From: email@example.com (Bob Wilson)
Subject: Re: Ringing in Discontinuous Buck regulator
Date: Tue, 24 Sep 2002 05:56:54 -0000
Organization: Your Organization
X-Newsreader: WinVN 0.99.9 (Released Version) (x86 32bit)
In article <3D8F348B.D8BBCEA7@singnet.com.sg>, firstname.lastname@example.org says...
>I am building a simple DC/DC Buck converter. I noticed that whenever,
>the current goes to Discontinuous mode, there will be a ringing waveform
>on the SOURCE pin of the switching Mosfet. This happens when the
>inductor current goes to zero. What I am curious is what is the
>mechanism that cause this ringing? Where is the Inductance and
>Capacitance coming from?
Yes, that is absolutely typical. It you think about it, it even makes
As the FET turns off, the inductor begins discharging its magnetic energy,
and since the original source of current (i.e. the FET) is now off, it
begins pulling current "up" through the freewheeling diode. As the energy in
the inductor is discharged, current through it (and through the diode)
linearly falls off. The circuit goes "discontinuous" the moment that the
inductor has discharged all its energy. Now, the diode stops conducting. But
the FET is still off as well. Thus, the FET/Diode end of the inductor "sees"
an open circuit (the FET is open and so is the diode since no current flows
through either of them).
But the non-conducting diode has capacitance to ground (and also there is
other stray capacitance as well). So once it goes discontinuous, you have a
series resonant tank circuit. The output side of the inductor still has the
low impedance of the load connectod to it, and the input end of the inductor
is connected through the series capacitance of the diode to ground. So, this
L/C tank sinply resonates at its own natural resonance frequency.
The resonance is damped, so it eventually dies out after a few cycles, if
the "off" time is long, or it is killed instantly the moment the FET turns
on again ....whichever comes first.
Because of the oscillation you are observing, it is generally not good
practice to design a Buck so it runs discontinuous. It is not damaging or
anything, but the main problem is increased EMI due to the large
amplitude of "ringing" (and also sluggish transient response).