Choosing the correct MOSFET for a low voltage cutoff circuit

Thread Starter

justinoberle

Joined Jun 16, 2021
36
I have a circuit that is draining my Li-Ion batteries far too low. In some cases it is destroying the batteries. Because of this, I would like to create a low voltage cutoff but I have a couple of questions. My power source is two 123-a 3.7V 800mAh batteries in series. I have a voltage regulator producing a constant and stable 3.3V for the circuit. I am considering using a MOSFET to switch the power off when the batteries get too low. However, I am unfamiliar with these and I am unsure which one to get. The circuit draws anywhere from 70mA to 200mA when on, which is in timed intervals. When off, it draws almost no current. And, as described above, it needs a constant 3.3V.
1) Do I need a low voltage cutoff(LVC) for each battery or can I place one LVC after the 2 batteries in series? I was worried one battery would still over drain.
If I only need one LVC, then the on voltage could range from 5V+ and the off voltage would need to be at 3.5V or less
1) Any suggestions on which MOSFET to choose here?
2) Should I use a BJT instead? And if so, suggestions on which one?
 

Wolframore

Joined Jan 21, 2019
2,407
Please supply a schematic. There are too many questions otherwise. Do you intend to charge it while in circuit? What sort of regulator? ... etc
 

Wolframore

Joined Jan 21, 2019
2,407
The linear voltage regulator is a little wasteful but it will work. 2 Li-ion cells is 8.4 - 6.4V. A cutoff can be put in I would use a logic level P channel with the gate controlled by the ESP. Mind the body diode. Calculate for circuit current and use one with low RDS. I keep AO3401 around and wouldn’t think twice about using it for this application. Or Si2301. Not sure what size format you need. These are SMT.
 

Thread Starter

justinoberle

Joined Jun 16, 2021
36
The linear voltage regulator is a little wasteful but it will work. 2 Li-ion cells is 8.4 - 6.4V. A cutoff can be put in I would use a logic level P channel with the gate controlled by the ESP. Mind the body diode. Calculate for circuit current and use one with low RDS. I keep AO3401 around and wouldn’t think twice about using it for this application. Or Si2301. Not sure what size format you need. These are SMT.
Thanks, although some people might use two 3V batteries instead of the 3.7V ones. This lowers the range to about 5V. Are those MOSFETS you mentioned able to handle this case?
 

Audioguru again

Joined Oct 21, 2019
3,866
The datasheet for an Energizer CR123A battery shows that it produces 2.8V at 28mA into 100 ohms for about 40 hours.
It is not an ordinary 3.7V (charged to 4.2V) Lithium rechargeable battery.

You said you have two of these battery cells in series (you said only 3.3V) then they must be in parallel. Two cells in series are charged to 6.6V and produce 5.6V at 28mA for 40 hours.

The datasheet does not say the minimum voltage but the voltage drops fast below 2.5V per cell.
 

Thread Starter

justinoberle

Joined Jun 16, 2021
36
The datasheet for an Energizer CR123A battery shows that it produces 2.8V at 28mA into 100 ohms for about 40 hours.
It is not an ordinary 3.7V (charged to 4.2V) Lithium rechargeable battery.

You said you have two of these battery cells in series (you said only 3.3V) then they must be in parallel. Two cells in series are charged to 6.6V and produce 5.6V at 28mA for 40 hours.

The datasheet does not say the minimum voltage but the voltage drops fast below 2.5V per cell.
No, they are in series and my voltage regulator reduces a stable 3.3V.
 

Wolframore

Joined Jan 21, 2019
2,407
Thanks, although some people might use two 3V batteries instead of the 3.7V ones. This lowers the range to about 5V. Are those MOSFETS you mentioned able to handle this case?
Look up the datasheets they are readily available. AO3401 can take 30v and will active with -2.5V @ 85mΩ Rds. Is that good enough?

R3 and R4 is an issue. They can be an order of magnitude higher (resistance) and still perform what you need.
 
Last edited:

Thread Starter

justinoberle

Joined Jun 16, 2021
36
Thanks! To be completely honest, datasheets can be a little overwhelming to me sometimes. For me, it is TMI and I can't seem to find what I am looking for in there. I try to read them very carefully, but it is a lot to take in, sometimes, not all the time.
You mentioned the linear reg was wasteful. What would you have gone with to get stable voltage? Just curious. And, my R3 and R4 are actually 1k and 2k, not 300 and 340 as shown in the schematic.
 

Martin_R

Joined Aug 28, 2019
103
Nice circuit. I would split the 10k sourc e-gate resistance in half and the junction of the resistors connect the 'on' switch to avoid shorting the supply of both buttons pressed
 

Thread Starter

justinoberle

Joined Jun 16, 2021
36
The linear voltage regulator is a little wasteful but it will work. 2 Li-ion cells is 8.4 - 6.4V. A cutoff can be put in I would use a logic level P channel with the gate controlled by the ESP. Mind the body diode. Calculate for circuit current and use one with low RDS. I keep AO3401 around and wouldn’t think twice about using it for this application. Or Si2301. Not sure what size format you need. These are SMT.
What is the breadboard friendly version of the AO3401?
 

Thread Starter

justinoberle

Joined Jun 16, 2021
36
View attachment 244895
I have used these in the past, just solder 3 pin header to them
Sorry, very late post but I just got these boards in and can now pick this back up. I have it working with a voltage regulator on the gate at 3.25V and the input voltage reduced by resistors for the source. This all works. When the source gets to about 3.6V the output is 0V. When it is above this, the output is whatever the reduced voltage is. I have 2 issues. I really want my regulated voltage as my output, not the resistor reduced voltage. And also, how do I make this reverse battery protected?

Here is a schematic. The source is 8V. Using resistors to drop it to 4.4V. If the source goes to 6.5V, the resistors drop the voltage to 3.6V which produces 0V from the drain. This is fine except I would like this to produce the regulated voltage from the drain if possible. If not, I would at the very least need the original voltage back so I can regulate it again to 3.25 for my system. Of course, unless it drops to low.
1629250479691.png
 

LowQCab

Joined Nov 6, 2012
1,350
A 10K Resistor is not going to allow enough Current for reliable operation.
This Circuit is basically all wrong,
and will just ruin the Battery by running it completely dead.

The Battery must be completely isolated from any Load when the Circuit is Off.
.
.
.
 

Thread Starter

justinoberle

Joined Jun 16, 2021
36
What in the actual [DELTED] are you doing?

Why are you regulating the gate and not the pass?
What is the pass? It has gate, drain, source. It was my understanding that if the source voltage is within the threshold or less than the gate, the drain is 0V. So, I have the gate regulated and as the source drops, once it drops to the threshold, the drain goes to 0V. Is this incorrect? If so, can I get some assistance in this schematic please? It works but the drain is unregulated, which is not what I want.
 
Last edited by a moderator:

Thread Starter

justinoberle

Joined Jun 16, 2021
36
A 10K Resistor is not going to allow enough Current for reliable operation.
This Circuit is basically all wrong,
and will just ruin the Battery by running it completely dead.

The Battery must be completely isolated from any Load when the Circuit is Off.
.
.
.
That was my 1st attempt at this. I will try to follow the schematic you posted above. If I wanted to exclude the switch would it look like this below, with everything in the red circle excluded? Also, is that 10k resistor from S to G necessary? I thought the P mosfet had something like this built in.
1629322464924.png
 
Top