diy solar

diy solar

Top balancing, recommendations (fastest or best practices)

Alfalfameister

New Member
Joined
Feb 21, 2020
Messages
50
Ordered from Amy (Luyuan instead of Xuba) 32 280Ah cells. They're already in my country, so should be delivered to my home soon.

Here's my plan:

PLAN A: (all while slightly compressed using rigid plywood at the ends, and loooong hose clamps top, middle, and bottom part).

1. Parallel them and "balance" them au naturel overnight - not really balancing, just getting the voltages in the same ballpark.

2. Disassemble then connect them 2P16S, slap a BMS (the Xiaoxiang / Overkill version), charge them with my 48v charger (maybe at 20A? 30A? can do up to 60A) at XXX voltage (was thinking 56?) until it says fully charged OR one of the cells hits HVD.

3. Disassemble them, then parallel them again (32P) and use a benchtop power supply (20 amp rating, but plan on using only 16A) and top balance to 3.6 (am thinking to just 3.60 for margin of safety (in case voltmeters are off by that little bit) instead of 3.65? Am I wrong to forego the last .05v?).

4. Finally, re-assemble them to it's final configuration (2P16S) and slap on the BMS again (120A rated BMS).

Questions for Plan A:

- is Step 1 necessary?
- what voltage and current should I charge them for Step 2?
- for Step 3, 3.60 or go all the way to 3.65

PLAN B:

1. Connect them all at 32P, use benchtop power supply at 16A and either 3.60 or 3.65V and just wait about two weeks (assuming cells arrived at 50% SoC).

Same question for Plan B: 3.60 or 3.65v?

With Plan A, I expect to save at least a week to 10 days, but Plan B looks a lot more straightforward (just time required). I figured, I waited 2 months for the batteries, hope they can last 3500-6000 cycles or about 10+ years (DoD 70-80%, likely from 90% SoC to 20% SoC, and charge ratings 100A max (or less than 0.2C for a 560Ah pack) and discharge of max 120A, but 80% of the time, likely at only 30A), so waiting two weeks to top-balance them with Plan B should be worth the wait.

Comments, suggestions, criticisms welcome (just don't use harsh words if they're criticisms, haha). :)

Thanks!
 
1. if less than 0.2V difference between cells, skip step 1

2. 60A, 3.60/3.65 really doesn't matter.

3. if you can get 20A, use 20A. However, you'll likely get well under 16A due to the very low voltage/high current voltage drop.

Plan B:

See #3 above.

EDIT: Will endorses the series charger w/BMS to expedite balancing.
 
Last edited:
Yes I would definitely series charge them or it may take forever to top balance them in parallel. Even if half full it will take more than 10 days if you set your power supply to 3.65 as the amps will decrease rapidly as voltage increases. Im willing to bet it will take about 14 days.
 
Hi,

When I receive my 18 x 280Ah EVE cells, I had planned to use the series method. I have no way to do parallel.

I'll build the bank, then connect it to my SCC, let the BMS do the top balancing. Start balancing at whatever it is they arrive at as a starting point, probably around 3.2V/cell (57.6V bank), slowly raise the voltage up to 3.65/cell (65.7V bank). Might take a couple of days. I can limit the charge current for the first few hours, though if I start first thing in the morning that will be low anyway.

The BMS should protect if there are any wayward cells. It will test the BMS at the same time.

I'll use an old(ish) phone to BT to the BMS, then use Teamviewer so I can watch it from the lounge room, rather than sitting in the garage all day.

Don't know if the BT app has an audible alarm for faults, hope so.

dRdoS7
 
When I receive my 18 x 280Ah EVE cells, I had planned to use the series method. I have no way to do parallel.

There is always the option to charge each cell fully before hooking them in series.

I'll build the bank, then connect it to my SCC, let the BMS do the top balancing. Start balancing at whatever it is they arrive at as a starting point, probably around 3.2V/cell (57.6V bank), slowly raise the voltage up to 3.65/cell (65.7V bank). Might take a couple of days. I can limit the charge current for the first few hours, though if I start first thing in the morning that will be low anyway.

I don't know what BMS you are planning on using. If the BMS has passive balancing I think it will take much longer than a couple of days. There are reasons parallel top balancing is recommended and there are different ways of accomplishing it.

I will link to a good article from Orion BMS that explains it.

 
Hi,

There is always the option to charge each cell fully before hooking them in series.

That's not an option for me, I have no way to charge an individual cell. I can charge them in 12V batches, but wouldn't have a suitable BMS.

I don't know what BMS you are planning on using. If the BMS has passive balancing I think it will take much longer than a couple of days. There are reasons parallel top balancing is recommended and there are different ways of accomplishing it.

