1
\$\begingroup\$

It seems that some buck converter designs use PWM to change the average voltage sent to the load : which means that for it's (ON) duty cycle, the load will see a higher voltage than required/specified in the it's data sheet.

My question is : is there a theorical maximum PEAK voltage of a PWM (ON) duty cycle ?
If i were to send a 3% duty cycle of 100 volts, to a load expecting 3 volts, wouldn't it break the load somehow ?
What about a 0.3% duty cycle of 1000 volts , etc ?

\$\endgroup\$
1
  • \$\begingroup\$ Ponder the switch ratings at 1 kV, 0.3% duty. Find DC-DC with low output ripple, low input and output ripple, zero ripple at selected terminals. \$\endgroup\$
    – greybeard
    Mar 10, 2023 at 20:29

2 Answers 2

0
\$\begingroup\$

In theory, just like PWM into first order RC filter will have output voltage based on duty cycle, same applies to second order LC filters. The components just need to be correctly selected to have low enough ripple.

In a buck converter, there is always an inductor and an output capacitor, and a buck converter monitors the output voltage and controls the PWM to keep output voltage stable, so it uses feedback.

As long as the inductor has properly selected inductance, and properly selected output capacitance, it will work, in theory.

Assuming you have 100 volts in and 3 volts out, the PWM duty cycle is 3%, but the PWM frequency and allowable output ripple define the inductor and capacitor values, so that the ripple currents and voltages in the components and output are low enough for what you need.

In practice, chips typically have limitations for their minimum on time, which defines how low the duty cycle can go. In practice finding a buck converter that can take 100V in will be difficult, and it will be difficult find one that support 3% duty, and thus much more difficult to find one that can take in 100V and supports 3% duty.

For steady state output (constant current) this will work. The difficulty is having a short on time to ramp inductor current up, if frequency is high to have low enough inductance to respond quickly enough to changes in the load current, so that it does not cause much under- and overshoots or ringing in the output voltage when there is a load step in the current.

\$\endgroup\$
2
\$\begingroup\$

First, most buck converters use an LC filter to get the average voltage of the PWM, which is regulated by a control loop. So the load does not see the peak PWM voltage.

If just using a PWM signal (as is often done for a resistive heater for example) the PWM voltage is chosen to be less than the maximum voltage that the load can handle. Otherwise there is certainly a risk of damaging the load.

\$\endgroup\$

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.