Li ION battery

[student4ever]

can some1 there tell me about the protection circuit of Li ION battery? are they just use switch? tqvm.

 

[Hero999]

They're a bit more complicatad than that. Li-ion batteries need to be protected from: overheating, overdischarge and overcharging.

 

[bountyhunter]

can some1 there tell me about the protection circuit of Li ION battery? are they just use switch? tqvm.

I helped design the controller IC that NSC was building to go into Sony's packs, I don't know if they ever released it. They control two series connected FETs acting as switches. They open the switches if any of these events happen:

1) Cell voltage gets too high or two low

2) too much current drawn

3) over temp condition

 

[student4ever]

protection

ooo i c, then normally they use transistor or relay to be the actuator in the protection circuit.

any other protection circuit thats industry use?

 

[bountyhunter]

ooo i c, then normally they use transistor or relay to be the actuator in the protection circuit.

any other protection circuit thats industry use?

Every one I have ever seen is solid state devices, never a relay. Motorola sells an IC that does it, might be in their catalog. The actual functions are pretty complex: they have to have a clock and enough logic smarts to time a current overload for a certain number of clock cycles, then open the FETs for a while and then "retry" intermittently after a time out period. These protection circuits are not trivial.

 

[student4ever]

suggestion

okey... actually i want to get some ideas for Li Ion protection circuit. i want to model it by using MATLAB. i've idea to use PLC, what ur opinion? with all the sensor ,PLC as the controller and actuator. is it practical? tqvm to all support.

 

[bountyhunter]

okey... actually i want to get some ideas for Li Ion protection circuit. i want to model it by using MATLAB. i've idea to use PLC, what ur opinion? with all the sensor ,PLC as the controller and actuator. is it practical? tqvm to all support.

My opinons are:

1) In my 33 years as an engineer, I never trusted models because they always lie if you wait long enough. My boss would tell me to use a model and I would answer:

"If I model it, I will still have to build it up to verify performance anyway. How about I skip the wasted first step and just give you real data?"

2) The LI protection ICs are very complicated because of all they have to do. They would not be easy to model. At NS, the stupid protection IC was so complicated (and had so many functions) they never could figure out a way to test the IC in the finite time necessary to get the test cost under control. IMHO, this would not be a model that has any representative value... if any model does.