Hi guys,

I'm hoping someone here has either seen this before or can see what I am missing!

So I had an RCT Axpert 3kVA 24V inverter installed for years, close to 6. But it became apparent that the 600W MPPT PV input just was not sufficient to make it viable as a long term Load shedding and power outage savior. Also it could only take the standard lead acid batteries - which I was replacing every 2 years and this just seemed counter productive.

 

So fast forward to a month ago, I decided it was time to upgrade the system.

I purchased a Kodak OG3.24 Inverter, 4 x 540W JA Solar Mono panels, and a lithium battery.

I simply disconnected the RCT from the power and PV breaker boxes respectively, and replaced it on the wall with the Kodak and reconnected the AC and DC wiring.

 

Now, when the power goes out the Earth Leakage trips, and the only way to reset the EL is to switch off the inverter, reset the EL and turn on the inverter again.

I have run through the wiring over and over, and I just cannot figure out where there would exist a potential difference between live and neutral to cause this to happen.

AC from Main DB leaves before the EL to AC Inverter DB where it goes into a change over switch, a 25A input breaker and then into the inverter.

AC from Inverter goes to a 25A output breaker, into the change over switch and back to the main AC DB where is feeds all the lighting circuits and one plug circuit (Which is not on EL - and need to be rectified I know)

If the change over switch is in the bypass position (i.e. incoming AC from main DB goes straight back to Main DB but also still goes to inverter AC in) and the AC breaker to the inverter is turned off the EL does not trip...

What boggles my brain is that the inverter is totally independent of the EL in the Main DB yet cause it to trip as soon as the input power to the inverter is removed (25A input breaker turned off or load shedding/power failure)... And this never happened when the RCT Axpert was in place...

 

Pointers of what to look for before I phone a sparky to come inspect everything??

Edited November 4, 20214 yr by Nic Holmes

23 minutes ago, 87 Dream said:

Can you take a pic of your AC input & AC output from the inverter connection. Also please take a pic of the inverter side sticker. This gives us a match to see what inverter model you have.

You seem to have identified that the fault lies within the inverter. 

Pretty sure we can troubleshoot this & spot a problem. 

87

Excuse the disaster that is the Inverter AC DB, its a wiring mess that has resulted from trying to trace through this issue and will be tidied up once the fault is found and rectified.

 

Attached is the data plate from the inverter, plus DB pics. Below is a schematic of the wiring.

 

27 minutes ago, 87 Dream said:

Do you have a sticker of the old inverter ? 

87

Previous inverter that was installed

18 minutes ago, 87 Dream said:

This link is also related to your question:

 

You had an old inverter & Voltronics changed from a floating earth to a EN safety. Hence you should be able to hook up an Axpert after your home's earth leakage & it works just fine. AC input fed from the grid after the Main DB Box earth leakage switch. The Earth leakage should stay up because it detects no fault. In your case you have gone from the sounds of it from a floating Earth to a more normal EN arrangement. 

This can be a needle in a hay stack. In you pics we cannot see the Neutral common Busbar.

I think it's better to rewire from the beginning. That way you know your cheques & balances. At the moment it's a mess to try & fault find.

87

Yeah I agree...

And the more I trace wires the more mess I'm making and the more confused I find myself.

 

So what you are saying is that I can take the power to the inverter after the main EL and not before it like it is currently wired? then following my schematic should be fine?

25 minutes ago, 87 Dream said:

It needs to be fed after the Earth leakage, some gurus that are about can also provide their comments. However, this is what is common practice in the modern Axperts that I have wired. 

I have not installed the floating earth Axperts. 

Maybe you will need some more assurance from the larger community here. 

87

I'm totally confused. I am a light current engineer and this floating earth/bonded story just doesn't compute in my brain

 

My next question that arises then, is if you can't have an earth leakage breaker on the output of the inverter how do you make your plug circuits that are fed from the inverter safe? as fault that trips the earth leakage will only put the inverter into battery mode and still supply power onto that circuit.

