Cable routing dilema

nd cable in a fixture I've made.

ut of a high voltage supply (up to 3000V, but it's used at 1500V).

for ac and 1000:1 for dc. The input of the ultravolt is two unshielded HV cables. The output of the power supply under test and the input to the plas tic box is HV coaxial connectors.

of me says to route the ground as short as possible to the ground of the ul trabox, the other part says that I should keep the ground next to the unshi elded wires to avoid discontinuities.

I'm guessing there won't be much difference either way. Do all the four connectors on the front panel have ground as the outer shel l? (no isolated connectors?) If so I might strap a ground line between all of them... (like the front panel was made of metal.) But that might be wrong too.

(sorry I can be of more help.) George H.

On Thu, 7 Apr 2016 03:11:24 -0700 (PDT), snipped-for-privacy@gmail.com Gave us:

The HV supply *may be* a full floating supply and if it is, your ground wire should not be grounded to the case at all.

What you should do is get a nice HV flatso resistor and make your own voltage divider pick point to take a voltage reading from and encapsulate it with the supply, thereby keeping the entire monitoring circuit small and proximal to that which it is monitoring.

You could also get some corona problems routing those lines right next to each other, despite the fact that they are on HV wire. I have seen corona poke (via electron migration boring) a hole through a 60kV HV wire before and it was only a 3kV source on the line.

I couldnt decide either, but the coax is not exactly 100% shielded, so a GND shield between the HV leads and the Coax might do more to eliminate corona leaking into the coax. But even that might be splitting hairs.

Cheers