Sol-Ark 15K Battery Voltage Regulation Issues

[Brightside]

I have a Dual 15K setup that has been working for over a year now. I'm grid tie with batteries, so I use a mix of batteries and grid.

The biggest annoyance I have is the discrepancy between my Battery set points and the actual measured battery voltages, when there is excess solar.

Without enough solar to power loads, the batteries charge to my Absorb and Float set points.

When PV input power is in excess of loads, the voltage will exceed my set points by 0.4-0.5V

So, with set points of 55.2V Absorb and 54.0V Float, I hit 55.7V and 54.4V respectively with excess PV.

I thought it may be fixed with a FW update, but it had no affect. Seems like it is something that could be fixed in the FW.

I realize it may not be a "big deal", but It's nice for the inverter to do what I want, all of the time, no just some of the time.

Anyone else have the same problem, are they all this way?

 

[Supervstech]

It could be the charge wires size.

 

[Brightside]

It could be the charge wires size.

The batteries are charged through 4 paralled 2/0 Copper cables. It happens with constant voltage and no current during float.

Also, it charges and floats properly at the correct set point voltages when grid charging without excess solar.

So on an average day while floating, if the sun come comes out, the batteries will charge from 54.0V to 54.4V, then when the clouds come the float drops back to 54.0V, and will continue this for the afternoon. Also during Absorb, it does the same thing, going from 55.2V to 55.7V and back dependent on excess PV.

 

[power.esrl3]

I'm glad you brought this up. I'm tired explaining this to Deye. Might as well try your luck.

Everything gets whacky when there is excess solar power.

 

[Brightside]

Good to know I'm not the only one. I don't ever see anyone complaining about this issue. I would figure more people would be talking about/noticing this issue to get it fixed. Unless people feel a 0.4-0.5V deviation from set point is ok. I brought it up a couple separate times with Sol-Ark, but it doesn't goes anywhere. I thought it would get traction after they had to roll back my FW, and I explained to them, the only reason I updated it was to try and fix this issue, which it did not.

The other issues I have is zeroing issues. I have a phantom(offset/not zeroed) 2A battery current on 1 inverter when no current is flowing. I really noticed it, when I took all my batteries offline except 1 100Ah battery(to deep cycle it, as it has a high resistance compared to the rest) I could not get it to float as it was stuck in absorb, because even with the actual battery current at 0.0A the inverter reads 2A, I had to double the bank size to 200A in the inverter settings to get it to float with a 2% tail current.

Another minor issue is the Voltage calibration of the inverter. My master reads 0.1V below the actual and the slave 0.2V below the actual, with no current flowing at the inverter terminals. At least this offset is consistent and only the master counts.

Another issue is MPPT current zeroing. I have a few MPPT's that don't zero. This affects power production measurement. They read 0.1A and 0.2A while the other 3 read 0.0A when disconnected or in the dead of night. This makes a big difference on gloomy winter days, where solar input is limited. When 3 channels have a total of 0.5A of non existent current, this makes you production #'s way off at 400V*0.5A for 200W of phantom power when the total peak power during the day may be 1kW.

It would be nice to have user zeroing and calibration settings built into the inverter.

 

[power.esrl3]

Good to know I'm not the only one. I don't ever see anyone complaining about this issue. I would figure more people would be talking about/noticing this issue to get it fixed. Unless people feel a 0.4-0.5V deviation from set point is ok. I brought it up a couple separate times with Sol-Ark, but it doesn't goes anywhere. I thought it would get traction after they had to roll back my FW, and I explained to them, the only reason I updated it was to try and fix this issue, which it did not.

The other issues I have is zeroing issues. I have a phantom(offset/not zeroed) 2A battery current on 1 inverter when no current is flowing. I really noticed it, when I took all my batteries offline except 1 100Ah battery(to deep cycle it, as it has a high resistance compared to the rest) I could not get it to float as it was stuck in absorb, because even with the actual battery current at 0.0A the inverter reads 2A, I had to double the bank size to 200A in the inverter settings to get it to float with a 2% tail current.

Another minor issue is the Voltage calibration of the inverter. My master reads 0.1V below the actual and the slave 0.2V below the actual, with no current flowing at the inverter terminals. At least this offset is consistent and only the master counts.

