Charging voltage for 3.8v Li-ion battery

[Delta]

Hello Forum!  This is my first post here (although I've been reading for a while, after having been lead here from numerous obscure Goooogle searches) and is a quick and simple query:

When using a constant voltage charging setup, what is the correct voltage to use to charge a 3.8v Li-ion battery?

The battery is a 5.7 Wh / 1500mAH item from a Sumsung mobile phone (link)

The USB socket on the phone is knacked, so I just want to knock up a quick charging circuit on the breadboard so I can take the battery out each night and charge it up.  I'll get a new phone eventually, honest...

I've searched around and found that 4.2v is optimum, does this sound correct?

Cheers.

[JacquesBBB]

You cannot use a constant voltage setup to charge a Li-ion battery !

You need to limit the current. With  your battery to about 1A max.
If you charge it with 4.2V, you will have a much too large current in your battery
at the begining of the charge, and possible  overheating of the battery.

The constant voltage of 4.2V can only be used once the battery is fully
charged to maintain the charge.   

[Delta]

OK, thanks for that.  Just to get some life in to the phone (so I can get my numbers at least - knew I should have stored them on the SIM card...) I will manually adjust the voltage whilst monitoring the charging current and keep it below 1A (probably lower).

Once I get to 4.2v is that the battery fully charged?

[JohnnyBerg]

Charging is fairly complicated. Have a look at the TP4056 datasheet, to see how it is done.

[JacquesBBB]

OK, thanks for that.  Just to get some life in to the phone (so I can get my numbers at least - knew I should have stored them on the SIM card...) I will manually adjust the voltage whilst monitoring the charging current and keep it below 1A (probably lower).

Once I get to 4.2v is that the battery fully charged?

If you want to be on the safe side,  and do a very quick and dirty setup, you can power at 4.2 V, but put a resistor in series
to limit the current to 0.5 A. That is put 10 Ohm in series for example -> 0.42A at most.
The resistor needs to be 2 W resistor .

If you have only  1/4 w resistors, put several in parallel or decrease  the current using Ohms law.

[IanB]

OK, thanks for that.  Just to get some life in to the phone (so I can get my numbers at least - knew I should have stored them on the SIM card...) I will manually adjust the voltage whilst monitoring the charging current and keep it below 1A (probably lower).

Once I get to 4.2v is that the battery fully charged?

You can put a resistor in the charging circuit to limit the maximum charging current. You should limit it to about 750 mA to be safe (0.5C). The battery will approach 4.2 V asymptotically but will never reach it exactly. You should normally stop charging once the charging current has reduced to a small value like 50 mA. The exact stop value is not critical, but don't leave the battery on charge after it has charged. Trickle charging is bad for lithium ion.

[IanB]

If you are in the middle of nowhere and have no other options you can try this: 

http://youtu.be/sEncZHchsA4

But really, don't try this unless it is an absolute emergency.

[JacquesBBB]

A  solution is to use one of these small boards  for about 1$
http://www.ebay.fr/itm/5V-Micro-USB-1A-Lithium-Battery-Charging-Board-Charger-Module-/400527882453?pt=LH_DefaultDomain_0&hash=item5d415278d5

They will charge the battery  at up to 1A current on a USB power supply, and automatically stop when the battery is charged.

[Delta]

Thanks for all the replies.  I just want to knock something together using what I've got lying around to get some life in to the battery just now.  I'll just make a quick adjustable LD1084 (LM317 equivalent) and tweak the voltage whilst monitoring the current in to the battery.

Would a constant current supply, with the max output voltage limited to 4.2v do the trick longer term?

[JohnnyBerg]

Would a constant current supply, with the max output voltage limited to 4.2v do the trick longer term?

It will dramaticly shorten the lifetime of your LiPo. As said before, the charge cycle is fairly complex:



Tolerance is tight!

[IanB]

Would a constant current supply, with the max output voltage limited to 4.2v do the trick longer term?

Yes, that is the recommended CC/CV charging protocol: constant current up to Vmax, then constant voltage until the current reduces to about 5% of Imax.

