RE: R: R: Conducted Emissions on Telecom Ports
The defined I/O coupling methods for EN55022:1997 do not appear to accurately depict real-world shielding provided by "twisted" pair wiring, almost as if the test method were rigged against passing EMI with T-P cable. Considerable study went into development of twisted pair connectivity rules for each ANSI/IEEE 802.x LAN technology, emissions, immunity, cable grade etc., including coupling (remember TokenRing was 4 and 16 MHz, and Ethernet was 10 MHz so the harmonics were there). David __ Reply Separator _ Subject: Re: R: R: Conducted Emissions on Telecom Ports Author: "Ken Javor" at ADEMCONET List-Post: emc-pstc@listserv.ieee.org Date:9/8/2000 5:31 PM The point is that RE/CE protect broadcast bands. Making an RE measurement (E or H field, regardless) from a LAN line a couple meters long is not representative of what you would measure if the LAN line were significantly longer, as it might be in situ. Therefore a CE measurement can be better correlated to predicted RE from a much longer line (at frequencies where the tested LAN line is electrically short. -- >From: Paolo Roncone >To: "'Ken Javor'" , "'Cortland Richmond'" <72146@compuserve.com> >Cc: "'emc-p...@ieee.org'" >Subject: R: R: Conducted Emissions on Telecom Ports >Date: Fri, Sep 8, 2000, 3:51 AM > > Ken and Cortland and many others that entered this subject : > > First, radiated emissions are best measured with radiated (not conducted) > measurements. There is a correlation between CM currents and RE but that's > not all (resonances, cable layout etc. count a lot). > Second, you say that CE are easier to measure than RE ? Agree if you talk > about emissions on AC power cords (as per CISPR22 and FCC part 15). But for > the new requirements on telecom ports, I suggest you to take a look at the > new (3.ed.) CISPR22 or EN55022 (sec. 9.5 + annex C.1) and may be you change > your opinion ! > Radiated emissions above 30 MHz are already covered. > If you wanna take care of lower frequencies (< 30 MHz) take a loop antenna > (remember the old VDE rules ?) and measure radiated H-fields with your > system in the same (typical) layout used for the higher frequencies (with > whatever cables you specify, UTS, STP etc.). I am sure that is much > quicker, easier and repeatable than all the nonsense (ISNs, CDNs, clamps, > current probes, capacitive probes, ferrites, 150 ohm resistors, signal > generators, impedance measurements, voltage measurements, current > measurements and more) in the new CISPR22. > As for the question of "outside world", I think in this ever more connected > world the border line between INSIDE and OUTSIDE is getting more and more > blurred, BUT I also think that a line must be drawn by the standard bodies, > otherwise it's gonna really get too much confusing (hope some > CISPR/CENELEC member gets it). > If we spill over the line (office, floor, building... whatever), emissions > requirements are triggered. But within that line it's to be considered an > "intra-system" (what's the system ? that's another good question to be > settled) interference potential and the manufacturer should take care of it > without need of enforcement because he has all the interest in making a > product (system) that works properly and reliably. > > One last point: based on David Sterner's note, looks to me that North > America has a pretty extensive Ethernet and-the-like network. I honestly > don't know if the FCC has already enforced emission limits on LAN ports. > Anyway, based on David's note looks like there are no complaits of > interference with TV and telephones. And please note, this is the very > bottom line of it. Emission limits should be intended to protect public > services ... and physics works the same on both sides of the Atlantic... or > not > > My personal opinion ... > > Paolo > > > > > > > -Messaggio originale- > Da: Ken Javor [SMTP:ken.ja...@emccompliance.com] > Inviato: giovedì 7 settembre 2000 18.43 > A: Paolo Roncone; 'eric.lif...@ni.com' > Cc: 'emc-p...@ieee.org' > Oggetto: Re: R: Conducted Emissions on Telecom Ports > > Although I don't work commercial EMC on a regular basis and I do not know > over what frequency range the telecom port CE are controlled (I assume here > 150 kHz - 30 MHz), I believe that there is a mistaken premise inherent in > the comments to which I am responding. The purpose of controlling common > mode CE on any port is not to prot
Re: R: R: Conducted Emissions on Telecom Ports
