A Machine with a “Memory” (Oct, 1937)

At first I thought this was a kind of Williams Tube, an early type of computer memory that used a grid of illuminated dots on a cathode ray tube to store data. However, according to the summary of this paper, it is basically what it looks like: a system that uses a camera to take a picture of an oscilloscope. Which made me wonder why they would call it a “memory” when it’s just a camera.

The reason I think, is that it uses the short term persistence of the oscilloscope image as a sort of buffer. When the event you’re interested in happens it will trigger the camera, giving you an image of the activity from before the event.

This is actually pretty handy and reminds me of the modern high-speed digital video cameras used on nature shows. They have to capture very unpredictable phenomenon that happen incredibly fast. By the time the photographer noticed, the event they care about has already happened. The trick to the cameras is that they are continually recording footage, keeping it for a short time in a buffer and then overwriting it. When an event happens that the photographer wants to capture, he presses the button and the camera just stops throwing away the old footage. This means the actual recording starts a few seconds before the button was pressed. This video from David Attenborough’s Life in the Undergrowth explains it perfectly. (I remembered the clip from watching the “making of” videos when I saw it a few years ago and wanted to link to it with this post. Apparently my google-fu is weak because I just spent the last hour searching before I found the right combination of keywords.)

By the way, if you haven’t seen Life in the Undergrowth, you should get a copy. It is mind-blowing and possibly my favorite nature documentary.

A Machine with a “Memory”

THIS machine has the faculty for picking up the results of electrical disturbances and registering them in its “mind.” A short time later they are illustrated on an oscilloscope, where photographs are made. The entire operation is automatic.

Of equal importance is the machine’s ability to “forget”—thus it is always ready to accept new electrical “ideas” that come to it.

It was perfected by W. E. Pakala (shown in the picture at the left) of Westinghouse.

2 comments
  1. DrewE says: February 28, 20138:21 am

    I think perhaps the memory part is some sort of a delay line or similar delay device (maybe a magnetic drum recorder from its appearance?), to enable the oscilloscope to display a signal before the trigger event. My Tek analog scope has a (short) built-in delay line for this very reason. It’s hard to decipher exactly what is going on from the layman’s description, though.

  2. Richard C says: February 28, 201310:19 am

    I did a little searching for patents with Mr. Pakala’s name, and terms like “oscilloscope”. It looks like this device is described in patent # 2186268 . As DrewE speculates, he’s using a delay, but I’ve never seen a delay line quite like this. He uses a series of capacitors on a rotor, turned by a motor at constant speed. One set of brushes charges the capacitors to “record” the voltage. After a bit of delay while the rotator rotates, another set of brushes at a different angle reads the stored charge and displays it on the oscilloscope. There’s also a facility for automatically opening and closing the shutter of the camera that makes the permanent record of the image.

    A modern digital storage oscilloscope makes this sort of delay function trivial to implement. It’s sometimes a lifesaver to “back up” and figure out what was going on just BEFORE a rare glitch happened, so I can understand why some people would be motivated to go to such lengths to create this sort of complex machine to accomplish the task.

Submit comment

You must be logged in to post a comment.