One reliability problem of tubes with oxide cathodes is the possibility that the cathode may slowly become "poisoned" by gas molecules from other elements in the tube, which reduce its ability to emit electrons. Trapped gases or slow gas leaks can also damage the cathode or cause plate-current run away due to ionization of free gas molecules. Vacuum hardness and proper selection of construction materials are the major influences on tube lifetime. Depending on the material, temperature and construction, the surface material of the cathode may also diffuse onto other elements. The resistive heaters that heat the cathodes may break in a manner similar to incandescent lamp filaments, but rarely do, since they operate at much lower temperatures than lamps.
The heater's failure mode is typically a stress-related fracture of the tungsten wire or at a weld point and generally occurs after accruing many thermal (power on-off) cycles. Tungsten wire has a very low resistance when at room temperature. A negative temperature coefficient device, such as a thermistor, may be incorporated in the equipment's heater supply or a ramp-up circuit may be employed to allow the heater or filaments to reach operating temperature more gradually than if powered-up in a step-function. Low-cost radios had tubes with heaters connected in series, with a total voltage equal to that of the line (mains). Following World War II, tubes intended to be used in series heater strings were redesigned to all have the same ("controlled") warm-up time. Earlier designs had quite-different thermal time constants. The audio output stage, for instance, had a larger cathode, and warmed up more slowly than lower-powered tubes. The result was that heaters that warmed up faster also temporarily had higher resistance, because of their positive temperature coefficient. This disproportionate resistance caused them to temporarily operate with heater voltages well above their ratings, and shortened their life.
Another important reliability problem is caused by air leakage into the tube. Usually oxygen in the air reacts chemically with the hot filament or cathode, quickly ruining it. Designers developed tube designs that sealed reliably. This was why most tubes were constructed of glass. Metal alloys (such as Cunife and Fernico) and glasses had been developed for light bulbs that expanded and contracted in similar amounts, as temperature changed. These made it easy to construct an insulating envelope of glass, while passing connection wires through the glass to the electrodes.
When a vacuum tube is overloaded or operated past its design dissipation, its anode (plate) may glow red. In consumer equipment, a glowing plate is universally a sign of an overloaded tube. However, some large transmitting tubes are designed to operate with their anodes at red, orange, or in rare cases, white heat.
Vacuum
The vacuum inside the envelope must be as perfect, or "hard", as possible. Any gas atoms remaining might be ionized at operating voltages, and will conduct electricity between the elements in an uncontrolled manner. This can lead to erratic operation or even catastrophic destruction of the tube and associated circuitry. Unabsorbed free air sometimes ionizes and becomes visible as a pink-purple glow discharge between the tube elements.
To prevent any remaining gases from remaining in a free state in the tube, modern tubes are constructed with "getters", which are usually small, circular troughs filled with metals that oxidize quickly, with barium being the most common. While the tube envelope is being evacuated, the internal parts except the getter are heated by RF induction heating to extract any remaining gases from the metal. The tube is then sealed and the getter is heated to a high temperature, again by radio frequency induction heating. This causes the material to evaporate, absorbing/reacting with any residual gases and usually leaving a silver-colored metallic deposit on the inside of the envelope of the tube. The getter continues to absorb any gas molecules that leak into the tube during its working life. If a tube develops a crack in the envelope, this deposit turns a white color when it reacts with atmospheric oxygen. Large transmitting and specialized tubes often use more exotic getter materials, such as zirconium. Early gettered tubes used phosphorus based getters and these tubes are easily identifiable, as the phosphorus leaves a characteristic orange or rainbow deposit on the glass. The use of phosphorus was short-lived and was quickly replaced by the superior barium getters. Unlike the barium getters, the phosphorus did not absorb any further gases once it had fired.
Transmitting tubes
Large transmitting tubes have carbonized tungsten filaments containing a small trace (1% to 2%) of thorium. An extremely thin (molecular) layer of thorium atoms forms on the outside of the wire's carbonized layer and, when heated, serve as an efficient source of electrons. The thorium slowly evaporates from the wire surface, while new thorium atoms diffuse to the surface to replace them. Such thoriated tungsten cathodes usually deliver lifetimes in the tens of thousands of hours. The end-of-life scenario for a thoriated-tungsten filament is when the carbonized layer has mostly been converted back into another form of tungsten carbide and emission begins to drop off rapidly; a complete loss of Thorium has never been found to be a factor in the end-of-life in a tube with this type of emitter. The highest reported tube life is held by an Eimac power tetrode used in a Los Angeles radio station's transmitter, which was removed from service after 80,000 hours (~9 years) of operation. It has been said that transmitters with vacuum tubes are better able to survive lightning strikes than transistor transmitters do. While it was commonly believed that at rf power levels above approx. 20 kilowatts, vacuum tubes were more efficient than solid state circuits, this is no longer the case especially in medium wave (AM broadcast) service where solid state transmitters at nearly all power levels have measurably higher efficiency. FM broadcast transmitters with solid state PA's up to approx. 15 kW also show better overall mains-power efficiency than tube-based PA's and with the advent of newer, more efficient transistors this power level will certainly increase.
Receiving tubes
Cathodes in small "receiving" tubes are coated with a mixture of barium oxide and strontium oxide, sometimes with addition of calcium oxide or aluminium oxide. An electric heater is inserted into the cathode sleeve, and insulated from it electrically by a coating of aluminium oxide. This complex construction causes barium and strontium atoms to diffuse to the surface of the cathode when heated to about 780 degrees Celsius, thus emitting electrons.
