Summer, 2014. I was at a friend’s house studying for university, when I received a call from a middle school teacher, the school I attended years ago. “We’re cleaning out the old computer equipment,” she said, “Is there anything that can be useful to you?”. I was interested to know if there was anything useful that could be donated to the local LUG (Linux User Group), so I asked for more information. “I’m not sure, but I see an Apple label here; I can’t say anything more.”, she replied.

Indeed, that computer was an Apple. But it wouldn’t be of much use to the LUG, since it was a computer from 1982. It was an Apple IIe, a bit dirty, together with the iconic green phosphor screen and one of those 5.25" floppy disk drives. “We have to throw them away to make room for another class,” she said, and apparently, there wasn’t even space left at the Interactive Museum of Computing, which was located in that school at the time. I gladly accepted, a piece of history like that cannot be sent to waste.

Once at home, I tried to connect everything. Despite the dirt, the scratches and the old age marks, it was working again. But the hitches soon came: first, one of the power supply capacitors, from the EMI filter, blew up with white smoke. This is a common issue that can be easily repaired. Then came the tragic discovery: the RAM were buggy. We discovered this by writing some simple BASIC programs; after a certain number of lines, the code was self-corrupting or was giving unexpected and incorrect results. Last, but not least, the floppy drive was missing a piece, a plastic part that squeezes the disk against the rotor to make it spin.

However, it still was a beautiful computer to look at. We carried it to an exhibition at the former Tesla bar in Montelupo Fiorentino, along with other vintage computers from other friends’ collections.

After that event, we put away the Apple II in our RetrOfficina, because none of us want to fix this series of faults, with the possible risk of further damage. Until last summer.

Memory Repair

With a lot of patience, in my spare time, I had started cannibalizing the RAM from the board of an Amstrad CPC464 that I had given up on (but that will be another story involving my friend giomba). That’s when I decided to do the same thing with the Apple II. Of course after confirming that there was indeed a memory problem.

In fact, I had never tried to deeply investigate the memory fault. I had assumed that testing the memory would require at least a program to be loaded via floppy, but I was wrong. The system ROM includes a self-test¹ software, including a routine that checks the memory by writing and reading consecutive patterns. To run the test, you simply have to turn on the computer, then press [Ctrl] - [Apple] - [Reset] simultaneously. The routine makes the screen flash a couple of times and then ends with a “KERNEL OK” message in case of success, or with an error message indicating which chip may have failed the test. In my case, the error message was clearly pointing his finger against “RAM”.

With a lot of patience, I desoldered all the chips so that I could check them individually using a homemade Arduino memory tester. Summing up, four chip out of 8 where damaged. Of course, instead of directly soldering new RAM chips, I placed some sockets to be able to confirm the correct operation of the new memories. A very wise decision, as one of the new chips I ordered was slightly out of specification and could not pass the self-test.

Cleaning and Retrobright

Years of disuse and dust had ruined the computer’s case. While waiting for the delivery of the new RAM, I washed the case with hot water and “Marseille soap” detergent, bringing it back in good condition without the need of any additional chemical tricks.

I did the same for the keys, cleaning them one by one. The space bar was in worse condition because someone had glued stickers on it. For this single piece, I wanted to try a retrobright solution using cream hydrogen peroxide. I followed some online advices, applying the cream to the piece, wrapping it in plastic wrap, and exposing it to UV light. The result after a full sunny day was poor: the piece had partially whitened leaving spots here and there. Probably, the plastic wrap was uneven. I decided to try a second application, this time placing the piece inside a harder plastic bag. It didn’t come out perfect, but now the leopard-like spots are less visible. However, to avoid making it worse, I settled for this result.

Finally, with a slightly abrasive sponge, I removed the dirt from the metal base. Unfortunately, the plastic cap covering the expansions holes were very aged, and many of them broke when I removed them. I put them apart in a bag to repair them in a second moment.

Experiments

At this point, it was time to turn on the machine and do something with it. But the problems weren’t over yet; the broken floppy drive was still broken. Fortunately, I learned about two very interesting projects:

• Apple Game Server - a collection of games in WAV format, ready to be loaded onto the Apple II directly from AppleSoft Basic via the cassette recorder jack.
• c2t - an application for generating WAV files from binary files, to be loaded via the cassette recorder jack.

Without a working floppy drive, this machine was still quite limited, but for now, it was enough to keep it minimally active and have some fun with it. Both literally and by trying to load some assembly code listings.

Various Notes and Technical References