I wonder what the planned lifespan was in terms of component selection. That beast is older than pretty much anything with electronics that I own that doesn't burn gasoline.
Arcade games and pinball machines are limited-life industrial equipment, meant to be discarded after 2-3 years max.
They generally used the cheapest components they could source as saving $2 on the BOP meant more profits, and it wasn't in their best interest to have old games competing with their new games.
The fact that many have survived 20+ years without even basic maintenance is a testament to late 70's engineering.
Mostly in the sense that we didn't know enough about electronics yet to make crappy transistors! They were all huge and built like tanks.
Transistors are still built with metal TO-22 cases like you see above.
They're built like tanks to dissipate heat, and you'l find those very same transistors in your "modern" stereos today. We also knew plenty how to make transistors in the 80's, having started manufacture of them in the 50's.
With regards to maintenance, most of these games had to be maintained monthly. Hence, the bad conditions of their circuit boards, traces, and evidence of poor/shoddy rework skills. The boards we're working on are completely covered in flux and bus wire. It sucks.
Transistors are indeed still made like that, and you _might_ find some like those in modern stereos, but not necessarily in the amplifier stages.
The last stereo that I owned had a single large block in the back that was basically a multi-transistor array in a single epoxy package. I think they decided that since you typically need matched transistors for both the driver and final amplifier stages, not to mention two sets of everything (one for both the left and right channels), it was somehow cheaper to fabricate a combined driver/amplifier module out of the same hunk of silicon rather than fuss around carefully selecting individual parts with similar gain/Q/whatever characteristics.
The end result is that while what's inside that epoxy blob might be the same kind of bipolar transistors they've been making for decades, if just one semiconductor in that module craps out, you'll need to replace the whole thing, and likely waste a lot of time and money in the process.
So while we haven't changed much about the basics of transistors themselves over the past 50-60 years, we've sure gotten a lot better at vendor lock-in.
On the flip side, although I can work on my Robotron because it has discrete components, half of the TTL chips are obsolete and NLA, so you have to scrounge parts from other dead machines - ColecoVisions are a particularly good source of chips - or hope some other hobbyist has figured out a modern solution.
It's not like I can whip off a 4116 RAM chip on a lathe...
Lots of decent, modern stereo receivers still use discrete components in the amplifier stage. IIRC, even the modern (and cheap) receivers made by Sherwood are using discrete amps.
Meanwhile, I once had a circa 1990 JVC receiver which had a monolithic IC amp module, running in class AB. It worked OK until the module died, and then it was useless.
There is a shift lately in home audio, as well as the pro market, toward class D PWM amplifiers. It has been gaining momentum over the past decade or two (mostly for efficiency, thus making power supplies and heatsinks/cooling cheaper and lighter). Many of these are built around a giant IC (like every single cheap Wal-Mart all-in-one surround system), but even with the dramatic shift in topology that class D brings, many manufacturers are still using discrete transistors at the output stage.
Myself, I'd rather own/hear/buy a good IC-based amp than an older discrete amp that was repaired poorly. I've got bad memories of spending many nights hearing the badness of what was once a nice, discrete Yamaha professional amp after someone at some point had replaced several (not all) of the MOSFETs with parts of a different make and model. It was an ugly sounding beast due to that half-assed fix.
At least I got paid to hear it. :-/
"We also knew plenty how to make transistors in the 80's, having started manufacture of them in the 50's." I suspect he was suggesting that we knew how to make transistors that _worked_ just fine. In the subsequent 30 years we've also gained the technology to make them so cheap-and-crappily that their MTBF is within 4 nines of the typical warranty period of the products expected to be made with them.
That's a lot of busted components - was there an electrical surge or something? I know caps can dry out over 20 years but those transistors should be fine.
Perhaps your computer curse has followed you to arcade games.
We have no idea what's wrong any more, and are now following the instructions in the FAQs which all say, "just start replacing god damned everything."
Have you contacted the nerds on sites like the KLOV Forum? I know people local to me who are experts at Atari vector systems but no one specifically in SF.
When it gets to random component replacement, it's pretty bad.
Asked some friends of mine, they recommend:
Ken and TJ - (408)390-PINS
These guys have restored both my cocktail and stand-up Tempest to life several times now:
There are 3 transistor packages there:
small, all plastic: TO-92medium, plastic with metal tab: TO-220largest, diamond shaped, all metal: TO-3
it seemed like your star wars machine was having some issues too? I was most sad however amply entertained instead by your lovely guests.
Yeah, Star Wars is on the list, too. It's fine for the first ten minutes or so after powering it on, but after that it starts getting jittery as hell.
Electrolytic capacitors not only have a finite service life, they have a *shelf life*. What's more, they have plenty of failure modes beyond losing electrolyte.
When I realized that fact about vintage electronics, which are out there on the far side of the bathtub curve where everything starts breaking again, I decided that if I'd pay to get it fixed, I'd turn it off when it wasn't needed.