Milling Machine trips 100 mA RCD

[BQW]

A neighbour of mine has a workshop attached to his Bungalow that contains a Milling machine.
This are powered from a single to 3 phase rotary inverter.
The Bungalow is feed by a TT system with the normal 100mA RCD next to the CU thats fitted with cartridge fuses.
The workshop has its own CU with a 30mA RCD, feed from the house, the 30mA RCD and has been tested and passed with a RCD tester. I did notice however the workshop did not have its own earth rod.
When powering up the the Milling Machine from the inverter it trips the 100mA RCD, at the main CU, not the 30mA one in the workshop CU
It`s then a very long story of intense fault finding, with no evidence of obvious faults or earth leakage issues so it would appear their may be some incompatibility issue with the Inverter and the 100mA RCD.
My question is,
1. Has any one come across a RCD incompatibility problem with rotary inverters before .
2. The 100mA RCD at the CU has not been professionally tested, I assume this is the property of the electricity supplier, would they charge to test and replace it.
All information appreciated.

 

[bernardgreen]

. The 100mA RCD at the CU

next to the CU thats fitted with cartridge fuses

How old is this device ?

Is it possibly a VOELCB and not an RCD

An RCD operates by a difference between the Live and the Neutral currents at the RCD, the voltage on the Earth wiring is not taken into account.

A VOELCB break the supply when the voltage on the Earth wiring in the house is more than 50 volts above the potential on the ground rod.

When inductive loads such as motors are involved VOELCBs the voltage on the Earth wiring can bounce and this can trip the voltage operated devices.

VOELCB have two "Earth" connections, one (marked E ) connects to the Ground Rod and the other ( often marked F ) connects to the Earth wiring in the house.

 

[BQW]

I think you may well be be on the right track, it is a old instal. I will go back and have a better look, two earth connections seems familiar.
Thanks, really appreciated.

 

[JohnW2]

When powering up the the Milling Machine from the inverter it trips the 100mA RCD, at the main CU, not the 30mA one in the workshop CU

If it really is a 100mA RCD, and if what you describe is happening repeatedly, then that would tend to pont a finger at that "100 mA RCD".

However, as bernard has said, if it is actually a VOELCB, then that rather changes things - so we look forward to hearing confirmation of what it is! If it is labelled/described as "100 mA", then it must be an RCD, not a VOELCB.

Having said that, even if it were a VOELCB, and given that we're talking about an inductive load, I'm not sure why it should trip in the absence of earth leakage (which would also trip the 30 mA RCD), unless it were faulty, since there should be no significant current in the earth wires. In other words, I don't quite understand what bernard means by "the voltage on the Earth wiring can bounce". The the absence of earth leakage, the only voltage which should appear on earth wiring would be due to 'incidental coupling' (inductive/capacative), and that would be not enough to trip any device.

Has this problem always existed with the milling machine, or has it only recently started happening?

Kind Regards, John

 

[bernardgreen]

since there should be no significant current in the earth wires.

If it is a VOLCB then the ""Earth"" wire ( CPC ) in the building which one assumes is connected to the frame of the invertor will be pulled towards the mid point of Live-Neutral ( ~115 volts ) by capacitors in the filtering of the invertor mains supply. If that capacitance passes enough current onto the CPC to create a 50 V potential different across the coil of the VOELCB then the VOELCB will operate,

"the voltage on the Earth wiring can bounce"

The CPC ( Earth wiring in the house ) is not connected directly to Ground but instead it is connected via the coil of the VOELCB and hence can be pulled away from Ground

 

[JohnW2]

... will be pulled towards the mid point of Live-Neutral ( ~115 volts ) by capacitors in the filtering of the invertor mains supply. If that capacitance passes enough current onto the CPC to create a 50 V potential different across the coil of the VOELCB then the VOELCB will operate ...

Fair enough - you are proposing a possible mechanism for the 'earth leakage' which I said would be necessary.

However, as I said, no matter what the mechanism of that 'earth leakage' (filter capacitors or whatever), if it is enough to trip a VOELCB (typically about 250 mA - 50V across a typical {if I recall correctly} ~200Ω coil), then the 30 mA RCD in the circuit certainly ought to trip (as well as, or instead of, the VOELCB).

Kind Regards, John

 

[bernardgreen]

Given the use of an invertor then something along these lines is probably happening,

The transformer / inverter between mains via RCD and the machinery will prevent any Earth leakage affecting ( tripping ) the RCD

 

[JohnW2]

Given the use of an invertor then something along these lines is probably happening, ... The transformer / inverter between mains via RCD and the machinery will prevent any Earth leakage affecting ( tripping ) the RCD

OK, now that I've pointed out that leakage to earth through mains filter capacitors adequate to trip the VOELCB would almost certainly trip the 30 mA RCD, you've moved to a completely different hypothesised mechanism, which is fair enough. However, as I wrote ...

(in) the absence of earth leakage, the only voltage which should appear on earth wiring would be due to 'incidental coupling' (inductive/capacative), and that would be not enough to trip any device.

Do you really believe that the impedance of the path of "Induced voltages between frame and stray paths to ground" could possibly be enough to put ~250 mA through the VOELB coil?

Kind Regards, John

 

[bernardgreen]

you've moved to a completely different hypothesised mechanism, which is fair enough.

I do not think I have moved to a different mechanism.

Yes I do believe the induced voltages could be enough to operate the VOELCB

 

[JohnW2]

I do not think I have moved to a different mechanism.

All I can say is that it sounds very different to me. You were originally talking about 'leakage' to earth through mains filter capacitors, which would equally affect the RCD and VOELCB. You are now talking about induced voltages in circuitry which is isolated from the RCD.

Yes I do believe the induced voltages could be enough to operate the VOELCB

I'm very surprised by that statement. You seem to be implying that if the frame of the machine were not earthed, then touching it (and earth) could well have fatal consequences - if the 'induced voltages' could put 250mA through a 100Ω VOELCB coil, they could presumably put 50 mA through a "1kΩ human being". Is that what you believe?

Induced voltages can certainly result in 'tingles', but I've never heard of them producing significant shocks, let alone fatal ones.

Kind Regards, John