Balancing large amounts of lithium batteries

[justanothercanuck]

I was thinking of replacing my e-bike battery with a lithium battery pack.  The original battery pack is a 4S 48v/20ah SLA pack, going on about 2 years old.  It still works, but I'll probably have to replace it at some point, as I'm lazy during the winters and have let the battery run itself under the 30v mark.

So back to the lithium...  I was looking at the Sony 18650 VTC5's, 2600mah capacity, rated for 30a loads.  The idea was to build two 24v 6S10P packs (2x 25.2v/26ah).  The reason I chose to do it in two packs was that the balancer/chargers I was looking at online seem to have a maximum of 6 cells per pack.  Not a big deal really, when it comes time to charge i can disconnect the two halves and charge them that way.

I get the idea behind why you should balance battery packs, to keep any individual cell from going under or over voltage.  In a series circuit each battery has a balancer wire, but if you're using batteries in parallel, you wind up with several batteries connected to a balancer wire.  There was a youtube video I was watching where someone harvested the 18650s from laptop batteries (which I'm going to skip -- I'd rather have fresh new batteries to start with) and his battery pack was 6S12P (crazy schematic mockup included below).  In a way, doesn't that defeat the point?

At what point does balancing seem pointless?

[BravoV]

In parallel (row level) at your circuit there, how the balancing process work at discharge or charging since they're connected directly at the same cable ?

Any weak cells will be deteriorated faster since it will receive uncontrolled current charge from strong cells, while strong cells also will experience the torture too as it will dump it's energy to charge the weaker cells, again uncontrolled.

Not an expert, just wondering.

[justanothercanuck]

In parallel (row level) at your circuit there, how the balancing process work at discharge or charging since they're connected directly at the same cable ?

Technically, it wouldn't.  At least, that's my understanding.  I guess that's another problem with building such a battery pack.

Maybe this isn't such a good idea after all...  but how would any other electrical vehicle approach the same problem?  I mean, Tesla must be doing something right(?) to get this stuff to work.

[BravoV]

Maybe this isn't such a good idea after all...  but how would any other electrical vehicle approach the same problem?  I mean, Tesla must be doing something right(?) to get this stuff to work.

Most of the usage of multiple li-ion cells are using them in series since its easier to maintain the balance, and then use the super efficient synchronous buck converter to step down the voltage.

[NiHaoMike]

In parallel, they would self balance. In series, you would actually need to balance them. LT makes a chip specifically for that application.

[kaptain_zero]

Considering that you are using this for a high power draw situation, I would suggest avoiding the LiIon 18650 type cells and go with LiFePO4 chemistry. It's great for high current and safety. In the past I think you could have harvested A123 cells from power tool packs (Dewalt brand rings a bell), but of course they have all switched to LiPo packs as they can be made to fit compact and oddly shaped spaces.

The problem with large "packs" of 18650 LiIon cells is, that if a single cell goes bad and enters a thermal runaway, it will likely take the rest of the pack with it. It does not take a large amount of imbalance to "get in trouble" with LiIon. A "Vent with Flame" episode can be nasty. LiFePO4 is more forgiving of abuse than LiIon and LiPo, something you would appreciate if it should ever happen.

Parallel charging cells at a slow rate using CC/CV would help minimize the need to balance charge, but discharging them in parallel, as mentioned earlier in this thread, would still allow a weak cell to to cause an imbalance. Considering how much variance there is in the capacities of 18650 cells from the same lot, I would not risk it.

This sort of thing "parallel discharging" etc. is being done today, but primarily by remote control aircraft hobbyists, who are acutely aware of the risks, and who take great care in maintaining their battery packs, including measuring imbalances/discrepancies in each cell of each pack of cells. Well.... that or watch their planes go up in smoke. 

I've looked at some of the battery packs from Golden Motors Canada, the prices are not for the faint of heart, but personally, my hind end (not to mention the garage or house) is worth more than the few dollars saved by trying to go cheap.  I only mention Golden Motors as I have perused their website. I have no idea of the quality of their products, and I'm sure there are plenty more suppliers now. I only remember them from a couple years ago when I was toying with the idea of an ebike for getting to work.

Regards

Christian

[justanothercanuck]

VTC5's aren't exactly cheap, they're like $10/battery, unless I can get some sort of distributor discount, especially considering I'd need to buy 120 of them, and maybe an extra 10 for giggles. 

I'll certainly look into what LiFePO has to offer though.

[kaptain_zero]

Just for giggles:  http://killacycleracing.com/about-us/killacycle/

I WANT one.... 

[nctnico]

The best thing to do is match the cells (measure internal resistance) which are parallelled. However if the discharge current isn't close to the cells' maximum then the internal resistance doesn't cause excessive inbalance between the cells. Starting with new cells is a good idea though.

[peter.mitchell]

a tip for paralleling cells, if you can, use the contact on the opposite end of the string of cells, eg,
this will discharge the parallel cells more evenly, rather than stressing the cell at the end of the string.