Why does my main house consumer unit trip when the inverter switched to AC backup?

[marconi]

Hello Marconi

The issue with the pv input is this.

When I measure the pv input from the roof at the point it connects into the cabinet input connections (lower left in picture) I get the expected 253 vdc.
When the inverter is switched on but the isolator in the cabinet is off (lower right in picture), the voltage from the pv remains constant, when I then turn on the isolator in the cabinet for the pv, the voltage goes up to between 350 and 450 vdc.

When this happened I then isolated the pv supply using the external isolator I have fitted before the cabinet. Which effectively removes any input to the inverter from the pv, I then measured the voltage at the input connection to the cabinet and found the 100 to 200 vdc appearing at the input of the pv in the cabinet. If I turn off the isolator in the cabinet the voltage drops to zero, this led me to conclude that the dc voltage was coming from the inverter out through the input terminals in the cabinet.
Now I have no idea of how this is possible as this would indicate that the inverter is working totally incorrectly or that the inverter is somehow sending power to the point it is supposed to be taking it in.

When the inverter is on and working it only shows 252 vdc input from the pv (see Right hand pic) on the front panel of the inverter and it was not until I re measured the voltages of all of the terminals in the cabinet whilst trying to find any indication of what may be causing the problems I am having I discovered this issue.

I am somewhat concerned that this may be down to a wiring issue in the cabinet or the inverter but it is difficult to determine which without removing more wiring. Perhaps when I come to remove the inverter this weekend I will measure the voltage at the input connection to the actual inverter with all wiring removed to see if it is as above, when there are no other routes for the voltage from the inverter itself.

Feel free to ask for more information if this is required

Many thanks as ever and Regards

Tonyboy

View attachment 106299 View attachment 106300

I will do the same measurements on one of my brother in law’s solar leccy vans and see what I measure. Don’t think it is a problem but best to be sure.

 

[ukiri]

Give us a clue, how old is the MEM distribution board and installation? 25 to 35 years old? And going off the PV system diagram in The Big Blue Book your array should be connected to the main earth bar, it`s not for lightning protection but electrical safety. Chances are you are getting a cumlative "overload " of the earth fault detection on the MEM rcd from all the circuits and devices connected. And if there`s a B type rcd in the inverter cab, the AC shouldn`t be upstream of it as indicated by the hieracy of rcds diagrams. You might even have to fit an isolating transformer. The suppliers instructions should have detailed all this.

 

[Tonyboy]

I will do the same measurements on one of my brother in law’s solar leccy vans and see what I measure. Don’t think it is a problem but best to be sure.

Hi Marconi
I am much obliged, I will do the tests tomorrow morning before I go out at lunchtime. Hopefully the results will elicit an answer.

Many thanks

Tonyboy

Give us a clue, how old is the MEM distribution board and installation? 25 to 35 years old? And going off the PV system diagram in The Big Blue Book your array should be connected to the main earth bar, it`s not for lightning protection but electrical safety. Chances are you are getting a cumlative "overload " of the earth fault detection on the MEM rcd from all the circuits and devices connected. And if there`s a B type rcd in the inverter cab, the AC shouldn`t be upstream of it as indicated by the hieracy of rcds diagrams. You might even have to fit an isolating transformer. The suppliers instructions should have detailed all this.

 

[Tonyboy]

Give us a clue, how old is the MEM distribution board and installation? 25 to 35 years old? And going off the PV system diagram in The Big Blue Book your array should be connected to the main earth bar, it`s not for lightning protection but electrical safety. Chances are you are getting a cumlative "overload " of the earth fault detection on the MEM rcd from all the circuits and devices connected. And if there`s a B type rcd in the inverter cab, the AC shouldn`t be upstream of it as indicated by the hieracy of rcds diagrams. You might even have to fit an isolating transformer. The suppliers instructions should have detailed all this.

