Apple Macintosh LC III

The Macintosh LC III is the (not surprisingly) third version of the LC computer introduced in October 1990. The original LC was part of a trio of low-cost Macintosh computers intended to challenge the growing dominance of PCs and also stand up against the Commodore Amiga and Atari ST. The low-cost trio included the Classic, the LC, and the IIsi. The Classic carried on the original Macintosh form factor, replacing the venerable Macintosh SE, and the Macintosh IIsi was a scaled-down Macintosh II. But the LC fit nicely between the two. In its small pizza-box case, it included many of the characteristics of the Macintosh II family, including color graphics, but offered limited expandability.

I attended college during the early 1990s. Our campus was Mac-based, and you would spot a Classic or an occasional 100 series PowerBook in the dormority rooms, but lucky students had LCs. Their color graphics and reasonable performance made it the computer to have. A friend of mine had an LC II, and though it was usually running Microsoft Word, it was capable of spinning up games or exploring America Online.

The LC (rumored to stand for “low cost” and codenamed Elsie) was the brainchild of two Apple engineers.  H. L. Cheug, from Singapore, and Paul Baker, who had worked on the Lisa.  Both left Apple to follow Jef Raskin’s ill-fated Information Appliance company but returned as Raskin’s project faded.  Cheug and Baker wanted to build high-quality Macs at a low cost.  Despite limited corporate support, they plugged away until competitive pressure forced Apple to consider low-cost solutions.  Cheug and Baker’s efforts resulted in the color LC that incorporated many advantages of the higher-cost Macintosh II line at a much lower price.  The LC’s construction was simple: a scaled-down logic board, narrow rectangular power supply, fan, speaker, floppy drive, and a hard drive (though a dual floppy arrangement was possible).  The layout was designed for automated construction.

The LC was targeted at the education market.  Though it offered limited expansion, it had a processor-direct-slot (PDS) that worked with an Apple IIe card, allowing schools to access the extensive Apple II software library while embracing the Macintosh platform.

While the LC was introduced with the Macintosh Classic, the LC III was introduced with the Color Classic in February 1993, making the LC III a solid bookend to the original.  And while the original LC could be outfitted with an Apple IIe card displacing the Apple II’s dominance in the education market, the LC III finished the job by coming out around the time the Apple IIGS was retired. 

By the introduction of the LC III, the LC line had become Apple’s most popular Macintosh, with industry analysts estimating Apple sold nearly one million LCs in 1992 alone.  Equipped with the base 4MB of RAM and 80MB hard drive, the LC III was priced at $1,349, $700 less than the LC II when it launched less than a year earlier.

The LC III was surprisingly robust.  While the original LC’s 68020 processor ran at 16MHz, the LC II moved to the more capable 68030, but it was limited to a 16-bit data bus, severely limiting throughput, and it lacked more sophisticated memory management features.  The LC III had no such limitations.  Its 68030 processor was clocked at 25MHz and utilized the full 32-bit bus.  This effectively doubled the speed of the computer.

MacWorld Magazine loved the LC III, claiming that there was no longer any reason to buy the Macintosh IIci or IIsi.  The LC III was as powerful at a much lower price.  Apple probably appreciated this opinion, as they had positioned the IIvx as a business-class 68030 replacement and were rolling out a growing list of Quadra systems offering the latest 68040 processor.

The LC III was one of the first Apple computers to utilize 72-pin SIMMs.  This higher-density memory module was commonly used on new PC systems of the day.  4MB of RAM was soldered on the logic board, and the single SIMM slot could accommodate 4MB, 8MB, 16MB, or 32MB modules—making the maximum allowable RAM 36MB.  512KB of VRAM was installed on the logic board, and 256KB could be added in the single VRAM expansion slot.

Unlike the earlier LCs, the LC III includes a socket for a 68882 math coprocessor, allowing it to stand toe to toe with the IIci and other higher-end 68030 systems.  The LC processor direct slot (PDS) remained; however, the slot was expanded from 96 pins to 114 pins, providing access to the processor’s 25MHz capabilities for accelerator cards.

I got this LC III for a reasonable price through an eBay auction. After I purchased it, the seller told me it had been a troubled restoration project, and it only booted occasionally. Unfortunately, it was also damaged in shipment. The post office sent an apology note with the crushed box, but the damage was done, as back corner of the LC III was smashed.

I slid this computer onto a storage shelf, where it sat until 2023 MARCHintosh came along.  I decided repairing this LC III was a good MARCHintosh project.  I started by thoroughly cleaning the outside of the case.  Apple’s Snow White design aesthetic is classic, but the plastic grooves are a chore to clean. I also needed to address the smashed back corner of the computer.  I found almost all the broken pieces in the shipping box and taped them to the computer.  So I was able to glue the puzzle pieces back together again.  I also filed down the plastic lip that makes up the top cover’s simple latch.  Like those in the Macintosh II family, the LC’s case is opened by pulling on two plastic tabs on the corners of the top lid.  I assumed my glue repair would not hold up well to the stress of the slight bending required to pop open the case, but the top cover is still held in place with the latch on the other side. 

