View Full Version : Sinclair ZX81 - bad video

July 28th, 2014, 01:16 AM
Started working again on my old ZX81. After changing the keyboard membrane for a new one I took on fixing the dark video (back porch missing) problem on the micro - had a good experience with RWAP Software (keyboard replacement) so I sent for their ZX Vid composite retrofit which also corrects the back porch problem; installation was not a problem - only when I tried testing the new setup did things not work as they should. I still had a very dark image on the CRT monitor (composite path) and the raster failed to lock drifting vertically, tried adjusting the pots on the new video board to not much of a success, managed to lock the screen in place but with a weird bright white background on the top 1/3 of the CRT and dark as it was on the bottom, still I was able to discern the "K" at the bottom line - a bot better than with the old RF modulator in place.

Decided to leave the ZX81 idle to see if something would warm up inside enough to cause a change, and it did: the white top portion gave way to end up with a full black screen as seen on the photo:


Fixing up a quick "HELLO WORLD" fill the screen program produced the following:


As bad as this is it is better than with the former RF in place.

The ULA chip gets quite warm, pressing on the board around its area will cause the display to garble and sometimes put out an all white screen - it goes back to the thing seen up on the pictures as soon as the pressing stops - leads me to believe two things: the ULA is defective and also that there is a cold solder joint somewhere on the mainboard - removed and reseated all the socketed chips to no avail.

I'll run my soldering iron on the board just to whiff out any possible solder caused issues, still I suspect that the chip is in bad shape, from what I have read on ZX81 web links these are hard to find, making it a better solution just to get a "new" ZX81/TS1000.

This is one of those moments I wish I had a scope at home, only way is to try and borrow some lab time at the place I work.

Any ideas or references to typical solutions with a problem like this are welcome! :rolleyes:

-- RP