Charging Lithium Batteries—rather Balanced and in Series or Parallel?

[Hironiemus]

Hi,

I gathered some information and questions about battery charging, it would be great if someone
could verify and answer these respectively.

Actually I am new to all this and after I read here for ages I decided to
post for the first time.

I wonder which charging method would be best for charging
lithium batteries—balanced and in series or parallel?

For the technical part here this is what I found,
how lithium batteries are to be charged:

Parallel Charging

• On one side all the plus poles of the batteries and on the other side all the
  minus poles of the batteries are connected.
  Therefore you can look at the batteries as they were one large battery.
• The maximum (constant) voltage is to be set as 4.2 V for Li-Ion/Li-Po and 3.7 V for LiFePO₄.
• The maximum (constant) current can be found by multiplying the number of
  parallel cells by their capacity and should not excess 1 C
  (For example 4×2000 mAh batteries would give an maximum charging current
  of 8 A).
• Different battery size/capacity can be charged at once but then the charging
  current has to be calculated by multiplying the capacity value of the battery
  which is the smallest in capacity by the number of batteries to be charged
  (i. e. 1×600 mA 3×2000 mA would give a maximum current of 2.4 A).
• When charging parallel the charger (ideally) stops at 4.2 V, but it only
  detects the average value so if the batteries do not charge at the same
  rate, some could be overcharged and others not fully charged.
• If you start charging, the batteries should not have more than 0.5 V difference in voltage,
  because one battery can charge the other one, which can come out bad …

Serial Charging

• When charging batteries in series the plus pole of one battery
  is connected with the minus pole of the next battery.
• You must use a balancer.
• The maximum (constant) current is set as if you would charge a single cell.
  (i. e. when charging 4×2000 mAh the current must not exceed 2 A since
  the whole current will flow through every cell).
• One balancer cable must be connected before and after every cell.
  (Number of balancer cables equals number of batteries plus one.)
• The balancer ensures that the voltage in every cell won't exceed the
  terminating voltage (4.2 V Li-Ion/Li-Po/ 3.7 V LiFePO₄).
  This is done by checking the voltage after every cell and
  delivering an compensation current—if necessary—to get the batteries to the same level.
  (Voltage after first battery should not exceed 4.2 V, 8.4 V after the second, and so on).
• Charging batteries with different capacities should be possible but only with
  a charging rate that the balancer can deliver as compensation.

On this basis it seems to me balanced serial charging is better than charging
parallel because there is no risk in overcharging single batteries.
Am I right with this?

Are the balancer connectors basically positive or negative?
So would I need to wire negative, 4.2 V, 8.4 V, positive or vice versa?

Of course all of this applies only to a hobby charger.

Can someone tell me which charging method does the Pila IBC uses?
I think I heard it uses parallel charging.

Please point out, if i missed something out or got something wrong.

Oh and before I forget: Hello to everyone at CPF. :wave:

 

[Lion251]

When charging batteries in parallel, the voltage of all batteries will be exactly the same the whole time. So, when the charger stops at 4.2V, all batteries will be exactly full.
The 'detects the average' which you describe when charging parallel will occur when you charge in series without balancing leads.
0.5V difference is too much for Lithium cells when there is an appreciable charge left. At 4.1V, the cell is 90% full, at 3.6V it is almost empty. With cell voltages above 3.6V, I would rather take a maximum voltage difference of 0.05V.

When charging in series, I would advise against charging batteries of different capacity (more than 10% difference or so) at the same time. This would give the balancer more work than it was designed to do.

 

[zer0ne]

LiFEPo4 should be 3.6v?? And I hardly sees anything higher than 0.7C for LiCo cylindrical type.

 

[rlcrisp]

I'm looking into this too for getting my first hobby charger and building my first balancing rig.

One of the things you said about balancing is "4.2V after the first, 8.4V after the second...etc" doesn't seem right. Since you have leads on each side of each cell the balancer looks at the potential across each cell individually (Is cell #1 at 4.2V? Is cell #4 at 4.2V?). As long as you wire each cell up with the correct polarity, I'm not sure it actually matters at all if you wire them into the balancer in order (1-,1+,2-,2+,3-,3+) vs (1-,1+,3-,3+,2-,2+) - somebody correct me if that is wrong.

