How to Compute MOSFET Switching Losses

There are two types of losses in MOSFET. One is conduction loss and the other one is switching loss. Conduction loss is explained in detailed here. Switching loss is not as easy as computing the conduction loss. There are many parameters to consider. You need to understand these parameters to select the correct values.  However, you don’t need any more to be an expert in power electronics as I will guide you on how to compute MOSFET switching losses.

MOSFET switching loss is a complicated thing. Switching losses are contributors to overall power dissipation in a continuously switching circuit like SMPS, DCDC converters, chargers, and inverters to name a few. Switching losses are not happening at static condition like when the MOSFET is conducting. It is the losses associated during turn on and turn off the MOSFET. Design engineers’ techniques will vary from each other. In this article, let me share with you the technique I used on how to compute MOSFET switching losses.

What are the things to Consider on How to Compute MOSFET Switching Losses

How to Compute MOSFET Switching Loss due to Gate Charge

The energy needed to charge the gate capacitor is given below

Energy = ½ C x V2

Energy is also can be expressed as

Energy = Power / Frequency

Equating both equations will yield

Power / Frequency = ½ C x V2

Charge of a capacitor could be expressed as

Charge, Q = C x V

Expressing C

C = Charge, Q / V

Then substitute C to the equated equations above to give the gate charge power loss

Power loss gate charge = ½ x Qtotal x V x Fsw

Where:

Qtotal is the total gate charge specified in the MOSFET datasheet

V is the applied VGS (gate to source voltage)

Fsw is the switching frequency

Below table is how IRFR320TRPBF-BE3 (MOSFET from Vishay) datasheet specified the total gate charge. Vishay is offering great options for MOSFETs that sit your design. Visit them at https://www.vishay.com/.

For worst case power loss, consider using the maximum value. For typical value, use the typical value instead. Sometimes datasheets will only indicate the maximum values though.

How to Compute MOSFET Switching Losses - Total gate charge

Gate charge loss is a function of voltage (VGS) and switching frequency (Fsw). This simply means that the applied VGS could not be exaggerated as it will result in higher gate loss. To guide you on what must be the proper VGS to apply, read How to Determine the Correct MOSFET VGS Threshold.

In the same manner, the switching frequency must be selected such that the gate loss is not too much that it will jeopardize the system efficiency and the MOSFET health.

How to Compute MOSFET Switching Loss due to Output Capacitance

The COSS will charge the moment the MOSFET is turned off. The energy needed during this time will cause power loss.

The energy of a capacitor is given in below equation

Energy = ½ C x V2

Energy is also can be expressed as

Energy = Power / Fsw

Both equations when equated will yield

Power / Fsw = ½ C x V2

Re-arranging the equation will give

Power loss output capacitance = ½ x COSS x V2 x Fsw

Where:

COSS is the output capacitance specified in the MOSFET datasheet

V is the drain to source voltage

Fsw is the switching frequency

The maximum power loss due to the output capacitance is maximum at the maximum value of the drain to source voltage. The choice of switching frequency is also a concern to minimize this switching loss.

The IRFR320TRPBF-BE3 MOSFET datasheet specified the COSS as below table.

How to Compute MOSFET Switching Losses - output capacitance

Switching Loss due to Rise and Fall Times

The last thing to consider on how to compute MOSFET switching loss that I want to share is the one due to rise time and fall time. For some, this is analogous to the COSS loss thus no need to consider separately. In my designs, I still consider this as separate source of switching loss though.

Below is the equation of this switching loss.

Power loss rise and fall times = ½ x (trise + tfall) x Irms x V x Fsw

Where:

trise is the rise time specified in the MOSFET datasheet

tfall is the fall time specified in the MOSFET datasheet

Irms is the rms drain current of the MOSFET. This is a circuit value.

V is the drain to source voltage

Fsw is the switching frequency

Below is the rise and fall times specification of the sample MOSFET.

How to Compute MOSFET Switching Losses - rise and fall times

The total switching loss is the sum of all the individual losses. If the datasheet provides the typical values of each parameter needed, use it if your goal is to derive the typical switching loss value. On the other hand, If the objective is to get the maximum switching loss, the maximum parameters in the datasheet would be selected.

Summarizing How to Compute MOSFET Switching Losses

I hope this article can give insights and guidance when you start your own design. Summarizing here the key points to consider on how to compute MOSFET switching losses.

1. Switching loss is happening during the time the MOSFET is either switching to on or off state.

2. Switching losses are dependent on the switching frequency.

3. Gate loss is due to the total gate charge. It is dependent on the VGS aside from the switching frequency.

4. Output capacitance loss is due to the COSS. The maximum COSS loss will happen at the maximum drain to source voltage.

5. The total switching loss of a MOSFET is the sum of the total gate charge loss, COSS loss and the loss due to rise and fall times.

6. The parameters needed on how to compute MOSFET switching losses are total gate charge, COSS, rise and fall times which are specified in the datasheet under dynamic characteristics.

Please do not forget to hit the share and like button. Should you have any comments, suggestions or questions to improve this article, feel free to leave comments. Let’s discuss! Electronics believer is also offering consultations. Kindly email to [email protected].

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.