 and serve to seat the core on the supports 24.

The high voltage windings 31 are wound upon an open tapered dielectric support 32 with adjacent convolutions extending along the outside surface of the support as shown in FIGURES 1 and 2. As previously described, the lower end of the high voltage windings is relatively close to the low voltage windings. This is possible since the end of the windings can be connected to substantially the same voltage. The voltage between the low voltage and high voltage windings increases from the bottom to the top of the windings. The high voltage winding is tapered so that the spacing between the windings increases to withstand the higher voltages.

The lower end of the support structure 32 is notched as indicated at 33 to provide space for circulation of dielectric fluid as will be presently described.

Since the transformer is adapted to operate at relatively high voltages, electric field concentrations are reduced wherever possible. In accordance with the present invention, the upper convolutions of the high voltage winding 31 which is subjected to the highest voltage and which would present a relatively sharp edge in comparison to the remainder of the winding convolutions are partially enclosed in an enlarged hollow conductor 34. This conductor is provided with a relatively large radius for reducing the concentration of electric field lines. Arcing or shorting at the secondary is substantially eliminated. The hollow conductor is connected to the last convolution of the high voltage winding and may itself form the last winding turn. In any event, this conductor has its path broken whereby it does not present a shorted turn.

By extending the low voltage winding a substantial distance above the hollow conductor or ring 34, the field concentration at the low voltage or primary winding is reduced to safe limits. Furthermore, the low voltage winding looks much like a plane surface to the outer wind ing and no high electric fields exist at the top turns because of the proximity to the outer windings. Usually, the inner winding is a low voltage, high current winding and large size wire is used. The low voltage and large diameter contribute to the low electric fields and, therefore, no special treatment of the top wire is required.

As viewed in FIGURE 2, the large diameter hollow conductor or rings for the two high voltage windings cooperate with one another to form a conductor extending through the core opening or window 36. This conductor is surrounded by the core and primary windings. The relatively large conductor forms in essence a center conductor while the low voltage windings and adjacent core material form an outer or surrounding conductor. It is well known that there is an optimum ratio of inner to outer diameters for coaxial structures to provide minimum electric field at the inner conductor. The electric field varies slowly with change in ratio of the center conductor to the approximate radius of its low voltage surroundings so that the exact diameter is not critical. However, the relatively large diameter of the hollow conductor serves to greatly reduce the electric fields at the top of the high voltage winding.

Due to the relatively high voltages between the high and low voltage turns, it is desirable to reduce any possible electric field concentration points. Such field points may occur at the low voltage winding edges 37. In accordance with the present invention, these edges are rounded to present a rounded edge and minimize electric field concentration. This again is comparable to a coaxial structure and the ratio of radii again is of importance and can be optimized.

As described, the windings are wound on open forms. The insulation between windings and convolutions can be an oil or gas dielectric. The capacitance is relatively low since the dielectric constant of oil or gas is considerably less than for solid insulating materials between the primary and secondary windings. Furthermore, it gives more eflicient cooling since the oil and gas can flow easily through and around the windings and through the slots 33.

The forms can be dispensed with entirely if a plastic support is formed as a web between windings. The plastic web will not project into high electric field regions and thus there would be no solid dielectric in the high electric field regions. An alternative is to place the supporting structure outside the windings where the electric fields are normally much lower than in the inside.

Pulses applied to the transformer are usually unidirectional with a flux density in the core starting from zero or a low value and increasing to a value dependent on the pulse time and duration but limited by saturation of the core. By passing a current through one of the existing windings or through a winding specifically for the purpose, the flux density can be started from a value near saturation in the reverse direction. This effectively doubles the swing available for each pulse. The result is for given performance specifications of a pulse transformer, a much smaller transformer can be built, or conversely, fora given size, considerably better performance can be obtained. Generally, there is provided an isolation inductance in series with the biasing current supply to prevent the current supply from short circuiting the pulse voltages.

The windings of the transformer are shown in FIG- URE 1 and occupy both legs of the core. The windings are identical but wound in reverse directions from each other to give the same voltage polarity at adjacent ends. Corresponding terminals of the winding may be connected for parallel operation. The connection is illus- 4 trated in FIGURE 3. Parallel operation reduces the leakage inductance. However, the capacitance increases. This connection can be advantageous in meeting a set of performance characteristics, i.e. in controlling the inductance and capacitance of the transformer to give optimum rise-time characteristics or high frequency response.

The two windings may be connected to be electrically in parallel for pulses but separate for carrying low frequency current such as a heater power for cathodes of an associated electron tube. Paralleling in this instance is achieved by using coupling capacitors 38. The capacitors maintain the two windings at nearly the same pulse or high frequency voltage for corresponding points along the windings. The capacitors are selected to have high enough impedance so that negligible D.-C. or low frequency current is drawn by the capacitors.

By modifying the secondary Winding, the transformer may be connected so that the sections on each of the legs of the core are operated in series. This is schematically illustrated in FIGURE 4 wherein the input is applied to the primary 41 which cooperates with a secondary winding, schematically illustrated at 42, which has a large conductor 43. The large conductor 43 is connected to the lower end of the secondary 45 disposed on the other core leg, which secondary cooperates with the primary 44. The secondary 45 is spaced from the primary at its lower end a distance which corresponds to the voltage between the primary and secondary windings. There can be provided an enlarged rounded conductor 46 for reducing electric field concentration if conditions warrant. The upper end of this secondary is spaced from the primary and includes an enlarged rounded conductor 47. The spacing between the conductor 43 and the adjacent secondary 45 of the second section is also maintained so as to prevent voltage breakdown between the transformer sections. The spacing of the secondary winding 45 is still tapered but does not taper to zero or near zero but to a value corresponding to the voltage difference at the bottom and outwardly to accommodate larger voltages at the top.

