[SI-LIST] Re: Switch Mode PS Noise

Declan,

Regarding cleaning up the probe response, often there are some ground 
loop currents flowing down the shield of the probe when making this 
type of measurement. It passes through the chassis of the scope and 
flows down the safety earth or down other probe sheilds to what the 
scope thinks is "ground". This "extra" current in the shield generates 
a greater voltage drop on it than occurs on the center conductor, 
which is typically driving a megohm. The goal is to have the current 
in the shield exactly equal, but opposite sign of the current in the 
center conductor.  This can often be accomplished by passing the probe 
through ferrite cores. The "real" probe signal currents don't see the 
core impedance because the net magnetism from them is zero. The 
"clamshell" EMI cores seem to work very well for this.

Regards,
Ron

>X-Original-To: si-list@xxxxxxxxxxxxx
>Delivered-To: si-list@xxxxxxxxxxxxx
>Date: Mon, 23 Oct 2006 07:14:24 -0700
>To: "Declan Hegarty" <declan.hegarty@xxxxxxxxx>
>From: steve weir <weirsi@xxxxxxxxxx>
>Subject: [SI-LIST] Re: Switch Mode PS Noise
>Cc: istvan.novak@xxxxxxxxxxx, si-list@xxxxxxxxxxxxx
>Mime-Version: 1.0
>X-Virus-Scanned: by amavisd-new-20030616-p10 (Debian) at 
avenirtech.net
>X-archive-position: 19907
>X-ecartis-version: Ecartis v1.0.0
>X-original-sender: weirsi@xxxxxxxxxx
>X-list: si-list
>X-Virus-Scanned: by amavisd-new-20030616-p10 (Debian) at 
avenirtech.net
>Content-Transfer-Encoding: quoted-printable
>X-esp: ESP<-125>=RBL:<-140> SHA:<15> UHA:<0> SLS:<0> BAYES:<0> 
SenderID:<0> Spam Dictionary (TRU10):<0> Obscenities Dictionary 
(TRU10):<0> Scam Dictionary (TRU10):<0> Adult Dictionary (TRU10):<0> 
Embed HTML Dictionary (TRU10):<0> Float Dictionary (TRU10):<0> 
esptweak:<0> HTML Dictionary (TRU10):<0> URL Real-Time Signatures:<0> 
Spam Dictionary 2 (TRU10):<0> 
SIG:<dZ58npA1AmUFBd9EhQYaCkMA3HIhniRIf4BPmWkbPD2DuyhTYRr-ETicZ2u8 
Er8-VPcZ5And0LMNfVN-TcdlfP4huP-wheNKJAAAAAAAAAAAAAAAAAAAAAAA 
AAAAAAAAAAAAAAAAAAAAKQiIfl40351Vo6Mm9NPtAkzeSMwaNOoA_K-ZqiAE 
Xk6S1qGv4AQFBkANcmiipISzAAp0jAVBxYJn2reSyZPuPKUTA>
>
>Declan, if you fashion a probe with minimal loop area between the 
>shield and the probe tip, you will minimize your noise pick-up.  If 
>you are using a probe with a ground clip attached to an inch or two 
>of wire, you will have lots of pick-up.
>
>When the switcher runs it could be imparting noise on the input in a 
>frequency range that the linear doesn't suppress.  An additional item 
>to check is comparison of the noise at the input with the output.
>
>Good luck.
>
>
>Steve.
>At 06:35 AM 10/23/2006, Declan Hegarty wrote:
>>Thanks very much for your input guys.  I think I've confused things
>>with a poor description of my circuit.  The switch-mode output here 
is
>>not the input to the linear regulator.  The IC I'm using, although
>>designed as a dual-ouput switch mode controller, also happens to 
have
>>a small linear regulator controller built in, comprising a feedback
>>pin (to a fixed internal reference of 1V), and a base drive pin.  So
>>electrically, the switch-mode and linear supplies should be
>>independent.  The input bias for the pnp comes from a different
>>source,
>>
>>However (possibly due to poor component placement ) I believe the
>>linear regulator external transistor is picking up switching noise.
>>When I shut the switch-mode output down, the linear regulator works
>>just fine.  When the switch-mode output is running, I can see 
voltage
>>spikes on the base of the pnp, synchronous with the switching
>>frequency.  I'm just trying to establish how much of this noise is
>>"real" and how much is spurious pickup by the probe.
>>
>>Regards,
>>Declan.
>>
>>On 10/23/06, steve weir <weirsi@xxxxxxxxxx> wrote:
>>>Declan, just to add a hopefully simple test to the suggestions 
Istvan
>>>offered:  break the connection from the switcher to the LDO, and
>>>source the LDO from a suitable bench supply.  If the problem is 
gross
>>>instability in the LDO loop, oscillations will still
>>>develop.  Compensation is your friend.
>>>
>>>Good hunting.
>>>
>>>
>>>Steve.
