Balancing Batteries - how to?

It occurred to me that I really don't know the 'details' of how chargers balance a battery pack.

I know that a balanced pack has all the cells within .1 (or better) volt of each other. I know that balance chargers have a discharge circuit. But I don't know how the balance is achieved.

Does the charger limit the charge to individual cells by controlling voltatage and amperage fed to each cell through the balance taps?

Do the discharge circuits kick in to limit the charge voltage to a cell that is too high, or is this a cycle type thing, where the pack is charged and then individual cells discharged to the correct voltage?

What would be the ideal way to do this? Charge each cell individually?

Thanks,
Jim.

When I began to learn about balancing I asked the same questions. Here's the thread with the answers.
http://endless-sphere.com/forums/viewtopic.php?f=14&t=29892

Both my RC charger and my Lifepo4 charger work more or less similarly. Pretty much your second guess.

The RC charger has the balancer built into the charger. The battery has no bms. The charger will drasticly slow down the rate of charge as the pack approaches full. As it does that, any cells that start to go too high voltage will start discharging, letting the slow ones catch up. If any cell gets really high, the charger will shut off for a few moments. The process repeats many times, and can take quite some time if you have a cell way out of balance.

The ping battery, or similar lifepo4 setup is slightly different. The charger is seperate from balancer, but works nearly the same. When the pack gets to a certain voltage, it shuts off for awhile. Meanwhile, any high voltage cells are getting discharged by the balancer that is part of the battery bms. A bit cruder than the RC charger, and potentially taking longer to balance a really out of balance pack.

You can take an RC charger, select 1s charging, and use it to quickly top off a really low cell. But by the time you need that, usually you have a cell in that pack that needs to be retired.

Some people with electric cars do exactly that, one charger for each 12v lead battery in the car.

Ok. I think I got it.

The pack is bulk charged while monitoring each individual cell voltage. Then 'high' cells are discharged to match up the volts of the rest of the cells. If you have a super performing cell, the charging stops, that cell is discharged, and then the operation picks back up again. Is this correct?

And, If you have one really low cell, are the rest of the cells discharged down to the low cell voltage in order to achieve 'balance'? (Can I assume that a balanced charger has some sort of reporting that says, "Hey, you got a bad cell here", or is it me and my voltmeter after I note the final charge voltage is low?)

Thanks,
Jim.

Also, the balanced chargers don't do any charging through the balance leads. Right? That would be a manual operation to try to fix a low cell.

Thanks,
Jim.

JimW said:

Also, the balanced chargers don't do any charging through the balance leads.

Click to expand...

Correct. No charging. The balanced wires are used for discharging during balancing.

JimW said:

Also, the balanced chargers don't do any charging through the balance leads. Right? That would be a manual operation to try to fix a low cell.

Thanks,
Jim.

Click to expand...

Correct, only when "rigged" via adapters by adventurous users.

One interesting thing is that some low cells are merely lower SOC when originally assembled. Trouble is, "bleed balance" method may never be able to properly match an equal capacity but slightly lower SOC cell to the others in the string. Actively bumping up a cell or several in series through balance wiring can sometimes deliver a better balance than the normal RC charger bleed method.

Of course, if you keep coming back to one low cell after discharge but higher than the rest after bulk charging, it's probably less capacity than others in the group and that cell capacity is limiting capacity of your pack.

Exactly. The balance leads normaly monitor cell voltage if hooked to the charger, and bleed away excess voltage if you have selected balance charge on the charger.

But you can connect to the balance leads of just one cell, select charge 1s, and then at low amps (1-2 amps) bring up a low cell quicker than waiting for the balance charge system to work. This can be handy when you have one that is way out of balance. For example, you set for balance charge, and hours later the charger times out before it finishes. In 30 min, you could likely bring up the low cell by charging it at 1 amp through the balance wires.

It's not so adventurous to do this with an RC charger that will shut off when it should. It's risky if you do it with a bare 5v power suppy. I wrecked two packs that way, fell asleep for a minuite,and the cells overcharged. Never again with a plain 5v phone ps.