Hello ukiri
Many thanks for the reply
Yes the MEM is probably at least 25 year's old. I have a separate earth rod for the batteries adjacent to the kit in the external cupboard and as all of the supplies from the solar consumer unit are completely separate from the MEM I was going to connect the panels back to this and not to the incoming supply earth.
As for suppliers instructions well I got the inverter and Bluetooth battery monitor manuals and nothing else, hence the ongoing thread and the advice I have been getting.
Many thanks for the reply and if the tests I do today do not help I will look into an isolating transformer. I am assuming this is something you fit between the ac input to the cabinet and the ac input to the invetrer and the isolation part is isolation of the earth back to the grid supply?
Regards
Tonyboy

 

[Tonyboy]

I will do the same measurements on one of my brother in law’s solar leccy vans and see what I measure. Don’t think it is a problem but best to be sure.

Good Morning Marconi

I have carried out the requested tests. Multi- meter set on 200ohms lowest setting.

Measured between N and E on output terminals reading was 00.2 not very high.
The measurement between the two cables with output mcb on and output rcd off, no reading at all just the decimal point.

Set up the inverter to charge from the AC input but I have left the PV isolated as I am still not sure if the readings at the input terminals are an issue.

Timer came to start and ac started charging the batteries at 10 amps (I have set this limit), mains in house not tripped and still charging as I type.

Not sure why there is no reading on the two cables (black and Blue) as the meter is showing 00.1 when I touch the terminals together as a test.

I am going to let the inverter charge the batteries for the period I have set to see if it keeps going.

I hope the results are able to help and I look forward to hearing what you think is needed next.

As ever many thanks for the help.

Regards

Tonyboy

 

[marconi]

What I am working up as a solution is a very fast ac changeover switch to connect the ac output socket to either the inverter or the incoming grid mains but never both. Would you be happy with the changeover which happens within a few cycles of the ac mains so should not be noticeable?

 

[marconi]

Good Morning Marconi

I have carried out the requested tests. Multi- meter set on 200ohms lowest setting.

Measured between N and E on output terminals reading was 00.2 not very high.
The measurement between the two cables with output mcb on and output rcd off, no reading at all just the decimal point.

Set up the inverter to charge from the AC input but I have left the PV isolated as I am still not sure if the readings at the input terminals are an issue.

Timer came to start and ac started charging the batteries at 10 amps (I have set this limit), mains in house not tripped and still charging as I type.

Not sure why there is no reading on the two cables (black and Blue) as the meter is showing 00.1 when I touch the terminals together as a test.

I am going to let the inverter charge the batteries for the period I have set to see if it keeps going.

I hope the results are able to help and I look forward to hearing what you think is needed next.

As ever many thanks for the help.

Regards

Tonyboy

These readings indicate the output N-E link is inside the Axpert box.

 

[marconi]

Good Morning Marconi

I have carried out the requested tests. Multi- meter set on 200ohms lowest setting.

Measured between N and E on output terminals reading was 00.2 not very high.
The measurement between the two cables with output mcb on and output rcd off, no reading at all just the decimal point.

Set up the inverter to charge from the AC input but I have left the PV isolated as I am still not sure if the readings at the input terminals are an issue.

Timer came to start and ac started charging the batteries at 10 amps (I have set this limit), mains in house not tripped and still charging as I type.

Not sure why there is no reading on the two cables (black and Blue) as the meter is showing 00.1 when I touch the terminals together as a test.

I am going to let the inverter charge the batteries for the period I have set to see if it keeps going.

I hope the results are able to help and I look forward to hearing what you think is needed next.

As ever many thanks for the help.

Regards

Tonyboy

Did axpert box enter bypass mode while charging battery from mains? If not sure send me picture of display.

 

[Tonyboy]

What I am working up as a solution is a very fast ac changeover switch to connect the ac output socket to either the inverter or the incoming grid mains but never both. Would you be happy with the changeover which happens within a few cycles of the ac mains so should not be noticeable?

Hello Marconi

I thank you for the reply.

I will admit straight away that I have no idea of what one of those does.

