Editing Delay-line memory (section)

== Acoustic delay lines ==

===Mercury delay lines===
[[File:Mercury memory.jpg|thumb|Mercury memory of [[UNIVAC I]] (1951)]]
After the war, Eckert turned his attention to computer development, which was a topic of some interest at the time. One problem with practical development was the lack of a suitable memory device, and Eckert's work on the radar delays gave him a major advantage over other researchers in this regard.

For a computer application the timing was still critical, but for a different reason. Conventional computers have a [[clock period]] needed to complete an operation, which typically start and end with reading or writing memory. Thus the delay lines had to be timed such that the pulses would arrive at the receiver just as the computer was ready to read it. Many pulses would be in-flight through the delay, and the computer would count the pulses by comparing to a master [[Clock signal|clock]] to find the particular bit it was looking for.

[[File:SEACComputer 010.png|thumb|Diagram of mercury delay line as used in [[SEAC (computer)|SEAC computer]]]]
[[Mercury (element)|Mercury]] was used because its [[acoustic impedance]] is close to that of the piezoelectric [[quartz crystals]]; this minimized the energy loss and the echoes when the signal was transmitted from crystal to medium and back again. The high [[speed of sound]] in mercury (1450&nbsp;m/s) meant that the time needed to wait for a pulse to arrive at the receiving end was less than it would have been with a slower medium, such as air (343.2&nbsp;m/s), but it also meant that the total number of pulses that could be stored in any reasonably sized column of mercury was limited. Other technical drawbacks of mercury included its weight, its cost, and its toxicity. Moreover, to get the acoustic impedances to match as closely as possible, the mercury had to be kept at a constant temperature. The system heated the mercury to a uniform above-room temperature setting of 40&nbsp;°C (104&nbsp;°F), which made servicing the tubes hot and uncomfortable work. ([[Alan Turing]] proposed the use of [[gin]] as an ultrasonic delay medium, claiming that it had the necessary acoustic properties.<ref>{{cite journal |last1=Wilkes |first1=Maurice V. |title=Computers Then and Now |journal=Journal of the ACM |date=January 1968 |volume=15 |issue=1 |pages=1–7 |ref=wilkes |doi=10.1145/321439.321440 |s2cid=9846847|doi-access=free }}</ref>)

A considerable amount of engineering was needed to maintain a clean signal inside the tube. Large transducers were used to generate a very tight beam of sound that would not touch the walls of the tube, and care had to be taken to eliminate reflections from the far end of the tubes. The tightness of the beam then required considerable tuning to make sure that both transducers were pointed directly at each other. Since the speed of sound changes with temperature, the tubes were heated in large ovens to keep them at a precise temperature. Other systems{{specify|date=September 2019}} instead adjusted the computer clock rate according to the ambient temperature to achieve the same effect.

[[EDSAC]], the second full-scale [[stored-program]] [[digital computer]], began operation with 256 35-[[bit]] [[Word (data type)|words]] of memory, stored in 16 delay lines holding 560&nbsp;bits each (words in the delay line were composed from 36 pulses, one pulse was used as a space between consecutive numbers).<ref>{{cite magazine |last1=Wilkes |first1=M. V. |last2=Renwick |first2=W. |title=An Ultrasonic Memory Unit for the EDSAC |magazine=Electronic Engineering |date=July 1948 |pages=209–210 |url=http://www.cs.man.ac.uk/CCS/Archive/misc/EDSAC/EDSAC%20Memory%20Image.pdf}}</ref> The memory was later expanded to 512&nbsp;words by adding a second set of 16 delay lines. In the [[UNIVAC&nbsp;I]] the capacity of an individual delay line was smaller, each column stored 120&nbsp;bits, requiring seven large memory units with 18 columns each to make up a 1000-word store. Combined with their support circuitry and [[amplifier]]s, the memory subsystem formed its own walk-in [[Room (architecture)|room]]. The average access time was about 222&nbsp;[[microsecond]]s, which was considerably faster than the mechanical systems used on earlier computers.

[[CSIRAC]], completed in November 1949, also used delay-line memory.

Some mercury delay-line memory devices produced audible sounds, which were described as akin to a human voice mumbling. This property gave rise to the slang term [http://www.rfcafe.com/references/popular-electronics/electronic-mind-remembers-popular-electronics-august-1956.htm "mumble-tub"] for these devices.