Thanks to all who replied. This has been very educational.

One point I'm still not perfectly clear on - If you have one really low cell, are the rest of the cells discharged down to the low cell voltage in order to achieve 'balance'? (Can I assume that a balanced charger has some sort of reporting that says, "Hey, you got a bad cell here", or is it me and my voltmeter after I note the final charge voltage is low?)


SamTexas - Thanks for the Link! I learned a lot.

Thanks,
Jim.

JimW said:

SamTexas - Thanks for the Link! I learned a lot.

Click to expand...

You're welcome.

JimW said:

One point I'm still not perfectly clear on - If you have one really low cell, are the rest of the cells discharged down to the low cell voltage in order to achieve 'balance'?

Click to expand...

Depends on how low that low cell is. But yes, the balancing scheme is to discharge all other cells until they reach the lowest cell voltage. Don't get your hope up too much on BMSs and balanced chargers. They barely work, if at all. In my opinion, BMSs are the weakest link in the Ebike realm. Not sure if you knew, but around here, "BMS" is lovingly called "Battery Murdering Systems".

"BMS" = "Battery Murdering Systems" LOL :lol:

Perhaps I have found a niche in which I can contribute a bit.

Thanks,
Jim.

most RC ballance chargers will keep the charged cells from going over-charged *( draining power via the ballance wires ) while the low one's catch up ( the charger applies power to the charge leads to the entire pack in series.. on a large ebike pack this can take a LONG time.. depending on the charger.. since i dont use a BMS i just tap into thatlow cell/group via the ballance plug and charge it directly to match the others ( saves hours of waiting )

the RC gizmo's that " ballance " only, without charging, like the battery medics, drain the higher one's to match the lower one's..

JimW said:

Perhaps I have found a niche in which I can contribute a bit.

Click to expand...

That would be a welcome news. If you're seriously considering a new BMS or balancing scheme, I sincerely hope that you abandon from the very start the idea of bleeding (discharging) to achieve balance. It's (in my opinion) a pathetic way to achieve balance. A BMS has control of EVERY SINGLE CELL in a pack, boosting the low cells is more efficient and quicker than bleeding the high cells at a extremely slow rate.

Also abandon the idea of powering the BMS with just the first few cells, thus creating an imbalance to the pack it's trying to balance! Also make sure to include a user switch to turn off BMS when not needed (mostly for long term storage). Last and not least, make sure the BMS works!

SamTexas said:

If you're seriously considering a new BMS or balancing scheme, I sincerely hope that you abandon from the very start the idea of bleeding (discharging) to achieve balance. It's (in my opinion) a pathetic way to achieve balance. A BMS has control of EVERY SINGLE CELL in a pack, boosting the low cells is more efficient and quicker than bleeding the high cells at a extremely slow rate.

Also abandon the idea of powering the BMS with just the first few cells, thus creating an imbalance to the pack it's trying to balance! Also make sure to include a user switch to turn off BMS when not needed (mostly for long term storage). Last and not least, make sure the BMS works!

Click to expand...

Very true.

Balance means getting every cell to the same state of charge, in percentage terms, where charge is defined as the usable energy available in the cell. Cell voltage is a very unreliable indicator of state of charge, except for two specific conditions.

The first is when the cells are nearly completely discharged, when the cell voltage under load suddenly drops. Discharging all cells to the same set voltage under load means they are all effectively at 0% state of charge and therefore balanced. This isn't generally a particularly useful thing to do (although Jack Rickard would disagree, I'm sure).

The second condition where voltage indicates state of charge reasonably accurately is when cells reach the same voltage under charge and the charge current has fallen to a low level through every cell in the pack. This is what most RC chargers try and do on a balance charge, by switching the charge current off, bleeding down the highest cells, then switching the current back on to charge the low cells (and the cells that have just been drained a bit). As has been said above, this can take a fair while. Charging cells with individual chargers will get them all to 100% state of charge more quickly, as will using high current shunts to bypass those cells that reach full charge first and still allow charge to those cells that need it.

