LI Ion Battery Charging Disconnect Circuit help Needed

Thread Starter

mishra87

Joined Jan 17, 2016
854
Hello Guys,
I wanted add a disconnect circuit using MOSFET switch with microcontroller as per below. uC will cut off charging using mosfet in series so user have control to charge the battery.


I wanted to use mosfet based switch and i tried to do LT Spice Simulation but i am confused if it is correct .
I assume battery will always have charge of 3V to 4.2V when user want to charge it. I am not sure if this circuit works well as per my thought.
battery cut off is 3V

Could anybody help me ?
LT Spice Circuit is attached here.

Thanks in Advance !!
1591878722959.png
 

Attachments

Last edited:

Thread Starter

mishra87

Joined Jan 17, 2016
854
Edit : V2 =4.2V from Battery charger IC
C1 = Li Ion battery (defined range is 3 to 4.2V but it could be empty)
 
Last edited:

Thread Starter

mishra87

Joined Jan 17, 2016
854
How are you testing for primary battery? ADC can only read volts...
Based on cell resistance.
My intention here to understand about LT Spice Circuit if this will work as per application since MOSFET drain is always 4.2V and Source could be Li ion battery usually 3V to 4.2 or 0V-4.2V

Lets forget about primary battery and analyse this circuit.

Regards,
 
Last edited:

Irving

Joined Jan 30, 2016
1,056
Well, 3.3v on base of transistor turns on MOSFET and 5v appears across C1 as expected... there's not much more to say...

Source is at 5v, Drain goes from 0v (after C1 discharged) to 5v when MOSFET turned on.
 

Thread Starter

mishra87

Joined Jan 17, 2016
854
Well, 3.3v on base of transistor turns on MOSFET and 5v appears across C1 as expected... there's not much more to say...

Source is at 5v, Drain goes from 0v (after C1 discharged) to 5v when MOSFET turned on.
Thanks for your reply & time.

I understood what you are trying to say. My doubts was since in place C1 i will use Li Ion battery. My lower cutoff for the battery will be 3V so i assume battery will charge from 3V to 4.2V when mosfet is turned ON.

In some cases battery may be completely drained at 0V.

I just wanted to understand the mosfet behavior. Since mosfet source is always at 4.2V from battery charger CC/CV chip and lets say mosfet drain is at 3V so source to drain voltage difference will be -2V and at some point of time when battery will be completely charge then source to drain will be at same potential i.e. 0V so i am not sure still mosfet remained turned ON and still battery will be on charging mode.

CC/CV charging chip terminate the charging.

Regards,
 

Thread Starter

mishra87

Joined Jan 17, 2016
854
I am not sure if people understand correctly the problem statement.

I simply wanted to charge Li battery as per post#1 block diagram.
Battery charger chip (CC/CV) -> Switch (MOSFET) -> Li Ion Battery

To acheive this i wanted to implement a switch circuit between Charger chip & battery.

Hope this will clear perception !

As per my understanding to run P-mosfet in saturation region below condition has to be met.

VSG > |VT| , VSD > VSG - |VT|

And second condition void when VS = VD = 4.2V (when is at 4.2V charge voltage)


Regards,
 

Irving

Joined Jan 30, 2016
1,056
Your NPN transistor pulls gate down to near 0v so MOSFET stays on since Vgs > Vth (Vth for that FET is 1.8V).

Where you might have an issue is when charger is in CC mode. While FET is off, charger volts will be 4.2v, but once FET turned on the charger output voltage (and therefore FET source) will be just above the Li-Ion battery volts and so Vgs is roughly battery volts. As Vth ~ 1.8-2v then it'll work down to 3v. If you let pack volts go below 3v then the pack could be damaged and I'd not want to charge them in the device. At 0v they are toast and dangerous to recharge in situ. Run away!
 

Thread Starter

mishra87

Joined Jan 17, 2016
854
Your NPN transistor pulls gate down to near 0v so MOSFET stays on since Vgs > Vth (Vth for that FET is 1.8V).

Where you might have an issue is when charger is in CC mode. While FET is off, charger volts will be 4.2v, but once FET turned on the charger output voltage (and therefore FET source) will be just above the Li-Ion battery volts and so Vgs is roughly battery volts. As Vth ~ 1.8-2v then it'll work down to 3v. If you let pack volts go below 3v then the pack could be damaged and I'd not want to charge them in the device. At 0v they are toast and dangerous to recharge in situ. Run away!
Thanks for your support !

