Charging Li-ion cells in series

Hi. I'm looking at the LM3622 datasheet

It's a simple Li-ion linear charger. Comes in 4v/8v Li-ion versions. In Li-ion topology, a lot of fuss is made about the cell's final voltage accuracy and how overshooting this will destroy the cell. But what would happen if two cells in series were chargeed with one unequally depleted? Wouldn't the other cell shoot way over its rated voltage, destroying it? It makes no sense that nowhere has this been mentioned. Is this a real concern? Should the two Li-ion cells in series be equalized when charging or charged separately?

TIA.

I've thought about doing this, but couldn't figger out how to make it happen automatically. Cells are in a PDA, so while I have access to the center point, don't have access to much else to control it. What method did you use to switch in the resistors?

Thanks, mike

--
Return address is VALID.
500MHz Tek DSOscilloscope TDS540 $2200
http://nm7u.tripod.com/homepage/te.html
Wanted, 12.1" LCD for Gateway Solo 5300. Samsung LT121SU-121
Bunch of stuff For Sale and Wanted at the link below.
http://www.geocities.com/SiliconValley/Monitor/4710/

Hi Mark,

For a brief but concise summary see Atmel's app note AVR450. It's about a uC based charger but explains charging methods of the different chemistry batteries.

Be careful since Li Ion batteries can explode when mistreated. As far as I know it is not a good idea to charge in a series configuration.

Regards, Joerg

Yes. There are various ways to get round this. For example, when on discharge you detect an imabalance, you connect a small load to the unbalanced cell. Or, do the same on charge, to shunt some of the charge current away.

Thanks everyone for the responses. It's amazing that no datasheet mentions this posibility. Then again, we are supposed to know this, right? :)

Here's an interesting equalization technique, wonder if it will work:

Mark Jones .-------------+--------+--> Vdd

Helios Studios ||--' EM MOSFET | | .--------. ||

You should take it into account. In one of my hobby project I use 2 cells taken from a used laptop pack in series. These are (slow) charged with a simple 78L08. I put a 1K resistor parallel to each cell during charging. Cell are thus not fully charged (4 instead of 4.2V) which gives me a margin of inbalance that will also be corrected up to 4mA by the shunt. There is some lost capacity due to the 4V, but it is minimal. Worked for me, simple and cheap.

Joop

Most commercial multi-cell packs have a pack protection module. Among its functions are detecting under/over voltage (of the series string as a whole), over-current on discharge, and cell_voltage_imbalance.

The module intervenes by effectively turning off a FET in series with the cells, thus preventing any significant current flow in or out. Cell imbalance is one of the causes of pack "failure" in laptops.

With a switch (dual pole/make). The GND is always connected to the cells. The cells center tap is switched in to the center between the resistors. The V+ is switched to the +8V of the 78L08 and the actual circuit.

The 78L08 seems to handle the 8V without noticable current draw. The shunt takes 4mA of course. But my circuit uses about 50mA so I found the added 4 mA neglectable.

Charging only happens when I plug in the transformer. It could all be done a bit nicer on all fronts but it works with affordable of-the-shelf components. My device is an LC meter which I do not use a very often. The li-ion makes sure that it always works because the self discharge is extremely low. A potential reduced capacity of the old cells, the sub optimal charge (4V) and the waist current (4mA) are all unimportant compared to full enough cells whenever I need the meter for a couple of minutes.

As for your PDA, do you really have multiple cells in there? Most mobile phones e.g. just have a single cell for simplicity. The are already enough rumours out there about exploded non-brand cells. Any explosion due to unsafe charge circuits with aging cells put in series is not what manufacturers would like to be linked to. They just use a dc/dc converter if needed at all to avoid the issue. For laptops the high power draw makes it more efficient to use a higher pack voltage. But then you see 3 of 4 cells in series and complex circuitry to safeguard the pack as well as provide SMB communication with the computer.

Joop

Sorry, I got carried away. To come back to your question, why not use a small relay to switch in the top and center cell connection whenever the mains power is available?

Joop