I recently re-discovered this fantastic accelerator demonstration using a bowl and a ping-pong ball from Todd Johnson at Fermilab.
The thing I think is great about this demo is two-fold. First, the principle of acceleration is pretty straightforward; for a particle with a constant charge you need to have the right polarity (+/-) on the strips at the right time to make it accelerate. The second great thing to show is the analogy of the bowl as the magnetic field. This is especially relevant to the machines I work on, called FFAGs, as the magnetic field really does increase radially, so the particles spiral outwards just like in the bowl as they gain energy. But they don't spiral out *too* much, because the field gradient keeps them confined to a fairly tight circle, just like the curvature of the bowl does for the ping-pong balls.
I really want to make one to use in a new public lecture, so I contacted Todd about how he made it. I hope I can convince someone to help me, although I'm still a bit worried about the danger of using 15kV in a public demonstration.
In case you want to re-create it yourself, here is what he told me about it:
"The power supply I used is a surplus unit which provides about 15KV from a 12VDC input, intended to be used in an "air ionizer". Sadly these devices have disappeared from the market. Several manufacturers make small DC/DC converters that look like they would do the job nicely, for example: http://www.ultravolt.com/products/348 These cost about $250 or so as I recall.
The clear "bowl" is actually a custom made part. I had found some serving bowls which worked well for the first couple of these I made but of course those also soon disappeared from the market. The key feature is that there should be a central flat spot which then meets the curved side with no discontinuity. Any faint corner or step will trap the ball and prevent it from accelerating, due to friction between the two points of contact at different radii it will then have. My hemisphere was fabricated by a company that makes acrylic domes, and they were able to accommodate my request for a flat spot in the center. A regular hemisphere will also work, however the ball will need a push to start it since with zero initial momentum it cannot travel from one set of electrodes to the other set (which would of necessity be uphill).
The ping pong ball is coated with a carbon-based conductive paint called "wire glue" which I found locally. I'm sure that coatings such as "Aquadag" would also work well. I gave the coating a protective layer of clear spray enamel (Krylon), which seems to make it extremely rugged and does not significantly affect the conductivity at the voltages involved.
Regarding the safety issue, it is not apparent from the video but there is an acrylic cover over the device which prevents onlookers from reaching in and contacting both sets of electrodes. The cover also improves the performance by allowing the air to begin to move along with the ball rather than remaining turbulent as it would if the top were open to the room."