Battery charger for wider voltage range

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I have bought some Li-ion batteries from China which are not what they purport to be.

Outside of the capacity issue (they are less than 1/3 of claimed), they have a very wide voltage range and can discharge to way beyond safe limits.

I have constructed a battery charger for a nominal 12V, in practice from about 10V to 13V.

However these batteries can drop to as low as 6.5V (I discovered after one year of non use).

Currently my battery charger has 4 positions, selecting the charge current, 100mA, 250mA, 500mA and 1000mA. However if the battery has drained down to say 6.5V and you select 1000mA, the charger will need to dissipate more than 10W of heat.

**** This question is about ideas to limit the power dissipated as heat. ****

The first idea that comes to mind is to try and limit the current when the voltage is too low.

Another idea would be to have a variable voltage supply to the charger to follow the battery charge level, so for example if the battery is at 6.5V then the supply to the charger should be 6.5V + internal losses = 11V. If the battery is at 12.5V then the supply should be 12.5V + internal losses = 17V.

Maybe there are other ideas or ways of achieving the same, and I do not know how to make a variable DC supply without losing heat anywhere, I mean it seems I will need to have a variable mains transformer ?
 
An easy way to avoid power loose is to detect the battery level, and switch the secondary of the transformer with a relay to a lower voltage. As voltage increases, relay gets higher voltage levels. Easy to do using a LM393 double comparator wired as window comparator. Also you need a tap or taps in the transformer's secondary.
 
It seems like a fan and/or better heatsink would be easiest.
What has confused me most is that if the charger puts out a regulated 12V, when connected to a 6.5V battery it will still output 12V.
I suppose it is the OP's second to last paragraph that has me stumped; it isn't correct to my mind. Charge voltage must be higher than the battery voltage.
If this is a constant-current charger, it needs to be designed for worst case, not kludged with a Variac, etc.
 
Last edited:
Sorry guys but none of the above seems like a practical solution to the original problem...

One way to do it is to construct the charger around a low Rds(on) MOSFET like say the IRF540 which has an Rds(on) of 33 milliohm and capable of 33A drain current so power dissipated isn't gonna be a problem if a normal heat sink is used. to control the current passing through the MOSFET you can use a pulse width modulation (PWM) with a fixed frequency and a variable duty cycle dependent on the battery voltage, to generate this signal you can use various techniques micro controllers, 555 timer or even a computer port. :)
 
Two words: Buck converter.

Linear tech do some sand that is reasonable for this, and you typically get better then 85% efficiency.

Seriously a current mode switcher is the way to do this, and is really not hard these days.

I would note that charging something claiming to be lithium that has dropped that low should be conducted with extreme caution, fire is very possible.

Regards, Dan.
 
Maybe stating the original problem

That's a question for the OP but i guess if you read his words carefully he makes himself pretty clear.

why none of the above seems like a practical solution would make your response come across a little better.

A "fan or a better heat sink" is hiding the problem with circuit design and not solving it. the problem is heat generated/energy wasted by certain circuit element, solution===>> use another element (and perhaps a better technique) that have a very low internal resistance with current handling capabilities way more than needed. Simple ain't it. ;)
 
Sorry guys but none of the above seems like a practical solution to the original problem...

One way to do it is to construct the charger around a low Rds(on) MOSFET like say the IRF540 which has an Rds(on) of 33 milliohm and capable of 33A drain current so power dissipated isn't gonna be a problem if a normal heat sink is used. to control the current passing through the MOSFET you can use a pulse width modulation (PWM) with a fixed frequency and a variable duty cycle dependent on the battery voltage, to generate this signal you can use various techniques micro controllers, 555 timer or even a computer port. :)

If I understand it correctly this technique lowers the average current when the battery is too drained? Is that not the same as in lowering the charging current (somehow) ?
 
It seems like a fan and/or better heatsink would be easiest.
What has confused me most is that if the charger puts out a regulated 12V, when connected to a 6.5V battery it will still output 12V.
I suppose it is the OP's second to last paragraph that has me stumped; it isn't correct to my mind. Charge voltage must be higher than the battery voltage.
If this is a constant-current charger, it needs to be designed for worst case, not kludged with a Variac, etc.

A battery charger is not simply a regulated X volt supply. It is more like a current limited X volt supply, which means that in many/most cases the current limiting circuitry is active. That further means that when the battery is down to 6.5V, the charger will output something like 6.6V-6.8V on the battery - just what is needed to reach and trigger the current limit. The current limit is chosen by the user, depending on the battery, and the choice of slow-fast charging programme. The voltage limit is also chosen by the user depending on how much strain he wants to put on the lithium-ion battery, the higher the voltage the greater the strain.
 
If I understand it correctly this technique lowers the average current when the battery is too drained? Is that not the same as in lowering the charging current (somehow) ?

The benefit from using MOSFET+PWM is you can pass high current to the load while controlling the amount of this current without the control element (the MOSFET) dissipating large amount of energy in the form of heat due to the small voltage drop across the MOSFET.

"Too drained battery" usually indicate a faulty battery or a battery near operational life's end i.e you shouldn't let it be too drained or its voltage drop beyond certain value.
Designing a charger circuit around a somehow faulty or very low quality battery is a futile and fruitless exercise.
 