Edited November 4, 20214 yr by Nic Holmes

The common problem with RCDs is that if you have a 2 earth neutral ties you create a potential difference which cause an imbalanced current to flow in the neutral line and the RCD does what it's supposed to and trips. Your new inverter could be introducing this second EN bond when inverting. Check pages 58-60 of wiring unlimited which is in the downloads section if you want more information.

5 hours ago, Nic Holmes said:

What boggles my brain is that the inverter is totally independent of the EL in the Main DB yet cause it to trip as soon as the input power to the inverter is removed (25A input breaker turned off or load shedding/power failure)... And this never happened when the RCT Axpert was in place...

Hi @Nic Holmes, I am of the opinion that your new inverter has an automatic  earth bonding  relay arrangement , something your old inverter didn't have.  

Further to this , if your inverter output circuits has a neutral that is wrongfully connected to the neutral of the circuits that is fed from the main ELB , then you will have this failure mode. In this case the following happens :

When the input grid to inverter is lost , the earth bonding relay bond neural to earth at the inverter.

However , the loads on the Main ELD have now two return paths, that of the neutral of the ELD , AND that of the inverter earth bond. Current flowing through the inverter earth bond is not seen by the ELD, and trips.

If you have a suspicion that this may be the case , then you can test it as follows:

 

1 isolate the inverter output from its loads by means of the circuit breaker. In this manner the load neutral is isolated from the inverter.

 

2 then remove incoming grid to input of the inverter.

 

3 if the ELD does not trip it's a dead giveaway you have mixed up neutrals 

Good luck

 

 

 

 

 

Edited November 4, 20214 yr by dropkick
Misspelling

Something I noticed on your drawing the neutral connected to the neutral bar which feeds the plugs is on the in house neutral in and not on the house neutral out.

There is no earth leakage on the drawing but there one which doesnt have a neutral and seems to be bypassed in the picture ... I am assuming it is not the earth leakage in the picture that is tripping? 

 

 

1 hour ago, isetech said:

Something I noticed on your drawing the neutral connected to the neutral bar which feeds the plugs is on the in house neutral in and not on the house neutral out.

I Agree with isetech the neutral connected to house neutral in seems bypassed as it goes straight to the neutral bar, I’m also thinking this neutral should maybe be connected to the house neutral out.

11 hours ago, isetech said:

Something I noticed on your drawing the neutral connected to the neutral bar which feeds the plugs is on the in house neutral in and not on the house neutral out.

There is no earth leakage on the drawing but there one which doesn't have a neutral and seems to be bypassed in the picture ... I am assuming it is not the earth leakage in the picture that is tripping? 

 

 

The purpose on the ELD in the Inverter AC DB was to feed the two plug circuits wired to the right of it. (2x10A) but it was also tripping and bypassed.

 

The Main ELD lives in another DB, and the two neutrals come from the same neutral bus within the main DB so whether its house out or house in, if you check for continuity between them they tied together anyway.

12 hours ago, dropkick said:

Hi @Nic Holmes, I am of the opinion that your new inverter has an automatic  earth bonding  relay arrangement , something your old inverter didn't have.  

Further to this , if your inverter output circuits has a neutral that is wrongfully connected to the neutral of the circuits that is fed from the main ELB , then you will have this failure mode. In this case the following happens :

When the input grid to inverter is lost , the earth bonding relay bond neural to earth at the inverter.

However , the loads on the Main ELD have now two return paths, that of the neutral of the ELD , AND that of the inverter earth bond. Current flowing through the inverter earth bond is not seen by the ELD, and trips.

If you have a suspicion that this may be the case , then you can test it as follows:

 

1 isolate the inverter output from its loads by means of the circuit breaker. In this manner the load neutral is isolated from the inverter.

 

2 then remove incoming grid to input of the inverter.

 

3 if the ELD does not trip it's a dead giveaway you have mixed up neutrals 

Good luck

 

 

 

 

 

I have just run this test.

Turned the output 25A breaker off, tripped the incoming 25A breaker to the inverter to put it into inverting mode and well.. ELD still trips...

8 minutes ago, Nic Holmes said:

I have just run this test.

Turned the output 25A breaker off, tripped the incoming 25A breaker to the inverter to put it into inverting mode and well.. ELD still trips...