>>>
>>>
>>>At 06:07 AM 10/23/2006, Istvan Novak wrote:
>>> >Hi Declan,
>>> >
>>> >If the LDO output is higher than the set point, the likely 
problem is
>>> >self-oscillation,
>>> >especially that you use a PNP transistor in the LDO.  The signal 
you see
>>> >at the
>>> >base of the PNP transistor may appear to be related to the 
switching
>>> >frequency
>>> >of the upstream converter, but unless you can see both waveforms 
on the
>>> >same
>>> >scope screen, synchronized, I would first suspect the 
self-oscillation
>>> >of the LDO.
>>> >
>>> >Suggestions:
>>> >- check the ripple at the output of DC-DC converter, which is the 
input
>>> >of the
>>> >LDO.  Supposedly you have buffer capacitors at that point, and 
>>> the switching
>>> >ripple should be below 100mVpp.  If bigger than 100mVpp, you need 
more
>>> >capacitance or better capacitors at the point.  If below 100mVpp:
>>> >- change the LDO, by adding a small series resistance (10-30 
ohms) to the
>>> >base of the PNP transistor.  This may already kill the 
self-oscillation,
>>> >if in
>>> >fact that causes your problem
>>> >- if this is not enough, you can overcompensate the LDO 
regulation loop.
>>> >If there is access to the feedback loop, try adding capacitors to 
>>> reduce the
>>> >loop bandwidth (where and how much, depends on the LDO).   You 
can
>>> >also try adding capacitance (nF range) in parallel to the 
base-collector of
>>> >the PNP transistor
>>> >- note also that LDO stability may be affected by the output load 
and
>>> >output capacitance and any extra phase shift in the sense and 
feedback
>>> >looop
>>> >
>>> >Once you know where the problem is coming from, we can think 
about
>>> >fixing it.
>>> >
>>> >Regards,
>>> >
>>> >Istvan Novak
>>> >SUN Microsystems
>>> >
>>> >Declan Hegarty wrote:
>>> >
>>> > >Hi folks,
>>> > >
>>> > >I'm currently investigating a power supply circuit based on an 
IC with
>>> > >two switch-mode outputs and a small linear regulator.  The 
linear
>>> > >regulator uses an external pnp transistor in a simple low 
dropout,
>>> > >common emitter configuration, and is currently running out of
>>> > >regulation, with an average DC level of 2.39V instead of the 
required
>>> > >1.8V.
>>> > >
>>> > >I believe that noise from one of the switcher outputs is 
coupling into
>>> > >the base of this transistor, causing current pulses at the 
output
>>> > >which are being averaged across the output capacitor to 
increase the
>>> > >nominal DC level.  When I shut down the offending switch mode 
output,
>>> > >the linear regulator works properly.  So to my question.......
>>> > >
>>> > >I'm using a fairly cheap and cheerful oscilloscope and probe.  
When I
>>> > >probe the base of the pnp transistor, I see 600mV spikes
>>> > >time-correlated with the switch-mode transitions from 0-12V.  
I'd like
>>> > >to understand how much of this noise is just pickup by the 
probe
>>> > >leads, and how much is actually present on the base pin of the 
pnp.  I
>>> > >wondered if the following was a valid approach:
>>> > >
>>> > >1.  Place the ground lead of the probe at a fixed ground point 
on the
>>> > >PCB, near the area of interest.
>>> > >2.  Probe a ground via as close to the pnp base as I can find.
>>> > >3.  Observe the noise on this measurement.
>>> > >4.  Now probe the pnp base.
>>> > >5.  Observe the noise on this measurement.  Subtract the noise
>>> > >observed in 3 to approximate the "real" coupled noise at the 
pnp base.
>>> > >
>>> > >Clearly this isn't going to be high precision, I know.  Any 
other
>>> > >suggestions for an improved understanding of what's going on 
here
>>> > >would be much appreciated.
>>> > >
>>> > >Thanks. D.
>>> > >
>>> > >
>>> >
>>> 
>------------------------------------------------------------------
>>> >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
>>> >
>>>
>
>------------------------------------------------------------------
>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
>  
>

------------------------------------------------------------------
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

Other related posts: