SE/30 with vertical jail bars - looking for a direction steer on my next steps

[lfletche]

Hey all, hoping someone can give me a bit of a steer on a problem I am trying to solve with an SE/30 I picked up last week. The logic board itself thankfully had a non-leaking original battery installed which I was relieved about - but the capacitors had done some pretty serious leaking, and in one case the capacitor was totally MIA.

A good clean, and full recap later the board is booting with chime perfectly - but I have video issues. I've included some pics here but you can see it is has permanent vertical jail bars. I've been doing lots of reading up on this issue and it seems to be a common enough problem, but I wanted to get a second opinion before I start next level of trying to fix this. Here are the things I have tried so far:

• Checked all traces to and from the VRAM chips at UC6 & UC7. Also checked all traces on UE4, UE7, UG6, UG7 & UI6. All check out fine.
• Swapped the chips installed at UE4, UE7, UG6, UG7 & UI6 as well as the video ROM with another working SE/30 board. Problem still happens.
• Checked all traces on UA8, UB8, UC8, UD8, UE8, UF8 & UG8. All check out fine, also confirmed there was no ground shorts on any pins that should not have had ground.
• Reflowed UE8. Didn't help.

My gut and the reading I have been doing is telling me the problem is with UE8, or perhaps one of the A, B, C or D chips next to it and that my next step is to lift the chips fully clean under them and then resolder back on. But I am also wondering how likely is it that one of these chips has actually gone faulty and I need to replace it? Is there anyway to test to confirm for sure if one of these is likely faulty?

The other interesting thing I have noticed is that when I let the SE/30 boot into System 6 and then open a window and move it around the window (from the parts I can see) becomes garbled. Closing the window and re-opening it clears that and the window appears correctly again (although still with bars on it), until you move the window again and then it becomes garbled again. This made me wonder if the problem was more likely with the VRAM chips?

Thanks for your help, any steer on the next direction I take in troubleshooting this is greatly appreciated.

 

[Drake]

You're in the right train of thought. It's typically an eaten trace between the two highlighted sections

 

[lfletche]

Thanks @Drake - I’ve since gone through and removed UA8 through UG8 from the board, throughly cleaned the board as it did still have some cap juice under the chips. I then went through and checked all the traces from all those chips to ensure they all get where they are meant to as per the schematics - everything checked out and was going where is is meant to.

I’ve now gone and put UA8 through UG8 back on the board - but this time I used known good chips. I have another SE/30 board that has some other weird issues that I am still working on - but it’s video works fine so I used this as an opportunity to clean under the chips on that board as well and have swapped UA8 through UG8 between the two boards.

Everything back on and cleaned up and the jail bars fault still continues on the original board with new chips.

I’m starting to wonder if those video memory chips that are soldered to the board are the fault here. Would one of them being faulty cause this issue?

 

[Drake]

@techknight any insight?
I'd think those are the culprit since the consistency of the missing segments

 

[YMK]

The other interesting thing I have noticed is that when I let the SE/30 boot into System 6 and then open a window and move it around the window (from the parts I can see) becomes garbled. Closing the window and re-opening it clears that and the window appears correctly again (although still with bars on it), until you move the window again and then it becomes garbled again. This made me wonder if the problem was more likely with the VRAM chips?

This sounds like a broken or shorted address line.

 

[techknight]

@techknight any insight?
I'd think those are the culprit since the consistency of the missing segments

I have gotten to the point where I replace those chips anymore. They easily get damaged from the capacitor leakage. UE8 is mandatory, and the rest are secondary.

Also I have found that sometimes the VRAM itself gets damaged on the data bit outputs and its missing a bit. From the look of the raster, you might be missing 2 to 4 bits consecutively, so its entirely possible that one of the VRAM ICs is bad in its entirety.

 

[lfletche]

Thanks @techknight - I have found a source for those VRAM chips so have a couple on order. I think that is my next step to try swapping them out, when I look at the current VRAM chips under a microscope on a few of the pins there looks like there are some signs of rust where the pin leg goes into the chip - so I makes me think you are correct and one of those chips has totally died.

Only thing I am not looking forward to is how to get those old chips off the board. I feel like the best way is going to be to try and suck the solder out of each hole?

 

[Drake]

Thanks @techknight - I have found a source for those VRAM chips so have a couple on order. I think that is my next step to try swapping them out, when I look at the current VRAM chips under a microscope on a few of the pins there looks like there are some signs of rust where the pin leg goes into the chip - so I makes me think you are correct and one of those chips has totally died.

Only thing I am not looking forward to is how to get those old chips off the board. I feel like the best way is going to be to try and suck the solder out of each hole?

lots of flux then ADD a bunch of new solder, hopefully you have a gun to suck it out ala FR301 or you'll have a hell of a time. Worse case just snip the legs away from the body and solder to the old ones and you'll avoid damaging the board further

 

[wottle]

So you didn’t replace UE8? I had a very similar problem that was solved with replacing UE8. https://tinkerdifferent.com/threads/se-30-jail-bars.2312/

 

[lfletche]

So you didn’t replace UE8? I had a very similar problem that was solved with replacing UE8. https://tinkerdifferent.com/threads/se-30-jail-bars.2312/

Nah I have swapped UE8 with a known good chip

 

[lfletche]

Ok, I think we might be on the right path with these VRAM chips.... Today I cut/disconnected the 5V leg of the UC7 VRAM chip (pin 12) and turned the machine on and jailbar fault looks exactly the same even though that whole chip has been disconnected from the 5V feed.

Then repeated the same test on the other UC6 VRAM chip and now I have a total blank screen which is to be expected.

Then I reconnected pin 12 of UC7 and still blank screen. Reconnected pin 12 of UC6 and jailbar is back.

So my current theory is UC7 chip has failed.

The replacements I have sourced just arrived today so I will look to swap it this weekend. Only thing I have noted is the installed chips are TMS-4461-15NL. The new ones I have are 12NL. As far as I can tell that means they are a little bit slower? Would that be an issue?

I did note my other faulty SE/30 board has two 12NL chips on it. So I guess just depends when it was manufactured?

 

[Patrick]

i'm supper newb at this.

but i think lower numbers are faster.

Access Time-Max

120.0 ns

vs

Access Time-Max

150.0 ns

i bet either would work.

that is, its ok to go with a faster chip.

 

[lfletche]

Success!!! The jailbars are gone - the UC7 VRAM chip was the culprit. I’ve replaced both UC7 & UC6 just for the sake of completeness but the fault was 100% the UC7 chip.

I ended up cutting the legs on the old chips a then heating up the remains of the old pins and pulling them out. Then just used solder wick to remove all the old solder and clean up the holes. I then installed IC socket strips rather then soldering the new chips directly to the board so that I can easily swap them out if my new chips had any issues. But so far, all good. This SE/30 is back in business!!

Thanks everyone for your help & suggestions. Especially big thanks @Drake & @techknight!

 

[Drake]

congratulations! the reward was friendship all along...
and a working computer.