Midnight solar DC. Breakers

Had a situation last night.

Just got in the house 7:30 pm.
Turn on lights and suddenly power went out.

I knew we weren’t running any heavy loads.

Went out to the inverter and the DC breaker going to master inverter was tripped.
Reset everything and got it back up.

Look at SA and saw something weird.

Inverter 1 voltage dropping while inverter 2 stay the same.

Remember no heavy loads.voltage just dropping on inverter 1 but not 2 and both going to same battery bank through midnight Solar combiner.

Each inverter has 250 amp Carling ( midnight) breaker at the combiner.
The voltage starting dropping few minutes before we got home to lowest I see is about 42v then the breaker tripped.

Not sure what to make of it.
No shorts or breaks in the 4/0 copper cable.
When I reset breaker it appears fine.

Lower voltage draws higher current.

But no big loads you say?

Does these breakers have Remote Trip?

Did BMS shut off, then reconnect, tripping breaker due to inrush?

Sag of voltage on one inverter not other could be due to flaky connection. When it decided to improve, had inrush from one inverter's caps to the other?
Or, maybe it sagged due to breaker open.

Did the breaker somehow end up with poor internal contact, leading to heat, voltage drop, then thermal trip?

Hedges said:

But no big loads you say?

Click to expand...

Nothing but some lights.

Hedges said:

Does these breakers have Remote Trip?

Click to expand...

No remote trip.

Hedges said:

Did BMS shut off, then reconnect, tripping breaker due to inrush?

Click to expand...

The battery voltage stayed constant on inverter 2 and no errors on any of the batteries.

Hedges said:

Sag of voltage on one inverter not other could be due to flaky connection. When it decided to improve, had inrush from one inverter's caps to the other?
Or, maybe it sagged due to breaker open

Click to expand...

It opened finally but until then just lost voltage.

The connections are tight and bright shiny copper.

Look like a breaker temporary failure ?

SparkyJJO said:

Did the breaker somehow end up with poor internal contact, leading to heat, voltage drop, then thermal trip?

Click to expand...

Not sure it’s been on for 6 month or more.
When I reset it looks fine.

Maybe inverter decided to shoot through H-bridge. And breaker saved it.
Not supposed to happen.

Hedges said:

Maybe inverter decided to shoot through H-bridge. And breaker saved it.
Not supposed to happen.

Click to expand...

I don’t know but it concerns me when I can’t find a reason for a failure.

Everything appears to be working normally today?

Agreed, flaky failures are concerning.

Two paralleled inverters? Did they loose synchronization and play tug-of-war? Did one connect to the grid, or grid connected to output of inverter?

Hedges said:

Agreed, flaky failures are concerning.

Two paralleled inverters? Did they loose synchronization and play tug-of-war? Did one connect to the grid, or grid connected to output of inverter?

Click to expand...

2 sol-ark 15s in parallel.

Grid is off so no way for them to connect right now.

Inverter 1 is master so when it went down 2 followed suit.

Synchronization looks consistent .

Are you able to catch the battery voltage with an external meter at all? I see you are using Solar Assistant which pulls from our ModBus readings. If there was a problem with the Sol-Ark voltage readings on the Master inverter, Solar Assistant would show those readings bad too. We would want to see external voltage readings in the future to see if they corroborate what the Sol-Ark is reporting or if the inverter is having a logic issue and showing false low values.

Make sure the inverters are up to date and maybe try and put some heavy loads on it to see if you are noticing extremely different voltage readings between the two inverters.

Separately, perhaps a BMS problem? Are you in Closed loop comms?

Carlos_Sol-Ark said:

Are you able to catch the battery voltage with an external meter at all? I see you are using Solar Assistant which pulls from our ModBus readings. If there was a problem with the Sol-Ark voltage readings on the Master inverter, Solar Assistant would show those readings bad too. We would want to see external voltage readings in the future to see if they corroborate what the Sol-Ark is reporting or if the inverter is having a logic issue and showing false low values.

Make sure the inverters are up to date and maybe try and put some heavy loads on it to see if you are noticing extremely different voltage readings between the two inverters.

Separately, perhaps a BMS problem? Are you in Closed loop comms?

Click to expand...

No BMS problems with 20 different batteries.
The slave inverter sees no issues.

Run voltage control.

I wasn’t able to read the voltage with a meter during the event however when reading with a meter the master inverter always reads .3 to .4 less than the meter and the slave inverter on the screen and SA.

This has always been an issue and I just ignored it.

The meter will say 53.4vdc when reading at the battery contacts on both inverters.

Master inverter says 53vdc on screen and SA.
Slave inverter says 53.4 on screen and SA.

I even read 53.4 V on the opposite side of the DC breaker in the inverter..


So master inverter has never read correct voltage from the get go.

I run the whole house on them daily.
Up to 24kw.

No it is not a wiring or BMS or connector issue as I read the same voltage at both inverter inputs.
It is just being misreported on the master inverter screen and SA.


I ask for update a couple months ago and I was told at current version.

I don’t think a firmware update is going to fix a faulty voltage sensor.

You are welcome to look in them and see what’s what.

Nobodybusiness said:

No BMS problems with 20 different batteries.
The slave inverter sees no issues.

Run voltage control.

I wasn’t able to read the voltage with a meter during the event however when reading with a meter the master inverter always reads .3 to .4 less than the meter and the slave inverter on the screen and SA.

