BIG 18650 battery pack

[Lupino.86]

I would like but a big battery pack using 20x (less or more) 18650 Li-ion 4.2V 2200mA... do you have tricks for me?
I have around 80 battery, then free you fantasy

Firse hope: 4.2V 44000mA
Second hope: 8.4V 22000mA
What suggested me? I use Q5 and P7 LED!


Thanks :wave:

 

[sdnative]

Well, a 20 pack is pretty huge. You really need that much juice??

3 of my 6X flat holders gets you up to 18 at least May be worth a look >> http://www.luminousdiy.com/Light Kits and Parts.htm#holders

~C

 

[Lupino.86]

Be right, I had only at that time 20 batteries while I am now having around 80 of it therefore free your imagination! :twothumbs
I don't know the various combinations of connection among batteries, here is the why of my post. :sigh:

I have this torch (on the bike I mount only the head, not the body) with this LED:
1x CREE Q5= 3W
1x CREE Q5= 3W
6x CREE Q5= 18W
1x CREE Q7= 10W
TOTAL POWER NEEDED= 34W/h


Great idea the holders in you link buy is very expensive! (3x 6battery= 52.35$):thumbsdow
I have thought about insert on my battery pack in a water bottle, but I don't know whether to connect the so many batteries!


Thanks for reply lovecpf

 

[timberwolf]

Depending on the exact chemistry (Cobalt or Manganese) you should minimize the number of cells in series. When you connect cells in series they might start to de-balance while being discharged, Cobalt cells are more sensitive to this effect then Manganese cells. They same is true for charging, if you want to charge the whole pack at once, then you might need balancing circuits for the Cobalt cells, Manganese cells should only be checked every 5 to 10 cycles.

And please keep in mind, 20 Cobalt cells will give a hell of a fire if something goes wrong.

 

[T0RN4D0]

He wants to have 4.2V, so its just a matter of sticking the 20 cells in parallel. You can do that, just make sure you don't short out anything because you are looking at some serious welding amps if you short that thing =)

Also, i hope you have a nice charger to juice that thing back up Even at 10 amps (thats what normal cheap-ish hobby chargers can do) it will take you 6 hours to restore most of the energy.

 

[kramer5150]

I would like but a big battery pack using 20x 18650 Li-ion 4.2V 2200mA... do you have tricks for me?

Firse hope: 4.2V 44000mA
Second hope: 8.4V 22000mA
What suggested me? I use Q5 and P7 LED!


Thanks :wave:

first hope is just 20 cells in parallel. You are going to need a high output CC/CV hobby charger though. It will need to be a good one too, that can handle the current output without over-heating. Overnight charging is something that I try to avoid, and do not recommend it with 18650 cells (even "safe" chemistry types).

Second hope is 10 cells in parallel, in series with 10 more in parallel. Again you are going to need a good CC/CV charger. If you want to charge all the cells together off the same charger you are going to need one that can balance charge the two parallel arrays. You will need to create the necessary wiring, and design the pack to physically accomplish this. I am not sure if you can balance charge 10 cells in parallel, I will let others chime in on this one. I think some CC/CV chargers only allow SINGLE cells in series, when balance charging. Again, don't quote me on that though.

Soldering bus bars onto cells is not the most straightforward thing. You are going to need a good soldering iron (a station would be ideal), one that can maintain its temperature when it contacts the cell. Cells are heatsinks, and will "freeze" your iron tip when the two come in contact, unless your iron can sustain its own temperature enough to allow the solder to flow. I use an 80 watt weller pencil iron, and it has worked fine for my RC hobbies. The generic "pencil" tips that are OEM for most irons and stations will not work very well. I have always found the more blunt tips with flats on them to work best... pic here.

Cell surface preparation is paramount. Iso-propyl alcohol (99% will suffice) and a dremel ball grind attachment to scuff the surface. A seasoned RC heli enthusiast once told me that solder flows where flux goes. So a little flux goes a long way to ensure good solder contact.

Deans battery bars and 12AWG wet noodle are pictured. I have seen guys make their own battery bars using tinned copper stock, cut to shape.

 

[Lupino.86]

Please don't read "20 batteries" as the maximum number of batteries that can use, I possess around 80 batteries!!! :devil:
I would like to build a battery pack with balanced elements that is if I have to use 18 or 24 batteries for me it is not a problem. I don't know if 20 batteries it is a good balanced number...

 

[Lupino.86]

Depending on the exact chemistry (Cobalt or Manganese) you should minimize the number of cells in series. When you connect cells in series they might start to de-balance while being discharged, Cobalt cells are more sensitive to this effect then Manganese cells. They same is true for charging, if you want to charge the whole pack at once, then you might need balancing circuits for the Cobalt cells, Manganese cells should only be checked every 5 to 10 cycles.

And please keep in mind, 20 Cobalt cells will give a hell of a fire if something goes wrong.

