As regular readers may recall, I recently attended KansasFest for the first time. In addition to being a big vat of awesomesauce, the conference left me in possession of something I’ve wanted for a very long time- an original Apple IIc Plus. The second computer I ever owned (after our family’s Apple II+) was a clone of the Apple IIc, called the Laser 128EX. The Laser was a great machine- fully compatible, lots of features the IIc didn’t have, and cost less than half as much. Yes, even in the 1980s, Apple was pricing their products at a premium level for fewer features. But then as now, Apple products had that special…. something. They were always lust-worthy. I spent thousands of hours on my Laser, and it was, on paper, better in every single way than the Apple IIc. It had three clock speeds (up to 3.6 Mhz!), internal AND external expansion slots, a numeric keypad, a parallel port, special function keys and status lights, and lots of other things that the IIc didn’t have. It was a terrific machine, and buying it was a perfectly rational choice. Despite all that, some part of me, deep in the lizard brain somewhere, always wanted a IIc. Well, thirty years later, now I gots me one. Boom.
Smuggled home from Kansas City in my luggage, wrapped up in every pillow, sock, and towel I could find, the new member of the family has been waiting for some love ever since.
I’ve been working on a software project for 8-bit Apple II machines, and it’s at the point where I need to test on real hardware. Just like modern devices, developing in a emulator is fine to a point, but sooner or later you have to test on real steel. It’s time to get the IIc Plus fighting fit.
The Apple IIc Plus was released in 1988, four years after the original Apple IIc. Yes, 1984 was also the year some other computer was released, but whatever. It was the year of the IIc as far as we’re concerned. The point is that this machine is 26 years old, well past its intended service life. I’m personally not a fan of “just plug it in and see what happens”. If a machine has been sitting in a basement for decades, or bought off eBay in an unknown state, there could be all manner of things wrong that would make it bad idea to apply power. In my case, the machine came from a collector who took good care of it, but it’s a good exercise to go through the machine anyway, so let’s do it.
The Apple IIc looks very much like a modern piece of Apple equipment. There’s a reason- it was the introduction of their so-called “Snow White design language”, which would define the look of the company and their products for a long time to come. Apple II computers always looked nice relative to most of their contemporaries, but the IIc was a giant leap into the future of computers as lifestyle products. It’s a lovely machine, and was totally mind blowing in 1984. It’s still a beautiful machine to look at today, which says a lot about Snow White. It’s not easy to design things that hold their aesthetic value for more than a couple of years.
What’s especially lucky about this example is that there is minimal yellowing. Computers from the 1980s have a nearly universal problem whereby the plastic they were made with changes color over time. It’s especially noticeable on lighter colors, so old Apples are hit the hardest by this effect. It’s believed to be caused by a fire-retrardant chemical in the plastic interacting with UV light. The more daylight the machine was exposed to in its lifetime, the worse the yellowing. There is a chemical process called Retr0brighting which can be used to reverse the effect temporarily. Done properly, it seems to return the machines to like-new appearance for a few years. After that, the yellowing process resumes. Presumably the Retr0brighting can be done again at that point, although the machines aren’t yet old enough for anyone to know for sure.
Now let’s take a look from behind.
The IIc was a turning point for Apple in a lot of ways. Until now, the back of an Apple II was just a bunch of blanking plates for use by aftermarket ports connected to expansion cards. The machine made no attempt to guess what you might want to use it for- it was a blank slate. Now, all of a sudden, the IIc has replaced all that with a hard-coded set of ports. Internally, the machine still views these as expansion cards, but you can’t change them. These port decisions weren’t made in a vacuum, mind you. By 1984, Apple II users had settled into patterns of common usage. Everyone pretty much had a similar set of expansion cards, and everyone had informally decided on which slots those cards should live in. You probably had a printer card in slot 1, a modem or serial card in slot 2, an 80-column card in slot 3, extra floppy drives in slot 5, and main floppy drives in slot 6. Slots 4 and 7 were a little more up in the air. If you had a mouse or sound card, it probably went in slot 4. If you had a RAMDisk or a hard drive controller, it probably went in slot 7. If you had anything weird or niche (and there were hundreds of weird and niche Apple II cards), it went in any slot you weren’t using.
