BMS when not fully charging a pack

[E-HP]

I testing my logic on how I charge my battery pack and whether the BMS can protect it while charging, with some assumptions about how the BMS behaves (which may or may not be accurate).

I charge my 14S pack to 4.15V per cell, or 58.1V, and discharge to 52V, ~3.7V per cell. Here's the logic I'm testing:

• If the typical balancing BMS behavior is to stop charging when any cell group reaches 4.2V, then under my routine, no cell group would charge greater than 4.2V.

• If the typical balancing BMS behavior is to not balance until the pack reaches full charge, or 58.8V, then the pack will not balance under my routine.

• If the pack is not balancing, then the most out of balance the pack can be would be for 13 groups to charge to 4.2V, and one group charged to 3.5V when the pack is charged to 58.1V. If this is the case, then when discharged to 52V, it seems like the lowest group would be at close to 3V (although the discharge curve for the cell type would come into play).

I’m trying to see how much risk I have of damaging my pack, using my charging/discharging routine, assuming this type of BMS behavior if I don't balance my pack often. I don’t think it would ever get as far out of balance as described, and I’ve only balanced the pack twice and it didn’t need it either time.

 

[amberwolf]

If the BMS has a cell-level LVC that actually works, (it should, but you'd have to test it to be sure), *and* you are discharging from a working discharge port that is capable of shutting off discharge when LVC is reached, then it shouldn't be possible to actually damage the cells even in the worst-case imbalance.


If the BMS has bluetooth then you could use it's app to check the cells' balance during operation and charge, and even static, to verify them periodically.


If your BMS doesnt' ahve any specs available for it's actual HVC, LVC, balance-start/stop points, etc., then you'd need to test to see what each of these actually is (safer than assuming, if the numbers are important to your particular application).


If yours is a resistive balancer (with a shunt and transistors for each cell group), then it is unlikely to balance at any point other than near full charge. (it could be setup to do so if it's designed to do so and is programmable, but otherwise it probably does not, instead using fixed-voltage trigger chips for each channel to turn the shunts on and off).

If yours is a capacitive charge-transfer balancer (no shunt/transistors per group) then it is technically possible for it to balance at any time, including just sitting there, even nearly-empty, if it is setup / programmed to do so.


Note that "then under my routine, no cell group would charge greater than 4.2V." is only true when the cells are balanced. If there is a significant imbalance and the difference in voltage is greater than 0.05v, then that extra voltage could end up in the lowest-capacity / highest-resistance group and charge to the 4.2v point while the rest are at 4.15v (except for the ones below that voltage that leave extra for the highest one(s)), and will continue to get higher in voltage until the BMS reaches whatever it's shutoff-HVC point is. If that's 4.2v, it'll turn off then. If it's higher, it'll go to whatever that is, turn off the charge port, and then, assuming yours is a balancing BMS, it will drain down (or transfer charge from) that group until it reaches whatever the stop-balancing voltage is, and also turn the charge port back on at whatever the restart-charge voltage is.

I testing my logic on how I charge my battery pack and whether the BMS can protect it while charging, with some assumptions about how the BMS behaves (which may or may not be accurate).

I charge my 14S pack to 4.15V per cell, or 58.1V, and discharge to 52V, ~3.7V per cell. Here's the logic I'm testing:

• If the typical balancing BMS behavior is to stop charging when any cell group reaches 4.2V, then under my routine, no cell group would charge greater than 4.2V.

• If the typical balancing BMS behavior is to not balance until the pack reaches full charge, or 58.8V, then the pack will not balance under my routine.

• If the pack is not balancing, then the most out of balance the pack can be would be for 13 groups to charge to 4.2V, and one group charged to 3.5V when the pack is charged to 58.1V. If this is the case, then when discharged to 52V, it seems like the lowest group would be at close to 3V (although the discharge curve for the cell type would come into play).

I’m trying to see how much risk I have of damaging my pack, using my charging/discharging routine, assuming this type of BMS behavior if I don't balance my pack often. I don’t think it would ever get as far out of balance as described, and I’ve only balanced the pack twice and it didn’t need it either time.

 

[E-HP]

Note that "then under my routine, no cell group would charge greater than 4.2V." is only true when the cells are balanced. If there is a significant imbalance and the difference in voltage is greater than 0.05v, then that extra voltage could end up in the lowest-capacity / highest-resistance group and charge to the 4.2v point while the rest are at 4.15v (except for the ones below that voltage that leave extra for the highest one(s)), and will continue to get higher in voltage until the BMS reaches whatever it's shutoff-HVC point is. If that's 4.2v, it'll turn off then. If it's higher, it'll go to whatever that is, turn off the charge port, and then, assuming yours is a balancing BMS, it will drain down (or transfer charge from) that group until it reaches whatever the stop-balancing voltage is, and also turn the charge port back on at whatever the restart-charge voltage is.

Thanks, all of this info is helpful. What you raise above is probably where I need the most clarification. Is the HVC a pack level cutoff, or is it cell level (cuts off charging the pack when one cell hits 4.2V)? That hasn't been clear in some of the posts I've read. Do cell level HVC BMSs exist? I think in the end getting a smart BMS that allows monitoring each group would be best for peace of mind, but I've also concluded that having well matched cells in the beginning trumps anything a BMS can offer. Still, it's good to know what the BMS doesn't do.

 

[amberwolf]

Thanks, all of this info is helpful. What you raise above is probably where I need the most clarification. Is the HVC a pack level cutoff, or is it cell level (cuts off charging the pack when one cell hits 4.2V)? That hasn't been clear in some of the posts I've read. Do cell level HVC BMSs exist?

LVC and HVC *have* to be cell-level in order for the BMS to do it's job. If it couldn't detect cell-level HVC and turn off the charge port, it could not protect against cell overcharge.

It could also have a separate pack-level HVC, but I don't know what yours has, or if this is a thing on many of them. (it's probably not common).


What the specific voltage used for various BMS trigger points is varies from BMS to BMS, and some are programmable (via bluetooth from a mobile app, etc), so if there is documentation (by the manufacturer or by another user that's done testing) for your BMS you can find out what they are without testing for them. Otherwise, testing is the only way to know for sure.

Trigger points for each channel (cell group) could include but may not be limited to the following, though they may use very different names for them, and specific behavior for each may vary:
HVC
LVC
Balance start (which is probably below HVC)
Balance stop (which is below Balance start)
Charge restart after HVC (which is below HVC)
Discharge restart after LVC (which is above LVC)

I think in the end getting a smart BMS that allows monitoring each group would be best for peace of mind, but I've also concluded that having well matched cells in the beginning trumps anything a BMS can offer. Still, it's good to know what the BMS doesn't do.

Well-matched cells being used well-within their limits (perhaps only the middle 50-60% is probably ideal, for capacity and for charge and discharge current limits, and any other spec they have (thermal, etc)), will probably work much better for much longer than a pack of unmatched cells being pushed hard to their limits a lot.

I don't use a BMS on my EIG cells; they are typically used in the above manner (except that I do almost fully charge them to about 4.13v/cell), and they stay balanced and perform fine for my usage on the trike, even though they were probably all manufactured over a decade ago, and I got them used.