Charging a Li-ion Pack

[richcj10]

I am using this pack in a product I am designing:
http://www.batteryspace.com/Polymer-Li-Ion-Battery-Module-25.9v-3650mAh-94.5-Wh-7A-rate-with-PCM-UN.aspx
I want to use a IC to charge the pack but I can't find anything that would do it.
Any Ideas?

[blasto9000]

Several major IC manufacturers offer complete battery charging ICs.  (I've used Linear before in a lithium ion application):
http://www.linear.com/products/battery_charger_ic

I see that your pack has seven cells but I don't see individual taps to each cell so that the health of each one can be monitored.  This is a safety issue in my mind, but then, I haven't used those prismatic type cells before (only the cylindrical type).

[richcj10]

The pack has a built in balancer.

[microbug]

One solution would be to get a few of these or these and wire them up in parallel. I think they output 500ma, so if you had 10 you could charge at a rate of 5A. This would only work if all your cells were wired in parallel, which they are not. Try looking here. Found what you are looking for: http://www.batteryspace.com/smart-charger-6.0a-for-25.9v-li-ion/polymer-rechargeable-battery-pack.aspx

[Paul Price]

All single chip Li ion charge management/charging chips are either capable of charging combinations of cells with only 3.7 and/or 7.4V battery combinations, so a single chip solution is not possible without some creative circuit adaptation not intended by IC application design.

I also use in a robotic project a 26V Li-ion battery I have created by series connecting seven 3.7V cells. I have programed a MCU as a both the charging chip and controller for the motor used in this project. The chip uses PWM to control a  MOSFET boost/buck convertor. The MCU does undercharge detection, constant current intermediate charging and finally transitioning into a constant voltage float charging circuit to complete charging. The MCU also indicates charge state progress with a single LED displaying different LED blink patterns to give feedback of the progress of the charging operation and indicate overheating of the batteries, pre-charge conditioning, full current charging, the nearly finished charge state and the fully charged charge state, as well as battery voltage and charger power supply faults. In this way the LI ion pack can be safely and most quickly charged.

Li ion batteries require very careful voltage monitoring and current regulation so as to not damage the battery during charging by excessive charge currents or voltages applied. A single fixed charge voltage/current setting is not acceptable if you don't want to at some time damage the battery.

A brainless(non-MCU) circuit that would work using several IC's would be to use a LM339 comparator to allow detection and adjustment of charge current according to battery voltage. A discarded old 30-32V HP printer supply would be ideal to power this type of charger. The output voltage could be regulated by a LM317 chip which would be itself further controlled by the LM339 sensing the battery voltages and currents at different states of charge, using diodes and possibly another LM339 to perform the logic/switching.  However if you look at the circuit sprawl and size of this contraption, a MCU starts to look like a better solution.

Since MCU chips are so cheap, it is much more practical to use one to control the whole operation reliably while providing some much desired feedback of the state of charge operation.  The only gotcha is you need to know how to program the beast.
. 

A simple way to charge the 26V battery you have is to create a step-up switching convertor circuit to convert a 12V to 18V wall wart power supply or a laptop power supply (18V to 21V at 3 to 4.5 amps) to be a voltage high enough to create the proper constant current at all levels of charging and this means all the while monitoring battery voltage level and current.

Care must be taken to not allow full charge current to be applied a fully discharged battery pack and not start full constant current charge levels until the battery has been charged to a usable voltage level, and  only then allow the maximum constant current charge, and then finally the constant current charge scheme must transition to a constant voltage float charge with current limiting to finish the charge operation with a specified termination current detection of full charge. Only a single MCU is necessary and sufficient  to do all these things easily with just one chip.

But what are you worried about??? The web page link you have inserted shows clearly that the batter pack has an internal battery management unit:


One PCM (10A limited, Manufacture part#  PCM-L08S12-306 (7S) / PCM-LO8S13-109 (0809)  is installed in the battery pack

 to balance charging and protect battery module from
Overcharge (> 29.6 V)
Over discharge ( < 16.1V)
Over drain ( > 40 Amp)
Short circuits
One 7 Amp polyswitch installed to limit discharge current at 7 Amp
Charge/ Discharge terminal: 6.0" long 16 AWG open wire

So all you need to do is create a 28 to 32V power supply to feed the battery pack. You could series connect two 12V unregulated wall warts at 12V with a three amp rating or use a 24V 3A transformer and rectifier/filter cap and use the unregulated output  to feed the battery. If you don't see any smoke after a few hours you're in business!

