Re: DOC Requirement for Year CE Mark is Affixed
I would agree with your interpretation, with a caveat in regard to your use of four digits for the year. We have, in the past, been told by the MHRA (the enforcing body for the Medical Devices Directive in the UK) that we were not allowed to put the year number on DofC's because it might be confused with the notified body (NB) number for a product where the NB number is a requirement (i.e. where the manufacturer's quality system has to be assessed). This was for a two digit year number so I imagine they'd take a very dim view of the use of four digits (since this is the standard format for a notified body number). Our (unofficial) response to this was to conclude that the MHRA don't know the rules which apply to products which are not medical devices very well, and to ignore them in these cases. However, we do take care to ensure the digits are not present on DofCs for medical devices - the MDD and the directives which require the digits of the year are mutually exclusive so this is the correct approach. Nick. At 19:09 -0600 2/6/08, Monrad Monsen wrote: The Low Voltage Directive 2006/95/EC states in annex III that a declaration of conformity (DOC) must contain the following elements: ... the last two digits of the year in which the CE marking was affixed. I note that a like statement is not listed for the EMC Directive. I interpret this requirement as listing the year in which the CE mark was first affixed to this model number product. If true, then the signature date on the DOC would not be adequate because a company may make a later change to a DOC and re-issue the DOC long after the first CE mark was affixed to this product line. I looked through the DOCs of others in our industry, and I do not see this being followed anywhere. What are your interpretations of this requirement? My interpretation is that the last two digits of the year statement is merely a minimum of two digits, but it allows for using a four digit year. Most people avoid using two digit years after the year 2000 crunch, so I'd prefer to use a four digit year. You can view the low voltage directive by going to http://ec.europa.eu/enterprise/electr_equipment/lv/direct/text.htmhttp://ec.europa.eu/enterprise/electr_equipment/lv/direct/text.htm; and clicking on the language version of choice (en for English, etc.). Again, the requirement is found in annex III. Thanks. -- Monrad Monrad L. Monsen Compliance Program Manager Storage Group Sun Microsystems mailto:monrad.mon...@sun.commonrad.mon...@sun.com 303.272.9612 Office - This message is from the IEEE Product Safety Engineering Society emc-pstc discussion list.Website: http://www.ieee-pses.org/ To post a message to the list, send your e-mail to emc-p...@ieee.org Instructions: http://listserv.ieee.org/request/user-guide.html List rules: http://www.ieee-pses.org/listrules.html For help, send mail to the list administrators: Scott Douglas emcp...@ptcnh.net Mike Cantwell mcantw...@ieee.org For policy questions, send mail to: Jim Bacher: j.bac...@ieee.org David Heald:emc-p...@daveheald.com All emc-pstc postings are archived and searchable on the web at: http://www.ieeecommunities.org/emc-pstc
Re: DOC Requirement for Year CE Mark is Affixed
In message p06240801c46a9cb4a264@[192.168.1.30], dated Tue, 3 Jun 2008, Nick Williams nick.willi...@conformance.co.uk writes: We have, in the past, been told by the MHRA (the enforcing body for the Medical Devices Directive in the UK) that we were not allowed to put the year number on DofC's because it might be confused with the notified body (NB) number for a product where the NB number is a requirement (i.e. where the manufacturer's quality system has to be assessed). This was for a two digit year number so I imagine they'd take a very dim view of the use of four digits (since this is the standard format for a notified body number). Our (unofficial) response to this was to conclude that the MHRA don't know the rules which apply to products which are not medical devices very well, and to ignore them in these cases. However, we do take care to ensure the digits are not present on DofCs for medical devices - the MDD and the directives which require the digits of the year are mutually exclusive so this is the correct approach. 'Mutually exclusive', so it's perhaps not surprising that the requirements for the content of DoCs differ between the MDD and the LVD. But it does seem to be illogical, like so much in EU legislation. A more fundamental question is 'What is the reason for requiring the two digits to be included?' It is also far from clear what 'first affixed' means, when you take into account engineering changes in the product, changes in the edition of the safety standard which is valid, changes in the Directive which applies... For example, products now having a DoC referring to 2006/95/EC may have had the CE mark first affixed under Directive 73/23/EEC. Is that the required date, or that on which the CE mark was 'first affixed' under 2006/95/EC? In view of all these uncertainties, I think it's not surprising that many DoCs do not include the date. -- OOO - Own Opinions Only. Try www.jmwa.demon.co.uk and www.isce.org.uk Either we are causing global warming, in which case we may be able to stop it, or natural variation is causing it, and we probably can't stop it. You choose! John Woodgate, J M Woodgate and Associates, Rayleigh, Essex UK - This message is from the IEEE Product Safety Engineering Society emc-pstc discussion list.Website: http://www.ieee-pses.org/ To post a message to the list, send your e-mail to emc-p...