Hi Nic , just be be sure, is that inverter output breaker double poled, ie it must isolated the neutral as well when tripped?

1 minute ago, dropkick said:

Hi Nic , just be be sure, is that inverter output breaker double poled, ie it must isolated the neutral as well when tripped?

Ah.... No its not, only on Live out!!

So, I think you might be onto something.

10 minutes ago, Nic Holmes said:

Ah.... No its not, only on Live out!!

So, I think you might be onto something.

Ok , so I think you know what to do , temporary decommission the Neatral output of the inverter , keep the output live circuit breaker off , and run the test on the inverter input again.

9 minutes ago, dropkick said:

Ok , so I think you know what to do , temporary decommission the Neatral output of the inverter , keep the output live circuit breaker off , and run the test on the inverter input again.

I will try that later today and feedback.

If that is the case and there are mixed up neutrals, any idea how to rectify it?

 

Thinking about it...

If I turn the changeover switch to the off position (ie Neutral from Inverter & Live from inverter isolated from house DB) it does not trip the ELD when input power to the Inverter is removed.... I think this answers it!

Edited November 5, 20214 yr by Nic Holmes
Brainwave...

1 minute ago, Nic Holmes said:

I will try that later today and feedback.

If that is the case and there are mixed up neutrals, any idea how to rectify it?

That's  just a lot hard work , you'll have to isolate and trace , that's where the hiring of electrician is an expensive exercise. You are LC engineer, so you can save yourself money here, if you do have the time.

But let's first ascertain , in no uncertain terms , that we have inadvertant neutral routing here .

 

We'll chat again.

One of the most common faults with earth leakage units ... mixing up neutrals and series connecting earth earth leakage units. 

Add in an inverter with bonded earth/neutral ... even semi skilled sparkies get confused.

Understanding the SANS regs helps with positioning of the earth leakage in the circuits. 

One of the most common comments I hear ... "but it has worked for years now suddenly".

Make sure the earth leakage devices are not in series and make sure the neutrals and allocated to the correct neutral bar ... then make sure you earth neutral bond is before the earth leakage ... as close to the inverter as possible. 

I still need to see one of those external neutral/earth bonding units.  

 

8 minutes ago, isetech said:

One of the most common faults with earth leakage units ... mixing up neutrals and series connecting earth earth leakage units. 

Add in an inverter with bonded earth/neutral ... even semi skilled sparkies get confused.

Understanding the SANS regs helps with positioning of the earth leakage in the circuits. 

One of the most common comments I hear ... "but it has worked for years now suddenly".

Make sure the earth leakage devices are not in series and make sure the neutrals and allocated to the correct neutral bar ... then make sure you earth neutral bond is before the earth leakage ... as close to the inverter as possible. 

I still need to see one of those external neutral/earth bonding units.  

 

Pure truth, I cannot agree with you more .

And I blame not the electricians , I blame the SANSes and the COCs of this world ,they always try to be too formal with their normatives , leaving doors open for confusion and ambiguity. 

I have seen some potentially dangerous installation instances on this forum.

9 minutes ago, dropkick said:

Pure truth, I cannot agree with you more .

And I blame not the electricians , I blame the SANSes and the COCs of this world ,they always try to be too formal with their normatives , leaving doors open for confusion and ambiguity. 

I have seen some potentially dangerous installation instances on this forum.

Oh and I also blame manufacturers that does not clearly spec exactly  what safety precautions their products have, and recommendations to the system integrator. Instead they bragging about fancy displays etc.

Edited November 5, 20214 yr by dropkick

1 hour ago, isetech said:

One of the most common faults with earth leakage units ... mixing up neutrals and series connecting earth earth leakage units. 

Add in an inverter with bonded earth/neutral ... even semi skilled sparkies get confused.

Understanding the SANS regs helps with positioning of the earth leakage in the circuits. 

One of the most common comments I hear ... "but it has worked for years now suddenly".

Make sure the earth leakage devices are not in series and make sure the neutrals and allocated to the correct neutral bar ... then make sure you earth neutral bond is before the earth leakage ... as close to the inverter as possible. 

I still need to see one of those external neutral/earth bonding units.  

 

I think I have just clicked as to what the issue is...