Apple codified these informal standards into the IIc, and for the most part it worked great. Sure, the loss of the total freedom to expand and hack was a deal-breaker for many, but the arrangement in the IIc was simpler, and worked the best most of the time for most people. That’s still the mantra of the company today. Apple products can be maddening if you’re trying to push outside the boundary of how they are expected to be used, but within that boundary they work better than most other products. That’s the compromise they introduced with the IIc. They’re not competing on feature count; they’re competing on overall user experience.
So, looking at the back of our machine, left to right, we have:
One last note on ports- there’s a slot above the game port for a Kensington lock. The IIc Plus was the start of Apple’s long relationship with Kensington locks- a de facto standard they continue to support to this day. This was a real boon to education and public institutions. The ability to lock a machine down to a desk in a non-destructive way was quite valuable in a lot of situations.
Wow, I really did not intend to blather that long about ports. There’s something broken in me, I think. Let’s get back to pictures, because reading is for chumps.
Here we get to the most controversial decision on an already controversial machine. The built-in floppy drive. The original IIc had a very nice internal 5.25″ floppy drive, and it made perfect sense. Fast forward to 1988, and Apple has decided the winds are blowing in the direction of 3.5″ floppies. That’s certainly true in a general sense. All 16-bit computers of the period were fully committed to the 3.5″ floppy, including Apple’s own Macintosh and IIgs machines. The Amiga, and Atari ST had also gone this route from the start. So why not the IIc Plus? Well, the 8-bit Apple II line already had a HUGE software library in 1988, all of which was on 5.25″ floppies. Furthermore, software publishers were now being asked to release software in two formats for the same machine- 5.25″ floppy and 3.5″ floppy. In theory, developers could package multi-disk software on to one disk for the IIc Plus. However, that extra storage was unlikely to be leveraged, because it would mean a lot of extra work by developers for a machine very very few people owned. It was an aggressive choice by Apple, but likely the wrong one in hindsight. A few software packages were released in IIc Plus-friendly format, but not many.
However, what was bad then is nice now. Nowadays, blank 3.5″ floppies are much easier to get. They are still being made new in some corners of the world, and you can buy them in stores. 5.25″ floppies, on the other hand, are ancient history. They are difficult to find new, and the ones that exist are so old that they are often moldy (yes, floppies grow mold and it’s nasty) and/or the magnetic medium is failing. Furthermore, the IIc Plus’s drive is quite nice. It’s a double-sided, variable-geometry 800k drive with power eject. That level of polish is very “Apple”, and it’s pleasant to use even today. This drive is the excitingly-named “Apple 3.5”, which is a simplified version of the “Unidisk 3.5”. The Unidisk had an onboard 6502 and RAM for buffering, which allowed slower machines (like an 8-bit Apple II) to keep up with the high data rate (2:1 interleaving!) of the 3.5″ drive. The “Apple 3.5” version has no onboard management, and relies on a faster system CPU to handle the data. This was fine for the Apple IIgs and Macintosh, but was a problem for the IIc Plus. Thus, the IIc Plus includes a custom ASIC and SRAM cache to manage this drive.
PC users in the audience might be wondering why these drives are 800k instead of 720k, as they are used to. Apple used variable-speed motors to cram more data onto the disk. Sectors towards the inside of the disk are physically smaller, so these drives can slow down the disk to fit more data into them, thus squeezing an extra 80k onto the disk. Apple has a long history of pushing the limits like this. In any case, 800k is a huge amount of storage for an 8-bit Apple II. It’s almost like having six 5.25″ floppy drives connected all the time. Losing the ability to run 5.25″ software is much less of a problem today than it was back then, because much of what we do is through emulated disk images or cracked software anyway. Unless you’re a hardcore user who wants to boot original software on original uncracked floppies, the 3.5″ drive is more upside than not, in my opinion. That said, you can always connect an external 5.25″ floppy drive as well.
