Do I need to top balance?

[timblack1]

I purchased 16 LFP cells and an Apexium battery box kit which includes a BMS from Jenny Wu. Everything seems to be in perfect condition. I am very pleased.

On my first voltage test after opening the boxes, 12 of the cells each have 3.256 volts. The other four cells vary by only one millivolt higher or lower than that. So currently they are balanced, but at a medium voltage. Their voltage is not at the top of the cells' voltage range.

The BMS includes an active balancing feature. I have not done an internal resistance or capacity test on these cells because I don't have the equipment to do so. [Update on 8-10-2024: I was wrong; this BMS uses passive, not active, balancing.]

Should I top balance these cells before assembling and using the battery? Or, should I trust the BMS to top balance them for me while or after they charge for the first time?

 

[Partimewages]

Fully charge and top balance. It's not always necessary but saves headaches down the road.

 

[Brett V]

Measuring 3.2 volts is no indication of them being at the same SOC, you have no idea of how close they are. Some will say slap it together and let the BMS top balance them for you but if you want it right, top balance them first.

 

[time2roll]

I would skip the old school parallel top balance. Assemble as a battery with the BMS. Limit the initial charge to monitor the balance as the battery approaches full. Once voltage can easily go to 55.20+ place the battery in service. Continue to monitor the cell balance as voltage is increased to the desired level.

 

[timblack1]

I would skip the old school parallel top balance. Assemble as a battery with the BMS. Limit the initial charge to monitor the balance as the battery approaches full. Once voltage can easily go to 55.20+ place the battery in service. Continue to monitor the cell balance as voltage is increased to the desired level.

Interesting. This sounds like real top balancing, because I would still actively and personally monitor each cell's voltage. The fact that it uses the BMS's active balancing to accomplish the goal of top balancing does not obviously show me that it is anything different than real top balancing. If I test each cell voltage directly using a multimeter rather than rely on the BMS's sensors and potential small voltage drops caused by the sensor leads and connections, this sounds to me like it could be nearly identical to "old school" top balancing. Am I missing something which would be a big reason not to go this route?

 

[740GLE]

IMO an active balancer of 2a does everything an old school top balance will do (except it will save you time and limit any possible damage due to human error). So many times people try and “top balance” and then somehow a cell or two gets up to 3.8v due to flakey bench top supplies.

Active balance just saves so many beach aches and works assuming you have enough time to keep the charge voltage where cells are above 3.45-3.55v. IMO no need to bring them up to 3.65 as long as cell delta looks good.

 

[k490]

If you want to do capacity testing nothing wrong with doing an old school top balance charge. The main issue is you have to buy a power supply that is capable of low voltage and high current to get them fully charged in a reasonable amount of time. Those are not your typical hobby power supply most are limited to 10 amps. You have to buy one when your done it will be a dust collector unless you have a use for it. This is why I bought an active BMS capable of 2 amps per battery. I had no interest in capacity testing I thought at first I would get a tester I realized when you buy diy batteries from China there is no way to return them. The cost of shipping them back is about what you paid for them. If you end up with a bad cell might as well just run them until there is a problem. I doubt I will buy DIY batteries again the cost of a ready to use battery from EG4 isn't all that much different.

 

[Bananaman321]

I purchased 16 LFP cells and an Apexium battery box kit which includes a BMS from Jenny Wu. Everything seems to be in perfect condition. I am very pleased.

On my first voltage test after opening the boxes, 12 of the cells each have 3.256 volts. The other four cells vary by only one millivolt higher or lower than that. So currently they are balanced, but at a medium voltage. Their voltage is not at the top of the cells' voltage range.

The BMS includes an active balancing feature. I have not done an internal resistance or capacity test on these cells because I don't have the equipment to do so.

Should I top balance these cells before assembling and using the battery? Or, should I trust the BMS to top balance them for me while or after they charge for the first time?

I recently just purchased those same apexium boxes and 116 EVE LF280K cells from Jenny Wu as well.

