• 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.

SE/30 Jailbar Pattern before and after recapping

DannyDorito

Active member
Hello! I know this is a common question, but I just recapped my SE/30 and am getting the Jailbar pattern. The computer completely works, has sound, is stable, etc. Just an issue with the video circuitry. I understand that usually this pattern is linked to an issue with a trace being eaten away and a single data line being gone. I've downloaded the schematics, but I get no continuity on any of the data lines when testing between the video ram chips and UE8 or the Multiplexers, so I feel I must be doing something wrong. After all, it's only one line of pixels out every 8. If I understand the SE/30 video system correctly, that means only one data line is out. I guess I just need help understanding what could be causing this if it's not a broken trace, or how to test what traces aren't making it since clearly I'm doing it wrong.

 
Last edited:

psa425

Member
I had the exact same issue a few years ago. I had a bad trace between UE8 pin 3 to UC7 pin 3. Here's a pic of the bodge, no more jail bars. Maybe your issue is the same line.
 

Attachments

  • UE8 Bad Trace.jpg
    UE8 Bad Trace.jpg
    136.2 KB · Views: 58

DJ68K

Member
I second what psa425 says: It's definitely a broken trace somewhere in the video RAM circuit. I fixed it on my SE/30 by toning out every single leg on those chips looking for broken continuity. As you said, there's 1 bad and 7 working pixels. It's an 8-bit data path (4 bits per VRAM chip) and one of those bits is stuck off the entire time. You could try reflowing the legs of UE8 but in my case I had to attach a bodge wire between one of the legs to the UC6 RAM chip.

Look at the SE/30 schematics and find UE8. Then use a multimeter in continuity mode and tone out every single leg on the chip to multiple destinations but especially the video RAM chips. Double and triple check before soldering in a bodge wire. This one took me a few hours to fully figure out but during that time I learned how to read and understand the schematics and it's a skill that's come in handy ever since.

Note that UE8 doesn't connect to the video ram with the system's data lines, it uses its own data lines. 4 pits per VRAM chip. If you're not getting any continuity, it's because you might be checking for continuity between UE8's video data line to a system data line. Check between VID(0)-(7) on UE8 to the four legs on each of the VRAM chips.
 

Attachments

  • IMG_0780.jpeg
    IMG_0780.jpeg
    2.2 MB · Views: 37
  • IMG_0783.jpeg
    IMG_0783.jpeg
    2.9 MB · Views: 28
  • IMG_0784.jpeg
    IMG_0784.jpeg
    2.9 MB · Views: 30
  • IMG_0785.jpeg
    IMG_0785.jpeg
    2.3 MB · Views: 44
Last edited:

MattH

New member
Hi, newbie to hardware restoration here, thanks for all of the good info on this forum!

I've got a similar problem on my SE/30, but with two-pixel-wide black bars. The system has sound and boots from my BlueSCSI after my amateur recapping (but ADB is dead for now, sigh). I was able to trace VID(0) through VID(7) and sure enough two of them had flaky or missing continuity. I reflowed the legs on UE8 and one of them seems to have continuity now, but the black bars are still two pixels wide. (The black bars were also present before the recapping.)

I think my next step is to try to jumper the two VID signals? And if that doesn't work, it means the pins might be corroded up inside the IC? UE8 was pretty green around the pins. Edit: I guess I should probably check that they aren't shorted to something else, huh. Can leftover cap juice under the IC cause that? I gave it an IPA bath but don't have an ultrasonic cleaner.

This might be unrelated, but: When new capacitors are installed, what should continuity testing sound like? Some of mine do the "brief beep" as expected, but others have continuity between both sides. I uninstalled a couple of those caps, and they test normally ("brief beep") out in the open air. Puzzled, I checked the bare SMT pads: +/- have continuity with each other even when nothing is installed. Is this because the caps are providing "reserve power" to guard against voltage drops, or does that perhaps indicate that something else is wrong? I'm still learning to read schematics, sorry if this should be obvious.
 
Last edited:

DannyDorito

Active member
My issue did end up being because of a broken vram trace. Now onto the ADB issue...

My experience is limited, but your issue sounds like cap juice messing with the connections. Try poking around with a jumper anyway, to see if you can jump the ram chips to UE8 on vid 0 and 7. That's how I figured it out, because my continuity was so flaky.
If you have a hot air station it might also be beneficial to remove UE8 and clean under it. It's a lot of work but it may be worth it.
It could also be that UE8 is bad and you'll have to source a replacement.
 

Kouzui

Well-known member
Gonna bump this thread, as I have the same issue.

ue8andfriends.jpg
se30jailbars.jpg

I've reflowed UE8.

I haven't done much continuity testing yet. I'll update if I fix it.
 

Attachments

  • Screen Shot 2024-02-21 at 8.57.38 PM.png
    Screen Shot 2024-02-21 at 8.57.38 PM.png
    608.6 KB · Views: 33

joeshay

New member
I got a weird jail bar after the recap, which is kind of different from others.
jailbar_se30.jpg
It only jail-bar on the gray desktop pattern but the withe image is un-break.
I did bug all the traces and nothing broke of data A() lines from Video ROM UK6 to UA8/UB8/UC8/UE8 and UC6/7.

Is this a broken chip issue or likely a broken line somewhere other than the video interface?
IMG_3434.jpg
 
Top