1
\$\begingroup\$

I am trying to understand the benefits of constant on time control (COT) over voltage or current mode control.

In voltage mode control, during a load transient (increase in load) an aggressive control loop would drive the duty ratio to 1 (until the output shows signs of recovery) and helps to reduce the undershoot.

I do not understand how COT is better in this regard since it will only produce a set of closely spaced pulses. How is it better than keeping the switch on continuously?

Advantages of Constant-On-Time Control in DC/DC Converters

\$\endgroup\$
1
  • \$\begingroup\$ May as well add this link to your question. \$\endgroup\$
    – jonk
    Commented Aug 15, 2021 at 20:48

1 Answer 1

3
\$\begingroup\$

I am just summarizing & translating what is said in the link. Thanks for sharing. I like it.

Drawback of the voltage feedback: The output voltage is (should be) filtered and fed back to the control loop (voltage level comparator). Thus, the feedback is delayed by the filtering, resulting less optimum (slower) transient response.

Drawback of the current mode: The PWM frequency is constant while the Pulse Width varies depending on the fed-back current. Thus, the control is always "PWM cycle time" delayed (set PW becomes effective at the next cycle, coming after the measurement cycle)

COT architecture compared to the Voltage Control mode:
Since the control relies on the immediate voltage drop (valley) of the current cycle, however it is a voltage feedback (eventually), the voltage has to be fed back without the filtering. (Filtering will defeat the COT mechanism. Later stage, that voltage can be filtered on the load side.) Thus, COT does not have the delay from the output filtering.

COT compared to the Current mode:
COT immediately starts generating pulse at the detection of the threshold (voltage dip). Thus the control becomes effective without waiting time. But,the current mode detects the current of the immediate cycle and wait until the next PWM cycle in order to place a control pulse. (The PWM controller cannot cancel the current cycle and restart the cycle.)

\$\endgroup\$
2
  • \$\begingroup\$ Do you know if COT can also be used in boost configuration? Most of the discussion in the paper seems to be assuming buck configuration. \$\endgroup\$
    – user57037
    Commented Aug 16, 2021 at 1:14
  • 1
    \$\begingroup\$ @mkeith , I am not really an expert in switching power supply area, but thought COT is an interesting idea. I thought that too, as you said, boost configuration could be difficult with COT. Now, I guess, the concept of the COT can be used in boost control. However, COT may loose the merit of single cycle response time, unless compromise other aspects. \$\endgroup\$
    – jay
    Commented Aug 16, 2021 at 2:42

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.