what sort of thing do I have to worry about or watch out for, building a lithium-ion battery pack?

The basis of my idea revolves around a design similar to this:
http://www.instructables.com/ex/i/16...3001143E7E506/

Her design calls for two nimh batteries, but I'm thinking adding an additional chip, which supports charging a Li-ion battery using USB ... basically adding a built-in recharge capability. I would like to use two lithium cells in parallel for a higher density... this way, I can connect the device to a usb host to recharge the device from the hose, or connect a usb gadget to the device to rechage the gadget from the device ... not sure how to handle the switching between charge and discharge modes

The chips I'm looking at (MAX1555 or MAX1811) will charge and maintain a single lithium ion or lithium polymer cell ... but it makes no mention of cells in parallel ... my thinking is, two cells in parallel should just look like one higher capacity cell, I mean, if I might have a single cell that is rated at 1000 mAh, but I could also use a single cell rated at 2000 mAh ... the datasheet doesn't mention any limits as to the capacity of the cell, so, to the chip, it'll look like I just have a big 'ol cell connected if I take two 2000 mAh cells and parallel them. of course, it would take longer to recharge all that capacity, especially with current limited to only 100mA (or 500ma for the 1811)

does this make any sense?

__________________
If you don't have a planet, what good are gold bars?

want to contact me directly? gmail gordonthree
check out my project website: http://projects.dimension-x.net
Favorite numbers:
09 F9 11 02 9D 74 E3 5B D8 41 56 C5 63 56 88 C0

 

Cells in parallel tend to be a bit 'dodgy', unless they are EXACTLY (100.00000%) the same then one will discharge into the other until they are equalised - this obviously means you're not getting the full efficiency from them. Likewise, if you charge them in parallel, one series of cells will tend to hog the current.

So be careful!.

__________________
PIC programmer software, and PIC Tutorials at:
http://www.winpicprog.co.uk

 

Can't you isolate the cells from each other using a diode for each one? Thus keeping them from charging/discharging into each other?

 

The Maxim chips don't look like they have any timeout/charge termination mechanisms. That is sorta distrubing, but when talking about 2AH worth of batteries, I think it's pretty dangerous. The standard termination is to disconnect the batteries when charge current drops to 1/10C - so in the case of 2AH worth of cells, 200mA. This does put a lower limit on charge current (or an upper limit on the capacity of battery you can attach to a charger).

In any case, from the safety perspective, you really ought to have a battery protection circuit.
http://focus.ti.com/docs/prod/folder...ucc3952-1.html
Essentially it'll pull the battery out of the circuit if it starts getting overcharged, overdischarged or shorted - which is definitely preferable to having the lipoly cell balloon on you - or the Li Ion from venting or exploding.


As for wiring batteries together, series configurations are definitely a pain because a charge balancing circuit is required for longer strings (they usually start appearing when there are 3 or more cells in series), but I've never heard of rechargeable Lithium cells being difficult to use in parallel.

I don't think a parallel configuration discharging would have any issues -the low impedance, low charging "losses", and the large change in voltage during the discharge should make the batteries track pretty well.

I might believe that charging would have issues, but I'm still doubtful.

 

I would think you want them to discharge into each other so that they might balance out voltages when disconnected better. Either way, it doesn't matter because you will load the batteries differently anyways if they aren't balanced, and that has nothing to do with batteries charging each other.

Plus batteries can put out a @#(&@ load of current. You would need a massive diode on each cell to handle the heat. Plus you have that annoying voltage drop.

 

The big problem is the the negative temp coefficient of a silicon diode means one battery will tend to charge or drain faster than the other. This is very very bad.

I believe many laptop packs parallel 2 cells together to get the required capacity. As noted, this can be a bit risky. A key here is to keep them in the same temperature when charging and discharging. For example, if one cell is buried deeper in the device it can get warmer than one on the outside for a given discharge rate. The warmer battery will have a slightly higher voltage thus will provide more of the current, leading to an imbalance condition.

__________________
I thought what I'd do was I'd pretend I was one of those deaf-mutes.

 

i can't see why you couldn't parallel them, did you get them from phones? because most modern phones have built in circuitry to prevent overcharging and short circuiting, i have a li-ion battery, when fully charged i could quite easily short the terminals and nothing will happen because the circuitry will shut off the current, wouldn't recomend unneccesary shorting though.

lithium is a very danegrous chemical, make sure your batteries are well sealed, because a little drop of water and you've got problems.

 

What if you only charge them in parallel, but you actually use them in circuits that place them in series?

Even if one cell charges faster, that shouldn't be an issue, because it would simply fill up faster. Once full, it could be made to pull less current from the supply while the slower battery continues to charge.

 

The problem is that as soon as you connect the cells in series, you need to add in some protection/balancing circuitry. The battery needs to be disconnected as soon as a single cell hits the minimum voltage (~2.5V), otherwise it'll be damaged in short order.

By this point there's a microcontroller or something monitoring the batteries, and the next logical step is to put a transistor in series with each cell which is used to shunt current around the cell while charging.