Discharging cells after charge to bring them to the same nominal terminal voltage can actually put the pack further out of balance, all it really does is give a nice looking voltage display................

"Charging cells with individual chargers":

Jeremy Harris said:

Discharging cells after charge to bring them to the same nominal terminal voltage can actually put the pack further out of balance,

Click to expand...

That's new to me Jeremy. Can you elaborate?

Example: I bulk charge my 10s pack to 41.0V. Once the charge is complete, I check individual cells' voltages. They are in the 4.08V - 4.11V range. So I manually discharge all cells to 4.08V. Did I just make my pack less balanced?

hehe.. yes it does...

SamTexas said:

Jeremy Harris said:

Discharging cells after charge to bring them to the same nominal terminal voltage can actually put the pack further out of balance,

Click to expand...

That's new to me Jeremy. Can you elaborate?

Example: I bulk charge my 10s pack to 41.0V. Once the charge is complete, I check individual cells' voltages. They are in the 4.08V - 4.11V range. So I manually discharge all cells to 4.08V. Did I just make my pack less balanced?

Click to expand...

OK, let's say we have 10 cells in a pack and they have a natural capacity variation between cells of just 2% (in practice I think that 5% variation might be more typical for many Chinese made cells). When they are all charged to the point where at the chosen cell full charge cut-off voltage every cell charge current has dropped to a similar very low level, then they will all be at 100% state of charge.

Now, if you discharge the cells to a nominal voltage, say your 4.08V, how can you be sure that you have taken exactly the same capacity out of each cell and that they are still at the same state of charge? The natural small variation in capacity, surface charge discharge rate, internal resistance, even small inter-cell temperature differences, between cells will mean that you will almost certainly take more charge from some cells to bring them to your nominal "balance" point than you will from others. The result will be that a pack that was in balance, with all cells at 100% state of charge, will now be slightly out of balance in all probability, with maybe a few percent difference in state of charge across the cells.

This all stems from the fact that cell voltage isn't a very reliable indicator of state of charge, and is one reason that I run an Ah meter on my bike to keep track of the capacity I have remaining in the pack.

I suppose it's also worth noting that cell balance doesn't really matter too much if you only use a part of the pack capacity between charges, as it's unlikely that it'll get out of balance enough to cause a problem with one cell/cell group going down prematurely. I only balance charge my folding bike pack once in every four or five months, most of the time I just bulk charge it as it's quicker (this pack is too small to be charged with my high current DC DC converter charger).

I still don't understand it. But thanks for the explanation.

SamTexas said:

I still don't understand it. But thanks for the explanation.

Click to expand...

Jeremy has made about 6-10 posts on this subject recently. I think if you read all of those including some posts from others near his posts, then I think you will understand.

Hmmm.

Would it be more important to have a pack with the voltages of each cell all the same, or a pack where the watts of power in each cell was the same (regardless of voltage reading).

I'm not sure how one would practically determine the total watts in a cell (except by discharging it).

If cells had differing watt capacities (due to age or mfg differences) how much difference is important, or is it only important not to over exhaust the lowest power cell in the pack?

I'm not sure I'm asking this question correctly (I think is somehow relates to the true state of charge of a cell - and how it can be determined). Is this somewhat addressed by some of those fancy chargers that output graphs of each cell?

Thanks,
Jim.

JimW said:

Hmmm.

Would it be more important to have a pack with the voltages of each cell all the same, or a pack where the watts of power in each cell was the same (regardless of voltage reading).

I'm not sure how one would practically determine the total watts in a cell (except by discharging it).

If cells had differing watt capacities (due to age or mfg differences) how much difference is important, or is it only important not to over exhaust the lowest power cell in the pack?

I'm not sure I'm asking this question correctly (I think is somehow relates to the true state of charge of a cell - and how it can be determined). Is this somewhat addressed by some of those fancy chargers that output graphs of each cell?

