RE: OATS vs anechoic - AC line impedance at RF

2011-02-15 Thread emc-p...@ieee.org 
It’s worse than just the filter.  How long are the wire runs from the
outlets on the turntable to the filter?  What is between the filter and the
outlets?  Circuit breaker panels in the turntable pit?  Or not?  How are the
wires run?  Different ways for different OATS facilities, not just between
different chambers.

 

This mess is why ferrite clamps on the cables leaving the measurement area
were incorporated for a time in CISPR 22.  They weren’t adequately defined,
so they got pulled in a later revision.  That, and what do you do if you have
100 cables leaving the EUT?  But the problem persists.  A very real problem
for repeatability between test facilities.

 

Ghery S. Pettit

 

From: emc-p...@ieee.org [mailto:emc-p...@ieee.org] On Behalf Of
ralph.mcdiar...@schneider-electric.com
Sent: Tuesday, February 15, 2011 3:47 PM
To: EMC-PSTC@LISTSERV.IEEE.ORG
Subject: Re: OATS vs anechoic - AC line impedance at RF

 


That's an 'eye opener' Derek.  The tests I've witnessed over the years always
have AC power coming through an LISN during RE testing, so at least that helps
standardise the impedance.   I wonder what effect the AC facility filter
upstream from the LISN has on source impedance.
___
_ 

Ralph McDiarmid  |   Schneider Electric   |  Renewable Energies Business  |  
CANADA  |   Regulatory Engineer 




From: 

Derek Walton  

To: 

EMC-PSTC@LISTSERV.IEEE.ORG 

List-Post: emc-pstc@listserv.ieee.org
Date: 

02/13/2011 06:07 AM 

Subject: 

Re: [PSES] OATS vs anechoic

 






HI Brian,

a number of things can be going on.  Just for a start, the chambers and OATS
have NSA done, if your problems are vertical, it's possible to have 8 dB
difference in NSA especially at the lower end: just keep that in mind.

HOWEVER, NSA doesn't always reflect what the EUT does on the site. An
observation I made testing a pump a while back was that the impedance of the
power cord as it "terminates" at the shield room/OATS ground plane influenced
greatly the radiated emissions.

What I mean by terminates is that power under the ground plane may have
filters, LISN's or nothing. BUT, the impedance there is going to influence the
emissions. In my experiment I added several large Ferrites ( almost the size
of a fist, just as the power came into the chamber from the Power line
filters. There is nothing in the standards that says what the impedance of the
power source should be for RE, so I believe this is legitimate. At 70 MHz
emissions dropped almost 12 dB. 

The way you describe outages in bands kind of implies that a cable is
resonating when you have excessive emissions, maybe the beads are worth a
try...

Cheers,

Derek. 

-Original Message-
From: Brian O'Connell 
To: 'Brent G DeWitt' 
Cc: EMC-PSTC@LISTSERV.IEEE.ORG
Sent: Sat, Feb 12, 2011 6:02 pm
Subject: RE: [PSES] OATS vs anechoic

Brent,





The power cord is dressed casual. Sorry.





Class II construction - no ground wire, hence no ground plane to reference.





Back at the office and am now viewing similar dissimilar results from two
indoor 


and two outdoor sites, so it probably has something to do with the H-field,
and 


the power cord radiator makes a little more sense to me, I think...





Brian 





> -Original Message-


> From: Brent G DeWitt


> Sent: Saturday, February 12, 2011 3:53 PM


> To: oconne...@tamuracorp.com <mailto:oconne...@tamuracorp.com> 


> Subject: RE: [PSES] OATS vs anechoic


> 


> 


> Hi again Brian,


> 


> How is the power cord dressed and terminated at the ground 


> plane?  Huge


> variations between test sites on how this is done!


> 


> Brent


> 


> > -Original Message-


> > From: Brian O'Connell [mailto:oconne...@tamuracorp.com
<mailto:oconne...@tamuracorp.com?> ]


> > Sent: Saturday, February 12, 2011 6:22 PM


> > To: EMC-PSTC@LISTSERV.IEEE.ORG <mailto:EMC-PSTC@LISTSERV.IEEE.ORG> 


> > Subject: Re: [PSES] OATS vs anechoic


> > 


> > Just finished 2d OATS site - was not able to detect much 


> of anything.


> > 


> > Went back to indoors fully anechoic, where the noise plot 


> is deafening.


> > 


> > This is a simple 100W unit with one main ac/dc converter and (6) dc


> > current modulators. Construction is mostly plastic, with 


> exception of


> > metal base plate. As it is Class II construction, there is 


> no ground


> > wire.


> > 


> > Bad stuff is at 40-80MHz and 200-500MHz. Did some isolation


> > troubleshooting - looks like the input power cord is the 


> main radiator.


