4 ms·

Here they had a series on Edsac, the first von Neumann machine. There they had a long mercury tube as main memory. The signal was turned into sound by a loudspe

Here they had a series on Edsac, the first von Neumann machine. There they had a long mercury tube as main memory. The signal was turned into sound by a loudspeaker and sent to pass through the mercury tube (sound is very slow when passing through mercury), then on the other end converted to digital and looped back to the loudspeaker for the next trip (in loop). They say that this hack was invented during WW2 to store/recall RADAR signals. https://m.youtube.com/watch?v=lXJ-tYqPARg https://m.youtube.com/watch?v=lXJ-tYqPARg The proper word for that is delay line memory https://en.wikipedia.org/wiki/Delay_line_memory https://en.wikipedia.org/wiki/Delay_line_memory

jonsen 8y ago
(sound is very slow when passing through mercury) From your link: “The high speed of sound in mercury (1450 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 m/s)...” The reason for using mercury: “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.”
- MichaelMoser123 8y ago
  yeah, the speed of sound in mercury is not that slow. Never trust anything you hear on youtube. https://www.engineeringtoolbox.com/sound-speed-liquids-d_715.html https://www.engineeringtoolbox.com/sound-speed-liquids-d_715...
  - majewsky 8y ago
    It makes intuitive sense, too. For sound to travel through a medium, the constituent particles have to interact with their neighbors to keep the sound wave moving. In a liquid, the particles are much closer to their neighbors (on average) than in a gas at normal pressures. Because of this stronger coupling, the average liquid (or solid) should have a much higher speed of sound than the average gas.
Animats 8y ago
The trouble with delay line memory was not the speed of sound. It was that you had to wait for the entire memory (or, on average, half of it) to go around to get to any data item. More memory meant slower memory. Electronic arithmetic was much faster than the memory system, so delay line computers were very memory-bound. But not too expensive, once steel rods replaced mercury tanks.

jonsen 8y ago

(sound is very slow when passing through mercury) From your link: “The high speed of sound in mercury (1450 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 m/s)...” The reason for using mercury: “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.”

MichaelMoser123 8y ago
yeah, the speed of sound in mercury is not that slow. Never trust anything you hear on youtube. https://www.engineeringtoolbox.com/sound-speed-liquids-d_715.html https://www.engineeringtoolbox.com/sound-speed-liquids-d_715...
- majewsky 8y ago
  It makes intuitive sense, too. For sound to travel through a medium, the constituent particles have to interact with their neighbors to keep the sound wave moving. In a liquid, the particles are much closer to their neighbors (on average) than in a gas at normal pressures. Because of this stronger coupling, the average liquid (or solid) should have a much higher speed of sound than the average gas.

Animats 8y ago

The trouble with delay line memory was not the speed of sound. It was that you had to wait for the entire memory (or, on average, half of it) to go around to get to any data item. More memory meant slower memory. Electronic arithmetic was much faster than the memory system, so delay line computers were very memory-bound. But not too expensive, once steel rods replaced mercury tanks.