As a EE, charging in series scares me a hell of a lot less than charging in parallel. You can get yourself into trouble pretty easily going the parallel route before you even turn your charger on...

 

[Hironiemus]

Hi thank you for your answers.

Lion251: Thanks, these were facts I was rather not so sure about.
About the parallel charging: What happens if one battery has an lower internal resistance,
should this battery not charge somewhat faster
because more current flows through this battery?

zer0ne: Yes, I think LiFePO₄ was rather 3.6 V,
I also read sometimes Li-Po is 4.1 V rather than 4.2 V.
I think 3.7 V was the nominal voltage of Li-Ion cells.
Are the AW Li-Ion cells all LiCo? Also I wonder how could ever charge an
10180 correctly it is only 90 mAh, I did not see any hobby charger that could
be set to an cc below 100 mA.

rlcrisp: You are probably right about better just connecting the right cables,
I hope the manual coming with the hobby charger I want to buy explains it
well enough. But I still think it should be 4.2 V, 8.4 V, etc.
Given the batteries all have the same resistance you should be able to get
a fraction of the whole voltage after every battery. For example with four
batteries it would be 1/4 of 16.8 V, after the second 1/2 of 16.8 V and so on,
like in a wire with a high resistance, where you can get any voltage
in between the positive and the negative voltage depending on where you
measure.

In your notation 1+ and 2- should be the same as 2+ and 3- should be.
so if you start at 0 V and go up to 12.6 V for three batteries,
1+/2- should be 4.2 V, and 2+/3- should be 8.4 V
which gives a difference of 4.2 V for the second cell.
Since you can do this for every cell it is essentially the same you stated
("Is cell #1 at 4.2V? Is cell #4 at 4.2V?").

So can I conclude balanced series charging is generally considered safer?
Is there any drawback in serial charging? What could you do with
empty slots in a charging cradle?

 

[pae77]

There is an extensive discussion of the merits, method and relative safety of parallel charging li-ion packs (general principles discussed apply to the single cells most flashaholics use as well though) with hobby chargers by very knowledgeable RC folks in this thread. I encourage anyone interested in the subject to review it carefully.

Here is a brief excerpt from the first post in the above cited thread:

". . . How is the charge current calculated?
Simply add the currents of the single packs.
Say, one wants to charge a 1Ah, a 2Ah, and a 3Ah pack together, each with 1C.
1A+2A+3A=6A.
So these packs are charged with 6A.

There may, however, be the issue, that the packs connected in parallel have a different inner resistance (IR). The older a pack, the higher the IR. This means, that such a pack will not accept the charge current as willingly as a brand-new pack with a very low IR.
Now in the hypothetical situation that one has a lazy old 5000mAh pack connected in parallel with a brand spanking new 500mAh pack, and decides to charge at 1C with 5.5Ah, the chances are good that the small pack will see more than its share of 500mA.
My advice is this:
Know your packs and their health. If you are pushing for the maximum rated charge current of the packs, be sure to connect only those in parallel that are not way off concerning age, capacity, and general health.
If you, however, slowly charge with a small current over night, I do not see a problem.
More on this subject.

Generally, when one has the time, I advise to charge LiPos as slow as possible. Let's say, you want to charge your lipos over night. In this case, there is no need to charge than at 1C or even 2C.
If you have 12h time, use it. Set the current to 1/10C.
I do this all the time. Less current means less problems, less thermal stress for the chargers and power supplys, less risk of anything malfunctioning.

Questions that might arise:

- How come this works with packs of different capacities?
This is because they are connected in parallel. This way, they will always have the same voltage. It is not possible to overcharge a single pack. One can parallel charge packs of any capacity one wants, no matter how big, small, or different they are.

Although all pipes have different shapes and volumes, they are always filled to the same level. The same applies to the packs' voltage levels.

- What happens if I connect packs with different states of charge?
A current will flow between the packs. However, this current is not very high, even if one pack is full and the other depleted. Yet, it should be avoided to connect packs with a state of charge differing over 50%

Here is a site that explains everything in a nutshell with great diagrams:
http://sites.google.com/site/tjinguy...allel-charging . . . ."