Out of curiosity may i ask about the battery specifications...voltage, actual capacity, specified capacity and number of cells. :)

Out of curiosity may i ask about the battery specifications...voltage, actual capacity, specified capacity and number of cells. :)

A 12V/9800mAh pack with 3 cells each weighs 54 g. Each cell is double the size of a normal 3.6V/1200mAh cell, so I presume each pack is at around 2400mAh. The cells are connected in series giving 12V nominal. Each cell has a small circuit onboard - this is the protection circuitry.

The capacity is purported to be 9800mAh but in my opinion it is more like 2400mAh if that.

Because the cells had discharged just sitting on the shelf, maybe the protection circuit was faulty, I could not charge the pack. I took the cells out, removed the circuits and charged them individually. It is now working OK, but without any protection.

The reason for the question above was when I tried to use the 12V charger on the discharged pack, it was getting rather hot because it was not designed to charge down to 6-7V and was thinking of ways of improving it.
 
Hi,

Its simply design. You probably can't charge the batteries properly
because 12V is not enough. Your charger sounds crude and wrong.

You probably want about about 14V for proper charging,
and different batteries have different optimum charging
currents related to their state of discharge.

They also have different trickle charge requirements.

They also have different maximum charging states
if used as standby and you want to maximise life.

You cannot charge a 12V battery properly with a
12V supply, a common DIY "boombox" mistake.
You get about 1/3 of real battery capacity.

rgds, sreten.
 
Last edited:
The protection boards themselves may draw a little bit of current Some cordless tool packs are infamous because the protection board draws power from just one set of cells, rather than across the whole pack, and the battery becomes unbalanced and fails.

Consider buying an iMax B6 smart hobby charger from eBay; they're basically clones of clones, but they seem to do the job for under $30 shipped. They'll do any type and number of cells, so you could charge cells individually at first, then the whole lot. Or if you connect the battery pack to the balance input, the charger will make sure the cells are evenly charged.

If the protection boards aren't working, Fasttech.com has a selection of PCMs for different cell arrangements and currents.

Many of the 18650 lithium-ion batteries from China are reputed to be salvaged from old battery packs sent for recycling. We might be better off to salvage batteries ourselves, then test them and use the best. (The B6 charger can do charge and discharge cycles to measure actual capacity.) (18 comes from the cell diameter, 65 from the length in millimeters)
 
A 12V/9800mAh pack with 3 cells each weighs 54 g. Each cell is double the size of a normal 3.6V/1200mAh cell, so I presume each pack is at around 2400mAh. The cells are connected in series giving 12V nominal. Each cell has a small circuit onboard - this is the protection circuitry.

The capacity is purported to be 9800mAh but in my opinion it is more like 2400mAh if that.

Thank you for the detailed answer. For a charge voltage of 4.2V/cell you're gonna need a charger that outputs a 4.2*3= 12.6V with a tolerance of +-150mv, going above that will reduce the battery life . At a charge rate of 0.5C that means a charging current of about 1.2A for the battery to remain cool during charging. The battery is fully charged when the voltage reaches the threshold of 4.2V/cell and the charging current drops to about 72mA. You don't have to fully charge a li-ion battery as that stresses the battery and reduces service life.

In the above calculations i used the actual 2400mAh capacity. Heat generated at this power level should be quite small.:)
 
Thanks for all replies.

Yes my charger is not absolute 12V and is not that crude :) It is configurable because I have not decided yet on a charging regime. Basically any charger you buy has a different programme. For example when your charger shows green light, you have no idea what voltage it has charged to, how fast and so on. Even the same model from same manufacturer I have seen that they charge to different levels, eg 11.8V, 12.2V and so on. I am not talking about super-sophisticated chargers, but rather those small plug-in things that come with the battery packs. The failure of which is what prompted me to make my own, currently on breadboard. I have also used my bench PSU to charge them, but it does not know when to stop :)

The final voltage is a matter of choice, as long as you do not exceed the 4.2V per cell. The charging current again a matter of choice, chould be anything to 1C, but I work with much less, again selectable.

As per message above #17 the charger stops when the current has dropped to a predefined level. Lithium-ions do not need to be charged to max voltage, actually they live longer if you do not charge them that much. However if they discharge too much they may become dangerous (internal shorts).

There are many features missing from my charger, for example a shut down timer, detection of an over-drained battery, excessive heat when voltage is too low, no battery temperature sensor, no smoke sensor :),

Once the protection circuitry has blocked a cell on the battery you cannot do anything other than throw the battery away, or what I did which is to remove the circuits and charge the cells individually and evenly.

I will buy the iMax B6 as suggested since it is so cheap :)

I attach picture of one such cell and a schematic of my charger.

U1 is the current limiter and selector via a switch.
U2 is the voltage controller, where the maximum voltage is selected via a trim pot.
D1 (crudely) serves to protect from funnies at the output. There is a small voltage drop which is accounted for when setting up U2, however I am not striving for 12.6V exactly, I usually stop at 12V so I do not care.
U3A amplifies the voltage drop on the current limiter resistor and references it to earth
U3B compares against set level and lights up a red/green led and drives the relay transistor on/off as required.
S1 is the Start momentary switch
S4 is the Stop momentary switch
The circuit is simplified, missing smoothing caps, relay protection diode etc
 

Attachments

  • battery-charger-Sep-2013.jpg
    battery-charger-Sep-2013.jpg
    123.9 KB · Views: 51
  • lithium-ion-battery-internal.jpg
    lithium-ion-battery-internal.jpg
    262.5 KB · Views: 43
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.