Choosing primary inductor and output capacitor values for a buck converter

[Michael]

Hi there - I've been working this through my head, and I haven't been
able to put my finger on it. Both your inductor and your output
capacitor affect Vout ripple on a buck converter. But are there
reasons to increase one and not the other, or anything like that? For
example, if you were to design the DC/DC converter to be able to
handle a large step in the output current without a large change in
output voltage, would there be one of these that you would want to
change more than the other?

Thanks!

-Michael

 

[John Popelish]

Hi there - I've been working this through my head, and I haven't been
able to put my finger on it. Both your inductor and your output
capacitor affect Vout ripple on a buck converter. But are there
reasons to increase one and not the other, or anything like that? For
example, if you were to design the DC/DC converter to be able to
handle a large step in the output current without a large change in
output voltage, would there be one of these that you would want to
change more than the other?

I think the generalities are:

For a given frequency, ripple voltage and load current,
higher inductance improves the losses in the switch, since
the peak current gets closer to the load current as the
inductance rises.

For a given frequency, ripple voltage and load current, the
transient response (settling time after a step load or input
voltage change) gets shorter as the inductance gets lower,
since the inductor current changes more in a single cycle.

Those generalities do not imply that changing the frequency
is not something you should also consider.

 

[Nobody]

Hi there - I've been working this through my head, and I haven't been
able to put my finger on it. Both your inductor and your output
capacitor affect Vout ripple on a buck converter. But are there
reasons to increase one and not the other, or anything like that? For
example, if you were to design the DC/DC converter to be able to
handle a large step in the output current without a large change in
output voltage, would there be one of these that you would want to
change more than the other?

The larger the inductor, the greater the voltage spike if you suddenly
disconnect the current (or the voltage drop if you suddenly increase the
current drain). The larger the capacitor, the lower the voltage change.

To allow for sudden drops in current draw, it helps if 0.5*LI^2 is
substantially less than 0.5*CV^2 (i.e. if the current drops to zero,
you can dump all of the inductor's energy into the capacitor without a
significant voltage increase).

For the case of the output current suddenly decreasing, you could just
dump the energy into an overvoltage protection circuit if it's likely to
be an infrequent occurrence (i.e. not a PWM load). For the case where the
current suddenly increases, you can't increase the inductor current any
faster than dI/dt = Vin/L, so any excess has to come from the capacitor.

tl;dr version: a lower L/C ratio = less voltage fluctuation for a given
current fluctuation.

 

[Tim Wescott]

I think the generalities are:

For a given frequency, ripple voltage and load current,
higher inductance improves the losses in the switch, since
the peak current gets closer to the load current as the
inductance rises.

For a given frequency, ripple voltage and load current, the
transient response (settling time after a step load or input
voltage change) gets shorter as the inductance gets lower,
since the inductor current changes more in a single cycle.

Those generalities do not imply that changing the frequency
is not something you should also consider.

I believe that more inductance also implies a bigger inductor, because the
amount of energy that you can store in a given volume of core is more or
less constant. Higher inductance at a given current stores more energy in
the core, so it requires a larger core to prevent saturation.

--
Tim Wescott
Control systems and communications consulting
http://www.wescottdesign.com

Need to learn how to apply control theory in your embedded system?
"Applied Control Theory for Embedded Systems" by Tim Wescott
Elsevier/Newnes, http://www.wescottdesign.com/actfes/actfes.html