Would this be something I could fit into the cabinet local to the inverter and controls?
The speed at which it changes over to some extent does not matter as it will not interfere with any items powered by the solar as they will only be things that are used when I am using the pv/batteries to power them.

My idea in the beginning was to have a method by which to charge the batteries overnight from the grid supplying the inverter, acting as a battery charger.

I would not have any need for the ac to supply any of my solar outlets as they would not be used at this time.
If such a switch could provide power to the inverter and the inverter could charge the batteries (if needed) and only do this then I would be quite happy with it.

I am assuming the issue is currently that when the grid supply is activated on by the inverter, it both supplies the inverter and then on to the output of the cabinet via the RCD/MCB in the cabinet in bypass mode before the consumer unit for
the solar.

I will have no need for the supply to be bypassed to the consumer unit and if it was just available from the inverter/charger for the batteries that would be more than satisfactory. My solar outlets are for a small number of items in the home that I would prefer to power by pv and in general would only be used when I am in the home and thereby reduce my consumption of the grid supplied power. Fanciful I guess but it was the premise of the initial purchase.

Please feel free to send me more information as and when you are able. I am off to work fairly soon and I will probably not be able to respond until late tomorrow or Friday late afternoon.

I am as ever very grateful for you help

Regards

Tonyboy

 

[Tonyboy]

Did axpert box enter bypass mode while charging battery from mains? If not sure send me picture of display.

Hello Marconi
have attached this, currently not charging as batteries are nearly full.

Regards

Tonyboy

 

[marconi]

Axpert Inverter - Who has stock - Page 5 - https://www.4x4community.co.za/forum/showthread.php/223750-Axpert-Inverter-Who-has-stock/page5

My train journey was spent trawling the internet for solar leccy using an axpert box exhibiting up front rcd trips when the box connects output LN to input LN in bypass/line mode. You will read in the link above something very similar to what we have been investigating in regard to an N and E connection on the output side and our discovery that it is done inside the box.

Others have had the same problem when this off grid solar leccy Axpertbox like in your set up is connected to the grid mains supply which already has N and E connected together before the main distribution board RCD.

Further evidence in this link above that there is indeed an internal NE connection to create TNS at the ac output - as we detect with the ohmmeter measurements.

As a final check would you turn off all power from solar, battery and mains and then reconnect the black and blue cables to the Axpert box ac output terminals.

Then with ac output mcb on and ac output rcd off repeat the charging procedure you just did and confirm the batteries are charging for a say a minute.

Now, open the main switch in your off grid consumer unit which supplies the black 13A outlets annd ensure all its mobs are off. Thus no loads are connected the cabinet ac output socket.

Close the cabinet ac output rcd - I reckon the home mem rcd should not trip. Go to off grid consumer unit and close its main switch. With all off grid mobs off there is still no load connected the cabinet ac power output socket so I reckon mem rcd will still say closed.

Have some loads plugged in to off grid sockets and turned on. Turn on the off grid mcbs one by one checking between each mcb whether the house mem rcd has tripped or not. Sooner or later I expect the closure of one of these mobs to apply a load which will create an unbalanced current through mem rcd causing it to trip.
I hope this is clear enough.

 

[marconi]

In the image below of your Axpert box's lcd display the arrow I have circled in blue appears when bypass/line is selected to indicate that the input ac LN are directly connected to the ac output LN. Another giveaway of bypass/line mode is when the input and output voltages and frequencies are exactly the same values. This, 248V and 49.9Hz in this instance. (ps: Nice healthy mains supply voltage - our home sits around 242V. If you ever find that the Axpert will not display incoming mains it might be because the cabinet's voltage relay detects it as too high or too low and so will not energise the ac input contactor - let me know and I will advise you what to do).

