Read r: Untitled Document text version

Dispenser cathodes for CRTs


DURING THE 40 YEARS of their existence, tungsten dispenser cathodes have been confined, for the most part, to high-power vacuum devices such as traveling-wave tubes, klystrons, and magnetrons. Although they perform better than oxide cathodes, CRT manufacturers showed little interest in using them because of their higher cost. After all, the oxide cathode worked well enough - and it was cheap. But that attitude is changing. High-resolution high-brightness CRT applications are proliferating, and they require electron sources with ever higher current densities. It has become obvious over the last few years that the oxide cathode cannot continue to meet these demands much longer. To better appreciate the historic change we are witnessing in CRT technology, let's take a closer look at the oxide cathode. Hot wires for hot electrons The oxide cathode was the first advance beyond the pure tungsten filament, whose operation had been explained by Richardson at the turn of the century. The oxide cathode was discovered accidentally in 1903 when Wehnelt observed that areas of a hot platinum wire emitted more efficiently after becoming contaminated with stop-cock grease. He assumed that metal-oxide impurities originating from the grease were contributing to the increased emission. To pursue this theory he conducted a detailed examination of various metallic oxides, and found that oxides of the alkaline earth metals, in particular, produced considerable electron emission. The oxide cathode had been born. Oxide cathodes are commonly produced by mixing together carbonates of barium, calcium, and strontium. The mixture is combined with a binder, usually nitrocellulose, and sprayed onto a nickel substrate. During a subsequent tube exhaust, the 2550-~m-thick coating decomposes to oxides, and the binder and carbon dioxide is pumped away. When heated, these oxides emit electrons at relatively low temperatures (below 700 degC). The oxide cathode is capable of delivering high current density, but only in very short pulses because of resistance heating and dielectric charging. Even at current densities that are only moderately high- less than 2 A/cm2 - the increased temperature depletes the oxide coating more rapidly, which reduces the cathode's lifetime. Modifications-the addition of scandium oxide by Mitsubishi and the addition of indium oxide by Sony - have extended the oxide cathode's current and life capabilities, but not beyond 2 A/cm2 and 20,000 hours. (This is a meaningful improvement. The anticipated lifetime of a standard oxide cathode at 2 A/cm2 is 10,000 hours or less.) Dispensing makes a difference The emission mechanism of dispenser cathodes differs significantly from that of oxide cathodes. In barium oxide (BaO) cathodes the electrons are emitted from donor sites created by oxygen vacancies in a semiconductor (BaO) matrix. In dispenser cathodes, the emission is from an activated metal surface assisted by a dipole consisting of barium over oxygen over the metal. The dipole significantly lowers the work function from 4.5 eV in the case of tungsten to approximately 2.0 eV. The work function is the amount of energy required to extract one electron from the surface of an object. In the case of a hot metal surface (such as a tungsten filament), the energy used to force the electron from the outer orbit of the atom is brute-force heat. In the case of a bariumactivated metal surface, a positively charged barium ion sits on top of a negative oxygen ion, which creates an electrical dipole that acts somewhat like an extracting grid and helps remove the electron from the surface. This reduces the heat energy needed to remove the electron. In the case of tungsten, the reduced energy requirement translates to an operating temperature that is approximately 10000c less. The basic difference in emission mechanism permits a dispenser cathode to operate with no restriction on pulse length. Because the cathode is a metal surface, there are no problems with resistive heating or dielectric charging. It has the same performance for dc as for short pulses. (But because a dispenser cathode's work function of 1.8-2.0 eV is somewhat higher than an oxide cathode's 1.3-1.5 eV, it requires an operating temperature of about 100~C higher.) History, short and sweet The original dispenser cathodes were developed in the late 1940s by Philips. They consisted of a porous plug of tungsten over a reservoir of barium carbonate. When the BaCO3 decomposed, barium and oxygen migrated to the surface through the pores in the plug and create the dipole mentioned previously. These "L-type" cathodes were the first of a class of dispenser cathodes called reservoir cathodes. Siemens improved on the concept by using pre-processed BaO in the reservior, thus creating the MK cathodes.

