FreeLists / si-list / [SI-LIST] Re: ESR too low?

[Bob McNamara <electronbob2@xxxxxxxxxxxxx>]

Hi Jonathan,

Yes, the ESR of the output capacitor affects the control loop
stability.  If you can change the frequency compensation of
the control loop, you could in theory find values that
optimize the loop response with the output capacitors you want
to use.  If not, you're stuck using capacitors with the higher
ESR (and even then, your 30 degrees of phase margin isn't
something to be real happy with).

Unfortunately optimizing the control loop seems to be in the
realm of "experts only".  The vendors should be providing
design tools to automatically optimize the output impedance
for a given output capacitance and ESR.

For type II compensation, there are only 2 caps and one
resistor to choose.  And type III compensation is still only 3
caps and one or two resistors.  A non-linear optimizer (such
as the Nelder-Mead simplex algorithm) can choose the optimal
values in the blink of an eye; all it needs is a model for the
switching regulator and the parasitics of the output
capacitance.

That may sound like an oversimplification, but I don't think
so.  Earlier this year I had a college intern use that exact
method to optimize the output impedance of a design that had a
nasty peak in its output impedance at 20KHz.  (We modeled the
power section of the switching regulator as a "black box"
transconductance based on step response measurements.)  The
result was excellent; the 39mOhm peak in the output impedance
was gone and the impedance was flat at about 12mOhm from 20KHz
on up to 100's of KHz where the large value ceramics started
to kick in.

I have no idea where the poles and zeros of the compensation
network were placed.  But I do know that the result had nearly
60 degrees of phase margin and more than 10dB of gain margin.

Now if only the vendors would supply such a tool...

Bob McNamara
Broadcom Corp.



--- "Fasig, Jonathan L." <fasig.jonathan@xxxxxxxx> wrote:

> Can anyone speak to the issue of bulk capacitor ESR
> being so low that it
> affects the stability of the voltage regulator?  I
> have a system with a
> well-know commercial 1.5V SMPS powering a board that
> was populated with
> numerous very-low ESR bulk caps.  The transient
> response rings beautifully
> for step load currents and the phase margin
> (supply+board) is typically less
> than 10 degrees (GM<4dB).  At certain low load
> currents the regulator
> oscillates spontaneously at 1.3KHz.  When these
> low-ESR caps are replaced or
> removed, the transient response settles nicely and
> phase margin approaches
> 30 degrees (GM>12dB) but ripple magnitude is
> affected.
> 
> Discussion or references appreciated.
> 
> Regards,
> jf
> 
> ------------------------------------------------------------------------
> 
> Jonathan Fasig                      Email: fasig.jonathan@xxxxxxxx    
> Mayo Foundation 
> 4001 41st Street NW 
> MSC Sn 2-132                        Phone:  (507)538-5464
> Rochester, MN 55901                 Fax:    (507)284-9171
> 

------------------------------------------------------------------
To unsubscribe from si-list:
si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field

or to administer your membership from a web page, go to:
http://www.freelists.org/webpage/si-list

For help:
si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field

List FAQ wiki page is located at:
                http://si-list.org/wiki/wiki.pl?Si-List_FAQ

List technical documents are available at:
                http://www.si-list.org

List archives are viewable at:     
                http://www.freelists.org/archives/si-list
or at our remote archives:
                http://groups.yahoo.com/group/si-list/messages
Old (prior to June 6, 2001) list archives are viewable at:
                http://www.qsl.net/wb6tpu