Can you wire LiFePO4 cells in parallel in a battery?

[Little-Acorn]

Forgive me for the ignorant-newbie question, but I get the impression this is being done. To give a simple example:

A guy has a bunch of 5AH LiFePO4 cells, and wants to build a 10AH battery. So he takes two of the cells, and wires them in parallel - that is, wires positive to positive, and negative to negative. Supposedly this gives him a LiFePO4 battery with the same voltage as a single cell, but with 10AH capacity. He then stacks up these parallel pairs in series, to get a battery of the voltage he wants, and winds up with (say) a 48V 10AH battery.

Maybe I'm wrong (would be unusual if I wasn't), but I thought you couldn't do that to low-impedance battery cells like LiFePO4 or NiCd or etc. The problem being, no two cells ever have EXACTLY the same voltage across their terminals. One cell might have 3.71V, while the next cell has 3.73V. If you put these two cells in parallel, you immediately get a current flow as the higher-voltage cell tries to charge the lower-voltage cell. Since the cells are very low impedance, the current can be quite large, and one cell can exhaust the other as it tries to charge it.

Am I full of it?

How, exactly, are LiFePO4 batteries built? I've heard rumors that a 20AH battery can be made by putting a bunch of lower-capacity (say, 5AH) cells together. If so, it would be necessary to wire sets of cells in parallel. If this is being done regularly, then obviously my caution is unwarranted.

Can you wire two (or more) LiFePO4 cells together in parallel, to increase the Ampere-hour capacity of the resulting battery? And then charge and discharge them as though they were a single, high-capacity cell?

 

[Russell]

How, exactly, are LiFePO4 batteries built? I've heard rumors that a 20AH battery can be made by putting a bunch of lower-capacity (say, 5AH) cells together. If so, it would be necessary to wire sets of cells in parallel. If this is being done regularly, then obviously my caution is unwarranted.

To build a 10Ah battery you could use large 10Ah cells or parallel 10 smaller 1Ah and the end result would be 10Ah. There of course would be more labor involved with the second method but if 1Ah cells is all you have and if labor cost isn't a big concern then so be it. All of Mr Ping's batteries for example are made using 5Ah cells for the basic building blocks so for a 36V/10Ah battery there are 12 parallel groups of two cells in series, for a 48V/20ah battery there are 16 parallel groups of 4 cells in series. Making different sized battery packs using the same basic cell does have advantages with handling inventory since only one cell has to be ordered rather than sourcing cells of say 10Ah, 15Ah, 20Ah and so on.

-R

 

[dogman dan]

My pingbattery uses 5 ah cells. 4 cells parallel in each group, then 12 groups. One bad cell in a group can drag down the others by leaving that group discharging the remaining cells at a higher c rate.

 

[torker]

Plus doesn't the bms take care of the diferential between batteries? On charge and discharge?

 

[fechter]

No problem putting LiFePO4 cells in parallel. Just make sure they are close to the same voltage when you connect them.
With NiCd or Nimh, you can't put them in parallel due to current hogging when they approach full charge.

 

[dogman dan]

On typical duct tape lifepo4, such as a ping, the bms treats the paralell group of cells like it's one cell. The paralelled cells will tend to balance themselves out to the same voltage unless one or more cell is dead. But at the end of the charge, the bms will try to get each group of cells fully charged, so none of the paralell groups of cells is undercharged compared to the others.

 

[El_Steak]

One bad cell in a group can drag down the others by leaving that group discharging the remaining cells at a higher c rate.

For diagnostic purposes, lets say I have a group of 4 cells: 3 are healthy and are at 3.5V. The forth one is bad and is at 2V.

If I check the voltage of the whole group in parallel with a multimeter, what reading will I get ?

3.5V (max)
2V (low)
3.125V (average)

 

[patrickza]

One bad cell in a group can drag down the others by leaving that group discharging the remaining cells at a higher c rate.

For diagnostic purposes, lets say I have a group of 4 cells: 3 are healthy and are at 3.5V. The forth one is bad and is at 2V.

If I check the voltage of the whole group in parallel with a multimeter, what reading will I get ?

3.5V (max)
2V (low)
3.125V (average)

As dogman said, the bad cell will drag the rest down. All the cells would have the same voltage, there is no way you could develop a situation like you describe. More likely for eg in a 48v setup you'd have 15 groups at 3.1v, and one group with the bad cell in it much lower, like 2.2v. If it goes too far you'd have to replace the whole group, but if you catch it before any damage is done you might only need to replace one cell.

 

[docnjoj]

I believe that 2 of my groups went "too far". Also the V2.0 Ping that I have has 4.0 a/h cells x5 for each group.
otherDoc

 

[dequinox]

No problem putting LiFePO4 cells in parallel. Just make sure they are close to the same voltage when you connect them.
With NiCd or Nimh, you can't put them in parallel due to current hogging when they approach full charge.

What is "current hogging" ? I read over in the ''series-parallel'' thread that this can be done...it just requires diodes and a bit of careful pack-building. A way to monitor the // strings is necessary too, right?

