Top vs bottom balancing?

Alec_J

Member
Joined
Mar 16, 2017
Messages
89
Could someone please describe top and bottom balancing? I understand what it is to balance packs, but I'm unclear on the meaning and reasons for top or bottom balancing.
 
i would like clarification as well my understanding is

bottom balance, bring all cells to lets say 2.75v and connect
top balance, bring all cells to full charge and connect

top balancing gives you the most capacity and is best for powerwalls
bottom balancing lets you run down to closer to 0% without killing cells, more useful in vehicles but you lose capacity

in a bottom balance scenario all cells hit bottom together
top balance they all hit 100% together
 
Its just what it says. Its either you make them even in the bottom or the top.

Ie either you start all cells at 4.2v and have them equal at top.

Bottom then you go to your lower setting lets say 3.2v and make them even there.

No matter what you will in theory get the same capacity. For high current systems the bottom type may be better in terms of saving the battery since when you get low all get low at same time and you will stop the application. If you have top balanced them 1 pack or a couple of them will go dead before rest do and you may kill it.

What you use for your home made powerwall all depends on the application. What is worst: Killing a pack due to charging it to high or taking it to low? Neither way you choose i recommend some kind of active monitoring that can disable discharge or charge when that pack that reaches the bottom or top end goes bananas :)


For passive balancing you almost always top balance since its quite useless to drain energy in the bottom.. You want to do that in the top end when you have excess.

Active balancing where you transfer energy that can be done in booth ends and therefore extend the capacity range slightly....
 
I suppose technically you could balance at any voltage, my charger will allow a balance charge at storage voltage although I don't remember using the option.

Think of top balance as "equalize all the cells at the highest voltage I wish to allow them" and for bottom balance "equalize all cells at lowest voltage I wish to allow them". It's not any specific voltage it's just the maximum or minimum you wish them to operate to.

When a really well matched and balanced low internal resistance pack goes off the curve it's obvious, there's no there there any more.
 
Have a question that relates to top and bottom balancing. The DIY pack builds that I have seen with 18650s, most are 100 or more cells per pack. Lets take for example100P14s that I saw for a DIY powerwall. For thepack there are 1400 cells, but lets just take one of the 100p that are then wired in series, ithas 100 cells. These are soldered with fuses, battery bars etc. how do you go about discharging each of the 100 cells for a bottom balance? And assuming this is part of a larger 100P14s pack, each 100 in each of the 14 in series would also have to be bottom balanced.

Is the only way to balance before building the 100p pack? And they you don't ever balance again?

Thanks in advance,
Robb
 
First of all its 14s100p and not the other way around. Its not 100 strings.. Its 14 packs in series.

1 pack of 100 cells are in parallel = they are balanced together. Cant be anything else.

No matter if you bottom or top you take your 14 pack and paralell at cel voltage. They will then balance to each other and thats where you have your balance point set. If its bottom or top or middle doesnt matter. All cells in parallel will equalize.
 
And it is the job of a BMS to balance the individualpacks in series while in use. (top or bottom if you choose a decent bms)
 
Cells in parallel is like having two (or more) tanks of water that have their bottoms all interconnected. No matter how much water you take out of one tank, the others will flow and balance out. No matter how much you dump into one cell/tank, it will flow out into the other cells/tanks in parallel. They cannot be in any other way.

Cells in series is like having a way way check valve between each cell/tank. Nothing can flow between cells/tanks in this mode
 
RobbMelanon said:
[...]

Is the only way to balance before building the 100p pack? And they you don't ever balance again?

Great question -- and great thread, thanks OP.

I am currently practicing my 'tuning' skills, in building packs that are capacity matched -- precisely matched. My theory is, that if I can top balance, then discharge to bottom, and its still balanced then heh ... that would be awesome, eh?

my aim is to run 10-20kWh without any real BMS. NB I aim to monitor these packs, and top them off at ~ 3.92V (normally).

so YES, to answer your question here, I believe that if you solder/weld up a pack, with mismatched (even 1%) capacities then you will be stuck with a battery that goes out of balance....



Cells in parallel is like having two (or more) tanks of water that have their bottoms all interconnected. No matter how much water you take out of one tank, the others will flow and balance out. No matter how much you dump into one cell/tank, it will flow out into the other cells/tanks in parallel. They cannot be in any other way.


um ... no, not quite.

in my experience this very rationale expectation is only true when charge/discharge rate is relatively low. think < 0.1C ... once the current gets up to 0.5C and certainly any more, then the cell's relatively different internal resistances start messing with everything. IE effective capacity changes, differently for each pack, and therefore battery goes out of balance.

so my practice/idea/tuning thoughts (above) are ONLY valid with loads and charging at relatively low rates.

if the cells were all homogeneous (IE new, from same-ish batch etc) then their internal resistances would generally all be matched, so then one can go higher in the C rate before things get wonky. but I'm using harvested (used) cells exclusively, so ... tuning practice for me.

