Film Capacitor

Abstract

A film capacitor is disclosed. In an embodiment a film capacitor includes a film comprising a blend of polypropylene and cyclo-olefin copolymer, wherein the blend includes an amount of at least two thirds by weight of polypropylene, and wherein the cyclo-olefin copolymer includes ethylene in a range of 23 weight % to 27 weight % inclusive and norbornene in a range of 73 weight % to 77 weight % inclusive.

Claims

1-11. (canceled)

12. A film capacitor comprising: a film comprising a blend of polypropylene and cyclo-olefin copolymer, wherein the blend comprises an amount of at least two thirds by weight of polypropylene, and wherein the cyclo-olefin copolymer comprises ethylene in a range of 23 weight % to 27 weight % inclusive and norbornene in a range of 73 weight % to 77 weight % inclusive.

13. The capacitor according to claim 12, wherein the blend comprises polypropylene in a range of 70 weight % to 90 weight % inclusive.

14. The capacitor according to claim 12, wherein the blend comprises polypropylene in a range of 78 weight % to 82 weight % inclusive.

15. The capacitor according to claim 12, wherein the polypropylene is a capacitor grade polypropylene.

16. The capacitor according to claim 12, wherein the film is metallized.

17. The capacitor according to claim 12, wherein the film is extruded and biaxially-stretched.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0016] FIG. 1 shows a film capacitor.

[0017] FIG. 2 shows a flow diagram representing the method for manufacturing the film of the film capacitor.

DETAILED DESCRIPTION OF ILLUSTRATIVE EMBODIMENTS

[0018] Film capacitors are electrical capacitors with an insulating plastic film as the dielectric. FIG. 1 shows a capacitor 1 comprising a dielectric film 2 which has been metallized on one side. The metallization forms an electrode 3 of the capacitor 1. The electrodes 3 of the film capacitor 1 may be metallized by applying a metal or an alloy of metals on the surface of the film. In particular, the electrodes 3 may comprise aluminium, zinc, gold, silver, magnesium or any appropriate alloy of these materials.

[0019] In the film capacitor 1 shown in FIG. 1, the films 2 are stacked on one another. Alternatively, two of the films 2 can be wound into a cylinder-shaped winding to form the capacitor 1. The winding can further be flattened into an oval shape by applying mechanical pressure.

[0020] The electrodes 3 are contacted by a contact layer 4, which is also referred to as schoopage. Moreover, the film capacitor 1 comprises terminals for electrically contacting the capacitor 1.

[0021] The film 2 consists of a blend of polypropylene and cyclo-olefin copolymer wherein the cyclo-olefin copolymer consists of ethylene and norbornene. The polypropylene has a greater percentage than the cyclo-olefin copolymer by weight of the blend. In particular, the polypropylene has a percentage by weight of two-thirds or more.

[0022] In the following, the manufacturing method for manufacturing the film 2 comprising a blend of polypropylene and cyclo-olefin copolymer is described. FIG. 2 shows a flow diagram representing the method for manufacturing the film. The manufacturing method uses state-of-the-art customary processes.

[0023] In a first step A, the polypropylene and the cyclo-olefin copolymer are blended together to form the blend. In the subsequent step B, the blend is melted and mixed to form a molten polymer. In the next step C, the molten polymer is filtered to form a filtered molten polymer. In the next step D, the filtered molten polymer is extruded through a flat die to form an extruded capacitor film. In the next step E, the extruded capacitor film is biaxially-stretched to form a biaxially-stretched capacitor film.

[0024] Afterwards, the biaxially-stretched capacitor film is metallized using customary processes. Before the metallizing, the film may be surface-treated by means of corona or flame. The metallization process is preferably carried out by Physical Vapor Deposition (PVD) in vacuum. The metal layer is applied at least on one surface of the film. The metal layer consists of any suitable metal, preferably aluminium, zinc, gold, silver or magnesium or appropriate alloys of the previously mentioned materials. The thickness of the metal layer usually ranges from 10 nm to 100 nm.

[0025] The capacitor has good self-healing abilities up to temperatures of 130 C. The cost of the base material for the film is higher than that of the reference capacitor because of the contribution of cyclo-olefin copolymer. But as the majority contribution to the weight is by polypropylene, the cost of the base material is moderate. As discussed above, the manufacturing process is based on state-of-the-art manufacturing steps. Thus, the production can be carried out in a cost-efficient manner.

[0026] In the following, life test measurements are described which compare capacitors according to embodiments of the present invention to a reference capacitor. The values of capacitance were measured on finished capacitors by a Keysight E4980AL Precision LCR Meter.

[0027] Table 1 provides a list of the capacitors used in the present tests.

TABLE-US-00001 TABLE 1 Blend composition COC composition Capaci- Polypro- Nor- Eth- tance at Loss pylene COC bornene ylene 1 kHz tangent Sample [%] [%] [%] [%] in F at 20 Hz 1 100% 0% 0% 0% 9.56 0.012% 2 80% 20% 75% 25% 9.33 0.016% 3 70% 30% 75% 25% 9.26 0.037% 4 80% 20% 80% 20% 9.79 0.016%

[0028] Sample 1 refers to a reference capacitor which comprises a film that consists only of polypropylene. Samples 2, 3 and 4 refer to capacitors according to embodiments of the present invention which comprise a film containing varying percentages of a commercially available high crystallinity capacitor grade polypropylene resin and complementary percentages of two commercially available cyclo-olefin copolymers of ethylene and norbornene. In particular, the blend according to sample 2 comprises 80 weight % polypropylene and 20 weight % of cyclo-olefin copolymer, wherein the cyclo-olefin copolymer consists of 75 weight % norbornene and 25 weight % ethylene. The blend according to sample 3 comprises 70 weight % polypropylene and 30 weight % of cyclo-olefin copolymer, wherein the cyclo-olefin copolymer consists of 75 weight % norbornene and 25 weight % ethylene. The blend according to sample 4 comprises 80 weight % polypropylene and 20 weight % of cyclo-olefin copolymer, wherein the cyclo-olefin copolymer consists of 80 weight % norbornene and 20 weight % ethylene.

