I worked for Tek during the mid 70's on the 4010 series. Field service in the Los Angeles area. Yes, the flood guns that cause the tube to store the image AND erase it would go weak after a time. The length of time seemed to vary for many reasons. I found that customers that left the terminal on all the time (Tek used them as terminals for their semiconductor test systems S-3200 series) they seemed to last a long time, BUT if you turned it off for any length of time, it would take a really long time to warm up and erase properly. The bigger terminals (4014 and 4015) had some adjustments for collimation (the shape of the flood beam during the erase pulse), but not the 4010 series. If it finally starts storing and erasing after a period of time, I would suggest that you live with it. The tubes are no longer available (probably obsolete for over 10 years). If you can find them, the tube for a 4012/4013 or the Tek611 (storage monitor, also used in the 4002A (somebodies favorite) would work in your 4010. If you have the hard copy unit (the dry silver paper printer), you may have trouble getting clear copies. There were times when you replace the tube with a new one and you couldn't get a good copy. You ended up replacing it a second time. It didn't take too much time to replace one after you had done it a few times. The computer interface. In the standard terminal, there was one slot for a data comm interface. There were two RS232 interfaces available. The standard full duplex and a deluxe i/f. The deluxe version had switch selectable baud rates and could do half duplex comm. Remember this was back in the days of 300-1200 baud modems, both acoustic couplers and AJ and Racal-Vadic modems. That was back in the days when you paid a 'buck a baud' for a modem, especially above 2400 baud. Tek also had a series of TTY Port interfaces. These were meant to be directly connected to a mini. They had them for DEC PDP-11's ,DG Novas, and others. The DEC used either a M7800 or M7856. The M7800 was easy to install, it had rotary switches and you simply selected the external clock position and made sure you had the right crystal to get 38k baud. The M7856 had to be modified to replace the 110 baud position (IIRC) with a clock from the terminal interface. The neat thing about these interfaces was that when you got to the bottom of the screen, they would stop the clock to the computer interface and the screen would 'freeze'. For those unfamiliar with the 4010 series of terminals, after you clear the screen and start to display information, when you got to the bottom of the screen, the cursor would position itself to the top middle of the screen and over-write down the middle. Then it would go back to the left margin and continue. So, you had to stop the computer from sending data to the terminal to keep from scrambling your display. The DG Nova interface/computer would really scream, usually in the 300k baud range. It did a great job with the big (4014/4015) displays. Cal Tech had several of them in their seismographic lab. They threw up a screen full of graphics almost instantly. Very impressive. The PDP interface was good for about 38k to 56k, depending on the processor. There were some modified products availbe to display a non-storing line of characters. IIRC, it wasn't available for the 4010, but was for the 4012 series. The 4012 had similar characteristic to the 4010 except it had a better character generator, lower case display and just generally a nicer terminal. I have mentioned the 4013 and 4015. These were versions of the 4012/4014 that had special character generators to be a terminal for the APL programming language. This language used graphical characters for its operators. An interesting language. I took a class in it at night. The IBM APL terminal was a 2741 (IIRC). For some oddball reason, it used 134.5 baud as a comm rate.