RCD tripping problem

[Porque223]

The load current magnitude is not supposed to influence tripping but at some high levels it probably does. This effect probably appears as a "rejection ratio" spec on the datasheets for the ICs used in these things.

 

[JohnW2]

The load current magnitude is not supposed to influence tripping but at some high levels it probably does. This effect probably appears as a "rejection ratio" spec on the datasheets for the ICs used in these things.

I doubt that the ICs have anything to do with it - it sounds as if you are talking about a situation in which the L-N current difference was being determined by some sort of comparator or operational amplifier. In an RCD, the 'difference current' is determined electromagnetically by the sensing toroid, and all the electronics know about is that sensed current difference. However, it might well be the case that some sort of electromagnetic saturation effects come into play at very high load currents - so that might affect sensing under high load current conditions (as will arise if the reason for the current imbalance is a low impedance L-E fault).

Kind Regards, John

 

[bernardgreen]

There will be a very small difference in phase between load current in the neutral and the same current in the live. The phase difference is a result of stray capacity to ground in the wiring. This difference will create a very small, insignificant, output from the sense coil in the RCD even when there is no earth leakage. With very high load currents this phase difference output from the sense coil may become large enough to affect the set point operation of the RCD.

 

[JohnW2]

There will be a very small difference in phase between load current in the neutral and the same current in the live. The phase difference is a result of stray capacity to ground in the wiring. This difference will create a very small, insignificant, output from the sense coil in the RCD even when there is no earth leakage. With very high load currents this phase difference output from the sense coil may become large enough to affect the set point operation of the RCD.

An interesting theoretical argument, but do you actually think that such a phenomenon would have any appreciable effect of the 'trip point' even at very high currents? In which direction are you postulating this effect would work?

Kind Regards, John

 

[Porque223]

If you have a spare, functioning RCD almost certainly you can coax behavior out of it that is "not as advertised."

If you can get a pinout for the internal IC you can tell how close it is to tripping by measuring a voltage on a pin.
Here's the only schematic I have
http://www.ti.com/lit/ds/snis158/snis158.pdf
but it's probably obsolete.

 

[bernardgreen]

do you actually think that such a phenomenon would have any appreciable effect of the 'trip point'

The effect would be insignificant and orders of magnitude less than over effects of tolerances in manufacture of the RCD.

even at very high currents? In which direction are you postulating this effect would work?

No comment. Saturday is not the day for detailed phase calculations.

 

[JohnW2]

do you actually think that such a phenomenon would have any appreciable effect of the 'trip point'

The effect would be insignificant and orders of magnitude less than over effects of tolerances in manufacture of the RCD.

That's what I suspected.

In which direction are you postulating this effect would work?

No comment. Saturday is not the day for detailed phase calculations.

That's exactly how I felt yesterday about Friday evening, so I hoped that you might already have done it!!

Kind Regards, John

 

[JohnW2]

If you have a spare, functioning RCD almost certainly you can coax behavior out of it that is "not as advertised." If you can get a pinout for the internal IC you can tell how close it is to tripping by measuring a voltage on a pin.

Yes, you could do that (e.g. measuring between pins 2 & 3 of the IC you mention below), but all you would essentially be doing would be measuring the output of the sense coil - and you'd have to do further work to determine what your answer meant in terms of 'how close to tripping' it was. The IC is not really doing anything sophisticated - it's simply acting as a threshold trigger for an SCR, hence the circuit breaker coil. I'm not actually sure how any of that would help (other than maybe being of academic interest). What matters is the overall function of the unit as a whole, which can be assessed with a suitable test rig on an intact and unopened RCD.

Here's the only schematic I have
http://www.ti.com/lit/ds/snis158/snis158.pdf
but it's probably obsolete.

That's interesting - although, as you say, it's over 20 years old. As above, there's nothing really unexpected about the IC itself, but the 'ground/neutral coil' shown in the Typical Application is interesting. As far as I can make out, the idea of that is to induce current into any path which arises due to a neutral-earth fault, thereby allowing the RCD to trip immediately in response to a N-E fault, even if there is no load on the RCD, or something along those lines. Modern RCDs (at least those used in the UK) do not appear to have that facility, so they will often only trip in response to a N-E fault if there is some load current flowing through the RCD.

Kind Regards, John