Question from an installer, I will forward questions.
Jag ska försöka.
The problem occurred during commissioning the off-grid system.
When the Victron Multi RS Solar were turned off and therefore the earthing relay in the unit was open I got around 60V between neutral and earthing in the sockets in the AC-system even when the main braker in the AC-switchboard was turned off and the RCD turned off.
To the main grounding rail we connected:
AC-switchboard with surge protection
Ground electrode
Victron Multi RS Solar
DC-switchboard with surge protection
Lightning protection in form of ring conductor electrode
DC negative on the 48V DC battery system
When we added a jumper between N-rail and PE-rail in the AC-switchboard before the RCD the problem disappeared.
I believe the problem wouldn´t have occurred if the earthing relay would have been closed in every case other than when a separate AC-supply unit have been connected, but maybe I’m wrong.
I now have turned off the earthing relay permanently and rely on the jumper in the AC-switchboard.
It is (n and e) always bonded at your first point of connection.
Where did the voltage come from when the inverter with no power source as it was off?
The other possible reason is capacitive coupling (ghost voltage) which means you are not using a low impedance or low z meter for testing electrical installations. Not the same a normal DMM
No, I find it strange as well. Just to clarify once again I haven’t personally been on this installation. Generally I always use the built in N-PE relay, but seeing as this system is never intended to have a genset or other external ac-source I felt like this solution is acceptable.
When and where does the MPPT in the multi rs disconnect? The PV is not isolated from the AC in the multi rs as I understand it? Is there a possibility that the potential from the PV is somehow induced on the neutral of the AC?
Panels and every other ground point are bonded to the main earth terminal, with a grounding rod for lightning protection as well.
I don’t see how this potential is possible to get with the inverter off. The only possibility is either from the batteries or from the panels.
Could there be a capacitive charge built up in the inverter that takes several minutes to discharge after it’s turned off that shows up between neutral and earth?
Panels remain live since you cant switch them off only disconnect at the system switchgear and isolate on the bottom of the RS. (Unless you have switchgear at the array.)
4.7 states that there is not full galvanic isolation between PV DC input and AC output, but is it reasonable to believe that the 60V measured between earth and neutral on the AC output is a result of this?
You are right and it’s reasonable to measure that voltage because there is a high impedance input on a DVM!
Try to fully disconnect PV panels and see if the voltage is still there. Use the PV switch on the Multi RS.
PV to AC-Out is not galvanic isolated…
Ah…
And don’t forget the safety Y capacitors between L ↔ PE and N ↔ PE that are generally found in any grid connected device…
They can play tricks on you when you make such measurements…
I’m pretty sure nothing else was connected to the ac output during measurements, but I’ve had issues before with high resistance multimeters when troubleshooting with electronics connected to the AC output
And another thing.
As in the case of Multi RS the battery is galvanic isolated from the AC-In and AC-Out, sometimes is not a good thing to connect its GND to PE…
Below, for reference, a generic schematic similar to Multi RS’s schematic.
Maybe will help with troubleshooting…
The plot thickens! Just spoke to the installer, he measured this voltage without the PV installed. Very strange. He also uses an advanced instrument, Eurotest Xde.
I really can’t think of any explanation. The batteries are galvanically isolated from the AC output. Measuring a potential between earth and neutral ac out on the inverter seems impossible.
I haven’t worked with Victron’s transformerless inverters, is there a possibility of charged capacitors leaking to earth for a couple of minutes?
Certainly there is the ground relay issue - this relay connects the chassis groud of the inverter to PE BUT only in case the AC connection gets opend. So no problem, too.
‘Problems’ meant by this are loops that ‘receive’ electro-magnetic fields. In electical design it is alway a goal to avoid loops like that.
Without doing a propper ‘single-point star-wise grounding’ you will always have the RCD tripping since filter currents are submitted the device_gnd and not get lead to ground.
