Motor Over Load Testing

http://ecmweb.com/mag/electric_nema_standards_overload/
http://www05.abb.com/global/scot/scot209.nsf/veritydisplay/41ba306d7b8fca1fc1257110002ea116/$file/2cdc106015b0201.pdf

thanks a lot for that helpful data, some testing procedure is to inject the 150%(for example) directly till the device trips.

some other procedure is to inject 100% for 1min then stop for 10 min then test the 150%.

the last method was not for ABB but it gives more accurate results.

The thermal overload relay (T25DU) curve goes from 1.05 x In to about 10 x In. You can use any value to test it. Not necessarily 1.50 x In. The lower the multiple, longer the time to trip,the more accurate the test.

ABB's superb automated assembly line at Heidelberg has so many quality checks on the way that they have >90% first-pass-yield. The line calibrates the relay at 2 x In. The curve shape having been set by the quality checks, one test point is enough.

You, as a user, should test at two points if you want to test at all. 1.5 x as you are doing, and 6 x In to simulate a locked-rotor condition.

so what is the procedure for performing a test on the hot curve

Feed 1.00 x In for 2 hours, then raise the test current to the desired value. This is hot curve testing.

One small thing ...remember, you must use a stabilised current source, single-phase, all three phases of the relay in series. As the relay heats up, its heater resistances increase, so if the current is not automatically increased to compensate, you will get erroneous results.

is this method recommended by ABB?

what about after rest from cold test, start the hot test

is this method recommended by ABB?

Not necessarily by ABB, this is common sense, and is used by all testing authorities.

what about after rest from cold test, start the hot test

i don't understand. The two hours of passing In is to heat it to the hot state. That can't be avoided.