For technical data, please refer to the Burroughs MBST Specifications reference.
Beam switching tubes are thermionic decimal counting devices used in high speed applications through the 1950s and 1960s. They feature counting speeds up to 10 MHz (versus 1 MHz for the fastest dekatrons) and high electrical efficiency. Many types are direct-drive, producing output signals which can be coupled directly to a Nixie tube with minimal intermediate components.
The magnetic beam switching tube, alternately referred to as a 'magnetron beam switching tube', 'crossed field counting tube' or 'trochotron', operates on the principle of crossed electrical and magnetic fields: an electron beam is emitted from a centrally heated cathode, spirals outward under the influence of a permanent magnet, and sequentially strikes a series of spade electrodes. The potential between the cathode and a selected spade causes the beam to lock onto that spade, which passes current to its adjacent target electrode. Adjacent to each spade/target pair is a switching grid electrode which controls the positional advancement of the electron beam. Early MBSTs have a large cylindrical permanent magnet which encircles the glass envelope, while later types have 10 discrete internal magnets which also function as the target electrodes.
Another somewhat similar device is the electrostatic beam switching tube, of which only one model was produced: the E1T. The E1T uses no magnets, and has a side-viewing phosphor screen which indicates count position.
|
Alfven-Romanus (Linear) Architecture
|
Ericsson engineers Hannes Alfven and Harold Romanus patented the first linear trochotron in 1946. These early trochotrons did not function as a ring counter, but were instead of linear construction, with an axially-positioned cathode. Although Ericsson developed at least one functional prototype, It was once believed that no linear trochotrons were ever commercially produced. This changed with the recent discovery of the Melz LP-4, an unusual Soviet design in a bizarre acorn-style envelope.
|
Backmark (Two Electrode) Architecture
|
The first cylindrical trochotron was invented in 1949 by Nils Backmark of L.M. Ericsson. Backmark's simple cylindrical trochotron was derived directly from the Alfven-Romanus linear trochotron, but introduced a centrally-positioned cathode which allowed fully circular ring counter operations.
The Backmark trochotron was intended for use as a selector in telephone switching systems, and had electrical characteristics which made it unsuited for computer use, namely the absence of a switching grid and a high amount of crosstalk. Both Ericsson and Burroughs experimented with two-electrode designs through this period; some designs featured ten discrete anode targets as specified in the original Backmark patent, while other designs had a single sleeve-type anode surrounding the spade cage. Ericsson also developed the RYG10, a MBST with a sleeve-type anode and phosphor-coated targets, which would allow the count position to be viewed in the same manner as a Dekatron.
Ericsson later produced a line of more conventional beam switching tubes under the Trochotron brand name, but only after incorporating a switching grid, developed at Burroughs several years after Backmark's original invention.
|
Fan-Kuchinsky (Three Electrode) Architecture
|
A somewhat improved magnetic beam switching tube was invented in 1952 by Saul Kuchinsky of the Burroughs Corporation. These early Kuchinsky tubes were still fairly undesirable in terms of electrical characteristics, but with the collaboration of Burroughs engineer Sin-Pih Fan, the Kuchinsky tube was approached with the following improvements: (1) reduce crosstalk noise to a minimum; (2) increase output current; (3) isolate beam locking, beam switching and other independent tube functions from each other; (4) provide a tube structure in which a change of potential at the output electrodes does not substantially tend to change the output current afforded by the beam over a wide range of output potential variations; (5) provide a tube structure which is capable of handling a signal having a high percentage of modulation without affecting stability of operation; (6) provide a tube structure in which the beam switching may be rapid and in which the unmodulated output signal approaches a square wave in form; (7) provide a tube structure which requires a minimum of external circuitry for operation. The product of these improvements is the Fan-Kuchinsky magnetic beam switching tube, suitable for use in computers and other precision applications.
The key structural advancement over the Backmark trochotron is the addition of a switching grid, which provides independent beam switching control. Like the Backmark design, the Fan-Kuchinsky tube has a large cylindrical magnet surrounding the glass envelope. The Fan-Kuchinsky tube would remain the definitive state of beam switching tube structure for several years. Kuchinsky and Fan also developed several other MBST-derived devices which apparently never made it to production, including a multi-decade counter, a single-tube BCD encoder and several types of beam switching flip-flop.
|
Kuchinsky-Wolfe (Four Electrode) Architecture
|
In 1960, Saul Kuchinsky designed a total of six different concepts for an internal-magnet device, with the goal of developing a smaller, lighter tube with improved means for providing a magnetic field. Burroughs engineer Roger W. Wolfe then added signal-improving shield grid elements which reduce crosstalk by absorbing stray electrons that would otherwise splatter on an adjacent target. These two improvements encompass the Kuchinsky-Wolfe magnetic beam switching tube, commercially known as the Beam-X Switch, the ultimate improvement of Backmark's original cylindrical trochotron.
Beam-X tubes were highly suitable for use as a Nixie tube driver, at a time when the demand for computer tubes was imploding. Burroughs was fond of noting in various publications that a Beam-X tube could replace 43 transistors and some number of passive components, and described the BX-1000 as the ideal Nixie driver. Nonetheless, by 1965 the rapidly encroaching transistor had nearly entirely eliminated the beam switching tube market, and by 1970 Burroughs had completely abandoned the beam switching tube in favor of Trixie transistor drivers and BIPCO integrated driver IC sockets.
In 1946, Philips engineer Adrianus van Overbeek designed a unique, albeit highly impractical decimal counting tube, the E1T. Many such impractical counting tube designs surfaced in the 1940s and 1950s, but the E1T was the only one to make it to production. Standing alone among the many successful dekatrons and magnetic beam switching tubes, the E1T is an electrostatic beam switching tube, equipped with a directional electron gun, focusing, acceleration and deflection grids, and a side-facing phosphor screen which indicates count position in a highly unintuitive fashion. Although the E1T has theoretical frequency limit of 1MHz, practical counting speeds are much slower. Nearly all examples of E1T, regardless of brand name, were produced by the Philips sample department; the E1T was a difficult tube to manufacture and never went to large-scale production.
