Batteries have become one of the most important devices in modern society with the rise in demand for portable devices, smartphones becoming ubiquitous, and even cars shifting towards battery technology. Batteries are also very important in satellites, storing energy to power the system when solar energy is not available.
Re-chargeable batteries can be charged by forcing an electric current through it. Different types of batteries have different properties when it comes to recharging so this blog will be focusing on lithium ion batteries (4.2V) as BLUEsat will be using these, and because it is one of the most common batteries.
Li-ion batteries have an intercalated lithium compound as electrodes, discharging when Li-ions move from the negative electrode to the positive electrode. Charging forces the ions back onto the negative electrode. Unfortunately, every time ions are shifted some ions react with the electrodes, decreasing the number of free ions in the battery. Overcharging and overheating will also severely reduced battery life and potential damage to the battery. A carefully designed charging process must be used to maintain maximum battery life and battery condition.
Charging a Li-ion battery occurs in two steps:
- Constant current: the charger applies constant current to the battery. The voltage applied increases steadily towards the voltage limit of the battery. Current should be kept between 0.5 and 1 C to prevent high temperatures.
- Constant voltage: the charger applies a voltage equal to the voltage limit of the battery, as current declines.
The constant current stage quickly charges the battery until battery voltage is about 4V. The charger switches to the constant voltage stage to prevent overcharging. The constant voltage is maintained until current falls to around 0.1C and then should be terminated when the battery reaches full charge. Most modern chargers can detect full charge and stop at this point.
If more than one cell is being charged, then the batteries need to be balanced to the same level of charge before the constant voltage phase. This can be achieved through using a balancing circuit. We are charging our batteries as a single unit at 8.4V.
Thien’s battery charging board
Thien, a former member (and ex-president) of BLUEsat, developed this battery charging board a few years ago, but it had not been tested until now. Although the board will be modified slightly – to fix any problems with the circuit and to make it compatible with the new battery unit – the way it’s supposed to work remains the same.
The circuit is built around the LTC4012-1, a battery charger controller chip. In the circuit diagram below, you can see:
- Connection points for the power source (Vin, top left), battery unit (VBAT, bottom right) and load (Vsup, top right)
- The LTC4012-1 chip (that whole yellow rectangle)
- Various resistors and capacitors
- MOSFETs (Q1, Q2 and Q3) and an inductor (L1)
- A diode for the BOOST supply
Let’s take a closer look at each of those categories…
These are the light green screw terminals on the actual circuit board. There are actually two connections on each block because one of them is ground.
Vin would be connected to the solar panels, either directly or, in the case of our CubeSat, through a MPPT converter (see the previous BLUEsat blog post on Maximum Power Point Tracking). VBAT is the connection to our battery unit, which is made up of two lithium-ion batteries in series. Vsup is the output that supplies power, from either the solar panels or the batteries, to the other parts of the satellite.
This chip has many pins, for voltage/current sensing, indicator outputs, forcing shutdown, supplying power to other pins, and more. We will elaborate on its function in the next section of this blog post.
Resistors and Capacitors
The LTC4012-1 can be programmed to control battery charging at a range of voltages and currents, not just the ones suitable for our particular battery unit. To select the required charging voltage and current, we have to connect appropriate resistors to certain pins. The datasheet for the LTC4012-1 provides a helpful circuit diagram for a typical application of the chip, along with suggestions for capacitor values and tables to look up the resistances we need.
The MOSFET (metal oxide semiconductor field effect transistor) is a device that basically acts like a switch. It can either block or allow current flowing from the drain terminal to the source terminal, depending on the voltage at the gate (compared to the source). Note that the symbol used in Thien’s circuit diagram includes a diode between the source and drain, indicating that the MOSFET still allows current to flow from the source to drain.
There are different types of MOSFET that can be made. For example, Q1 and Q3 are both enhancement mode MOSFETs which means that the switch is normally off. The directions of the arrows show that Q1 is n-channel, requiring a positive gate voltage to turn it on, while Q3 is p-channel, requiring a negative gate voltage compared to the source.
In Thien’s circuit, Q3 is used to maintain a certain voltage needed between two pins of the chip (DCIN and CLP), and to prevent current flowing backwards from the batteries to the solar panels. Q1 and Q2 are controlled by logic inside the LTC4012-1 to create a PWM waveform that is smoothed out by the inductor (L1). This means that the charging voltage and current won’t be affected by Vin being higher than required.
Instead of SW, the diode was meant to be connected from INTVDD to BOOST, as shown below:Together with the inside of the chip and charge on the capacitors, this part of the circuit generates a voltage even higher than the chip’s internal 5V supply. This voltage would be needed at the gate of Q1 to turn on the MOSFET when its source terminal is already at a high voltage.
This chip is the one we are using for our battery charging circuit and it provides a good introduction into how lithium-ion battery charging is actually achieved. It is a battery charging chip which is able to provide the constant current and constant voltage features necessary for charging lithium-ion batteries. This chip is also designed specifically for lithium ion batteries as it features precision internal resistors to supply 4.1V per cell. (The LTC 4012-2 provides 4.2V per cell).
The chip achieves constant current by quickly switching on and off the connection to the power supply and averaging the current over time at the desired limit. In constant voltage mode, the switching is intended to maintain a steady voltage rather than a steady current. The current going in to the batteries is determined by the batteries’ own capacity to take in current at this point, and the chip does not directly reduce the current flow anymore.
The MOSFETS in Figure 1 are both n-channel MOSFETs, corresponding to Q1 and Q2 in our circuit. The two MOSFETs are controlled by the two pins, TGATE, and BGATE, which are switched on and off quickly. The FETs are alternately switched on, so that when the top FET is ON, the bottom FET is OFF, and vice versa to ensure that the system power supply Vsup is never shorted to ground.
When the top FET is ON and bottom FET is OFF, the system supplies power to the inductor (L1) and the batteries, which increases the current flowing to the batteries (ICHRG). When the FETs are switched (the bottom FET is ON, and top FET is OFF), power is no longer supplied to the inductor (L1). As inductors resist the change in current, the inductor starts discharging and will attempt to keep the current ICHRG relatively constant.
The switching of the FETs is controlled by pulse width modulation (PWM), shown in Figure 2. While the top FET is turned ON, ICHRG is constantly increasing. The value of ICHRG is measured through the voltage drop across the sensing resistor VSENSE. When this voltage is higher than a programmed value, the FETs are switched for a fixed period of time to discharge the inductor, before switching to charge the inductor again via the system power supply.
While in constant current mode, the chip sends an active signal on the Input Current Limiter pin (ICL).
In voltage charging, the switching of the top FET and the bottom FET are no longer coupled with ICHRG. They are instead controlled by the voltage as measured from VBAT, which is then compared with the value programmed by the pins FVS0 and FVS1. While this voltage is kept constant, the current ICHRG is determined by the amount of current the batteries can absorb, which in turn is dependent on their state of charge. As the batteries approach full capacity, the amount of current they absorb drops towards zero. The chip has no inherent ability to shut off the current going into the batteries. However, it is able to send a signal on its CHRG pin to indicate when current has fallen to a tenth of programmed value. This will let us know that the lithium ion batteries have finished charging.