Interestingly, Apple did think through the consequences of this decision somewhat. The internal drive is mapped to slot 5, which is unexpected. The original IIc’s drive is mapped to slot 6, which is the typical location for the default floppy drive. This was a clever choice, though, because the Apple II Autostart ROM scans the slots from #7 downward when looking for a disk to boot. By mapping the external floppy drive to slot 6, the IIc Plus has the rather unique ability to boot from the external drive if one is connected. That means you can still boot dedicated 5.25″ Apple II software which would not run from a 3.5″ floppy. There were warehouses full of said software in common use at the time, so this simple “hack” to maintain some measure of compatibility was a clever solution.
Less talk. More pictures.
Now we can see the keyboard, the power supply, and the floppy drive. A pretty simple beast, but this was the height of miniaturization in 1988. It’s also worth noting how nice this keyboard is. Apple has always made great keyboards (although I don’t love their latest ones with flat-top keys). The IIc keyboard looks great, and is wonderful to type on.
Steve Wozniak was particularly adamant about socketing chips as much as possible, because he was a hacker at heart. He knew some people would want to hack on the machines, because that’s what he would do. If chips are socketed, it opens up many doors for modifying and tinkering with basic tools. Nowadays, in a world of nearly-microscopic-pitch SMT circuit boards, hacking is much more challenging.
Woz has been quoted as saying the IIc was his favorite Apple II, despite the more closed nature of it. I like to think his influence is still visible here. There are three socketed chips- the 65C02 CPU, the memory management unit, and the monitor ROM. If you had to pick only three chips to socket for hackability, those are decent choices. There’s also a “memory expansion interface” (which in fact gives pretty decent general access to the bus), an “internal modem connector” (which is a copy of the slot 2 serial port), and an extra header marked “J14”. That last one is unique to the IIc Plus- it contains a few extra signals normally available to expansion slots in a IIe. The idea was that, by making more general signals available near the memory expansion slot, more general expansion options might be possible. The IIc Plus didn’t last long enough in the market for us to find out. I intend to explore this interface at a later date, because I think the possibilities are vast.
Someone has made a couple of terrific high-res scans of the IIc Plus motherboard, so I’ll link to those here. There’s a front half, and a back half. Take a look if you want to see some of Apple’s better engineering in the late 1980s. Also, here’s a great reference for all the custom ASICs in all Apple II computers.
This is a good time to mention the IIc Plus’ signature feature- the hardware accelerator. During the glory days of the Apple II, there were a few companies making accelerators for the CPU. This is much more complex a challenge than simply overclocking like modern PC hackers do. The Apple II architecture is a delicate dance of hard-coded timing requirements and all the systems are tightly coupled. Things as diverse as the expansion cards, video generator, and floppy drives all depend intimately on the CPU running an exact speed. It’s a side effect of Woz’s passion for eliminating glue logic and making chips do multiple jobs to simplify the design. To accelerate these machines, you have to run the CPU faster while not disturbing everything else. Systems such as video and floppy controllers still need to think the machine is running at 1Mhz. As such, hardware accelerators were always complex affairs, generally involving an entire expansion board with a special ribbon cable that replaced the CPU (which, thanks to Woz, was always socketed). The two heavy-hitters in the accelerator market were Applied Engineering with their Transwarp products, and Zip Technologies with their Zip Chip. Apple liked the Zip Chip so much that they actually licensed it for the IIc Plus. Thus, the IIc Plus has the option to run at 1Mhz or 4Mhz, with little or no loss of software compatibility. That was a real boon when using productivity software and such. For games, it wasn’t much use, because many games used simple cycle-counting for timing animations and such. They often became unplayable if the machine ran faster. It’s worth noting that the Laser 128EX had an accelerator long before the IIc Plus came out, but it was less compatible. I can tell you from much experience that the high-speed setting was mostly useless, as software became very unstable. It had a lower 2Mhz setting that worked quite well, though. I ran most software in that mode.
In any case, many of the chips you see (including the big QFN ASIC) are related to this. There’s a 16k cache, and various special controller chips that make up the accelerator system. Roughly a quarter of the logic board, in fact, is devoted to this!
