[Amps] SHV / proper cable for DC energy transfer (not for RF energy transfer)

[LB3HC]

I have found that the Radiall SHV part number  R317072000 for 50 ohm
coax with crimp connection is speced and tested to 12kV (Note: when
mated, not when unmated). The spec is signed but not verified. Last
signature date, 1979. Even if the spec is old, this should be more
trustable that purchasing a cheap Cina copy off eBay. This is a safety
critical component.

Datasheets:
http://uk.rs-online.com/web/search/searchBrowseAction.html?method=getProduct&R=2127438&cm_vc=prev_uk

Coax:
RG58 is delivered in several different formats. Foam core, solid core,
alu sheething, copper sheeting, steel core with thin copper, steel
core with thin copper and silver, stranded copper core, solid copper
core. Normally coax is rated after its RF properties, not its DC
properties.

Application in HV-RF deck for linear amplifier:
Energy transfer. High voltage. Direct current. No RF, relatively high
current 1,5A
Comment: This causes no skin effect and uniform current density in the
whole core.

Problem:
Since RF currents above a certain frequency does not penetrate the
core of the cable, the manufacturers save money by using a steel core
with thin copper outside. We are running DC and therefore I don't
think I like the idea of using a steel core for this application. I
think a proper cable for this application has a stranded or solid
copper core.

The manufacturer of the connector in question specifies the use of
URM76 cable manufactured by Belden. This cable is specced to 4kV DC.
If we run 3650 V DC, we have a safety margin of approx 10%.
Datasheet: http://www.farnell.com/datasheets/477134.pdf The spec has
not seen any revisions and is currently in version 1, dated
25.10.2007, unsigned.

I am inclined to use Belden cable and Radiall SHV connectors. However,
is this safety margin wide enough? In case the current limiting
resistor for B+ is in the HV deck and proper fusing is designed in, it
should be operator safe at least.

What do you think?

LB3HC