However, proper chargers or charging chips include extra safeguards such as using a very low current to bring the voltage up if the battery voltage starts out too low, and a latching power off once the battery is charged to avoid long term trickle charging.

[Rick Law]

Thanks for all the replies.  I just want to knock something together using what I've got lying around to get some life in to the battery just now.  I'll just make a quick adjustable LD1084 (LM317 equivalent) and tweak the voltage whilst monitoring the current in to the battery.

Would a constant current supply, with the max output voltage limited to 4.2v do the trick longer term?

I was doing that for a while, yeah, it will do the trick but poorly.

The LM317 load regulation is not that great.  Under load, your 4.2v will be pulled so far down thus current become so low that it takes much longer than healthy (not for the battery but for me, life is not that long, you know...)

So, I manually re-adjust the output to bring in a better current flow while keeping the loaded voltage as high as possible while below 4.2V.  Trouble is, as the battery charges, the decreased load increases the voltage - eventually load decreases to a point where the voltage exceeds 4.2V so it needs constant readjusting to keep it at around 4.2V.

Getting tired of such constant readjustment, I hire an Italian guy call Arduino to do the job.  "Mr. A" brought along his two assistant called ADC and DAC.  They are brothers, I think.  Easy thing for Mr. A's team.    With the ADC brothers, one is blind and just adjusts the voltage, the other is a reader who just monitor the voltage and current then tell his brother to step up the voltage or step down.  You see, ADC was born an ADC, it can read voltage and current like a charm.  But DAC was born a PWM, cross dressed to become a DAC...  Enough gossip about DAC's private life.  I will just add that DAC is quite a leader.  Where the DAC leads, the LM317 will follow.

The contraption works nicely but far too complex for a simple task.

So, I decided to fire them, hired a Chinese guy called TP4096 TP4056.  For 1/10 of the cost, the job gets done - simple, no blinking lights, no LCD screen...  It just gets done.    "Mr. A" now has a different job.

EDIT- may be silly to edit a dead thread but so as not to confuse myself or others... it is TP4056, not TP4096.  I confused myself when I had to search what I wrote...

[BravoV]

However, proper chargers or charging chips include extra safeguards such as using a very low current to bring the voltage up if the battery voltage starts out too low, and a latching power off once the battery is charged to avoid long term trickle charging.

... and apart from the timer, better one also includes temperature sensor sensing the battery's temperature as another layer of safeguard.

[JacquesBBB]

So, I decided to fire them, hired a Chinese guy called TP4096.  For 1/10 of the cost, the job gets done - simple, no blinking lights, no LCD screen...  It just gets done.    "Mr. A" now has a different job.

These are the boards I was mentioning. They do the job at a very cheap price so I can concentrate on my projects.

If you want to use in a separate was a li-ion cell, you can also use a equivalent board that monitors as well  the
discharge of the battery. Taking care that your stuff does not draw too much current and
stopping when you reach a too low voltage value to  preserve the battery.
this will be a little bit more expensive,  going from 1.2 $ to up to nearly 2$ as here
http://www.ebay.fr/itm/1pc-Micro-USB-1A-Battery-Charging-Board-Charger-Module-Protection-board-Arduino-/331474895754?pt=LH_DefaultDomain_3&hash=item4d2d71838a

see picture below.

I have systematically attached such boards to my  li-ion cells, so I just plug them in a USB power supply when needed.
I found that there was really no point saving on these very cheap boards. On the other hand, all my batteries come
from  discarded laptop computers. Very often computer batteries are  discarded but  only one cell may be in bad shape.
All the remaining ones are still  usable.

[JohnnyBerg]

Getting tired of such constant readjustment, I hire an Italian guy call Arduino to do the job.  "Mr. A" brought along his two assistant called ADC and DAC.  They are brothers, I think.  Easy thing for Mr. A's team.    With the ADC brothers, one is blind and just adjusts the voltage, the other is a reader who just monitor the voltage and current then tell his brother to step up the voltage or step down.  You see, ADC was born an ADC, it can read voltage and current like a charm.  But DAC was born a PWM, cross dressed to become a DAC...  Enough gossip about DAC's private life.  I will just add that DAC is quite a leader.