Thanks for tricks :kiss:
I use this green cased battery from Sony:
SF US18650GR
K G511MLI1ON

I have build an external battery for my Acer laptop using 2x2S4P and work great charge well 4 hours added in battery mode :twothumbs

 

[Lupino.86]

He wants to have 4.2V, so its just a matter of sticking the 20 cells in parallel. You can do that, just make sure you don't short out anything because you are looking at some serious welding amps if you short that thing =)

Also, i hope you have a nice charger to juice that thing back up Even at 10 amps (thats what normal cheap-ish hobby chargers can do) it will take you 6 hours to restore most of the energy.

Don't worry for soldering, I have expercience

For charger at the moment I have HYPERION EOS5i + balancer HYPERION LBA6 5s... I know the charger capacity and I hope to divide the 18650 in various branches (for example a big package composed in turn by 4 branches composed from 4 batteries in series or in parallel), it's a valid idea???

 

[T0RN4D0]

Huh, i guess you could build 2s10p and balance them just as any 2s pack. My guess is that it only comes down to the balancer/chargers capabilities. If you have a charger that can whip out 20 amps than just parallel as many as you want and you're good to go. If you have something that puts out 200W, but has a 7A limit, then put then in series and put balance wires on it. Again, you need to have some priorities. How important is the voltage in the final application, do you care how long it takes to charge etc...

Youre gonna have to do your homework tho, having 80 cells doesn't do you any good if you don't know how to use them. :thinking:

 

[Lupino.86]

Huh, i guess you could build 2s10p and balance them just as any 2s pack. My guess is that it only comes down to the balancer/chargers capabilities. If you have a charger that can whip out 20 amps than just parallel as many as you want and you're good to go. If you have something that puts out 200W, but has a 7A limit, then put then in series and put balance wires on it. Again, you need to have some priorities. How important is the voltage in the final application, do you care how long it takes to charge etc...

Do you believe that dividing in two your package batteries, would I can charge it with my HYPERION EOS5i + balancer HYPERION LBA6 5s?
I hope two package:
1s5p+1s5p and another 1s5p+1s5p = your 2s10p ...it's a valid idea?
In this way I charge 4 pack (one by one) composed by 5 battery :naughty:

Youre gonna have to do your homework tho, having 80 cells doesn't do you any good if you don't know how to use them. :thinking:

Hehehe, good words! holy words!!! :thumbsup:
But don't worry, safe first of all

 

[Lupino.86]

Anybody answer? :sigh:

 

[timberwolf]

For a 1s5p, thats simply 5 batteries in parallel you won't need a balancer for charging.
You only need a balancer if you charge more than one cell in series.

However with that many cells in parallel you really risk a runaway situation if even one cell goes bad. Without any protection we are talking about a quite amazing chain reaction...

The HYPERION EOS5i has an output capabillity of 5A, for an 5x2200mAh = 11Ah pack it will take approx. 2,5hours to reach 90% capacity thats, about 10hours for you whole pack.

Just a tip: Building such a big battery just for fun isn't a really good idea, especially with a technology like Lithium-Ion Cobalt.

Maybe take a look at this video:
http://www.valence.com/technology/battery_safety/battery_safety_video

And yes this can happen even without anybody shooting at your pack

 

[Lupino.86]

For a 1s5p, thats simply 5 batteries in parallel you won't need a balancer for charging.
You only need a balancer if you charge more than one cell in series.

I didn't know him/it, good to know him, thanks :kiss:
Then, Could I charge the batteries in 2s10p form? No risk?

However with that many cells in parallel you really risk a runaway situation if even one cell goes bad. Without any protection we are talking about a quite amazing chain reaction...

Really? Please help me!!! I would not
Recommend me the best solution and protection (for example, circuits), thanks in advance

The HYPERION EOS5i has an output capabillity of 5A, for an 5x2200mAh = 11Ah pack it will take approx. 2,5hours to reach 90% capacity thats, about 10hours for you whole pack.

Time is not a problem
I will need 2days for recharge 4pack of 1s5p... no problem, no problem

Just a tip: Building such a big battery just for fun isn't a really good idea, especially with a technology like Lithium-Ion Cobalt.

Why? I will use this big pack on mountain bike. I use so many 18650 battery because I want a lot of duration of light so that to be able to pedal for a whole night without worrying to replace me the batteries in the four torches

Maybe take a look at this video:
http://www.valence.com/technology/battery_safety/battery_safety_video

And yes this can happen even without anybody shooting at your pack

OMG

 

[timberwolf]

A decent protection with that many batteries seems undoable for me, you have many cells in parallel and therfore can't measure single cell voltages.

I think a best effort solution would be to monitor each cells temperature and to shut down the whole pack if it rises. But that might allready be to late.

Seriously I personly wouldn't even think about building such a big lithium cobalt battery just because of the fire/explosion risk.

