Hi,

I'm looking for the final view about tying neutral and earth on the output side of my Infini/Replus. 

When my inverter is online with grid and is grid tied then the output "sees" the tie on the supply side anyway. 

But when grid goes off the 2P "safety" relay opens and now the neutral on the output side floats. 

My sparky is here today and he says his course said that the output should have neutral/earth tied. 

When the inverter is grid-tied this seems harmless since on the input side there is an existing tie anyway. 

In offline mode the tie will tie the output neutral to earth which is a common earth with the supply side. 

I don't think it would be harmful in either mode but I am scared to damage my inverter with some unexpected issue? 

Thanks 

6 minutes ago, Elbow said:

I'm looking for the final view about tying neutral and earth on the output side of my Infini/Replus.

I'm going to be neutral on this. 😋

Then I'll be a live wire, and get more active 🗯️ 😀. 

As we've discussed with Axperts, the issue is that multiple connections between neutral and earth cause problems with (a) non-fault current in the earth wires, and as a consequence (b) false tripping of safety switches.

Ideally, the inverter should do the connection itself, since it knows when the connection to the AC-in neutral is off. Otherwise, it's best to use some sort of relay external to the inverter, switched with a signal from the inverter.

I don't know the Infinis well, but I don't think they do the neutral out to earth connection automatically, and I don't know if they have a suitable signal, as the Axperts do (but you have to give up your generator start signal).

The double connection is not harmless, even though it seems so at first glance.

Edited March 28, 20197 yr by Coulomb

Hi

Just to add, after reading most of the threads, here i decided to tie the output neutral and earth together after testing with a short wire, and all looks fine.

My issue was that the earth leakage tripped whenever the grid was restored, and this fixed the nuisance tripping when grid restores. I've tested the earth leakages after the inverter (by pressing the test button) and they seem to work, so pretty happy with the result.

My other thought was to use a contactor, but since the Infini 3KW does not have a relay for this, I couldnt think what contactor to use, as the inverter engages the grid only a few seconds after the grid comes on, so a contactor would disengage too quickly.

Thanks, 

So my sparky tied neutral and earth on the output since in his view that was correct.  and we tested both online and offline and no smoke was seen. 

But I will take your advice and look for a signal that I can use to switch a relay to connect the neutral to earth on the output only when the safety relay is open. 

One way would be to look in the data coming on the modbus data to tell me when the inverter has isolated its input. 

If you tie the earth to neutral on the Inverter output while in bypass mode  the upstream feeding E/L is supposed to trip.

7 minutes ago, Jaws said:

If you tie the earth to neutral on the Inverter output while in bypass mode  the upstream feeding E/L is supposed to trip.

My inverter is fed before the main board RCD.  Which I think is usual.

19 minutes ago, Elbow said:

My inverter is fed before the main board RCD.  Which I think is usual.

In that case it wount trip but the reason why the only leagally correct way to do it is with a relay / contactor as per Coulomb is because now you are creating a TNC system after the point of supply.

 

 

 

We are only allowed TNS or TN-C-S systems

 

 

 

Once the grid goes off and the inverter disconnects from the Grid only then must you combine earth and neutral.  -Now the inverter becomes the new source of energy and your earth leakage right after the inverter becomes the point of control

Edited March 28, 20197 yr by Jaws

Hi @Jaws - 

 

i do understand that it’s a problem, but of course as things stand it’s a problem to link neutral and earth since that creates the multiple points issue when the inverter is tied, and it’s a problem to not have the link since neutral floats when the inverter is isolated.  And floating neutral apparently creates a hazard (not too obvious to me why why since an RCD would still detect imbalance and trip if you touched the floating neutral and took current down to earth?)

So I will see about a relay.

In the meanwhile, which is the lesser of two evils?

Its always hard to give advice over the damn internet.

But if you have a solid earth bond on the inverter and an E/L installed directly after Earth Neutral bond  (its perfectly save).

These damn standards are not always clear and can be differently interpreted.  The whole thing with TNC and the warning was that you run only two conductors by combining the earth and neutral.  So the intent why they don’t want to do it is because:

1. Its much saver to have a separate earth and neutral conductor , which you have to the inverter.
2. Not to interfere with upstream E/L protection (which you dont have)  as an E/L does exactly that check for leakage current to earth.  So if the live conductor touches earth it will trip but also if the neutral touches earth it will also trip, as some of the current will now flow back trough the earth instead of all the current back trough the neutral.