The point is that RE/CE protect broadcast bands. Making an RE measurement (E or H field, regardless) from a LAN line a couple meters long is not representative of what you would measure if the LAN line were significantly longer, as it might be in situ. Therefore a CE measurement can be better correlated to predicted RE from a much longer line (at frequencies where the tested LAN line is electrically short. -- >From: Paolo Roncone >To: "'Ken Javor'" , "'Cortland Richmond'" <72146@compuserve.com> >Cc: "'emc-p...@ieee.org'" >Subject: R: R: Conducted Emissions on Telecom Ports >Date: Fri, Sep 8, 2000, 3:51 AM > > Ken and Cortland and many others that entered this subject : > > First, radiated emissions are best measured with radiated (not conducted) > measurements. There is a correlation between CM currents and RE but that's > not all (resonances, cable layout etc. count a lot). > Second, you say that CE are easier to measure than RE ? Agree if you talk > about emissions on AC power cords (as per CISPR22 and FCC part 15). But for > the new requirements on telecom ports, I suggest you to take a look at the > new (3.ed.) CISPR22 or EN55022 (sec. 9.5 + annex C.1) and may be you change > your opinion ! > Radiated emissions above 30 MHz are already covered. > If you wanna take care of lower frequencies (< 30 MHz) take a loop antenna > (remember the old VDE rules ?) and measure radiated H-fields with your > system in the same (typical) layout used for the higher frequencies (with > whatever cables you specify, UTS, STP etc.). I am sure that is much > quicker, easier and repeatable than all the nonsense (ISNs, CDNs, clamps, > current probes, capacitive probes, ferrites, 150 ohm resistors, signal > generators, impedance measurements, voltage measurements, current > measurements and more) in the new CISPR22. > As for the question of "outside world", I think in this ever more connected > world the border line between INSIDE and OUTSIDE is getting more and more > blurred, BUT I also think that a line must be drawn by the standard bodies, > otherwise it's gonna really get too much confusing (hope some > CISPR/CENELEC member gets it). > If we spill over the line (office, floor, building... whatever), emissions > requirements are triggered. But within that line it's to be considered an > "intra-system" (what's the system ? that's another good question to be > settled) interference potential and the manufacturer should take care of it > without need of enforcement because he has all the interest in making a > product (system) that works properly and reliably. > > One last point: based on David Sterner's note, looks to me that North > America has a pretty extensive Ethernet and-the-like network. I honestly > don't know if the FCC has already enforced emission limits on LAN ports. > Anyway, based on David's note looks like there are no complaits of > interference with TV and telephones. And please note, this is the very > bottom line of it. Emission limits should be intended to protect public > services ... and physics works the same on both sides of the Atlantic... or > not > > My personal opinion ... > > Paolo > > > > > > > -Messaggio originale- > Da: Ken Javor [SMTP:ken.ja...@emccompliance.com] > Inviato: giovedì 7 settembre 2000 18.43 > A: Paolo Roncone; 'eric.lif...@ni.com' > Cc: 'emc-p...@ieee.org' > Oggetto: Re: R: Conducted Emissions on Telecom Ports > > Although I don't work commercial EMC on a regular basis and I do not know > over what frequency range the telecom port CE are controlled (I assume here > 150 kHz - 30 MHz), I believe that there is a mistaken premise inherent in > the comments to which I am responding. The purpose of controlling common > mode CE on any port is not to protect equipment at the other end of the > cable, or other co-sited cables, but rather to control radiated emissions in > a frequency range in which CE are easier to measure than RE. In turn, the > purpose of controlling RE is to protect broadcast radio reception. > -- >>From: Paolo Roncone >>To: "'eric.lif...@ni.com'" >>Cc: "'emc-p...@ieee.org'" >>Subject: R: Conducted Emissions on Telecom Ports >>Date: Thu, Sep 7, 2000, 9:45 AM >> > >> >> Hi Eric, >> >> I 100% agree with you. The scope of emissions standard should be to protect >> the "outside" (i.e. public) environment from interference. So only ports >> that connect to public telecom networks should be covered by the standard. >> The problem is (as pointed out in one of the previous notes) that the new >> CISPR22 / EN55022 standard clearly includes LAN ports in the definition of >> telecommunications ports (section 3.6) no matter if they connect to the >> "outside world" or not. >> >> Regards, >> >> Paolo Roncone >> Compuprint s.p.a. >> Italy >> > --- This message is from the IEEE EMC Society Product Safety Technical Committee emc-pstc discussion list. To cancel your subscription, send mail to:
RE: R: R: Conducted Emissions on Telecom Ports