Failure modes
Catastrophic failures
A catastrophic failure is one which suddenly makes the vacuum tube unusable. A crack in the glass envelope will allow air into the tube and destroy it. Cracks may result from stress in the glass, bent pins or impacts; tube sockets must allow for thermal expansion, to prevent stress in the glass at the pins. Stress may accumulate if a metal shield or other object presses on the tube envelope and causes differential heating of the glass. Glass may also be damaged by high-voltage arcing.
Tube heaters may also fail without warning, especially if exposed to over voltage or as a result of manufacturing defects. Tube heaters don't normally fail by evaporation like lamp filaments, since they operate at much lower temperature. The surge of inrush current when the heater is first energized causes stress in the heater, and can be avoided by slowly warming the heaters, gradually increasing current. Some tubes intended for series string operation of the heaters across the supply will have a definite controlled warm-up time to avoid excess voltage on some heaters as others warm up. Directly-heated filament-type cathodes as used in battery-operated tubes or some rectifiers may fail if the filament sags, causing internal arcing. Excess heater-to-cathode voltage in indirectly heated cathodes can break down the insulation between elements and destroy the heater.
Arcing between tube elements can destroy the tube. An arc can be caused by applying plate potential before the cathode has come up to operating temperature, or by drawing excess current through a rectifier which damages the emission coating. Arcs can also be initiated by any loose material inside the tube, or by excess screen voltage. An arc inside the tube allows gas to evolve from the tube materials, and may deposit conductive material on internal insulating spacers.[10]
Degenerative failures
Degenerative failures cause the performance of the tube to slowly deteriorate with time.
Overheating of internal parts, such as control grids or mica spacer insulators, can result in trapped gas escaping into the tube; this can reduce performance. A getter is used to absorb gases evolved during tube operation, but has only a limited ability to combine with gas. Control of the envelope temperature prevents some types of gassing. A tube with very bad internal gas may have a visible blue glow when plate voltage is applied.
Gas and ions within the tube contribute to grid current which can disturb operation of a vacuum tube circuit. Another effect of overheating is the slow deposit of metallic vapors on internal spacers, resulting in inter-element leakage.
Tubes on standby for long periods, with heater voltage applied, may develop high cathode interface resistance and display poor emission characteristics. This effect occurred especially in pulse and digital circuits, where tubes had no plate current flowing for extended times.
Cathode depletion describes the loss of emission after thousands of hours of normal use. Sometimes emission can be restored for a time by raising heater voltage either for a short time or a permanent increase of a few percent. Cathode depletion was uncommon in signal tubes but was a frequent cause of failures of monochrome television cathode-ray tubes.[11]
Other failures
Vacuum tubes may have or develop defects in operation that makes an individual tube useless in one device, but which may not prevent its satisfactory operation in another system. Microphonics refers to internal vibration of tube elements, which modulates the signal from the tube in an undesirable way; sound or vibration pick-up may affect the signals, or even cause uncontrolled howling if a feedback path develops between a microphonic tube and, for example, a loudspeaker. Leakage current between AC heaters and the cathode may couple into the circuit, or electrons emitted directly from the ends of the heater may also inject hum into the signal. Leakage current due to internal contamination may also inject noise.[12]
Computer vacuum tubes
Colossus
Colossus (and its successor Colossus Mk2) was built by the British during World War II to substantially speed up the task of breaking the German high level Lorenz encryption. Based on 1500 vacuum tubes, Colossus replaced an earlier machine based on relay and switch logic (the [Heath Robinson]). Colossus was able to break in a matter of hours messages that had previously taken several weeks. Colossus Mk2 used a total of around 2000 vacuum tubes. Colossus was the first ever use of vacuum tubes on such a large scale for a single machine. The largest project previously had used just 150 tubes and had proven to be extremely unreliable. The main design problem at Colossus's inception was how to make vacuum tube based equipment reliable when the tubes were used in large numbers.
The Colossus computer's designer, Dr. Tommy Flowers, had a theory that most of the unreliability was caused during power down and (mainly) power up. Once Colossus was built and installed, it was switched on and left switched on running from dual redundant diesel generators (the wartime mains supply being considered too unreliable). The only time it was switched off was for conversion to the Colossus Mk2 and the addition of another 500 or so tubes. Another 9 Colossus Mk2s were built, and all 10 machines ran with a surprising degree of reliability. The 10 Colossi consumed 15 kilowatts of power each, 24 hours a day, 365 days a year—nearly all of it for the tube heaters.
Whirlwind
To meet the reliability requirements of the early digital computer Whirlwind, it was necessary to build special "computer vacuum tubes" with extended cathode life. The problem of short lifetime was traced to evaporation of silicon, used in the tungsten alloy to make the heater wire easier to draw. Elimination of the silicon from the heater wire alloy (and paying extra for more frequent replacement of the wire drawing dies) allowed production of tubes that were reliable enough for the Whirlwind project. The tubes developed for Whirlwind later found their way into the giant SAGE air-defense computer system. High-purity nickel tubing and cathode coatings free of materials that can poison emission (such as silicates and aluminium) also contribute to long cathode life. The first such "computer tube" was Sylvania's 7AK7 of 1948. By the late 1950s it was routine for special-quality small-signal tubes to last for hundreds of thousands of hours, if operated conservatively. This reliability made mid-cable amplifiers in submarine cables possible.
World War II
Near the end of World War II, to make radios more rugged, some aircraft and army radios began to integrate the tube envelopes into the radio's cast aluminium or zinc chassis. The radio became just a printed circuit with non-tube components, soldered to the chassis that contained all the tubes. During WWII in 1942, rugged metal vacuum tubes were mounted in anti-aircraft shells. These proximity fuzes made anti-aircraft shells 6 times more effective. In the fall of 1944, artillery shells with proximity fuses were used. The tiny tubes were later known as "subminiature" types. They were widely used in 1950s military and aviation electronics.
No comments:
Post a Comment