Moved: Conversation on Top and Bottom Balancing

ArthurEld

Photon Sorcerer
The one comment I will pull out of my self imposed spoiler quarantine, is that: if there is actually some logic/science behind the discharge down to 2.0V beyond ('its worth a try') it may benefit us and others to understand that logic. Some people (that I mostly trust, but aren't infallable) have observed positive benefits to completing full cycles to ~3.65 every so often. But I've never heard of anyone even muse about the opposite (full--or in this case 'over'--discharge) having any benefit whatsoever. Maybe they just want to test a lower voltage to rule out any slight measurement errors.
I am in the process of bottom balancing my 48V battery to 2.5V. I will top balance it when I am done.
I don't know if there is a benefit from doing that.
I am fairly sure that bottom balancing brings the cells in sync pretty good. It is way better than not balancing at all.
My theory is that it is good to go through the entire discharge/deflate charge/inflate process in the restraining fixture.
It may be good for other things too.
It doesn't take that much time and I am enjoying the heck out of learning about these cells.

edit: I am fairly sure the 2.5V-3.65V recommendations mean that you are less likely to damage cells if you stay in that range.
It doesn't mean that you will definitely damage your cells if you go out of that range.
 
Last edited:

Dzl

Perpetual Newbie & Unofficial Forum Librarian
Staff member
Moderator
I am in the process of bottom balancing my 48V battery to 2.5V. I will top balance it when I am done.
I am fairly sure that bottom balancing brings the cells in sync pretty good. It is way better than not balancing at all.
I don't think there is any harm in this, but I also can't see any potential benefit (yet for some non rational reason the idea still appeals to me). In very simple terms, balancing umatched cells =
  • Top balance: minimum imbalance at the top, maximum imbalance at the bottom
  • Bottom balance: minimum imbalance at the bottom, maximum imbalance at the top
They are opposites. Doing one undoes the other.

cut.png

My theory is that it is good to go through the entire discharge/deflate charge/inflate process in the restraining fixture.
Is this the same 'restraining fixture' the cells will be calling their permanent home?

It may be good for other things too.
It doesn't take that much time and I am enjoying the heck out of learning about these cells.
I like this attitude (y) inquisitiveness and curiosity are two of the things that make this forum a fun and interesting place to participate
 
Last edited:

ArthurEld

Photon Sorcerer
I don't think there is any harm in this, but I also can't see any potential benefit (yet for some non rational reason the idea still appeals to me). In very simple terms, balancing umatched cells =
  • Top balance: minimum imbalance at the top, maximum imbalance at the bottom
  • Bottom balance: minimum imbalance at the bottom, maximum imbalance at the top
They are opposites. Doing one undoes the other.

View attachment 33854
Exactly what I was saying. Bottom balancing is way better than not balancing because it pulls the states of charge into balance.
Is this the same 'restraining fixture' the cells will be calling their permanent home?
That doesn't matter much in my opinion. I just think the cell needs to get used to a limited space. That space isn't going to change much when in a fixture. But it can change a lot without a fixture.

edit: here's an image that might help-
1611145884342.png
 
Last edited:

Dzl

Perpetual Newbie & Unofficial Forum Librarian
Staff member
Moderator
Exactly what I was saying. Bottom balancing is way better than not balancing because it pulls the states of charge into balance.
Well yes, but if you will be:
..bottom balancing my 48V battery to 2.5V. I will top balance it when I am done.
then the top balance will undo/reverse the bottom balance. I think maybe I am misunderstanding you. From the above statement it sounded like you were going to be bottom balancing and then immediately turning around and top balancing. But I think I may have misinterpreted your meaning?
 

ArthurEld

Photon Sorcerer
Well yes, but if you will be:

then the top balance will undo/reverse the bottom balance. I think maybe I am misunderstanding you. From the above statement it sounded like you were going to be bottom balancing and then immediately turning around and top balancing. But I think I may have misinterpreted your meaning?
The operating capacity becomes much larger whether you top or bottom balance. I think it would be better for the cells if you moved the operating capacity to the middle of all of the cells. But that is too complicated.
1611146606246.png
 

Solarfun4jim

Solar seduced :-)
The operating capacity becomes much larger whether you top or bottom balance. I think it would be better for the cells if you moved the operating capacity to the middle of all of the cells. But that is too complicated.
View attachment 33857
I would agree that a 'middle' operating capacity would be ideal, except that many of the shunts and bms's need to get the 'full charge' to the original top balance level, to recalibrate them back to 100% SOC values. Still evaluating if i can lower the top balance, to be done at say 3.45 instead of 3.65 and then re-configuring the shunt to read that as the full charge?