Thanks,
Jim.

Click to expand...

Cell voltage is generally a poor indicator of state of charge, so having cells at the same voltage (other than at the two conditions I've already mentioned) doesn't mean they are balanced, if we accept that balance means bringing all cells in a pack to the same state of charge.

State of charge refers to the usable capacity in the pack, or the cells, in terms of the percentage of usable energy (not power) left in it. A fully charged pack, with the cells all balanced, would have a 100% state of charge, a fully discharged pack that could no longer deliver usable power would be at 0% state of charge. It's worth noting that whereas the fully charged pack will have all cells at 100%, the fully discharged pack doesn't mean all cells are at 0%, in all probability only one cell in the pack will be at 0%, the lowest capacity cell and the one that has activated the LVC (assuming you have one).

If you have, for example, a 10 Ah pack, with a capacity variation between cells of +/- 1%, then when the pack is balanced at 100% state of charge the highest capacity cell might have have 10.1 Ah available, whereas the lowest capacity cell might only have 9.9 Ah available. Your usable pack capacity is set by this lowest capacity cell, at 9.9 Ah.

JimW said:

Hmmm.

Would it be more important to have a pack with the voltages of each cell all the same, or a pack where the watts of power in each cell was the same (regardless of voltage reading).

Click to expand...

Quick correction first: Not "watts of power" but energy content or the number of Whs in each cell.

In a perfect world, each and every cell in a pack has the exact a) same # of Whs, b) same chemistry and c) same internal resistance. In that perfect world, the cell voltage is a very reliable indicator of SOC. So when all cells are at the same voltage, the pack is perfectly balanced.

In the real world cells used in a pack have slightly (very slightly) different Whs, they don't age identically so their internal resistance are also slightly different. This leads to different rates of energy consumption (charging) and energy release (discharging), in spite of the fact that they all experience the exact same current during charge or discharge.

So, to be a perfect in a non-perfect world, we would NOT look at the voltage to determine the SOC. Instead we would have a sophisticate instrument to accurately measure the number of Whs going out of EACH cell during discharge. In other words, we need one watt-meter for each and every cell. With that data in hand, we know exactly how many Whs we need to put back into each cell at charging time. When charging is complete the pack is balanced, and we did not even look at voltage.

Back to the real world: We can't afford to have one watt-meter for every cell. We don't have a charger that understand the concept of SOC based Whs and NOT voltage. So we assume that we live in a perfect world (same capacity, chemistry and internal resistance) and we use voltage to determine SOC. It's not perfect, but it's PRACTICAL.

That's how I understand it.

SamTexas said:

Back to the real world: We can't afford to have one watt-meter for every cell. We don't have a charger that understand the concept of SOC based Whs and NOT voltage. So we assume that we live in a perfect world (same capacity, chemistry and internal resistance) and we use voltage to determine SOC. It's not perfect, but it's PRACTICAL.

That's how I understand it.

Click to expand...

Very dangerous to assume that cell voltage equals, or even approximates to, state of charge. There are a host of reasons (and I've already mentioned them) for cell voltage not giving an accurate, or even an approximate, indication of the state of charge of a cell.

I keep saying this, but I'm not at all sure it's being fully understood. There are two, and only two, circumstances where cell voltage can be used as a reliable indicator of state of charge. The first is when the cell voltage drops rapidly under load at full discharge, the second is when the cell charge current drops to a very low value when the cell voltage is capped at the fully charged voltage.

Any intermediate cell voltage measured cannot be assumed to give an indication of state of charge, in fact you could have the same cell terminal voltage for maybe a 20 to 30% variation in state of charge.

also varies by chemistry.... and cell brand..

AT REST...

Lifepo4 have such a flat voltage profile, a cell can read 3.2v at rest and be almost empty.. and 3.2v nearly full..

Lipo cell at 3.7~3.8v and be 30 to 80% SOC..

-0-0

apply a load and this changes .. drastically.