When you have a bunch of people gathered around a table for a “Dungeons & Dragons” session, you have to expect that things are not always going to go smoothly. After all, people who willingly create and immerse themselves in an alternate reality where one bad roll of the dice can lead to the virtual death of a character they’ve spent months or years with can be traumatic. And with that trauma comes the search for the guilty — it’s the dice! It’s always the dice!
Eliminating that excuse, or at least making it statistically implausible, is the idea behind this radioactively random dice roller. It comes to us from [Science Shack] and uses radioactive decay to generate truly random numbers, as opposed to the pseudorandom number generators baked into most microcontrollers. The design is based on [AlphaPhoenix]’s muon-powered RNG, but with a significant twist: rather than depending on background radiation, [Science Shack] brought the power of uranium to the party.
They obtained a sample of autunite, a weird-looking phosphate mineral that contains a decent amount of uranium, perfect for stimulating the Geiger counter built into the dice roller. Autunite also has the advantage of looking very cool under UV light, taking on a ghostly “fuel rod glow,” in the [Homer Simpson] sense. The decay-powered RNG at the heart of this build is used to simulate throws of every standard D&D die, from a D4 to a D100. The laser-cut hardboard case holds all the controls and displays, and also has some strategically placed openings to gaze upon its glowing guts.
We really like the design, but we have to quibble with the handling of the uranium ore; true, the specific activity of autunite is probably pretty low, but it seems like at least some gloves would have been in order.
How am I supposed to teach this thing to roll well if I can’t intimidate it by smashing bad dice with a hammer?
Smash the radioactive source instead. Duh.
100% stupenously fabulous awsome groovy doolally brilliant project.
Since radiation safety and minerals are my thing –
The use of autunite in this project is not a good idea. Autunite is particularly well known for its tendency towards becoming metamict – that is to say that the radiation destroys the crystals and liberates the residue as toxic, micronised, radioacive dust.
Far better to use uraninite or torbernite as your source as they lack the self-destructive nature of autunite.
On the other hand, radioactive decay is an excellent randomness generator.
Radiation safety for the novice (free download):
https://www.academia.edu/31501150/Here_be_Dragons_The_Care_and_Feeding_of_Radioactive_Mineral_Species_Feb_2017_
Would the Americium of a smoke detector work?
The build would need a different detector. Might be legal implications of reusing a smoke detector source. The principle would work fine and you should potentially expect hundreds of events a second.
Not a good idea if you’re Young Sheldon and intend to also build a nuclear reactor.
(“Where are the noooklear wessels ? “)
It’s microscopic and won’t glow under UV light, which would make this a lot less cool.
Love the title of that safety document. A sense of humour in academia does seem to be a little hard to find at times…
My thoughts go to “tickling the dragons tail”.
https://www.military.com/history/tickling-the-dragons-tail.html
here is a better link to what “tickling the dragons tail” actually was:
https://en.wikipedia.org/wiki/Demon_core#Second_incident
The clicks from the vial made them scared, so they thought they would like do stuff with the dust.
And set a click bait title about how it might give them cancer. Y’know for the lols.
The hackaday.io page has no info except telling people to watch the video, which has no links to build files or code.
Why is an an irresponsible closed build being given a hackaday signal boost?
This build may not be that dangerous, but it sends really bad messages.
Are you assuming that this site is visited by mostly Instructables-level readers?
Oh, there probably are a few. But I would advise them to only read here for the learning and not for the doing.
…until they learned enough for their doing to be safe for themselves and their surroundings.
Stuff like this is why I leave the radiation and chemical hacks to the experts, even when some youtuber says it’s safe!
> radioactive decay is an excellent randomness generator.
Assuming that you do all the processing right, and the detector is not sensitive to interference, yes. However, using the same level of care, it’s easier to get equally excellent random numbers from an avalanche noise generator. Or just use the built-in random generator in a suitable microcontroller.
Thanks for the link, Alysson. I downloaded it and will read it.
