Digging up treasures in mom’s basement
Over the holidays, I took the rare opportunity to dig through the stockpile of childhood belongings in my parents’ basement. Thanks as always, mom and dad, for judiciously moving that stuff from house to house and storing it carefully all these years. I had two singular preservation goals- My Apple IIgs motherboard, and my 5¼” floppy disks.
The IIgs motherboard was a concern because it’s the one and only Apple II model with a realtime clock, which means it has a battery. As regular readers will know from my extensive berating on the subject, batteries get old and leak. When they do, they take everything in a 6″ radius down with them. Many a retrocomputer has been irreparably destroyed by battery neglect. In the case of the IIgs, the battery is soldered to the motherboard, so you need to cut it off. Of course, this means you lose whatever 25-year-old NVRAM settings you had, but that’s a small price to pay to avoid PCB Cancer. There are only two kinds of motherboard batteries- those that are leaking and destroying the machine, and those that are imminently leaking and destroying the machine. As it turns out, mine was in the latter category, and in fact the motherboard looked brand new. Two quick snips, and now I know it will stay that way until I can get around to resurrecting the machine properly.
With that out the way, it was time to press on to the more interesting of the two missions. I had set out to rescue my 5¼” floppy disk collection. The rated life on these disks was only about 20 years, so the clock is very definitely ticking on all that remain. If they were stored in a humid environment, they likely have mildew on them at the very least. More on that later. At worst, the magnetic medium is literally flaking off the mylar base, and whatever treasured bits were stored therein are lost forever. Floppies are a dying medium, and a great deal of interesting human history does not exist in any other form. Nowadays, our precious data goes through a sort of constant migration. You move it from hard drive to hard drive as you upgrade computers, and increasingly it finds its way on to the internet, where magic Cloud pixies presumably back it up and propagate it ad infinitum. Floppies were before all this. This data was occasionally backed up… on to other floppies. For the most part however, it’s all locked in a moment in time. If we don’t make an explicit effort to physically carry it over the air gap to the internet, an entire generation of digital human history will be lost forever. I wanted to do my little part in this effort, and indulge in some nostalgia along the way. Join me, won’t you?
The first challenge was getting them home. I needed not only the disks themselves, but a floppy drive with which to read them. My current Apple II setup (unusually enough) does not currently include a 5¼” floppy drive.
Now that everything has made it home, the next step is to find a way to read the disks. Without that, my “preservation” effort will consist primarily of admiring them on the shelf until I’m carted away by the authorities under suspicion of severe emotional distress. “Bury them with meeeeee,” I would cry as they tightened the straps on the gurney. A dark scenario, I concede, but don’t think it can’t happen to you.
Where were we- oh yes, the floppy drive. Here’s the fine specimen that has survived 30 years in captivity:
Then as now, Apple products were expensive and sought after. Those of us who couldn’t afford the real fruit bought the next best thing- V-Tech’s “Laser” brand of Apple II compatibles. In fact, Lasers were superior in many respects, and thanks to learning from the Franklin company’s legal trouble with their Apple II clones, V-Tech’s products were clean-room reverse-engineered and thus immune to Apple’s formidable ranks of lawyers. This particular drive was purchased to supplement my Laser 128EX, which was a clone of the Apple IIc. When I sold the Laser to finance my upgrade to a second-hand Apple IIgs, I kept this drive. The IIgs that I was buying (from some creepy guy I met on a BBS, no less) did not come with any 5¼” drives, so I needed a way to keep access to my existing collection. I have a couple of original full-height Apple Disk II drives somewhere, but they are buried with my original Apple II+. That’s an excavation for another day.
This drive should work just fine with my shiny new-to-me Apple IIc Plus, but I want to give it a checkup first. These disks I’m trying to preserve are on the hairy edge of failure as it is, so it behooves me to make sure the drive I read them with is in tip-top shape first. Let’s crack this baby open!
Interesting foot note (hurr hurr) right off the bat- notice the feet. Apple products used a fancy deluxe high-tech rubber of some sort that looked and felt very nice, but over time degrades into a horrible goo. V-Tech used… cork. Still as perfect as the day they were made. Cheaper, too!
I’ve literally never opened this drive in 30 years of ownership, so I’m guessing about how to get in there. When in doubt, start removing screws from the outside in. The six around the perimeter seem like good candidates.
