How important is balancing a battery pack?

Hello,

I was having a discussion with someone.

"Balancing LiFePO4 cells to absolutely the same charge voltage is not that important, the charge curve shows that there is very little difference between a 3.55V and a 3.65V charge for example, fussing that detail will not make a meaningful difference to the capacity delivered. All these RC devices are cool if you like to spend time maintaining a battery, but such methods are not fail safe, they don’t protect your battery if you make a simple mistake and everybody makes mistakes occasionally. IMO a battery that takes care of itself and is protected in a fail safe manner is infinitely better."

What are your thoughts on this quoted statement? I won't reveal who said it, only that I don't really know if it's a solid concept to follow.

http://www.youtube.com/watch?v=a03cOIp5Euw Hehe, thanks.

I think they are balanced close enough if they are between 3.55v and 3.65v.

All the ones I have tested don't take in much over 3.55v.

Even a cheap B6 charger will get them that close or closer.

One of the goals in building a good battery pack is to build one that is easy for a Hyperion 1420 to balance.

It can get them balanced very close quickly if the cells are a close match.

You can watch the cells being charged on a display and watch the voltages on the cells rise together and see the quality of your cell matching job.

It’s not a giant problem if they are out by more than 0.1V within the same cell-group. In a large pack I think it matters less than in a small pack where the final voltage is more affected.

During maintenance I look at two parameters:

• Cell-to-cell balance within a cell-group (unit battery) of which I try to pull within 0.03V when balancing, and
• Cell-to-cell balance across the entire pack. Here I do not allow more than 0.05V after balancing.

In the real world, I’ve seen the disparity across the pack be more than 0.15V out – which certainly robs distance. Think of it this way: The charger will stop when peak voltage is reached, but it could be that one cell is running hot, and another at the other end is running cold: The hot cell can exceed HVC without a BMS, the overall capacity of the pack is not fully topped off, and the cold cell(s) act like a drain because they contribute less current over the course of discharge.

The short story is that if you are going on a short ride, it doesn’t make that big of a difference. However if you are depending on that pack to get you far down the road, it is wise to balance as best you can to achieve top performance. :wink:

Checking every now and then... KF

LiFePo's don't balance them and you will find a weak battery in a short time.

LiFePo's seem to need more balanceing than LiPo's. LiPo's might drain at differant rates on a graph but seem to charge up to almost balanced with just bulk charging.

LiFePo's don't seem to do that. They may drain the same or differant, but time to charge them, you better balance them with a BMS if you don't want a ruined battery pack in short time.

I run both types, mostly now LiPo's since they are stronger and lighter for the same capacity. I can bulk charge them from almost 80% DOD and not need to balance them for a long time.

If I try that on my LiFePo's I see high cells way before others start to peak. So I need a BMS to manage them properly. It is also much safer than LiPo's as long as you have a BMS and proper charger that works togeather.

Dan

If you start their life with a good balance, and keep it up for a couple more in the early days, you'll find LiCo doesn't need much more balancing after that. But yes, I would argue it is important to balance a pack, especially when you first get it.

I'd say the really important thing is knowing if you have an inbalance or not. Even with a bms automaticly balancing your pack, nothing substitutes for knowing if it works or not. Even with a simple setup like a typical lifepo4 with bms pack, you need to know when it's time to leave it charging for a few days if a large imbalance should occur. That's one reason I like the signlalb bms. If all the led's don't light, leave it plugged in longer.

It's definitely important to make sure a new pack gets balanced initially. Call it breaking in the cells or just call it balancing it, a few light cycles followed by charging on a balancing charger, or longer than normal periods on a charger +bms balancing it will insure that you get off to a good balanced start before you try to set any personal distance records with the new battery.

One lifepo4 cell at 3.55v and another at 3.65 would be what I call balanced. As you correctly state, above 3.55v it doesn't make that much difference because you are looking at surface charge that will either dissipate itself or dissapear instantly with a very small load. From 3.55v to 3.65v there is not much watthours in it.

But if the battery had one lifepo4 cell at 3.35 and another at 3.45, the same .1v difference would be quite significant. In that case, there would be some wh difference between the two cells. Because of the flatness of the lifepo4 discharge curve, it could be a very singificant difference if they are that different at that point in the discharge.

Lico is a bit different IMO. From a fully charged 4.2v, good sound cells won't self discharge much more than .05v. Apparently less overcharged because of the risk, there is more wh involved when a lico cell is at 4.15, and another is at 4.05.

As for importance, of an inbalance, the main important thing is to not let an inbalance drift further and further out of balance with each cycle. No, balancing every cycle is not needed with any chemistry if the cells are good. But once cells start ageing, a weak cell might require balancing every cycle to keep it from drifting into a very unbalanced situation that will not only reduce your total wh capacity, but also perhaps put extra load on paralelled cells that increases the wear and tear on them.

http://www.endless-sphere.com/w/index.php/Balancing:_What_is_it_and_when_is_it_necessary%3F

Regarding the title question:

It depends on the quality and age of the cells, and how well they are matched to each other.

If they are new, high quality cells, they will likely stay in balance on their own, if they started out that way.

As they age, some will lose capacity or change internal resistance faster than others, and at that point, balancing may become very important. But balancing still won't bring back the capacity loss--it just helps ensure that all cells get fully charged.


With lower quality or poorly matched cells, they may not even start out balanced, and may require balancing every time they are charged.


If a cell's characteristics cause it to rise in voltage faster than the others, whether they are fully charged or not, and the BMS of the pack has cell-level HVC, then it will cause the pack to stop charging, and the other cells will not be fully charged (possibly that cell, too).

