Some strange things have been used as computer memory over the years. Previously I have written about the vacuum tube memory of ENIAC and mercury delay line memory used in some early computers. Here I want to tell you about a really odd memory technology that was used early on.
For many years, until very recently, the Cathode Ray Tube (CRT) monitor was the standard computer display, as well as television sets. If you are over 21, you most likely remember these large, heavy monitors. You may also remember that if you turned off a CRT monitor or TV in a dark room, the screen would continue to glow for some time. And you may even remember that the screen usually had a considerable static charge on it after being used. It turns out that CRTs had some interesting properties. What is interesting to us is that CRTs were used as early computer memory by taking advantage of one of those strange properties. As a side note, the name Cathode Ray Tube dates back to the late 1800s when the principals of operation were first discovered. It is rather incorrect, but has stuck around for over a hundred years despite that. It is an interesting story how the name came about, but that's another story for another time. Let's see how CRTs were used as memory.
The CRT is an interesting device. It is a vacuum tube that works on the same principal as any other vacuum tube. A large glass "tube" that is somewhat bell shaped is evacuated to cause a vacuum. At the back (narrow) end is placed a cathode, or negative terminal. The large, somewhat flat display area at the front has another terminal (anode) that is positive. The cathode is heated to a point that causes the electrons to "fly off" the metal terminal and create a cloud of electrons around it. Because of the difference in voltage applied to the cathode and anode, the cloud of electrons is attracted to anode and fly toward it in a stream. They prefer the straightest, easiest path and without any coercing will hit the anode right in the middle. Perhaps you have seen an older CRT monitor or TV when the power first comes on or goes off with a bright spot right in the middle. The Cathode is created in a shape that minimizes the size of the electron stream and is therefore called an "electron gun."
We now have a stream of electrons from the cathode to the anode. But how does it display a picture? The anode (display) area is also covered with a phosphorescent chemical compound that glows when struck by the electrons. The intensity of the electron stream determines how bright it glows: stronger means brighter. To get more than just a small spot in the center, the electron stream is deflected from side to side and up and down. In most monitors and TVs the electrons are deflected by a magnetic field. Around the sides of the tube, near the skinny cathode end, are some large coils of wire. The coils have current passing through them that creates a magnetic field which attracts and repels the electrons, causing their path to curve. By changing the strength of the field the beam of electrons can be made to hit any part of the screen. In some types of CRT, normally used with oscilloscopes and "vector monitors," the coils of wire are replaced with large metal plates. A voltage on those plates attracts and repels the beam similarly to the magnetic field. The electrostatic plates are able to deflect the beam faster than changing the magnetic fields, so are used where high speed is needed. But in either case, the beam is able to reach any point on the screen and create light and dark spots. With a TV and most monitors, the beam is "scanned" from left to right to create a line, then from top to bottom to create a series of lines, creating an entire image on the phosphor.
That is neat and all, but how is it used as memory? Before I get to that let me tell you that some of this information was found at radiomuseum.org and you can get more details and some neat pictures there. Check them out for some really neat information on early radio and electronics technology. So, to use something as a "memory" it has to be able to store information. It turns out that when the electron beam strikes the anode, it leaves a small amount of electrostatic charge. Remember I mentioned the static charge on a CRT? You may have experienced that before. The charge is localized to where the beam hits, and will bleed away in a short time. But it can be detected if the beam hits the same spot again before it bleeds away. The voltage measured at the outside surface of that spot on the screen will be just a bit higher when the beam hits it again because of the added static charges. A one or a zero can be stored for a short time by either lighting or not lighting a small area (or lighting a smaller or larger area) and then reading the voltage on a second scan. By dividing the screen up into storage areas and placing a grid of voltage measuring contacts over the display, the screen can store data for a short time.
Most computers today use dynamic RAM (DRAM) that stores charge in a capacitor. The charge will also bleed away in a short time and must be refreshed periodically by reading the data and rewriting it to keep the information. Much like modern DRAM, the CRT had to be refreshed. The same was true with the mercury delay lines we looked at before. So the electronics were made to feed back the read out signal back to the input and refresh the display. An interesting point about CRT memory was that a second CRT could be added without the sensor grid over the face. That one would have the screen visible and you could see the contents of the memory! That is a great debugging tool. A typical CRT could store a thousand or so bits. A few dozen would be used to store a few kilobytes of data, which was a considerable improvement over the accumulators of ENIAC or the mercury delay lines.
I hope you are enjoying this series of posts. Please leave some feedback so I know if this is interesting or not. And if you think others may enjoy what I write, please spread the word.
Thanks.
No comments:
Post a Comment