• Updated 2023-07-12: Hello, Guest! Welcome back, and be sure to check out this follow-up post about our outage a week or so ago.

Mac SE/30 Video Issues - Suspect UC6 VRAM chip

I was in the middle of fixing up a Mac SE/30 when I noticed the image began looking glitchy (was trying to get MODE32 working since the Rominator II is out of stock right now; So I had two BlueSCSIs plugged in along with a floppy emu).
Like so:
IMG_0362.jpeg
I have recapped the system, and have checked the continuity of each cap, so I believe that is on the up and up. I removed the ROM to force the Mac to push black and white horizontal bars to the CRT; that looked correct, so I believe that the analog board is okay as well. Aside from the visual oddities, the system works basically fine. I traced out all the multiplexer chips and UE8 and I couldn't find any broken traces. Finally, I decided to use my hot air station to heat up all the chips to see if it would temporary improve things. I found that if I heated up UC6 that the image went back to normal which was great. However, over time, the image would degrade back to the glitchy state. Below is a picture of it working after heating up UC6.
9A26AE75-887D-4EB5-8036-0E02C8144958_1_105_c.jpeg

I should mention that the board had very little in the way of corrosion and that I have given it multiple IPA baths, plus a fair amount of deoxit.

At this point, I was wondering what my steps forward should be. I have also checked all the SO/VID pins on UC6 against UE8.
 

Phipli

Well-known member
Have you tried reflowing the solder in UC6? Heat can cause movement due to expansion and the intermittent behaviour might be as a dry joint expands and contracts. Continuity testing can apply enough pressure to make the pin contact, so in some cases isn't reliable.
 
Can you please post some high res pics of your logicboard?
Sorry, should have done this in the first place. Please excuse any liquid or grease that you see on the board, it is either deoxit or flux which I haven't cleaned yet. Also, please excuse the bodge job around C9, that pad tore off completely.
 

Attachments

  • IMG_4553.jpeg
    IMG_4553.jpeg
    3.7 MB · Views: 19
  • IMG_4554.jpeg
    IMG_4554.jpeg
    2.6 MB · Views: 20
  • IMG_4555.jpeg
    IMG_4555.jpeg
    2.1 MB · Views: 19
  • IMG_4556.jpeg
    IMG_4556.jpeg
    2.3 MB · Views: 17
  • IMG_4557.jpeg
    IMG_4557.jpeg
    2.5 MB · Views: 15
  • IMG_4558.jpeg
    IMG_4558.jpeg
    2.5 MB · Views: 12
  • IMG_4559.jpeg
    IMG_4559.jpeg
    1.9 MB · Views: 14
  • IMG_4560.jpeg
    IMG_4560.jpeg
    3.5 MB · Views: 11
Have you tried reflowing the solder in UC6? Heat can cause movement due to expansion and the intermittent behaviour might be as a dry joint expands and contracts. Continuity testing can apply enough pressure to make the pin contact, so in some cases isn't reliable.
I haven't tried reflowing the solder around UC6 yet. I'll try that when I get back from work today.
 

Chopsticks

Well-known member
may even be worth just replacing UC6 with a new chip, those 74F253’s are pretty cheap to buy. also do keep in mind that if there was any capacitor leakage around any of those IC’s best practice it to actually remove and clean under them. Also one of then pins on UC6 looks a little suspect so assuming it’s not just the lighting and camera angle its Possible corrosion has made its way inside the ic. So give it a reflow and recheck continuity to all pins on that chip long term I’d still replace it.
 
I tried reflowing the pins on UC6. There seemed to be very little solder there, so I added a bit more. Did not seem to make much of a difference if at all. I'll go back and recheck continuity on the chip. For now, I have ordered a new NEC uPD41264C-12 to replace the TI chip that is currently in UC6's spot. I'll looking into buying some 74F253s to replace all the U*8 chips.

Edit: Does anyone know a good socket for the VRAM? If I am gonna remove it, I'd like to make it socketed too.
 
Quick addendum:
I have attached a video of the mac in action. I have noticed that in the columns of the video, there seems to be a duplicate cursor, if I move the cursor around, as if the those two memory addresses are getting the same information.
 

Attachments

  • IMG_4570.MOV
    95.8 MB
Additional update: I tried heating up the chip and surrounding solder to fully rule out a bad joint; after doing so, heat no longer fixes the video output, so safe to say the VRAM (or half of it) is cooked (pun intended).
 

Chopsticks

Well-known member
Top