[muvideo]

One PCM (10A limited, Manufacture part#  PCM-L08S12-306 (7S) / PCM-LO8S13-109 (0809)  is installed in the battery pack

 to balance charging and protect battery module from
Overcharge (> 29.6 V)
Over discharge ( < 16.1V)
Over drain ( > 40 Amp)
Short circuits
One 7 Amp polyswitch installed to limit discharge current at 7 Amp
Charge/ Discharge terminal: 6.0" long 16 AWG open wire

So all you need to do is create a 28 to 32V power supply to feed the battery pack. You could series connect two 12V unregulated wall warts at 12V with a three amp rating or use a 24V 3A transformer and rectifier/filter cap and use the unregulated output  to feed the battery. If you don't see any smoke after a few hours you're in business!

Paul, I agree with the the first part of your post, but reading the last part, as I understand,
beware that the PCM module is not a charger, I would not rely on it, it is only a battery
protection, and will kick in only when battery parameters (overvoltage undervoltage ecc) are off.
The balancing provided is limited to 85mA, so pumping uncontrolled voltage into the battery
will overcharge it and the PCM will disconnect at 4.35V per cell (worse condition), too high:
http://www.batteryspace.com/prod-specs/5673.pdf
The battery still needs a proper charger, current and voltage limited to 29.4V, if the battery
will be charged at need and disconnected from the charger, it can be a dumb CC/CV circuit,
the only protections, as you said, is to limit charging current if the voltage is lower than 3V
per cell, an not charge at all if the voltage is too low (to be decided, but I would not charge
a li-ion with less than 1.5V per cell).
If the battery will remain connected to the charger, it's better to use more intelligent design,
it needs to disconnect the charge when the battery is full, and reconnect only if the voltage
drops under some predefined level.
To the OP, consider also discharge limits, your circuit should disconnect the battery when
the voltage is too low, the PCM will disconnect at 2.5V per cell , that is an "emergency"
level, better that your circuit disconnects at 3V (or higher depending on how much current
you are taking from the battery and it's internal resistance). 
Leaving some residual charge into the battery, not charging to full capacity, and keeping
it cool will lenghten it's service life.

Ciao!

[Neilm]

PCM circuits are there to prevent the battery pack catching fire. Generally they do this by stopping over charge and over discharge. I have seen some of these that "protect" by blowing a special fuse open circuit. Once activated the pack has to be thrown away (or hacked to repair).

Also, overcharging Li-batteries absolutely kills them. I have seen an article that stated if the normal cell (4.2V) had a life of 3000 charge cycles, over charging by half a volt would reduce that to less than 100. I don't know how true that statement was but I would take note of it unless you can afford to keep replacing the battery

Neil

[Paul Price]

The attached circuit will slow charge the Li bat pack without damage by over current or over voltage.
You can sometimes find an old HP printer power supply that outputs 32V at approx. 300ma, otherwise you can always series connect two 12V unregulated wallwarts or add a bridge rectifier to a 24V AC 1amp transformer. You don't even need to filter the supply output with a filter cap.

[muvideo]

The attached circuit will slow charge the Li bat pack without damage by over current or over voltage.
You can sometimes find an old HP printer power supply that outputs 32V at approx. 300ma, otherwise you can always series connect two 12V unregulated wallwarts or add a bridge rectifier to a 24V AC 1amp transformer. You don't even need to filter the supply output with a filter cap.

IMHO better use an LM317 for this job. By the way it can be
also current-limited easily with an npn transistor.

[NiHaoMike]

Also, overcharging Li-batteries absolutely kills them. I have seen an article that stated if the normal cell (4.2V) had a life of 3000 charge cycles, over charging by half a volt would reduce that to less than 100. I don't know how true that statement was but I would take note of it unless you can afford to keep replacing the battery

And conversely, using a lower charging voltage (4-4.15V/cell) enhances lifetime. I recommend doing so (4.1V charger chips are pretty common) unless you really need the extra capacity. For your pack, a bench PSU set to 28.7V (coincidentally very close to the 28.8V peak charging voltage of a 24V lead acid battery) and 2A (or less) would work nicely.