 

 

24 minutes ago, Jaws said:

because now you are creating a TNC system after the point of supply

I’m curious why you describe it as a TNC system since I don’t have a 2 wire live and neutral as per your diagram?  Your diagram shows the live phases and a single PEN.

 All my plug circuits etc are protected by RCD whether on the grid side or the inverter out.

So on my inverter out the neutral earth is tied before the RCD and so it would trip if "return" current gets around the RCD via me and earth.

So I feel it’s safer than a floating neutral until I can get the relay worked out.

 

23 minutes ago, Elbow said:

I’m curious why you describe it as a TNC system since I don’t have a 2 wire live and neutral as per your diagram?  Your diagram shows the live phases and a single PEN.

 All my plug circuits etc are protected by RCD whether on the grid side or the inverter out.

So on my inverter out the neutral earth is tied before the RCD and so it would trip if "return" current gets around the RCD via me and earth.

So I feel it’s safer than a floating neutral until I can get the relay worked out.

 

Agreed.

The reason its working in your installation is because you supply your inverter from a circuit with no earth leakage protection.

If you supply the inverter from an Earth Leakage protected circuit then you need the Neutral Earth Relay on the ouput of the inverter otherwise the supply Earth leakage unit will trip.

 

SANS 10142 is rather clear on this topic:

 

 

 

Edited March 28, 20197 yr by Jaws

2 minutes ago, Jaws said:

Agreed.

The reason its working in your installation is because you supply your inverter from a circuit with no earth leakage protection.

If you supply the inverter from an Earth Leakage protected circuit then you need the Neutral Earth Relay on the ouput of the inverter otherwise the supply Earth leakage unit will trip.

 

 

So to think aloud - if I make a relay that can bridge N-E - how should the control logic go?

1) at the same time the safety relay opens the bridge relay should close.

2) or the bridge relay closes after a short delay? How short?

3) or the bridge relay closes before the safety relay opens?

Elbow wait Im still trying to get past your counter argument that you can earth the neutral down stream , as long as the down stream is not after an earth leakage device - so far I have nothing to say its not allowed.

If you read TNS as three seperate conductor L + N + E     and TNC as two conductor  L + PEN you run three wires its still TNS

Or does TNS means from a point down Neutral and eath are always seperate ....(not just in terms of condcutors but also in terms of connection)

Its been a long day , let me sleep over it. 

Edited March 28, 20197 yr by Jaws

Ok found the relevant sections

 

 

So P3 below is what most Axpert users will go for,   but I think it makes the assumption that the UPS / Inverter will automatically connect Earth and neutral when it starts inverting.

 

 

 

 

 

 

 

Edited March 28, 20197 yr by Jaws

20 minutes ago, Elbow said:

1) at the same time the safety relay opens the bridge relay should close. 

2) or the bridge relay closes after a short delay? How short? 

3) or the bridge relay closes before the safety relay opens? 

The bonding relay must close AFTER the safety relay opens (otherwise it might trip upstream RCDS... which you don't have, but that is the reason). It opens before the safety relay closes for the same reason. The exact time is not very important, but should probably be a short interval, as fast as is reasonably doable.

4 minutes ago, Jaws said:

Ok found the relevant sections

That was the section I alluded to earlier. With a TN-C-S installation, if your inverter supplies the whole installation and is reasonably close to to the supply point (where the bond is installed and the wire goes to the earth spike), then you can connect the neutrals together (this has the effect of bridging the neutral on the input of the inverter onto the output of the inverter), causing input, output, everything to be permanently bonded.

Demo regarding current flowing on the ground wire ...

3 hours ago, plonkster said:

... then you can connect the neutrals together (this has the effect of bridging the neutral on the input of the inverter onto the output of the inverter), causing input, output, everything to be permanently bonded.

But then the problem is if you have double pole breakers (as seems to be common in South Africa), then you are disconnecting the inverter neutral from earth whenever the AC input is isolated, or the breaker trips. Surely this is a problem too?

4 hours ago, Elbow said:

And floating neutral apparently creates a hazard (not too obvious to me why why since an RCD would still detect imbalance and trip if you touched the floating neutral and took current down to earth?)