The only lithium technology I would build such a big battery is lithium managanese(LiMn) or even better lithium iron-phosphate (LiFePo).
Lithium cobalt isn't the right technology for such a big battery, that's also the cause why big lithium batteries for cars are coming out that slowly, their solution is either using LiMn or LiFePo(the tesla uses that chemistry) cells I by now don't know of any car that uses LiCo cells at all.

 

[Lupino.86]

A decent protection with that many batteries seems undoable for me, you have many cells in parallel and therfore can't measure single cell voltages.

doesn't any circuit of protection exist for batteries in parallel?
possible that circuits exist for only batteries in series?????

I think a best effort solution would be to monitor each cells temperature and to shut down the whole pack if it rises. But that might allready be to late.

A control of temperature is easy to be implemented, (can recycle a simple thermostat of boiler) if can get by only with that is good! Very good!!!:thumbsup:

Seriously I personly wouldn't even think about building such a big lithium cobalt battery just because of the fire/explosion risk.

Excuse if I ask you him but the risk is sure to 100% or happens to a person on 100? Are all the people that have realized my same idea in precedence dead or cripples? :duh2:
Can I build a package batteries to use a couple of times it or do I risk to die burnt surely while I am pedaling?
Excuse me if I insist, but I think that if the package batteries is done with care and criterion (temperature warning, discharge voltage limit, protection against the water, falls and short circuit) I don't believe that can explode without a valid motive!
Do I think well or no?:thinking:
How do you explain me that use from more than one year a small package batteries formed by 3 groups by 2 batteries every and have not I ever had problems? It form a beautiful package from six batteries that it gives power a 128 5mm LED lamp that I use every evening!
Even of I say that the protection exists against only the bumps and the water!!! :huh: Protection don't exist against the overheating, short circuit and position!!!
Am I fortunate simply??? :naughty:
Or does a valid explanation exist? :thinking:

The only lithium technology I would build such a big battery is lithium managanese(LiMn) or even better lithium iron-phosphate (LiFePo).
Lithium cobalt isn't the right technology for such a big battery, that's also the cause why big lithium batteries for cars are coming out that slowly, their solution is either using LiMn or LiFePo(the tesla uses that chemistry) cells I by now don't know of any car that uses LiCo cells at all.

I have understood well thing you intend to say, you have been very clear, but unfortunately the batteries "LiMn - LiFePo" are more expensive and it doesn't suit me to purchase her because I possess around 80 batteries type "Lithium cobalt".
If can implement the integrated protection in the protected batteries with a circuit I do gladly it! :twothumbs

 

[timberwolf]

Ok I try to keep it simple:
1.) It is practical impossible to measure the voltage of cells which are directly connected in parallel.

2.) The risk it highest when charging the battery-pack.

3.) A thernostat is way to slow and unsensible to protect you against anything, and you need 20 sensors dierctly at the cells not just only one.

I won't go any more into detail, I refuse to support building such a battery pack which practical is a bomb and a danger to you and others.
And I bet those 80 cells aren't even from the same batch, so you have big differences right from the start.
Seriously, forget about it.

 

[SilverFox]

Hello Lupino.86,

Let's take a look at the safety of a large battery pack.

If you connect 20 cells in parallel you will have a 3.7 volt (nominal) 44 amp hour battery. If your load is 34 watts, your battery should give you about 4.8 hours of run time.

There should be no problems with a set up like this...

UNLESS, a cell shorts out.

What do you think would happen if a cell shorted out in the pack?

Tom

 

[timberwolf]

Just a short addition, this shorting of cells is very likely to happen due to variances in their capacity when they are not from the same batch.
The lower capacity cells are those which quite sure will short out and taking all the other paralleled cells with them...

 

[Lupino.86]

Ok I try to keep it simple:
1.) It is practical impossible to measure the voltage of cells which are directly connected in parallel.

Yes, is true. :thumbsup: In parallel mode i have like one big battery (sum of battery amperage, same fix voltage)... I know this

2.) The risk it highest when charging the battery-pack.

Make me think that the only way not to have risks both that to separate one to one the single batteries to charge! Only so I don't have risks in phase of recharge, true? oo:

3.) A thernostat is way to slow and unsensible to protect you against anything, and you need 20 sensors dierctly at the cells not just only one.

WOW, should I connect a thermostat for every battery? It's an idea immediately to be discarded!
However I like the idea of the single thermostat... I believe really that will put it in practice because it is really profit and necessary

I won't go any more into detail, I refuse to support building such a battery pack which practical is a bomb and a danger to you and others.

I understand very well your fears and the dangerousness of these batteries (the video has been very clear! ). I am me the first one that abandons a project if it reveals dangerous, however understand me that I have written in this proper forum to look for in thousands even a solution profit of others impassioned as us. lovecpf

And I bet those 80 cells aren't even from the same batch, so you have big differences right from the start.

To be precise I have around 60 batteries Sony and less than 20 batteries Sony Fukujima STG... the amperage is equal, also the type G5 is the same

Seriously, forget about it.

Is not there really nothing to be done?
Any circuit of regulation or control it doesn't exist?