We will make comparisons between these three conduction modes, identify where or when they are best suited, and provide practical ADI circuit examples in this 6th and final part of the series, focusing on DCM and CCM in SMPS circuits.
Generally, discontinuous mode is used under light load conditions and then transitioned to continuous mode when the output current increases. The output current at which the change from CCM to DCM occurs differs for each product and depends on the power supply specifications, inductance, and IC control method as detailed in the previous parts of this series.
In addition, products that employ forced PWM control (even during PWM control) operate in CCM without changing to DCM even during light loads.
DCM and CCM Basic Comparisons
Figure 1: DCM and CCM Basic Comparisons
Switch-mode power supplies (SMPS) like buck, boost, inverter, and flyback circuits have the flexibility to operate in either DCM or CCM. These modes are influenced by various factors such as the inductor size, load current, and duty cycle. Changes in these operational conditions can cause the circuit to switch between DCM and CCM.
The designer can ensure the circuit remains in a desired mode. For instance, by selecting a larger inductance (L) than what is needed for the application, the designer can ensure that the inductor remains charged throughout the operation, thus maintaining the circuit in CCM.
Some circuits have a control circuit that is designed to lock the operation into a specific mode, whether CCM or DCM. These are referred to as forced continuous conduction mode (FCCM) or forced discontinuous conduction mode (FDCM), respectively.
CCM or DCM
Recognizing or detecting when and where CCM or DCM occurs in a circuit is one thing, but influencing one mode versus the other is another challenge. We have seen a circuit originally designed or tuned to operate in CCM can fall in DCM and vice-versa when conditions (such as Vin, switching frequency, load current) change, with the consequences of Vout not behaving as expected.
How do we ensure that the circuit designed in one mode remains in that mode, regardless of changes in conditions? Various methods exist.
Ensuring Conduction Mode by Over-/Undersizing Inductance Value
For a simple buck converter, as discussed in part 4, that CCM, DCM, or BCM can be ensured when the inductor value is made 2x or 3x from the critical value Lcrit:
Figure 2: Forced DCM or CCM by Over-/Undersizing L Value
Lcrit =R*(Vin-Vout)*d*T/2 Vout
* If L >> Lcrit (by a factor 2x or 3x): the converter is in CCM
* If L << Lcrit (by a factor 2x or 3x): the converter is ensured to be in DCM
* If L = Lcrit: the converter is in BCM
Allow Inductor Current to be Negative (Forced CCM or Forced PWM)
To avoid the coil current being discontinuous (blocked at zero), it must continue to drop in the negative values. This means current will be re-injected into the input, increasing wasted power dissipation and thus lowering the efficiency. The advantages are:
* The original transfer function Vout/Vin = d is conserved
* There is less ripple and noise generated by the converter
Figure 3: Forced CCM by Allowing Negative Coil Current
To allow the coil current to flow back to the input stage, the main switch must be bidirectional and sized accordingly
Figure 4: Bidirectional Switch to Allow Coil Reverse Current
Forced Conduction Mode Circuits Example
Most ADI SMPS circuits offer the user the option to set the conduction mode to PWM, forced PWM, forced DCM, forced BCM, PFM, etc.
The MAX17633A/B/C are typical examples of that ability.
Figure 5: MAX17633B Ability to be Set in Forced PWM and DCM Operation
The selection is realized in the switching oscillator and control block. On the MAX17633, it is performed by pin 11 (MODE/SYNC). You must connect MODE/SYNC to SGND for constant-frequency PWM operation at all loads. And connect MODE/SYNC to INTVCC for DCM operation at light loads.
Figure 6: MAX17633 Functional Diagram Showing the Conduction Mode Selector
Conclusion
Comparing DCM and CCM, DCM causes more ripple current and inductor losses. It is easier to stabilize since the same situation occurs at the end of each cycle (the current goes to zero systematically). This mode is well-suited for light or no load.
CCM is well-suited for constant and heavy loads. It offers less output ripple and gives a clean output voltage. The main problem is the reverse coil current when the load is light and/or the inductance is too small. The behavior is also more complex, as the situation at the end of every cycle depends on both the output voltage and coil current.
See blogs in the DCM & CCM in SMPS series.