I no longer use the cell repacker on this site, as it 'only' (great job though) works with matching capacity. since I'm only making smallish batteries (1-3kWh) at the moment, which don't have enough cell in each pack (low P count) then the standard deviation on internal resistances in each one is far too high. hence me developing another way to 'tune' the construction to _try_ to match this as well as pack capacities.


*DCkiwi puts on fireproof jacket, so flame away* ; P

ps as I noticed on one of Jehu's vids, a smart guy at EV West said the trick is to have it bottom balnced, then charge it, and the battery still balanced at the top. this is the trick, Im quite convinced. its not about one, or the other, really. albeit they both can be useful in their own way..
 
I'm currently Running my 4.6kWh bank as a Peak energy offset power source without a BMS. I created it using tested 1s packs of 8 cells that are the same as as close to the same Ah quantity as possible. I have 56 of these in my 7s setup. I top balanced (@4.1v/cell) maually and got them within 8mV per cell and I watch them during the discharge and as long as I keep the pack above 3.2V/cell I stay within 40mV spread. Now if I start to discharge more than that it gets worse pretty quick.

Just food for thought.
 
sil5er said:
Cells in parallel is like having two (or more) tanks of water that have their bottoms all interconnected. No matter how much water you take out of one tank, the others will flow and balance out. No matter how much you dump into one cell/tank, it will flow out into the other cells/tanks in parallel. They cannot be in any other way.


um ... no, not quite.

in my experience this very rationale expectation is only true when charge/discharge rate is relatively low. think < 0.1C ... once the current gets up to 0.5C and certainly any more, then the cell's relatively different internal resistances start messing with everything. IE effective capacity changes, differently for each pack, and therefore battery goes out of balance.

The cells that are in a parallel pack will feed each other at a very low resistance. We aren't talking about connecting a 4.2V cell with a 3.8V cell or of greater distance. The difference is of milliVolts of one another, if any at all. So, as you stated, "this very rationale expectation is only true when charge/discharge rate is relatively low" is exactly what I was referring to and what is going on here.


The only way this won't be true is if there are so few cells in parallel, and a high load is applied to them, that they can't balance themselves out rapidly. However, once the load stops, they will balance themselves out with each other, even at a higher resistance, and even at higher voltage difference (which, voltage can't be different due to Laws involved)
 
Korishan said:
sil5er said:
[...]

in my experience this very rationale expectation is only true when charge/discharge rate is relatively low. think < 0.1C ... once the current gets up to 0.5C and certainly any more, then the cell's relatively different internal resistances start messing with everything. IE effective capacity changes, differently for each pack, and therefore battery goes out of balance.

The cells that are in a parallel pack will feed each other at a very low resistance. We aren't talking about connecting a 4.2V cell with a 3.8V cell or of greater distance. The difference is of milliVolts of one another, if any at all. So, as you stated, "this very rationale expectation is only true when charge/discharge rate is relatively low" is exactly what I was referring to and what is going on here.

The only way this won't be true is if there are so few cells in parallel, and a high load is applied to them, that they can't balance themselves out rapidly. However, once the load stops, they will balance themselves out with each other, even at a higher resistance, and even at higher voltage difference (which, voltage can't be different due to Laws involved)

true dat.

but in the context of the OP's questions, I think its still fair to point out problems with balancing when high ( >0.2C?) dis/charge rates are involved. To use your analogy, its like some of the cups get holes in and start leaking water. which then creates a further problem when load/charge is removed, and pack internally 'balances' ... further compounding the issue of a pack with different capacity from its neighbours, thus throwing the battery out of balance.

yes, this is very definitely more of a problem on smaller batteries, which are easier to load up.

if it was only one pack, IE a 1s 'battery' then this would all be academic. balancing by definition is only in play when there are multiple packs in series. the aim is to have a battery that stays balanced from top to bottom of its charge, and back again.
 
I studied the whole bottom balancing concept onsome of the Ecar/Ebike videos on Youtube this weekend. It looks like the guys involved with that aspect of battery packs are very concerned about bottom balancing first then bringing the whole pack up and top balancing ...then maintaining an on-going monitoring of the pack for long term while in actual use. The procedure I got was to discharge the pack with a load to a safe bottom balance before top balancing and this was done at a low discharge rate that could be monitored. They then top balance. But again it is done as a whole pack ...meaning cells in parallel are looked at as just one cell and are all brought to a safe discharge voltage that matches other cells in parallel. What I wonder is if this is a good procedure ..why not discharge each cell to a specific safe voltage of say 60% capacity....before building a pack. Then once the pack is built you should have every cell be the same voltage and now you top balance for actual use. It seems to me then every cell would be balanced both in parallel and in series. Thoughts anyone?
 
