Jul 132012
 

With KansasFest 2012 only a few days away by now, I started thinking seriously about whether I would try to bring a machine with slots along with me. I don’t want to travel with anything particularly rare, but I thought it might be worth it to bring a ][+ into working order for that purpose.

Easier said than done.

I have several ][+es as this point. The Bell & Howell ][+ and A2S2-10087 are both right out, they're not traveling. But the ][+ we bought in 1982, or the ][+ I got in its box, or even the 8050 "terrarium" seemed like possible candidates. Except I know the terrarium has some kind of RAM problem that would require some diagnosis, so I turned to the other two, which have for a while been kind of disassembled in my lab space.

The first step I took was to re-assemble the ][+ I grew up with. There are a couple of known problems with the motherboard. First, when I got it, it was partly covered in a stubborn bunch of goop, which I eventually managed to get out via a shower. In the process, I'd damaged one of the 74LS194 chips in the lower left corner—it had gotten trapped in the goop and lost a pin when I extracted it. The pin was actually still there, in the lower right hole of the socket, though I'd kind of forgotten about that.

iiplus-74LS194-socket

So, I knew I’d need to replace that. Also, the character generator ROM had been pretty well covered in goop as well, but I have a spare that I could put in its place. I took it out to look at it, and it didn’t seem as bad as I remembered, though, so as a first step, I just replaced the 74LS194 by taking it from the II+ with the box (similar era), and put a spare 4116 RAM chip in the place where the 16K language card would plug in. (As alluded to above, I actually forgot that the broken pin was still in the socket, but it didn’t seem to affect anything, and actually it also seems to be gone now, last I checked, so I guess it came out on its own.) Another problem that I am a bit more concerned about is that one of the RAM sockets came up a bit when I was cleaning the motherboard some time ago, and though it felt like I got it back down again, I may well have severed some important connections. What kills me about this is that I didn’t document which one that was! Don’t do that. It means that if there is a problem related to this, I’m going to have to check each socket with a voltmeter or something to find out where the connection is broken.

I crossed my fingers, though, plugged in the keyboard, and powered it on.

homeiiplus-d40e

Two problems here. First, it crashes into the monitor, so something’s not right with the world. Second, the keyboard doesn’t do anything. I don’t actually remember whether the light came on, but certainly none of the keys worked.

Time not really being on my side, I moved to plan B. I tried plugging in the lid and keyboard from the box-II+, and the keyboard did respond. But it was dirty. Also, I didn’t have any standoffs in the box-II+ pan, because I’d stolen them to re-assemble the terrarium. Otherwise, I’d have taken the spare II+ board and just mounted it in the box-II+ pan (since there was known to be a problem with the box-II+ motherboard anyway). But without standoffs, I had to do this with my childhood II+’s pan instead. The standoffs from the childhood II+ were not transferrable, they are kind of square whereas the holes for the standoffs in the other pan were round. So, I mounted the spare II+ board into the box-II+ pan, and plugged the box-II+ keyboard into it, which, though it was dirty, seemed to work.

I start to see how motherboards, pans, and case tops can kind of get separated. I’m trying to document with little post-it notes (and here) everyplace where I’ve made a substitution, so that I might someday be able to reassemble them as they were.

One bit of troubleshooting I tried was to swap the keyboard encoder boards, too, to see if I could use the childhood II+’s keyboard with the box II+’s encoder board. That didn’t seem to help, so I swapped them back. I’ll address the childhood II+’s keyboard/encoder later. Now I needed to clean the box II+’s keyboard.

I went all-out. Not content to simply take up the keys and clean them and the area underneath, I took the entire keyboard all the way apart. They keys came off, to be individually cleaned, and I unscrewed the PCB from the keyswitches.

iiplus-keyboard-apart

The PCB was a bit grimy. It has a plastic cover with holes in it that can also come up, so I took off the plastic and wiped down both sides with alcohol, then wiped down the PCB. All clean. Not a super complicated mechanism. The keys, when pressed, connect two pads on the PCB, through the window in the plastic.

iiplus-keyboard-mech

Reassembling the thing turned out to be harder than I’d anticipated, though. Between the holes for the screws, there are little plastic pegs by the keyswitches that need to fit into the holes in the PCB. It took a few tries, and I actually kind of bent one of the keyswitch connectors because I hadn’t noticed that it had jumped off the peg, and so the peg bent the metal because it wasn’t aligned with the hole. It took some trying to unbend it and get it back over the hole, and maybe I succeeded ok, and maybe I didn’t. It’s also difficult to put these keyboards back together, because you can’t put much pressure on the key switch (on the would-be top of the keyboard) as you reassemble it, or else the keyswitch will jump out a bit and become kind of “permanently pressed.” So I re-assembled this keyboard probably about four times before I basically got everything lined up.