 

[Tonyboy]

In the image below of your Axpert box's lcd display the arrow I have circled in blue appears when bypass/line is selected to indicate that the input ac LN are directly connected to the ac output LN. Another giveaway of bypass/line mode is when the input and output voltages and frequencies are exactly the same values. This, 248V and 49.9Hz in this instance. (ps: Nice healthy mains supply voltage - our home sits around 242V. If you ever find that the Axpert will not display incoming mains it might be because the cabinet's voltage relay detects it as too high or too low and so will not energise the ac input contactor - let me know and I will advise you what to do).

Hello Marconi
just back in, so I am going to carry out the test laid out in the previous reply in a few minutes.

Quite interesting to see others have had some issues with the NE link.

Whilst laying awake the other night I was thinking of how to rectify the issue with my limited knowledge. It is somewhat simplistic but I thought I would share it with you so you could tell me why it is not possible.

Having disconnected the L and N output from the inverter for out tests I wondered if it was possible to insert a timer in between the inverter output and the protection devices the inverter serves on the board, so that at a time point prior to the AC being timed to come on for charging, say 1 hour before, the output was disconnected on the timer and therefore would not affect the rcd and mcb on the output of the cabinet as the output would effectively be off, Then at a time when the AC charging stopped charging the batteries again say 1 hour later the timer then re connects the L and N to the output of the cabinet ready for use through the solar CU. In both circumstances there will be no load on the output of the inverter prior to or after the charging has taken place.
I had even considered just fitting a double pole isolator in the cabinet on the L and N of the inverter output with which to mechanically separate it from the incoming ac supply that is activated by the timer function in the inverter.
I will get back to you as soon as I have done the tests.
Regards

Tonyboy

 

[pc1966]

I wondered if it was possible to insert a timer in between the inverter output and the protection devices the inverter serves on the board, so that at a time point prior to the AC being timed to come on for charging, say 1 hour before, the output was disconnected on the timer and therefore would not affect the rcd and mcb on the output of the cabinet as the output would effectively be off, Then at a time when the AC charging stopped charging the batteries again say 1 hour later the timer then re connects the L and N to the output of the cabinet ready for use through the solar CU.

Is the issue not fundamentally one of when you have no grid AC that you need the N & L linked?

If that is detected as part of any supply transfer switch, which I assume HAS to be there so you don't back-feed to mains, can't it operate a relay to link N&E when running on the inverter?

 

[marconi]

Hello Marconi
just back in, so I am going to carry out the test laid out in the previous reply in a few minutes.

Quite interesting to see others have had some issues with the NE link.

Whilst laying awake the other night I was thinking of how to rectify the issue with my limited knowledge. It is somewhat simplistic but I thought I would share it with you so you could tell me why it is not possible.

Having disconnected the L and N output from the inverter for out tests I wondered if it was possible to insert a timer in between the inverter output and the protection devices the inverter serves on the board, so that at a time point prior to the AC being timed to come on for charging, say 1 hour before, the output was disconnected on the timer and therefore would not affect the rcd and mcb on the output of the cabinet as the output would effectively be off, Then at a time when the AC charging stopped charging the batteries again say 1 hour later the timer then re connects the L and N to the output of the cabinet ready for use through the solar CU. In both circumstances there will be no load on the output of the inverter prior to or after the charging has taken place.
I had even considered just fitting a double pole isolator in the cabinet on the L and N of the inverter output with which to mechanically separate it from the incoming ac supply that is activated by the timer function in the inverter.
I will get back to you as soon as I have done the tests.
Regards

Tonyboy

I have in mind something along those lines. But as you have shown by your persistence investigating this problem already we need to work out the 'whys' first lest we contrive a bodge solution that masks something unsafe.

Are you on good terms with your neighbours?

 

[Tonyboy]

In the image below of your Axpert box's lcd display the arrow I have circled in blue appears when bypass/line is selected to indicate that the input ac LN are directly connected to the ac output LN. Another giveaway of bypass/line mode is when the input and output voltages and frequencies are exactly the same values. This, 248V and 49.9Hz in this instance. (ps: Nice healthy mains supply voltage - our home sits around 242V. If you ever find that the Axpert will not display incoming mains it might be because the cabinet's voltage relay detects it as too high or too low and so will not energise the ac input contactor - let me know and I will advise you what to do).

