Hi FilipB,
Well i thought that page you linked explained it pretty well,
but i suppose you need a background in thermionic devices for
it to be easily understood.
So i will try to illustrate its operation from a slightly
different view point.
When the filament is running, you should be able to see it in
the dark, with nothing else lit, as faint orange glowing lines.
When hot in a vacuum, the surface electrons make a sort of cloud
around the wire.
If something nearby becomes positive, some of these electrons
will will shoot off towards it. That leaves no shortage of them
behind, as more will travel along the wire.
However, there are grids in the way, the potential on the grid
can stop this flow of electrons.
So by having in this case, eleven grids covering different areas
of the phosphor-ed section, you can pick which section will be
active.
This is like a matrix-ed arrangement that is often used with LEDs.
Being as only one section under a grid is active at a time,
this means that individual parts of each section can be common-ed
up.
If you look at the pic that ive lifted from that page, you can see
that the 7-segment numbers are each covered by a grid, that is to
facilitate the matrix-ing that i mentioned.
These pretty little displays will usually operate on fairly low
voltages, maybe as low as four or five volts.
When experimenting to draw up the grids and the elements, use a
series resistor, these little glowing segments don't need much
current, the grids need hardly any current at all.
So, to sum up, to pick a segment, its grid has to be plus, and
the phosphor element has to be plus, the filament is regarded as
cathode and is taken negative. Sometimes its centre-tapped, that
would probably be for AC use, the c/t being for negative and both
ends balancing with the ac. But these filaments are such low
currents that i would feed it DC anyway.
I hope this has clarified it a little.
Regards John