"Occasional" 240v

[coopec]

Sarabande

The maximum wattage of any tool powered by the cigarette socket would be 15 X 12 = 180 watt. You could set up a separate 12V Power Outlet with heavier power cables and a 30 amp (?) fuse (Like I'm doing for my yacht cockpit to power spotlights etc)

If I am wrong I'm sure someone will correct me!

Clive

 

[PaulRainbow]

In my view RCD and earthing is less important and a GRP boat as the hull and deck is insulated anyway.

Its a different matter on my steel boat as if the hull becomes live it would kill some one.

I cannot understand why the mains earth would be connected to the battery negative but alot of people think it must be.

On my steel boat my earth is connected to the hull through a GI and my batteries are isolated from the hull to reduce any galvanic currents that could flow.

You GRP boat owners dont have to consider that much as everything is isolated from each other.

Part of the problem here is that you have not wired your own boat according to current regs, in that you have not bonded the AC ground wire to the DC negative, not that you are required to follow those regs.

The OP does not have a steel boat, or a house. If he connects the earth and negative in the plug it will make no difference to the operation of his RCD, because the "earth" wire doesn't go anywhere, he'll just be connecting any conductive casings, that would be grounded in a normal installation, to AC negative. All that does is create a short circuit if the positive touches a case and the fuse will blow. If, on the other hand, he connects the AC "earth" to DC negative the RCD will work as it is designed to do. As he has no shore power connection he doesn't increase the risk of galvanic corrosion and does not require a galvanic isolator. If he fitted shore power, he would need a galvanic isolator.

 

[sarabande]

Sarabande

The maximum wattage of any tool powered by the cigarette socket would be 15 X 12 = 180 watt. You could set up a separate 12V Power Outlet with heavier power cables and a 30 amp (?) fuse (Like I'm doing for my yacht cockpit to power spotlights etc)

If I am wrong I'm sure someone will correct me!

Clive

I've looked closely at cigarette sockets as a power source. 15A fused is the normal capacity, though there are stronger ones.
Providing heavier duty power cables and a 30A fuse, and deploying a 400 power drill is almost guaranteed to melt the internal circuitry of the plug unless it is one of the HD ones. Much better to use a proper two pin socket, examples of which can be found on the Bulgin website.

 

[matthewriches]

I made a quick and dirty video to demonstrate a test...

 

[sarabande]

Marvellous video, very valuable, many thanks for taking the time to put it all together. Makes me realise how little I really know !


I take on board the comment about "exercising the RCD". Very well proven.


I believe the new regs about Consumer unit cases being metal are now in force. Boats may in future need to spec aluminium ones.

 

[PaulRainbow]

I made a quick and dirty video to demonstrate a test...

If that was installed correctly on a boat, the AC earth wire would be connected to the DC negative. I'm not surprised it didn't trip in your test, because the earth wire doesn't go anywhere.

Could you try that again with AC earth bonded to DC negative ?

 

[superheat6k]

Following adverse suggestions on here I deleted my thread responses, but having considered the matter I felt this was a little trite, so here they are reinstated, obviously now out of context with the thread order, but for those who may be interested in having the entire thread available here is post 4. The others deleted will follow shortly.



If I were fitting a consumer unit rather than using say an extension lead occasionally then I would suggest the install should be done properly. Once done it is likely you would want to use shore powers supplies sometimes, so would have more than one source of power. As such then full pole separation of the individual power sources is essential, on both live and neutral e.g a 1 - Off - 2 two pole switch isolator (NB not a 12v battery isolator - must be rated for at least 400volts). Also the neutral from the Inverter should be grounded at the inverter, and before the isolator, ideally inside the inverter casing. Simpler place for the RCD is as a part of the consumer unit, and these are normally the size of two single pole MCBs, so a four MCB unit should provide all your needs.

Full supply isolation also means there is no risk of the supply plug from the inverter having a its prongs live, although I would do away with the 3 pin socket in favour of a more permanent fixed connection to the inverter.

Proper practice is to bond the earth to the boats bonding circuit, which is also normally connected to the -VE negative, but the bonding circuit should not be used as an active negative conductor. If you do then start to use shorepower more frequently then a galvanic isolator in the shorepower Earth line would reduce (but not eliminate) any likelihood of galvanic corrosion.

 

[superheat6k]

This followed post 13 ...

Indeed the RCD doesn't require an earth as it is a balance scale device :- Current in = Current out with a normal limit of 30 ma. If the current does not balance it defect means it has found another route of return and trips.

This requires the neutral must be connected to earth to provide the return path. Without it might as well leave the RCD in its box.

