I have posted this before.
A quicky
For 600 VAC and lower systems
I perform these test at 1000 VDC after verifying no loads are connected. Time frames very with type of test, go-no go short conductor runs 30 seconds to a minute (when the meter reaches a stabilized reading) and or long runs of cable or busway, which can require 5-10 minutes. Minimum readings are published in NETA specifications. The often-quoted one (1) megohm is, in my opinion a mistake, 5 megohm would be a minimum under certain conditions (buried cable, emergency situations on 208/120 VAC systems). 50 megohms is a minimum we prefer with 100 megohms being the norm. BUT on new systems one would expect and hope to see full scale of the meter. At one time this full scale was typically 2000-4000 megohms, today that varies with the manufacture some meggers now read into gigohms and teraohms. One item to consider is comparison readings, such as a 3-wire circuit ?L1? reads 2000 megohms; ?L2? reads 200 megohms and ?L3? reads 2000 megohms. While not a major issue I would want to know why ?L2? was low. This would hold true for any type of circuit with similar conductors.
When Meggering it is important to verify all loads are turned off epically where there is sensitive electronics. You can damage this equipment or at a minimum get false readings.
Typically start with the lowest voltage your megger has, starting at 50, 100 or 250 VDC will minimize any chance of damaging equipment.
For trouble shooting single phase, I would megger Line to ground, line too neutral, neutral to ground. Typically when trouble shooting this is a go no go test allow megger reading to stabilize, no reason to wait one minute if the reading has stabilized.
For new houses with no loads on (and smokes and similar equipment disconnected]
And the neutral bond lifted, I megger the panel with the meter lifted and all CB?s off for one minute, ?L1? to neutral and ground, ?L2? to neutral and ground, Megger neutral to ground.
As for exposed copper at terminals, this does nothing to the readings unless there is a problem with the terminal, such as some conducting medium between the terminal and ground or another polarity.
We megger everything and document all readings:
All new installations at a standard time frame (minimum one minute). We do gear separate from feeders prior to termination.
All equipment serviced once again at one-minute minimum, we like to megger when de-energized and prior to re-energizing to verify we did not screw up. Plus if you get low readings with the first reading the customer MUST be aware of this, because human nature would be to blame you if you just megger at the end of service.
Houses after a lightning strike, with separate readings for the panel, then individual branch circuits, then the completed system.
Neutrals for all services worked on to determine if there are down stream shorts.
