Can this device be installed in an installation and comply with B.S. 7671?

There is no reason why you should not use such a technique to defeat additional protection for part of an installation downstream of an RCD if you wish to do so - in effect the system injects current in the NE loop, to match and cancel the current in the L-E loop. It is not much more complex than an audio power amplifier and a sense coil. The power amplifier part only needs to deal with a few volts, maybe 10s at most, and current equal to those to be cancelled.


As the problem with EMC filters is that they allow L-E current to flow, injecting counter current  in this way is a good way of reducing currents in RCDs and earth rods when no fault is present

Would you put one of these filters

Lindgren 6000 series fixed installation filters

Leakage Current at 250 VAC 50 Hz​  <10% of Rated Current

In a TT system without something to tame the earth leakage ? - the leakage on a 60A filter is not to be sneezed at.

Even these N900 series manage to leak the best part of an amp to earth in normal use.

  The usual wheeze is to trim the capacitances from the 3 phases, so that the resultant leakage is near zero, but this active approach should allow results that are as almost  as good on a single phase system, or a system where only 2 phases are used.


The key thing however is to be very clear about how the part that is not under RCD protection is to be kept safe -  if it is all in the one box, and the effect is that a piece of closed class 1 equipment has a low earth leakage, no one will question how this is achieved, but this technique allows the wiring that is after the active current cancellation to be part of the fixed wiring, and that is I think a novel step.

I'd want to see the bulk of the installation being low leakage and RCD protected, so the magic current injection box as close as possible to the filters, and those in turn close to the inverter - and all of that, I'd like to be enclosed,  such that additional protection by RCD is not really needed, as it may not have it.


I agree if you do not allow cancellation at 50Hz, then the load side circuitry is still RCD projected to a degree, but then the dominant leakage current is not being ameliorated either, which sort of defeats the best point of the technique somewhat.

There is no reason why you should not use such a technique to defeat additional protection for part of an installation downstream of an RCD if you wish to do so - in effect the system injects current in the NE loop, to match and cancel the current in the L-E loop. It is not much more complex than an audio power amplifier and a sense coil. The power amplifier part only needs to deal with a few volts, maybe 10s at most, and current equal to those to be cancelled.


As the problem with EMC filters is that they allow L-E current to flow, injecting counter current  in this way is a good way of reducing currents in RCDs and earth rods when no fault is present

Would you put one of these filters

Lindgren 6000 series fixed installation filters

Leakage Current at 250 VAC 50 Hz​  <10% of Rated Current

In a TT system without something to tame the earth leakage ? - the leakage on a 60A filter is not to be sneezed at.

Even these N900 series manage to leak the best part of an amp to earth in normal use.

  The usual wheeze is to trim the capacitances from the 3 phases, so that the resultant leakage is near zero, but this active approach should allow results that are as almost  as good on a single phase system, or a system where only 2 phases are used.


The key thing however is to be very clear about how the part that is not under RCD protection is to be kept safe -  if it is all in the one box, and the effect is that a piece of closed class 1 equipment has a low earth leakage, no one will question how this is achieved, but this technique allows the wiring that is after the active current cancellation to be part of the fixed wiring, and that is I think a novel step.

I'd want to see the bulk of the installation being low leakage and RCD protected, so the magic current injection box as close as possible to the filters, and those in turn close to the inverter - and all of that, I'd like to be enclosed,  such that additional protection by RCD is not really needed, as it may not have it.


I agree if you do not allow cancellation at 50Hz, then the load side circuitry is still RCD projected to a degree, but then the dominant leakage current is not being ameliorated either, which sort of defeats the best point of the technique somewhat.