Hi guys,

I have an Axpert MKS 5k inverter with Pylontech batteries (US3000 x 2) and eight 330W panels. From the main supply, power goes through an earth leakage to the inverter input and the geyser. There is also a stove connected which is bypassing the EL, but that must have been done 20 years ago. The output of the inverter also goes via another earth leakage to the live trip switches (lights and plugs) and an isolated neutral bar. I am using a common earth.

The problem I have is that the secondary earth leakage trips randomly. It trips quite a lot when the inverter switches from utility power to solar/battery. I have never seen a trip when it switches from solar to utility at night. Recently it has also tripped when plugging in the treadmill and some other exercise machine. Again this doesn't always happen.

This weekend I made sure I didn't mess up the neutral connections when splitting off the stove and geyser during installation and all looks good.

Any ideas what else I can try?

how old is the EL?

The transients when the inverter switches over can cause some devices to draw a little more earth-leakage current for a short time and thus trip the earth leakage breaker. I have found that replacing the earth leakage breaker with a Schneider one solves the problem about 90% of the time.

44 minutes ago, Dex_ said:

how old is the EL?

It is only a few months old.

37 minutes ago, Stanley said:

The transients when the inverter switches over can cause some devices to draw a little more earth-leakage current for a short time and thus trip the earth leakage breaker. I have found that replacing the earth leakage breaker with a Schneider one solves the problem about 90% of the time.

I can try a new one, because the one I have in currently is one of those ACDC products, so I'm not sure how reliable they are.

I found that the Schneider brand definitely seems to be less sensitive to transients and high-frequency components than some other brands.

7 hours ago, Pipes said:

one I have in currently is one of those ACDC products

I am almost certain that is your only problem. You should see the internal heat those units generate if you connect them after an inverter.  

7 hours ago, Pipes said:

I can try a new one, because the one I have in currently is one of those ACDC products, so I'm not sure how reliable they are.

Crappy quality.

We’ve replaced 4 off them.

Finicky rcds tripping whenever can cause damage to other things like servers.

Happened to me.....

A lot has been posted on trips and rcds.

Don't assume the one rcd will trip before the other one does.

It all depends on the actual ma trip sensitivity per rcd.

 

Edited September 16, 20196 yr by Johandup

8 hours ago, Stanley said:

The transients when the inverter switches over can cause some devices to draw a little more earth-leakage current for a short time and thus trip the earth leakage breaker. I have found that replacing the earth leakage breaker with a Schneider one solves the problem about 90% of the time.

Found something similar. I assume this is a voltronic that actually has a bonding relay on the output. Many of the earlier ones didn't have it.

So when the transfer switch switches, there is a very short time in the middle when earth and neutral is not bonded, maybe about a second. During this time any EMI filters or Surge Suppressors you have on the output will absorb some charge. Most EMI stuff is simply three capacitors forming a triad between live, earth and neutral, so as you unbond earth and neutral the potential across these rises very quickly to around half of the supply voltage. Surge arrestors are usually the same, a small triad consisting of MOVs (metal oxide varistors) and MOVs act like small capacitors too.

So about a split second later, the inverter bonds neutral and earth back together again, and this causes the neutral/earth capacitor/MOV combos on that line to create a little discharge burst, and this is often enough to throw the RCD.

The solution is a better RCD, as Stanley said. How much better? Well... what I did was I bought progressively better RCDs until I found one that did the trick. Look for an RCD that is resistant to nuisance tripping. Almost all of them have some resistance, but the cheapies will have the least.

I eventually bought an ABB type A-APR, and it was insanely expensive, but there are some peculiarities about my inverter which yours doesn't have, so I doubt you'd have to go this far.

Further, also look at the type of te RCD. Type-AC is the cheapie, 30-year old tech. Type-A is better, and not too expensive.

(Hint, don't confuse the residual current tripping behaviour with the overcurrent behaviour (where the device also includes overcurrent). You could have a type-AC RCD that also has a type-C overcurrent curve. Just to really confuse you... 🙂 ).

Edited September 16, 20196 yr by plonkster

7 hours ago, plonkster said:

Found something similar. I assume this is a voltronic that actually has a bonding relay on the output. Many of the earlier ones didn't have it.

So when the transfer switch switches, there is a very short time in the middle when earth and neutral is not bonded, maybe about a second. During this time any EMI filters or Surge Suppressors you have on the output will absorb some charge. Most EMI stuff is simply three capacitors forming a triad between live, earth and neutral, so as you unbond earth and neutral the potential across these rises very quickly to around half of the supply voltage. Surge arrestors are usually the same, a small triad consisting of MOVs (metal oxide varistors) and MOVs act like small capacitors too.

So about a split second later, the inverter bonds neutral and earth back together again, and this causes the neutral/earth capacitor/MOV combos on that line to create a little discharge burst, and this is often enough to throw the RCD.

The solution is a better RCD, as Stanley said. How much better? Well... what I did was I bought progressively better RCDs until I found one that did the trick. Look for an RCD that is resistant to nuisance tripping. Almost all of them have some resistance, but the cheapies will have the least.

