Over at davebr’s “My Tektronix Memories” page, he has a really great video of the Tektronix 4010 terminal demonstrating various features of the DISSPLA graphics software. This is a video taken of the screen of the terminal as it renders the demonstration file at 2400 baud. This gives you a very good representation of what using these terminals was like. You watched a bright spot move across the screen as it stroked the graphics onto the storage tube. If you look carefully, you can see the alphanumeric cursor blinking in the lower left side of the screen as the file transmits a bunch of NUL characters to the terminal in order to implement a delay in the presentation. The bright full-screen flash is the mechanism used to erase the screen before drawing the next picture.
So how long does the phosphor hold the ‘glow’ on these tubes? Would the image eventually fade away?
Holding the image on the screen for long periods of time can cause screen burn. There is a “hold and view” circuit in the terminal that will dim the display after a period of inactivity of about 90 seconds, but the screen still retains the image. Pressing any key (usually the shift key is the best choice) or receiving any data from the communications port will switch the circuit from “hold mode” to “view mode” and show the retained image. There are warnings in the manual about not leaving an image on the screen for too long or risk damaging the display by “burning” the image into the tube. From this I infer that the image will be retained for as long as power is supplied to the tube. The image is displayed not through phosphor retention as in a traditional raster dynamic refresh display, but from retaining the written electric charge pattern on the inside of the tube. I suppose the charge might eventually migrate away or dissipate, but this would be an effect that would take many hours or possibly even days.
The image retention time is in theory infinite. I used to work in the Storage CRT Engineering department at Tektronix and many tubes would retain an image for months. It was a true Bi-Stable system, where the flood guns needed to view the stored image also maintained it in the light or dark state. If any area gained or lost charge, it would be corrected automatically. If this sounds too much like magic, look up ‘secondary emission’ to see how firing electrons can be used to create positively and negatively charged areas.
Manufacturing variability, aging of the tube, etc. meant that in the long term there would be some fading up of dark areas or losing some of the stored image. There was a spec on retained storage time, but it was very conservative.