@ieee.org Instructions: http://listserv.ieee.org/request/user-guide.html List rules: http://www.ieee-pses.org/listrules.html For help, send mail to the list administrators: Scott Douglas emcp...@ptcnh.net Mike Cantwell mcantw...@ieee.org For policy questions, send mail to: Jim Bacher: j.bac...@ieee.org David Heald:emc-p...@daveheald.com All emc-pstc postings are archived and searchable on the web at: http://www.ieeecommunities.org/emc-pstc
RE: DOC Requirement for Year CE Mark is Affixed
According to Annex III, there is no mention as to a date that the DoC is signed by the signatory. Therefore, is the the last two digits of the year in which the CE marking was affixed (for the first time) the same as the date a signatory signed a DoC, if it appears on the DoC? Best regards, Ron Wellman From: emc-p...@ieee.org [mailto:emc-p...@ieee.org]On Behalf Of Nick Williams Sent: Tuesday, June 03, 2008 12:20 AM To: Monrad Monsen Cc: emc-p...@ieee.org Subject: Re: DOC Requirement for Year CE Mark is Affixed I would agree with your interpretation, with a caveat in regard to your use of four digits for the year. We have, in the past, been told by the MHRA (the enforcing body for the Medical Devices Directive in the UK) that we were not allowed to put the year number on DofC's because it might be confused with the notified body (NB) number for a product where the NB number is a requirement (i.e. where the manufacturer's quality system has to be assessed). This was for a two digit year number so I imagine they'd take a very dim view of the use of four digits (since this is the standard format for a notified body number). Our (unofficial) response to this was to conclude that the MHRA don't know the rules which apply to products which are not medical devices very well, and to ignore them in these cases. However, we do take care to ensure the digits are not present on DofCs for medical devices - the MDD and the directives which require the digits of the year are mutually exclusive so this is the correct approach. Nick. At 19:09 -0600 2/6/08, Monrad Monsen wrote: The Low Voltage Directive 2006/95/EC states in annex III that a declaration of conformity (DOC) must contain the following elements: ... the last two digits of the year in which the CE marking was affixed. I note that a like statement is not listed for the EMC Directive. I interpret this requirement as listing the year in which the CE mark was first affixed to this model number product. If true, then the signature date on the DOC would not be adequate because a company may make a later change to a DOC and re-issue the DOC long after the first CE mark was affixed to this product line. I looked through the DOCs of others in our industry, and I do not see this being followed anywhere. What are your interpretations of this requirement? My interpretation is that the last two digits of the year statement is merely a minimum of two digits, but it allows for using a four digit year. Most people avoid using two digit years after the year 2000 crunch, so I'd prefer to use a four digit year. You can view the low voltage directive by going to http://ec.europa.eu/enterprise/electr_equipment/lv/direct/text.htmhttp:/ /ec.europa.eu/enterprise/electr_equipment/lv/direct/text.htm and clicking on the language version of choice (en for English, etc.). Again, the requirement is found in annex III. Thanks. -- Monrad Monrad L. Monsen Compliance Program Manager Storage Group Sun Microsystems mailto:monrad.mon...@sun.commonrad.mon...@sun.com 303.272.9612 Office - This message is from the IEEE Product Safety Engineering Society emc-pstc discussion list.Website: http://www.ieee-pses.org/ To post a message to the list, send your e-mail to emc-p...@ieee.org Instructions: http://listserv.ieee.org/request/user-guide.html List rules: http://www.ieee-pses.org/listrules.html For help, send mail to the list administrators: Scott Douglas emcp...@ptcnh.net Mike Cantwell mcantw...@ieee.org For policy questions, send mail to: Jim Bacher: j.bac...@ieee.org David Heald:emc-p...@daveheald.com All emc-pstc postings are archived and searchable on the web at: http://www.ieeecommunities.org/emc-pstc - This message is from the IEEE Product Safety Engineering Society emc-pstc discussion list.Website: http://www.ieee-pses.org/ To post a message to the list, send your e-mail to emc-p...@ieee.org Instructions: http://listserv.ieee.org/request/user-guide.html List rules: http://www.ieee-pses.org/listrules.html For help, send mail to the list administrators: Scott Douglas emcp...@ptcnh.net Mike Cantwell mcantw...@ieee.org For policy questions, send mail to: Jim Bacher: j.bac...@ieee.org David Heald:emc-p...@daveheald.com All emc-pstc postings are archived and searchable on the web at: http://www.ieeecommunities.org/emc-pstc
Re: DOC Requirement for Year CE Mark is Affixed
