Battery protection circuit schematic. Charging/discharging over-temperature protection

Protection Mechanisms For Batteries And Their Charging Systems

We take batteries for granted and often use them recklessly without taking care of them and their charging systems. This results in their shorter life and sometimes outright failure when we need them the most. The protection mechanisms described here could protect the batteries and their chargers even when these are misused.

Our modern-day lives are quite dependent on batteries, which come in all forms, shapes, and internal chemistries and are found almost everywhere. Batteries that can only convert their chemical energy into electrical energy and their electrochemical states cannot be revived are called primary batteries. Those whose electrochemical states can be revived back, using electrical current, are called secondary batteries or rechargeable batteries.

The secondary batteries are quite interesting in their electrical properties. When these are being discharged they behave like an energy source, but during charging they act as electrical loads. Rechargeable batteries come in different chemistries. The most prolific ones are the various types of lithium-ion (Li-ion) batteries found in almost all portable devices, such as mobile phones, laptops, tablets, toys, etc. We shall be looking at the various battery protection mechanisms used in electronic circuits for such rechargeable batteries.

battery, protection, circuit, schematic, charging, discharging

The lead-acid batteries are either sealed or tubular. These are also used very often in domestic, industrial, and automotive settings and, as of now, form the backbone of electrical energy storage applications, such as domestic and industrial inverters, emergency lighting, and various automotive applications.

The other two most used chemistries for rechargeable batteries are the NiMH (nickel metal hydride) and NiCd (nickel cadmium) batteries. These were till recently used in still photography cameras, however their use in these areas has diminished and taken over by Li-ion batteries. These are still used quite widely in railway applications as battery backup for the coaches.

It is interesting to note that all the above battery chemistries have been known for at least the last 40 years, with the lead-acid ones being known for more than 150 years now. From an electronics circuits design standpoint, the protection mechanisms that we shall discuss apply to all types of secondary (or rechargeable) batteries. Some protections are required during the charging process, while others make sense only during the discharge process.

Thus, some protections are implemented as part of the charger, while others are implemented as part of the battery management system that oversees the charging and discharging process of the battery.

The protection mechanisms we shall discuss are listed under Table 1. It is important to note that these are listed for the complete battery and not for individual cells. Protection of each cell would be part of another article on battery management systems.

Before we get into the details, it would be prudent to discuss the typical charging stages for a battery. Generally, battery charging is divided into three distinct stages. These can then be subdivided into as many as seven different stages based on the state of charge (SoC) of the battery. The SoC is a figure of merit of a battery’s charge capacity and is quite elusive to measure accurately. However, for a large part of the charge-discharge curve of the battery, the SoC can be estimated using the battery terminal voltage.

The three charging stages are bulk, absorption, and float stages. Some battery chemistries are quite sensitive to impressed voltages and hence only expect very slight float voltage and usually require the charging circuit to disconnect once a certain terminal voltage is reached.

battery, protection, circuit, schematic, charging, discharging

The most common charging algorithm used is the CC-CV method, which stands for constant current-constant voltage method. While some chemistries only require a CC method, the others get charged using only the CV method. Variations in the impressed voltage and the permissible charging current result in more charging stages.

Charging over-voltage protection

Charging over-voltage protection mechanism must ensure that during the charging process the battery is not subjected to voltages higher than the maximum recommended value. While the battery is in the CV charging phase, over-voltage protection is a continuous control problem, where a complicated PID loop is used to ensure that the voltage is fixed and does not exceed the voltage set point for that phase. Typically, this is achieved using a voltage regulator design, either using a discrete fixed or adjustable voltage regulator IC, or by using embedded code for a microcontroller which produces control signals to keep the voltage to the battery constant.

In the CC charging phase, over-voltage protection is a simpler limiting control problem. Here a comparator circuit in hardware or comparison operators in the microcontroller code can be used to ensure that the impressed voltage is lower than the maximum permissible voltage. Generally, both continuous control and limiting control functions are integrated into the charger to provide reliable control.

battery, protection, circuit, schematic, charging, discharging

Charging over-current protection

This protection mechanism ensures that the current flowing into the battery is kept below a maximum permissible value. It is quite clear that one cannot push current into a load unless the impressed voltage is set to a value such that the required current flows against the load resistance. Thus, voltage control is an essential part of current control.

