If problem has disappeared, shouldn't those 4 screens look the same? (edit: sorry missed that RND(1) function - perhaps it would be better to run something that should give a known pattern?).
Anyway: if ROM or main RAM is faulty, machine wouldn't be stable but quickly hang. So if you can type (perhaps not see) & run BASIC programs, hear sound with a game cartridge etc, you can safely assume that problem is in VRAM or VDP (with VRAM being the usual suspect).
aha, ok, well, the problem now is find vram to replacement, is easy? where? expensive?
The first screen is correct. It shows horizontal and vertical lines every 5 points.
The strange thing is that the rest of screens fails the same way. Though the colors are random, the last two lines of characters seem to have the same failure.
@RetroTechie: the 4 screens should look the shame, random colors, but the same aspect.
Have you allready tried tapping gently underneath the MSX to see if there are changes in the patterns on the screen. If tapping does not change the output then the chips itself are screwed otherwise it is a bad solder.
I thinks, there is a problem with the REFRESH LINE to the VRAM chips.... and not that the chips are wrong (all vram chips failing at once?) nah, is more probably that just the refresh signal is broke (soldering or the track), and that is all.
the symtons of the refresh being bad, is: first clean CORRECT screen, destroying with time, and after 10 seconds, total corruption of the screen
So, the MSX only needs a bit of attention, and none part to buy.
and HOw identify the refresh line in the pcb?
check solder on /CAS and /RAS pin 5 and 16 on memory chips
re-solder all the VRAM chips, may be a poor contact, or change all memory vram's
OK I try it and i comment the results... A lot of thanks