In message nbenipnmgoglomnalfjlaefbcnaa.rwell...@wellman.com, dated Tue, 3 Jun 2008, Ronald R. Wellman rwell...@wellman.com writes: According to Annex III, there is no mention as to a date that the DoC is signed by the signatory. Therefore, is the the last two digits of the year in which the CE marking was affixed (for the first time) the same as the date a signatory signed a DoC, if it appears on the DoC? No, for the reasons indicated in my follow-up to Nick's message. Apart from changes to the product's internals or the applicable standards since the mark was 'first affixed', the DoC signatory may change. -- OOO - Own Opinions Only. Try www.jmwa.demon.co.uk and www.isce.org.uk Either we are causing global warming, in which case we may be able to stop it, or natural variation is causing it, and we probably can't stop it. You choose! John Woodgate, J M Woodgate and Associates, Rayleigh, Essex UK - This message is from the IEEE Product Safety Engineering Society emc-pstc discussion list.Website: http://www.ieee-pses.org/ To post a message to the list, send your e-mail to emc-p...@ieee.org Instructions: http://listserv.ieee.org/request/user-guide.html List rules: http://www.ieee-pses.org/listrules.html For help, send mail to the list administrators: Scott Douglas emcp...@ptcnh.net Mike Cantwell mcantw...@ieee.org For policy questions, send mail to: Jim Bacher: j.bac...@ieee.org David Heald:emc-p...@daveheald.com All emc-pstc postings are archived and searchable on the web at: http://www.ieeecommunities.org/emc-pstc
Placement of LISNs for Conducted Emissions Testing to CISPR22/FCC part 15
Our company sells ITE systems that are housed in commerial 19 racks. The system is designed to be fault tolerant and redundant, so each rack has two AC mains cords. We do our own EMI certification testing (we're an accreditted lab), and our typical EUT consists of two of these rack, so there are 4 AC Mains cords to test, which we connect to 4 LISNs. A modified version of this system is now in the works, where the dual AC mains cords are replaced by multiple cords (with lower current per cord). The design now would have 10 AC mains cords out of one rack, and 8 from the other. This means the two rack EUT would have 18 AC Mains cords to test. In a perfect world, where cost were no object, we would have 18 LISNs, since this is the most efficient for testing -- set it up once, and test everything. Our question is how to place a relatively large number of LISNs and satisfy the standards' requirement of the 80cm spacing of the EUT and LISN. Specifically: 1) Are we allowed to place LISNs around all sides of the EUT, maintaining the 80cm spacing (i.e. have LISNs at the front face of the EUT, and run the mains cord from the back to the LISN)? 2) Are we allowed to stack LISNs on top of each other, as long as the LISN is bonded to the ground plane? Since we have to test each cord in turn, we could reduce the number of LISNs by combining a number of the cords not currently being tested through a second (or third) LISN. The downside of this is having to re-plug the cords after each cord is tested, which requires shutting the system down and restarting, which is a non-trivial task (and takes longer than it does to test one cord). John D. Flavin Teradata TCP Engineering 17095 Via del Campo San Diego, CA 92127 john.fla...@teradata.com V: (858) 485-3874 F: (213) 337-5432 - This message is from the IEEE Product Safety Engineering Society emc-pstc discussion list. Website: http://www.ieee-pses.org/ To post a message to the list, send your e-mail to emc-p...@ieee.org Instructions: http://listserv.ieee.org/request/user-guide.html List rules: http://www.ieee-pses.org/listrules.html For help, send mail to the list administrators: Scott Douglas emcp...@ptcnh.net Mike Cantwell mcantw...@ieee.org For policy questions, send mail to: Jim Bacher: j.bac...@ieee.org David Heald: emc-p...@daveheald.com All emc-pstc postings are archived and searchable on the web at: http://www.ieeecommunities.org/emc-pstc
Re: Placement of LISNs for Conducted Emissions Testing to CISPR22/FCC part 15
In message 809376ab0310a746b991380cb827f8e6017d1...@susday7659.td.teradata.com, dated Tue, 3 Jun 2008, Flavin, John john.fla...@teradata.com writes: Our question is how to place a relatively large number of LISNs and satisfy the standards' requirement of the 80cm spacing of the EUT and LISN. Specifically: 1) Are we allowed to place LISNs around all sides of the EUT, maintaining the 80cm spacing (i.e. have LISNs at the front face of the EUT, and run the mains cord from the back to the LISN)? I don't see why not. While you can't assume that you can do anything that isn't specifically prohibited, I think your proposal doesn't have a fatal technical defect. 2) Are we allowed to stack LISNs on top of each other, as long as the LISN is bonded to the ground plane? I think that is less proof against criticism. The inductances of the ground straps of the upper LISNs might affect the results. I think CISPR/I experts (there are some who lurk here) will bear in mind for the future that they have to allow for products with 18 mains leads. At least.(;-) -- OOO - Own Opinions Only. Try www.jmwa.demon.co.uk and www.isce.org.uk Either we are causing global warming, in which case we may be able to stop it, or natural variation is causing it, and we probably can't stop it. You choose! John Woodgate, J M Woodgate and Associates, Rayleigh, Essex UK - This message is from the IEEE Product Safety Engineering Society emc-pstc discussion list.Website: http://www.ieee-pses.org/ To post a message to the list, send your e-mail to emc-p...@ieee.org Instructions: http://listserv.ieee.org/request/user-guide.html List rules: http://www.ieee-pses.org/listrules.html For help, send mail to the list administrators: Scott Douglas emcp...@ptcnh.net Mike Cantwell mcantw...@ieee.org For policy questions, send mail to: Jim Bacher: j.bac...@ieee.org David Heald:emc-p...@daveheald.com All emc-pstc postings are archived and searchable on the web at: http://www.ieeecommunities.org/emc-pstc
