MG4 Discharge Cable

[Martinonline]

@Everest, I can follow all your recommendations and agree your full safe solution. I would use a V2L carefully.

However, in respect of earthing arrangements: surely if there is no earth return then anyone touching a live appliance would not be offering a circuit to earth (hence why the RCD does not work). However, if there is an earth return the RCD should trip. There is a 'but' I think in that when the person lets go of the live appliance they could have a residual static charge which will need to be discharged.

Please correct me if I have misunderstood the situation.

I assume the MG4 has overload protection so appliance circuit shorting will trip the supply. Otherwise an RCBO is advisable.

 

[Everest]

However, in respect of earthing arrangements: surely if there is no earth return then anyone touching a live appliance would not be offering a circuit to earth (hence why the RCD does not work).

For a single appliance there is no issue. A fault in appliance A might make the metal case live from the inverter. You are safe touching the appliance since no current will flow. But you don't know it's 'live' and, as there is no ground path, the RCD won't trip.

But if you connect appliance B and appliance B develops a fault, this time from the inverter's Neutral and you touch appliance B, no problem, as above.

But touch A and B at the same time you get full 230V across your body, with the RCD assuming you are just another valid load

Maybe not common, but unsafe - esp. if one appliance is (say) your boiler and all taps and radiators become live.

I assume the MG4 has overload protection so appliance circuit shorting will trip the supply. Otherwise an RCBO is advisable.

I would add MCB with RCD or RCBO anyway for fault protection - though that will need a load of 3 to 5 times the current rating to trigger a Type B MCB quickly.

 

[CrustEE]

On the subject of earthing, this article explains that inverter power sources cannot necessarily have a N-E connection without the potential of causing serious damage to the inverter.

MythBusters #6

This month, James Eade looks at his specialist subject of temporary power and, in particular, the earthing of temporary generator sets.

electrical.theiet.org

This is supported by measurements that I made using a multimeter at the MG V2L via a cheap ebay adapter. No load connected.
L to N: 222Vac
E to N: 26Vac
E to L: 196Vac
This does indeed indicate that the E (PE) connection is not tied to N and not to a mid-way point either.

I recently had my consumer unit changed and ran three items in the house from the MG V2L. This was only partially successful because the MG discharge stopped by itself on several occasions. I have tried to recreate this failure since, and could not. The first occasion was during wet weather (drizzle), and the second was when it was dry. The conclusion I have come to is that MG inverter does have internal protection but I can only speculate that it might be some sort of current leakage detection.

Subsequent to this, I experienced MG charging drop-outs during wet weather when using a two year old type 2 cable. I replaced the cable and the drop-out problem appears to have gone away. Of course, I don't know if the car was stopping the charge or Zappi. Either way, I surmise that damp in the system will cause current leakage.

What we really need to know from MG is:

• What are the V2L output parameters?
• What internal protections are provided?
• What is the recommended earthing connection (if any) when powering a house appliance which is not double insulated (such as a fridge)?
• Does MG recommend any additional protection at the house, such as an RCBO?

 

[Everest]

On the subject of earthing, this article explains that inverter power sources cannot necessarily have a N-E connection without the potential of causing serious damage to the inverter.

This is supported by measurements that I made using a multimeter at the MG V2L via a cheap ebay adapter. No load connected.
L to N: 222Vac
E to N: 26Vac
E to L: 196Vac
This does indeed indicate that the E (PE) connection is not tied to N and not to a mid-way point either.

Good of you to have made and detail those readings, thanks. Can you confirm whether the 'E' in those measurements was the earth connection on the V2L lead or the house earth? Also, it would be interesting to repeat said measurements with a 5600 ohm resistor between L and E as that would trigger a 40mA earth leakage current, triggering a typical 30mA RCD. I suspect there is some induced voltage on the lines, but not enough current to be an issue (but that is only guesswork).

With regard to inverter damage, just for clarity, I was not proposing (in post #20 above) to bond N and E on the V2L connection, but to bond the N output to real earth at the termination of the V2L connection, prior to the RCD and prior to the house's critical loads circuit. Then, within the house's critical load circuit there should be an earth rod providing a protective earth reference. Just effectively creating a TT earthing arrangement. This is needed unless the house already has TT earthing, which occurs mostly in rural location with overhead supply.

