To find this out I've made a computer program to simulate the movement of charged particles in an electric and magnetic field (in vacuum).
Coulomb's law, Biot-Savart formulism and Leapfrog integration is used (non-relativistic).
First I came across a configuration that confines only positive ions (or electrons). But I discovered that this was quite similar to the already existing Penning trap and would not be very effective to produce fusion, because of the nonneutrality of Penning trap plasmas.
Then I changed it a bit and now it's possible to trap both positive deuterium ions and electrons. The design just needs a constant magnetic field of about 1 tesla and a couple of rings and small spheres with a voltage of about +/- 200 kV. Depending of the polarity of the rings & spheres, there are two possible configurations. The ions and electrons are generated in the program with random speeds up to 3E6 m/s in all directions (and still keep confined).
But.., this is only a simulation with very few particles (no more than 500 or 1000), the simulation time is very small and it doesn't prove that any fusion would take place..
If you are interested:http://www.valgetal.com/physics/Droomid ... g.%206.jpg
(1e configuration)http://www.valgetal.com/physics/Droomid ... 0fig.7.jpg
(2e configuration)http://www.valgetal.com/physics/Droomid ... oom11.html
(calculations, screenshots etc.)https://www.youtube.com/watch?v=bvvv1YLbuLY&t=11s
(video capture of a simulation)
Thank you for your time and regards