exxos's Atari, Amiga & retro forum

I'm trying to fix my friend's 1040 STF. After running for a while, it starts to show rogue pixels in some places and occasionally it crashes.

The ST is fully stock, except for an installed Gotek FDD replacement.

Let me start with an example of what's going on:

Here are some of my observations:

- The corrupted pixels aren't changing as long as the part of the screen they're in isn't redrawn.
- The problems show up in games but I've seen it happen on the desktop as well.
- The problems only start to arise after warming up the ST for an hour or so.
- There are occasional crashes, often in predictable scenarios (specific demo screens and games).
- Some games appear unaffected while other quickly show the said problems, this pattern seems to be rather steady.
- RAM tests tend to fail.

And here some of my analyses:

- If you look at the pixels on the bottom part of the screen (see example above), you can easily jump to the conclusion that there must be noise or some other distortion on a single data line, as the pattern is repeating exactly every 16 pixels which is exactly the width of the data bus.
- Given the fact that the corruption is only visible in the part of the screen that doesn't get redrawn every frame and that the pixels don't change location, I think the RAM isn't faulty, it must be the transport to it that has a problem. The shifter simply displays what's in RAM and does so consistently, nothing random there.
- Since there are crashes as well, this probably isn't just video related.

This has been going on for a while now and I've tried the following:

- Recap the PSU (Exxo kit)
- Replace the mainboard 4700uF capacitor with Exxos' replacement cap
- Reseat the MMU and GLUE chips (cleant the socket and legs)
- Reflow the MMU and GLUE socket legs on the bottom side of the PCB
- Firmly press the shifter and any other socketed chips

So what am I looking at now, an electronics problem with noise on a data line? I do have a scope but I have only little experience using it (this may be the perfect occasion to learn though). I do have average soldering experience and only basic understanding of electronics.

What do you guys suggest I try next?

The STF video is pretty bad to start with, there might not be actually anything wrong with it. Someone (cant remember who sorry) actually investigated the video issue and routed more wires for 5V and 0V to the shifter and it improved things. Though that is mostly interference type issue.

Bad pixels may well be bad RAM. https://www.exxosforum.co.uk/forum/viewt ... ?f=25&t=28

exxos wrote: 26 Sep 2018 22:45 The STF video is pretty bad to start with, there might not be actually anything wrong with it. Someone (cant remember who sorry) actually investigated the video issue and routed more wires for 5V and 0V to the shifter and it improved things. Though that is mostly interference type issue.

Well, video quality is perfect otherwise, no visible ghosting, interference or anything. The motherboard is a C103088 REV. 2.A so it's not Silly Style.

exxos wrote: 26 Sep 2018 22:45 Bad pixels may well be bad RAM. https://www.exxosforum.co.uk/forum/viewt ... ?f=25&t=28

When the ST is still cold, there are no problems. Does that make sense when having bad RAM?

Here's a picture of the mainboard (C103088 REV. 2.A).

Those style boards normally are OK video wise.. you sure its just not a bad cable or bad connection on the cable ? Could be some bad contacts on MMU or shifter socket, or even bad chips. RAM is likely cause.. and easy to leave testing.

Here's a RAM test result:


And here's a video of a demo screen going all wrong:

exxos wrote: 26 Sep 2018 23:00 Those style boards normally are OK video wise.. you sure its just not a bad cable or bad connection on the cable ? Could be some bad contacts on MMU or shifter socket, or even bad chips. RAM is likely cause.. and easy to leave testing.

I don't see how any cable could cause this, the pixels stay put in the first screenshot (they don't move anywhere and no additional pixels appear), so definitely no video cable problem.

We've cleaned the MMU socket and legs with alcohol. I must say I haven't reseated the shifter yet. But given the consistent pattern (not changing until the particular part of the video ram is written to) I can't imagine the shifter doing being the cause. I mean, the shifter doesn't know what RAM is overwritten every frame and what part is not, am I right? This really sounds like something is going wrong between the MMU and the RAM.

sandord wrote: 26 Sep 2018 23:04 Here's a RAM test result:

Showing bad bits in RAM then...

exxos wrote: 26 Sep 2018 23:10 Showing bad bits in RAM then...

I was thinking that when there's a problem with the MMU or something like that, RAM tests would fail, even if RAM was OK. Maybe I'm being too optimistic but hey ;)

If that address is in the upper half of the 1MB area, you could disable the upper bank leaving 512K and try again...