EDIT...Apologies cinergi for usurping your thread...just thought it was an interesting diversion. You do throw up lots of good topics for discussion. (y)
 
Last edited:

Hedges

I See Electromagnetic Fields!
I deleted this because I am over thinking and this is cinergi's thread

You got to be quick to quote me because I keep changing my mind

Commenting on your deleted comment ...

If a mismatched battery pack is bottom balanced, then when charged to a target voltage (all the charge controller knows), BMS would disconnect due to a runner.
It would take communication between BMS and SCC to regulate operation to limit charging of lowest capacity cell.
 

ArthurEld

Photon Sorcerer
Commenting on your deleted comment ...

If a mismatched battery pack is bottom balanced, then when charged to a target voltage (all the charge controller knows), BMS would disconnect due to a runner.
It would take communication between BMS and SCC to regulate operation to limit charging of lowest capacity cell.
I'm not sure about that Hedges. Once the cells state of charge are in sync you need to set the charge limits to stay within the usable range.
In my mind you shouldn't be setting the charge limits where cells are running.
But the preferred method for our purposes is top balance and my comments aren't an argument against top balance.
The picture shows what is going on. If you don't balance you have seriously limited capacity. Whether you balance the top or bottom has less significance.

edit again: the operating range is always going to be determined by the lowest capacity cell so the operating range for that cell will be the same whether you top or bottom balance
 

cinergi

1.21 Jigawatts
I'm not sure about that Hedges. Once the cells state of charge are in sync you need to set the charge limits to stay within the usable range.
In my mind you shouldn't be setting the charge limits where cells are running.
But the preferred method for our purposes is top balance and my comments aren't an argument against top balance.
The picture shows what is going on. If you don't balance you have seriously limited capacity. Whether you balance the top or bottom has less significance.

edit again: the operating range is always going to be determined by the lowest capacity cell so the operating range for that cell will be the same whether you top or bottom balance

What @Hedges said is correct if you use a CA/CV charge profile. If you bottom balance, and the cells are different enough in capacity, you'll have a BMS disconnect when one of the cells reaches 3.65 and the rest are still at 3.4. You could try to charge to an average VPC of 3.35 to guesstimate your stopping point or use SoC to terminate the charge (or best: a CAN Bus solution which terminates the charge when the highest cell reaches a preset value). LFP isn't particularly conducive to bottom balancing due to the flat voltage curve. If your cells are closely matched, then sure ... but not with these EVE cells we're getting...

You're right, for storage situations like Solar, top balance is better.

And with that, I'll fire off a new post that brings us back on topic :)
 
  • Like
Reactions: Dzl

ArthurEld

Photon Sorcerer
What @Hedges said is correct if you use a CA/CV charge profile. If you bottom balance, and the cells are different enough in capacity, you'll have a BMS disconnect when one of the cells reaches 3.65 and the rest are still at 3.4. You could try to charge to an average VPC of 3.35 to guesstimate your stopping point or use SoC to terminate the charge (or best: a CAN Bus solution which terminates the charge when the highest cell reaches a preset value). LFP isn't particularly conducive to bottom balancing due to the flat voltage curve. If your cells are closely matched, then sure ... but not with these EVE cells we're getting...

You're right, for storage situations like Solar, top balance is better.

And with that, I'll fire off a new post that brings us back on topic :).
cinergi and Hedges, the operating range is based on the lowest capacity cell. That range doesn't change whether you top balance or bottom balance. You will set the exact same voltages either way.
 

Hedges

I See Electromagnetic Fields!
cinergi and Hedges, the operating range is based on the lowest capacity cell. That range doesn't change whether you top balance or bottom balance.

Yes but ... The voltage at the top of the range differs. And that is what SCC uses to detect full.

You will set the exact same voltages either way.

I don't think so.

Cinergi said the same thing I was thinking - If top balanced, all cells move up towards knee of the curve and target charge voltage is reasonably distributed across them. If bottom balanced, one runner will shoot up the knee while the rest remain on the flat. That means a much smaller delta in pack voltage is produce when approaching top end of operating range.

I think with bottom balance you'll get the BMS disconnecting your system, and it stops working (assuming single port, not separate charge/discharge ports.)
 
