Sol-Ark 15k won't charge batteries higher than 110A.

[hpeyerl]

Sol-Ark 15k with 8 x Shoto 48100 (16 cell lifepo4) open-loop.

All 8 batteries are in a rack with 3 big copper bus bars at the back of the rack. 1 bus bar for grounds from each battery, one for positives, and one for negatives. Previously I had a single 2/0 from each of the positive and negative bus bars to one set of the terminals on the inverter. I had the charge limit configured to 150A and discharge limited to 80A.

Last week I added a second pair of 2/0 cables from the bus bars to the inverter.

Each of the positive 2/0 cables has a T-class fuse.

This week I added another 9 panels bringing me up to around 18kw worth of panels. (We bought an EV).

Prior to adding the second set of 2/0 cables and 9 panels, I had noticed that my battery charging was definitely limited to some ceiling by looking at the graphs in my SolarAssistant but it was nowhere near the 150A I had configured. It seemed like the Sol-Ark wasn't charging at 150A, more like 110A. As a result, since there was nowhere to put excess PV, the PV graph also looked flattened at the peaks. If I turned on a bunch of extra load (my 30A compressor), then the PV production would rise to meet the new demand, until I turned the compressor back on. This confirmed to me that the battery charging was being held back somewhere

After adding the new 2/0 cables, I bumped up the charge limit to 275A.





Yesterday was the first day with the new panels and a mostly sunny day. I noticed (from SolarAssistant) that the batteries were still only charging at 6200W even though there was plenty of PV. I turned on the compressor again just to confirm there was excess PV and it popped up to almost 12kw.

It's pretty apparently in the graphs that the charging is being limited to around 6200W and that there is excess solar available.



Since my batteries are not closed-loop, I've got the Sol-Ark configured for 'Batt-V'. It appears to be ignoring the 'Max A Charge (0-275A)' setting.



Has anyone encountered this before? Have any ideas?

 

[hpeyerl]

ok. I'm a retard.

I'm using TOU and had the Power limited to 6000W in there.

I'm positive that's where the problem is (was?).

I couldn't see how to delete this thread, otherwise I would have.

 

[hpeyerl]

Well it's mid-day and that didn't solve the problem.. my max has gone up from 6200 to maybe 6300W of charging but there's still excess solar that's being suppressed (verified by turning my compressor on).

I'm back to being stumped.

 

[Lt.Dan]

Compare your settings in Solar Assistant to mine, and see what you get:

 

[TM48]

Your BMS's may be limiting the charge current. Do you have any software to view the BMS settings?

 

[Lt.Dan]

Your BMS's may be limiting the charge current. Do you have any software to view the BMS settings?

It's open loop. The BMS cannot throttle current without communication. It should be an "all or nothing".

 

[robby]

Can you post a screen shot like this from the day when you turned on the 30A compressor and got extra PV power?

I am fairly certain this is a setting issue but I would like to see exactly what is happening on the graph.
BTW why do you have the battery internal resistance set to 8 mohm? I use 2 and I know some people use 1 mohm.
I don't think that would make a difference but I was just wondering.

 

[robby]

Okay never mind, I think I found your problem.
You need to change a setting in your battery packs.
See page 21 of the User Manual for your answer.
Helios 48100 Manual

 

[TM48]

It's open loop. The BMS cannot throttle current without communication. It should be an "all or nothing".

Some can be limited in software for discharge and charge...not dynamic but still limited.

 

[hpeyerl]

Okay never mind, I think I found your problem.
You need to change a setting in your battery packs.
See page 21 of the User Manual for your answer.
Helios 48100 Manual

Interesting! I wonder what register that is. Closest I see is 0x1013 "Average charge current setting (percentage of battery charge current target value)[Ibat_set]" and my register dump shows "Ibat_set: 1500 0x5dc"

I'll dig into this more tomorrow. This is a good lead!

(I'm not sure where the 8ohm came from. I'm not sure I would have set that myself but Carlos was helping me some last year).

 

[robby]

It's open loop. The BMS cannot throttle current without communication. It should be an "all or nothing".

Dan most of the BMS units do have current limit settings that will be used no matter how it is connected.
In closed loop mode it just sends that Info to the Inverter so that they are on the same page.
See the first three minutes of this video for more info.

 

[robby]

Interesting! I wonder what register that is. Closest I see is 0x1013 "Average charge current setting (percentage of battery charge current target value)[Ibat_set]" and my register dump shows "Ibat_set: 1500 0x5dc"

I'll dig into this more tomorrow. This is a good lead!

(I'm not sure where the 8ohm came from. I'm not sure I would have set that myself but Carlos was helping me some last year).

