Overview of Conduction Modes in a DC/DC Converter
If you look around the internet and in application notes for switching DC/DC converter example designs, they are almost always presented as operating in continuous conduction mode (CCM). What exactly is this operating condition, and how is it affected by component selection? The answer is a bit complex and depends on the switching converter topology, but there is a simple process that is used to determine the required component values that will ensure continuous conduction. Outside of continuous conduction (critical or discontinuous conduction), the converter can exhibit somewhat different behavior, which might be undesirable.
In this article, we’ll look at how the discontinuous conduction mode can arise and how component values + parasitics can play a role.
What Are DC/DC Converter Conduction Modes?
DC/DC converter conduction modes refers to the switching action exhibited by the inductor current in the converter’s switching stage. There are three conduction modes
- Continuous conduction mode: the inductor in the switching regulator will always have non-zero current.
- Critical conduction mode: the inductor current switches ON as soon as the inductor current drops to zero. There is a specific duty cycle that is needed to cause this.
- Discontinuous conduction mode: the inductor current drops to zero and stays at zero until the next pulse in the PWM signal. This can happen if the duty cycle is too short or the load is too small.
Most designs you’ll find online from semiconductor vendors are intended to operate in the continuous conduction mode. The same ideas and switching behavior of inductor currents can be found in other power circuits, namely a PFC circuit.
The two other modes are not necessarily bad or undesirable. In fact, in discontinuous mode, you can have stable regulation at the required output voltage, and you will have lower switching losses because the switching FET will be off for some period. However, because of the discontinuous current in the inductive section, there can be some additional noise compared to the continuous conduction mode case.
To keep a switching converter operating in the continuous conduction mode, you need to select components so that the inductance and capacitance in the output section are above some critical value. If these values are too small, the inductor current will quickly fall to zero. The diode or rectifier forward voltage must also be low enough to allow forward conduction during switching at the required inductor current, otherwise your diode will block conduction even when forward biased.
Conduction modes in a switching DC/DC converter.
What can happen when we are not in continuous conduction mode? And how to parasitics play a role? Let’s look at an example.
Component Parasitics Create Oscillations
One problem with the above process is that it can be difficult to determine the real behavior of the switching regulator when parasitics are present. The dynamic behavior of the output voltage from a switching regulator can be examined in simple simulations with a theoretical FET driver. An example buck converter is shown below.
This buck converter can operate in the discontinuous mode, and there will be component parasitics that influence the output voltage.
As we will see, this converter can operate in the discontinuous mode. The components selected above will also have parasitics that produce deviations from ideal behavior. In real converters, it is possible to observe oscillations due to ESL in the capacitors, parasitic inductance in the dI/dt loop, and the winding capacitance in the inductors. These components have self-resonant frequencies that correspond to an oscillation in the transient response.
Oscillations can be observed in the transient response in two instances:
- During switching, where an oscillation can be observed on the output voltage or inductor current
- During startup, where the output voltage suddenly rises from zero to its nominal DC value
If we just look at the case with a small inductor (3 uH) and low load of 10 Ω, we can see clear oscillation in the inductor current with 200 mA peak amplitude on the falling edge. The turn-on portion of the circuit is very severe, with huge overshoot reaching -17.6 A. In these conditions, the converter is operating in the discontinuous mode.
In discontinuous mode, there is an oscillation on the output due to the transient response in the LC section during switching.
In discontinuous mode, when the inductor current hits zero, the output can exhibit an underdamped oscillation. Just as is the case with output filtering, the simplest solution is to add some small amount of resistance (approximately 1 Ohm) on the output filtering capacitor. You can spot this behavior in a transient analysis simulation.
Now if we include parasitics in the simulation, we can see additional transient behavior. We can see both effects in the example below. When the winding capacitance (1 nF) and capacitor ESL value (10 nH) are included in the simulation, there are additional underdamped oscillations on the output voltage and the inductor current. The parasitic inductance in the capacitor accounts for two contributions: trace inductance, which is typically 5-10 nH/inch, as well as the ESL value, which is typically ~1 nH when vias and pads are considered.
Parasitics create additional transient behavior.
From these examples, we have an important conclusion: it’s critical to understand the parasitics in the circuit as these can cause the design to operate in an undesirable mode (discontinuous conduction) and with excessive oscillatory behavior on the output during switching.
More Complex Topologies
Switching regulators that go beyond the standard buck, boost, or buck-boost circuits with a single element or a half-bridge switching stage can have more complex formulas for the minimum component values. If you intend to operate in continuous conduction mode, these formulas are an appropriate place to start to get some idea of minimum component values, but you should still simulate your regulator with an idealized load. Parameter sweeps with the component values will help you determine when the converter is operating in continuous conduction mode.
Everything we’ve talked about so far has focused on switching regulators, but the same ideas apply in bridge circuits for motor control, where a PWM signal is being used to modulate the current sent to motor windings. The only difference is we do not use a capacitor to smooth the output to a nominal DC value, we simply allow the PWM signal to turn fully off and the current to periodically fall to zero. Oscillations in motor driver circuits arise for the same reasons as in switching regulators: it is due to parasitics exciting underdamped transients as discussed above. When designing these circuits, it’s important to use transient analysis simulations to spot these oscillations.
When you need to perform simulations for evaluating power electronics circuits, use the best simulation and modeling features found in PSpice, the industry’s best circuit design and simulation software 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.