Source: http://www.rcgroups.com/forums/showthread.php?t=932319

 

[45/70]

". . . How is the charge current calculated?
Simply add the currents of the single packs.
Say, one wants to charge a 1Ah, a 2Ah, and a 3Ah pack together, each with 1C.
1A+2A+3A=6A.
So these packs are charged with 6A.

Real quick, that doesn't sound right to me. In theory, the 1A pack would receive a 2C (2A) charge. I did not have time to check out the link, but will later.

EDIT: The proper way to do this when charging dissimilar capacity packs, or cells, is to determine the maximum charge rate for the lowest capacity cell, or pack. Then make certain that the charge rate does not exceed that maximum. In this case, for a LiPo pack of 1A, the maximum rate would be 1A. So, with three packs being charged, 3x1= 3A rate to charge the three packs in parallel.

Dave

 

[pae77]

Well, I tend to agree with you about the charge rate.

Despite the initial formula provided by the poster I quoted above, he later goes on to recommend charging at very low rates if time allows, even as low as 1/10thC. Personally, I generally prefer charging Li-ion cells low and slow, usually at about .25C unless I'm in a hurry (which is pretty rare).

 

[rlcrisp]

Hi thank you for your answers.

rlcrisp:
But I still think it should be 4.2 V, 8.4 V, etc.
Given the batteries all have the same resistance you should be able to get
a fraction of the whole voltage after every battery. For example with four
batteries it would be 1/4 of 16.8 V, after the second 1/2 of 16.8 V and so on,
like in a wire with a high resistance, where you can get any voltage
in between the positive and the negative voltage depending on where you
measure.

In your notation 1+ and 2- should be the same as 2+ and 3- should be.
so if you start at 0 V and go up to 12.6 V for three batteries,
1+/2- should be 4.2 V, and 2+/3- should be 8.4 V
which gives a difference of 4.2 V for the second cell.
Since you can do this for every cell it is essentially the same you stated
("Is cell #1 at 4.2V? Is cell #4 at 4.2V?").

So can I conclude balanced series charging is generally considered safer?
Is there any drawback in serial charging? What could you do with
empty slots in a charging cradle?

You're right, you only need one wire connection to the balancer for between cells (like you said, 1+ and 2- are at the same potential).

In a perfect world (batteries have same IR and stored energy) we could just charge in series and assume that they all have the same potential. Since that is never the case, we do a balanced charge so that the charger can attempt to correct for IR and storage state (I haven't been able to read about how much imbalance they can correct for - how much extra sourcing or sinking current each balance wire will carry in a typical balancer).

Since in series the same current flows through each cell (that is essentially the definition of series - all the current flowing through one series element has to flow through the other series elements. Parallel is the dual definition - voltage across all parallel elements has to be the same which is why that is much scarier charging method. When you connect two cells in parallel with different voltages, the only thing limiting an infinite amount of current flowing from one to the other is the two internal resistances!), it only makes sense to charge at the highest rate which the weakest cell can accept. Technically, balanced charging is not in series since the whole point of the balancing wires is to push extra current through the lower voltage cells and less through the higher voltage cells so that the cells are carrying different current.

Not knowing how your cradle is configured, as long as you connect the cells in series with balancing cables and complete the circuit with the charging terminals then it won't matter if you have empty slots. If the empty slots are keeping the circuit from being completed then no current will flow.

Nothing I've said should be accepted as good practice or safe, in theory everything I said is correct, I believe. However, LiIon charging is dangerous so I can't be held accountable. Anybody with more experience please correct my mistakes.

 

[Hironiemus]

Hi,
My cradle is not configured in any way because I do not own one, yet.
But today I bought a hobby charger and I will use it with a cradle
from Conrad electronics.

By the way the manual of the charger says charge termination will be at 3.7 V
for LiFe, 4.1 V for Li-Ion and 4.2 V for Li-Po.

One remark about the parallel charging though: I understand how to calculate
the charge rates but then all this would really only apply if all the cells have
the same inner resistance at any charge state, or more ideally the inner resistance is the same and does not depend on the charge state.
This is probably what you were refering as similar and healthy cells.

