Search results

Made that for a while, it's open source! https://github.com/buserror/libmui

So, having replaced the PSU with another, and still having the wavy video issue, that narrowed down the problem to the analog board proper. So I recapped it. I also changed the 4 pin connector housing/plug that was completely perished. I didn't find any obvious 'bad cap' -- definitely not...

Oops didn't last. 10 minutes of stable video and the problem came back :-/

Ok tested. And that transistor WAS the problem; video is nice and stable now, so the fixed anode cable worked fine... I probably need to work a bit more on that board, the 4 pin connector is burned as many are, and it could use a new fan, but at least the video is great!

In other news, Q2, that famous transistor, was also replaced. Can't test right now as my wife's on a conf call and not going to like if I smoke/trip the place up :-)

Well that was fun. Anode wire broke in the cup as I pulled the cup out... I thought I was screwed, but I managed to remove the cable from the cup, the glue was pretty weak with age! Cleaned the clip, removed 1cm of extra insulation and reassembled it. As good as new, I hope.

So, I got my replacement transistors... is there a schematic or a hint on *which* transistor to change, for someone who doesn't know the layout or gory details of how the analog board works? :-)

I recapped my SE/30, reflowed the big connector on the analog board. Deoxed the trim pots... Machine is stable, but the video isn't! The video 'jitters' a bit left/right -- not necessarily the whole screen, it can be the center of the video while the border stay in place... but it varies a...

I just got a LC475 from @Kai Robinson -- it has that issue -- did anyone found a likely cause? I could rework the audio chip but if there's an alternative solution...?

On pre-g3/g4 it was often a good idea to convert strategically from C to asm. But I would never 'write from scratch' in asm; I'd make the code in C first, disassemble it and see how bad/good it was. VERY often just changing the C code a bit was enough to get the optimiser to do the 'right thing'...