I figure if they use Americium-241 pellets to make commercial smoke detectors, there's no reason they cannot use such pellets for hardware random number generators. I read an unconfirmed anecdote that casinos often use them in their digital gaming machines. Even though they don't have nearly as high of a bitrate as the optical HRNGs, radioactivity is quite reliably random.

There have been a few hobbyists who built their own but that's a bit too risky for me right now.


This is an old idea:


I have never seen anybody selling this kind of devices. My assumptions about the reasons:

It contains radioactive material (no matter, was it a capsule from an already approved home appliance or even that appliance "as is") and thus must go through the full inspection & approval process for devices that contain radioactive material to be legally kept for sale.

Every subprocess for the handling of the radioactive capsule also must be approved and have a bunch of certificates.

It's well possible that this device gets its certificate, but the process is far too heavy for a device that is not very often asked in local supermarkets.

|improve this answer|||||
  • \$\begingroup\$ I'm aware that it's not a new idea. However, I don't see why it hasn't been done as it's useful(even though not to most ordinary people). \$\endgroup\$ – Mr X Jan 24 '17 at 20:24
  • \$\begingroup\$ @Mr X If the circuit is kept simple - no parallel systems, only one detector - the generation rate of numbers at a given bit resolution can be miserable The detector generates slowly fading tail after each hit like the ionization tail in the old Geiger-Muller tube. To keep the detected pulses apart enough, not plenty of detections/sec can be allowed. \$\endgroup\$ – user287001 Jan 25 '17 at 23:23

Not the answer you're looking for? Browse other questions tagged or ask your own question.