 

[dennis]

Forgive me for the ignorant-newbie question, but I get the impression this is being done. To give a simple example:

A guy has a bunch of 5AH LiFePO4 cells, and wants to build a 10AH battery. So he takes two of the cells, and wires them in parallel - that is, wires positive to positive, and negative to negative. Supposedly this gives him a LiFePO4 battery with the same voltage as a single cell, but with 10AH capacity. He then stacks up these parallel pairs in series, to get a battery of the voltage he wants, and winds up with (say) a 48V 10AH battery.

Maybe I'm wrong (would be unusual if I wasn't), but I thought you couldn't do that to low-impedance battery cells like LiFePO4 or NiCd or etc. The problem being, no two cells ever have EXACTLY the same voltage across their terminals. One cell might have 3.71V, while the next cell has 3.73V. If you put these two cells in parallel, you immediately get a current flow as the higher-voltage cell tries to charge the lower-voltage cell. Since the cells are very low impedance, the current can be quite large, and one cell can exhaust the other as it tries to charge it.

Am I full of it?

How, exactly, are LiFePO4 batteries built? I've heard rumors that a 20AH battery can be made by putting a bunch of lower-capacity (say, 5AH) cells together. If so, it would be necessary to wire sets of cells in parallel. If this is being done regularly, then obviously my caution is unwarranted.

Can you wire two (or more) LiFePO4 cells together in parallel, to increase the Ampere-hour capacity of the resulting battery? And then charge and discharge them as though they were a single, high-capacity cell?

http://www.batteryuniversity.com/partone-24.htm is a good place to start for a newbie...
the answer is yes you can...
http://www.endless-sphere.com/forums/viewtopic.php?f=2&t=7813 this is an example of 4 x 36V 10Ah LiFEPO4 battery connected in series...each 10Ah battery has to be charged separately. see http://www.batteryuniversity.com/partone-9.htm http://www.batteryuniversity.com/partone-16.htm
I hope this answers your question...

 

[dogman dan]

Back to El Steak's question, I don't know that I can exactly answer it, but here's what I think is true.

Measuring the capacity of each cell group with a load test would be the bestdiagnostic. Measuring voltage of each group after discharging for awhile will show a bad group if it's bad enough, but sometimes that just shows a group that did not get a fully balanced charge and may be ok

What happens with the group with one bad cell is the bad cell reduces the capacity of that group, so that group will fully discharge sooner than the others. The other thing that happens with the low capacity group is that ,say you have a 4p pack, the weak group is in effect a 3p group. That makes the c rate higher on those three cells compared to the 4p groups. Yet another thing that drags down that group with a bad cell. Also that 3p group will have a deeper depth of discharge. I think if you get a bad cell, you have to replace the whole paralell group to fix it right.

 

[timhellsten]

Background:
I have 2 48v 30AH LiFeP04 batteries.
they each have a BMS system.

they are Duct tape batteries from a china supplier.
out of the batteries there is a Red (+) and Black (-) as well as multiple white small wires all of these wires lead to a beige box the Battery Management System.

coming out of this BMS is a Red (+) and Black (-) as well as a hooded male connector for a plug so it can be charged.

My question:
I would like to connect these batteries in parallel so that I have a 48v 60AH battery...

so my questions...

A. do I just connect Positive to Positive and Negative to Negative and then connect those to the input of the device they power?

if I do that when I charge them do I connect both plugs to the house power to charge each battery at the same time... ?

how do I charge both batteries.... Thanks

 

[Megsy73]

Forgive me for the ignorant-newbie question, but I get the impression this is being done. To give a simple example:

A guy has a bunch of 5AH LiFePO4 cells, and wants to build a 10AH battery. So he takes two of the cells, and wires them in parallel - that is, wires positive to positive, and negative to negative. Supposedly this gives him a LiFePO4 battery with the same voltage as a single cell, but with 10AH capacity. He then stacks up these parallel pairs in series, to get a battery of the voltage he wants, and winds up with (say) a 48V 10AH battery.

Maybe I'm wrong (would be unusual if I wasn't), but I thought you couldn't do that to low-impedance battery cells like LiFePO4 or NiCd or etc. The problem being, no two cells ever have EXACTLY the same voltage across their terminals. One cell might have 3.71V, while the next cell has 3.73V. If you put these two cells in parallel, you immediately get a current flow as the higher-voltage cell tries to charge the lower-voltage cell. Since the cells are very low impedance, the current can be quite large, and one cell can exhaust the other as it tries to charge it.

Am I full of it?

How, exactly, are LiFePO4 batteries built? I've heard rumors that a 20AH battery can be made by putting a bunch of lower-capacity (say, 5AH) cells together. If so, it would be necessary to wire sets of cells in parallel. If this is being done regularly, then obviously my caution is unwarranted.

Can you wire two (or more) LiFePO4 cells together in parallel, to increase the Ampere-hour capacity of the resulting battery? And then charge and discharge them as though they were a single, high-capacity cell?

http://www.batteryuniversity.com/partone-24.htm is a good place to start for a newbie...
the answer is yes you can...
http://www.endless-sphere.com/forums/viewtopic.php?f=2&t=7813 this is an example of 4 x 36V 10Ah LiFEPO4 battery connected in series...each 10Ah battery has to be charged separately. see http://www.batteryuniversity.com/partone-9.htm http://www.batteryuniversity.com/partone-16.htm
I hope this answers your question...

Thanks for the info, interesting reading