===Magnetostrictive delay lines===
[[File:Torsion wire delay line.jpg|thumb|Torsion wire delay line]]

A later version of the delay line used [[wire|steel wires]] as the storage medium. Transducers were built by applying the [[magnetostriction|magnetostrictive effect]]; small pieces of a magnetostrictive material, typically [[nickel]], were attached to either side of the end of the wire, inside an [[electromagnet]]. When bits from the computer entered the magnets, the nickel would contract or expand (based on the polarity) and twist the end of the wire. The resulting [[torsional]] wave would then move down the wire just as the sound wave did down the mercury column.

Unlike the compressive wave used in earlier devices, [[Torsion (mechanics)|torsional]] waves are considerably more resistant to problems caused by mechanical imperfections, so much that the wires could be wound into a loose coil and pinned to a board. Due to their ability to be coiled, the wire-based systems could be as long as needed, so tended to hold considerably more data per unit; [[kilobit|1&nbsp;kbit]] units were typical on a board only 1&nbsp;[[square foot]] ({{nobr|~30 cm × 30 cm}}). Of course, this also meant that the time needed to find a particular bit was somewhat longer as it travelled through the wire, and access times on the order of 500&nbsp;microseconds were typical.

[[File:Highgate Wood 100 micro second delay line store.JPG|thumb|100-microsecond delay-line store]]

Delay-line memory was far less expensive and far more reliable per bit than [[Flip-flop (electronics)|flip-flops]] made from [[vacuum tube|tubes]], and yet far faster than a [[latching relay]]. It was used into the late 1960s, notably on commercial machines like the [[LEO&nbsp;I]], [[Highgate Wood Telephone Exchange]], various [[Ferranti]] machines, and the [[IBM 2260|IBM 2848 Display Control]]. Delay-line memory was also used for video memory in early terminals, where one delay line would typically store 4 lines of characters (4&nbsp;lines × 40&nbsp;characters per line × 6&nbsp;bits per character = 960&nbsp;bits in one delay line). They were also used very successfully in several models of early desktop [[electronic calculator]], including the [[Friden, Inc.|Friden]] EC-130 (1964) and EC-132, the [[Olivetti]] [[Programma&nbsp;101]] desktop [[programmable calculator]] introduced in 1965, and the Litton [[Monroe Epic]] 2000 and 3000 [[programmable calculators]] of 1967.

===Piezoelectric delay lines===
[[File:ultrasonicdelayline.jpg|thumb|An ultrasonic [[analog delay line]] from a [[PAL]] color TV; it delays the color signal by 64&nbsp;μs. Manufacturer: VEB ELFEMA Mittweida ([[East Germany|GDR]]) in 1980]]
A similar solution to the magnetostrictive system was to use delay lines made entirely of a [[piezoelectric]] material, typically quartz. Current fed into one end of the crystal would generate a compressive wave that would flow to the other end, where it could be read. In effect, piezoelectric material simply replaced the mercury and transducers of a conventional mercury delay line with a single unit combining both. However, these solutions were fairly rare; growing crystals of the required quality in large sizes was not easy, which limited them to small sizes and thus small amounts of data storage.<ref>{{cite book |title=Glass Memories |publisher=Corning Electronics |date=1963 |id=RRP 8/63 5M |url=https://archive.org/details/TNM_Glass_computer_memories_-_Corning_Electronics_20171206_0185}}</ref>

A better and more widespread use of piezoelectric delay lines was in European television sets. The European [[PAL]] standard for color broadcasts compares the signal from two successive lines of the image in order to avoid color shifting due to small phase shifts. By comparing two lines, one of which is inverted, the shifting is averaged, and the resulting signal more closely matches the original signal, even in the presence of interference. In order to compare the two lines, a piezoelectric delay unit to delay the signal by a time that is equal to the duration of each line, 64&nbsp;μs, is inserted in one of the two signal paths that are compared.<ref>{{cite thesis |first=F.T. |last=Backers |title=Ultrasonic delay lines for the PAL colour-television system |date=1968 |type=Ph.D. |location=Eindhoven, Netherlands |publisher=Technische Universiteit |url=https://pure.tue.nl/ws/files/3648245/161551.pdf |pages=7–8 }}<br/>{{cite journal |first=F. Th. |last=Backers |title=A delay line for PAL colour television receivers |journal=Philips Technical Review |volume=29 |issue= |pages=243–251 |date=1968 |doi= |url=http://www.extra.research.philips.com/hera/people/aarts/_Philips%20Bound%20Archive/PTechReview/PTechReview-29-1968-243.pdf }}{{Dead link|date=May 2024 |bot=InternetArchiveBot |fix-attempted=yes }}</ref> In order to produce the required delay in a crystal of convenient size, the delay unit is shaped to reflect the signal multiple times through the crystal, thereby greatly reducing the required size of the crystal and thus producing a small, rectangular-shaped device.