Insulation Resistance test on DC cable

[john_gopt]

Hi all,
I had a question posed to me the other day which made me think. I am used to doing megger (Insulation resistance) tests on AC circuits where I have a definate ground to connect to, but what if I want to test a DC cable.

To simplify the problem I have a long straight conductor and have access to both ends (close enough to attach the megger), imagine a loop. If I connect the live to the conductor and the black test lead to the insulation (or maybe wrap a new conductor around the insulation so the clips don't damage it) at the opposite end of the cable would this give me a good IR measurement?

Am I just getting tired and not thinking straight?
John

 

[7anoter4]

I think it is good idea but I would add a graphite painting. See-for instance:

http://www.solianiemc.com/products/...tive-paints/?gclid=CNrf7ODhxtACFUUW0wodSrIImw

 

[HCBFlash]

So you have a single, unshielded, conductor passing through a non-conductive raceway, and an ungrounded system? I doubt it, and I hope not!

.


Me wrong? I'm just fine-tuning my sarcasm!

 

[qman5]

@HCBFlash

Are those all the assumptions that you will now treat as fact? Read and comprehend the question before you post.

@john_gopt

I see no errors in this methodology. It makes absolute sense to me.

 

[HCBFlash]

That method will tell you about the resistance characteristics of your cable's insulating material, as installed, but will not tell you that the insulation is undamaged physically, or even that the conductor is actually isolated and insulated from other conductors. As you describe it, there would be no need to have test leads at each end of the cable. You would need a shield, conductive sheath, or conductive raceway to become part of your test circuit to determine the insulation resistance, as installed, between the conductor and whichever of those separate conductive paths present, whether or not any part of that circuit/system is grounded.

.


Me wrong? I'm just fine-tuning my sarcasm!

 

[itsmoked]

This will leave the cable dangerously charged won't it?

Keith Cress
kcress -

http://www.flaminsystems.com

 

[electricpete]

So you are going to connect one lead of the tester to the conductor and you're trying to figure out where to connect the other lead (as a "ground reference" for the test in my loose terminology).

Like hcbflash, I agree it is necessary to understand the physcial environment of the cable. If it's shielded, then of course use shield as ground reference. If it's unshielded but runs through a conduit then you'll want to include that conduit in your ground reference. If you find it necessary/helpful to connect some other metalllic strap around the cable near the termination, that's fine… just connect that extra strap to the conduit so that both items together will form the ground reference.


Bear in mind there are two basic types of flaws you may be looking for:
type 1 - through-insulation defect somewhere along the length.
type 2 - leakage from the termination to groundplace… may track along surface of the cable to the nearest ground.

To detect type 1 ideally you'd like the outside of the cable in intimate contact with ground all along its length (ideal but difficult... use some judgement). To detect type 2 you'd like to add a strap nearer to the terminations than the location nearest the termination that might contact ground during service. The insulation resistance reading of course varies with the ground reference chosen.


=====================================
(2B)+(2B)' ?