Skip to main content

Why Does My MOSFET Oscillate?

MOSFET oscillation

MOSFETs are some of the most important and ubiquitous devices in electronics, both in integrated circuits and as discrete components. The usage of a MOSFET is very simple: they act as a fast switch that can be triggered by applying a voltage to the gate terminal, and the gate voltage can be at logic levels. This allows very high gain to be achieved when the device is modulated ON, and these devices can switch quickly for applications like pulse drivers.

Anytime there is a fast switching event that drives a MOSFET, there is the possibility of exciting an underdamped oscillation in the load circuitry. This would be more commonly known as ringing, and it is desirable to prevent this from happening in many circuits. To identify potential MOSFET oscillations, there are many parasitic elements in a PCB layout and component mounting method that have to be identified. We’ll illustrate simple ways that parasitic oscillations are prevented in this guide.

MOSFET Ringing and Oscillation

When a circuit is being driven with a MOSFET, and the MOSFET is quickly switched ON with a pulse applied to the gate terminal, there are two possible responses.

  • Ringing
  • Oscillation

Ringing is the initial transient response, and it will decay to zero due to passive resistance in the MOSFET circuit. Just like in an RLC circuit or one of the standard oscillator circuits, a circuit that is being driven by a MOSFET can exhibit oscillations, and the type or frequency of the oscillation can vary depending on the type of load in the circuit.

All MOSFETs designed for power delivery will have three factors in common that will contribute to unintentional oscillations when the MOSFET and its load circuit are excited:

  • Large transconductance gain
  • Parasitic capacitance
  • Parasitic inductance

The parasitic inductance in the MOSFET typically comes from the component leads and the traces connected to the circuit, and the interconnections between terminals can create a circuit with feedback. When this interacts with capacitances in the load or MOSFET body, the result is a parasitic oscillation. Such an oscillation can have large overshoot, which would damage the load, the MOSFET, or both.

Driven Oscillations

Power MOSFETs can be driven in AC or DC, with the typical AC driving being a PWM signal or an AC sinusoid. In general, however, any periodic AC waveform (even a modulated waveform) is appropriate for driving a MOSFET. The table below describes how driven oscillations result in MOSFET circuits and how they can coexist with ringing.

Driving Method

Response Type

Description

Fast pulse

Ringing

Transient response with decaying oscillation

PWM pulse stream

Ringing

Repeated transient excitation on edges with DC delivery on average

Continuous AC wave

Driven oscillation

There will be an initial transient response, but this decays and the response settles to an AC wave

Sigma-delta modulated pulse stream

Ringing + oscillation

Decaying transient response on edges, but sinusoidal power delivery on average; this is very similar to a simple PWM signal

When a power MOSFET is being driven in a steady state (purely an AC sinusoid or DC driving), then any initial oscillation will decay and will not persist. In these cases, the transconductance approaches zero and the MOSFET conducts according to its load line characteristics.

When Does Parasitic Oscillation Result?

Parasitic oscillation results from unintentional driving through positive feedback, and this occurs due to parasitics in the MOSFET component. A power MOSFET is susceptible to parasitic oscillation when the load current suddenly changes (switching or short circuit), or during the transient switching period where the transconductance is increasing quickly. Oscillations that are kicked off by a switching event will also decay slowly, so they look like ringing, but they can be sustained when the excitation is repeated.

MOSFET waveform

Parasitic oscillations observed in a MOSFET driver circuit due to positive feedback. [Source: Rohm]

Types of Stability

A better way to describe what is happening in a MOSFET circuit is through linear stability analysis. This would be performed by looking at the transfer function of the circuit being driven with a MOSFET.

Note that this is still an LTI problem; we do not have an issue of nonlinear stability here, so the transfer function approach is still appropriate. In other words, the MOSFET’s gain is just ratio, it is not an increasing function of the power delivered to the load. Therefore, there are three types of stability observed in MOSFETs:

  1. Unstable decaying orbit (driven with a gate pulse)
  2. Stable limit cycles with transient rise/fall (driven with sinusoidal gate signal)
  3. Stable limit cycles that are a linear combination of two or more unique stable limit cycles (composite sinusoidal signals)
  4. Stable limit cycles with superimposed decaying orbit limit cycles (sigma-delta modulated signal)

Point #1 above describes ringing, while points #2 to #4 above each describe a driven oscillation as the voltage applied to the gate changes. The transfer function, and in particular the poles in the transfer function, can be determined using pole-zero analysis in a circuit simulation program like SPICE.

When you’re ready to create and simulate your MOSFET-driven circuits and identify oscillations, you can design and simulate your circuits with the simulation tools in PSpice from Cadence. PSpice users can access a powerful SPICE simulator as well as specialty design capabilities like model creation, graphing and analysis tools, and much more.

Subscribe to our newsletter for the latest updates. If you’re looking to learn more about how Cadence has the solution for you, talk to our team of experts.