We will discuss here the three main contributors of power losses of the MOSFET when used in switching converter or switch mode power supply. The main MOSFET efficiency factors are drain to source on state resistance, gate charge and the equivalent output capacitance. Drain to source on state resistance or simply RDSon has a positive temperature coefficient (increases with temperature rise). These factors will be discussed in detailed below, so keep reading.
MOSFET Efficiency Factors
1. RDSon is the Dominant Efficiency Factor for MOSFET
This is the equivalent drain to source resistance of the MOSFET. This will vary with temperature. Say at 25’C ambient temperature the RDSon is 100mΩ but when the operating temperature increase to 100’C, the RDSon may become 130mΩ or higher. Figure 1 is a graph of a junction temperature versus the normalized RDSon of a particular MOSFET.
As you can see in Figure 1, the RDSon of a MOSFET will increase with temperature rise. By the way, do not be mistaken in the y-axis of the graph. It is not an absolute RDSon value but rather a multiplier to the initial RDSon. Let say the initial RDSon specified is 10mΩ, then at 100’C the resulting RDSon will be 1.3*10mΩ which is equal to 13mΩ.
The smaller the RDSon of a MOSFET, the better since conduction loss is just the product of the square of the current and RDSon as below.
Conduction loss is the power dissipation of the MOSFET when it is conducting statically. Statically means DC operation or continuous operation (no switching or transitions between on and off states).
Yes the smaller the RDS the lower the conduction loss. However, for switching converters, conduction loss is just a one piece of the puzzle. So don’t be fascinated right away for the lower RDSon.
2. Gate Charge (Qg) of MOSFET Contributes Input Losses
Another parameter that is one if the main MOSFET efficiency factors is the gate charge. In order to turn on a MOSFET, you need to provide a voltage across its gate-source terminals. This applied gate-source voltage will not right away meet the threshold level since the input capacitance or the gate capacitance must be charged first.
The larger the input capacitance, the longer the time needed before the MOSFET operates in the hard saturation. Now, what is gate charge? From coulomb’s law, charge Q is equal to the product of capacitance and voltage (Q=CV). Gate charge is coined from this equation and it is widely used by MOSFET vendors in their datasheet.
The energy stored in the input capacitance is can be defined as
Energy is the product of power and time (E=P*t). We can modify the above equation as below
Some design engineers do not include the constant one half in the power dissipation formula to consider the worst. Actually the constant one half is given already from the energy equation of a capacitor. If you are going to derive the energy of a capacitor, you will end up in integrating the area under the intersection of the current and the voltage.
The area of intersection is actually forming a triangular shape that is why there is a one half factor. In our analysis above we consider that the energy store in the input capacitor is the one that will be transferred to the MOSFET itself that is why we use the energy equation of the capacitor in deriving the power loss.
Another analysis to get the power loss due to gate charge is by using the charge equation in terms of the product of current and time as below.
So, the power loss due to gate charge is dependent to the switching frequency, applied gate-source voltage and the gate charge itself. The higher the gate charge is, the higher is the loss.
3. MOSFET Output Capacitance (COSS)
This is the effective output capacitance of the MOSFET. This is dependent to the drain voltage (dynamically). For the input capacitor, the power loss is generated during turning on of the MOSFET while for the COSS, the loss is generated during turn off of the MOSFET. We can start the derivation of the power loss by considering the energy of the capacitor or the COSS.
From the equation, the power loss due to COSS is also dependent to the switching frequency, drain-source voltage and COSS itself. The higher the COSS of the MOSFET, the higher is the power loss.
Another contributor to switching converter losses is the rise time and the fall time of the MOSFET. This parameter is not as significant as the three factors above. I will leave this to you to derive the equation for this.
To conclude this topic, there are two types of losses in a switching converter; conduction loss and the switching losses. Conduction loss is dominated by the RDSon of the device while switching losses are governed by gate charge and COSS. In selecting power MOSFET for switching applications, you may do analysis in your design if what is dominant; the conduction or the switching losses. If the former dominates, prioritize RDSon in your selection. If otherwise, select a device with lower gate charge and COSS even though the RDSon is little bit higher.
Once again, I say this to you; there is no substitute for actual testing. Above explanations and computation is just reference for you to have a starting point in your device selection. Actual result may tell differently because there are other factors that we did not considered in our analysis. Cheers!