Note that while it may look like any old PC floppy drive inside, this is a GCR (Group Code Recording) device. In the early days of spinning magnetic media, there wasn’t yet agreement on how the binary bits should be encoded into magnetic flux changes. IBM PC floppy drives (and all modern ones since) use MFM (Modified Frequency Modulation) encoding. They are two different ways to solve the problem of how to encode bits as tightly as possible, without violating the physical limits of how closely you can place magnetic particles of opposite flux. Floppy disk encoding techniques is a very deep (and largely forgotten) topic that I don’t have the space to go into here. Suffice it to say, there’s good reasons Apple II/Commodore/etc disks are fundamentally incompatible with PC floppy drives, and these reasons go deeper than just the formatting of the disks.
Speaking of depth, let’s continue downward into our humble Laser drive.
Take a look at how this upper shell is mounted. Those are brass standoffs bonded into the plastic, which accept machine screws from below. This is Engineering For Serviceability, and I love to see it. Nowadays, if there were screws at all, they would be self-tapping into plastic posts, and you’re lucky if they survive two insertion cycles. This device has been built with the expectation that it will be in service for a long time, and may need to be opened on occasion. On the one hand, this is the kind of engineering in products that I miss. On the other hand, this is inefficient, because these drives were probably all obsolete before they ever needed opening. Nowadays we manufacture seemingly with lower quality, but you can argue that we actually are acknowledging the reality of the lifetime of these products, and the end result is cheaper to buy if we do. Why build something to last fifty years if it’s only actually useful for five? If 85% of its engineered life is spent in a recycling bin, we have wasted the resources used to make it so robust. You can argue that the most efficient product dies, on average, precisely at the moment it is no longer useful, and not a moment later. In any case, this over-engineering means we can enjoy these devices today, so I’m not complaining.
The devil is in the details when evaluating quality. Notice how the return spring for the hub clamping mechanism has a rubber tube over it. This is likely to prevent it from snagging on the various moving bits below it. The drive would probably work fine for a long time without that, but this is Engineering for Longevity. With this tube, the mechanism will work a little better for a little longer. I love to see this!
The clamp appears to move freely, the spindle spins, and everything looks quite ship-shape. Let’s clean the head while we’re in here.
Let’s take a look at the circuit board while we’re here. It’s not really a “driver board”, like you would see in a PC floppy drive, or even the contemporaries of the day (Commodore, Atari, etc). These were “dumb” drives, in the sense that the Apple II’s main CPU controlled everything. And I do mean everything. The reading and writing of disks was literally done with clock-cycle-counted code running on the main CPU, which has direct control of all motor, head, and stepper activity. Steve Wozniak did this because it was cheap and very performant for the user. Apple II drives are blazing fast compared to all others of the day. This approach also created a paradise for authors of copy-protection systems, because they could do all manner of crazy stunts, from arranging bits in a spiral pattern, to inventing new sector/track geometries, to completely altering the layout and meaning of bit streams. If you want to see how deep this rabbit hole really goes, follow modern retrocracker 4am on Twitter. The cat and mouse game between Apple II crackers and the copy protection programmers was an epic battle of technical wits of the sort that we may never see again.
Anyways, back to our PCB.
This Sony chip is interesting, because it’s a fairly primitive early surface-mount device. SMT was uncommon in mass-produced hardware at this time. The pitch is quite large (looks like 50mils), and this one appears to be hand soldered. The board itself would likely have been wave-soldered from the bottom, so this chip would have to be added afterwards. I’m sure they would have preferred to have this on top to save the extra assembly step, but we’ll see below why that wasn’t done.
I have not succeeded in identifying this chip. A1073 looks like the date code (10th week of 1973, which would be about right), which leaves 7D01 as the part number. This has eluded my google-fu, but it is directly connected to the read/write head. That suggests some sort of GCR encoder/decoder, perhaps. First reader to locate the datasheet for this chip gets a cookie FedExed to them. I promise!
Unscrewing the PCB now, we get a look at the front. It’s a lovely thing.
The horizontal chips are all off-the-shelf TTL logic- as simple as the earliest 8-bit computers themselves. The vertical chip on the left is a general-purpose Mitsubishi transistor array, no doubt related to controlling the stepper and spindle motors. Perhaps it’s wired up as a pair of H-bridges. Here’s the datasheet for that, if you’re as interested in the gory details as I am. If you’re not, you’re probably reading the wrong blog. If that’s the case, how on earth did you make it this far into the article? Seriously- go outside or something. It’s a nice day out there (I assume).
