Random Mac Classic failure

[gfuller]

I didn't foresee this happening.

My 1991 Macintosh Classic has been mostly reliable since I recapped its analog and logic boards. (The only problems were caused by a loose memory expansion card, easily fixed.) I've been using it near daily for a little over a month or so — powered on for about 4-8+ hours a day with no issues. (Most of the time it's sitting idle in the background with the AfterDark skyline screensaver running.) That is until today.

Machine was running perfectly this morning, with me doing some work in PageMaker and FreeHand. Unfortunately, while I wasn't paying attention, the machine decided to fail to a strange striped pattern. Doing a soft restart (via restart button) just placed the machine into the familiar "chime cycle" I experienced when the machine's AB died before I had it recapped. Nothing changed after a hard restart. Sometimes it will make loud buzzing noises through the speaker, and one time I believe it even tried to access the floppy (without a disk inside). It also seems to go to the interrupt entry prompt. One time it actually gave me a Sad Mac error, which is my first in a while.

https://twitter.com/i/web/status/1659628376334712849

I've had this machine recapped just a couple years ago, both analog and logic boards. I've also replaced its PRAM battery with a modern replacement, as I did want to try to keep the time intact. (And it has been keeping time somewhat well.)

I immediately suspected the aforementioned memory expansion card, as maybe I didn't get it completely seated back when I reassembled the machine after my BlueSCSI installation. I removed the bucket, removed the expansion card and reinstalled it to see if it would boot up. No dice... the machine booted into System 7, but immediately crashed and started going haywire (lines/patterns on screen, buzzing from speaker). I attempted again, but same result.

https://twitter.com/i/web/status/1659635822025490432

I've looked around and nothing looks suspicious right off the bat. All cables on the logic board are tight.

https://twitter.com/i/web/status/1659640327341240323

Any ideas on a possible culprit? Hopefully it's something not too difficult as I'm limited on space and don't even have all my equipment with me, not to mention my lack of advanced knowledge and equipment. (I'm worried that me not wearing an ESD-bracelet while installing the BlueSCSI may have done some damage? I made sure to frequently ground myself on chassis ground... not too many grounding options here.)

Any help would be greatly appreciated.
Thanks,
Garrett

 

[Mu0n]

a ground wired ESD-bracelet is not a bad idea, but it's very likely not the culprit here. Some people have made videos about actively zapping chips and computers with ESD and it turns out you really have to work for it to make it happen.

Here's my first round of obvious things to check but must be ruled out before people can offer other suggestions:

-is the interrupt switch always turned on due to mechanical abnormality, forcing this repetitve state of booting? you can test for continuity while the computer turned off, logic board out and check for continuity

-what are you getting as voltages out of the external disk drive port while it's running, here's the pinout so you can check with a multimeter. if your DMM probes are too thick, then consider using a simple wire poking into the right pins and use crocodile leads to your DMM.

From Inside Macintosh vol 3:

from a another forum

-can you get us a highest res as possible clear picture of the logic board for us to inspect?

-a problem appearing after prolonged idling could be a marginal connection finally failing due to temperature dilation. Best case scenario: reseat every chips/RAM that's in a socket properly (possibly treating the metal junctions with at least alcohol rubbing, or deoxit if you have it - watch Adrian's Digital basement's work to see him use at least every other video). Worst case scenario, you have solder points to resolder because they've developped cracks. Sometimes, they're real hard to see to the naked eye, you'll only see them with a very zoomed in view of a good macro capable camera or microscope. They tend to develop under every place on the underside of a board that has the big white J connector sockets. Working on these machines often means you disassemble the boards and disconnect these cables. The act of taking the machine apart can help push these cracks to develop failures in the long run. It happened with my childhool Mac Plus, its video was suddenly only displaying in a tiny vertical line. Resoldering the cable solder points between logic and analog board fixed my precious baby to its former glory.

The crack in question:

 

[Mu0n]

I watched your latest video and I think more and more that your interrupt switch is stuck.