While was studying more about this circuit people use below for polarity reversal for battery so why can't it be used used for charging purpose ?
1591960684663.png

Regards,
 
Last edited:

Irving

Joined Jan 30, 2016
1,056
Not sure I understand the meaning of your question.

As I said before, this will work for charging battery starting at 3v or even 2v, but once battery goes much below 3v I'd be wary of charging it in a device. I definitely would not charge in the device if its significantly below 3v; the risk of thermal runaway is too high and I'd be charging the battery alone outdoors or in a protective environment!
 
Last edited:

Thread Starter

mishra87

Joined Jan 17, 2016
854
Not sure I understand the meaning of your question.

As I said before, this will work for charging battery starting at 3v or even 2v, but once battery goes much below 3v I'd be wary of charging it in a device. I definitely would not charge in the device if its significantly below 3v; the risk of thermal runaway is too high and I'd be charging the battery alone outdoors or in a protective environment!
Thanks for reply !

I think there some gap in understanding.

From your post i understand that if my battery voltage is 2V the mosfet will get turned ON and battery will be charged till 4.2V then battery charger chip terminates the charging.

The Idea was to use mosfet based circuit that user will have the control in his hand by using uC.

When i say battery charger chip : e.g. https://www.monolithicpower.com/en/...2/document_type/Datasheet/lang/en/sku/MP2660/

These are very intelligence chip which will define the charging profile of battery based on battery voltage. So these chips always monitor the battery voltage first and then set the charging profile.
When you say Thermal runway/protection environment, does it mean because of mosfet is connected between charging chip & battery hence it needed to be protective.

As per my understanding MP2660 will take care all the things only just i wanted a connect a switch between MP2660 and battery ! MP2660 will decide the charging profile of battery.

Then my doubt was, Will P MOS work if S & D of mosfet will have same voltage level i.e. 4.2V output from MP2660 and 4.2V from battery.


Now my question are :

1. Will this p-mosfet be on always(even battery completely drain/battery voltage could be anything) to charge the battery at 4.2V and MP2660 will terminates the charging ?
2. Is there any risk to charge the battery using mosfet (neglect the charging time due to Rds(on) of mosfet) ?
3. I had impresseion mosfet will not work when both Source & Drain will be at 4.2V. (i.e. 4.2V output from MP2660 and 4.2V from battery)

Thanks again !
 

Irving

Joined Jan 30, 2016
1,056
Explain why you need the mosfet when the MP2660 has its own switch? What does the this added mosfet give you?

answers to questions.
1. Yes, as long as MP2660 output is >~2v and NPN transistor is on.
2. No, no risk, but giving user additional control could leave battery in an indeterminate state of charge.
3. Mosfet is on as long as Vgs > Vth. If Vd = Vs no current flows but mosftet still on.
 

Thread Starter

mishra87

Joined Jan 17, 2016
854
Explain why you need the mosfet when the MP2660 has its own switch? What does the this added mosfet give you?

answers to questions.
1. Yes, as long as MP2660 output is >~2v and NPN transistor is on.
2. No, no risk, but giving user additional control could leave battery in an indeterminate state of charge.
3. Mosfet is on as long as Vgs > Vth. If Vd = Vs no current flows but mosftet still on.
Thanks for your support !

I think i got my answer.

Regarding "giving user additional control could leave battery in an indeterminate state of charge". I can use more intelligence chip which will delicately monitor battery voltage and then it will take decision on charging. see attachment.

I have done some more research on p mosfet and below is the understanding. mosfet will always turned on in saturation.
because it will always meet below two condition

1. Vgs > Vt
2. Vsd > Vsg - Vt
1591972774733.png1591973668908.png
 

Attachments

Last edited:

Audioguru again

Joined Oct 21, 2019
2,048
I guess you did not read about charging and discharging Lithium batteries on The Battery University.com and you did not look at charger IC datasheets.
1) If a Lithium battery voltage is about 2.5V or 3V or less, it is dangerous to suddenly normally charge because some of its ions might have converted into metallic Lithium and will cause an explosion or fire. If a battery charger IC detects the low cell voltage then it attempts charging with a low current. If the voltage does not rise or if the temperature rises then it stops charging with a warning indication.
2) A battery cell is nowhere near a full charge when its voltage reaches 4.2V. When it becomes 4.2V then it is fully charged and disconnected later when it charging current becomes much less.
 