Re: Placement of LISNs for Conducted Emissions Testing to CISPR22/FCC part 15
First a direct response to the question posed, then a challenge to the premise upon which it is based. The second comment, if legitimate, is more important than the first. You could use one dual LISN, or one LISN per current-carrying power conductor, if you had eighteen different make-before-break switches that would allow each power cord to draw current either from the LISN power output port, or the LISN input power side. If you want to go with eighteen LISNs, I think it is technically acceptable to stack them, but you want the ground strap to maintain a lower than 5:1 length-to-width ratio, so that likely means stacking no more than three high. But here’s an interesting and likely unwelcome thought, which I invite other forum members to comment upon. The point of meeting a conducted emissions requirement is to protect radios operating below 30 MHz that might be powered from the same branch circuit, or in the case of class A which likely applies here, operated within some distance of the equipment, but plugged into a different branch. If the equipment in your two racks operates simultaneously, it isn’t obvious to me that you are even allowed different LISNs – presumably all your rack equipment plugs into the same branch circuit, which should be represented by a single pair of LISNs. Immediate problem solved, but potentially more noise to filter, especially if power supplies operating off each cord operate at same switching frequencies. Ken Javor Phone: (256) 650-5261 From: Flavin, John john.fla...@teradata.com List-Post: emc-pstc@listserv.ieee.org List-Post: emc-pstc@listserv.ieee.org List-Post: emc-pstc@listserv.ieee.org Date: Tue, 3 Jun 2008 16:40:10 -0400 To: emc-p...@ieee.org Conversation: Placement of LISNs for Conducted Emissions Testing to CISPR22/FCC part 15 Subject: Placement of LISNs for Conducted Emissions Testing to CISPR22/FCC part 15 Our company sells ITE systems that are housed in commerial 19 racks. The system is designed to be fault tolerant and redundant, so each rack has two AC mains cords. We do our own EMI certification testing (we're an accreditted lab), and our typical EUT consists of two of these rack, so there are 4 AC Mains cords to test, which we connect to 4 LISNs. A modified version of this system is now in the works, where the dual AC mains cords are replaced by multiple cords (with lower current per cord). The design now would have 10 AC mains cords out of one rack, and 8 from the other. This means the two rack EUT would have 18 AC Mains cords to test. In a perfect world, where cost were no object, we would have 18 LISNs, since this is the most efficient for testing -- set it up once, and test everything. Our question is how to place a relatively large number of LISNs and satisfy the standards' requirement of the 80cm spacing of the EUT and LISN. Specifically: 1) Are we allowed to place LISNs around all sides of the EUT, maintaining the 80cm spacing (i.e. have LISNs at the front face of the EUT, and run the mains cord from the back to the LISN)? 2) Are we allowed to stack LISNs on top of each other, as long as the LISN is bonded to the ground plane? Since we have to test each cord in turn, we could reduce the number of LISNs by combining a number of the cords not currently being tested through a second (or third) LISN. The downside of this is having to re-plug the cords after each cord is tested, which requires shutting the system down and restarting, which is a non-trivial task (and takes longer than it does to test one cord). John D. Flavin Teradata TCP Engineering 17095 Via del Campo San Diego, CA 92127 john.fla...@teradata.com V: (858) 485-3874 F: (213) 337-5432 - This message is from the IEEE Product Safety Engineering Society emc-pstc discussion list.Website: http://www.ieee-pses.org/ To post a message to the list, send your e-mail to emc-p...@ieee.org Instructions: http://listserv.ieee.org/request/user-guide.html List rules: http://www.ieee-pses.org/listrules.html For help, send mail to the list administrators: Scott Douglas emcp...@ptcnh.net Mike Cantwell mcantw...@ieee.org For policy questions, send mail to: Jim Bacher: j.bac...@ieee.org David Heald: emc-p...@daveheald.com All emc-pstc postings are archived and searchable on the web at: http://www.ieeecommunities.org/emc-pstc - This message is from the IEEE Product Safety Engineering Society emc-pstc discussion list. Website: http://www.ieee-pses.org/ To post a message to the list, send your e-mail to emc-p...@ieee.org Instructions: http://listserv.ieee.org/request/user-guide.html List rules: http://www.ieee-pses.org/listrules.html For help, send mail to the list administrators: Scott Douglas emcp...@ptcnh.net Mike Cantwell mcantw...@ieee.org