Let’s move on to the power supply. Aside from satisfying my need to take things apart, the point of this exercise is primarily to evaluate the health of the power supply. This is one of the most common failure points of retrocomputers, and when they go, they have an unpleasant tendency to take other elements down with them. A simple visual inspection will suffice. We want to make sure the capacitors aren’t leaking, and there’s no sign of shorts or burning anywhere. Signs of excessive heat could be indicative of failing components that are working harder than they should. We can also check the internal fuse while we’re here.
The power supply is very elegant design. It has two clips holding it in place, and it literally lifts straight out, with no fasteners involved. The high-voltage connector is concealed under the shielded box, making it a safe machine to work with the top removed. Nice!
The top of the power supply is easily opened by removing two screws.
As you as can see, this power supply is in incredibly good shape. This suggests either the machine didn’t see much use, or perhaps it was replaced at some point. It’s also possible the capacitors were replaced by a previous owner. There’s one oddity- the capacitor in the back right corner looks like one that wasn’t made very well from the factory. The plastic wrapping doesn’t extend very far above the top, as they usually do. This is mildly concerning, but the cap appears in perfect shape, so I’m going to err on the side of leaving things alone. With old electronics, there’s always a chance of creating a problem by touching things, so I think it’s best to leave well enough alone here. I’ll keep an eye on it, and if I ever have the supply out for another reason, I’ll replace that cap for peace of mind.
This supply looks like quite a nice design (thanks, Sony!). There’s good airflow, and any component that dissipates a lot of heat is elevated above the board. For example, all the power resistors have at least a 1/2″ of air space underneath. This is a nice trick that I leveraged myself when rebuilding one of Johnny’s power supplies. The IIc Plus has no fan, so this kind of attention paid to passive cooling is good to see.
The next major system that may need some love is the floppy drive. Floppy drives are complex mechanical devices that need constant care to keep running, so it’s a safe bet that any drive you find at a garage sale will have some issues.
As expected, it’s dusty and a little cruddy. Floppy drives are the perfect storm for attracting crud- they are exposed to the outside world, they have greased surfaces, and foreign objects are regularly inserted into them with extreme prejudice. They’re always dirty. This drive appears to be in good condition though, so cleaning things up with some Q-tips and fishing stray hairs out of the mechanisms is all this one seems to need.
We should also take this opportunity to clean the heads. Even when the drive is used with the utmost care, the heads will get dirty. That’s normal function.
This is a double-sided 800k drive, so it has two heads. Most older Apple IIs have single-sided drives. In that case, the head is on the bottom, and the top is a pressure pad. Don’t try to “clean” the pad, you’ll just make a mess. The head can be cleaned with some isopropyl alcohol on a Q-Tip. The heads on this drive are quite clean, once again suggesting the previous owner took good care of it (aside from the Coke spilled in the air vents, that is).
This drive is another interesting cross-roads for the industry. You can see the transition to modern manufacturing in a few places.
I’ve barely touched on the dark art of vintage floppy drive maintenance. I could easily fill an entire blog post on this subject. Instead, for a more complete treatise, I direct you to Tony Diaz’s session from KansasFest 2010.
Now, as we like to say in the automotive world, assembly is the reverse of removal. Put everything back together, and it’s time for the moment of truth. Does it power up, with no pops or magic smoke?
Huzzah! A bit anticlimactic, I confess, but there are some things to note about this boot:
So, that’s good progress. We know the machine powers up, and boots far enough that the ROM has scanned the drive and recognized the absence of a disk. That means it’s running code, and thus the machine is completely “up”. Now what? Well, obviously, we need a display. However, the IIc Plus outputs composite video, and there is literally nothing in my house that can render composite video. I have three LCD monitors, an LCD TV, and a half-dozen other display devices, but not a single one has a composite video input. They just don’t put that on things anymore. This is the real crux of the challenge of retrocomputing- the machines themselves are still as functional and useful as ever. However, all the support devices (storage, displays, joysticks, keyboards, communications) have moved on and left them behind.
Next time, we’ll get in to solving this problem! We may even have to warm up the soldering iron (shocking!).