Ha, little fall down the memory hole. I had a Harris 20mhz 286 back in the late 80's. I thought that thing was a beast at the time. Paired it with a Conner 100MB disk and I remember my brother laughing derisively, "Who the hell needs 100 megabytes?"
I'm Japanese, and the 80286 at 10MHz was huge for Japan's PC-98 scene. The V30 handled backward compatibility while the 286 ran much faster than what we had before. This project brought back memories—the 286 was the chip of my era, and it's great to see people still exploring its capabilities decades later.
I'm not a fan of breadboards, they tend to be unreliable even for trivial circuits. We need something more affordable and practical for home PCBs[^1]. Why is it that nobody has invented a tin 3D printer, or at least a 2D version of it, i.e. a tin plotter?
[^1]: I'm discouraged from home-etching by the chemicals and the dark-room part of the process.
For the 2D version, you might not need very much custom. Use a regular pen plotter and use a pen with conductive ink. These both exist today, though personally as a hobbyist PCB designer, I can get 2-layer and 4-layer boards cheap enough from JLCPCB or Oshpark or PCB Unlimited that I don't bother trying to make them myself.
For $2,500 the Carvera Air makes very nice 2 sided pcbs with solder mask. Though even in raw materials it is hard to match a finished board from China if you can wait a couple of months.
I've seen a similar project a while ago and thought this was about the same thing at first: [1][2]
Both essentially built a DIY chip tester for a 286 and both built around a Harris 80C286.
If I understood it correctly, the goal behind this project seems simulating the rest of the PC, purely for the challenge and learning experience, documenting the process of building the chip tester (and getting mildly philosophical in the process).
The other project was more directly interested in the 286 itself, undocumented instructions, corner cases in segmentation behavior, instruction cycle timing, etc. and also trying to find out if there are any difference between the Harris and Intel variants.
Looking at the amount of wires going into this, my instinct is that this cannot scale, in 5-10 years this won't be doable for a Pentium chip, at least not as an at home hobby project. But I actually think it could go the other way, and in 5-10 years you'll be able to do this at home for far more sophisticated kit, unlocking crazy amounts of reverse engineering possibilities that were once thought of as near impossible, or at least only possible for a nation state scale setup.
I'm pretty sure if you really want to, you could do something like this as a hobbyist with a Pentium right now.
Instead of futzing with wires on a breadboard you could simply designing a PCB up front, throw the design over the fence at JLC or PCBWay, insert coin, wait patiently at the mailbox, solder your scavenged Socket 7 onto the board.
The days of toner transfer and aquarium pumps are already long gone. Getting production quality, one-off, multi layer PCBs done as a hobbyist is dirt cheap these days, no government budgets required.
> you could simply designing a PCB up front, throw the design over the fence at JLC or PCBWay, insert coin, wait patiently at the mailbox
It blows my mind that I can use free-as-in-beer Free-as-in-Speech software to design a PCB, email it to a dude in China, and get a finished working professional-looking PCB back in my hand within a week, for the price of a couple of coffees. And if I want the components stuck on too, it'll cost a little extra, maybe three coffees it costs now.
If I want it really quickly then for the price of a decent takeaway curry I can have it flown over next day. What the actual hell?
Edit: the slowest part of "next day" is when it hits the UK, and if I could guarantee it just gets delivered to DHL's Edinburgh depot I could drive down there in two hours.
Ha, little fall down the memory hole. I had a Harris 20mhz 286 back in the late 80's. I thought that thing was a beast at the time. Paired it with a Conner 100MB disk and I remember my brother laughing derisively, "Who the hell needs 100 megabytes?"
I'm Japanese, and the 80286 at 10MHz was huge for Japan's PC-98 scene. The V30 handled backward compatibility while the 286 ran much faster than what we had before. This project brought back memories—the 286 was the chip of my era, and it's great to see people still exploring its capabilities decades later.
Japanese PC-98 games have an aesthetic that’s so unique. There’s this one Twitter account that posts images from visual novels from that era and they all look so pretty: https://x.com/PC98_bot Also on bluesky https://bsky.app/profile/pc98bot.gang-fight.com
Cool! But the wires!