RE: Placement of LISNs for Conducted Emissions Testing to CISPR22/FCC part 15
One LISN per power cord is acceptable. One power cord per LISN is required for the power cord being measured. That way you know that the emissions being measured are from that cord, and not another one. This was simpler with the old design – two cords comes out of the cabinet, each to its own LISN. Now John has to contend with a bunch of cords. This is addressed in CISPR 22, article 9.5.1, which states: “The mains cable of the unit being measured shall be connected to one artificial mains network (AMN). Where the EUT is a system, which is a collection of ITE with one or more host units, and each item has its own power cable, the point of connection for the AMN is determined by the following rules: a) Each power cable that is terminated in a power supply plug of a standard design (IEC 60083 for example) shall be tested separately.” There is no question that the power cords are tested one at a time. A later paragraph in 9.5.1 calls for one or more additional AMNs for the additional power cables. Article 7.2.1 of ANSI C63.4:2003 has different text that conveys the same message. So, if your thought is to be accepted, both ANSI C63.4 and CISPR 22 (and probably other standards, as well) will have to be changed. Given the success in reducing or largely eliminating interference from ITE that the current standards have demonstrated over the past 20+ years, I doubt that will gain much traction. At least, I certainly hope not. :-) Ghery S. Pettit, NCE Convener, CISPR SC I WG3 Member, C63 SC 1 From: emc-p...@ieee.org [mailto:emc-p...@ieee.org] On Behalf Of Ken Javor Sent: Tuesday, June 03, 2008 2:09 PM To: Untitled Subject: Re: Placement of LISNs for Conducted Emissions Testing to CISPR22/FCC part 15 First a direct response to the question posed, then a challenge to the premise upon which it is based. The second comment, if legitimate, is more important than the first. You could use one dual LISN, or one LISN per current-carrying power conductor, if you had eighteen different make-before-break switches that would allow each power cord to draw current either from the LISN power output port, or the LISN input power side. If you want to go with eighteen LISNs, I think it is technically acceptable to stack them, but you want the ground strap to maintain a lower than 5:1 length-to-width ratio, so that likely means stacking no more than three high. But here’s an interesting and likely unwelcome thought, which I invite other forum members to comment upon. The point of meeting a conducted emissions requirement is to protect radios operating below 30 MHz that might be powered from the same branch circuit, or in the case of class A which likely applies here, operated within some distance of the equipment, but plugged into a different branch. If the equipment in your two racks operates simultaneously, it isn’t obvious to me that you are even allowed different LISNs – presumably all your rack equipment plugs into the same branch circuit, which should be represented by a single pair of LISNs. Immediate problem solved, but potentially more noise to filter, especially if power supplies operating off each cord operate at same switching frequencies. Ken Javor Phone: (256) 650-5261 From: Flavin, John john.fla...@teradata.com List-Post: emc-pstc@listserv.ieee.org List-Post: emc-pstc@listserv.ieee.org List-Post: emc-pstc@listserv.ieee.org Date: Tue, 3 Jun 2008 16:40:10 -0400 To: emc-p...@ieee.org Conversation: Placement of LISNs for Conducted Emissions Testing to CISPR22/FCC part 15 Subject: Placement of LISNs for Conducted Emissions Testing to CISPR22/FCC part 15 Our company sells ITE systems that are housed in commerial 19 racks. The system is designed to be fault tolerant and redundant, so each rack has two AC mains cords. We do our own EMI certification testing (we're an accreditted lab), and our typical EUT consists of two of these rack, so there are 4 AC Mains cords to test, which we connect to 4 LISNs. A modified version of this system is now in the works, where the dual AC mains cords are replaced by multiple cords (with lower current per cord). The design now would have 10 AC mains cords out of one rack, and 8 from the other. This means the two rack EUT would have 18 AC Mains cords to test. In a perfect world, where cost were no object, we would have 18 LISNs, since this is the most efficient for testing -- set it up once, and test everything. Our question is how to place a relatively large number of LISNs and satisfy the standards' requirement of the 80cm spacing of the EUT and LISN. Specifically: 1) Are we allowed to place LISNs around all sides of the EUT, maintaining the 80cm spacing (i.e. have LISNs at the front face of the EUT, and run the mains cord from the back to the LISN)? 2) Are we allowed to stack LISNs on top of each other, as long as the LISN is bonded to the ground plane?
