Keeping the Keck Telescopes Running | Hacker News Reader

Keeping the Keck Telescopes Running

darkerview.com

76 pointssohkamyung7 comments

I always wondered how old machines get fixed when replacement parts are no longer in production. This was very educational.

A lot of these types of installations are bespoke, at least in large part.

I used to have a friend that worked at a particle accelerator lab, they had a complete machine shop on site so that they could make repair parts, or new parts as needed for various experiments. These are not things you will ever buy at Amazon or even a place such as McMaster-Carr.

I'm personally curious to see what will happen in 30 years. Its one thing to keep computers from 20-30 years ago running because generally the components are fairly simple, but not anymore.

The act of unboxing a prototype PCB for the first time, soldering all the components up by hand at the hot air station and then switching it on for the first time (and cringing in anticipation of magic smoke or expensive bangs) is one of my favorite things, speaking as a software engineer who dabbles with hardware design in smaller projects for work.

It encourages really deep thought and full exploration of your design before "compilation" (spinning a PCB), something that I don't get as much with the quick turnaround run-error-fix-run cycle of the interpreted languages I develop in day to day.

It reminds me more of my really early days playing with Gentoo on a Pentium II with 128MB of RAM, where running make had a real cost on a large project. It really encouraged you to think changes through instead of just slapping some changes in and running it to see if they were correct or not.

That’s a great web site. Worth the time reading through his posts.

I spent the last two weeks laying out a PCB with iPhone-like density; now I hate laying out PCBs.

It’s therapeutic until you spend too long on it, then it’s drudgery.

High desity PCBs are a challenge. Most of the time as hobbyists or for engineering prototyping that density is not necessary. Giving yourself 10% or 15% extra space is hugely beneficial to your sanity.

Production boards are a different story, but that isn’t likely to be the first iteration.

Article:

> The details in a PCB layout seem endless, you have to check everything thrice. Even after doing so you just have to accept that there will be some small error. Hopefully something you can live with.

At the robot club I give these encouraging words to people trying PCB design for the first time: “PCB design is like golf. Low scores are better, but holes-in-one are rare.”

A couple of wires tack-soldered on can salvage most boards.

Thank you for your service to science.

Reply on news.ycombinator.com