When the battery is in CC charging phase, this protection becomes part of a continuous control strategy, which aims at keeping the current value constant within the given battery terminal voltage envelope. This is generally achieved using a linear or switching regulator design, like the voltage control strategies shown above, where the quantity being controlled is current. While there are tens of ways to create a constant current source, only the simplest ones are presented below for reference.

In CV charging phase, the protection becomes a simpler limiting control problem, where a simple comparator circuit in hardware or comparison operators in the microcontroller code does the task. If the current flowing into the battery (or the load) increases beyond a pre-set limit, the designer can either choose to shut down the charging supply or reduce the impressed voltage to keep the current flowing within a limit.

Most switching power supplies provide overload protection (which is basically over-current protection), whereby the power supply shuts down the output for some time and then rechecks by starting the supply again. In the world of power supplies, this is called the hiccup mode.

In the worst case, the hiccup mode, with its repeated start-stop cycles, can cause permanent damage to some of the control electronic elements, such as op-amps and other control ICs. If the power is only being used for battery charging, then this might be alright. In other cases, where the power supply is shared by other sub-systems, it is best to use a linear current limiting provision of some kind to ensure longevity of the control electronics.

Circuit Specifications

The idea was actually requested by one of the dedicated readers of this blog, Mr. Saurav, as explained below:

Looking for some ideas/help/suggestions. I have installed a 2.2 kw off grid solar system, using loom solar panels, excide battery and excide solar inverter. The inverter has this pre-setup priority, first solar, then grid, last battery. I have disconnected the mains supply to the inverter, so for me it is solar then battery. To this overall setup, I have added an ACCL with grid as secondary.

So in the evening, whenever there is no solar and the battery is out of charge, it falls back to grid power.

This setup has one problem. ACCL switches to mains power at night, when the battery is completely drained out or deeply discharged and that’s what I don’t want.

I want to turn off the battery power, when the battery has 20% remaining power or the battery is at a certain voltage. That way battery life can be better.Is this something doable? Do we have something readily available for this? Or do we need to build something for this?

The Design

The circuit design for the proposed battery deep discharge protection circuit can be witnessed in the following diagram:

As can be seen, the circuit has a very components, and its working can be understood through the following points:

There are a couple of power transistors coupled with each other where, the base of the TIP36 transistor forms the collector load of the TIP122 transistors.

The base of TIP122 is biased through a resistor/zener diode network, where the zener diode ZY determines the cut off voltage for the TIP122.

The zener diode voltage is selected such that it matches the critical low voltage value of the battery, or any value at which the draining of the battery by the load is required to be stopped.

As long as the battery voltage stays above the zener voltage, or the voltage at which the cut-off needs to happen, the zener diode keeps conducting which in turn keeps the TIP122 in the conducting mode.

With TIP122 conducting the TIP36 gets the required base current, and it also conducts and allows the battery current to pass to the load.

However, the moment the battery voltages reaches or drops below the zener voltage which is also the deep discharge voltage level, causes the zener diode to stop conducting.

When the zener diode stops conducting, the TIP122 base voltage is cut off and it switches OFF.

With TIP122 now switched OFF, the TIP36 is unable to get its base bias current, and it also switches OFF turning off the battery current to the load.

The procedure effectively prevents the battery from further draining and depleting below its deep discharge level.

The indicated load can be any specified load, such as an inverter, a motor, an LED lamp etc.

Using a MOSFET

The indicated TIP36 can supply a maximum current of 10 amps to the load. For higher current, the TIP36 could be replaced with a P-Channel MOSFET such as the MTP50P03HDL, which is rated to handle at least 30 amp current.

When a MOSFET is used in place of the BJT TIP36, the 50 ohms resistor can be replaced with a 1K resistor or a 10K resistor, and the TIP122 can be replaced with a BC547.

Battery Over Current Protection

Protection is provided using the two control pins OD and OC (which stand for over discharging and over charging respectively). These two controls attach to the gates of two MOSFETS and stop current flow to the battery if there is a problem.