Re: Placement of LISNs for Conducted Emissions Testing to CISPR22/FCC part 15
John, Way back when, I worked on something similar. One power cord of sufficient size could run the whole thing but the cleanup crew had a tendency unplug it to run the vacuum cleaner. So redundant cords were put into place, which if plugged into the same branch circuit, all stopped when the floor buffer over powered that branch. Now the cords had to plugged into different branches... and so on... Fault tolerant and redundant? I took that to mean that any one of the power cords could be pulled and the system would still run. If so, then a little musical chairs with the power cords should get them rearranged so the system never shuts down. If a number of these power cords are meant as parallel connections to reduce the current per cord then I would say that they are indeed just one cord and plug all of one parallel set into one LISN. Quickly you'll see that presents a challenge in that now there needs to be plug strip to accept these cords. The solution is left to the creative lab tech. Is the 80cm length for the power cord under test only, leaving the other cords as long as necessary, or it the 80cm for all the power cords. I thought it was just the one cord (or set of cords) under test. All this just to keep the building wiring from broadcasting the conducted noise from a system. Ken Javor ken.ja...@emccompliance.com wrote: First a direct response to the question posed, then a challenge to the premise upon which it is based. The second comment, if legitimate, is more important than the first. You could use one dual LISN, or one LISN per current-carrying power conductor, if you had eighteen different make-before-break switches that would allow each power cord to draw current either from the LISN power output port, or the LISN input power side. If you want to go with eighteen LISNs, I think it is technically acceptable to stack them, but you want the ground strap to maintain a lower than 5:1 length-to-width ratio, so that likely means stacking no more than three high. But here’s an interesting and likely unwelcome thought, which I invite other forum members to comment upon. The point of meeting a conducted emissions requirement is to protect radios operating below 30 MHz that might be powered from the same branch circuit, or in the case of class A which likely applies here, operated within some distance of the equipment, but plugged into a different branch. If the equipment in your two racks operates simultaneously, it isn’t obvious to me that you are even allowed different LISNs – presumably all your rack equipment plugs into the same branch circuit, which should be represented by a single pair of LISNs. Immediate problem solved, but potentially more noise to filter, especially if power supplies operating off each cord operate at same switching frequencies. Ken Javor Phone: (256) 650-5261 From: Flavin, John john.fla...@teradata.com Date: Tue, 3 Jun 2008 16:40:10 -0400 To: emc-p...@ieee.org Conversation: Placement of LISNs for Conducted Emissions Testing to CISPR22/FCC part 15 Subject: Placement of LISNs for Conducted Emissions Testing to CISPR22/FCC part 15 Our company sells ITE systems that are housed in commerial 19 racks. The system is designed to be fault tolerant and redundant, so each rack has two AC mains cords. We do our own EMI certification testing (we're an accreditted lab), and our typical EUT consists of two of these rack, so there are 4 AC Mains cords to test, which we connect to 4 LISNs. A modified version of this system is now in the works, where the dual AC mains cords are replaced by multiple cords (with lower current per cord). The design now would have 10 AC mains cords out of one rack, and 8 from the other. This means the two rack EUT would have 18 AC Mains cords to test. In a perfect world, where cost were no object, we would have 18 LISNs, since this is the most efficient for testing -- set it up once, and test everything. Our question is how to place a relatively large number of LISNs and satisfy the standards' requirement of the 80cm spacing of the EUT and LISN. Specifically: 1) Are we allowed to place LISNs around all sides of the EUT, maintaining the 80cm spacing (i.e. have LISNs at the front face of the EUT, and run the mains cord from the back to the LISN)? 2) Are we allowed to stack LISNs on top of each other, as long as the LISN is bonded to the ground plane? Since we have to test each cord in turn, we could reduce the number of LISNs by combining a number of the cords not currently being tested through a second (or third) LISN. The downside of this is having to re-plug the cords after each cord is tested, which requires
Re: Placement of LISNs for Conducted Emissions Testing to CISPR22/FCC part 15