You need the neutral connection to earth to force the imbalance. Otherwise, the RCD would not always register the current difference. For example, maybe both active and neutral conductors are approximately equally capacitively coupled to some active conductor (whose neutral return is earthed), separate from the device's conductors. Then you could get a shock, real current to ground, and the current through the RCD will be about the same, and so it won't trip.

When one of the conductors is earthed, none of that current will flow through you, assuming that you are also earthed. So all the capacitively coupled current will be registered as an imbalance, protecting you.

For the Axpert I see there is an option

On 2019/03/29 at 8:30 AM, Jaws said:

For the Axpert I see there is an option

Not on all models. Some have it built in (they made one of the relays double-throw), others with new anough firmware has the setting as above, and others don't have it at all.

Edited March 30, 20197 yr by plonkster

 

So I tossed and turned on Thursday night thinking about neutral - earth ties.  My XYL asked me why I couldn’t sleep and I was embarrased to say why!  I’m sure when I did she was thinking “who is this weird person I’m married to??” 🤣

I drew it all up in the 4 possible configs and thought about it and I see why you’d want the tie there when the inverter is isolated from the grid.  And equally why you’d want it not there when the inverter is grid-tied.

(This also starts to give me some insight into why CoCT might be cautious about hybrid inverters - they have a lot of different modes to think about).

Anyway - I understand why I’d want to use a contactor to make the neutral earth tie but only when the inverter is isolated from the grid.

Normally no current flows on the neutral to earth tie, right?  But the whole point is that in fault conditions it will pass a lot of current so as to quickly trip the overload breaker.  My inverter output has a 32A 2P breaker on it.

So what rating would the contactor need to be?

So now I’m looking for suggestions of an affordable DIN-mounting contactor.  It only needs to have single pole switching.  For me it would be easiest if the coil/solenoid was low voltage but I could switch that in turn with a low power relay.

The working principle would be that when the inverter safety relay opens, isolating the inverter, the contactor would make the neutral to earth connection.  Ideally I’ll find a signal from the inverter, failing that switching based on the AC input being present or not is pretty close.

Thanks,
Elbow

 

 

19 minutes ago, Elbow said:

So what rating would the contactor need to be?

Continuous it need only be larger than the max that might go through there. Though the inverter has a 32A on its output, the bond must be made BEFORE this breaker, so size it to the max of the inverter or the breaker on the inverter input (whichever is larger).

I'm curious how you're going to switch it. I think the only fool-proof way (that I could think of) was to open it up, locate the safety relay coil terminals, and take a signal from there. Use that to carefully feed a SSR (that ought to be safe, those things are usually an opto-isolator with a Thyristor) and switch the contactor with that, using the inverter's output AC to power the coil.

8 minutes ago, plonkster said:

Continuous it need only be larger than the max that might go through there. Though the inverter has a 32A on its output, the bond must be made BEFORE this breaker, so size it to the max of the inverter or the breaker on the inverter input (whichever is larger)

Well input breaker is 32A too.  But of course any breaker will allow more than the rated breaking current for a time.

10 minutes ago, plonkster said:

I'm curious how you're going to switch it. I think the only fool-proof way (that I could think of) was to open it up, locate the safety relay coil terminals, and take a signal from there. Use that to carefully feed a SSR (that ought to be safe, those things are usually an opto-isolator with a Thyristor) and switch the contactor with that, using the inverter's output AC to power the coil.

So that was my plan.  Block diagrams show the safety relay controlled by two detection signals coming from redundant control logic.  I need to test to see what I find.  If it is ordinary digital voltage level with a driver circuit to drive the relay coil I might be able to use it quite easily to switch a low power relay - with a 5v coil - and use that in turn to switch the contactor.

I have a second unit which I can look at in the lab.

No detailed design yet.

I did write to Voltronic to ask about circuit diagrams (and also their advice about N/E bonding - we shall see what they reply (if they reply).  

 

1 hour ago, plonkster said:

others with new enough firmware has the setting as above,

It appears in main firmware version 72.70.

Quote

and others don't have it at all.

True, but they could update to 72.70 or 73.00. [ Edit: but this is for one line of Axperts, not Infinis, the subject of this topic. ]

Edited March 30, 20197 yr by Coulomb