Both measurements are of course not really applying a short circuit  but rather a change in load that creates a small but measurable voltage drop, from which the current that might have flowed if the full 230V or whatever had been present,

The anti-trip passes less than 30mA RMS (or it would trip) and does a complex switching pattern to try and eliminate being fooled by system voltage changes during the test.

The traditional meter uses a much higher current - ten amps up to about 25 are common in testers for 32A circuits so the voltage drop from the test is much larger and the magnitude of any mis-read so much less.

However, 100:1 is well beyond this - so there are a few possible explanations.

1) a corroded contact that was high Z during the ATT testing, but then blasted clear and 'welded shut' by the higher test current

2) an RCD or RCBO coil - less likely as then the tests on the  non- ATT setting should have been cut short by it tripping. The coll inductance cancels for L-N loop (if it did not the RCD would trip) but not on an L-E loop  Certain ATT pulsed waveforms are high frequency and can generate a significant back EMF in that series inductor, yet to the 50Hz the inductance is not really significant, This can fool some staff into reporitng faults when there is not one. Certainly an un balanced final  circuit is a bad idea and  can fool certain types of test kit.

3) There is noise or waveform distortion that fools the meter, and another meter or even the same on another day, may read very differently.

Mike

Both measurements are of course not really applying a short circuit  but rather a change in load that creates a small but measurable voltage drop, from which the current that might have flowed if the full 230V or whatever had been present,

The anti-trip passes less than 30mA RMS (or it would trip) and does a complex switching pattern to try and eliminate being fooled by system voltage changes during the test.

The traditional meter uses a much higher current - ten amps up to about 25 are common in testers for 32A circuits so the voltage drop from the test is much larger and the magnitude of any mis-read so much less.

However, 100:1 is well beyond this - so there are a few possible explanations.

1) a corroded contact that was high Z during the ATT testing, but then blasted clear and 'welded shut' by the higher test current

2) an RCD or RCBO coil - less likely as then the tests on the  non- ATT setting should have been cut short by it tripping. The coll inductance cancels for L-N loop (if it did not the RCD would trip) but not on an L-E loop  Certain ATT pulsed waveforms are high frequency and can generate a significant back EMF in that series inductor, yet to the 50Hz the inductance is not really significant, This can fool some staff into reporitng faults when there is not one. Certainly an un balanced final  circuit is a bad idea and  can fool certain types of test kit.

3) There is noise or waveform distortion that fools the meter, and another meter or even the same on another day, may read very differently.

Mike