Rob wrote:

>> Don't they shield the electronics from radiation? I suppose that it
>> is not very effective.  Also, I would think that tubes would be
>> somewhat susceptible as well due to charge excess build up on the
>> elements/ ionization of the gas inside (if it uses one).

> There are nuclear hardened versions of IC's and transistors.

  To some extent.  But that is confusing radiation exposure with EMP and
it is the latter that was really under discussion.

> I believe the problem with semiconductors is that radiation impulse
> flattens out the hump on the junction making the device a conductor
> (resistor).

  The problem is a transient with a very high rise time and field (i.e.,
volts per metre).

  A high rise time is otherwise known as a very broad spectrum, with
appreciable power components at frequencies likely to be used in
communications equipment.  Transistors poorly tolerate substantial power
dissipation, and even T-R switching diodes are likely to fail in the
kilowatt range.  Valves OTOH are limited by their emission current - no
more will flow.

  Many volts per metre means that really amazing voltages can build up
along the length of the plane, particularly across any loops.  On a
practical note, even if you use fibre optics, you still require power
reticulation within a plane.

  Conductive shielding (copper, silver, aluminium) *is* effective
against EMP (just watched "Small Soldiers" at the movies last night by
the way!), but power has to get in somewhere.  Coaxial capacitors and
threshold devices can be deployed of course.
--
  Cheers,
        Paul B.