I also purchased an extra of the same bms's and I planned to check all the cells for resistance and use the four weakest to build a 12 volt battery which I will keep as functional spare parts if I ever have a problem with one of the big batteries

I plan to clean and assemble all the bus bar's under compression with carbon conductive paste and then fully charge and use the method in the video below to balance the cells manually

I'm not really concerned with capacity checking them they come with a data sheet should be able to match them pretty closely with that though they tend to run within only a few Ah of one another from what others have posted.

Not that I feel it's really going to matter how often do we fully deplete these batteries. I figure if I have a pack running short down the road I can just switch a cell out with the the 12v battery during annual maintenance.

Good luck with your build

 

[hardtop]

I didn't balance my 4 EVE 304 grade A cells. No problems at all so far with 35 cycles on the pack. Right now cell voltage difference is 0.007 and always under .01. It may be a problem later like someone else said but idgaf I'll worry about it then.

I didn't want to spend time with top balancing because looking at all the articles on how to do it, the effort seemed like a big pita, super technical, detailed, involved, painful and I needed to buy a charger and carry it with me wherever I go overlanding (since I carry everything I own) and I'm so glad I didn't do it lol. But yeah, I will suffer the consequences of my actions!

 

[Bananaman321]

I didn't balance my 4 EVE 304 grade A cells. No problems at all so far with 35 cycles on the pack. Right now cell voltage difference is 0.007 and always under .01. It may be a problem later like someone else said but idgaf I'll worry about it then.

I didn't want to spend time with top balancing because looking at all the articles on how to do it, the effort seemed like a big pita, super technical, detailed, involved, painful and I needed to buy a charger and carry it with me wherever I go overlanding (since I carry everything I own) and I'm so glad I didn't do it lol. But yeah, I will suffer the consequences of my actions!

Cells are regulatory tested before shipping and should come balanced in a nearly full state of charge though this can drift apart if stored for an extended period time because not all cells discharge equally in no load storage.

That being said it should not be necessary and it should balance itself out over time but this only happens in the full state of charge above 3.45v per cell

It is something we could keep an eye on there is no sense in checking it unless at a full state of charge though because of the flat charge curve.

If a connection were to loosen at one of the bus bar locations over time and heat be dissipated during high charge/discharge it would cause the cells to drift apart possibly more than the BMS could keep up.

This is why it is recommended to check all stud torque annually on non-welded stud packs and DIY installations

Hope this helps learning every day

 

[timblack1]

https://diysolarforum.com/resources...ls-using-a-low-cost-benchtop-power-supply.65/ includes "Appendix B: Step-and-Pause top balance" and "Appendix D: Top Balance using an active balancer," which are basically the same as some of the recommendations here which are not the "old school" top balancing method.

 

[timblack1]

I also purchased an extra of the same bms's and I planned to check all the cells for resistance and use the four weakest to build a 12 volt battery which I will keep as functional spare parts if I ever have a problem with one of the big batteries

I'm interested to find out if you find any large differences in their internal resistance.

I plan to clean and assemble all the bus bar's under compression with carbon conductive paste and then fully charge and use the method in the video below to balance the cells manually

Ray Builds Cool Stuff's videos on the Apexium 280 battery box and other battery builds have been very helpful to me. I just got some MG 847 in the mail, and plan to use it to reduce resistance in the connections. I plan to sand the contacts WITH the MG 847 to prevent any new oxidation from forming immediately after the sanding. I think that should work. (I originally posted this paragraph in another thread, but am adding it here also since it is more directly a reply to your comment above.)

I hope your builds go well. If you learn tips and tricks you haven't seen yet regarding the Apexium boxes I'd be interested to hear about them here. I'm filming my build and hope to share it on YouTube someday. If I'm slow enough I might be able to incorporate your ideas into my build.

 

[HRTKD]

I did the "old school" top balance on the original eight cells that I bought in 2020 and put into service as two 4s batteries. Zero issues with those cells. They stay well balanced. But I think a large part of it is that they are high quality cells, not B grade rejects.

My newest set of 18 cells are in progress using the "old school" top balance method. If I had my inverter/charger when the cells arrived I probably would have put together a 16s battery and brought those cells up to 3.55v or 3.6v. However, I'm planning to balance all 18 cells and use the 16 that are matched for the final battery. So the traditional top balance method is what I'm going with, even if it is slower than watching glaciers move.

