diy solar

diy solar

48V battery - top balance cells in two batches?

balachai

New Member
Joined
Mar 13, 2021
Messages
67
I've just purchased 32 CALB cells, 163 Ah each. Like the ones in this pic, with the raised terminals.

I've got a 50A single-cell charger, but not enough busbars to connect all the cells in parallel. So, is this OK?
- Connect 16 of my cells in parallel.
- Charge them to 3.65V.
- Repeat with the remaining 16 cells.
- Connect them all in parallel-series to make up my 48V battery?

PS. I got these cells instead of the cheaper no-name option because they should be better quality, and are in-stock from a legit physical store which gives an 8-year warranty to 80% SOH. The QR code is intact but my phone won't read it.
 

Attachments

  • Cell QR code LoRes.jpg
    Cell QR code LoRes.jpg
    77.1 KB · Views: 9
I've just purchased 32 CALB cells, 163 Ah each. Like the ones in this pic, with the raised terminals.

I've got a 50A single-cell charger, but not enough busbars to connect all the cells in parallel. So, is this OK?
- Connect 16 of my cells in parallel.
- Charge them to 3.65V.
- Repeat with the remaining 16 cells.
- Connect them all in parallel-series to make up my 48V battery?

Yes.

PS. I got these cells instead of the cheaper no-name option because they should be better quality, and are in-stock from a legit physical store which gives an 8-year warranty to 80% SOH. The QR code is intact but my phone won't read it.

I’d urge you to check the capacity of your cells sooner rather than later.

After they are top-balanced, configuring them into a 32S string and discharging through a 2000W heating element is the procedure I found to most easily and quickly distinguish any weak / low-capacity cells from the stronger / full-capacity cells: https://www.amazon.com/Camco-02203-...ocphy=9032080&hvtargid=pla-350581636011&psc=1

A 2kW element like that has a resistance of ~7 Ohms and will discharge a 280Ah 32S ~102V battery @ 14.62A / 1.5kW under 20 hours...

I’ve only discharged an 8S battery through mine (@ 3.7A / 94W where hearing in air worked fine, so if you decide to put 1.5kW DC through a 2kW heating element, please consider using a water bath rather than air cooling (no idea how much heat will be generated nor whether air cooling will be safe).
 
I've adopted a slightly different methodology on Top Balance which may suit you as well. I use a TekPower TP1540E CC/CV Bench Power supply. I use Ring Terminals not "clips" for the charger wires. Clips are not that good and can fool readings and/or heat up due to resistance.

I top charge each cell to 3.60V and allow to saturate till amperage drawn drops to 2A and sits for 30 minutes.
Then they are set aside to settle while I do the others.... They will settle down to 3.450-3.50 and pretty much stay there. That's normal for LFP.
Once the set is done, I set 8 cells in Parallel (I'm 24V so ) then parallel the entire set to 3.60 and allow saturation till amps drawn drops to below 2A and sits there for an hour. Then turn off power and allow to settle 12 hours hour or so, That usually leaves them around 3.500 collectively, at which point they can be asssembled into a proper pack.

One 280AH Cell from 3.29V to 3.60 full saturation takes about 12 hours starting with over 30A CC which gradually decreases and flips to CV.
Once all cells are Top Charged and assembled into a Parallel Set for Top Balance, it only takes about 3 hours to fully top balance 8 cells.
It is a little more tedious but in the end more efficient and you see each cell independently during the Top Charge process.

Remember the primary Power Curve for LFP is 3.000-3.450. The curve is a cliff fall below 2.900 and a cliff climb without gain above 3.500.
 
I’d urge you to check the capacity of your cells sooner rather than later.
That was going to be my next question - whether a capacity check is necessary. :) I guess it's a good idea to make sure the product is correct.
heating element
That's a great option thanks, 'cause I don't have a usable 48V inverter yet.
A 2kW element like that has a resistance of ~7 Ohms
Cool.

I use Ring Terminals not "clips" for the charger wires.
Yep I was planning on doing that too.
I top charge each cell to 3.60V
Sounds good to me - long life is more important than high capacity.

you see each cell independently during the Top Charge process.
What info do you get from doing each cell individually, and how does that help?
 
Last edited:
At that voltage how would I measure amp-hours to determine battery capacity? I can measure amps with my Fluke clamp meter, but not amp-hours.

You can estimate by taking lots of voltage readings and integrating (W=V^2/R) but that’s really not the goal of that setup.

The goal is to quickly sort the cells from lowest-capacity to highest-capacity.

When a first cell drops to 2.5V (or is clearly heading that way well before the others), you pull it out of the string and start charging it for individual capacity measurement with your 150W tester.
(While then continuing to drain the remaining cells to identify the 2nd-weakest cell, etc...).

There is little point in knowing the exact capacity of all of your cells.

What you want to know first and foremost is whether you have any noticeably deficient cells (and you want to know that quickly so you can attempt to get your vendor to address the deficiency),

And what you want to know second is roughly how well-matched the discharge curves (and along with it capacity) of your remaining cells are
 
Update: I've top-balanced each batch of 16 separately to 3.6V. Now the two batches have settled:
- Batch A: 3.507 V
- Batch B: 3.569 V

Is that close enough to assemble them into a proper 48V pack, with each cell pair including one cell from batch A and one from batch B?
 
Late probably, but you mentioned in the OP you broke the set into 2 batches due to not having enough bus bars. Watch how this guy does it...

 
Late probably, but you mentioned in the OP you broke the set into 2 batches due to not having enough bus bars. Watch how this guy does it...

Thanks Sojourner. His threshold for cell-cell unbalance before connecting in parallel is 200 mV. My unbalance is only 62 mV, so maybe I'm OK to connect up?

But I like how he gets around a lack of busbars - putting twisted copper cable under the nuts, and tightening them down. I might just do that... :) Current will be very low anyway when finishing the top balance.
 
Update: I've top-balanced each batch of 16 separately to 3.6V. Now the two batches have settled:
- Batch A: 3.507 V
- Batch B: 3.569 V

Is that close enough to assemble them into a proper 48V pack, with each cell pair including one cell from batch A and one from batch B?
My two strings were similar. I'm not doing parallel cells, but I put the strings in parallel with a fuse between them and connected without fireworks. All the cells fairly quickly balanced to the same voltage.
 
My two strings were similar. I'm not doing parallel cells, but I put the strings in parallel with a fuse between them and connected without fireworks. All the cells fairly quickly balanced to the same voltage.
Great, thanks for the info!
 
Thanks Sojourner. His threshold for cell-cell unbalance before connecting in parallel is 200 mV. My unbalance is only 62 mV, so maybe I'm OK to connect up?

But I like how he gets around a lack of busbars - putting twisted copper cable under the nuts, and tightening them down. I might just do that... :) Current will be very low anyway when finishing the top balance.

If I were to do the copper wire method of connecting/balancing cells, I'd probably use a length of ground wire from a romex cable instead of the way the guy in the video twisted that smaller wire. Also, I'd cover the positive terminals with something when connecting the negatives and vice versa, for safety.
 
Good points, thanks. Maybe rather than stripping the whole cable I'll just strip intermittently at spacing to match the battery terminals.
 
Back
Top