If the neutral is not bonded to earth somehow on what is a battery sourced supply and a fault develops there is a risk of electrical short circuit and a remote possibility of a shock from the system should a second fault develop somewhere else.

For example a metal cased tool or device develops an internal short to the case on one of the supply cables, live or neutral. The case will now be at live or neutral potential, and because the supply is not tied to earth either could sit anywhere between earth and 240 v potential.

Now touching the faulty tool would not cause an electric shock, but then another device on the system, perhaps a damaged extension lead cable for instance, also develops a fault, this time between the non already fault connected conductor and the protective earth inside the dodgy lead.

A full short circuit has now developed, but because the RCD is still seeing a balanced current flow then it will not operate. At best the inverter will trip, but a 600 w inverter could easily send enough power through a partial short circuit to cause a fire.

Bonding the neutral to the earth circuit provides a return path for either conductor foreign a fault circuit, allowing the RCD to operate as desired.

 

[superheat6k]

This is the final post deleted, with which at least one forumite disagrees, although in my view and practically what I have said here is perfectly valid, although not completely correct from a purist viewpoint. It does adequately explain the

From a prior post - "Neutral is generally NOT GROUND in UK domestic power distribution.
There is current flowing in the neutral wire from the transformer to the house. This wire has both resistance and inductance. Therefore the two ends are at different potentials."

But the neutral is grounded at the source on UK domestic supplies - normally at the star point of the local Delta Star step down transformer at the substation. The star point provides the neutral between the three distribution phases. It is earthed to the transformer case and normally by some form of substantial ground plate or embedded spike. Effectively the local transformer at the substation is the supply source in a domestic situation. The neutral cable itself is just a wire - it will not have any inductance within and of itself, although by definition the power flowing will having an element of inductance from the induction coils that form the transformer. In regard to grounding this is irrelevant because the grounding is after the inductive part of the circuit. Inductive loads on domestic supplies are fairly negligible.

So although current will run through the neutral, and could be slightly out of phase with the voltage (power factor likely to be better than 0.98), the supply is as far as is feasible balanced between the three distribution phases to the various premises being supplied from any particular transformer, with the ideal being that no current at all runs through the neutral star point at the transformer, although in practice some imbalance will almost aways exist. Even on a long run from the transformer, due to the size of the distribution cables the neutral potential (voltage) will be fairly close to earth potential. The Americans refer to the Neutral as the Grounded conductor, with the Protective Earth as the Grounding conductor.

It is simple to measure the voltage between neutral and earth with any decent multimeter, and invariably you will see 0 volts.


I leave it to others to decide whether the previously deleted posts are nonsense, as one forumite has suggested.

 

[Lon nan Gruagach]

I made a quick and dirty video to demonstrate a test...

If the RCD didn't trip with your outgoing live to earth 30mA test then the inverter is not a centre tapped output. It would be more accurate to call it a floating output.

 

[Lon nan Gruagach]

This followed post 13 ...

Indeed the RCD doesn't require an earth as it is a balance scale device :- Current in = Current out with a normal limit of 30 ma. If the current does not balance it defect means it has found another route of return and trips.

This requires the neutral must be connected to earth to provide the return path. Without it might as well leave the RCD in its box.

If the neutral is not bonded to earth somehow on what is a battery sourced supply and a fault develops there is a risk of electrical short circuit and a remote possibility of a shock from the system should a second fault develop somewhere else.

For example a metal cased tool or device develops an internal short to the case on one of the supply cables, live or neutral. The case will now be at live or neutral potential, and because the supply is not tied to earth either could sit anywhere between earth and 240 v potential.

Now touching the faulty tool would not cause an electric shock, but then another device on the system, perhaps a damaged extension lead cable for instance, also develops a fault, this time between the non already fault connected conductor and the protective earth inside the dodgy lead.

A full short circuit has now developed, but because the RCD is still seeing a balanced current flow then it will not operate. At best the inverter will trip, but a 600 w inverter could easily send enough power through a partial short circuit to cause a fire.

Bonding the neutral to the earth circuit provides a return path for either conductor foreign a fault circuit, allowing the RCD to operate as desired.

Herewith is a real life happenstance that, I think, will demonstrate how no earth bonding makes life go pear shaped.

I was called to a neighbour's house, they had just had a nasty zap while stoking the wood fuel stove. The shock was between a wet floor and the stove.
A visual inspection showed that the pipes to the back boiler were properly bonded to the earth from the mains supply.
Prognosis: the wet floor was live..... Not really possible, is it.
Eventually the fault, rather the faults since there were more than one. Were :
The domestic earth wasn't connected to any thing outside the house.
The kettle had developed. Live to earth fault.
Result was the domestic protective earth was in fact live.

 

[VicS]

I made a quick and dirty video to demonstrate a test...