 

[morelightnow]

I prefer series charging with a balancer. I tried LifePo4 without a balancer and it failed miserably. One battery got charged the other did not. My other batteries are Ni-Mh and they all get charged individually.

 

[PapaLumen]

Hi,
My cradle is not configured in any way because I do not own one, yet.
But today I bought a hobby charger and I will use it with a cradle
from Conrad electronics.

Which cradle are you thinking of from Conrad? I guess they dont do an 18650 specific holder?
Thanks for any info.

 

[VidPro]

one tiny flaw in this theory
(For example 4×2000 mAh batteries would give an maximum charging current
of 8 A).

when you parellel multiple cell items and ONE cell item in the parellel set will no longer charge proper, and starts letting off heat instead, that ONE will be hit with much and much power.
so when things go Bad, this scenario could make them go very bad.
i could even create an example putting a BAD cell that wont "go up in voltage" in a parellel configuration hit it with 1C for the parellel set and demonstrate that because the voltage will not rise easily on the bad one, a large majority of the charge will be hitting that ONE cell.

usually when i do parellels, i keep the Max current of charge less than the 1C for the whole set, because of that. Preferably when making a safe parellel arrangement , when your Blind to the batteries, the max charge would still be 1C for One battery 2A , of the parellel set, not 8A. that way when one battery fails and is just (mostly) heating, it is less likly to reach thermal overload.

and so PAEs point
My advice is this:
Know your packs and their health. If you are pushing for the maximum rated charge current of the packs, be sure to connect only those in parallel that are not way off concerning age, capacity, and general health.
is very important

 

[KiwiMark]

Real quick, that doesn't sound right to me. In theory, the 1A pack would receive a 2C (2A) charge.

No, that doesn't work - if the 1Ah pack was getting 2A then the voltage would rise faster than the 2Ah or 3Ah pack which would make the current slow down - this setup would auto regulate so that from the 6A more current would go into the higher capacity packs. However you cannot say that somehow each pack will magically get exactly the current flow that you have decided it will. The higher capacity packs will get more amps and the lower capacity ones will get less, but it wont be exactly in proportion to their capacity.

Personally I wont charge different batteries at the same time on one single channel charger. I will sometimes charge a matched set like 3 cells from my Mag 3D that have been charged to the same level and then used in the same light in series. Even those wont be identical, but they will be close to each other (same brand, same age, same treatment). I prefer to charge in series with balance leads though - that way each cell gets the same amperage unless the voltage varies and then the balance leads do the job of giving the cells the right current. One thing I do is use 2 hobby chargers, that way with dissimilar cells I can charge one on the first charger and a 2nd cell on the second charger - this lets me charge a bunch of different batteries twice as fast. Generally I am not needing to charge dozens of batteries at the same time anyway.

 

[Hironiemus]

… usually when i do parellels, i keep the Max current of charge less than the 1C for the whole set, because of that. …

I think this is safer but somewhat kills the advantage of parallel charging,
since it would take the same time or even less if you would charge
them one after the other. Except if you can′t change the batteries
because you leave them unattended.

… One thing I do is use 2 hobby chargers, that way with dissimilar cells
I can charge one on the first charger and a 2nd cell on the second charger -
this lets me charge a bunch of different batteries twice as fast.
Generally I am not needing to charge dozens of batteries at the same time anyway.

Yes, the safest way is probably one battery per channel, although the most
expensive. The chargers with several true independent channels I found
would have been at least triple of what I paid now.
(I figure this is not exactly what you do, but you do it one type per channel.)
But as you said it, most of the time you don′t need boatloads of charged
batteries for flashlights.

Which cradle are you thinking of from Conrad? I guess they dont do an 18650 specific holder?
Thanks for any info.

I thought of the Voltcraft cradle (order id. 201004 - 62), which can be
modded to fit 18650s. I was looking for the greatest flexibility though,
therefore I opted for this cradle with the adjustable spring. If I was looking
for a dedicated solution for 18650s I might have ended up using an Pila IBC.
But you also might have a closer look at Xtar. DigiKey makes battery holders
for 1-4 18650s otherwise you can mod a DX/KD charger as a cradle.