Headrc said:
But again it is done as a whole pack ...meaning cells in parallel are looked at as just one cell and are all brought to a safe discharge voltage that matches other cells in parallel. What I wonder is if this is a good procedure ..why not discharge each cell to a specific safe voltage of say 60% capacity....before building a pack. Then once the pack is built you should have every cell be the same voltage and now you top balance for actual use. It seems to me then every cell would be balanced both in parallel and in series. Thoughts anyone?

Time. It takes less time to discharge the whole pack than individual cells. Plus, it's a lot easier to manage the load. It's more difficult to keep a constant 300mA over 1 cell, than it is 3A over 10 cells. Each cell still sees 300mA, but it's easier with the 3A load vs the 300mA load.
Also, every cell in parallel will 100% be at the same voltage, regardless. It not possible for them to not be the same. Some smart guys math/theory/principle.

The reason why the eVehicle and eBike world do more of the bottom and top balancing is because they need to know exactly if they have 5mins left, or 15mins left, or 30mins left on a pack voltage. With lithiums, the SoC curve is so flat for so long, it's hard to tell based on voltage and overall capacity alone.

Plus, doing a bottom balance with a load on the whole pack, you can quickly see if a pack is having an issue as it'll drop load voltage quicker/farther than a fuller voltage. Example, a weak cell could drop from 4.2V to 3.9V under 2A load. But at 3.4V it could drop to 2.8V under the same load.
 
The thing with packs getting out of balance is due to packs' internal resistance being different. If the difference is too high, the packs with higher internal resistance burn some energy internally as heat; the higher the current, the more they'll burn and they will keep getting more and more out of sync as high current is being applied.

If -on the other hand- you manage to get all the packs on the same internal resistance, even if the capacities are different, you could keep them in sync through many cycles, by carefully calculating the top or bottom battery voltage, depending on whether you bottom or top-balanced respectively.

However, with packs containing lots of cells (say 50 or more), internal resistance averages out and it really cannot differ that much between the battery's packs, so big packs matched in capacity rarely get out of sync.
 
Headrc said:
Thoughts anyone?

Well, it doesn't make much sense to do both. Top balancing the battery after bottom balancing it leaves you with a top balanced battery and the bottom balancing it was rather pointless.
Top and bottom balancing follow two different concepts that don't work together.

Top balancing means the cells will drift on the discharge and will balance on charge while bottom balancing is the opposite.
A bottom balanced battery can be charged with any CCCV charger without balancing if you calculate in advance how long you need to charge it so you don't put too much energy in or have a automated cutoff once the first cells reaches the set end of charge voltage. The cells will have different end of charge voltages when you do that but all of them have the same energy stored and will have equal voltages when discharged.
A top balanced battery works the other way round, the cells will drift on discharge and you have to stop discharging at the point where the first cell hits the end of discharge voltage. All remaining cells will have some charge left at this point which can't be used. When fully charged all cells will have the same voltage, but not the same amount of energy stored.
 
Yeah, the greater the number of cells in parallel, the easier to balance, sort of. As long as you don't have a leaker in the mix, then it should be fine.

Also, with the greater the number in parallel, the harder to imbalance them as the amps gets spread out over the cells. So, even if a cell(s) have high resistance, 200A across 100 cells is 2A/cell, but across 300 cells is only 660mA / cell. So that makes a huge difference too. If you plan on large amp draws quite often, go with larger packs (or multiple series in parallel). In this case, almost, quantity is better over than quality (in reference to IR)
 
Ok ...this a; makes sense to me. Except for what Darkraven just stated. As I understand it from the Ecar community, the bottom balancing is to make sure your top balancing is equal voltage at the top of the charge curve. That argument is made here on this video at around 5:30.https://www.youtube.com/watch?v=rQC_9TwyjTk
 
My understanding was that if you bottom balance, then when you get to the top it should already be in balance. You dont do both bottom and top balancing (i.e. after discharging all to the same voltage at the bottom you dont charge all to the same voltage at the top). A perfect bottom balancing should result in a pretty good top balance. If you actively top balance after that, then you wont know exactly how much capacity the system has, unless you are using a monitoring system that includes a shunt or other current sensor that computes capacity.
 
Back
Top