Overcharge protection voltage: 4.3V (typ) ± 50mVOvercharge release voltage: 4.1V (typ) ± 50mVOverdischarge protection voltage: 2.4V (typ) ± 100mVOverdischarge release voltage: 3.0V (typ) ± 100mV

Overcurrent detection voltage: 0.15 (typ) ± 30mVShort circuit detection voltage: 1.35V (typ).

Note: The above over current and short circuit voltages are measured across the on-resistance of the MOSFET.

DW01A Datasheet

The crucial part of the DW01A operation is the controlled dual MOSFET (N Channel); Specifically the RDS(ON) resistance of the N Channel MOSFET.

In the datasheet it states that the threshold current for overcurrent detection is determined by the turn-on resistance of the charge and discharge control MOSFETs.

There are problems in designing a current detection device this way, because as it also says in the datasheet:

turn-on resistance of the MOSFET changes with temperature variation due to heat dissipation, It changes with the voltage between gate and source as well

It is designed this way because it is a very cheap method.

  • The threshold voltage levels fall outside the normal charging voltages of a battery charger, and so do not interfere with the normal charging process.
  • The exact short circuit current value does not matter (as long as it is reasonable i.e. not 100A! it can be made to be 3A (see calculations below).
  • The current limit reduces as the MOSFET’s selected RDS(ON) gets worse. this is a good; If you use a high RDS(ON) value, the current needed to trigger the short circuit is smaller. Also increasing temperature increases RDS(ON).

This is a fail safe device, so as long as the values chosen fall outside the normal operating state of the charging battery, it will provide short circuit protection even if the exact charging-cut off value changes with temperature and voltage.

You should simulate, analyse and test the MOSFET operation to make sure it is acceptable for your application.

Dual MOSFET N Channel Datasheet

Using RDSON as the Current Limit

The current limiting voltage threshold is detected by a comparator, when the voltage at the CS pin reaches 150mV. The comparator voltage will be reached when the voltage drop across the resistance of the two (switched on) MOSFETs reaches 150mV. this is caused by more current flowing through the two MOSFETS and is therefore the voltage drop across 2 x RDS(ON). You can find RDS(ON) values in Figure 6 of the datasheet for the 8205, which is labelled Rdson On-Resistance(mΩ) vs ID- Drain Current (A) and shows the curves for various Vgs values.

Since the battery voltage is close to 4.5V using that curve gives RDS(ON) as 20m⦠which results in a short circuit current of 3.75A (0.15/(220e-3)).

When the battery discharges it will be closer to 2.5V giving RDS(ON)as 25m⦠resulting in a short circuit current of 3A (0.15/(225e-3)). Once triggered, the DW01A the discharge MOSFET (OD) is turned off. It is only released when the load is removed.

Two Overcurrent Threshold levels

There are two over discharge values (the one above) 150mV and 1.35V. The reason for the second one is that both are associated with activation delays. For the 150mV one the delay is 10ms, while for the second the delay is 5us.

So for an extremely large short circuit the activation delay is much faster.

When the short circuit current detector has been activated, you, must remove the load, before the DW01A allows current to flow again (OC MOSFET turned on).

Does Size Matter (or is it a tight fit)?

Since this is a family site, I’m assuming you’re referring to the length of the battery after the circuit is added. Good question! And in SOME circumstances it does.

Both these protected cells are 16340 size. The bottom one is wrapped.

The added electronics will stack an extra millimeter or two onto the length of the cell. If being used in LED flashlights, most manufacturers design the tubes to accommodate variable sizes depending on the type of battery used. Where length seems to vary more, is with 18650 batteries. Some users have found some batteries to be a tight fit, both in length and width. I own a couple like that too. But it’s usually only with ONE particular flashlight (for instance). When I switch to a different light, the cell fits perfectly.

What About Unprotected Batteries?

I’m glad you asked! You can certainly opt to buy those if that floats your boat. It’s all a matter of safety. Unprotected cells will work exactly the same way. You just need to be EXTRA careful and be diligent about checking voltage during usage.

As an alternative, this article describes another type of lithium chemistry (IMR) that is not only more commonly found in unprotected form, but is actually safer to use.