I eventually bought an ABB type A-APR, and it was insanely expensive, but there are some peculiarities about my inverter which yours doesn't have, so I doubt you'd have to go this far.

Further, also look at the type of te RCD. Type-AC is the cheapie, 30-year old tech. Type-A is better, and not too expensive.

(Hint, don't confuse the residual current tripping behaviour with the overcurrent behaviour (where the device also includes overcurrent). You could have a type-AC RCD that also has a type-C overcurrent curve. Just to really confuse you... 🙂 ).

Thanks for the advise. My inverter is fairly new, I think it is a Nov 2018 model, so I trust it has the bonding relay. The hunt is on for a better RCD😀

On 2019/09/16 at 1:21 PM, Pipes said:

From the main supply, power goes through an earth leakage to the inverter input

There is not suppose to be an earth leakage on the input to the inverter:

On the "eskom" non essential circuit, a earth leakage is required, but the supply to the inverter must be taken before the earth leakage. An earth leakage is also required after (at the output)  of the inverter.

What is happening is that the inverter is bonding the neutral and earth, which is causing the tripping.

27 minutes ago, JacquesVDM said:

There is not suppose to be an earth leakage on the input to the inverter:

On the "eskom" non essential circuit, a earth leakage is required, but the supply to the inverter must be taken before the earth leakage. An earth leakage is also required after (at the output)  of the inverter.

What is happening is that the inverter is bonding the neutral and earth, which is causing the tripping.

Jacques, if this was an issue, wouldn't it trip the EL at the input to the inverter?

41 minutes ago, JacquesVDM said:

What is happening is that the inverter is bonding the neutral and earth, which is causing the tripping.

I do not agree with this. I install Earth Leakage units on the input of the inverter on all my installations. Not one trips. The Inverter only bonds earth and neutral once the grid falls away, by that time there is no power on the input El to sense the "Connection" and trip. 

1 hour ago, JacquesVDM said:

There is not suppose to be an earth leakage on the input to the inverter:

On the "eskom" non essential circuit, a earth leakage is required, but the supply to the inverter must be taken before the earth leakage. An earth leakage is also required after (at the output)  of the inverter.

What is happening is that the inverter is bonding the neutral and earth, which is causing the tripping.

O_O really? no EL on the input side of the inverter? i don't agree.

28 minutes ago, Dex_ said:

O_O really? no EL on the input side of the inverter? i don't agree.

There are rules about what should be protected by an RCD (anything feeding sockets, and also water heaters aka geysers). There is no rule that says you must protect an inverter's input with an RCD. You must however protect the output, but again, you don't have to if it doesn't feed sockets or any other load that explicitly requires this protection.

You must however earth the inverter and you must have overcurrent protection. Non-negotiable.

My own opinion: If the inverter is installed in a place where people can get access to it, touch it, that sort of thing, then consider installing some level of residual current protection. My inverter is in my garage. High enough up that small kids can't ouch it, but teenagers can. I therefore have an RCD on it. Because of the way it tests TN bonding, that RCD is a 100mA unit (not 30mA). It is okay because I'm not feeding sockets or geysers from there 🙂

 

5 hours ago, Pipes said:

Jacques, if this was an issue, wouldn't it trip the EL at the input to the inverter?

Not if the inverter is connected before the earth leakage.

5 hours ago, Jaco de Jongh said:

I do not agree with this. I install Earth Leakage units on the input of the inverter on all my installations. Not one trips. The Inverter only bonds earth and neutral once the grid falls away, by that time there is no power on the input El to sense the "Connection" and trip. 

What I also noted and had to do, was to connect my earth and neutral  together at the inverter output. Otherwise I would have about 80V between N and E.

5 hours ago, Dex_ said:

O_O really? no EL on the input side of the inverter? i don't agree.

I don't see a problem. The frame of the inverter is earthed, so no chance of a shock.

On 2019/09/16 at 1:21 PM, Pipes said:

Any ideas what else I can try?

I have seen the power store has a earth leakage option relay for about R300 online. I think this relay bonds earth and neutral and might prevent the nuisance tripping. 

 

10 hours ago, Gerrie said:

I have seen the power store has a earth leakage option relay for about R300 online. I think this relay bonds earth and neutral and might prevent the nuisance tripping. 

I am confused. Why would an Earth Leakage relay bond neutral and earth to prevent a trip if bonding Earth and Neutral creates a trip?

Edit: Okay, I went and searched for this, and can see why it might be confusing. this is not really an Earth leakage. Is an NER- Neautral earth relay, use to bond the Earth and neutral in battery mode. Its just an external remedy for something a few other inverters already have build in as a standard. 

 

When off-grid, the earth and neutral need to be bonded before the earth leakage breaker so that the breaker will actually function correctly (i.e. trip when there is an earth fault), so some inverters incorporate a relay that bond neutral and earth at the output when they disconnect from the grid. Bonding the earth and neutral doesn't cause the trip so much as certain appliances that draw excessive earth leakage current during the change-over. This is usually solved by replacing the earth leakage breaker by one that is less sensitive to short-duration transients.

