Imbalance When Empty?

rg12

100 kW
Joined
Jul 26, 2014
Messages
1,584
Brand new pack, 4C chinese brand (very high quality and consistent cells with 15mOhm IR) all balanced until depleted to cut off and at about 3.15V average per cell, the cells have a 0.10V imbalance.
Never watched packs going empty through this BMS and my question is if it's normal.
IMG-20210109-WA0055.jpg
 
Normal in my view. Close to 3.0 to 3.2 volts, a very small capacity usage cause huge drop in voltage. This is not nearly the case when you are closer to nominal (3.6-3.7V) voltage where capacity usage cause only small variation in voltage. So cell voltages will be much closer togoether when you recharge the battery and batt would seem perfectly balanced.

Matador
 
Normal because the voltage graph is at the very end, at the cliff of the discharge curve.
Any very small difference in:
IR, tab weld resistance/cold tab weld/hot weld, variance in thickness of tab metal, variance in solder thicknesses, wire gauge used, cells at the beginning or end of "S"eries....

.....between the cells will have a more drastic effect then it would at the left part of the charge, more charged part.

These are 25R but the graphs are all similar.

111111.jpg
 
Thats what I thought also but I forgot to mention that it's a 16P pack which logically should make all the little imbalances and capacity changes between cells all averaged.
 
markz said:
IR, tab weld resistance/cold tab weld/hot weld, variance in thickness of tab metal, variance in solder thicknesses, wire gauge used, cells at the beginning or end of "S"eries....

.....between the cells will have a more drastic effect then it would at the left part of the charge, more charged part.

Yes thats how it is, and one more reason could be the BMS.
Lets say the BMS does balance the pack at 4.10V with accuracy of 0.01V and then the pack get discharged to the end of the curve, that small difference from the beginning could increase to 0.10V then.
 
rg12 said:
Thats what I thought also but I forgot to mention that it's a 16P pack which logically should make all the little imbalances and capacity changes between cells all averaged.
So is this a 52v, 16s, 16p pack, ?
And those voltage readings are from each 16 cell parallel group ?
If so, those voltages tell is little about the individual cell status.
What cells are they ?
 
This depends on the BMS, however, the general rule is that balancing won't take place again until high on the charge curve. The very slight imbalance you are seeing is perfectly normal in my experience. Be concerned if the pack is fully charged and out by more than 0.1v (again can differ for each bms)
 
Yes about the BMS, but there is a wide variety of BMS's out there from generic crap, to half decent, to pretty good to really good.
There are many different topologies for BMS circuit design that can have pro's and negatives.
The BMS also drains a small amount of current even when not in use, again depends on quality if it continues to drain until pack is no good. And how much power BMS sips, I dont know but I'd guess milli or micro amps, 24/7/365. Some batteries will have an on/off button, my Grintech battery does. Using the battery until LVC kicks in on the BMS, then letting it sit would drain it dry. Charge back to storage voltage 3.60V or 3.70V no matter what, then store it, or wait until your going to use it then charge up to full.
 
You all may be interested to hear what GM thinks about cell voltage variation. Back in 2017, when the Chevy Bolt came out, their service manual said that all cells in their 96s3p packs, from 4.16V down to 3.28 V, should be within 0.03 V. If any cells fell outside that value, the battery section containing those cells would be replaced. Fast forward four years, with lots of failing packs, and a few fires, they now say that all cells should be within 0.07 V of the AVERAGE cell voltage.
They have lowered their standards 5X. Yeah, they are in deep doo-doo.12-29-18-3.jpg
 

Attachments

  • 11-15-20 back.jpg
    11-15-20 back.jpg
    51.8 KB · Views: 417
madin88 said:
Yes thats how it is, and one more reason could be the BMS.
Lets say the BMS does balance the pack at 4.10V with accuracy of 0.01V and then the pack get discharged to the end of the curve, that small difference from the beginning could increase to 0.10V then.

Yup. Our 2017 Bolt, like most until half way through 2019, at 100% SoC per GM, is never closer than 0.02 V. The 2020s we have seen are now at 0.006 V spread.

12-19-20 full charge.jpg
 
With bike packs esp, there can be current paths that aren't evenly spread thru a group by the nickel strips under load, that won't rebalance themselves. It can also be a BMS where it's low current cutoff for deciding when to end the change can be a little above the trickle one group needs to get evened out.

The problem with doing it is usually voiding your battery warranty, but I've had several problem batteries I opened up, and charged the groups one at a time, through a settable RC charger with a plug I made to go into the balance plug without shorting anything out. That tended to help, getting them all nice and truly equalized at the top.
 
Back
Top