Check people discussing it:

Macintosh SE FDHD 0000000F and 0000000D error

How to fix this error? Does anyone have experience?

68kmla.org

 

[Ron's Computer Videos]

Has the board been ultrasonically cleaned?

 

[gfuller]

I watched your latest video and I think more and more that your interrupt switch is stuck.

Check people discussing it:

Macintosh SE FDHD 0000000F and 0000000D error

How to fix this error? Does anyone have experience?

68kmla.org

I've seen this mentioned several other times by various people, and it seems like an easy fix.

I tested the machine just a couple minutes ago with the bucket off, eliminating the mechanical case portion of the switch in case one was stuck. Machine exhibited the same issue without the cover, chiming twice and freezing(?) on the first try and going into an infinite "chime cycle" on the second. The plastic pieces inside the case appear to be intact. However, while I can hear an audible "click" and feel a tactile response on the restart (front) button, I don't really feel or hear a click on the interrupt button. The interrupt button feels like it isn't really pressing anything.

I didn't remove the logic board completely, but I test pressed both buttons directly. Both buttons seemed to click and had a tactile feel.

Knowing that others have had similar issues resulting in the same exact error code because of the interrupt switch, that seems to tell me it's most likely the culprit. I'm going to order a multimeter to test for continuity on the switches. (With that said, will both switches need to be removed from circuit to test properly? My soldering is really bad.)

With the multimeter I'll be able to check for voltages also.

As for bad/cracked solder, that's always a possibility and could explain why it might've happened after the machine had been running for some time. That'll be something I also consider.

Regarding the logic board, attached to this post is a photo of both sides of the board that (hopefully) can help if needed. The logic board was ultrasonically cleaned back in May 2020 when I had it recapped.

 

[Mu0n]

However, while I can hear an audible "click" and feel a tactile response on the restart (front) button, I don't really feel or hear a click on the interrupt button. The interrupt button feels like it isn't really pressing anything.

That looks like 2 cracks, though I can't be 100% sure since the picture feels movement-based blurred.

 

[pocketscience]

You should be able to test those switches without removing them from the board. Sounds like a stuck switch from your description though, but a continuity test will confirm..

 

[gfuller]

I checked for continuity on both switches, and it was exactly how I expected: With the switch not actuated, both switches tested open. With the switches pressed/actuated, both switches tested closed. Unless there's something I'm missing (as both are normally-open switches) that rules the switches out. I tested the machine right after making the test to verify that it was indeed experiencing the same fault as earlier this week.

My recap guy has provided some guidance, and my next look will be at voltages coming off the analog board. I'll continue working on this project tomorrow

 

[gfuller]

Update regarding this project...

Voltages checked out fine on the hard drive power connector — 5.04 volts red-black, 12.16 volts orange-black, with the machine continuing to exhibit issues. I've also retested the reset and interrupt switches and both tested open when not pressed, to be expected.

While I had the logic board removed from the computer, I went ahead and did a visual inspection. Meanwhile, I also found the Classic logic board schematics and traced the reset switch back to the UA5 sound chip right next to it. Upon closer look, lo and behold, some of the pins on UA5 look a little suspicious.

A neighboring chip, UB5, also looks a little iffy to me. Both chips have pins that appear to have some corrosion/oxidation on them. I inspected all other chips on the board and those were the only two that really stood out as potentially problematic.

Unfortunately, both chips are surface mount and I just don't have the experience or equipment to remove and clean them. So I'll be sending the board off to be repaired with the LCD board in my upcoming PowerBook 165 restoration project.

While writing this post, I discovered that I had the reset and interrupt mixed up, and thus forgot to investigate the interrupt issue. Reset goes to the UA5 sound chip mentioned above, and I strongly believe it may be the issue. However, what about the strange interrupt dialog glitches? According to the schematic, the NMI switch goes to IPL2. Maybe it's problematic as well?

I'm wondering if this board should be re-cleaned? It was supposedly cleaned back when I got it recapped.