And then I plugged it in, and it worked much less well than before. Most of the second row of keys (ASDF etc.) didn’t do anything when pressed, and also some of the keys in the vicinity of “X” were not responding either.

It’s a simple mechanism, but it’s a fiddly, fiddly device. I don’t know if I broke anything seriously, I suspect it’s just that something in there is not aligned properly, so some of the keyswitches are not completing the connection they need to when pressed. But if I’m going to get this to work, I’m going to have to disasemble and reassemble yet again. I have yet to try this.

Up to this point, I’d been kind of a snob about keeping original pieces together, looking upon eBay auctions with disapproval where the pan and the motherboard don’t match. But I’m really starting to see how that can happen very easily in the attempts to just get a machine working. If the motherboard doesn’t work, just set it aside for later, swap in a known-working one for now.

So, this was not a particularly successful attempt. I still don’t have a combination made with these machines that resulted in a working one. The closest is the spare motherboard in the pan of my childhood II+, with the box-II+’s top and keyboard attached, but the keyboard does’t work and I haven’t gotten further in the troubleshooting than that. I think at this point, I’ve pretty much decided that I won’t be bringing a II+ with me to KFest, there’s not really sufficient time to get one working (and it would be complicated to bring anyway).

Jul 122012
 

One of the things I’d had on my list for a while is to clean up all of my floppy drives. I have a lot of floppy drives. I have a lot of the old style (A2M0003) Disk ][ drives.

A2M0003s

Those are pretty straightforward to take apart and clean. They're actually quite similar inside to the Apple /// drives, though perhaps even a little bit simpler to deal with.

I also have 3-4 of the newer A9M0107 drives. I intended to clean those up today, but instead I cleaned one. They're really much more of a challenge to work with than the A2M0003s.

A9M0107

A9M0107-under

(Update: The smarter thing to have done here would have been to watch Tony Diaz's 2010 KansasFest presentation on disk drive maintenance before having undertaken this. Turns out, for example, that you don't need to take the bottom [or front] off if you’re only looking to clean the drive head. I was kind of after a total cleanup, but a couple of steps can be skipped for more frequent drive head cleaning. Of course, figuring it out as I went, at least given that I didn’t break anything in the process, was instructive too.)

Unfortunately, I didn’t take a lot of photos as I was doing this, but it’s quite difficult to get at the drive head. The first thing you do is remove the four screws from the underside, visible in the picture above. Then you need to get the top, bottom, and front plastic off. You can kind of angle out the bottom from the front, once you’ve disengaged the power cord, but it requires a little bit more force than one would like to exert. Getting the top casing off requires sliding it horizontally a bit, because it has little tabs that hold it on (it can slide once the screws from the back have been removed, inside the recesses on each side).

A9M0107-back

Once the plastic casing is off, you are confronted by a metal shield. A serious metal shield. I don’t have a picture of it on, but you can see it in the background a couple of pictures down from here. You have to remove the ground wire screw to free it. The shield is on really tightly, I actually had to get out pliers to pull it off. It is not connected by anything but friction (though note that it has a couple of little hooks on the front end that go into slots, so it has to be slid forward a bit before it can be lifted off). With the shield removed, you can see the analog board.

A9M0107-analog

Two further screws to remove, and then you really need to disconnect all three of the disconnectable wires, because they are strung so tightly that you won’t really be able to get the analog board out without doing that. Once the analog board is out, you are confronted by a little cardboard cover. In the background you can see the metal shield from a couple of steps ago.

A9M0107-cardboard

The cardboard cover lifts out, after which you need to disconnect the other grounding wire and remove the metal shield that was under the cardboard cover. Almost there. With these removed, you can finally see, and clean, the drive head.

