DC + 18ms and DC - 7.5ms. So all passed with flying colours, trip times about 20ms at 30 mA all round. 10 MA of DC changed the AC 180 to 21 mA and DC+ to 12 mA

Are these ramp tests, results in mA RMS at time to trip, or times in milliseconds ?

Either could make sense in the context but imply subtly different things.


In  any case I am heartened that you too see the blinding current is comparable to the trip current , as that backs up my observations - so 20mA may pull a 30mA RCD slightly off beam on a 30mA test, but it will still fire at a fault current of 100mA, and still needs a jammer comparable to the higher test current to blind it.


The failure to trip on a supply with a PSSC of <16 amps, tells me this is an electronic RCD and internally there is a fairly chonky actuator coil pulled in by a power transistor or triac, not a delicate hairspring loaded mechanical balance. The actuator current needed to fire ones of this sort can be several amps, and the internals are not as fine or as fussy as the traditional passives.


Please take care not to burger what looks like a decent PSU, - note that at the moment of opening, you do not know if the L or N arc extinguishes first, so there is a risk of a mains voltage or part of it appearing on the PSU for a short time if the contact with the PSU accross is the one that opens first.

Also,  even with the volts low, you are shunting the N sense coil with the unknown output impedance of the PSU - which should be quite low and may give some imbalance to begin with.


If there is a known series resistance, say 100 ohms,  (even an old kettle element or something would do) then the fault current if subjected to mains is limited, and the current share between coil and PSU is very much in favour of the coil at a few milli-ohm versus the PSU and resistor combo.


I have also spoofed an LNE with a fake E and a very local NE bond for such tests, from an L and N only - so upstream RCDs are not affected. But of course only after double checking polarity is right.


Perhaps I too need to take a pic, but my set  up is, shall we say, rather more agricultural, and I'd want to tidy up first, or folk will chitter about the exposed live bits.

DC + 18ms and DC - 7.5ms. So all passed with flying colours, trip times about 20ms at 30 mA all round. 10 MA of DC changed the AC 180 to 21 mA and DC+ to 12 mA

Are these ramp tests, results in mA RMS at time to trip, or times in milliseconds ?

Either could make sense in the context but imply subtly different things.


In  any case I am heartened that you too see the blinding current is comparable to the trip current , as that backs up my observations - so 20mA may pull a 30mA RCD slightly off beam on a 30mA test, but it will still fire at a fault current of 100mA, and still needs a jammer comparable to the higher test current to blind it.


The failure to trip on a supply with a PSSC of <16 amps, tells me this is an electronic RCD and internally there is a fairly chonky actuator coil pulled in by a power transistor or triac, not a delicate hairspring loaded mechanical balance. The actuator current needed to fire ones of this sort can be several amps, and the internals are not as fine or as fussy as the traditional passives.


Please take care not to burger what looks like a decent PSU, - note that at the moment of opening, you do not know if the L or N arc extinguishes first, so there is a risk of a mains voltage or part of it appearing on the PSU for a short time if the contact with the PSU accross is the one that opens first.

Also,  even with the volts low, you are shunting the N sense coil with the unknown output impedance of the PSU - which should be quite low and may give some imbalance to begin with.


If there is a known series resistance, say 100 ohms,  (even an old kettle element or something would do) then the fault current if subjected to mains is limited, and the current share between coil and PSU is very much in favour of the coil at a few milli-ohm versus the PSU and resistor combo.


I have also spoofed an LNE with a fake E and a very local NE bond for such tests, from an L and N only - so upstream RCDs are not affected. But of course only after double checking polarity is right.


Perhaps I too need to take a pic, but my set  up is, shall we say, rather more agricultural, and I'd want to tidy up first, or folk will chitter about the exposed live bits.