I'm not a fan of breadboards, they tend to be unreliable even for trivial circuits. We need something more affordable and practical for home PCBs[^1]. Why is it that nobody has invented a tin 3D printer, or at least a 2D version of it, i.e. a tin plotter?
[^1]: I'm discouraged from home-etching by the chemicals and the dark-room part of the process.
For the 2D version, you might not need very much custom. Use a regular pen plotter and use a pen with conductive ink. These both exist today, though personally as a hobbyist PCB designer, I can get 2-layer and 4-layer boards cheap enough from JLCPCB or Oshpark or PCB Unlimited that I don't bother trying to make them myself.
One maybe unrecognized problem, of breadboards is not so good for high speed digital circuits due to capacitance
Bad Obsession Motorsport actually went this route, you can see their custom printer on their youtube[1] channel
[1]: https://www.youtube.com/watch?v=FzrZoVKT8gM
For $2,500 the Carvera Air makes very nice 2 sided pcbs with solder mask. Though even in raw materials it is hard to match a finished board from China if you can wait a couple of months.
In my experience it's a couple of weeks, not months from China.
wirewrap circuit construction would be more reliable. However high speed (if needed ) could be a problem ? from crosstalk.
I've seen a similar project a while ago and thought this was about the same thing at first: [1][2]
Both essentially built a DIY chip tester for a 286 and both built around a Harris 80C286.
If I understood it correctly, the goal behind this project seems simulating the rest of the PC, purely for the challenge and learning experience, documenting the process of building the chip tester (and getting mildly philosophical in the process).
The other project was more directly interested in the 286 itself, undocumented instructions, corner cases in segmentation behavior, instruction cycle timing, etc. and also trying to find out if there are any difference between the Harris and Intel variants.
[1] https://forum.vcfed.org/index.php?threads/286-cpu-experiment...
[2] https://github.com/dbalsom/arduinoX86
Here's a project that also does this to ensure cycle-accuracy for their emulator: https://github.com/dbalsom/arduinoX86
Looking at the amount of wires going into this, my instinct is that this cannot scale, in 5-10 years this won't be doable for a Pentium chip, at least not as an at home hobby project. But I actually think it could go the other way, and in 5-10 years you'll be able to do this at home for far more sophisticated kit, unlocking crazy amounts of reverse engineering possibilities that were once thought of as near impossible, or at least only possible for a nation state scale setup.
I'm pretty sure if you really want to, you could do something like this as a hobbyist with a Pentium right now.
Instead of futzing with wires on a breadboard you could simply designing a PCB up front, throw the design over the fence at JLC or PCBWay, insert coin, wait patiently at the mailbox, solder your scavenged Socket 7 onto the board.
The days of toner transfer and aquarium pumps are already long gone. Getting production quality, one-off, multi layer PCBs done as a hobbyist is dirt cheap these days, no government budgets required.
I don't know about Pentium, but there's definitely some homebrew 486 projects out there, e.g.:
https://maniek86.xyz/projects/m8sbc_486.php
> you could simply designing a PCB up front, throw the design over the fence at JLC or PCBWay, insert coin, wait patiently at the mailbox
It blows my mind that I can use free-as-in-beer Free-as-in-Speech software to design a PCB, email it to a dude in China, and get a finished working professional-looking PCB back in my hand within a week, for the price of a couple of coffees. And if I want the components stuck on too, it'll cost a little extra, maybe three coffees it costs now.
If I want it really quickly then for the price of a decent takeaway curry I can have it flown over next day. What the actual hell?
Edit: the slowest part of "next day" is when it hits the UK, and if I could guarantee it just gets delivered to DHL's Edinburgh depot I could drive down there in two hours.
Reminds me of the Tamagotchi Matrix:
https://hackaday.com/2015/11/24/building-the-infinite-matrix...
So they're stepping a 286 very slowly and sorta... bit-banging the I/O pins on it?
Love it. No notes.