Live is taken to the inverter from before the Earth Leakage in the main DB, but if I recall, Neutral is connected to the Neutral bus on the output of the Earth Leakage unit in the Main DB, this same Neutral bus is then also connected to the Inverter AC output on the return feed.

I shall have to open up the main DB and have a good look at how it is connected.... I must highlight, the original connection was done by a sparky, I simply replaced the Inverter and found a "new problem"

17 minutes ago, Nic Holmes said:

I think I have just clicked as to what the issue is...

Live is taken to the inverter from before the Earth Leakage in the main DB, but if I recall, Neutral is connected to the Neutral bus on the output of the Earth Leakage unit in the Main DB, this same Neutral bus is then also connected to the Inverter AC output on the return feed.

I shall have to open up the main DB and have a good look at how it is connected.... I must highlight, the original connection was done by a sparky, I simply replaced the Inverter and found a "new problem"

Ha ha , I wondered this morning if the problem could be as simple as the Neatral of the inverter output has been wired to the ELD downstream neutral bar 🙂

If that would be the case , you will have to remove all neatrals from the busbar , and trace which is the inverter one . Make yourself a half-decent buzzer to trace , don't use your fluke meter, the input impedance is too high , so it will show connections that are not  'hard' connections. Inject at least 30 milliamps from buzzer. 

On 2021/11/05 at 10:12 AM, dropkick said:

Ha ha , I wondered this morning if the problem could be as simple as the Neatral of the inverter output has been wired to the ELD downstream neutral bar 🙂

If that would be the case , you will have to remove all neatrals from the busbar , and trace which is the inverter one . Make yourself a half-decent buzzer to trace , don't use your fluke meter, the input impedance is too high , so it will show connections that are not  'hard' connections. Inject at least 30 milliamps from buzzer. 

So...

Finally I managed to get to the main DB and well, this is a total disaster!

1. The inverter is connected downstream of the ELD, both Neutral and Live, Both input to Inverter and Output from Inverter!

2. EVERYTHING in the house is connected downstream of the ELD....

3. Cannot make heads or tail of the wiring!!

 

 

59 minutes ago, Nic Holmes said:

So...

Finally I managed to get to the main DB and well, this is a total disaster!

1. The inverter is connected downstream of the ELD, both Neutral and Live, Both input to Inverter and Output from Inverter!

2. EVERYTHING in the house is connected downstream of the ELD....

3. Cannot make heads or tail of the wiring!!

 

 

That is indeed a mess. Did the sparky provide a COC when it was last changed?

Those breakers jumpered with 1.5mm or 2.5mm wiring (hard to tell)? Guess it doesn't matter though, they really shouldn't be jumpered like that from my understanding.

I'm guessing the cable tied L/N/E on the right is what's feeding your inverter AC input? If so - definitely looks like it's after E/L.

Probably an opportune time to separate your essential and non-essential loads and get a separate DB for essential loads.

Edited November 8, 20214 yr by JoeyhZA

Hi @Nic Holmes what is the current status of your situation.do you require further support? 

You may noticed that in my conversation with Coulomb I have made sure that the operation of earth bonding relay in Axpert is well understood and there is no doubt in my mind that this is the cause of your problems. You will have to implement split neutrals , one for the main ELD and a seperate neutral for your inverter output.

Then your downstream ELD will also work correctly.

 

And the split neatrals is also  recommended / required as per SANS.

Edited November 11, 20214 yr by dropkick

14 minutes ago, dropkick said:

Hi @Nic Holmes what is the current status of your situation.do you require further support? 

You may noticed that in my conversation with Coulomb I have made sure that the operation of earth bonding relay in Axpert is well understood and there is no doubt in my mind that this is the cause of your problems. You will have to implement split neutrals , one for the main ELD and a seperate neutral for your inverter output.

Then your downstream ELD will also work correctly.

 

And the split neatrals is also  recommended / required as per SANS.

Thanks for the follow up.

 

I am in the process of employing the services of an electrician with experience in the Axpert King inverters, I cannot figure out where all these neutrals go for the circuits I need, so time to pay someone to fix it correctly and get a new CoC!!