New cells arrive today! Spend 100 hours top balancing...

[Tulex]

A 15v charger doesn't allow me to charge a 48v battery. Even if i split the battery in half, that's still 24v.

Also (correct me if I'm wrong) but it would only ever be 60a if the battery was 100% dead. Otherwise the current would be less and less as the voltage approached the max.

I've never had the need to charge a 48V battery, my inverter does that. The 60a part is correct, but that's the same with a 10a version. All I can say is that I had 48 cells in parallel and topped them off very quickly with my 15v 60a charger.

 

[Tulex]

Link?

The one I got is linked in my post earlier. Ok, it's a large little more......

 

[sunshine_eggo]

Top Balancing LiFePo4 Cells using a low cost benchtop power supply.

Jun 9, 2020

Top Ballancing

• FilterGuy
• How-to PDF's or Ebook's

 

[Tulex]

BUT, I'm moving to the camp that top balancing is a waste of time and money. I did a lot or research before building my batteries. Some very knowledgeable people insist it's a must. And, because there is no real downside other than a few dollars and time, I went with it.

Well, my batteries got way out of whack a while back from not being fully charged for a lot of cycles. I forced a few full charges from grid, and they were back to normal. This with passive BMSs. So, top balance if you want, but it's most likely not necessary at all. Seems to me the only people who need to do it are people who don't use a BMS, and that's not something I would do.

 

[kc130up]

I'd buy a 3KW 57V rectifier for ~$100 delivered, use that to bulk charge then use the power supplies you already own to bring up low cells while it's in pack form.

Link?

 

[kc130up]

What BMS are you planning to use in the final build?

IMO all this top balancing with leads and individual chargers to perfectly charge the cells is absolutely pointless.

I’m sure it’s fun to play around and check the current and voltage of each pack. Jot down and log the voltage rise over time.

But for me, build the battery with BMS and charge it, let the JK 2a active balancer do its thing.

The battery is there for solar storage, charging it from the grid just feels dirty.

My second battery I built was an 8s 280k ver3, after fully assembled, I only tossed maybe 30ah at it with my cheap 30v 10a charger, to verify things were working. I then just connected it parallel with my 80% SOC, voltage delta was maybe .10v 3.40 vs 3.30. Initial inrush was only about 200w into the new battery. After about a week or so both batteries were nice and balanced and fully charged to 3.45v/cell.

AP21S002. It does not have an active balancer.

 

[42OhmsPA]

Link?

I went with the preprogrammed 57V, there are other options and you can program them to different settings; I haven't gotten that far yet. (Tons of info in the chargenectifier thread)

I just found this on AliExpress:
$49.92 | Emerson R48-3000e3 Communication Switching Power Supply Module 48V Induction Heater Power Supply Disassemble With interface

https://a.aliexpress.com/_mKGrPvv

 

[sunshine_eggo]

BUT, I'm moving to the camp that top balancing is a waste of time and money. I did a lot or research before building my batteries. Some very knowledgeable people insist it's a must. And, because there is no real downside other than a few dollars and time, I went with it.

Well, my batteries got way out of whack a while back from not being fully charged for a lot of cycles. I forced a few full charges from grid, and they were back to normal. This with passive BMSs. So, top balance if you want, but it's most likely not necessary at all. Seems to me the only people who need to do it are people who don't use a BMS, and that's not something I would do.

This depends on the following things:

All cells are nearly identical capacity.
All cells are at the same relative SoC upon receipt.
All cells need the same Ah input to achieve 100% SoC within about 0.1-0.2%.
All cells have nearly identical self-discharge (SD) rates.
Sufficient time has NOT passed since last charge to allow SD rates to negatively impact balance more than the 0.1-0.2%.

The 0.1-0.2% number is the threshold where it's hard to get all cells in the 3.45-3.65V range at charge termination. is about the max you can tolerate with a ~50mA passive BMS getting them balanced in a few cycles at elevated voltage OR that will come into balance with normal cycling in a reasonable timeframe - maybe a week or so.

I have done detailed testing of two sets of matched EVE cells from Amy Wan, and BOTH sets required balancing. I've assisted with other builds requiring it. We've seen a huge number of new builds requiring balancing. We've seen multiple brands of batteries including Battleborn, EG4, SOK, Trophy, and all other common brands have balance issues upon receipt.

Your "way out of whack" batteries were likely not significantly more than the 0.1-0.2% SoC deviation.

Cells received from China have not been cycled or charged for at least 60 days, so you can imagine there is a natural deviation of SoC due to SD variation.