In the early 1950s, Philips developed impregnated-dispenser ("B-type") cathodes. The B-type cathode consisted of a porous plug whose pores were filled with barium calcium aluminate, a ceramic material [Fig. 1]. As a result, each pore became it's own reservoir. A number of improvements have been made to this concept over the years. The most significant occurred in the 1960s when a surface coating of osmium ruthenium alloy was added to the tungsten, greatly reducing the work function. The resulting cathodes are generally referred to as "M-type," and the reduction in their work function from 2.1 to 1.8 eV produced a 100'C reduction in operating temperature for the same emission (or an increase in current density of 2.5 times at the same temperature). In 1978, the U.S. military became concerned about the reliability and life expectancy of (primarily) space-based defense systems utilizing microwave tubes. The Army, Navy, and Air Force organized the Tri-Service Cathode Workshop specifically to address cathode development efforts. This workshop has been held every other year since 1978 and has chronicled the development of cathodes. The chronicles reveal that modern surface-analysis techniques, such as Auger and XPS, have helped us to a better understanding of the operation and theory of dispenser cathodes. The significant developments reported during this period were: Mixed-metal-matrix cathodes, where an enhancing metal such as osmium or iridium is added to the porous metal (Varian). Controlled-porosity dispenser (CPD) cathodes (NRL, Varian, Hughes, Semicon Associates). Top-layer scandate cathodes, conventional impregnated cathodes with a thick surface layer of a special barium, tungsten, and scandium oxide compound (Philips). High-current long-life reservoir cathodes, an improved version of the original MK cathode (Varian). Sputter-coated scandate cathodes utilizmg a thin surface layer of the same compound used by Philips in their top-layer scandate cathodes (Hitachi). Iridium-coated cathodes, a version of the "M" cathode utilizing an optimized tungsten-iridium surface (Toshiba). Optimum-alloy-coated cathodes, co-deposited optimized alloy of osmium and tungsten (Varian). Scandium-oxide-doped oxide cathodes (Mitsubishi). Many of these developments are important for CRTs. Dispensing in CRTs Cost has been the primary factor limiting the use of dispenser cathodes in CRTs. Nonetheless, some CRTs have utilized dispenser cathodes because - despite their high cost - they were the only way of achieving the performance required in certain applications. Among these have been avionics displays, camera tubes, and oscilloscopes. Thomson-CSF, Tektronix, RCA, Hitachi, Philips, Toshiba, and many other companies have experience in one or more of these areas. The promise of more commercial applications is driving current activity. As a result, several CRT companies have developed in-house capabilities for producing dispenser cathodes. Philips, Matsushita, Toshiba, Sony, Thomson, and Hitachi all have some capability. Philips has developed a pressed and sintered scandate dispenser cathode. Matsushita, Sony, and Thomson can produce conventional coated impregnated cathodes - osmium-ruthenium coated at Matsushita, Thomson, and Sony; iridium coated at Toshiba. Hitachi produces conventional impregnated cathodes coated with a tungsten scandium-oxide compound. Except for Philips' pressed and sintered scandate - a modification of top-layer scandate in which the aluminate is pressed along with the other ingredients and sintered instead of being impregnated - all of these cathodes are of the conventional impregnated dispenser type. The process for making such cathodes begins with tungsten particles, typically between 4 and 6 mm in diameter. The particles are compressed into small pellets that are sintered to an 80% density

at temperatures in excess of 2000'C. These pellets are impregnated with a material generally made up of barium, calcium, and aluminum oxides in a molar ratio of 5:3:2 or 4:1:1. This process requires a temperature - usually in excess of 1500~C-sufficient to melt the aluminate so that it can infiltrate the porous pellet. The excess molten material is removed from the surface, and the pellets are coated with either osmium/ruthenium, iridium, or tungsten/scandium oxide to lower the system's effective work function. These cathodes have the advantages of long life and high current. Disadvantages include longer activation times perhaps two to four times that required for oxide cathodes and the CPD cathode described later - and an initially high barium evaporation rate that can produce stray emission.

A controlled-porosity dispenser (CPD) cathode specifically designed for use in CRTs has been recently developed by Semicon Associates, a major independent dispenser cathode manufacturer [Fig. 2]. This classic reservoir dispenser cathode was described at the recent 1991 SID Symposium in Anaheim. Instead of a porous tungsten structure whose pores are filled with aluminates, this cathode has a thin 50-µm membrane with a precise array of laser-drilled holes directly behind the 0-1 aperture. Behind the foil, in a separate reservoir, is the barium source in the form of a ceramic pellet. When heated, the barium compounds in the pellet partially decompose, producing a partial pressure behind the foil. The barium migrates through the holes-or pores-to the surface and forms the required monolayer. This reduces the work function to a low 1.73 eV, which permits electrons to be extracted at a relatively low temperature. The barium is always a fixed distance from the surface, unlike an impregnated cathode in which the barium supply recedes from the surface with use.

With its uniformly spaced pores - 5-µm diameter on 20-mm centers-and constant migration distance from the reservoir, this CPD cathode has extremely uniform and constant emission. The same features also give it an activation time that is quite rapid. Most users have reported being able to adapt the cathode directly into the standard tube-processing schedules they use for oxide cathodes. Either tungsten/rhenium or molybdenum/rhenium can be used as the emitter surface. Because some of the more expensive manufacturing steps used for impregnated cathodes are not present, CPD cathodes offer an opportunity for lower cost. Potential disadvantages of this type are (1) more barium evaporation than in an oxide cathode, and (2) possible stripping of the dipole from the activated metal surface by ion bombardment, which can produce an emission slump under severe loads. It is obvious that better tube processing will be required for all dispenser cathodes. Table I compares the features of the various cathode types. A CPD in your CRT? With the cost of a dispenser cathode projected to go as low as from $2 to $3, long-life high-performance CRTs can be expected to use dispenser cathodes in applications such as projection tubes, high-resolution monitors, and HDTV.


r: Untitled Document

4 pages

Report File (DMCA)

Our content is added by our users. We aim to remove reported files within 1 working day. Please use this link to notify us:

Report this file as copyright or inappropriate


You might also be interested in

r: Untitled Document
Microsoft Word - Chapter 6 cathodes_022208.doc