Lithium Ion (Pouch) battery - Over/Under voltage charge protection

[DeanA]

I am considering a 3.7V 1000mAh Lithium Ion Pouch battery for a project.  Same as this:
https://core-electronics.com.au/polymer-lithium-ion-battery-1000mah-38458.html
It says it has "PROTECTION CIRCUIT - Built into the battery is an over-voltage, over-current, and minimum voltage protection circuit"
So I guess it is fine to charge from say a 5V source with a current limiting resistor and the in-built detection circuit will cut off charge when the battery reaches the overvoltage detection point?  Or do I need an external circuit to detect overvoltage and turn off charge to the battery?

[Mr.B]

The protection circuit is just that – a protection circuit.
You should not be using it as a charge management circuit.
You would be well advised to use a proper charge management chip for LiPo batteries.
They are readily available and cheap.

Edit to add: And always purchase your LiPo batteries with the built in protection circuit as well as adding your own proper charge management.

[uer166]

A protection circuit is not a charger, and it is not fine to use it as such.

Those BMS ICs trigger after over voltage already happened to a certain extent, so use a proper charging system for the cell.

[RoGeorge]

So I guess it is fine to charge from say a 5V source with a current limiting resistor and the in-built detection circuit will cut off charge when the battery reaches the overvoltage detection point?  Or do I need an external circuit to detect overvoltage and turn off charge to the battery?

Not fine.  You still need a dedicated charger circuit (search for Li-Ion charger IC, there are many chips that can do the charging just right).  If you have more than one cell, then you'll need a balancer, or BMS (Battery Management System).

Those protections embedded in the battery are against accidents, like a malfunctioning charger, or a malfunctioning load.  The threshold for the protection are set outside the normal (repetitive) limits supported by the battery.

If you count on the protection board to terminate the charge/discharge for you, then the battery will be abused at the end of each charge/discharge cycle, and will wear out much faster, or even worst, eventually the abused battery might burst into flames and burn down the entire building. 

[KerimF]

For a small rechargeable battery, I may try to also find out how many diodes (as 1N4007) connected in series can protect it (by connecting them in parallel to it) from overvoltage.

[RoGeorge]

Diodes won't work.

[KerimF]

Diodes won't work.

I believe you have surely good reasons to say it.

But, on my side, I got a ready-made charger from USA (7.2 V nominal). The batteries survived for 3 months. My first thought was that the batteries were somehow faulty already when I purchased them. So, I replaced them with new ones. The same happened. Fortunately, the third set of batteries have survived for about 5 years so far after I protected them with diodes (working as a power zener in parallel whose breakdown voltage decreases when the temperature increases).
   

[RoGeorge]

I meant, the diodes won't work the same as (or as good as) a charger IC.

The datasheet for Li-ion tells to charge at constant current first, then at top voltage (usually 4.2V/cell) to continue the charging until the charging current drops under a certain limit.  For example, this type of cells has to be floated at 4.2V until the charging current drops under 50mA, then the charge must be terminated:  page 4/9 https://www.dnkpower.com/wp-content/uploads/2022/07/dnkpower-LG_daHB2-18650-datasheet.pdf.  Keeping the 4.2V applied even longer is not good for the battery.

Applying more than 4.2V/cell will damage the battery.  The 4.2V limit is very tight, a few tens of millivolts extra is already too much.  This one, for example https://www.ti.com/product/BQ24091 has 1% voltage accuracy.  Of course, the charging can be stopped at a lower than 4.2V, just to be safe, but then the battery will store less energy.

Diodes are not all the same, the voltage drop varies from one diode to another, varies with the current through the diode, and varies with the external temperature.  The series diodes might have been an acceptable workaround for that battery-killer charger, but that is more like a hack, not something accurate and easy to reproduce.

In contrast with the diodes, a dedicated charger IC can properly terminate the charging, can read the battery thermistor, has a safety timer that disconnects in case the charging takes too long, and so on.

[SiliconWizard]

Yes, please use a dedicated charging IC for anything Li-ion. Handling the charging properly is described in the info RoGeorge gave, and still it's just scratching the surface. A Li-ion battery is not as easy to handleand forgiving as a NiCd or NiMh battery - if you were used to these, forget about that.

Purely from a rhetorical standpoint, you can, more or less, charge a Li-ion battery with just a voltage source and a series resistor.
Still, you need to make sure your voltage source is 4.2V within about 50mV - which will already require a more elaborate voltage source than you'd probably want to care about if you're thinking of cheaping out on a proper charging IC.
Alternatively, for safety reasons, you can of course settle for a lower max voltage. For instance, 4V regulated within +/-5% (which should not be too hard to do with a cheap linear reg) should do. You can go for 3.9V for a bit more margin.

The issues though with this approach are the following:
- Since you'll have to limit the max voltage to lower than 4.2V, you won't get a full charge.
- Since the charging current will vary with the cell voltage, if the cell is fully discharged, it will start with the max. current and the current will gradually decrease. So to keep the max. charging current within specs for the cell, you'll have to select the series resistor appropriately (typically, a fully discharged Li-ion cell should be around 3V, so if you use a 4V source, your starting current with be (4-3)/R = 1/R. The downside is that charging the battery will take much longer this way (due to the current dropping as the cell charges up). Also, starting with the max current is not a good idea if the cell is in deep discharge state. Dedicated ICs do check the cell voltage before starting to charge, and if in deep discharge, they'll either decide to fail (if the cell voltage is too low to be safe), or "trickle charge" it with a low constant current until it raises above a certain threshold (about 3V), past which they'll switch to a higher current. With a simple series resistor approach, attempting to recharge a deeply discharged cell may result in something not pretty.
- Dedicated ICs do detect an end of charge condition and stop charging at this point. A simple series resistor approach won't, and will keep "trickle charging" the cell until you unplug it from the charger. That's usually guaranteed to not end up very well after a while (many batteries will tend to swell).

So, while this could work if you really have nothing else at your disposal and must charge a Li-ion battery in some emergency situation, under supervision, I would certainly not recommend doing that routinely and leaving it unsupervised.

[tszaboo]

Diodes won't work.

I believe you have surely good reasons to say it.

But, on my side, I got a ready-made charger from USA (7.2 V nominal). The batteries survived for 3 months. My first thought was that the batteries were somehow faulty already when I purchased them. So, I replaced them with new ones. The same happened. Fortunately, the third set of batteries have survived for about 5 years so far after I protected them with diodes (working as a power zener in parallel whose breakdown voltage decreases when the temperature increases).
 

Because an UL test report of any Lithium battery will tell you the safe charging conditions. With current, voltage, cut-off and a whole bunch of temperature requirements. What you describe wouldn't. In fact what you made is dangerous, and you are liable for criminals charges in case you placed it into a product. Batteries are not toys.

[KerimF]

Because an UL test report of any Lithium battery will tell you the safe charging conditions. With current, voltage, cut-off and a whole bunch of temperature requirements. What you describe wouldn't. In fact what you made is dangerous, and you are liable for criminals charges in case you placed it into a product. Batteries are not toys.

Very true, thank you.
I used to apply practical solutions for my own needs only (in this case, for my own batteries).