Lithium Ion charging ICs

[MrAl]

Here's an updated drawing with some safety warnings...
I also updated the set point voltage to 4.15v instead of 4.18v as the original design had.
Note also that this design as is charges at a much lower current than most. To increase current (and decrease charge time) the resistor R1 has to be lowered.

 

[Mohamed Najiullah]

Here's an updated drawing with some safety warnings...
I also updated the set point voltage to 4.15v instead of 4.18v as the original design had.
Note also that this design as is charges at a much lower current than most. To increase current (and decrease charge time) the resistor R1 has to be lowered.

Can i use this same circuit if i use 4 NiMH cells in series instead of LiPO batteries?

 

[Hero999]

No you can't NiMH cells can't be charged like Li-ion batteries.

As I've said repeatedly, if you use three AA NiMH batteries and a 200mA solar panel, you don't need to worry about limiting the voltage or the current.

All you need a diode to prevent the batteries from discharging back into the solar panel. I'd recommend a Schottky diode because the voltage drop is lower. If you're using a hand crank dynamo, you'll need to connect it via a separate diode, otherwise the lots of the power from the solar panel will go into powering the motor.

 

[Mohamed Najiullah]

Actually i'l be using 4 AA NiMH batteries of 1.2V each.. I read somewhere that NiMH are also sensitive to overcharging and will be destroyed if overcharged for a long time. So how do i figure out when the batteries r fully charged?

 

[audioguru]

Use a battery charger IC. It shuts off when it detects that the battery is fully charged. It also detects that a battery is already charged then refuses to over-charge it like a simple timer charger will do.

 

[MrAl]

As I've said repeatedly, if you use three AA NiMH batteries and a 200mA solar panel, you don't need to worry about limiting the voltage or the current.
All you need a diode to prevent the batteries from discharging back into the solar panel. I'd recommend a Schottky diode because the voltage drop is lower. If you're using a hand crank dynamo, you'll need to connect it via a separate diode, otherwise the lots of the power from the solar panel will go into powering the motor.

Hi there Hero,


I have to wonder, did you say what you really wanted to say there? You realize that
200ma for extended periods is enough to ruin the NiMH cells in a fairly short time like
3 months right? Are you perhaps basing this on the full sun shining only for short
periods per day?

The charge regimen for charging NiMH cells incorporates either minus delta V, zero delta V,
max temperature cutout, (or a combination of those), and also a default timer.
I realize that the charge current is usually higher but 200ma is still enough.
For example, 90ma will kill the cells in about 6 months.
I guess at 200ma for 24 hours a day 7 days a week would not happen with a solar charger,
so figure 8 hours a day so that's about three times the life at 200ma, or 9 months. Not
too bad i guess.

Come to think of it, a solar source might require special thought as to how often to allow the
charger to actually charge. With the sun going down at night and then coming back up the
next day, the charger would probably go through the algorithm of minus delta V again, and
as each day passes it would go through this over and over again even if the cell hasnt been
drained each day. That might have a detrimental effect on the life of the cell too.

 

[Hero999]

I have to wonder, did you say what you really wanted to say there? You realize that
200ma for extended periods is enough to ruin the NiMH cells in a fairly short time like
3 months right? Are you perhaps basing this on the full sun shining only for short
periods per day?

200mA into a 2500mAh battery is only 0.08C and as you said, the panels are only going to be putting out 200mA for a couple of hours at mid day so I think it's highly unlikely that the cells would be ruined.

 

[MrAl]

Hi again,


Well, 150ma 24 hours a day 7 days a week would take out sub C sized NiCd's too, so i thought i would mention it. That's the reason why cells dont last the full expected time length in many cases. Same with drill packs.
With this app yes it is a little different, because the sun wont be hitting the panel 24/7 at least not on this planet so it may be ok for a year or two.
If i had a 200ma panel i think i would be willing to sacrifice two cells to try this. I had done that in the past with another type of device (90ma) and two cells, and sure enough three times in a row both cells would not take a charge after about 6 months. As i was saying, at 200ma that's about 3 months maybe, but at 1/3 of a day it's 9 months. At 4 hours a day that comes out to about a year and a half, which isnt too bad i guess. It becomes more important when the cost and the number of cells goes up. Two cells dont cost much, but 20 cells would be more of a problem. With the NiCd drill packs and other devices that run from multiple NiCd's sub C cells the cost becomes important. I want more than a year out of my sub C cells, i want five years and that's what i get with a specially designed charger. With a charge 24/7 the cells only last a year.

 

[Mohamed Najiullah]

Have i mentioned that i live in Southern India, where there r only three seasons.. Hot, hotter and hottest.. Even at mid winter we have a min temperature of 25 degrees celsius and a max of 30.. In summer the temperature goes up to 45 degrees..

And btw i thot only overcharging destroyed the batteries.. But will continuous charging, (despite the fact that the cells will be dicharged atleast once a day, ) also destroy the batteries?

 

[Hero999]

What about limiting the voltage?