 

[45/70]

No, that doesn't work - if the 1Ah pack was getting 2A then the voltage would rise faster than the 2Ah or 3Ah pack which would make the current slow down - this setup would auto regulate so that from the 6A more current would go into the higher capacity packs. However you cannot say that somehow each pack will magically get exactly the current flow that you have decided it will. The higher capacity packs will get more amps and the lower capacity ones will get less, but it wont be exactly in proportion to their capacity.

If you charge three cells in parallel at 6A, each cell will receive 2A, regardless of the cell's sizes or capacities. The voltage of all three cells will be the same at all times. With the cells paralleled, it would be impossible for any individual cell to have a voltage different from the other two. That is the nature of a parallel circuit.

Dave

 

[Selectron]

If you charge three cells in parallel at 6A, each cell will receive 2A, regardless of the cell's sizes or capacities. The voltage of all three cells will be the same at all times. With the cells paralleled, it would be impossible for any individual cell to have a voltage different from the other two. That is the nature of a parallel circuit.

Dave

You're correct in saying that the voltage of the three cells would be the same at all times, but you're mistaken about the current - if you charge three cells in parallel at 6 amps then it doesn't automatically follow that each cell will receive 2 amps. Far from it actually - the current flow through each cell would vary dynamically according to the state of charge and internal resistance of each cell at any given instant - and those factors would of course be constantly changing. About the only thing you can say with certainty is that the current leaving the charger at any given instant would equal the sum of the currents feeding into each of the three cells.

This could get complicated very quickly, but an easy way to grasp that the cell charge currents would not necessarily be equal would be to picture three resistors in parallel, hooked up to a power supply set at, say, 10 volts. Resistor one is one ohm, resistor two is two ohms and resistor three is three ohms. The voltage across the three resistors is identical - 10 volts across each - but resistor one has a current flow of 10 amps, resistor two has a current flow of 5 amps, and resistor three has a current of 3.3 amps, so a current of 18.3 amps flows from the supply but it does not divide equally between the three resistors.

That example was for a constant voltage source feeding a purely resistive network, but in real life, when charging cells, you need to consider that each cell is also both a voltage source and a current source, possessing internal resistance, and capable of both sourcing and sinking current, and in addition to current flow from the power supply, there will be inter-cell currents flowing which will vary dynamically throughout the charging process.

 

[SilverFox]

Hello Selectron,

Li-Ion chemistry is such that when the cells are hooked up in parallel, they balance the state of charge while they are balancing the voltage. This means that the current will be equally divided between the cells. Differences in internal resistance will have a slight effect during the CC phase of the charge, but these differences tend to even out during the CV phase of the charge.

The major issue is with a cell shorting out. To guard against this, you need to take good care of your cells and keep them from over charging and over discharging. In addition you never charge unattended and keep an eye on the cells while charging.

You can further eliminate risk by limiting the number of cells you charge in parallel.

My choice is to, in general, limit the amount of cells I charge in parallel to 3. With 18650 cells, I charge at 6 amps and am very comfortable doing so. If you mix it up so you are charging R-CR123 cells along with an 18650 cell, I adjust the current to 3 times the capacity of the R-CR123 cell. For example: When paralleling an R-CR123 cell with two 18650 cells, I charge at 1.5 amps.

There are concerns with safety while charging Li-Ion cells, but parallel charging doesn't do much to increase the safety issues. Series charging can be more dangerous, and series charging using balancing raises the concern of a balancer malfunction. All in all parallel is a safe way to charge Li-Ion cells.

Tom

 

[45/70]

Yeah, I should have mentioned that this characteristic is specific to charging Li-Ion cells in parallel, as Tom said. For example, with nickel based chemistry cells, the current will not balance out and you will have problems.

Dave

 

[Mr Happy]

Hello Selectron,

Li-Ion chemistry is such that when the cells are hooked up in parallel, they balance the state of charge while they are balancing the voltage. This means that the current will be equally divided between the cells.

I really think VidPro and Selectron are correct about this. The extent to which the current is equally divided between parallel cells depends on the similarity of the cells. If the cells are of the same brand and history, they will balance pretty well. But if they are of different brands, capacities, histories or sizes then you cannot predict very well what will happen.

I would always advise people to tread very carefully in this area.