Restoring Vintage Computers | Hacker News Reader

Restoring Vintage Computers

curiousmarc.com

88 pointsdefied17 comments

Seeing his list of restored devices, including an HP2645, recalled to me how we hacked the original HP2640A glass teletypes to turn them into little desktop computers. 2640s came with a set of self-tests that you could run from an HP3000 minicomputer. Didn't take us kids long to figure out that the role of the HP3000 was to download a string of bytes representing 8008 code, that the terminal ran to test itself. They had 1K byte and 4k byte memory cards in the terminal. So we reverse engineered the 8008s interface to the screen and keyboard IO system, and were soon building little screen games. This was 2 years before the Apple II debuted, and when the only "home" computers you could get were big boxes like the Altair 8800. We basically had stumbled across the PC, but weren't smart or ambitious enough to do anything besides play dumb screen games with it.

I got an HP 4951b serial protocol analyzer off eBay, and want to do something similar with it. It’s based on an NSC-800 CPU, which feels like a cross between a Z80 and an i8085. My explorations are at https://github.com/InPermutation/HP4951b

How cool! What kind of games?

Most of our effort went into a solitaire version of pong, where you used arrow keys to move the "bat" along the bottom of the screen, and the "ball" ricocheted symmetrically off the top. If I recall correctly, the ball sped up over time, and we introduced some randomness in ricochet to make it a challenge you'd eventually lose. But honestly, it was a long time ago, so how much of what I remember actually worked well, I wouldn't bet.

It did teach me 8008 assembler though. Used that later build a primitive RT OS with multi-level interrupts that operated a cross-bar switch for connecting KSR 33 teletypes to different computers in the lab. That was fun. We wrote the OS in concurrent pascal pseudo-code, then hand compiled into assembler.

Very fun story, thanks.

> We wrote the OS in concurrent pascal pseudo-code, then hand compiled into assembler

I love low level programming, but I never considered that technique back in the day. Kind of brilliant.

This guy is amazing. Of course being a million millionaire help but his skills and experience are something I just dream about.

Watch the series where they repair an Alto computer, the disk packs and bring it up. Also repairing IBM 029 keypunch.

Amazing. And his colleagues are just as amazing.

His Apollo Guidance Computer restoration videos are equally amazing. They actually got the AGC to run. Incredible.

Testify! His "elevator music" explanation tangents alone are worth the price of admission. The Apollo ranging stuff was fascinating.

I have a couple of machines lurking in my collection which worked fine the last time I turned them on, but that was 15 years ago now... I really should do the research on how to make sure I don't accidentally damage them by powering them on if there's a dried up capacitor lurking in the power supply or something. (One of them is acutually genuinely semi-rare: the ICL 3300 was produced as a prototype but was cancelled before it went into production. The PERQ 1 is older and arguably cooler, but more common.)

This happened to my BBC micro, but after I'd cleared away the explosion and bought a replacement cap kit on ebay it worked fine.

If any of them have batteries, you should remove those whenever they go into storage. Batteries have the most chemicals, so they do the most damage when they leak. Big electrolytic capacitors are second.

If you get into it, it's interesting how many sort of sub-niches there are and the amount of available info. Like, for example, bleaching yellowed plastic...there's a whole "retr0brite" culture. Or being able to recreate old undocumented PAL chips because somebody took the time to feed in every possible input and map the outputs.

I don't get the whole 'retr0brite' thing. It's 30+ year old plastic. Exposing it to oxidizers will just make it more brittle.

It's only really popular because "The 8-Bit Guy" makes 45 minute videos painstakingly showing you this process, because he lacks the technical ability or background to tell you anything insightful about the computers he's working on---usually common 80s home computers that were poorly stored in someone's attic for decades.

I started buying and restoring Commodore computers about a year ago.

A Commodore 64 was my first computer and it’s so satisfying diagnosing which chip is bad, replacing it and seeing the computer start up successfully.

I'd love to find an old TI DS-990/10 system. Used one in 1979 for a few years. Good times, implementing Kernighan and Plauger's code from Software Tools in Pascal.

should have [video] in the title.

Why? It's not a link to a video.

Reply on news.ycombinator.com