Fairlight-CMI

Hi.

I think that this comes from one defective bit in a buffer or shift register or even an interrupted printed-wire-connection on the PCB, rather than from a video RAM chip.

Because it happens in every vertical line, and in every horizontal place (I guess every 8 pixels?). So multiple RAM chips would need to be affected all in the same place to cause such a consistent pattern, or one RAM chip would need to output one bit so strongly that it would override anything any other RAM chip would output onto the same line at some other point in time.

I fixed an earlier video board of a CMI II several months ago, and if I remember it right, there is a parallel to serial shift register on the way from RAM (8 bits parallel) to video out (8 bits serial). Clamping one bit on the input of that register (?) to (I guess) low (?), would give a pattern similar to what you photographed (clamping one input bit after another would do the trick for either of the 8 bits in the sequence, and maybe I did this only after I had removed another socketed IC right before that shift register? Unsure.)

Probably, your problem can be either on the way from CPU to all RAMs or on the way from all RAMs to that shift register.

You can find out whether it's CPU to RAM by writing a few different values to Video RAM in the debug console, and reading them back again. If you get them all back ok, then the problem is most probably between RAM and the shift register. Take into account possible auto xy update logic etc when doing the test.

Sorry it's been some time ago I did that work; I can look it up in my notes, but rather after 17th of Dec.

Greetings, Joerg

--- In Fairlight-CMI@yahoogroups.com, Gordon JC Pearce <gordon@...> wrote:
>
> dvdborn wrote:
> > Since a while my IIx has vertical lines across the screen.
> > 
> > You can see a picture of it here: http://kunst.phl.be/~dvdborn/fairlight/vertical_lines_IIx.JPG
> > 
> > Does anybody know which component in particular is the culprit?
> > 
> > David
> > http://dvdborn.blogspot.com
> 
> Dying video RAM.  It's possible it's also some of the glue logic 
> surrounding the video RAM, or the shifter, but the dynamic RAM is pretty 
> fragile.
> 
> Gordon MM0YEQ