You should be able to get BMS software for it and just change it with a PC and a USB to RS-485 converter.
You are limited to 120 Amps in the current default settings.

 

[hpeyerl]

You should be able to get BMS software for it and just change it with a PC and a USB to RS-485 converter.
You are limited to 120 Amps in the current default settings.

I'll see if I can find the software. I don't have a Windows PC but can figure something out. I've been using a modbus cli tool with an rs485 USB transceiver and I have a PDF of all the registers.

Thanks.

 

[Lt.Dan]

Interesting, I thought all the current limiters were only used in special scenarios and only for short periods.

Learn something new every day

 

[rickst29]

It's open loop. The BMS cannot throttle current without communication. It should be an "all or nothing".

As robby described, I think that all BMS units impose maximum limits for charge and discharge for the pack. In some high-current situations, they can't respond to a short circuit fast enough to protect themselves (that's why you also need fast-acting fuses),. But in extended periods of high current, they will be successful -- and maybe in a bad way, with cycling between 'allowed' versus 'disallowed' charging states.

I don't the BMS which is used in your pre-packaged "battery packs". My own batteries (16S-2P, EVE 280Ah-V3 cells) will use large ANT BMS units. They allow up to 400A continous discharge each, but that high value is probably not "default", and the maximum charging limit is almost certainly a lot lower. I got bigger-than-needed BMS units to to keep them running cool and well below rated capacity levels.

On my cells, the maximum allowed charging rate is also temperature dependent. My BMS units can not vary the limit according to temperature, and the maximum rate (0.5C = 140 Amps per pack) cannot be allowed below about 8 degrees C, or above 30 degrees C. I allow a minimum charging temp of only 5 C, and EVE specifications show that I need to use a lower charging rate (.25 of Capacity, around 70A max per pack and 140A total).

My parallel EVE packs are 540Ah, two thirds of your total size. But If you follow my guidelines from EVE specifications, you don't want more than 210A total anyway- so your 'shortfall' is smaller than you think. That's for the case of very capable EVE cells (V3) with 3000N compression, your 'pre-built' battery packs might not be compressed or contain cells as good as mine.

8 small packs with cheaper 100Ah cells, uncompressed, and generic BMS units not-yet-programmed probably "play it safe" by imposing severe restrictions on charge current, maybe lower than 25A limit which I'd consider to be reasonable. But Sol-Ark charging is also questionable, and it is utterly weird that their User Interface programming offers the same high maximum for CHARGING as DISCHARGING. I recommend that you should not exceed about 200A total.

I also wonder if conversion (from higher voltage MPPT strings to lower voltage DC charging) has a lower limit than 12,000 watts from battery discharge into AC. For the case of 51 volts batteries, 12kW charging power is only 235A at 100% efficiency, so the 285A limit in the specification-setting GUI panel is appears to be bogus.

 

[robby]

I'll see if I can find the software. I don't have a Windows PC but can figure something out. I've been using a modbus cli tool with an rs485 USB transceiver and I have a PDF of all the registers.

Thanks.

In the meantime lower the Inverter Charge rate down to 110A

 

[robby]

Interesting, I thought all the current limiters were only used in special scenarios and only for short periods.

Learn something new every day

I use to think the same thing.
I don't think it's a good idea to let the BMS keep trying to throttle the current as it must be generating a lot of heat from the Mosfets.

 

[Tulex]

Can you post a screen shot like this from the day when you turned on the 30A compressor and got extra PV power?
View attachment 296638
I am fairly certain this is a setting issue but I would like to see exactly what is happening on the graph.
BTW why do you have the battery internal resistance set to 8 mohm? I use 2 and I know some people use 1 mohm.
I don't think that would make a difference but I was just wondering.

I believe 8 mohm is the default setting. I never changed mine, and that's what it's set at.

 

[robby]

I believe 8 mohm is the default setting. I never changed mine, and that's what it's set at.

Fortress Power had told me to use 2 mohm.
It could be because the power goes through relays and not Mosfets.

 

[hpeyerl]

We have a winner!

It was indeed the register I mentioned above which was set to 15A (which x8 is 120A). I bumped the register on each pack up to 30A and then bumped the inverter up to 175A. I managed to see over 150A of charging earlier but since it appears to be laundry day, it's sporadic. But these are definitely the highest charge rates I've seen so far.

Thanks everyone and especially Robby who pegged it perfectly!

For anyone who might find this thread in the future:

for  I in $(seq 2 9); do echo $I;  bin/modbus -s $I -b 9600 -p 1 -P n  /dev/ttyUSB0 h@4115/H=0xbb8 ; done