Another issue is MPPT current zeroing. I have a few MPPT's that don't zero. This affects power production measurement. They read 0.1A and 0.2A while the other 3 read 0.0A when disconnected or in the dead of night. This makes a big difference on gloomy winter days, where solar input is limited. When 3 channels have a total of 0.5A of non existent current, this makes you production #'s way off at 400V*0.5A for 200W of phantom power when the total peak power during the day may be 1kW.

It would be nice to have user zeroing and calibration settings built into the inverter.

Yeah! You nailed everything, and with the MPPT too. It baffles me that no one is complaining about these issues, no wonder Deye will not focus on this.

We might need Will Prowse's or Andy's exposure for this to get fixed.

Its not ok. It complicates everything such as balancing logic, and whatever algo bms has. Even performance and battery health wise, its not ok if it doesnt respect set parameters, both in inverter and bms.

 

[DougfromdaUP]

My 12k has the same voltage overshoot problems, off by .3 volt usually. I gave up talking to them and just adjust my control setting to account for the overvoltage.

 

[power.esrl3]

My 12k has the same voltage overshoot problems, off by .3 volt usually. I gave up talking to them and just adjust my control setting to account for the overvoltage.

Yikes.

How about those people with no controls on their BMS? Usually set at 57.6V? Imagine the damage it will cause.

They sleep sound at night thinking they are doing fine because their inverter and battery is "compatible" because it is communicating.

 

[widget4145]

Hi, I posted a related topic of this issue. This issue seems to be related with Time Of Use being enabled.

This is very wrong. This should not happen. This is a huge issue that should be fixed as soon as possible.

 

[Brightside]

Yikes.

How about those people with no controls on their BMS? Usually set at 57.6V? Imagine the damage it will cause.

They sleep sound at night thinking they are doing fine because their inverter and battery is "compatible" because it is communicating.

I experienced this when I first tried closed looping my batteries to the solark. Batteries called for 56.0V to absorb. Solark ramped them up to 56.5V. This pushed them all into Over Voltage Cell disconnects at 3.90V per cell. Needless to say, that was the last of the closed loop battery comms.

 

[Tulex]

I'm not seeing this with my setup. But then, being closed loop with Seplos BMS, the BMS prevents it.

 

[2TrevorJ]

I'm not seeing this with my setup. But then, being closed loop with Seplos BMS, the BMS prevents it.

Same...the Sol-Arks hit/maintain the BMS requested voltage +/- 0.1V even with wild swings in PV production.

My batteries have a voltage display that matches the Sol-Ark display, Solar Assistant monitoring (via Sol-Ark comms), and my multimeters (again; for the most part - +/-0.1V).

 

[Brightside]

Same...the Sol-Arks hit/maintain the BMS requested voltage +/- 0.1V even with wild swings in PV production.

My batteries have a voltage display that matches the Sol-Ark display, Solar Assistant monitoring (via Sol-Ark comms), and my multimeters (again; for the most part - +/-0.1V).

Are both of you grid tied or off grid?

If grid tied, do you sell all excess production?

 

[Tulex]

Grid tied but not selling.

 

[Brightside]

Interesting. That's definitely different than how mine is working.

 

[2TrevorJ]

Are both of you grid tied or off grid?

If grid tied, do you sell all excess production?

On grid and selling

 

[Nobodybusiness]

Not seeing this on my 2 15ks but maybe it’s setup differently.

Grid tied but grid has been off since March 1st.

 

[Ampster]

My batteries have a voltage display that matches the Sol-Ark display

That is because in closed loop communicationt, the SolArk displays the voltage which the battery reports to it via the communication port. Before that my SolArk always reported about 0.3 to 0.4 volts higher than my multimeter on the battery terminals or from the battery reports. That voltage acuracy and the more acurate SOC reporting are two of my reasons for favoring closed loop communication.

 

[widget4145]

This is weird really.

Closed or open loop, doesn't matter in my case, there is always 0.5v offset below the requested voltage. Unless I disable TOU or when there is no grid.

Gridtie with selling option.

 

[Solar Guppy]

Unrealistic expectations of accuracy it would seem, 0.5V of 50 volts is 1%, that close to laboratory grade equipment and for 16S batteries 31 millivolts per cell, so its the setting that should be adjusted or the battery and it would be way to close if this knocks a BMS offline.