Not going to argue chapter and verse of the standards, nor make a case they should be changed – been there, done that, waste of time. However, they have departed significantly from the original intent and that leaves one asking: “Why are we even doing this?” The original intent of conducted emission requirements was two-fold. One was protecting against ripple on the bus that could affect radio reception by direct conduction through the power supply from the mains. That (FCC) limit was originally 48 dBuV for class B. Secondarily, that limit protected against electromagnetic radiation from the mains due to rf currents, and that protected against radiated interference in a frequency range where it would be quite difficult to make radiated measurements. If you have a single device that plugs into a single branch circuit, albeit through multiple power supplies and multiple power cords, then all those cords should indeed plug into a single pair of LISNs. Now if the power supply connected to each one of those cords is totally different from each and every other power supply, there is likely no harm done, because in any single 9 kHz channel, you likely won’t get any superposition. But if the design were such that all p ower supplies were identical, then their emissions would add in quadrature (assuming they are not phase-locked) and 18 of them would yield 12.5 dB more signal in each 9 kHz bandwidth occupied by a clock harmonic than for a single power supply running off a single LISN. Another issue, prevalent at and around 150 kHz, is rectification harmonics. These will be directly proportional to total current draw, so that splitting the total current amongst many LISNs makes it much easier to meet the limit, and this isn’t justified if the equipment plugs into a single branch circuit. This isn’t a problem in the USA, because we have no BCB receivers below 530 kHz, but it affects European BCB reception in the LW band from 150 – 300 kHz. My opinion, FWIW, is that if the standards ignore the original intent, then the standards are an end in themselves with degraded contribution to the original intent to foster controlled levels of rfi. Ken Javor Phone: (256) 650-5261 From: Pettit, Ghery ghery.pet...@intel.com List-Post: emc-pstc@listserv.ieee.org List-Post: emc-pstc@listserv.ieee.org List-Post: emc-pstc@listserv.ieee.org Date: Tue, 3 Jun 2008 14:21:31 -0700 To: Ken Javor ken.ja...@emccompliance.com, Untitled emc-p...@ieee.org Conversation: Placement of LISNs for Conducted Emissions Testing to CISPR22/FCC part 15 Subject: RE: Placement of LISNs for Conducted Emissions Testing to CISPR22/FCC part 15 One LISN per power cord is acceptable. One power cord per LISN is required for the power cord being measured. That way you know that the emissions being measured are from that cord, and not another one. This was simpler with the old design – two cords comes out of the cabinet, each to its own LISN. Now John has to contend with a bunch of cords. This is addressed in CISPR 22, article 9.5.1, which states: “The mains cable of the unit being measured shall be connected to one artificial mains network (AMN). Where the EUT is a system, which is a collection of ITE with one or more host units, and each item has its own power cable, the point of connection for the AMN is determined by the following rules: a) Each power cable that is terminated in a power supply plug of a standard design (IEC 60083 for example) shall be tested separately.” There is no question that the power cords are tested one at a time. A later paragraph in 9.5.1 calls for one or more additional AMNs for the additional power cables. Article 7.2.1 of ANSI C63.4:2003 has different text that conveys the same message. So, if your thought is to be accepted, both ANSI C63.4 and CISPR 22 (and probably other standards, as well) will have to be changed. Given the success in reducing or largely eliminating interference from ITE that the current standards have demonstrated over the past 20+ years, I doubt that will gain much traction. At least, I certainly hope not. :-) Ghery S. Pettit, NCE Convener, CISPR SC I WG3 Member, C63 SC 1 From: emc-p...@ieee.org [mailto:emc-p...@ieee.org] On Behalf Of Ken Javor Sent: Tuesday, June 03, 2008 2:09 PM To: Untitled Subject: Re: Placement of LISNs for Conducted Emissions Testing to CISPR22/FCC part 15 First a direct response to the question posed, then a challenge to the premise upon which it is based. The second comment, if legitimate, is more important than the first. You could use one dual LISN, or one LISN per current-carrying power conductor, if you had eighteen different make-before-break switches that would allow each power cord to draw current either from the LISN power output port, or the LISN input power side. If you want to go with eighteen LISNs, I think it is technically acceptable to stack them, but you want
