When testing branch circuits to assure that there does not exist a connection between the neutral and ground conductors, what is the acceptable resistance allowed between the conductors?
What is the proper instrument to use for this measurement?
I have received different opinions from testers, inspectors and engineers.
In addition to your opinions, I am looking for documentation that will define the requirement for performing the test.

I have a couple of opinions, and they’re worth every penny you paid for ‘em.

Where ground-fault sensing is needed, downstream neutral-ground connections will desensitize the type of ground-fault protection needed for 480Y/277V service/feeder overcurrent devices. For modern GFCIs, a neutral-ground connection should trip the upstream GFCI breaker or receptacle. {That may be the case with AFCIs, too.}

Testing for downstream neutral-ground connections can be done with gizmos intended for the purpose. Alternately—at the panelboard—branch circuits could be tested with an ohmmeter while lifting individual neutrals and checking for continuity. Also, at the receptacle a temporary hot-neutral load could be applied while looking for an at-the-plug increase in neutral-ground voltage. Too much N-G voltage may indicate an excessively long circuit. Too little voltage change may catch “cheating” though a locally shorted connection.

This is a test that is being performed at the request of the electrical inspector prior to energizing the system.
See the following additional comments.
When testing branch circuits to assure that there does not exist a connection between the neutral and ground conductors, what is the acceptable resistance allowed between the conductors?
What is the proper instrument to use for this measurement?
I have received different opinions from testers, inspectors and engineers.
In addition to your opinions, I am looking for documentation that will define the requirement for performing the test.

Many jurisdictions in Southern California require the following test to assure compliance with 250-24 (a) 5. At each panel-board the neutral feeder is removed from the neutral buss for that panel. An ohmmeter is placed across the neutral and ground busses. A reading of infinity is required to assure that no connection exists between the neutral and ground of the branch circuits. If necessary the neutrals are removed from the neutral buss and the neutral are tested one at a time.

‘ship — I don't think there is a magic number that defines what's good and what isn't. Have SoCal AHJs specified one? Again, a degree of common sense should apply, but sometimes numbers get interpreted/applied way past their original intent.

I have a personal example of how my bright ideas were well out of bounds. Years ago at a IAEI-chapter meeting, a presentation was delivered on flood restoration of residential housing in a low-lying area.

There was considerable discussion on what was considered acceptable and what wasn’t. Part way through the discussion, I asked why a megger was not used to simply check all the wring before power up, for that was my universal, kneejerk, looking-down-a-drainpipe response in an industrial setting of unit substations and medium-sized motors. Well, it turned out my comment was way out of place, for almost none of the older wiring would have stood up to the potentially extreme sensitivity of a 500V megger test. I ‘ate some crow’ that time.

"Highly situational" may be a fitting buzzphrase, here.




[This message has been edited by Bjarney (edited 07-23-2003).]