48 minutes ago, Stanley said:

Bonding the earth and neutral doesn't cause the trip so much as certain appliances that draw excessive earth leakage current during the change-over.

This is another concept that is new to me. Why would anything draw more leakage current during changeover. Earth leakage current is just what it says. It is leakage current from Live or neutral to earth normally caused by insulation issues, why would changing over affect this condition. 

If you are referring to the effect that DC ripples and some Surge Arrester circuits and Transients might have in causing False trips, then i might agree that the changeover might play a role, but not in the amount of earth leakage current that a device might draw. 

I think @plonkster explained it quite nicely above. Many appliances (mostly electronics with switched mode power supplies etc.) have capacitors and / or MOVs from live and neutral to earth. During the change over, neutral and earth are not bonded so earth is floating, this means that the capacitors and MOVs form a voltage divider that makes the earth terminal of the appliance float somewhere around half-way between live and neutral. When you then bond the neutral and earth, those capacitors (and to some extent MOVs because they have capacitance too) then discharge to earth through the bond and this causes a small spike in earth leakage current which can trip the breaker. Also, if there is any high frequency ripple on the inverter's AC output (due to poor filtering etc.) that can also cause higher than normal earth leakage current when running from the inverter because those same capacitors and MOVs (due to their capacitance) have a lower impedance at higher frequencies. So in some cases you may have trips when running from the inverter and not just at the change over. You can get special earth leakage breakers designed to be tolerant to high frequencies and transients.

22 minutes ago, Stanley said:

that can also cause higher than normal earth leakage current when running from the inverter because those same capacitors and MOVs (due to their capacitance) have a lower impedance at higher frequencies.

Agree, higher "Leakage current"

22 minutes ago, Stanley said:

So in some cases you may have trips when running from the inverter and not just at the change over. You can get special earth leakage breakers designed to be tolerant to high frequencies and transients.

Not "leakage current" , but transients and high frequencies.  And for that you get different Class/type RCD's that is specifically designed to deal with these kinds of "False" Trips. 

That is the point I was trying to make. 

1 hour ago, Stanley said:

draw excessive earth leakage current during the change-over.

Not all of these conditions can be seen as "Drawing Excessive earth leakage current", becuase its not really current that's been drawn... 

11 minutes ago, Stanley said:

You can get special earth leakage breakers designed to be tolerant to high frequencies and transients.

@Stanley do you perhaps have a model number or some other detail on the Schneider earth leakage breaker that I need to look out for. I need to find one here in Pretoria, so just need to make sure I buy one that will actually work.

@Pipes, not really, when we have had these sort of problems in the past, we just went to one of our local suppliers here called Plumstead Electrical and bought the Schneider brand earth leakage breaker. I haven't needed one for quite a long time so I don't have any specifics. I do remember that it is much cheaper than the special breakers that are specifically designed to prevent these nuisance trips due to transients and seems to solve the problem most of the time.

3 hours ago, Stanley said:

I think @plonkster explained it quite nicely above. Many appliances (mostly electronics with switched mode power supplies etc.) have capacitors and / or MOVs from live and neutral to earth. During the change over, neutral and earth are not bonded so earth is floating, this means that the capacitors and MOVs form a voltage divider that makes the earth terminal of the appliance float somewhere around half-way between live and neutral. When you then bond the neutral and earth, those capacitors (and to some extent MOVs because they have capacitance too) then discharge to earth through the bond and this causes a small spike in earth leakage current which can trip the breaker. Also, if there is any high frequency ripple on the inverter's AC output (due to poor filtering etc.) that can also cause higher than normal earth leakage current when running from the inverter because those same capacitors and MOVs (due to their capacitance) have a lower impedance at higher frequencies. So in some cases you may have trips when running from the inverter and not just at the change over. You can get special earth leakage breakers designed to be tolerant to high frequencies and transients.

This is an excellent explanation, it goes even beyond mine. I have forgotten about the high frequency effect of filters. A capacitor has an ESR value (effective series resistance) and this is frequency dependent. It is usually chosen exactly so that it tends to shunt higher frequencies (noise) but pass only tiny amounts of the 50Hz base frequency. They do however increase your "standing loss" too. Even at a clean 50Hz sinusoidal signal, all those filters eventually add up to a couple of milliamps. In fact, a small earth leakage is considered "normal" nowadays.

I read somewhere that the reason why we have 30mA RCDs is precisely because if they made the limit any lower people would stop using them due to all the nuisance tripping.

Back when I had all my RCD problems, I temporarily removed the RCD to test it. The leakage peaked at 40mA when the inverter closed the backfeed relay, but was comfortably below 10mA under normal conditions.

I badly wanted to put a scope on it to see what that transient looked like, but 1) I don't own a storage scope (yet), and 2) I would also need a good current clamp. You're talking around 10k+ of equipment to satisfy my curiousity. I blindly bought an RCD with transient immunity and it worked, so like doctors often do: Since the medicine fixed the problem, it follows that the problem was a transient during switching.

The reason why I know it has to do with EMI filters and surge suppressors, is during testing I would disconnect those and I found that it does lower the 40mA spike I observed.