[0029] The blends of samples 1 to 4 have been biaxially-stretched into capacitor films of a thickness of 8 m by customary processes. The films have been vacuum-metallized to obtain a sheet resistance of 20 Ohm/sq. Then, the films have been transformed into metallized film capacitors comprising a rolled, flat-pressed element inside a plastic box sealed with a potted epoxy resin by customary processes identical for all samples.

[0030] Table 1 shows the average values of capacitance for at least 20 capacitors from each sample. The capacitance is measured at 1 kHz. The capacitors according to samples 2 to 4 show a similar capacitance to the capacitor of sample 1 which comprises the film of pure BOPP. The performance of samples 1 to 4 under operational stress caused by temperature and by a DC voltage have been evaluated through two life tests at different temperatures. Namely, the first test has been carried out at a temperature of 120 C. and the second test has been carried out at a temperature of 130 C. Five capacitors per sample have been tested in each life test. The capacitance at 1 kHz and the loss tangent at 1 kHz of the capacitors have been monitored by regular measurements every 160 hours during the test. A capacitor has been considered as failing the test when it showed an irreversible short-circuit. Such a capacitor has therefore been removed from the test after the failure. A failure indicates that the capacitor has failed to self-heal at that point of time.

[0031] Table 2 shows the conditions under which the first life test has been performed.

TABLE-US-00002 TABLE 2 Voltage Steps applied in long-endurance test A DC Voltage in V Field in V/m Time in Hours 2080 260 1000 2240 280 1000

[0032] The test comprised two steps of increasing voltage as described in Table 2. During the first step which took one thousand hours of the test, a DC voltage of 2080 V has been applied, resulting in a field of 260 V per m. During the second step which took the subsequent 1000 hours of the test, a DC voltage of 2240 V has been applied, resulting in a field of 280 V per m.

[0033] Table 3 lists the elapsed times of the test at which irreversible breakdowns affected the capacitors from each sample and gives the estimated mean time to failure (MTTF).

TABLE-US-00003 TABLE 3 Hours to failure in long-endurance test A at 120 C. (in increasing order) Average 1st. 2nd. 3rd. 4th. 5th. MTTF Sample failure failure failure failure failure Hours 1 346 442 541 709 709 549 2 no failure after 2.000 hours of test >2.000 3 no failure after 2.000 hours of test >2.000 4 130 202 322 346 442 288

[0034] It can be gathered from Table 3 that the capacitors according to samples 2 and 3 did not show any failure after 2000 hours of test. The capacitor according to sample 4 showed an average MTTF of 288 hours. The reference capacitors according to sample 1 showed an average MTF of 549 hours. Accordingly, samples 2 and 3 show a mean time to failure that is at least three times higher than that of the reference sample. It might even be much higher than three times because the failures in sample 1 took place at the first 1000 hours of the test and the voltage stress was increased in the subsequent second 1000 hours of the test.

[0035] The second life test was carried at a temperature of 130 C. and comprises four steps of increasing voltage as described in Table 4.

TABLE-US-00004 TABLE 4 Voltage Steps applied in long-endurance test B DC Voltage in V Field in V/m Time in Hours 655 82 168 1000 125 336 1400 175 603 1665 208 448 1960 245 778

[0036] As can be seen in Table 4, the voltage stress has been increased in each phase of the test as a DC voltage of 655 V is applied in a first phase, then the voltage is increased to 1000 V in a second phase, then the voltage is increased to 1400 V in a third phase, to 1665 V in a fourth phase and finally to 1960 V in a fifth phase.

[0037] Table 5 lists the elapsed times of test at which irreversible breakdowns have affected the capacitors of samples 1 to 4.

TABLE-US-00005 TABLE 5 Hours to failure in long-endurance test B at 130 C. and increasing V.sub.dc (in increasing order) 1st. 2nd. 3rd. 4th. 5th. Sample failure failure failure failure failure 1 672 1107 1300 1300 1559 2 no failure after 2.333 hours of test 3 2001 2333 4 1300 1300 1300 1540 1835

[0038] Again, the capacitors according to sample 2 did not show any failure. Only two of the capacitors of sample 3 showed failures. Each of the failures of sample 3 occurred in the last phase of the test. The capacitors according to sample 4 showed failure later than the capacitors according to the reference sample 1.

[0039] Altogether, the life tests show that samples 2 and 3 based on blends containing respective 20 and 30 weight % of a cyclo-olefin copolymer with 75 percent by weight of norbornene and 25 percent by weight of ethylene clearly outperform reference sample 1 which is based on pure BOPP, not containing any cyclo-olefin copolymer. Moreover samples 2 and 3 also outperform sample 4 which is based on a blend containing 20% of cyclo-olefin copolymer with 80% by weight of norbornene and 20% by weight of ethylene. The outperformance is achieved by the lack of internal irreversible short-circuits that allow samples 2 and 3 to continue under test conditions. Accordingly, samples 2 and 3 are preferred embodiments. Samples 2 and 3 have in common that each of them comprises cyclo-olefin copolymer which consists of 75% by weight norbornene and 25% by weight ethylene. Sample 2 is the more preferred sample as sample 2 outperforms sample 3 in the second life test.