Questions about protected Lithium-ion batteries. Ask away, and you shall receive the best answer possible!

This link features an amazing selection of both protected and unprotected batteries at excellent prices!

battery, protection, circuit, schematic, charging, discharging

Learn About 18650 Lithium-ion Batteries As far as LED flashlights go, most users will tell you that a light which…

Things You Should Know About Lithium-ion Batteries Batteries are an excellent source of portable power, and are literally everywhere. Since the invention…

Different Types of Rechargeable Batteries Rechargeable batteries are a good investment, for whatever the devise they might be intended for.…


As an input connector, I chose micro USB – it is very popular and you can use a jellybean phone charger to supply 5 V (however, it must be capable of supplying the required current – 1 amp for charging and 1 amp for the output!). I also included two pads, 5 V and GND, so you can solder wires to these directly, if you want. And also I routed out the D and D- pins from the USB connector, so if you for example want to provide power and also use the same USB to transfer data, you can solder to these pads. Just keep in mind that the D and D- are differential, so you need to keep wire lengths the same (and ideally use a shielded, twisted pair for longer cables).

The charging is realized using the very popular TP4056, which offers CC-CV charging with trickle charge. The battery is charged to 4.2 V with a user-adjustable current up to 1 amp. The charging current is determined by R3 according to the formula I = 1200 / R3. The datasheet states absolute maximum for charging current as 1.2 amps.

The TP4056 provides two open-drain outputs for status LEDs. These are routed out in such a way that you can either solder in a 1206 SMD LED, legs of a normal 5 mm LED, which can then be bent sideways, or cables, which can then be connected to a microcontroller (then you need to set up your micros pins as inputs with pullups). And lastly, there is a thermal pad beneath the TP4056 – do not forget to put some thermal grease before you solder in the IC.

Protection and load sharing

The protection itself is realized using two ICs – the first one is a dual MOSFET, I used the 8205A, which uses TSSOP8 package (if you are ordering these, do not mistake them with the 8205S, which are SOT23-6 package). The seconds IC is the protection controller – you can either use the DW01, which is more common, but the overdischarge voltage is 2.4 V, which I think is too low for me. Therefore I prefer to use the FS312, which is exactly the same, just the overdischarge voltage is set to 2.9 V, which is more suitable for the average lithium battery.

Note – calculating the overcurrent/short circuit current: the protection IC uses both MOSFETs in series as a sense resistor. And since their resistance can vary quite a bit, it is impossible to determine the exact trip current. But according to my calculations, with DW01 and 8205A it is around 2.5 A.

Note – starting the protection circuitry: when you first connect a battery, you need to “enable” the protection circuitry by plugging in the USB charging for just a brief moment (literally half a second will do).

The load balancing circuitry is also very important – it would be impossible to charge the battery and use the connected device at the same time (since the charging process depends on measuring battery current, if you tried drawing current from the battery at the same time, you might get it stuck in a CC loop and damage the battery). So the load balancing circuitry makes a low resistance automatic switch, which will basically power the boost converter from the 5 V source instead of the battery when it is present. So the 5 V input will be divided between the battery charging IC and the boost converter section. If you are interested to see how the load sharing part works, see my GitHub, there is a folder with a few pictures as an explanation.

Boost converter

I used the MT3608 boost IC, since I had good experience with it before. In this revision, I highly optimized the layout of all the switching components, so the efficiency is usually close to 80 % (see below), which is pretty good for something this simple and cheap. My goal was to achieve 800 mA output across the whole Li-Ion battery voltage range (e. g. 3.0 – 4.2 V), and I did succeed. And with voltages above circa 3.3 V, it can even output stable 1 amp. In fact it can output 1 amp across the whole Li-Ion battery voltage range, but the protection circuitry will kick in and prevent you from drawing more than 2.5 amps from the battery.

Additional sources

When I published the previous article about this, I got quite a few people ask me if I am willing to sell these boards. Since I ordered V1.2 boards and parts in large numbers, I am willing to sell a few of these – assembled and tested or just parts, depends on you. Here’s a link.

  • Eagle drawings (schematics and PCB) and partlist on GitHub
  • original Instructables article
  • EEVBlog – charging lithium batteries

Leave a Comment