Is this logic correct? (Charging Li-ION Protected Cells)

[ravishi]

So I have a project going and have an idea of how everything is going to work, but I wanted to run it by the users of this forum to see if I am not about to do a grave mistake.

I have two small solar panels that are about 2 Watts each (8.5V, 260mA Short Circuit). I want to charge two 18650 Lithium Batteries using these cells while still providing power to a small load of about 250 mA's max. The batteries I bought (http://dealextreme.com/details.dx/sku.5776) are protected cells and have overcharge protection. That is the reason I chose these over NiMH since I would need a seperate overcharge IC and circuitry. Is that a safe assumption?

If so, my circuit would be like this:
- The two solar panels in parallel with blocking diodes to give 8.3V and 500 mA current
- The two batteries in parallel to increase capacity and then use a step up voltage regulator to output 5V
- A load of 250mA connected to the voltage regulator
- A seperate DC converter (6V 500mA) to charge the batteries when there is no sun

I actually did a test for about two hours in the sun to charge the batteries. They were not fully discharged. After the two hours, I read the voltage on the cells and it read 3.95V. That worried me considering they are rated for 3.7V.

 

[Turbo Guy]

Li-ion 3.7 is the nonimal voltage. They charge to 4.2 volts. Not all protection devices are the same however most Li-ion protected cells have ones which disconnect at 4.25-4.35 per cell and also at somewhere around 2.5 volts. Most also have some type of over current protection during both charging and discharging.

Go to a site such as allbattery.com and look up the specs. for some of their protected LiIon cells.

 

[matrixshaman]

Safe voltage to end a charge for Li-Ion is usually considered to be 4.20 volts maximum. Anything over that shortens the life. Anything much less and they are not fully charged. I usually pull mine around 4.10v to 4.20 volt. So the ones you charged are safe but not really fully charged yet. Sounds like a nice setup. The output voltage of your setup is not as critical to charging as is the current setup. I believe 500ma sounds safe enough if you can verify that is the maximum available in full Sun light.

 

[Benson]

They can charge up to 4.2V, so they're not unsafely overcharged yet, but most cell protection PCBs are not designed for use as charge cutoff:

• They're designed for occasional emergency protection, and may fail when used repeatedly this way.
• They are typically set high enough that the cell will suffer reduced life from mild overcharging.
• When they cut out to stop charging, they'll also stop providing current to the load until the external application of voltage is released, so you may pick up some ugly transients as the light on the solar panels drops to where they can't support the load, and then the fully charged battery cuts back in.

This is really not a good way to do it, unless you know your batteries' protection is designed to serve as charge termination, and you've run some tests to make sure the solar -> battery switchout behaves well enough. Otherwise, I think an explosion is unlikely, but you're definitely risking shortened battery life, and there is some risk that the protection fails the wrong way and permits overcharge.

 

[Turbo Guy]

They can charge up to 4.2V, so they're not unsafely overcharged yet, but most cell protection PCBs are not designed for use as charge cutoff:

True however some of the cheap charger do inteed charge until the cells protection device terminates the charge and many use such chargers on a regular bases.


You are also correct that limiting charge to 4.1 per cell will greatly increase the life of the cells as will limiting the discharge to 3.0 volts per cell.

 

[mdocod]

- The two solar panels in parallel with blocking diodes to give 8.3V and 500 mA current

that part sounds reasonable...

- The two batteries in parallel to increase capacity and then use a step up voltage regulator to output 5V

Total stored energy is a function of both amps and voltage deliverable, wiring batteries in series or parallel does not change the amount of energy stored. I would be tempted to wire them in series and test to see if the charging stops when the pack reaches 8.3V with that diode installed.. A step-up regulator is going to introduce significant inefficiency, it's easier to step voltage down efficiently. The cells in a series configuration however would occasionally need to be re-balanced on a regular charger, testing would be required to determine how often re-balancing would be needed.

- A load of 250mA connected to the voltage regulator
- A seperate DC converter (6V 500mA) to charge the batteries when there is no sun

I wouldn't rely on the PCB in those cells to terminate the charge from a dummy power supply. However, if you WERE going to use the PCB to terminate the charge in this fashion, you must increase the charge speed so that when the charge is terminated at ~4.35V, the resting voltage of the cell after termination is at or under 4.20V if possible. I would suggest ~1.5-2A charge rate to achieve this behavior. Charging a LiCo cell at 0.25C (600mA/2400mAH) and then using the PCB to terminate the charge would undoubtedly result in a final resting voltage around 4.25V give or take, this would reduce cycle life of the cell dramatically.

-Eric

 

[ravishi]

Thanks to everyone who replied. I am sorry for not following up, I haven't had much time to work on this project.

So it looks like I need to better protect my batteries so I have been thinking of a way to do it. I didn't mention but I have a microcontroller that I am using in my design which allows me the possibility of reading the voltage of the batteries.

So my new idea is to do this, same setup as before:
2 solar panels in parallel(7.5V, 260+260= 520mA)
Step up regulator (Boost, switching, 90% efficiency)
2 Parallel batteries Li-Ion

And now I would add a voltage divider circuit of two 500kΩ in series with a tap between the two which goes into an input of the A/D. This battery voltage line will be half of the current battery voltage. Then, if this voltage goes above 2.05V, I could terminate the connection between the solar panels and the batteries. I wouldn't turn it back on until it went to 2.00V to reduce flicker.

However, what I am struggling with is designing the switch that will close and open the connection between the batteries and solar panels. I am forgetting my basic transistor knowledge. I have used NPN transistors but the problem with those is that you need the base voltage high to bias the transistor. Since the solar panel voltage is 7.5V and the microcontroller can output 5V, I don't think it will work. Is a PNP transistor what would solve my problem or is there something else?