Anyway, we have now answered the mystery of why they bothered to hand-solder an SMD on the back of the board. Routing space on top is tight (for 1980) here, and the cost-savings of a single-layer PCB must have outweighed the additional labor.
We’ve found our first authentic artifact. Under the PCB there’s a little gold… something. Recognize it? If you’re a genuine computer user from before 1990, I bet you do.
For anyone who may not be from this generation, these stickers were used to cover a notch cut in the upper right and/or upper left of the floppy disk. These notches told the drive whether it was allowed to write to the disk or not. In theory you could cover the notch with this sticker and protect the data on it. In practice, floppy disks are made of some secret government nuclear plastic that nothing sticks to. Disk drives served two purposes- accessing floppy disks, and internally storing write protect stickers. It’s remarkable this drive has somehow only accumulated one. Besides, those write protect stickers were remarkably ineffective against the usual causes of data loss- vacuum cleaners, televisions, dogs, and angry older sisters.
Okay, we’ve inspected what we can inspect, cleaned what we can clean, and told some young people to get off our lawn along the way. Installation is the reverse of removal, as they say. One more thing to do before testing on real disks. We should verify the speed. The easiest way to do that is with a venerable Apple II utility called Copy II Plus. This uberprogram is a swiss army knife of things an Apple II user might need to do related to disks, including making Totally Legal Backups For Safety And Personal Use of many commercial software programs. To use Copy II Plus, we’re going to need a copy of it (the irony of that is not lost on me). It’s easily located as a disk image on the Asimov archive, in 800k 3.5″ form suitable for my Apple IIc Plus.
Using ADT Pro as formerly described on this blog, we can turn this disk image file into a real bootable 3.5″ floppy disk.
Using Copy II Plus’ Verify Drive Speed option, we can check that our drive is within spec.
It’s good enough for now, but adjusting the speed to be closer to exact would be a good idea. Any marginal disks are more likely to read if our drive is right in the middle of the spec. However, I was unable to locate an adjustment anywhere in this drive. Believe me, I looked! I’d never heard of a floppy drive with no speed adjustment, but maybe this is an area where V-Tech cut some corners. Honestly, the speed adjustment is mainly needed for the older full-height drives which had belt-driven spindles. The belts would stretch over time, causing the drive to gradually slow down. This was really nefarious, because a slow drive could still read its own disks, but those disks would not be readable by an in-spec drive. So when you finally noticed and fixed the drive, some arbitrary set of your old files would no longer be readable. Newer devices like this one are direct-drive, so there’s no belt to slip or stretch. Much better. I’ve never actually seen a direct-drive device that needs speed adjustment, but it would be nice to be able to tweak this one to be perfect. Oh well.
Okay, our drive gets an official Blondihacks Clean (Enough) Bill of Health. What disk should we try in it? Well, there are 240 aforementioned options to choose from, but let’s grab the one off the front of the stack. As you can see, this is a lovely original- not a Totally Legal Backup For Safety And Personal Use, which was much more the norm.
I don’t actually remember this game. I have very fond memories of the original Infiltrator, which was a combination helicopter flight sim and 3rd-person stealth game. I had totally forgotten there was a sequel, never mind that I evidently owned it! Okay, Infiltrator II, let’s see what you got…
For the full Apple II booting effect, let’s try one more, this time in video form. The disk we’ll try is next in the stack- a Totally Legal Backup For Safety And Personal Use of Gemstone Warrior. This is a terrific game! Lots of depth, great animation, and very polished start to finish.
You know what the best part of this test was?
Now that’s some robust technology right there. That handful of magnetic bits survived 30 years in my mom’s basement, ready to come back and remind me how much I liked this game, and how I’ll probably never be as good as Trouba and Peter, whoever they are. But hey, at least I’m better than NIGhTShADE. Suck it, NIGhTShADE. I kicked your ass at Gemstone Warrior when I was 10.
There’s so much more of this kind of digital archeology stuff to come- I dug up some real comedy gold that you won’t want to miss. Stay tuned next time when we get to really dig in to that floppy pile. Don’t copy that floppy! Unless it’s a Totally Legal Backup For Safety And Personal Use.