A9M0107-head

Let’s hope it was dirty, so all of this was worth it. Reassembling is no picnic either, but it’s pretty much all of that in reverse. Just remember all the steps. It took quite a lot of convincing to get the power cord and its associated donut to sit close enough in to allow me to get the bottom plastic case back on, but eventually I managed it. I also somehow managed to yank out the drive light cable once everything was fully assembled. Fortunately, it was possible to squeeze it in the very small space and guide it toward its pins with a screwdriver, so I did get it re-secured, but it was an anxious minute as I contemplated the possibility that I’d have to get it all the way back down to the analog board again.

I’ll clean the rest of these some other day.

Jul 102012
 

I do not intend to get into video game collecting, but the Atari 2600 was a significant enough machine that it seems like I can make an exception for that one. I saw one going by on eBay at one point, and my interest was piqued enough to look into it. And today, my specimen arrived, which is actually a pretty nice one.

There were a lot of Atari 2600s, and they vary in their collectibility and rarity. The trick is that people who have them and sell them on eBay rarely know what they’ve got. Some people know that there are different models, but even still many people seem to get it wrong. Although I can’t guarantee I haven’t missed something somewhere, I think I basically have it now.

The original run of Atari 2600s lasted a year, and they were produced in 1977 (and I believe a little bit into 1978; they were introduced in October 1977 and the first model was only produced for a year). These first ones were manufactured in Sunnyvale, CA, and featured six switches on the top. After the first year, the case was redesigned a little bit, reducing the weight of the bottom of the case, and most of the manufacturing was moved to Hong Kong. After that, a further redesign reduced the number of switches on the top to four, with the player difficulty switches being moved to the back of the case. So, the distinctions among the early models are between 6-switch and 4-switch varieties, and, among the 6-switch varieties, between the heavy case and the lighter case. So, the first model, produced in the first year, is known among collectors as a “heavy sixer” and the next model, with the lighter case, is known as a “light sixer”. In general, the most collectible ones are the heavy sixers, which are produced in Sunnyvale. Light sixers are generally not produced in Sunnyvale, but there do seem to be quite a number of them that were, usually some kind of promotional models (although not really all that rare). The heavy sixers are still the rarest of the models as I understand it. Sears also marketed its own branded version of the 2600, which also came in at least heavy and light sixer varieties.

Some who know that the six-switch models are more collectible, but yet not understanding the difference between the heavy and light sixers, will refer to what they have as “heavy sixers” even when they’re really “light sixers”—searching around on eBay is likely to yield several examples of that. So, first of all, if you want to get one of those first year models, do not trust what the seller says it is. Trust the pictures. They will at least tell you what kind of case it is. Unless the seller is fairly sophisticated in the ways of the Atari collector (in which case you’re likely to find that the asking price is high), you won’t be able to be sure that the internals are original, but it’s pretty easy to see the difference between the light and heavy sixers’ cases. The most telling place to look is right next to the Atari logo in the front right of the machine, where you’ll be able to see either that the side is thin and with a more angled curve (a light sixer) or thick and with a more even curve (a heavy sixer). And of course, there need to be six switches on the top.

Armed with this information, I started looking around at auctions on eBay, because it’s relatively common for people to sell them indiscriminately, so even though the heavy sixers are rare, they do appear. I was fortunate enough to find one before much time had elapsed, unmarked as being such (being sold at the same price as 2-3 other light sixer systems, without any differentiation between them). So, here is mine.

heavy-sixer-setup

Mine’s actually got a pretty low serial number as well. I don’t completely understand the serial numbering scheme, but my impression is that the “E” series was the first one, and predates the others, which can end in various letters.

atari2600-68258e

There are various points of differentiation between machines, again I don’t really know exactly what drives them. One is that quite a number of early machines have a slot in the case for “channel select” (which is intended to distinguish between channels 3 and 4 I believe, although the options are labeled “A” and “B”) but without any switch inside. Some don’t even have the hole. Mine has the hole but not the switch.

cx2600-channel-select

As usual, my first priority was to take everything apart and clean it up. There are six screws to remove from the back (four of which are kind of at an angle), and then the top comes off. Here’s the CX-2600, topless.

cx2600-topless

Affixed to the case containing the motherboard was a piece of paper with a cryptic number affixed to it. I don’t think there is a consensus about what these numbers mean. But mine says “99 3390057. #”.

