[ackalker]

Ah, memories of one of my very first lab assignments in digital circuits engineering class (back when computer science was still busy splitting off from applied mathematics and electrical engineering). We were given nothing but a breadboard, some chips with basic logic gates (NANDs, multiplexers/demultiplexers), an EPROM, access to an EPROM burner (w/ hex keypad!), some LEDs, switches, resistors, jumper wires and a power source.

The assignment was to build a traffic light simulator, set the whole thing running, change the traffic lights as a result of switch inputs acting as sensors and a simulated interval timer.

Some students were baffled by this (lectures hadn't caught up with lab assignments at that point): how could you build a small processor using only logic and an EPROM? There's no memory or registers to keep state!

This is what differentiates combinatorial logic from sequential logic: feedback. Use some of the EPROM's data outputs along with logic gates and switch outputs (using the multiplexers / demultplexers) as address inputs to the same EPROM.

Sweet memories of solving Karnaugh maps, Quine-McCluskey minimization, logic hazard mitigation, etc. Good times.

[gnarbarian]

I took a similar class. I loved it. we started out with a project that would make leds perform a cylon style pattern [1]

For the final project we had to implement a multiplier (booths algorithm) using nothing except breadboards, wires, and a few very basic ICs that had Ands nors etc.

I loved that class and it really helped me understand how a processor actually works.

[1]https://68.media.tumblr.com/46f8f8a9ffd3f4bf900e1710c4e08fd0...

[groby_b]

> I loved it

You misremember ;)

I took a similar class, and like for everybody else I ever talked to, half of the wires were somewhat broken and caused intermittent connections. Solving the puzzle of building the system was fun, and quickly done - finding the &#(@ broken cables and replacing them was weeks of tedium.

I suppose in a way it was perfect preparation for a career in software engineering. Solve the core problem, and then spend significantly more time on actually making it work in a world the adamantly refuses to be pure and perfect.

[mycall]

My solution? Dumpster diving for wire.

[knodi123]

> Quine-McCluskey minimization

haha, I'm remembering how I felt like I was getting away with murder because I took 4 courses in college that spent time covering truth tables and complex boolean statements, and 3 of those taught Quine-McCluskey. I felt like an adult visiting an elementary school. (briefly, until we finished that chapter)

[ue_]

I'm doing ECE in my first year on university now, and so far the labs are pretty boring, except for where we are making our own CPUs using xilinx then loading those onto an fpga. We're way ahead with state machines etc. in the digital lectures.

Analogue is the hard part for me, I guess you didn't do so much of that in digital systems engineering.

[gravypod]

Is this how Shenzhen I/O was done back in the day?