 

[timblack1]

It appears I was wrong to think this BMS has active balancing. Rather, it appears it uses passive balancing. I've corrected my posts above accordingly.

 

[Fluff34567]

@timblack1 and @Bananaman321 would you mind telling me what you paid for your battery kits please? I'm thinking of buying the same kit and looking for approximate figures .. pls message me if better for you.
Thanks !

 

[timblack1]

@Fluff34567 I recommend you ask Jenny Wu directly at jennywu896@gmail.com, for 2 reasons:

1. She gave me multiple options for the cell quantity, capacity, manufacturer, and chemistry (she offered sodium batteries), and the case, BMS, and cables she provided are designed for the cells I selected. Your requirements could be different from mine, so you might actually order different components.

2. I asked her for prices over the course of a little over 2 years while I sorted out when and how I could get this project done, and the prices she quoted changed over time. The best thing about that is that the cell prices kept coming down! So, you are likely to get a better price than I did.

 

[timblack1]

If you purchase this kit with the Seplos 10E BMS, Andy at the Off-Grid Garage on YouTube has a series of videos about configuring that BMS which I'm finding very helpful; I'm working through this one right now:

The whole battery kit is working quite well for me, but I'm uncertain whether it's charging to full capacity, and sometimes I think it's not balancing. When it stopped charging around 90% recently the cells were about 70 millivolts apart, but right now all the cells are within 4 millivolts of each other, so it may have balanced the cells, they they are in a mid-range SOC so it might not be possible to determine that at present. So I'm still working through getting the settings right.

Specifically, the BMS reports that it's charging to 90% capacity, but it doesn't seem to be charging at my intended absorption voltage of 56.4V when it gets close to the 90% capacity mark (instead it's in the 54V range, closer to my intended float voltage of 54.4V), so I'm unsure whether it's actually charging up to a true 90% capacity. Maybe it charged at 56.4V, (estimated it) hit 90%, and switched to float voltage and started balancing, which would be perfect. I'm just not sure yet whether that's what happened; I want to actually see it happen.

 

[Bananaman321]

I use 55.2v absorption and 54v float here is why.

 

[timblack1]

I use 55.2v absorption and 54v float here is why.

Thank you! After watching that video, I changed my charge profile to:

Bulk/absorption V: 55.2
Float V: 54.4
Inverter cutoff V: 42.8

 

[Bananaman321]

I set my inverter cut off or back to grid at 46v in low voltage warning at 48v. 46 volts 2.87v / cell is around 10% stated charge I don't really want to go lower than that with my pack if you're running an S16 LFP

 

[Ampster]

46 volts 3.87v / cell is around 10% stated charge

Check your math. 48 divided by 16 is 3. I think you mean 2.87 not 3.87
FWIW I would never go that low.

 

[timblack1]

I set my inverter cut off or back to grid at 46v in low voltage warning at 48v. 46 volts 3.87v / cell is around 10% stated charge I don't really want to go lower than that with my pack if you're running an S16 LFP

Hm. https://www.mobile-solarpower.com/design-your-own-12v-lifepo4-system.html says "Inverter Cut-off: 42.8V-48V." I didn't know whether to pick the high or the low end of that range, so I picked the low end. But looking closer at the voltage curve and charts, I think 3.0V (48V pack) is as low as I should go per cell. So I set the charge profile to:

Bulk/absorption V: 55.2
Float V: 54.4
Inverter cutoff V: 48

It's my impression from Andy's video above that this absorption voltage will lengthen the time it takes to reach 100% state of charge, and the whole charge profile should give me less amp hours than full capacity. I'll let it cycle a few times at these settings to see if that is the case.

 

[Bananaman321]

Check your math. 48 divided by 16 is 3. I think you mean 2.87 not 3.87
FWIW I would never go that low.

I apologize for that I was on the phone and hit the wrong button I corrected it thank you

 

[littleharbor2]

Cells are regulatory tested before shipping and should come balanced in a nearly full state of charge

Not true. Cells are always stored and shipped in a reduced SOC, generally 40 - 50 %. It is not good for LFP cells to sit idle with a nearly full SOC..