cx2600-mb-tape

Elsewhere, floating around in the case, was a green sticker that says “118152″—again, I don’t think there’s a consensus on what these numbers mean, but these stickers are not unusual. And the numbers probably do mean something.

cx2600-118152

One more number I found in there was on a sticker on the RF modulator. This says “7807″—and I suspect that this might be the “build date”, in which case it was built at the end of February, 1978. Which, I might add, is a bit later than I’d expected. This is the only reference to 1978 I saw inside, although at least one of the chips we’ll see later was marked “7752″ which would indicate a production date (for the chip) at the very, very end of 1977. So, it’s highly unlikely that this machine was built in 1977, we already knew from that that it must have been assembled in 1978.

cx2600-7807

Opening up the casing to get at the motherboard, here’s the bottom of it. And yet another receipt-like piece of paper sticking out from underneath, which we’ll see in a moment. One note about the screws here: the one in the rear-right corner (when you hold it with the screws on top and the ribbon cable coming out the side furthest from you) is significantly longer than the other three, so it’s worth keeping track of which screw came out from where.

cx2600-motherboard-under

The motherboard says “Rev 8 C010433″:

cx2600-rev8-c010433

Flipping over the motherboard now, we see the receipt paper taped to one of the chips. The paper reads “9913800152.#” and it is taped to a chip that says “7747″ on it (47th week of 1977).

cx2600-motherboard

The other two major chips read “7752″ and “7748″

cx2600-77xx-chips

So what to conclude from all of this? Well, I think the best guess is that it was assembled in early 1978 (probably at the end of Feburary), and so would seem to have been relatively late in the production of the heavy sixers. Nevertheless, it does seem to have a relatively low serial number on the case. Perhaps the case and the internals don’t entirely go together, too, although they are not separated by much time.

Onward, to the cleaning. One thing that I found somewhat surprising was the fact that the painted panel behind the switches is actually a separate piece. So it is in principle possible to replace (or clean) the panel separately from the rest of the top.

cx2600-top-disassembled

The paint on mine is not perfect—actually, I’ve really almost never seen one that doesn’t have the same kind of paint wear this one does.

With everything cleaned and reassembled, I tried out a couple of cartridges. I plugged the 2600 into a set-top DVD recorder (so I could tune to channel 3), which sends its output into the Wings personality card in my Power Mac G3. Lo and behold, I was playing Pac-Man and Pitfall!. Badly.

cx2600-pacman-play

cx2600-pitfall-play

The setup I got was pretty complete. It came with two joysticks, two paddles, and quite a pile of cartridges. The cartridges I got are: Atari: Combat, Video Olympics, Warlords, Pac-Man, Missile Command, Defender, Human Cannonball, Video Pinball, Asteroids, Surround, Big Bird’s Egg Catch. Sears (Tele-Games): Adventure, Night Driver. Activision: Skiing, Fishing Derby, Pitfall!. Imagic: Atlantis. M Network: Astroblast. Sega: Buck Rogers—Planet of Zoom. Parker Bros: Return of the Jedi—Death Star Battle, Spider-Man, Q*Bert. Again, some cartridges are collectible, some cartridges are super-common, and going into this I had no idea which was which (nor did it really inform much my bidding on this machine). It is possible to get at least an idea of how common these things are from the AtariAge Atari 2600 Rarity Guide, though. For the most part, unsurprisingly, the cartridges I got are pretty common. A couple turned out to be “scarce” or “scarce+”, although they’re not in great condition.

Apparently these are “scarce” in the versions I have: Adventure (though mine has writing all over it and is missing the label from the end), Fishing Derby, Spider-Man (label fell off).

adventure-sears-cart

fishing-derby-cart

spider-man-cart

And these are “scarce+”: Atlantis, Big Bird’s Egg Catch (has something affixed to the label), Sega Buck Rogers: Planet of Zoom, Star Wars: Death Star Battle (label fell off).

imagic-atlantis-blue

big-birds-egg-catch

buck-rogers-cart

death-star-battle-cart

Still, a pretty decent acquisition, I think. I’m confident this is basically the first generation machine and I have a few somewhat hard-to-find cartridges, as well as some of the ones I actually remember playing in the distant past.

I also think that’s the last of my Atari acquisitions, but the 2600 was important enough in history to justify having this one.