Been back in the trade a few years now & a regular lurker here (236 I & II & C, 2391, time served etc, 17th) - I have always managed to track down answers using the search however need some help on this one.
New kitchen install in a 2 bed bungalow involving a board swap to RCBO's as only three circuits (lights, ring & cooker).
2nd fix day went fine, all tests carried out with satisfactory results - cooker hood flicked off & on to check function before leaving - job done.
Call via kitchen company I'm subbing for that RCBO is tripping out when fan speed increased on cooker fault. Go back & double check IR tests (all over 200 MOhms), test RCBO again (X1 19.7 X5 7.8 ramped at 27 mS).
Earth leakage clamped & when fan speed increased see a jump to 70 mA - aha I think faulty cooker hood.
Replacement arrives but fitters call to say same thing happening.
Go back again & found that if I connect hood to short leg of ring back to board it trips - connect it to other longer side of ring it doesn't! So remove all other circuits from board & same thing happens. Then swap over to similar length cooker circuit (2M) & it trips! Connect to short leg via extension lead it doesn't trip. This all after double checking IR etc. Swapped RCBO's , double checked connections in consumer unit.
I have installed numerous kitchens recently & never seen this fault. I realise that the common factor here is the short length of the circuit but can't work out why a common Bosch cooker hood would generate 70mA leakage because of it!
By Marconi This is what I think is happening based on the information provided. I have highlighted above the key statements which I have considered. I have to make some assumptions, for example I reckon the fan speed is adjusted by switching in/out capacitors which are connected in series with the motor: The sequence of switching is OFF/SLOW/MEDIUM/FAST corresponding to a progressive reduction in series capacitance. I suspect also that the cooker hood socket may be the first socket in the short 2m leg and the supply is TNC-S although the reasoning holds for TN-S. Studying my home's cooker hood, the switch operation is a slider mechanism: I reckon the contacts are break before make but not fast in opening and closing so arcing is possible/likely but short lived.
When the fan is turned to the SLOW position (- as in the quick test of operation mentioned above) the transient current through the path L-N is balanced thus the RCBO does not operate. The motor is spinning and current is flowing through its windings (which are inductive). Energy is stored in the magnetic field of the windings and there is a back EMF close to supply voltage less that dropped across the capacitor and any resistance. This is a steady state of operation. All along the windings of the motor there is capacitance to the earthed metalwork of the armature: at mains voltage the earth leakage is small and much less than would trip the RCBO.
When the fan is switched up in speed to MEDIUM, the situation is once again transient as the motor speeds up to a higher steady state speed. The switch opens before closing the contact to the next smaller value series capacitor. In this short interval, the applied EMF of the mains is momentarily removed which is driving the current through the motor and creating the magnetic field. The magnetic field starts to collapse thereby inducing an EMF in the winding of the motor which acts to maintain the flow of current L-N. This EMF is in series with the supply EMF: a large voltage across the contacts causes an arc to form, which is drawn as the contacts separate and extinguished. The windings EMF is now effectively connected across the capacitance of the winding to earthed metalwork. The sudden step change in potential difference across this distributed capacitance resulting from the opening of the switch - is now the EMF induced by the collapsing magnetic field, which is much higher than the applied mains (and corresponding back EMF of steady state operation). This step change has frequency components much higher than 50Hz so even though the distributed capacitance my not be large, at these frequencies the capacitive reactance is low - large high frequency current (ie 70mA) flow in the series loop formed by winding, distributed capacitance, earthed frame, CPC, CPC-N connection (Local if TNC-S or remote if TN-S), N back through RCBO and back to winding. There is thus an unbalanced current passing through the RCBO - no similar current in L through RCBO- which it detects and causes it operate.
When the contact eventually makes for MEDIUM, the mains supply is reconnected to the motor windings which causes a further transient, again in voltage across the winding and its distributed capacitance and there is once again current flow through the winding capacitance but now through a series path formed by L, winding, distributed capacitance, earthed frame, CPC, CPC-N connection - no similar current through N of the RCBO this time. Making the contacts also establishes again an L-N loop current through the windings which restores a new steady state at higher speed.
So why does the length of cable between the fan and the RCBO matter? The first thing to realise is that Twin and Earth Cable has different electrical characteristics depending on frequency. All quoted conductances and insulation resistances for example are at DC and low frequencies. At higher frequencies something called skin effect matters; the tendency of the current to flow only in the outer perimeter of the conductor which means at higher frequencies the resistance is higher (conductance is lower). It is why lightning conductors are generally broad tapes of copper (or hollow copper tubes) to create the maximum perimeter area for conductance of the lightning pulse. Thus a longer length of T&E will, at high frequencies, have a higher overall resistance end-to-end.
Next, the way the conductors are arranged in T&E, L-CPC-N, means there is more capacitance between the live conductors and CPC than between the live conductors. A capacitor relies on an insulator between to conducting plates - a dialectric material. A dialectric material is not a perfect insulator hence insulation resistance. The insulation resistance of a long length of T&E is lower than for a short length and dialectrics become weaker as frequency increases thus capacitance decreases.
Earlier I explained how I think in moving from SLOW to MEDIUM SPEED a short pulse of current flows in an N-CPC loop via the RCBO when the switch opens and then another short pulse of current via a L-CPC loop when the switch closes.
These currents are of course caused by short pulses of voltage, the former applied within the N-CPC loop and the latter applied within the L-CPC loop.
For the short length of T&E, there is not as much skin resistance and the impedance between its N conductor and CPC is higher so the pulse of voltage is effectively applied across the N-CPC connection via the N of the RCBO. Along the longer length of T&E there is more skin resistance and lower impedance between the N and CPC which means when the fan is connected only to the longer length of T&E, insufficient is present across the N-CPC connection via the RCBO N path for it to operate - the major part of the current in the loop flows between the conductors of the T&E. Similar reasoning can explain a higher current in the L-CPC loop for the short length of T&E.
A further thought on why the longer length attenuates the pulses of voltage is there are almost certainly more sockets in the longer length of ring with loads connected which will attenuate further the pulses propagating back towards the RCBO when the fan switch is operated SLOW-MEDIUM.
These are my musings on this head-scratcher. Without testing and recording some results for analysis that is all they can be.