This has always been an issue and I just ignored it.

The meter will say 53.4vdc when reading at the battery contacts on both inverters.

Master inverter says 53vdc on screen and SA.
Slave inverter says 53.4 on screen and SA.

I even read 53.4 V on the opposite side of the DC breaker in the inverter..


So master inverter has never read correct voltage from the get go.

I run the whole house on them daily.
Up to 24kw.

No it is not a wiring or BMS or connector issue as I read the same voltage at both inverter inputs.
It is just being misreported on the master inverter screen and SA.


I ask for update a couple months ago and I was told at current version.

I don’t think a firmware update is going to fix a faulty voltage sensor.

You are welcome to look in them and see what’s what.

Click to expand...

If the master inverter is consistently 0.4V off I can send a calibration command to fix this for you, but I don't think that is related to the shutdown and large voltage drop you saw. Sometimes the voltage sensors just read slightly differently. I'm hoping its just a one-off event, but as far as troubleshooting that large voltage drop-off, I think the only thing today is to try to get more data when it happens again. If it happens again.

If you are connected to wifi and have the master inverter SN number DM it to me and I can recalibrate the master to read 0.4V higher for you.

@Nobodybusiness what did SA show for amps/watts on output side of each inverter? This is a stretch here, but it wouldn't be impossible for inverter 1 to have been pushing power to inverter 2. That would/could have caused voltage at inverter 1 terminals to drop and inverter 2 terminals to stay roughly the same. (Because inverter 1 would/could be drawing lots of power off the DC bus and inverter 2 would/could be pushing lots of powet to DC bus, but because batteries would see a small net loss the voltage wouldn't actually rise on inverter 2. Although I guess it theoretically could.....)

Kind of a complicated thought process there.... but simply checking SA's recorded "critical load" and/or "battery current" on each inverter would rule it out.

The only way this would/could be the issue is if the inverters were to run very slightly out of sync with each other. Low power draw is more likely to cause power "recycling" than high power draw on stacked inverters. I have seen power "bounce" happen on double Schneider XW setups. 0-400 watts is the worst point on Schneider XWs and sometimes causes quite a bit of power "bounce" between the 2 units, to the point of causing lights to flicker.

Cmiller said:

what did SA show for amps/watts on output side of each inverter? This is a stretch here, but it wouldn't be impossible for inverter 1 to have been pushing power to inverter 2.

Click to expand...

How does that work?
Only tied together at the outputs.

Cmiller said:

That would/could have caused voltage at inverter 1 terminals to drop and inverter 2 terminals to stay roughly the same.

Click to expand...

If it is pulling enough power to bring down the voltage on a 6000ah battery then inverter 2 would have seen the same thing as they are hooked to the same battery bank?

Cmiller said:

The only way this would/could be the issue is if the inverters were to run very slightly out of sync with each other. Low power draw is more likely to cause power "recycling" than high power draw on stacked inverters.

Click to expand...

It was definitely low power as just some lights were on.

Whatever the cause, a reset of both inverters has seemed to corrected it.

I don’t like not knowing what caused this though.

Nobodybusiness said:

How does that work?
Only tied together at the outputs.

Click to expand...

If one inverter is just slightly "ahead" as far as the phase angle, it will have slightly higher voltage at any given time as compared to the other one. This causes power to flow to the lower voltage unit and get "recycled" by pushing back towards the 48v battery bus. Load sharing is also related to just plain the output voltage of each unit. This can only be tested with the outputs disconnected from each other and then checked in the inverter itself.

On Schneider XWs there is a specific method to adjust output volts, to improve load sharing.

Nobodybusiness said:

If it is pulling enough power to bring down the voltage on a 6000ah battery then inverter 2 would have seen the same thing as they are hooked to the same battery bank?

Click to expand...

You can still have a different voltage at the battery terminals of the 2 units if they have differing DC amps. E.g. one has high positive amps while the other has high negative amps.

Nobodybusiness said:

It was definitely low power as just some lights were on.

Click to expand...

As I mentioned, with Schneider XWs we saw more load sharing issues under very low load, than under high load.

Nobodybusiness said:

Whatever the cause, a reset of both inverters has seemed to corrected it.

Click to expand...

Nobodybusiness said:

I don’t like not knowing what caused this though.

Click to expand...

I know what you mean. Sometimes it's almost better to have a complete failure than a random glitch!

@Cmiller is this a potential issue on a pair of single phase inverters on a split phase system as well or isolated to stacked split phase units?

Swap breakers between inverters, just in case this happens again. If problem moves its the breaker, if not then inverter.

G00SE said:

@Cmiller is this a potential issue on a pair of single phase inverters on a split phase system as well or isolated to stacked split phase units?

Click to expand...

This would apply any time the outputs are in parallel. For instance when using 2x victron 120v inverters that are 180° phase angle from each other to produce 120/240V split phase, you dont have to deal with this. Now if you have 2x single phase units (whether they are 120V or 230V) that are running in parallel on a single leg and sharing the load, then you have this potential, as well as when stacking split phase units directly.

We never had much issues with Schneider XW stacked units, and when we did it was mostly flickering issues. On Sol-Ark stacked setups we haven't run into this issue at all, but we also don't have very many stacked Sol-Ark systems out there yet.

Many of the "old style low frequency" inverters (Sch, Victron, Outback...) have output voltage adjustment options, which can help this.