The BMS has active balancing @ 2A, so hopefully it will handle it.

Only way to see if it works is to try it.

dRdoS7
 
Only way to see if it works is to try it.

I suggest spending the 50-100$ on a bench top power supply so you can charge individual cells. Given how much money you spent on cells, it doesn't make sense to skimp on necessary tools/equipment. If you are charging a pack that's way out of balance, the BMS may be switching the pack on and off many times an hour. Your best bet is to adjust the charger upwards until one cell hits 3.6V, then leave it until the pack comes into apparent balance, then repeat.
 
Hi,

That's not an option for me, I have no way to charge an individual cell. I can charge them in 12V batches, but wouldn't have a suitable BMS.

The BMS has active balancing @ 2A, so hopefully it will handle it.

Only way to see if it works is to try it.

dRdoS7

2A capable doesn't mean it will do 2A. It's still subject to Ohms Law. A small voltage difference will result in a small current. HOWEVER, if you create a max difference scenario, e.g., right before the highest cell hits cut-off or is near 3.65V, AND lower your current to 2A or less, the balancers may be able to work more optimally. In other words, charge in series to max and then decrease charger input to stay just under max to optimize balancer operation.
 
I charged 8 of the 280 Ah cells in parallel with my benchtop power supply unit in less than 24 hours. This was after letting them rest in parallel for a few hours.

I don't know if it's that my PSU (30v, 10a) is more powerful (doubtful) or the cells were already at a good voltage level, but the charge took much less time than I expected.
 
I charged 8 of the 280 Ah cells in parallel with my benchtop power supply unit in less than 24 hours. This was after letting them rest in parallel for a few hours.

I don't know if it's that my PSU (30v, 10a) is more powerful (doubtful) or the cells were already at a good voltage level, but the charge took much less time than I expected.

Sounds like your cells had to already be close to full.
8 x 280 = 2240 Ah
10A / 2240 = 0.45% per hour
24hrs x 10A = 240Ah (or roughly 10% of total capacity in 24hrs)

Unless maybe the 10A rating is @ 30V and the PSU can supply substantially higher current at 3-4V volts? I didn't think this was how these Benchtop supplies work, but I don't really know.
 
I don't think it was providing the full 10 amps when I had the voltage set at 3.60. A photo I have of the charge process shows 5.65 amps. My records show that all the cells started at 3.32 volts, out of the shipping boxes.
 
Hi,

I suggest spending the 50-100$ on a bench top power supply so you can charge individual cells. Given how much money you spent on cells, it doesn't make sense to skimp on necessary tools/equipment. If you are charging a pack that's way out of balance, the BMS may be switching the pack on and off many times an hour. Your best bet is to adjust the charger upwards until one cell hits 3.6V, then leave it until the pack comes into apparent balance, then repeat.

Why do you think the BMS would be switching on and off? I'm not sure, but I thought the BMS takes from the highest cell, and puts it to the lowest. I could be wrong.

This from the description:

The basic principle is to use the super polar capacitor as the energy temporary storage medium, charge the battery energy with the highest voltage to the super polar capacitor, and then release the energy of the super polar capacitor to the battery with the lowest voltage, so as to reciprocate until all the battery voltages are equal. The constant current DC-DC technology ensures that the current is constant whether the battery is charged or discharged.

I did say I will slowly increase the voltage from 3.2V/cell to 3.65V/cell. I will be able to see individual cell voltages, if there's a big difference, I won't be raising the voltage.

2A capable doesn't mean it will do 2A. It's still subject to Ohms Law. A small voltage difference will result in a small current. HOWEVER, if you create a max difference scenario, e.g., right before the highest cell hits cut-off or is near 3.65V, AND lower your current to 2A or less, the balancers may be able to work more optimally. In other words, charge in series to max and then decrease charger input to stay just under max to optimize balancer operation.

If there's only a small voltage difference, that means the cells are reasonably balanced.

dRdoS7
 
I'm not sure, but I thought the BMS takes from the highest cell, and puts it to the lowest. I could be wrong.

This is true of an active balancing BMS, most BMSes are not active balancing, they use resistors to burn off charge from the highest cells (this includes all the most popular ones here: Daly, Chargery, SBMS0, Ant BMS, JBD Smart BMS/Overkill).
 
Last edited:
Hi,

This is true of an active balancing BMS, most BMSes are not active balancing, they use resistors to burn off charge from the highest cells (this includes all the most popular ones here: Daly, Chargery, SBMS0, Ant BMS, JBD Smart BMS/Overkill).

