Williams tube

from Wikipedia, the free encyclopedia
Williams-Kilburn tube with a hinged metal plate in front of the fluorescent screen for reading

The Williams tube , also known as Williams-Kilburn tube by Frederic Calland Williams and his colleague Tom Kilburn called, is a cathode ray tube , as memory random access in the first tube computers was used in the 1950s. It is considered to be the first computer memory that allowed any write and read access.

The Williams tube was developed in 1946 and 1947 at the University of Manchester in England and was used in the following years in the most powerful computer systems at the time, such as the SWAC . Due to the low reliability and the rapid aging in operation, the Williams tube was only used to a limited extent and was replaced by the so-called core memory as working memory from the mid-1950s . The Selectron was developed independently and in the same time frame in the USA , with an identical area of ​​application.

function

Williams tube from the tube computer SWAC for storing 256 bits

When a point is projected onto the anode of a cathode ray tube by an electron beam, this point remains visible for some time through the phosphorescence , depending on the phosphor used in the tube . If the energy of the electron beam is above a certain limit value, which among other things depends on the phosphor used, electrons are emitted from the phosphor layer as a result of the secondary emission. These fly a short distance away from the anode inside the tube, but are attracted by the positive charge of the anode and then land in the vicinity of the point in the fluorescent layer.

This leads to a charge gradient in the vicinity of the point: the region immediately at the point is depleted of electrons and thus slightly positively charged, a ring a little further outside the point where the electrons of the secondary emission hit is slightly negatively charged . Due to the poor electrical conductivity of the phosphor layer, this state remains for a certain time until it is balanced and can be used for information storage in this time frame. The writing of the information has to be repeated cyclically if the information is to be stored over a longer period of time, similar to the refresh cycle in DRAMs .

With the charge distribution, the information of one bit can be written per point . The storage density is limited by the size of the anode screen and the necessary spacing between the individual regions in order to avoid crosstalk . Typically around 200 bits to 2000 bits can be stored on the screen of a Williams tube. The optional addressing of the individual pixels (bits) is implemented by controlling the beam deflection in the cathode ray tube, which focuses the electron beam on the addressed point on the luminous layer.

Reading process

Luminous layer of a Williams tube that stores 256 bits in the form of 256 points or lines

The reading process, which only reads the status of one point at a time, is made possible by another effect when writing: By writing a bit and the resulting distribution of charge around the point, a short current pulse occurs in immediately adjacent electrical conductors when the image area was not previously described, i.e. there is no charge gradient between the image point and the outer ring area. If, on the other hand, the writing is carried out in a point already described above, there is no significant current pulse in adjacent conductors. To implement the read function, a thin metal plate is attached directly in front of the fluorescent screen and connected to the read amplifier . As a result, a so-called consuming reading can be used to determine whether the state of the point was set or reset before the reading process.

The technical difficulty is that consuming reading leaves all points on the luminous layer in the set state, including those bits which were not previously set. It is therefore necessary to reset memory positions that have not been set to the discharged state after reading, i.e. to delete them. This can be achieved by setting an additional write pulse just next to the actual storage point after consuming reading. With an appropriate selection of the distance and strength, the resulting secondary emission and the forming ring of charges due to the geometric displacement lead to charge extinction in the actual storage point. In practice, this erasure operation is carried out, since it is technically easier to implement from the control, by writing a straight line over the point, whereby a pattern of points and lines is optically formed on the fluorescent layer of the cathode ray tube in the memory matrix, as shown in the figure below. A point represents the charged state of a memory cell, a line represents the discharged or reset state.

Web links

Commons : Williams tubes  - collection of images, videos and audio files

Individual evidence

  1. ^ Early computers at Manchester University. (No longer available online.) Archived from the original on August 28, 2017 ; accessed on August 17, 2016 . Info: The archive link was inserted automatically and has not yet been checked. Please check the original and archive link according to the instructions and then remove this notice. @1@ 2Template: Webachiv / IABot / www.cs.man.ac.uk
  2. Patent US2951176 : Apparatus for storing trains of pulses. Registered December 10, 1947 , published August 30, 1960 , inventor: Frederic Calland Williams.