> > 


> > The EMC engineers at all three sites are both baffled and 


> fascinated


> > and want a sample

Re: OATS vs anechoic - AC line impedance at RF

2011-02-15 Thread emc-p...@ieee.org 

That's an 'eye opener' Derek.  The tests I've witnessed over the years always
have AC power coming through an LISN during RE testing, so at least that helps
standardise the impedance.   I wonder what effect the AC facility filter
upstream from the LISN has on source impedance.
___
_ 

Ralph McDiarmid  |   Schneider Electric   |  Renewable Energies Business  |  
CANADA  |   Regulatory Engineer 




From:   Derek Walton  
To: EMC-PSTC@LISTSERV.IEEE.ORG 
List-Post: emc-pstc@listserv.ieee.org
Date:   02/13/2011 06:07 AM 
Subject:Re: [PSES] OATS vs anechoic






HI Brian,

a number of things can be going on.  Just for a start, the chambers and OATS
have NSA done, if your problems are vertical, it's possible to have 8 dB
difference in NSA especially at the lower end: just keep that in mind.

HOWEVER, NSA doesn't always reflect what the EUT does on the site. An
observation I made testing a pump a while back was that the impedance of the
power cord as it "terminates" at the shield room/OATS ground plane influenced
greatly the radiated emissions.

What I mean by terminates is that power under the ground plane may have
filters, LISN's or nothing. BUT, the impedance there is going to influence the
emissions. In my experiment I added several large Ferrites ( almost the size
of a fist, just as the power came into the chamber from the Power line
filters. There is nothing in the standards that says what the impedance of the
power source should be for RE, so I believe this is legitimate. At 70 MHz
emissions dropped almost 12 dB. 

The way you describe outages in bands kind of implies that a cable is
resonating when you have excessive emissions, maybe the beads are worth a
try...

Cheers,

Derek. 

-Original Message-
From: Brian O'Connell 
To: 'Brent G DeWitt' 
Cc: EMC-PSTC@LISTSERV.IEEE.ORG
Sent: Sat, Feb 12, 2011 6:02 pm
Subject: RE: [PSES] OATS vs anechoic

Brent,





The power cord is dressed casual. Sorry.





Class II construction - no ground wire, hence no ground plane to reference.





Back at the office and am now viewing similar dissimilar results from two
indoor 


and two outdoor sites, so it probably has something to do with the H-field,
and 


the power cord radiator makes a little more sense to me, I think...





Brian 





> -Original Message-


> From: Brent G DeWitt


> Sent: Saturday, February 12, 2011 3:53 PM


> To: oconne...@tamuracorp.com  


> Subject: RE: [PSES] OATS vs anechoic


> 


> 


> Hi again Brian,


> 


> How is the power cord dressed and terminated at the ground 


> plane?  Huge


> variations between test sites on how this is done!


> 


> Brent


> 


> > -Original Message-


> > From: Brian O'Connell [mailto:oconne...@tamuracorp.com
 ]


> > Sent: Saturday, February 12, 2011 6:22 PM


> > To: EMC-PSTC@LISTSERV.IEEE.ORG  


> > Subject: Re: [PSES] OATS vs anechoic


> > 


> > Just finished 2d OATS site - was not able to detect much 


> of anything.


> > 


> > Went back to indoors fully anechoic, where the noise plot 


> is deafening.


> > 


> > This is a simple 100W unit with one main ac/dc converter and (6) dc


> > current modulators. Construction is mostly plastic, with 


> exception of


> > metal base plate. As it is Class II construction, there is 


> no ground


> > wire.


> > 


> > Bad stuff is at 40-80MHz and 200-500MHz. Did some isolation


> > troubleshooting - looks like the input power cord is the 


> main radiator.


> > 


> > The EMC engineers at all three sites are both baffled and 


> fascinated


> > and want a sample unit to play with. I am very happy that they are


> > amused, but I have to ship a product and cannot afford to 


> support the


> > Southern California EMC industry.


> > 


> > thanks,


> > Brian


> > 


> >  > -Original Message-


> >  > From: emc-p...@ieee.org  
[mailto:emc-p...@ieee.org]On   


> Behalf Of Brian


> >  > O'Connell


> >  > Sent: Saturday, February 12, 2011 12:25 PM


> >  > To: EMC-PSTC@LISTSERV.IEEE.ORG  


> >  > Subject: OATS vs anechoic


> >  >


> >  >


> >  > Good People,


> >  >


> >  > Have tested at several sites. All are 'big' names and have a


> >  > good reputation.


> >  >


> >  > The 10m fully anechoic chamber data indicates from +5dB to


> >  > +11dB in several frequencies; while the 10m OATS says the


> >  > unit passes Class B, min -4.2dB.


> >  >


> >  > Huh?!? What is the physics behind the difference between an


> >  > outside site and absorbers? Where are the Borg when you 


> need them?


> >  >


> >  > Brian





-





This message is from the IEEE Product Saf