  • Like
Reactions: Dzl

cinergi

1.21 Jigawatts
cinergi and Hedges, the operating range is based on the lowest capacity cell. That range doesn't change whether you top balance or bottom balance. You will set the exact same voltages either way.

This might deserve it's own thread, but you're not seeing my point. Yes, the total useable Ah capacity will be the same, but you will cause a HVDC every time you charge; you'll never be able to complete the charge without doing what I previously mentioned. The charger will see 13.6 volts at the battery but one of the cells will be over 3.65 and your battery gets disconnected.
 
  • Like
Reactions: Dzl

ArthurEld

Photon Sorcerer
I don't want to drag this out forever and I think top balancing is better.
But, you have to set your charging range based on the lowest capacity cell .
It doesn't matter if you line the other cells up to the bottom of that cell or the top of that cell.
I'll drop it now
 

Hedges

I See Electromagnetic Fields!
I don't want to drag this out forever and I think top balancing is better.
But, you have to set your charging range based on the lowest capacity cell .
It doesn't matter if you line the other cells up to the bottom of that cell or the top of that cell.
I'll drop it now

Do you program a charging range in terms of amp-hours? Does your charge controller track ampere-hours (including amps out as well as amps in to the battery), and stop charging when it computes that the battery has reached a certain state of charge? Not using voltage as the limit?

That's the only way I could see bottom-balancing working.
 

ArthurEld

Photon Sorcerer
Do you program a charging range in terms of amp-hours? Does your charge controller track ampere-hours (including amps out as well as amps in to the battery), and stop charging when it computes that the battery has reached a certain state of charge? Not using voltage as the limit?

That's the only way I could see bottom-balancing working.
Ok, you dragged me back in. hehe
I do not program in Ah.

It is difficult to make other people see things as I do. But I think we agree that the charging voltages in the charge controller are set based on the lowest capacity cell.
The lowest capacity cells voltages don't change if it is bottom balanced.

I think it is better to top balance because discharging will probably use higher amps and that could cause one of the other cells to run to the limit. So, by top balancing you push the cushion to the bottom end.
 
Last edited:

Luthj

Photon Sorcerer
Bottom balancing only makes sense for:

  • Packs which operate nearly continuously below 80%. EVs are a great example.
  • High current packs (1C or higher, especially for charging current) (EVs again)
    • High charge currents can drive big voltage differentials when charging. Bottom balancing can be useful for aged packs where a simple high voltage termination is more desirable than extended balance time.
  • Packs that don't supply constant loads, and/or have charger control instead of disconnect.
    • The reason for this, is that HVD is undesirable for off grid packs. So a charger control method other than disconnect is necessary to prevent HVD at the top of the charge due to balance issues.
  • Packs with manual balancing only.

The issue of course is how do you maintain this bottom balance? You either need a custom balance profile/logic (which is not common in BMS systems), or you need to manually balance. If you have a typical passive balancing system, the cells will start moving towards a top balance regardless. An active balancer can work with a bottom balance, but you need to spend a bit of time at low SOC for them to work.

Which brings up another point. Balancers need time to work. How long do most packs spend sitting in the low SOC range (below 10%)? Not much usually. However most packs do spend quite a bit of time above 90%, which lends to a top balanced method.

Which condition will the pack encounter more? LVD or HVD? Many/most packs aren't bumping against the LVD very often, so top balancing to prevent undesirable imbalance/behavior ( HVD) at higher SOC makes the most sense.

Other than a few corner cases, I think bottom balancing is a leftover from the early days of DIY LFP packs. Many of which were running without a BMS, and instead just used a pack HCD/LVD, and operated in the area between the knees for safety. Performing a bottom balance with a few resistors was easier than trying to work with non adjustable chargers and a top balance.
 

fafrd

Solar Addict
I don't want to drag this out forever and I think top balancing is better.
But, you have to set your charging range based on the lowest capacity cell .
It doesn't matter if you line the other cells up to the bottom of that cell or the top of that cell.
I'll drop it now
Correct. Ideally, your weakest cell hits maximum voltage at 100% SOC that is slightly above the voltage of all other cells (slightly to avoid HVD) and also hits minimum voltage at 0% SOC that is slightly below the voltage of all other cells.

If you’ve achieved that, your getting as much capacity from your pack of cells as they can deliver (barring the use of a high-current active-balancer which adds the ability to rob from the rich to give to the poor ;).
 
Top