Once I pulled out the logic board, I realized it also needed a good cleaning, and I carefully examined the damaged pads and the replaced capacitors.  The person who sold me the computer had removed the circular aluminum electrolytic capacitors and installed modern rectangular tantalum capacitors.  However, I suspect they used lead-free solder—and lots of it.  The board fit nicely in my ultrasonic cleaner, and after a ten-minute bath, I let it dry for a day or two.

I set about reworking the installed caps.  I had to add new solder to get the lead-free solder moving.  I then wicked up as much as I could.  I decided to do this to improve the look of the board, as I assumed the new caps worked fine, but after removing the solder on one of the caps, I found the leg of the prior electrolytic cap was still on the pad and had shifted under the new tantalum cap.  This explained why the capacitor wasn’t sitting flush with the board.  Since such a thing could create a short between two pads, I completely removed and reinstalled any capacitor that wasn’t positioned correctly.

With the other caps squared away, I focused on the missing pads under C13. This is a larger 100uf capacitor, but luckily, I had a replacement tantalum. The existing pad and traces had been wiped out. However, they ran to nearby vias. Given the small distance, and since the missing traces ran under the capacitor, I cut thin strips of copper tape and affixed them between the vias and the pad locations. After checking continuity on both sides of the board, I solder the cap onto the copper tape. It was helpful that the tape had adhesive on the back, keeping it from sliding around while I soldered the cap. It was tricky, but I was pleased with the result. I then added copious amounts of solder mask on top of the exposed portions of the copper tape to keep it from shorting and providing another means of affixing the tape and the attached capacitor to the board.

I then opened the power supply.  Though many capacitors are inside, they all looked good to me.  Unlike another TDK PSU I’ve used with an LC II, this one had all Nichicon capacitors. The other TDK used a mixture of Nichicon and Elan branded capacitors, and all the Elan caps had leaked badly. But this PSU was surprisingly clean, making me wonder how long this computer had been used.

When later testing the power supply, I discovered a problem. While the 5V and 12V rails were good, the -5V produced only -2.2V. I assumed I needed to recap the PSU despite its good visual condition, but then a forum thread at 68kMLA alerted me that C22 on the logic board may need to be installed opposite of the silkscreen. Several had reported the TDK PSU (and maybe others) did not provide the proper -5V power when the capacitor was installed with the positive end of the capacitor matching the positive markings on the board. This could be because the -5V rail is… well… negative, making a backward position appropriate. This problem seemed to arise when using tantalum capacitors rather than electrolytics. Poor power on the -5V rail would likely only affect serial communications, so those not using the modem or printer ports might not notice the problem. After flipping around C22, the PSU provided a solid -5.1V.

I also cleaned and lubricated the floppy drive.  It was pretty dirty, but after spraying generous amounts of isopropanol alcohol and then reapplying grease to the moving parts, the eject mechanism operated smoothly.  I opened the auto-eject gearbox, and I was surprised to find the one disastrous gear was still intact, and after poking at it, it didn’t fall apart.  I decided to leave it in place, though it could stop working at any moment.

The computer didn’t come with a keyboard or mouse, but I took this opportunity to clean up a period-correct Apple Keyboard II.  It had been stored in a barn or other unpleasant place, as it was filthy, with grass and other debris inside.  It was also a bit rusted.  After disassembly, I washed all the plastic parts and keycaps and treated the rusting metal with several anti-rust agents.  Perhaps I should have repainted the rust spots, but it looks good, and the keyboard now works.

After putting things back together, I was ready to power it up. I reattached the existing hard drive and plugged in an external BlueSCSI. I heard a cheerful chime, and the screen lit up. I was then greeted with a message telling me System 7.01 on the BlueSCSI would not work on this computer. This also told me the internal drive wasn’t working as it should have been the default boot drive. After making a new BlueSCSI image with System 7.1, I was ready to try it again. Only it still didn’t work. I then remembered several systems around this time required System Enablers. After checking, I discovered that the LC III needed System Enable 003, so I found the correct file and dropped it into the System Folder on the BlueSCSI. The internal hard drive started working during this trial-and-error with the external BlueSCSI. Perhaps it took a few start-up attempts to overcome stiction inside the drive. The internal drive had a fresh install of System 7.5 with a few games installed. Likely the work of the prior owner as he attempted his restoration.

I’m glad I showed this little LC III some love.  It is one of the great LC models.  A mere eight months after its introduction, in October 1993, Apple gave the LC III a speed boost, bumping it from 25MHz to 33MHz.  It also received a new slicker case and a slight name change to the LC III+.  However, its real replacement, introduced simultaneously with the LC III+, was the LC 475 with its 68LC040 processor.  This makes the LC III and the Color Classics the last marque Macintosh computers to utilize the 68030 processor.  Later budget-driven LCs and Performa models would still use the 68030, but those computers lacked much appeal except to starving students.

I ran the system through a full suite of Snooper and MacTest diagnostics.  It passed with flying colors.  I will keep the computer near my workbench, as its small size and convenient design make it a solid test device for drives, monitors, and other peripherals.