Even if a battery is truly top balanced and then discharged to the hazmat required SoC of no more than 30% max, then let is sit for at least 90 days before it gets to the end user. Even slight SD variations can take cells out of balance in this case.

Lastly, the first step of the balancing guide is to build the battery with the BMS and charge the battery to full. If it doesn't need balancing, don't balance it!

 

[robbob2112]

BUT, I'm moving to the camp that top balancing is a waste of time and money. I did a lot or research before building my batteries. Some very knowledgeable people insist it's a must. And, because there is no real downside other than a few dollars and time, I went with it.

Well, my batteries got way out of whack a while back from not being fully charged for a lot of cycles. I forced a few full charges from grid, and they were back to normal. This with passive BMSs. So, top balance if you want, but it's most likely not necessary at all. Seems to me the only people who need to do it are people who don't use a BMS, and that's not something I would do.

This depends on the following things:

All cells are nearly identical capacity.
All cells are at the same relative SoC upon receipt.
All cells need the same Ah input to achieve 100% SoC within about 0.1-0.2%.
All cells have nearly identical self-discharge (SD) rates.
Sufficient time has NOT passed since last charge to allow SD rates to negatively impact balance more than the 0.1-0.2%.

The 0.1-0.2% number is the threshold where it's hard to get all cells in the 3.45-3.65V range at charge termination. is about the max you can tolerate with a ~50mA passive BMS getting them balanced in a few cycles at elevated voltage OR that will come into balance with normal cycling in a reasonable timeframe - maybe a week or so.

I have done detailed testing of two sets of matched EVE cells from Amy Wan, and BOTH sets required balancing. I've assisted with other builds requiring it. We've seen a huge number of new builds requiring balancing. We've seen multiple brands of batteries including Battleborn, EG4, SOK, Trophy, and all other common brands have balance issues upon receipt.

Your "way out of whack" batteries were likely not significantly more than the 0.1-0.2% SoC deviation.

Cells received from China have not been cycled or charged for at least 60 days, so you can imagine there is a natural deviation of SoC due to SD variation.

Even if a battery is truly top balanced and then discharged to the hazmat required SoC of no more than 30% max, then let is sit for at least 90 days before it gets to the end user. Even slight SD variations can take cells out of balance in this case.

Lastly, the first step of the balancing guide is to build the battery with the BMS and charge the battery to full. If it doesn't need balancing, don't balance it!

My new cells — I built the battery with a BMS — charged as full as it would until I got cell over-volt alarms. There were a number of runners. Then, with the 2-channel supply (each set to 3.65V), I walked the leads through the whole pack two cells at a time.

It took about an hour for most cells — some took 30 minutes, some longer — to top them up. While I was doing this, I turned off balancing — no sense burning off power while trying to top them all out.

Once they were all only drawing 0.5 amps or so and at the right voltage, I turned balancing back on and finished charging the whole pack overnight. It only pulled 6 amps.

Then I let it rest — everything was within 0.005V.

Next, I drained the pack to 50% and recharged it. The cells stayed within 0.1V of each other and finished at roughly the same.

 

[Tulex]

This depends on the following things:

All cells are nearly identical capacity.
All cells are at the same relative SoC upon receipt.
All cells need the same Ah input to achieve 100% SoC within about 0.1-0.2%.
All cells have nearly identical self-discharge (SD) rates.
Sufficient time has NOT passed since last charge to allow SD rates to negatively impact balance more than the 0.1-0.2%.

The 0.1-0.2% number is the threshold where it's hard to get all cells in the 3.45-3.65V range at charge termination. is about the max you can tolerate with a ~50mA passive BMS getting them balanced in a few cycles at elevated voltage OR that will come into balance with normal cycling in a reasonable timeframe - maybe a week or so.

I have done detailed testing of two sets of matched EVE cells from Amy Wan, and BOTH sets required balancing. I've assisted with other builds requiring it. We've seen a huge number of new builds requiring balancing. We've seen multiple brands of batteries including Battleborn, EG4, SOK, Trophy, and all other common brands have balance issues upon receipt.

Your "way out of whack" batteries were likely not significantly more than the 0.1-0.2% SoC deviation.

Cells received from China have not been cycled or charged for at least 60 days, so you can imagine there is a natural deviation of SoC due to SD variation.

Even if a battery is truly top balanced and then discharged to the hazmat required SoC of no more than 30% max, then let is sit for at least 90 days before it gets to the end user. Even slight SD variations can take cells out of balance in this case.

Lastly, the first step of the balancing guide is to build the battery with the BMS and charge the battery to full. If it doesn't need balancing, don't balance it!