In very high voltage, high impedance applications, the capacitance of the transformer can be of such magnitude that the energy necessary to charge it becomes an appreciq able portion of the total useful energy. This is the case, for instance, in transformer designs for one or two million volts pulse and load resistance in the order of tens of thousands of ohms. This situation may be alleviated by dividing the transformer into a plurality of individual sections, each section supplying a fraction of the total voltage with the various transformer sections connected in series, each fed by energy from the next unit below it in voltage. In this way, no one transformer has to have in: sulation for more than its particular fractional part of the total voltage.

FIGURE 6 schematically shows four units, as an example, stacked in series. FIGURE 5 schematically shows the physical arrangement of transformers for accomplishing this connection. The individual transformers may be of the type described above. The first transformer includes a core 51, a primary winding 52 cooperating with the core, and a tapered secondary winding 53. The transformer has in addition a low voltage, low impedance secondary winding 54. For convenience, it can be a one turn secondary. This winding is riding at the high voltage with respect to the primary. It is contoured at the outer edge with a large diameter conductor 55 as well as the edge that meets the high voltage winding. Since the voltage any where on this secondary is nearly the same as for the top winding of the secondary, the spacing between the secondary and the low voltage secondary is tapered just as the spacing between the primary and secondary. The primary 56 of the second transformer is fed from the winding 54 by a low inductance transmission line. The low inductance transmission line minimizes contribution to leakage inductance of the low voltage winding. The leakage inductance of the low impedance secondary contributes greatly to the overall leakage inductance of the transformer and affects its operation. The second transformer includes a secondary winding 58 and a low voltage secondary winding 59. Again, this is connected by conductors 61 to the next transformer in series.

Referring to FIGURE 6, there is shown the connection of a plurality of such transformers. It is seen that the secondaries 53, 58, etc., are connected in series to provide a relatively high voltage between the output terminals 62 and 63. Each of the transformers above the first transformer are established at their respective voltages whereby the voltages between the windings remain low even though the transformers themselves are at a relatively high voltage.

Series transformers, as explained above, reduce capacitance in two ways. First, the individual transformer each has less capacitance than one large transformer having similar overall characteristics. Secondly, the individual transformer capacitances are connected in series as are the transformers so that the effective capacitance is greatly reduced.

Leakage inductance is increased because the leakage inductance of the individual units is added in series. In addition, the leakage inductance of the low 'voltage secondaries is also added. For instance, the transformer next to the top unit, FIGURE 6, has a certain leakage inductance between its primary and low voltage secondary. This leakage inductance is, in effect, in series with the output of both transformers at the top. Its contribution to the overall leakage inductance is then four times its actual leakage inductance. Similarly, the third unit from the top contributes nine times its leakage inductance to the whole. For this reason, it is important to make the leakage inductance of the low voltage secondary as low as possible as well as keeping the inductance of the transmission line interconnecting the transformers low.

This series of transformers will have characteristics of a transmission line. The reason for this is that each secondary is, in effect, a resonant circuit having inductance and capacitance. Furthermore, these resonant circuits are coupled to each other in succession. Any succession of coupled resonant circuits can support travelling waves. A sharp rising pulse applied to the primary of one is propagated through the series of transformers. If the transformers are not properly terminated when it reaches the top, it reflects back to the start of the bottom transformer and the resulting reflections affect the output waveform of pulses or high frequency response. It is possible to add at least a partial termination at the top transformer such as illustrated by the resistor 65 and capacitor 64. Since the higher frequencies cause the most trouble, the termination can be effective for higher frequencies without causing excessive losses.

Thus, it is seen that there is provided a transformer used singly or for series operation which is relatively compact and which is capable of withstanding relatively high voltages while providing broad band frequency and good pulse response.

I claim:

A high voltage transformer combination comprising a plurality of sections each including a winding pair, each section comprising a core, an elongated primary winding including a plurality of adjacent convolutions each having substantially the same diameter wound on said core, an elongated secondary winding including a plurality of adjacent convolutions surrounding each of the primary windings to form a winding pair, the convolutions of each of said secondary windings increasing progressively in diameter from one end to the other to provide a tapered winding so that the spacing between the primary and secondary windings increases from one end of the windings to the other of each of said winding pairs, conductive means carried adjacent the large end of the secondary winding of each winding pair and connected to the last convolution for reducing the concentration of electric field at said end of the same, a low voltage secondary winding surrounding the secondary winding of each of said winding pairs to form a section; the low voltage secondary winding of the first section being connected to the primary winding of the next section and the low voltage winding of each successive section being connected to the primary winding of the next successive section, and means connected to the secondary winding of the first section and of the last transformer section of the transformer combination to provide a high voltage output.

References Cited by the Examiner UNITED STATES PATENTS 891,496 6/1908 Luschka 336-185 X 1,062,046 5/1913 Smith 336-185 X 1,585,448 5/1926 Weed. 1,816,448 7/ 1931 Terman 336-224 2,937,349 5/1960 Camilli 336- X 3,026,492. 3/ 1962 Narbut 336-70 OTHER REFERENCES Bozorth, R. M.: Ferromagnetism, D. Van Nostrand Co., Inc., Princeton, New Jersey, 1951 (pp. 538-553). LAILAMIE E. ASKIN, Primary Examiner.

JOHN F. BURNS, Examiner.

W. M. ASBURY, Assistant Examiner.

Images