If the charger has balancing circuitry of one type or another, it can use that to equalize voltage on the cells, then resume charging, until all cells are as full as they can get, by whatever method the BMS uses to detect that (normally cell-level HVC).


In the case of a cell that has lower capacity but shows a higher voltage than the others when it reaches it's full charge (or it reaches full charge first), it might not make any difference to the pack if voltage is *not* drained down on that cell and charging does not continue: Assuming that this cell is the lowest-capacity cell in the series string, and that none of the others has charged to less than this cells' present capacity, you will still get pack-level/cell-level LVC at the same point whether you balance the pack or not, in theory. In reality, I don't know that for certain as I have not tested the theory.

Regarding the implied question in the OP:

A cell-level protection, LVC for while riding, is important for all chemistries, if you really want to protect the battery and lengthen it's lifespan.

Pack-level LVC only helps if the cells are all the same capacity and characteristics (not just "balanced" which really only means they're the same voltage when fully charged).

If you have cells with different internal resistance, or different capacity, then with a pack-level LVC, if it is set to x-number-of-cells * whatever-LVC-for-a-cell-is, then some cells will be BELOW that cell-level LVC and some will be above it, once pack-level LVC is reached. If the difference in cells is significant, that could be a disaster for the overdischarged cells.

Let's say that you have 20 5Ah RC LiPo cells in series. They're old, and one of the cells only has 4Ah now, and another only 4.2Ah, while the rest still do 5Ah. Let's say you put your cell level LVC so that they're usually only discharged to what would amount to 4.5Ah anyway, for a safety margin. When you discharge that pack to a full 4.5Ah, using a pack-level LVC (or worse, just watching the watt meter for Ah usage!), you have just used up 0.5Ah more than that one cell had in it to start with--it's been severely overdischarged.

I don't know exactly what voltage it will be at, at that point, but it could be possible that in this or a worse scenario to bring it below 0V, and actually reverse-charge it. At that point, "dramatic failure" of the cell could be possible. I wouldnt' want to be recharging it afterward, in my house, unprotected by the curtains, that's for certain.


So regardless of any kind of balancing being done, it could be extremely important to have some sort of cell-level LVC that is reliable and will cut off pack discharge whether you think you want it to or not. Unless you are in a riding situation where your life is at stake and overdischarging the pack to the point of irreversible damage no longer matters, I think you would want to protect the pack from this kind of potential problem.

That says it all.

Balancing is just a tool that is used to attempt to have a pack with all the cells as full as possible, so that the lowest capacity cell in the pack, which limits the entire packs capacity, is as fully charged as possible. Any undercharged cell may limit capacity even more.

But what's really cruicial with most chemistries, is stopping before you are 100% discharged. A really unbalanced pack is hard to know when to stop, unless you have cell level monitoring and are watching, or have cell level lvc.

Then with lifepo4, just watching cell level monitoring is harder, because lifepo4 tends to drop off the cliff so abruptly. So a bms with actual cell level lvc is good for those who will deeply discharge the pack. With lico, you see it slowly getting closer and know when it's coming, so it's not so hard to know when to stop in time.

Just wanted to say for a neewbie this thread was excellent reading and well explained, ES has been an excellent tool for me

"Unless you are in a riding situation where your life is at stake and overdischarging the pack to the point of irreversible damage no longer matters"

Made me laugh. Makes me think of a scene from some random action movie.

when reading these replies, keep in mind that the voltage expressed on the terminals of the cell has not been shown to correlate to state of charge. a battery cell charged to 3.55V is not fully charged. it has to reach 3.65V to be fully charged. some cells will self discharge so fast that they cannot be pushed to 3.65V by the balancing current.

when we talk about balanced, we are talking about having the pack composed of cells with equal amounts of charge stored in each cell. the only way to make sure you fully charge a cell is to charge it until the voltage at the terminals reaches 3.65V.

if the terminal voltage drops after that, it does not mean it is uncharged. it will still have the full capacity minus whatever bleeds off to self discharge from the fully charged 3.65V, which is minor compared to usable energy. so even if a cell has 3.37V on it it can still hold the full capacity of charge.

dnmun said:

when reading these replies, keep in mind that the voltage expressed on the terminals of the cell has not been shown to correlate to state of charge. a battery cell charged to 3.55V is not fully charged. it has to reach 3.65V to be fully charged. some cells will self discharge so fast that they cannot be pushed to 3.65V by the balancing current.

Click to expand...

Bullshit again. A123 26650 M1 datasheet specifically says 3.6V as the maximum charge voltage at 25 degree C.

dnmun said:

when reading these replies, keep in mind that the voltage expressed on the terminals of the cell has not been shown to correlate to state of charge. a battery cell charged to 3.55V is not fully charged. it has to reach 3.65V to be fully charged. some cells will self discharge so fast that they cannot be pushed to 3.65V by the balancing current.

when we talk about balanced, we are talking about having the pack composed of cells with equal amounts of charge stored in each cell. the only way to make sure you fully charge a cell is to charge it until the voltage at the terminals reaches 3.65V.

if the terminal voltage drops after that, it does not mean it is uncharged. it will still have the full capacity minus whatever bleeds off to self discharge from the fully charged 3.65V, which is minor compared to usable energy. so even if a cell has 3.37V on it it can still hold the full capacity of charge.

Click to expand...

Some tend to settle down to about 3.41v when fully charged.

http://www.batteryspace.com/prod-specs/6450.pdf