↩ go back to index

Macintosh Classic II Repair — Part 1

September 2, 2023

Part 1: You are here

Part 2

Part 3 (Final)

Written on a MacBook Pro while listening to Blondie's Autoamerican.

Yesterday I got a free, partially functional Macintosh Classic II, courtesy of the Media Archaeology Lab.

If you're willing to go to Boulder, Colorado to pick one up, they might still have some available!

I didn't think to write any posts here about it until just now, so this story was mostly taken from my thread on the Fediverse:

https://tilde.zone/@nytpu/110991723840955445

I'm definitely planning to write a Part 2 once the parts I ordered arrive and I get to soldering, and maybe a Part 3 follow-up once I start actually doing stuff on the machine. There'll definitely be no need for additional parts detailing other issues I needed to fix, right? Right?

But at any rate, when I first got the machine home—after stopping at the thrift store to get an IEC power cord because I have no spares lying around—and turned it on, I got the checkerboard screen, as expected since that's what they said it was doing. Nominally that screen indicates the RAM is bad, but given its age a lot of people agree that it's more likely there's bad capacitors on the logic board.

The Mac on a desk, turned on with the display showing a black and white checkerboard pattern.The Mac on a desk, turned on with the display showing a black and white checkerboard pattern.

Because I'm a weirdo the first thing I was struck by was that this was the first time I've ever seen a monochrome CRT in real life. Color CRTs have a shadow mask or aperture grill that create the illusion of pixels, even though analog displays do not have pixels. The device drawing on the display may use pixels, such as a computer; but the displays themselves only have scanlines and do not have discrete pixels. This is especially visible with monochrome CRTs that have raw phosphors with no disruption, so you can see all the beautiful discrete scanlines (it helps that AFAICT the checkerboard is generated by the analog circuit itself so it doesn't use discrete pixels like the computer normally does).

Here's some of my attempts to show this but as many people can attest, it's very difficult to photograph CRTs. Especially when you're too lazy to get your actual camera and just use your phone.

Closeup of the checkerboard pattern, so you can see the discrete scanlines and see that there's no discrete horizontal pixels. It's out of focus and the white balance is very blue.Closeup of the checkerboard pattern, so you can see the discrete scanlines and see that there's no discrete horizontal pixels. It's out of focus and the white balance is very blue.

A second closeup like above, but with better white balance but visible tearing due to the CRT refreshing at a different rate than the phone camera.A second closeup like above, but with better white balance but visible tearing due to the CRT refreshing at a different rate than the phone camera.

Similar to the previous picture, only bluer againSimilar to the previous picture, only bluer again

Cosmetically, the computer seems to be mostly okay condition, especially if I give the case a good washing when its next taken apart. There is unfortunately some of what I call school IT graffiti on the side for Denver Public Schools. Hopefully it'll come off with some isopropyl alcohol and melamine foam; but I'm honestly okay with having it if it doesn't come off with minimal effort. Gives it some character perhaps, like I say about all my retro stuff when trying to cope about their poor condition.

Before we get to opening it up, since this is the first monochrome CRT I've had I couldn't help myself:

The screen displaying the checkerboard, and I hold a magnet up to it to show the image wobbling slightly, since electrons have a charge and are affected by the magnet.

Opening the Thing

The outer case has four obvious screws, but of course they're Torx T15. The largest driver in my electronics sets is T10, and the smallest driver in my normal tool set is T20; so I had to make a trip to the hardware store to get a T15 driver.

The bottom screws came out fine but then because Apple had to Think Different™, the top screws are underneath the handle and almost impossible to access without making an abomination like I did:

A small drill (because normal-sized drills didn't fit) with a driver extension in the computer. I pull out the drill to reveal another extension attached to the first, and then finally the already 2 inch driver itself. Even this contraption barely reached the screws.

Rest assured, these screws will be taped to the top of the machine until I'm confident I won't need to open it for a while (the case fit is really tight and combined with the bottom screws it holds on just fine).

But with it open (and my fears about the CRT mostly allayed) I got a look at the logic board and the issue immediately became apparent:

A can-shaped electrolytic capacitor labelled “U4 10 16V” with visible leakage out on the PCBA can-shaped electrolytic capacitor labelled “U4 10 16V” with visible leakage out on the PCB

A set of four electrolytic capacitors, three labelled as above and one labelled with a rectangle followed by “2T”, then “1 50V”. There is leakage visible surrounding all four capacitors.A set of four electrolytic capacitors, three labelled as above and one labelled with a rectangle followed by “2T”, then “1 50V”. There is leakage visible surrounding all four capacitors.

Turns out every single capacitor on the logic board was leaking. Luckily there seems to be no visible corrosion caused by the large amount of leakage. The PRAM battery also did not explode which would probably render the computer unrepairable—a possibility I wasn't even aware of until someone on Fedi replied good thing it's not battery bombed! Also luckily, the power supply/analog board seems to be in perfect condition still. While I probably will recap the analog board Someday™ to be safe, it's not mandatory to get the machine functioning.

While it'll be a pain to replace every single capacitor, with them being SMD and rather jam-packed in there, it is the best outcome possible because if it wasn't the capacitors I'd be SOL; either knowing what the issue is but being unable to repair it without a whole second machine to poach parts from, or not even knowing what the issue is in the first place.

Planning the Repair

Luckily it turns out that only three variants of capacitors were used on the logic board, so I just took a picture of the labelling and tallied how many of each I needed. But then I realized I have no clue what the labelling on the caps means. Using my sheer mastery of electrical engineering I figured that the second line of both was the capacitance in microfarads (μF) given the size and general labelling conventions, and naturally the third line meant 16/50 volts. But the rectangle symbols and alphanumeric codes were a puzzle to me, and I was worried that it meant some sort of special kind of cap or something. I found parts on Mouser with the exact same labelling other than the alphanumeric code, and the datasheet and pictures proved they were externally exactly identical so I assumed the code was mostly meaningless for my purposes; and this was confirmed after asking on Fedi and finding this page:

A Capacitor Wiki(!) page about the Mac Classic II

I ended up ordering fifteen of these (three more than needed as spares)

five of these (two spares)

and four of these (two spares)

I also tossed in a replacement PRAM battery and was amused that thirty years later, it was made by the same manufacturer with the same part number, the same specifications, and the same shrink wrap label.

Since it unfortunately didn't come with the original keyboard or mouse, I also picked up the ADB↔USB Wombat that lets you use modern USB peripherals with the Apple Desktop Bus port on old Macs. Funnily enough, it also lets you go the other way and use ADB peripherals on USB systems, so maybe if I ever buy a real ADB keyboard & mouse I could be one of those Model M weirdos who uses old keyboards with modern computers ;)

The Wombat is a bidirectional ADB-to-USB and USB-to-ADB converter for keyboards and mice.

Turns out that I don't have any spare keyboards or mice kicking around so I also picked up a Logitech MK120, the best cheap keyboard and mouse IMO; Logitech makes some good cheap stuff. Also a USB hub because the Wombat only has one USB port and expects you to use a hub.

So that's where the repair is at now, just waiting for parts to arrive. Mouser is very fast to ship so luckily I think the parts will arrive first, followed by the keyboard & mouse, and I expect the Wombat will doddle since it's not made or shipped by Professional BigCorp. Not that I'm expecting anything to arrive soon since it's labor day weekend. It'd be lovely if the caps would arrive by noon Wednesday since I have some free time and could get started then…

I'll post another update once I get back to it. More for myself than anyone else, here's a list of what I need to do: