This is awesome. I built a metronome from a Forest Mims book and parts I bought at RadioShack back in the day, so it's nice to see the timer in all it's glory. One of my biggest surprises when I reported for duty on a submarine in the early 80's was that most of the electronic systems were discrete components.
The other day, an HN post showed RadioShack.com has changed hands and the home page is all about a new type of crypto. Just checked, still is.
I’m surprised that RadioShack link still works, including the Add to Cart and Checkout buttons! Maybe the new owners are keeping the old e-commerce side alive?
I was wondering that too. I did some road trips in 2020 and was surprised to run across a functioning RadioShack in Montpelier, Idaho [1] so it's not dead. There's even a Tandy Leather store in Portland [2]. According to Wikipedia, the eCommerce portal is a going concern along with 500 franchises [3] and the branding, which is now owned by Tai Lopez and Alex Mehr who are taking into crypto-currency. The Radio Shack saga continues. Probably not interesting to younger people, but I have to admire the staying power of the brand when companies come and go all the time. Especially when I consider the Fry's Electronics in Silicon Valley are all shut down or selling junk on consignment. I don't know what happened there.
> The thing is that this is not a good design (the 555). I had a few years of experience, I’d say about five years, but I had no teacher, and I had to learn it by myself.
> You know it was really the beginning of design, so looking at it now, I would say “I wouldn’t do it like that again”. But nobody has actually changed it - it is still the same. They have shrunk it, they shrunk the dimensions. It was a 10 micrometer design, that was the standard size. You could make it in four (micrometers) now. This is more die per wafer, but nobody has changed the arrangement or the schematic.
> As a timer, you know, you trigger it and it runs for a certain time, it is very good. It has a temperature coefficient of like 23 parts per million. Over a large temperature range, that’s like .1 %. Its very stable. In free running mode, as an oscillator, its not so good, about 150 parts per million. And that you could improve down to about 10 ppm. So you could make an improved product. I’m amazed and stunned that in 30 years, somebody hasn’t looked at the schematic and said, “I can make this better”, so for the same area and same cost, and then they have a better product. Nobody has done that.
Of course he could make it better. Everybody can. The issue is if it makes sense. The only engineering science where improving something is (stil) cheap is SW. Although SW seems to be cursed and almost every improvement they make is a complete redesign.
555 is a cheap signal generator. It is very good at what he does. When you need a better one you can have a uC ( which you need to program and test) or a synthetiser ( which is more expensive). No need to thrash it just because it is old.
Eh, he said it could be improved for the same area and (per-unit) cost, and I trust he knew what he was talking about. Amortized across all the number of units sold, the redesign cost is very little with modern methods.
I enjoyed reading the interview. Besides this snippet, he talked about how modern design differs from back when he made the 555. It's always super impressive to me when someone keeps up with their craft through such huge technological changes.
People complain about thin computers these days... but this, right here, is the core of the issue: we need thicker ICs, only then will we finally get back our much missed "luggable" computers. No more electron tunnelling, electro migration, super scalar nonsense. We need good ol' fashion sturdy, reliable, hand drawn ICs.
Since the original 555 was developed in 1971, I'm wondering if the inventor[1] or Signetics, the company where he worked, would have made a similar physical mock-up in those days before electronic circuit simulation software existed.
Yes, almost certainly. Designs were routinely breadboarded into the 1980s. Maybe discrete transistors, maybe a mix of transistors and some simpler ICs. Even big designs got the treatment -- the Intel 4004 creators speak of a breadboard prototype in one of the oral histories.
My impression is that such prototypes were something given up quite reluctantly. It's the best way to verify your design works, after all. But eventually ICs developed enough. Either so tiny in process, that they behave electrically very different from circuits with macro components. Or simply so large in design, that it would cost more to build a discrete component prototype than just do a prototype mask run.
Probably, although by then you could get op amps and flip-flops as ICs. A 555 timer is two comparators, a flip-flop, an inverter, and a power transistor. If you have those building blocks, you can make one easily.
What a great project Idea! I have often looked at the internal schematic of the 555 and wondered if it could be made with discrete components - now I know that it is possible :-)
Its the primary (20kV) for a tesla coil. The 555 toggles a power MOSFET which drives the primary of a car ignition coil making a nice crackly 20 kV spark.
The 555 was my first introduction to electronics. I still have the 555 book my dad gave me back in the 80s!
Feel free to double check my math, but a modern M1 CPU with its ~16 billion transistors would be about 380 million times bigger (imagine a square ~1.7 miles on a side)
It would take an electric signal about 40ns to travel from the center to one corner and back, and a signal certainly needs to do more than that in a clock cycle, so the physical limit on its operational frequency would be well below 25 kHz. If you want comparable computing power to a regular M1, you'd need to build a tower of them at least 100,000 high. On the plus side, you could build a city on the stack and no one would need heating in the winter. The cluster should be powerful enough to coordinate the traffic lights. Synergy!
> The cluster should be powerful enough to coordinate the traffic lights.
This probably says more about how bloaty modern software development is, than the capabilities of your hypothetical macro-M1. You wouldn't need a cluster. Just one should do. The entire traffic light system of San Francisco ran on a few PDP-8 controllers, once. About as fast as the macro M1 at 25 kHz. (Even at 25 kHz, with parallel execution, it'd still be completing hundreds of thousands of 64-bit-data instructions per second, running circles around something like the original IBM PC or even the early Macintosh.)
With hardware so cheap and software stacks so deep these days, we often grossly overestimate the number of cycles really necessary to complete a task.
https://www.radioshack.com/products/getting-started-in-elect...