THIS is the link to my post with the issue I had. I think I was about .3v out of balance on at least one of the batteries.

Correct me if I'm wrong, but if all cells are not nearly identical capacity or identical self discharge rates, you would be fighting to keep them balanced even after a top balance.

And to be fair "to be fair", what you point out as the first step in the guide is actually on page two, section 3,5.

I don't see any harm in top balancing. In reality, it's nice to start with a battery in ideal condition. But, and maybe I've been lucky getting cells from 3 different places, these places seem to be matching cells very well, all have come in at the same state of charge for each batch. I'm guessing that I would have been just fine not top balancing.

 

[740GLE]

AP21S002. It does not have an active balancer.

Can you source a BMS with active balancer?

I don’t know why one would purposely chose a BMS that doesn’t solve potential issues.

 

[740GLE]

I don't see any harm in top balancing. In reality, it's nice to start with a battery in ideal condition. But, and maybe I've been lucky getting cells from 3 different places, these places seem to be matching cells very well, all have come in at the same state of charge for each batch. I'm guessing that I would have been just fine not top balancing.

Only issue is see is the 200lbs meat popsicle that could mess something up. Such small differences in connections or terminal prep will actually unbalance the cells. Bump the power supply and it’s now at 3.80v Or potential to drop a cell or wrench on exposed terminals.

Again that’s just my opinion, I’m in the camp of I want to “play” with my cells as little a possible.

 

[Tulex]

Can you source a BMS with active balancer?

I don’t know why one would purposely chose a BMS that doesn’t solve potential issues.

Because passive balancing is fine if you have matched cells. I think where people get to the point that they need active balancing is when they have a mishmash of cells.

 

[Tulex]

Only issue is see is the 200lbs meat popsicle that could mess something up. Such small differences in connections or terminal prep will actually unbalance the cells. Bump the power supply and it’s now at 3.80v Or potential to drop a cell or wrench on exposed terminals.

Again that’s just my opinion, I’m in the camp of I want to “play” with my cells as little a possible.

Well, I'm closer to 210, but yeah, I get it, it's part of why I've moved to the camp of it not being necessary.

 

[robbob2112]

My cells are all the same size and from the same vendor -- at least these two banks.... The third bank is from 18650 battery store, but over a month apart -- saga - they kept arriving with bent cells so I have 8 damaged cells to keep -- but 4 from the first batch were fine, then then 2 batches damaged - then a good batch after they custom packed them -- According to the QR code they were mfg on the same day so I have high hopes they will be matched...

NOTE - my BMS is a JBD with passive balancing -- little bank of resistors, 1 per line --

 

[740GLE]

Because passive balancing is fine if you have matched cells. I think where people get to the point that they need active balancing is when they have a mishmash of cells.

Agreed they can be made to work if you’re able to keep them in the “balance zone” enough for the cells health.

Are the passive balancing BMS cheaper? It kinda makes sense to have a BMS to have the best features to reduce your interactions/concerns.

It’s like buying a car and specifically searching one out that doesn’t have cruise control, cause you don’t use control.

Better to have it and not need it, than not have it but need it

 

[sunshine_eggo]

THIS is the link to my post with the issue I had. I think I was about .3v out of balance on at least one of the batteries.

Correct me if I'm wrong, but if all cells are not nearly identical capacity or identical self discharge rates, you would be fighting to keep them balanced even after a top balance.

Not necessarily fighting it, but you will need a certain minimum balancing time for a given period, e.g., let's say your SD delta yields a 1Ah difference per month (that's a pretty big number, my worst case cells were 2.5Ah over 5 months total). That means you'll need 20 hours of balancing per month. If you're in absorption 2hr/day, you only need to get fully charged an average of once every three days with two hours of balancing each time.

And to be fair "to be fair", what you point out as the first step in the guide is actually on page two, section 3,5.

The way I read it is Sections 1 and 2 are information only. The first time you're instructed to actually DO something is section 3.

I don't see any harm in top balancing. In reality, it's nice to start with a battery in ideal condition. But, and maybe I've been lucky getting cells from 3 different places, these places seem to be matching cells very well, all have come in at the same state of charge for each batch. I'm guessing that I would have been just fine not top balancing.

Mostly agree. It's a crap shoot, and it's why it's important to order from reputable suppliers to enhance your chances. Even with Amy Wan, my first batch of cells with all at 3.28-3.29V were between 28 and 40% SoC. The second batch was much more consistent at 31.5-33.0% SoC. The first batch was the older cells with single M6 internal threads. The second were LF280K with the dual threaded terminals - about 3 years, 2 months apart.