Re: Placement of LISNs for Conducted Emissions Testing to CISPR22/FCC part 15
tongue in cheek mode ON Are we not as EMC engineers supposed to interpret the standards to our economic advantage so we can ship products without delays for redesign and still have some rational for explaining why we did what we did? After all, if there is no foul, there is no fault, because I haven't heard any complaints. That is, the FCC has not yet written me that letter, again. If the license holder have a problem, I'll hear about it. If not, keep shipping. HAMS be damned. Oh wait, news flash, the military uses low frequencies?? who ever heard of such a thing? Why in the world would they want to talk to a submarine? It's underwater! tongue in cheek mode OFF, and time for another beer. c ya, --- Ken Javor ken.ja...@emccompliance.com wrote: Not going to argue chapter and verse of the standards, nor make a case they should be changed been there, done that, waste of time. However, they have departed significantly from the original intent and that leaves one asking: ³Why are we even doing this?² The original intent of conducted emission requirements was two-fold. One was protecting against ripple on the bus that could affect radio reception by direct conduction through the power supply from the mains. That (FCC) limit was originally 48 dBuV for class B. Secondarily, that limit protected against electromagnetic radiation from the mains due to rf currents, and that protected against radiated interference in a frequency range where it would be quite difficult to make radiated measurements. If you have a single device that plugs into a single branch circuit, albeit through multiple power supplies and multiple power cords, then all those cords should indeed plug into a single pair of LISNs. Now if the power supply connected to each one of those cords is totally different from each and every other power supply, there is likely no harm done, because in any single 9 kHz channel, you likely won¹t get any superposition. But if the design were such that all power supplies were identical, then their emissions would add in quadrature (assuming they are not phase-locked) and 18 of them would yield 12.5 dB more signal in each 9 kHz bandwidth occupied by a clock harmonic than for a single power supply running off a single LISN. Another issue, prevalent at and around 150 kHz, is rectification harmonics. These will be directly proportional to total current draw, so that splitting the total current amongst many LISNs makes it much easier to meet the limit, and this isn¹t justified if the equipment plugs into a single branch circuit. This isn¹t a problem in the USA, because we have no BCB receivers below 530 kHz, but it affects European BCB reception in the LW band from 150 300 kHz. My opinion, FWIW, is that if the standards ignore the original intent, then the standards are an end in themselves with degraded contribution to the original intent to foster controlled levels of rfi. Ken Javor Phone: (256) 650-5261 From: Pettit, Ghery ghery.pet...@intel.com Date: Tue, 3 Jun 2008 14:21:31 -0700 To: Ken Javor ken.ja...@emccompliance.com, Untitled emc-p...@ieee.org Conversation: Placement of LISNs for Conducted Emissions Testing to CISPR22/FCC part 15 Subject: RE: Placement of LISNs for Conducted Emissions Testing to CISPR22/FCC part 15 One LISN per power cord is acceptable. One power cord per LISN is required for the power cord being measured. That way you know that the emissions being measured are from that cord, and not another one. This was simpler with the old design two cords comes out of the cabinet, each to its own LISN. Now John has to contend with a bunch of cords. This is addressed in CISPR 22, article 9.5.1, which states: ³The mains cable of the unit being measured shall be connected to one artificial mains network (AMN). Where the EUT is a system, which is a collection of ITE with one or more host units, and each item has its own power cable, the point of connection for the AMN is determined by the following rules: a) Each power cable that is terminated in a power supply plug of a standard design (IEC 60083 for example) shall be tested separately.² There is no question that the power cords are tested one at a time. A later paragraph in 9.5.1 calls for one or more additional AMNs for the additional power cables. Article 7.2.1 of ANSI C63.4:2003 has different text that conveys the same message. So, if your thought is to be accepted, both ANSI C63.4 and CISPR 22 (and probably other standards, as well) will have to be changed. Given the success in reducing or largely eliminating interference from ITE that the current standards have demonstrated over the past 20+ years, I doubt that will gain much traction. At least, I certainly hope not. J Ghery S. Pettit, NCE Convener, CISPR SC I WG3 Member, C63 SC 1 From: emc-p...@ieee.org