Hello again Marconi

I followed the instructions and got the AC to charge the batteries with the USB timer set and it was charging for a short while.
I followed the instructions one at a time and then turned on the Solar Consumer unit. I then ran a kettle on the kitchen circuit and the microwave on a separate circuit serving the extension this took the output to 3Kw so I added an induction hob which took the output to 5.1 Kw, the max of this inverter being 5.5Kw I believe.

At this point the MEM rcd has not tripped and until I get another couple of sockets added to the next MCB circuit I can only add to the two I have at present with more items. I am surprised the mem did not trip but this may be as a result of only 2 mcb currently pulling power through the consumer unit.

Let me know your thoughts regarding this test and I look forward to trying some more experiments before I remove the unit to return to the supplier.
I have been testing all of my pv wiring in a last ditch attempt to see what is causing my dc.dc overcurrent issue but with no obvious problems with the wiring I am resigned to returning it to the supplier next week.

regards and thanks as ever for the help

Tonyboy

 

[marconi]

When do you intend to return the unit?

No reply yet from Voltacon - will chase up next week.

I will check out the 'lie of the land' at home with regards to my wife and consider visiting you if you wish. Perhaps you might ask your wife/partner if she would agree to me visiting bearing in mind you both have a new baby to look after. I will stay at a travelodge or go home the same day.

 

[Tonyboy]

Is the issue not fundamentally one of when you have no grid AC that you need the N & L linked?

If that is detected as part of any supply transfer switch, which I assume HAS to be there so you don't back-feed to mains, can't it operate a relay to link N&E when running on the inverter?

Hi pc1966
Currently the system is only suppled with ac when the inverter is set to USB and the timer is used to charge the batteries overnight.
When this is happening, there are no loads on the output for the inverter to supply so if the ac output was disconnected the ac would only be charging the batteries. As you can see from the previous I have been very lucky to have been given a lot of advice and I am still working towards a solution.
Unfortunately I am not an electrician as you may have guessed so I am working through some tests I have to see if the problem has a simple solution.
The whole system is off grid and has no capability of feeding power back to the grid. The cabinet and controls installed shuts off the supply to the inverter if the ac power fails and isolates the cabinet from the grid. The output of the inverter is to a separate consumer unit supplying separate solar supplied outlets.
Many thanks for the reply
Tonyboy

 

[marconi]

It seams to me now like their are two problems - one to do with the Axpert box ac output N-E internal link and the other to do with the 'ac input' supply somewhere along its full extent to the supply transformer.

 

[Tonyboy]

When do you intend to return the unit?

No reply yet from Voltacon - will chase up next week.

I will check out the 'lie of the land' at home with regards to my wife and consider visiting you if you wish. Perhaps you might ask your wife/partner if she would agree to me visiting bearing in mind you both have a new baby to look after. I will stay at a travelodge or go home the same day.

Hello again Marconi

I was planning to take it back next week as I am on holiday for a week, but I can wait longer if necessary.
I do not have a baby by the way, Well unless my wife has a big secret to disclose, so I will speak to my wife re a visit if you think this would be a help.

Currently I do have a very unwell father, so we are quite often out to assist him at present. I don't know how far away you are from Somerset and I think it should be a last resort if you were happy to do so, and I would certainly expect to cover your costs for such.

In response to the previous reply I missed, no I do not have any contact with the neighbours, suffice to say our house is 200 years old and it has a sperate supply from the pole over the road into my consumer unit and I believe the same applies to the house next door. I have no wish to let my neighbours know what I am installing.

In the interim what were your thoughts about finding a way to disconnect the output from the inverter as a means of isolating the input and output from the cabinet? I have read through the manual but could not see a way to disable line or standby mode and this I assume would stop the output being applied to the cabinet controls and thereby the output to the consumer unit when in USB timed mode.

Regards and thanks

Tonyboy