 

[kc130up]

I've never had the need to charge a 48V battery, my inverter does that. The 60a part is correct, but that's the same with a 10a version. All I can say is that I had 48 cells in parallel and topped them off very quickly with my 15v 60a charger.

If I had enough solar to charge the battery quickly, I would connect it to the inverter, set the BMS to stop at 3.6v, then top balance each cell to 3.65v to finish up.

It's been raining in the midwest for 2 weeks so very little sun!

However the weather forecast for today was wrong and its partial sun, so I went ahead and temporarily assembled the 16s and will connect it to the inverter as soon as I find a longer positive cable...



The BMS is on and working, with all cells showing voltage. I wired it correctly this time

 

[Tulex]

Didn't catch that your inverter is off grid. So Yeah, different story.

 

[740GLE]

If I had enough solar to charge the battery quickly, I would connect it to the inverter, set the BMS to stop at 3.6v, then top balance each cell to 3.65v to finish up.

It's been raining in the midwest for 2 weeks so very little sun!

However the weather forecast for today was wrong and its partial sun, so I went ahead and temporarily assembled the 16s and will connect it to the inverter as soon as I find a longer positive cable...

View attachment 290474

The BMS is on and working, with all cells showing voltage. I wired it correctly this time

So who’s gonna be the “BlueSeas switches aren’t rated for 57v batteries” police?

I’m in the camp they are fine, just don’t use it to break load, have a fuse pop first, or shut down via BMS before trying to open the switch.

I’m jealous of your soldering iron

 

[kc130up]

I often think about all the 48v rated hardware being used for effectively 58v circuits.

I have never tried to use the switch to break a load, (not that my EG4-3K can do more than 25a) anyway. I aways turn off the breaker to the 120v loads, turn off the inverter, then turn off the battery.

I have a related voltage/hardware bus bar question in the newbie section that I am (still) waiting on someone to answer about using 12v 250a bus bars on a ~58v 50a circuit.

Are there 60v switches one could get?

Secret, I have never used that iron (yet). A tenant moved and left it. I tested it and it worked, so I kept it

 

[740GLE]

I often think about all the 48v rated hardware being used for effectively 58v circuits.

I have never tried to use the switch to break a load, (not that my EG4-3K can do more than 25a) anyway. I aways turn off the breaker to the 120v loads, turn off the inverter, then turn off the battery.

I have a related voltage/hardware bus bar question in the newbie section that I am (still) waiting on someone to answer about using 12v 250a bus bars on a ~58v 50a circuit.

Are there 60v switches one could get?

Secret, I have never used that iron (yet). A tenant moved and left it. I tested it and it worked, so I kept it

Your tenant has good taste.

It’s a hot debate about the blue seas switches, many say they are UL listed for 48v but not much more. People have done voltage drop tests and they do have a descent drop, which isn’t ideal for 12v systems. I’m using two in my 24v system.

Victron switches (of similar style) claim up to 60 or 72v but aren’t UL listed, so it’s questionable.

As long as they aren’t used to break load your fine.

 

[kc130up]

Didn't catch that your inverter is off grid. So Yeah, different story.

My "solar shed" pretends to be off grid, but when the going gets tough, i either run a extension cord to the house or carry the Solix F2000 into the house, charge it, then take it back to the shed.

I have an "grid" 60' away with a 250a service

 

[Flunkyboi]

I've setup my "lab" for top balancing my new batch of 280ah cells. :D
View attachment 290054

After my first experience using the "crappy" power cords that come with the chargers, I made new leads using old PC power cables. Used all 3 wires from each to create one lead each.

While I impatiently wait for Fedex to deliver, I'm doing napkin math and 36w into a 4S pack of 280ah cells is like 13 hours (assuming they are 1/2 charged). I am somewhat impatient so I got 4 chargers to prepare for today, but still, that is days and days.

So I'm thinking, why not use a 12v charger with high current to get the cells "higher", then top balance from 3v to 3.65?

The cells can take .5c so no issues with too much current.

Brilliant idea, bad idea, or something worse than a bad idea?

One thing I really like about this board is it's refreshing to watch posters embracing their full nerd. Most in my circle want to talk football 24/7.

 

[Skypower]

I’m in the camp of, if you’re going to compress/fixture the cells, assemble them as they are to be used and charge them. If you need to get an active balancer to make that happen, so be it, even if the balancer is only temporary. I’ve seen cells swell a bit on the first charge. I don’t feel that it’s a good idea to compress once that’s happened.