Nivag, we started chasing why we are not getting a full screen of 'E' characters.
My guess is that the character row counter (H11) was misbehaving (either internally or as a result of the signals being fed to it).
I came to this conclusion by frame stepping the video that
@Desperado posted and I could see the bulk of the screen (75 to 80% or so) had an 'E' character, with the remaining screen characters having (what appeared to me) to be the wrong part of the 'E' character displayed.
We looked at the data input and output of F9 and those seemed as solid as a rock. Basically, at the static levels of the PETSCII for the character 'E'.
I then set
@Desperado the task of just looking at some of the simple logic signals around this schematic to (a) get him used to using his oscilloscope again and (b) looking for anything untoward.
In probing around, we found that measuring E11 pin 2 (CLK8) had a disastrous affect on the circuitry. This has led both Hugo and myself to suspect G5 is not driving too well, or when G5 was replaced, the soldering, PCB tracking or whatever is not too good.
Now what seems to have happened is that what WAS a temperature-related issue is now a permanent issue? So, we need to identify why...
G5 being partially faulty could account for a number of observed issues with this machine. However,
@Desperado doesn't have any more spares of this device.
I hope that catches you up with today's work?
Dave