Will Li-Ion cells in a pack equalize?

I am afraid the cells will even progressivelly differenciate.

Both overcharging or undercharhing cells decreases capacity and life.

Even when guarded by circuit, staying at fully charged, being weaker cell, has bad sideeffects for capacity and life.

e.g. shell life for about half charged cell is said to be at least 2x longer than for fully charged.

VERY...although when you're ready to charge is when it counts. Did you install MATCHED cells?

A more difficult issue is, "how do you get the controller to recognize the new capacity?" And for some controllers with volatile parameter storage, how do you get it to power up at all. Some have suggested that if the pack is not completely dead, you should keep the controller powered up while you exchange the cells, Maybe only need one of the taps, but unless you know which, it's better to power all the taps.

Some controllers have reset pins...although you have to reverse-engineer the pack to figure that out. The result of a reset varies from "omygoditworks" to "doorstop mode".

Interesting contradiction. For MATCHED cells, with different charge levels, your assertion suggests that the charge level (not capacity) will converge over time. But that's critically dependent on your definition of shelf life.

This sounds like a nightmare, so how on earth do they make them in the first place and store them so they are equal when somone buys the things? Brian

rst

Brand new cells are easier to match. Li-Ion degrades with age (approx.

On Thu, 20 Aug 2009 09:30:26 -0700, linnix wrote (in article ):

How is this done?

"approx. 1 year" what? They'll be dead? They take that long to equalize? What happens at 1 year (approx)?

No one mentioned mixing...

They are individually charged before wiring and programming the controller.

What

I said life of a year, not equalization.

My mistake.

Still unclear...

You say Li-Ion cells have life of 1 year?

I ( Poutnik ) have written only following......

You may have noticed I did not provide any definition.

I was just trying to say that speed of secondary chemical reactions, decreasing both capacity and charge level, is function of charge level, supposing the same initial capacity.

when on shelf, Charge level will converge AND at the same time capacity diverge...

when on use, cells with higher initial charge level will be closer to top charge state, comparing to others. Being at this state, the cells get older quicker, losing both level capacity by quicker secondary reactions. (I do not say losing charge level and capacity has the same rate )

Lithium based cells do not have any nivelization mechanism like hydrogen recombination, known at NiXX cell.s

equalize? What

I would say 1 year life means not very well treated cells.

Welcome to the world of battery chemistry.

This has been a known problem for years.

There are a few patents on the subject, and a few web sites on how to do this.

I have worked with Li-Po batteries with a separate controller and power logic to balance banks of batteries.

Not simple or cheap.

good lack and have fun

don

And most laptop battery controllers are poorly (cheaply) designed.

Look at the chips that are available for charging LIon cells. If you have a crappy charger, they probably rolled their own design.

Fortunately, my Lenovo T400 notebook allows me to set custom low and high charge level, related to current battery capacity.

On Thu, 20 Aug 2009 16:43:21 -0700, don wrote (in article ):

URLs?

Please share your experience. After all, this is what USENET is all about! :-)

Very. Matching is a key element in obtaining maximum pack cycle life. Ideally all cells should be from the same batch *and* subjected to a pre-packing charge (or discharge) regime to achieve identical cell open circuit voltage after a rest period.

No.

Think "Nickel-current" and lithium-voltage"

Ni-xx chemistries are charged in a current mode, and overcharged, a direct result of which is that the state of charge (SoC) of series cells equalise.

Li-xx charging is (current-limited) constant voltage - which will not equalise the SoC of series-connected cells. On the contrary, minute differences in cells will progressively amplify as the charge-discharge cycle count builds up. Normal pack protection modules monitor cell voltages and intervene when they exceed a preset threshold. Also as these differences increase, the usable pack capacity reduces simply because the low-voltage shutoff is invoked when the weakest cell reaches its lower limit, and charging should transition from constant-current mode to voltage limit mode when the healthiest cell (by now, with the highest SoC) reaches the preset upper limit - although in some designs this transition is determined by the *average* cell voltage provided no cell exceeds a safety upper limit.

The only good news comes in two parts: (a) parallelled Li-xx cells tend to equalise quite well ("think voltage"); and (b) if you can design an appliance to use a single cell voltage (i.e. ~3.7V nominal) then this issue goes away. Example: cellphones.

They have all the patents and other info.

don

Yes, laptops should have dual batteries. Except for the backlight inverter, not many components need 12v.

Hopefully this is done independent of the operating system. If you can set charging via the OS, consider yourself a hacker target.

Well. maybe. If you need high capacity, you either get it from higher capacity cells or a stack of lower capacity cells. I don't know what limits cell capacity, but there does seem to be a limit. So stacking and bucking gets longer operating time.