I have been making progress on restoring the VT100 I acquired some time ago. The RAM is now OK and the keyboard interface is working. However, I was still not getting any output on the screen and I know that the NVRAM is not working. I have obtained a replacement NVRAM chip but I want to fix the video first.

When I started to look more closely, the first thing I discovered was that the fuse on the monitor board for the +12V supply was open circuit. Using a bench power supply with current limiting I connected just the yoke and gave the board 12V. There was a faint glow from the yoke and it drew about 100ma. On advice I temporarily used a 12V car bulb in place of the fuse in order to test how much current gets drawn when the monitor board is fully connected to everything. It drew about 1.4A and the voltage on the output side of the bulb was 6.9V, so there was no dead short.

I also tested the flyback transformer with a ringing tester. Most of the windings seemed OK when compared to a VT101 that I have. One of the windings still rings, but not as much as the VT101. These are the scope traces for the two flybacks:

I tried to test without the car bulb acting as a current limiter. However, the yoke filament did not glow and I fancied I could smell some burning, possibly coming from the flyback transformer, so I switched it off very quickly. I fear that one of the flyback windings may have failed, but I don’t know.

On further advice I decided to connect an external monitor to the external video output on the basic video board. I found an old LCD screen that works just about well enough with a little video converter device I have. I verified it worked first with my working VT102 and I got a display. However, when I tried it with the VT100 I did not get the expected image. On startup I got some random looking patterns, and when it had finished starting up I got a blinking column down the left of the screen. I also noticed the smell of a chip that is getting hot, but I couldn’t locate it.

I checked that the DMA was retrieving the right data from the screen RAM. This seemed to be OK and the right data was being presented to the character generator ROM. However, my logic analyser doesn’t make it easy to check the validity of the ROM outputs. I could see though that pressing keys altered the output of E10, the Video Shift Register. It seemed to me though that the DC012 was probably outputting what it was being given, as something did get displayed and it did vary a little depending on what keys were pressed.

It was while checking the data sent to the DC012 that I realised that it was E10, the Video Shift Register, that was getting hot. In fact after a few minutes it would be too hot to touch. The same chip on the VT102 does not do this. I tried replacing E10, but the new part also gets hot. I am still not clear about the cause of this.

After a while I realised that the vertical flashing bar was just the cursor being displayed and being stretched the full height of the screen. I imagined that perhaps the first line of characters was somehow being displayed repeatedly across all lines of the screen, except that the characters I typed did not appear on later lines. After asking on the Vintage Computer Federation forums for advice I realised that perhaps it wasn’t the first line of characters that was being repeated, but the first line of pixels. I tested this by typing some different characters that had pixels at the top of the character cell, sure enough those pixels appeared across the full height of the screen. So now I knew that it was the first line of pixels that was being repeated.

The next thing to do was to check the Scan Count being output from the DC012 Video Control chip. This seemed to be either all zeroes or all ones. On the working VT102 it counts up as expected. I concluded that the DC012 must be faulty. It could even be why E10 gets hot, although the DC012 does not get hot.

Turning again to the Vintage Computer Federation forums, a very generous member offered to send me two DC012 chips. I took out the old one, replacing it with a socket, noting that the DC011 was already socketed. This fixed the issue and now I have the external video working, as shown in the image at the top of this post. I noticed that E10, the Video Shift Register, is still getting very hot and the new DC012 is a little warm too.

The current situation is the following:

• I still don’t have an image on the VT100 screen as I removed the monitor board and have taken it apart. I think the flyback transformer has failed, but I can investigate this a bit further, knowing that the video signal is now correct. I am not hopeful though, because although the DC012 was producing the wrong image, the video signal itself was correct.
• The NVRAM chip probably needs replacing.
• The Video Shift Register, E10, is still getting very hot and I need to work out if this is normal.