Surely if you limited the voltage to 1.4V per cell, you could float charge them all day without any problems?

So three NiMH cells couldbe charged to 4.2V using a TL431 and a transistor regulator (see the TL431 datasheet).

 

[Mohamed Najiullah]

The cells are of 1.2V each...

 

[Hero999]

That's the nominal voltage, the batteries should be charged to a higher voltage, I suggested float charging to 1.4V per cell.

 

[audioguru]

On Energizer's website there is a Ni-MH Battery Applications Manual that shows this charging voltage graph:

 

[Hero999]

I seen that before.

How do you charge it from a solar panel when the current can randomly vary between 0mA up to 200mA?

How would you know that the sudden decrease in voltage is caused by the battery being fully charged and not by the sun going behind a cloud?

 

[MrAl]

Hi again,


The short answer is that you dont know what the decrease in voltage is from, but then again you dont have to know this anyway. If the voltage decreases, that will shut the charging off anyway in both cases. If you really want to, you can monitor panel voltage and current. Note also that temperature measurement may or may not work, because 200ma, if that is even available, may not be enough to heat the cell up much. If the current is lower like 100ma that may not do it at all.

The real question i think is how would we detect end of charge anyway? Notice the graph that audioguru was nice enough to post in his previous post. In that plot we see that for a charge rate of 1C the charge 'hump' is quite pronounced, but for lesser current levels it is barely noticeable. That's very typical of NiMH cells and so that presents yet another problem with charging these kind of cells with a solar panel where the max current is much less (200ma) than the typical 1C charge rate (1000ma to 2500ma depending on cell size and capacity).

This leads me to believe that NiMH cells are not the best choice for solar applications like this. Li-ion are better anyway, as long as the charger is designed right. If the user still wants to use NiMH then perhaps just build the simpler charger with the diode and resistor and let them overcharge once in a while, if that even happens since the cells will probably be used day to day.
Another idea is to use a micro controller and monitor the charge time and discharge time and try to charge the cells only until they are full and that's it. Wouldn't be perfect, but it's a start i guess. Cells would still have to be allowed to overcharge a little, but not as much even if the cells were not used every day. We could also think about this for a while and maybe come up with a better idea.

Here is a link but i dont think they mention the solar charging problem:
**broken link removed**

 

[Mohamed Najiullah]

Hi again,


The short answer is that you dont know what the decrease in voltage is from, but then again you dont have to know this anyway. If the voltage decreases, that will shut the charging off anyway in both cases. If you really want to, you can monitor panel voltage and current. Note also that temperature measurement may or may not work, because 200ma, if that is even available, may not be enough to heat the cell up much. If the current is lower like 100ma that may not do it at all.

The real question i think is how would we detect end of charge anyway? Notice the graph that audioguru was nice enough to post in his previous post. In that plot we see that for a charge rate of 1C the charge 'hump' is quite pronounced, but for lesser current levels it is barely noticeable. That's very typical of NiMH cells and so that presents yet another problem with charging these kind of cells with a solar panel where the max current is much less (200ma) than the typical 1C charge rate (1000ma to 2500ma depending on cell size and capacity).

This leads me to believe that NiMH cells are not the best choice for solar applications like this. Li-ion are better anyway, as long as the charger is designed right. If the user still wants to use NiMH then perhaps just build the simpler charger with the diode and resistor and let them overcharge once in a while, if that even happens since the cells will probably be used day to day.
Another idea is to use a micro controller and monitor the charge time and discharge time and try to charge the cells only until they are full and that's it. Wouldn't be perfect, but it's a start i guess. Cells would still have to be allowed to overcharge a little, but not as much even if the cells were not used every day. We could also think about this for a while and maybe come up with a better idea.

Here is a link but i dont think they mention the solar charging problem:
**broken link removed**

Mr.Al
could i use the same charging circuit that u gave for the LiPo charging?

 

[MrAl]

Hi there,


NiMH cells require a different charging method and so need another type of circuit. So, sorry but no, you can not use that charger for NiMH cells.

The only advice i can offer at this time is to use Hero's idea of a resistor and diode to limit charge current. There is a chance that the NiMH cells will over charge on occasion, but that it shouldnt affect the life more than limiting to about 1 year or two instead of five years or something like that. You might have to replace the cells more often that's all. If you are only using four cells this shouldnt be too expensive.

With a typical NiMH charger you would never want to charge them continuously at 200ma as that would shorten the life quite a bit, but with a solar charger you wont be getting full charge current 24 hours a day 7 days a week so even at 200ma the cells should last at least a year or two. That's not too bad considering that there is no good way to detect the end of charge for NiMH batteries (AA size) at only 200ma current level.

I dont know if you are into using microcontoller ic chips or not, but you could monitor the charge and discharge and adjust the charge time using the measured data. That would be a much more complex design though and require a well written algorithm to function properly for years to come. I really dont think you want to be bothered with this much complexity. If someone wanted to build one just for the experience that would be different of course as it would be an interesting circuit and algorithm.