http://www.techlib.com/science/ion.html - old and gold. A friend tried building these ion chambers, but apparently getting high enough isolation between the container and the electrode was very hard.
http://physicsopenlab.org/2016/04/24/diy-cherenkov-detector/ - one for detecting muons (a PMT taped inside a thermos)
And a fun little application note from Maxim to build a gamma detector from a photodiode : https://www.maximintegrated.com/en/design/technical-document...
Looking for the datasheets for the pin diodes I've got in the parts drawer to see if I can make the Maxim circuit. I know there's some very low noise op-amps down there.
The ion chambers are impressive because even lower-cost. But they also detect the presence of people, touches and what not. :) My favourite aspect of silicon sensors like those PIN diodes is that you get the absorbed particle energy as well. In particular with alpha particles that gets quite interesting when measuring for example old ceramics painted with uranium glaze ('Fiestaware').
A class of more difficult projects is particle accelerators. Many people built fusors, but fewer people had built electrostatic and linear accelerators. In addition to technical difficulties, X-ray hazards from an accelerator is much greater than a fusor.
Cyclotrons are the hardest, a successful build is extremely rare. Experimenters reported that there were still considerable difficulties even with access to a college lab. In 2010, there was a "Small Cyclotron Conference"  - a meetup event for experimenters to share experience. It featured many fascinating presentations, make sure to see .
Particle accelerators are cool though, despite the obvious danger. If you're dealing with just electrons, IIRC you need to get to about 6MeV to overcome the nuclear binding forces and initiate a decay. You can make a hell of a lot of X-rays below that of course, but personally one of my life's goals is to cause an atom to split :) No time for that now, but that's what retirement is for. As a bonus, radiation is a lot less damaging when you're past 50, so unless you get high enough to induce acute radiation poisoning you'll probably be fine.
You can actually make a fairly low-tech accelerator with just a large Van De Graaff generator. There are even designs which replace the belt with a chain(https://en.wikipedia.org/wiki/Pelletron) and reach fairly high energies(for a home unit anyway).
Super-cold bottom (-26c or colder)
Technically all of the above would be present in a neck of a really cold beer, as it is being opened. But I don't think my fridge can make the beer cold enough. Soda obviously doesn't work because it contains no alcohol vapor. Also soda is solid at -26 c, unable to release vapors whereas beer will still be a liquid (I think probably depends on the beer).
Bottle didn't break though the contents were sadly inaccessible - seemed to be an opaque white viscous slurry.
I was hoping this article would cover building a particle accelerator in a pool, but it only covers the detection part of the system.
Ugh! I wonder what other hidden features are lurking on my laptop hardware. I am conflicted because while I understand perfectly well why market forces compel hardware manufacturers to add these "features", I don't like the idea of not being able to trust my primary computing hardware.
The audio subsystems contain the basic DC blocking and anti-aliasing features that you'd expect, but do not seem to add any other processing unless you ask for it (e.g., equalization). At least, nothing that would be noticeable in either steady state or transient analysis.
So far as I know, apps like Skype and Zoom do their own processing.
These "features" are features because they better map the physically raw input to the actual desired input for the most common use case. Sound is only sound because of human perception; if the hardware is sensitive to EM interference that the human ear isn't sensitive to, that's noise (literally).