The 10 bit ADC from the Arduino is by far not accurate enough to measure the voltages on the LiPo. Powered from 5V, the resolution is only about 5mV. And I am not even talking about accuracy.
Not an easy job for your Mr. A 
Actually a no go 

[ccayser]

I might be wrong, but newer Samsung models have 3.8V Li-Ion battery, not 3.7V battery. The full charge voltage for 3.7V battery is usually 4.2 and if I remember correcty then for Samsung 3.8V batteries it was 4.3 or 4.35V. So the general Ebay charger won't work as well and also just charging it with external power supply won't also charge it up to it's full capacity. I also would not charge it with 1A current, as it is recommended not to charge above 1C (in this case 150mA) when charger don't monitor battery temperature.

[JacquesBBB]

I doubt Samsung has invented an new chemistry for his batteries. So the Battery should be exactly the same as the 3.7V.

Now deciding at which level you charge them is just a matter of choice, and  0.1V does not matter.
Maybe they decided to charge them a little bit more in order to extend the autonomy,  even if the life time decreases.

In any case, if you use a standard charger, you will charge the battery to 4.2 V which will be nearly filled to max, even if your
phone will be able to top it then to 4.35V.

[JohnnyBerg]

Now deciding at which level you charge them is just a matter of choice, and  0.1V does not matter.

Hmm .. I think it does matter. Microchip also thinks it matters, they have 4 different version of the MCP73831.
There is even a different chip for 4,35 and 4,40V, so only a difference of 50mV!

[JacquesBBB]

Hmm .. I think it does matter.

When I say that it does not matter, I do not say that there is no difference. If You fill a bottle, you can fill it more or less.
Putting 4.35V is more than 4.2V.

What I say,  (and this is in what sense it does not matter) if you fill it to 4.2V, you will already have plenty
of power to run your phone, and more important, you will not hurt your phone in any manner.

Exactly as if you have  not completely topped your bottle.

[ccayser]

I doubt Samsung has invented an new chemistry for his batteries. So the Battery should be exactly the same as the 3.7V.

Now deciding at which level you charge them is just a matter of choice, and  0.1V does not matter.
Maybe they decided to charge them a little bit more in order to extend the autonomy,  even if the life time decreases.

In any case, if you use a standard charger, you will charge the battery to 4.2 V which will be nearly filled to max, even if your
phone will be able to top it then to 4.35V.

Unfortunately I disagree. When dealing with Li-Ion batteries the charge voltage matters and it matters a lot. Li-Ions are charged by their voltage and it must be in 1% of their specification. For example general 18650 batteries full charge is 4.2V. So the charger profile is set so it starts to charge at constant current(also important value, not to exceed the battery limits or rule of thumbs). When healthy battery reaches 4.2V it is usually about 80% full and the charger continues to charge in constant voltage mode till the charging current is about 10-20% of initial charging current(usually it is presettable).
And yes, there are higher than 4.2V maximum voltage batteries on the market: http://industrial.panasonic.com/www-data/pdf2/ACI4000/ACI4000CE74.pdf

When I say that it does not matter, I do not say that there is no difference. If You fill a bottle, you can fill it more or less.
Putting 4.35V is more than 4.2V.

What I say,  (and this is in what sense it does not matter) if you fill it to 4.2V, you will already have plenty
of power to run your phone, and more important, you will not hurt your phone in any manner.

Exactly as if you have  not completely topped your bottle.

Yes, I agree.

[Delta]

Blimey!  What a lot of discussion this has created!  It's been very edumacational for me anyway.

I'm currently charging it with a constant current at 300mA, and the voltage across it is coming up nicely.  When it gets to 4.2v I'll switch to constant voltage and leave it 20 mins or so until the current has dropped.

I've ordered a couple of those TP4056 modules, hopefully they'll be here soon.  I can't monitor temperature of the battery as due to my bodged "holder and contact assembly" I can't get to the NTC terminal (although at least I now know why I was reading 2.5k between that terminal and negative), so I don't fancy charging at 1000mA - what would you recommend? 500mA?  It looks like changing this will involve unsoldering an SMD resistor and trying to solder on to the pads... :-/

Thanks for all the advice, hopefully I can put this old Nokia away again...