I don't see (but open to thoughts) as to how that could be in any way detrimental to the inverter as there is no connection to the car's earth (= chassis) and the car itself is not grounded - unless standing in a deep puddle or covered in snow.

I recently had my consumer unit changed and ran three items in the house from the MG V2L. This was only partially successful because the MG discharge stopped by itself on several occasions. I have tried to recreate this failure since, and could not. The first occasion was during wet weather (drizzle), and the second was when it was dry. The conclusion I have come to is that MG inverter does have internal protection but I can only speculate that it might be some sort of current leakage detection.

Interesting - what where the loads? Were they inductive (e.g. fridge / freezer)? There could be restriction on power factor as well, as mentioned in your link.

What we really need to know from MG is:

• What are the V2L output parameters?
• What internal protections are provided?
• What is the recommended earthing connection (if any) when powering a house appliance which is not double insulated (such as a fridge)?
• Does MG recommend any additional protection at the house, such as an RCBO?

I suspect such info will not be available, given that they don't (it seems) even say what the output power limit is. From what I have read, there is overcurrent protection, determined on the resistor value in the V2L lead.

I doubt that the MG would even recommend powering a house from the V2L output either, due to a) the limited supply current and b) the complexity of earthing and safety. My best guess (and what I would say if I were MG) is that the V2L output is designed for powering temporary devices that are double-insulated for camping type activities and hence no earth requirements are relevant.

Would be good to find out more though and to know if any others have gleaned more info from MG.

 

[Eclectic]

I made my own from a charging cable. I used a resistor to give 7kW output, though I don't ever come close to that.

Used on my ZS Trophy LR to run washing machine/tumble dryer (before I got solar and Powerwall 2) and for mowing the front lawn as the outside power point is round the back of the house

It's always worked very well, and the flat format 13 amp socket fits through the letterbox to use as an indoor extension.

 

[CrustEE]

Can you confirm whether the 'E' in those measurements was the earth connection on the V2L lead or the house earth?

Directly on the adapter without lead.

Also, it would be interesting to repeat said measurements with a 5600 ohm resistor between L and E as that would trigger a 40mA earth leakage current, triggering a typical 30mA RCD.

I can do. It will take a bit of time though.

With regard to inverter damage, just for clarity, I was not proposing (in post #20 above) to bond N and E on the V2L connection, but to bond the N output to real earth at the termination of the V2L connection, prior to the RCD and prior to the house's critical loads circuit. Then, within the house's critical load circuit there should be an earth rod providing a protective earth reference. Just effectively creating a TT earthing arrangement. This is needed unless the house already has TT earthing, which occurs mostly in rural location with overhead supply.

Could you sketch a circuit of this for my full understanding?
FYI, I am in a rural area with overhead power cable, I am not sure how the original earth was provided to the property, but subsequently I had the solar system installed with an EPS circuit and a local earth rod.

I don't see (but open to thoughts) as to how that could be in any way detrimental to the inverter as there is no connection to the car's earth (= chassis) and the car itself is not grounded - unless standing in a deep puddle or covered in snow.

Using the diagram from the linked paper;

Ignore the box around the inverter as it is irrelevant, if you were to connect a single "line" output (nominally designated neutral) to CPC conductor it shorts the IGBT emitter back to the mid-point of the DC stack. Once the IGBT gate is turned on, too much current will flow in the loop.
The cure for this is to ignore the PE presented by the MG and connect N to E as it enters the house.

Interesting - what where the loads? Were they inductive (e.g. fridge / freezer)? There could be restriction on power factor as well, as mentioned in your link.

A fridge, wifi and an LED light. This is a good point because the fridge compressor is not always on.
I am going to experiment more because the worst grid loss we have had in this house lasted 56 hours and the house battery was not man enough to see us through.

 

[Everest]

Ignore the box around the inverter as it is irrelevant, if you were to connect a single "line" output (nominally designated neutral) to CPC conductor it shorts the IGBT emitter back to the mid-point of the DC stack. Once the IGBT gate is turned on, too much current will flow in the loop.

Could be, but that would depend on the design of the car's inverter and where and if it were grounded to the vehicle.

The cure for this is to ignore the PE presented by the MG and connect N to E as it enters the house.

Errm... that's exactly what I was saying

With regard to inverter damage, just for clarity, I was not proposing (in post #20 above) to bond N and E on the V2L connection, but to bond the N output to real earth at the termination of the V2L connection, prior to the RCD and prior to the house's critical loads circuit.