I did say in one of my replies that it has active balancing. That was one reason for choosing it. Also has 16-24S cells configuration, and a higher possible charge rate than many I looked at. My second choice was the ANT BMS, but that had only 50A charging, and a low balance current.



According to that page it is made by Heltec, branded JK, but I couldn't find it on their site. I searched the model number (JK B2A24S-20P) on the web, and found the site below, but currently only Balancers are listed.


dRdoS7
 
It doesn't matter how the bms balances. If you have a large imbalance one cell will exceed the voltage limit if the current is high enough, thus causing the bms to panic and drop the charger. BMS is not a magic cure all, it has real limits, which the installer needs to understand.

The recommended process for building a pack is done so for a specific reason. Skipping steps without a comprehensive understanding is more likely to waste time or risk damage in my view.

That balance board doesn't have disconnect functions, so I hope you have a full feature bms in the mix...
 
Last edited:
I did say in one of my replies that it has active balancing. That was one reason for choosing it. Also has 16-24S cells configuration, and a higher possible charge rate than many I looked at. My second choice was the ANT BMS, but that had only 50A charging, and a low balance current.

Oops, I'm sorry, I forgot about that. I didnt take the time to remind myself of the details before responding.

According to that page it is made by Heltec, branded JK, but I couldn't find it on their site. I searched the model number (JK B2A24S-20P) on the web, and found the site below, but currently only Balancers are listed.

The Heltec brand (manufacturer?) is kind of confusing. Their website seems like an afterthought, and has a picture/advertisement for the active balancing BMS but doesn't actually list it/sell it, their Alibaba and Aliexpress pages do list the BMS you have (or at least a close cousin--black metal case, active balancing).
 
Hi,

It doesn't matter how the bms balances. If you have a large imbalance one cell will exceed the voltage limit if the current is high enough, thus causing the bms to panic and drop the charger. BMS is not a magic cure all, it has real limits, which the installer needs to understand.

The recommended process for building a pack is done so for a specific reason. Skipping steps without a comprehensive understanding is more likely to waste time or risk damage in my view.

That balance board doesn't have disconnect functions, so I hope you have a full feature bms in the mix...

If you have read my posts, you will see that all of this will be done before the bank is being put into use. So it doesn't matter if it does switch off.

Can you list those specific reasons, and why parallel balancing the only method to top balance? Note that it's "recommended", not mandatory.

Not that I knew it at the time, and I had already decided to use the series method, but did you also see post #2?

Also, if you follow the link to the BMS I have bought, you will see it does what a BMS does, and has the proper protections. You may only have followed the second link, which is just a balancer (as I mentioned in the post).

Finally, have you heard of the Darwin Awards? ?

Oops, I'm sorry, I forgot about that. I didnt take the time to remind myself of the details before responding.

I've occasionally read a recent post on a forum, and replied without realising there are quite a few pages before it that I haven't read. [oops!]

The Heltec brand (manufacturer?) is kind of confusing. Their website seems like an afterthought, and has a picture/advertisement for the active balancing BMS but doesn't actually list it/sell it, their Alibaba and Aliexpress pages do list the BMS you have (or at least a close cousin--black metal case, active balancing).

It's the same one. I bought via Aliexpress, found it cheaper there.

dRdoS7
 
Last edited:
Decided to do Plan B. Though not necessary, I'm doing the "step balancing" - to 3.5v first, then to 3.6v.

Plan B a lot slower, but only have to disassemble and reassemble once. I'm still finishing up other stuff anyway (both solar related and not).

Would have gone straight to 3.6v, but so I can leave it unattended for several days, limited it to 3.5+ (tad under 3.6, just so I can watch/monitor it to 3.6).

Lowest cell was like 3.29, highest was 3.34. When finally connected 32P, the whole pack was 3.31-something.

After 2 hours and again at 8 hours, but taken while being charged (vs at rest), it's been "stuck" at 3.361v. I expect it to be stuck like that for several days (my fault for being "patient" (or rather, "lazy") and doing Plan B instead of Plan A).
 

Attachments

  • 20200917_155938~2.jpg
    20200917_155938~2.jpg
    109.5 KB · Views: 44
  • 20200917_170318~2.jpg
    20200917_170318~2.jpg
    171.2 KB · Views: 44
Why step balancing?

What's your current?

I don't believe you have to step balance (like some say - 3.4, then 3.5, then 3.6/3.65).

I'm just doing it so I can physically be there to watch it go from 3.55 to 3.6 (just personal satisfaction, ha ha).

Max current of my benchtop DC power supply is, in the manual, 20A. Just using 16-17A bec not sure if these inexpensive Chinese power supplies can do 20A continuous. So, derating it about 20%.
 
Back
Top