[ccayser]

Safest way would be charging at 1C rate. If your battery is 1500mAh, it should be safe to charge at 1.5A. But I would recommend to charge it within the range to half an amp(0.1C to 0.3C). You can monitor the temp with your hand, if you currently charge it just be sure that it is not going very hot. For me it is safe if it warms up a little, but anything beyond that is a no go for me. And also do not leave it unattended and near the stuff that catches fire easily.

[amyk]

Best way to determine the correct maximum voltage is to put it in the phone and let it charge to completion. Lithium-ion cells have been marketed as anywhere from 3.6v to 3.8v depending on chemistry but 4.2V is going to be safe for any of those as they're the cobalt/manganese type.

The notable exception is lithium phosphate, which will have 3.6v maximum voltage and often marketed as "3v" nominal.

[Rick Law]

Getting tired of such constant readjustment, I hire an Italian guy call Arduino to do the job.  "Mr. A" brought along his two assistant called ADC and DAC.  They are brothers, I think.  Easy thing for Mr. A's team.    With the ADC brothers, one is blind and just adjusts the voltage, the other is a reader who just monitor the voltage and current then tell his brother to step up the voltage or step down.  You see, ADC was born an ADC, it can read voltage and current like a charm.  But DAC was born a PWM, cross dressed to become a DAC...  Enough gossip about DAC's private life.  I will just add that DAC is quite a leader.

The 10 bit ADC from the Arduino is by far not accurate enough to measure the voltages on the LiPo. Powered from 5V, the resolution is only about 5mV. And I am not even talking about accuracy.
Not an easy job for your Mr. A 
Actually a no go 

Yeah, that is exactly how "Mr. A" got me to grow his team and build his empire.  I ended up hiring ADS1115 (16bit ADC) and MCP4725 (12bit DAC)...  I got those to learn, so, I don't regret buying them, but the whole contraption just got too complicated for day-to-day use.







That said, this was planned as a learning exercise to begin with.  Building this contraption and getting it to work was a very worth while experience.

The off-spring of this contraption became an add on digital control CC and CV for a buck-module with XL4051E1 (LM2596 work alike) and served as my power supply for a while.  That is why the ADS1115/MCP4725 are on a separate proto-board.  I can remove my AdaFruit modules in one short and plug it into my baseboard housing the XL4051 module.

That XL4051 digital cc/cv contraption too is now replaced --- by a MingHe B3603.  The B3603 costs less than the AdaFruit ADS1115.

The XL4051 module is is now a simple CC/CV module with my add-on analog CC which I occasionally used to feed the TP4096.  Building the analog current limit is from experienced learned building this string of contraptions.

So, I decided to fire them, hired a Chinese guy called TP4096.  For 1/10 of the cost, the job gets done - simple, no blinking lights, no LCD screen...  It just gets done.    "Mr. A" now has a different job.

These are the boards I was mentioning. They do the job at a very cheap price so I can concentrate on my projects.
...

Yeah, I use a similar board.  Same old story, the home team got canned and some cheap foreign labor got the job....

Note:  There are clones out there with 4056ES.  They are not as precise as the TP4096.  I blew the TP4056.  My replacement is a 4056es where 1A=980mA, and 100mA(stop current)=70mA.

EDIT - replaced all 4096 with 4056 above, I called it TP4096 by mistake.

[JohnnyBerg]

So, I decided to fire them, hired a Chinese guy called TP4096.  For 1/10 of the cost, the job gets done - simple, no blinking lights, no LCD screen...  It just gets done.    "Mr. A" now has a different job.

These are the boards I was mentioning. They do the job at a very cheap price so I can concentrate on my projects.
...

Yeah, I use a similar board.  Same old story, the home team got canned and some cheap foreign labor got the job....

A perfect demonstration why I made my own smd prototype board. 

It has Mr. A on it,  Mr. TP4056 can join the party, and there is even room for Mr. ADS1115 and MCP4725

I made a "behind the scenes" thread

There is even room for a load seperator, otherwise the battery never get's filled