SE/30 display issues; random pixel artifacts on screen

[roj_matt]

I recently recapped and refurbished an SE/30 to the point where the sound is loud and clear, and everything else seems to work great except for the display itself.  During my refurbish process I replaced the CRT, as the old one had severe burn-in.  Things appear much clearer on the new one, but random artifacts appear on the screen.  As you can see in the photo, there are some white pixels right by the mouse pointer, the MacintoshHD window (and its icon) and other places on the screen.  As I move the mouse pointer or drag the window elsewhere, the pixel artifacts follow the movement.  I also get flickering pixels (not visible in the photo) that get worse the longer I have the machine powered on.  It kind of looks like there is some kind of interference, but I don't have any other electronics powered on nearby.

I'm pretty sure something is faulty on either the analog board or the yoke/CRT, or both.  I have a working SE/30 that I've used to test this machine's logic board, and it works great. 

I haven't replaced any components or done anything else on the analog board--it's in the same state as when I received this machine.  I did try adjusting the focus, cut-off, height, etc.  No change.

I've visually inspected the analog board, and there are no obvious smoking guns (leakings caps, cracked solder joints, residue, etc.).  I've read up on other posts where replacing the flyback transformer fixed certain display issues (but none of the issues I can find are exactly like mine).  Can someone confirm that this might be my problem?  Any other ideas where I should start as far checking/replacing components? 

 

[apm]

Flipped pixels like this would pretty much have to be a digital issue. There's just no way an analog failure in the flyback, yoke or CRT would produce clean pixels with the wrong value like this.

If the same logic board works in your other SE/30, then check the voltage levels, especially +5V. Maybe it's out of range and causing glitches on the logic board. On the analog board, the only component I could imagine causing this would be the inverter U1, but even that seems pretty unlikely.

The white pixels track objects on the screen, but don't appear in the middle of the grey pattern on the background. It suggests to me a problem on writing to the VRAM UC6 and UC7 (check the continuity on the address lines perhaps).

 

[roj_matt]

Flipped pixels like this would pretty much have to be a digital issue. There's just no way an analog failure in the flyback, yoke or CRT would produce clean pixels with the wrong value like this.

If the same logic board works in your other SE/30, then check the voltage levels, especially +5V. Maybe it's out of range and causing glitches on the logic board. On the analog board, the only component I could imagine causing this would be the inverter U1, but even that seems pretty unlikely.

The white pixels track objects on the screen, but don't appear in the middle of the grey pattern on the background. It suggests to me a problem on writing to the VRAM UC6 and UC7 (check the continuity on the address lines perhaps).

Thanks, apm!  I'm an idiot.  So seeing that you mentioned that it sounds more like a digital problem than an analog problem, I went back and swapped logic boards between my working and semi-working machine, just to double check.  The problem followed the logic board.  I could have sworn that I tried this bad board in my other machine....  So this really is a digital issue.  I will check the address lines and report back.

 

[Paralel]

Yep, glitchy video issues like this are almost always related to video RAM issues.

 

[techknight]

Bad VRAM. Seen it before. 

 

[roj_matt]

Flipped pixels like this would pretty much have to be a digital issue. There's just no way an analog failure in the flyback, yoke or CRT would produce clean pixels with the wrong value like this.

If the same logic board works in your other SE/30, then check the voltage levels, especially +5V. Maybe it's out of range and causing glitches on the logic board. On the analog board, the only component I could imagine causing this would be the inverter U1, but even that seems pretty unlikely.

The white pixels track objects on the screen, but don't appear in the middle of the grey pattern on the background. It suggests to me a problem on writing to the VRAM UC6 and UC7 (check the continuity on the address lines perhaps).

I checked continuity with all 24 pins on UC6 and UC7, and everything tested ok.  As techknight and Paralel have stated, it must be bad VRAM (which means UC6 and/or UC7 may need to be replaced...).  I was REALLY hoping it would be a bad trace, or something like that that's easy to fix...

Bad VRAM. Seen it before. 

What am I up against here?  It seems like the most difficult thing here would be acquiring 2 replacement ICs for UC6 and UC7.  My initial Google searches reveal that these chips aren't easy to come by (but I could be wrong).  I've sifted through some of the other topics, and apparently you have replaced VRAM chips before.  Do you know where I can find these chips?

 

[roj_matt]

So there are one or two SE/30s on eBay for parts only. I'm considering buying one just to get the VRAM chips. I'd prefer to buy the chips directly if they are available, to avoid the risk of buying a broken machine that won't fix my problem. Does anyone know if they are still available somewhere? My usual electronics sources (Mouser, Jameco) don't stock any, which makes me think that the chips aren't made any more.

 

[techknight]

I actually have a few battery destroyed SE/30 boards around here somewhere, might be able to steal them from that. 

Also you can get the chips on ebay.

 

[roj_matt]

I actually have a few battery destroyed SE/30 boards around here somewhere, might be able to steal them from that. 

Also you can get the chips on ebay.

Thanks!  I found MT42C4064-12 chips on eBay, but not the MT42C4064-15 (which are currently used on the board).  Comparing spec sheets, the only difference I can see is max access time (120 ms vs 150 ms).  Is it safe to assume that the -12s will work just in place of the -15s?

 

[techknight]

yes. But not the other way around. 

 

[roj_matt]

So I successfully replaced UC6 and UC7 with MT42C4064-12 chips.  After all the soldering and desoldering, I was excited to hear the familiar chime after all was said and done, but was extremely disappointed that the problem persists.  In fact, I'm wondering if there might be multiple issues.  White or black pixels still track cursor and window movement on the screen, but there is also a lot of pixel flickering.  It kind of reminds me of "snow" on old analog TVs when a TV signal is weak.  The other strange thing is that this flickering matches hard drive activity.  So there seems to be an electrical issue somewhere.  Since I can't seem to upload video files, here is a Dropbox link to a video that shows what I'm seeing: https://www.dropbox.com/s/t92izeq7j80jzuu/se30_pixel_issues.mp4?dl=0

I have a working SE/30, and I see the problems when I install the board into either machine. 

Any help or suggestions are greatly appreciated!  My wife is wanting me to downsize my vintage computer collection, and I need to find a good home for this SE/30, but I'd like to get it working like new again, and not make this someone else's problem.

 

[Paralel]

You might need to end up replacing the other RAM chips.

Also, if its timed to HD activity, that would strongly suggest analog board instability. If it isn't recapped, it should be. Without the video issue, the analog board instability would most likely show up as a screen "wobble" when the HD is active.

 

[techknight]

IF that didnt work, try UE8. 

 

[roj_matt]

IF that didnt work, try UE8. 

Techknight, chalk another one up to your number of SE/30s saved.  I replaced UE8 with a new 74LS166AD, and my video is like new again!

Thank you thank you!

 

[uniserver]

yeah UE8 is a problem with these se/30's - good that you finally figured it out.

bad UE8 can give you all kinds of strange issues.   Cap goo leaks into it and that is when bad things start to happen.