I don't understand why the RCD did not trip when you connected your tester between RCD live out and inverter (centre tap) earth

Surely that is causing an imbalance . The test current flowing through the live but not through the neutral connection.

Because the inverter has a centre tap earth there is actually no difference between the so called live and neutral , both are in fact live at approx 100volts with respect to the "earth" as you demonstrated with the preliminary voltage measurements, I would expect the RCD to trip when the tester is connected between so called neutral and the earth connection. This time because current would be flowing in the "neutral" connection, but not the live connection, creating an imbalance.

Why then did the RCD not trip when the tester was connected between live output and centre tap earth. You seemed surprised yourself in the video.

 

[PaulRainbow]

If the RCD didn't trip with your outgoing live to earth 30mA test then the inverter is not a centre tapped output. It would be more accurate to call it a floating output.

You forget, in the test Matt did the "earth" connection doesn't go anywhere, it's not an earth, it doesn't go any further than the busbar in the consumer unit.

Most of you seem to be overlooking the simple fact that the AC "ground" connections need to be bonded to the DC negative.

 

[VicS]

You forget, in the test Matt did the "earth" connection doesn't go anywhere, it's not an earth, it doesn't go any further than the busbar in the consumer unit.

Most of you seem to be overlooking the simple fact that the AC "ground" connections need to be bonded to the DC negative.

He expalined that the " earth" connection was a centre tap. It would not have made any difference to his demonstrations if he had also connected it to an earthing stake in the ground


You don't bond the AC ground to the DC negative if the DC system is a fully insulated two wire system.

 

[BabaYaga]

the simple fact that the AC "ground" connections need to be bonded to the DC negative.

The requirement, as I read ISO13297, is not to connect AC protective earth to DC negative per se, but to "the main grounding/earthing point of the craft".
In many cases the negative terminal on the engine block is used for this purpose, in which case both DC negative and AC protective earth are joined at this point in order to connect to "ground" (the water). But the DC system might also be of the fully insulated two-wire type, that is, no connection to ground/earth.

 

[coopec]

Sarabande

Providing heavier duty power cables and a 30A fuse, and deploying a 400 power drill is almost guaranteed to melt the internal circuitry of the plug unless it is one of the HD ones.

I wasn't going to be so patronizing to mention that! But there are plenty of 12V power outlets that will handle 40A

 

[matthewriches]

So far we have shown that touching live and neutral at the same time would not trip the RCD. We knew that, as expected, as confirmed by superheat6k and rogershaw.... that's just how they work.

If the inverter is a centre tap, as it appears, you absolutely wouldn't connect a neutral - earth link as you would be shorting half the system! Very bad idea. If it is floating with caps causing the look of a centre tapped earth, then bonding the Earth and Neutral would be of benefit. The DC and AC appears isolated on this inverter, testing at over 60Mohm. There is of course a chance that the system has a phantom centre tap caused by a load of charged capacitors. Remember, I'll be testing as many as I can find.

I also have a plug in RCD to prove further that an imbalance is required. We'll also try a dead short to see what happens. Of course the MCB should blow.

The earth connection from the consumer unit busbar goes through the flex into the inverter and into the circuit board. On the live and neutral connections, we should be using double pole MCBs also.

 

[Interlude]

I seem to remember from the video that Active to Neutral was >220 V or so, while Active to "Ground" and Neutral to "Ground" were both less than 100 V and different (~90 and ~100?), indicating that the inverter output is floating, and it is stray capacitance between the A, N and G lines that his causing the readings. It not centre tapped, as that would result in equal A to G, N to G each being half of A to N. I would be bonding N to G, ideally so it can't be removed by unplugging something.

 

[solent clown]

the inverter I have has an external ground pole on it's chassis. I was planning to use that to bond ac/dc, and to run the neutral to for the rcd trip. I had taken for granted that would work with this unit, but clearly I need to investigate it's specification first.
Of course I could just keep things simple, and not bother with the rcds, but this is developing into an interesting thread, especially with Matts vids, and all the other inputg from various knowledgable chaps, so I will also log our attempts and results.

 

[matthewriches]

This may appear to be the case based on the readings. Of course, the test will be to have another play, bond the neutral to earth to see what happens. And try a few other inverters to see what they produce too.

I seem to remember from the video that Active to Neutral was >220 V or so, while Active to "Ground" and Neutral to "Ground" were both less than 100 V and different (~90 and ~100?), indicating that the inverter output is floating, and it is stray capacitance between the A, N and G lines that his causing the readings. It not centre tapped, as that would result in equal A to G, N to G each being half of A to N. I would be bonding N to G, ideally so it can't be removed by unplugging something.