Irving

Joined Jan 30, 2016
1,056
I can use more intelligence chip which will delicately monitor battery voltage and then it will take decision on charging. see attachment.
MP2660 has all the intelligence you need, its a perfectly good chip for managing the battery and controlling charge & discharge. I still can't fathom why you think you need to do it differently? Seems to me you are reinventing the wheel unnecessarily!
 

Thread Starter

mishra87

Joined Jan 17, 2016
854
I guess you did not read about charging and discharging Lithium batteries on The Battery University.com and you did not look at charger IC datasheets.
1) If a Lithium battery voltage is about 2.5V or 3V or less, it is dangerous to suddenly normally charge because some of its ions might have converted into metallic Lithium and will cause an explosion or fire. If a battery charger IC detects the low cell voltage then it attempts charging with a low current. If the voltage does not rise or if the temperature rises then it stops charging with a warning indication.
2) A battery cell is nowhere near a full charge when its voltage reaches 4.2V. When it becomes 4.2V then it is fully charged and disconnected later when it charging current becomes much less.
Thank you so much for your reply and participation on thread !

I will read the article to understand the charging & discharging and i can understand your concern about safety.
I do have NTC for cell temperature measurement.

Lets discuss a bit more and may be this thread will be useful to others as well.

For a time being i remove mosfet.
Now if i connect a battery and then i connect battery charger and switch on charger. The battery charger IC eg. mp2660 will first monitor the battery voltage & and based on battery voltage it will start charging the battery as per below . Nothing is written here that battery will explode because IC manage all 5 conditions of battery charging.
Why do you think if IC is having over discharge protection.
1591987810897.png
1591988355346.png

If i connect mosfet as per above-
Charger chip will perform same activity as per above accept few mV voltage drop due to mosfet Rds(on). This voltage drop could leave charging IC in indeterminate stage and could have more charging time. I do not see any safety concern.
This voltage drop issue can be corrected by below functionality of charger IC. Why do i bother about Rdson drop if i monitor the battery voltage with individual cell.(Although Rds(on) should be 10mohm pmosfet) see below

1591988566105.png

I hope you will get some perception about battery charger IC.

Regards,
 

Thread Starter

mishra87

Joined Jan 17, 2016
854
MP2660 has all the intelligence you need, its a perfectly good chip for managing the battery and controlling charge & discharge. I still can't fathom why you think you need to do it differently? Seems to me you are reinventing the wheel unnecessarily!
Thanks !
I am also not in much favor to implement this circuit. The idea to use this mosfet circuit to have control in hand.
Lets say if user connect non rechargeable battery and mp2660 could not recognize the battery type and it measure battery voltage and start charging to non rechargeable battery. So to avoid this problem i wanted to give this protection. Frist i will identify the battery type and then i will take decision on charging based on battery type.

At the moment i do not know, how can one identify the battery type.

if you think, this can happen .
 

Audioguru again

Joined Oct 21, 2019
2,048
Is your battery only a single cell that has fairly low current? Then the MP2660 will be fine as a charger and low voltage disconnect device. You must limit the input voltage and charging current to prevent the MP2660 from overheating.
Do you have I2C to control it?
 

Thread Starter

mishra87

Joined Jan 17, 2016
854
Is your battery only a single cell that has fairly low current? Then the MP2660 will be fine as a charger and low voltage disconnect device. You must limit the input voltage and charging current to prevent the MP2660 from overheating.
Do you have I2C to control it?
Yes, my capacity is few hundred mAh.
Yes i do have I2C control.

Regarding MP2660 overheating additionally it has thermal shutdown.
MP2660 will be used in host mode !

Regards
 

Irving

Joined Jan 30, 2016
1,056
MP2660 will detect battery as non-chargable/faulty as it applies a charging current and looks for a rise in volts. If it doesn't see that it will stop charging. It will stop charging if battery temp rises above a limit and also if its core temp rises too much.

I don't understand why/how you think user would swap rechargables for primary cells? Build sensible capacity single Li-Ion cell in to start with. Normally primary/recharge option would be AA/AAA Alkaline v NiMH but thats a dead technology route...
 
Top