Many, many years ago (more than 40, now) I acquired a wonderful sample of autunite crystals at a mineral show in Portland. Nicely mounted on a blue foam base and in a plastic snap-cover box. Over the years, the foam in the bottom completely degraded down and had to be replaced several times. My wife makes me keep the whole thing in the shed that is about 1000′ away from our home, now!
I originally used it with a Geiger counter I’d designed and built as a teenager, by the way. (The physicist who had designed the tube itself is the person I actually bought it from — it’s part of what got me started in physics when I was young.)
I used the autunite sample back in the late 1980’s when I was working on extremely sensitive diode detector systems (pyrometry, at signal levels sometimes approaching low femptoamps and with a noise level that allowed me to see photon/boson flocking effects at attoamp levels.) We collected data over a weekend and would periodically see sudden bursts in the detector system (kept in an ice bath for thermal stability.) A physicist and I were considering the source when I mentioned the possibility of cosmic ray events being detected and I was able to find a research paper from France characterizing what to expect near sea level. The average periods matched up well. We’d see anywhere from 1000 to 10000 electrons stripped up in each event. So I got my autunite and we would measure while I would rotate the diode detector relative to the autunite source. No question! The event size (number of charges stripped up) and shape would change vs the relative angle. With the autunite on-edge relative to the autunite (rather than on-face) would yield the larger charge bursts as the cascade traveled through a longer path in the depletion region.
Anyway. Yes. It’s nicely radioactive. I can attest to that fact.
Nice! Must get on with finishing mine. It’s all in pieces waiting for me to CNC the oak block I’m mounting it all in. I’ve got for pitchblende as my source, but I’ll keep it contained in a glass vial.
Given there is user input for each die roll, it’s easier (and safer) to sample a 16 bit micro second counter whenever the user presses the roll button. You can seed the pseudo random number generator function with this value and get values for each of the dice if more than one die is rolled at a time. That’s how my digital dice towers work: https://hackaday.io/project/186489-resto-mod-audible-digital-dice-towers
But, I assume “easiest way” was not to goal here. Looks like a fun project I wouldn’t trust myself to attempt.
Maybe a great idea would also be to actually simulate the phsyics of the dice. Use a random generator like this for a seed. Use a pressure-sensitive button for user-input. Combine the pressure-sensitive input with the random number, and simulate the physics of the dice.
Meh. I think it’s probably more ‘security by obscurity’ than actually adding more randomness. Random is already random. :)
Although… You could show the simulation of the die on a screen, for the fun of it.
And you could present the pressure-sensitive button as a feature. Giving the person that presses it the illusion that they actually have some sort of control over the dice. Like with real life dice. :P
There are easier ways to get random numbers. Any noisy source such as a floating input to a high precision ADC could source random numbers. Since this device requires user input to press the “roll” button, it’s even easier to sample a free running fast (1 uSec period) 16 bit counter/timer whenever the user presses the button. Use that number to seed the GNU clib rand() function and get the numbers for the die results. That’s how my digital dice towers work.
But, I assume “easy” wasn’t the goal here. It’s a very cool project. I would not trust myself to play with radioactive samples (especially without gloves). So, I’m glad someone else demonstrated this.
Gloves would not be my worry. It’s a few seconds of exposure, and you can always wash your hands.
Radioactive particulate dust in the lungs, now that would be my worry.
If someone wants to duplicate this, it’s trivial to harvest the americium from an old smoke detector. It’s already mounted and wouldn’t need to be touched, just moved and attached to the detector. Instead of the gamma radiation emitted by uranium, americium decay produces alpha particles, which are blocked by the wooden box and the air inside it.
If a glowing tube of rocks is desired for maximum Homer effect, there are plenty of non-radioactive minerals that fluoresce under UV light that could be housed in the tube.
As the article points out, the radioactivity actually gives you random numbers, that is the cool part. Your alternatives are “good enough” for dice rolling (although the lengths that some dice nerds go to to demonstrate bias in, and to therefore deride the use of, decorative designs can be a bit OTT) but won’t stand up to scrutiny – start with reading about the discussions around the RNG used in Linux, where non-predictability is the basis of your key generation & security.