All, As Ghery reported before (lost in the recent threads, but copied below), the closely related issue cf definition creepage is being addressed by CISPR SC G and is already is CDV stage but not yet FDIS. If it isn't already too late, this might be the best or only opportunity we'll get for bring the issue up for discussion in a CISPR committee within the next couple of years. This would be the opportunity to bring together in some way (?) the designers of Ethernet and the CISPR committee, so whatever the outcome we can agree the issue was examined with due engineering dilligence. Regards, Eric Lifsey Please respond to "Pettit, Ghery" To: Eric Lifsey/AUS/NIC@NIC, emc-p...@ieee.org cc: Subject: RE: Conducted Emissions on Telecom Ports This sort of question has already come up in CISPR SC G (the owner of CISPR 22). There is a CDV (Committee Draft for Vote) being prepared that, if adopted as a Final Draft International Standard (FDIS), will put a halt to the definition creep that has been happening with this issue. The text doesn't get rid of LANs as a telecom port, but it does prevent administrations from calling things like RS-232 (yes, Australia has tried to justify this as a telecom port), USB, 1393, etc telecom ports. Nothing happens fast in the IEC, so don't hold your breath waiting for this change to happen, but we are working on it. When the CDV comes out there will be a voting period on it and if it passes, it will then be re-issued as an FDIS for final vote. I wouldn't expect any final action for at least a year or more. Ghery Pettit Intel Member, US CISPR G TAG --- This message is from the IEEE EMC Society Product Safety Technical Committee emc-pstc discussion list. To cancel your subscription, send mail to: majord...@ieee.org with the single line: unsubscribe emc-pstc For help, send mail to the list administrators: Jim Bacher: jim_bac...@mail.monarch.com Michael Garretson:pstc_ad...@garretson.org For policy questions, send mail to: Richard Nute: ri...@ieee.org
RE: R: R: Conducted Emissions on Telecom Ports
Paolo, You bring up an interesting point about FCC. FCC recognizes CISPR 22:1985 is as an alternative test method. The 1985 version does not specify emissions on LAN or telco. FCC Part 68 specifies conducted emissions only on mains cables over 450kHz to 30MHz with slightly different limits. There seems to be considerable interest in requesting a review of the need for conducted emissions requirements for LANS, not to mention installation cost (STP cost differential, clumsy routing, earthing considerations). What is our next step to get a formal review? David __ Reply Separator _ Subject: R: R: Conducted Emissions on Telecom Ports Author: "Paolo Roncone" at ADEMCONET List-Post: emc-pstc@listserv.ieee.org Date:9/8/2000 6:51 AM Ken and Cortland and many others that entered this subject : First, radiated emissions are best measured with radiated (not conducted) measurements. There is a correlation between CM currents and RE but that's not all (resonances, cable layout etc. count a lot). Second, you say that CE are easier to measure than RE ? Agree if you talk about emissions on AC power cords (as per CISPR22 and FCC part 15). But for the new requirements on telecom ports, I suggest you to take a look at the new (3.ed.) CISPR22 or EN55022 (sec. 9.5 + annex C.1) and may be you change your opinion ! Radiated emissions above 30 MHz are already covered. If you wanna take care of lower frequencies (< 30 MHz) take a loop antenna (remember the old VDE rules ?) and measure radiated H-fields with your system in the same (typical) layout used for the higher frequencies (with whatever cables you specify, UTS, STP etc.). I am sure that is much quicker, easier and repeatable than all the nonsense (ISNs, CDNs, clamps, current probes, capacitive probes, ferrites, 150 ohm resistors, signal generators, impedance measurements, voltage measurements, current measurements and more) in the new CISPR22. As for the question of "outside world", I think in this ever more connected world the border line between INSIDE and OUTSIDE is getting more and more blurred, BUT I also think that a line must be drawn by the standard bodies, otherwise it's gonna really get too much confusing (hope some CISPR/CENELEC member gets it). If we spill over the line (office, floor, building... whatever), emissions requirements are triggered. But within that line it's to be considered an "intra-system" (what's the system ? that's another good question to be settled) interference potential and the manufacturer should take care of it without need of enforcement because he has all the interest in making a product (system) that works properly and reliably. One last point: based on David Sterner's note, looks to me that North America has a pretty extensive Ethernet and-the-like network. I honestly don't know if the FCC has already enforced emission limits on LAN ports. Anyway, based on David's note looks like there are no complaits of interference with TV and telephones. And please note, this is the very bottom line of it. Emission limits should be intended to protect public services ... and physics works the same on both sides of the Atlantic... or not My personal opinion ... Paolo -Messaggio originale- Da: Ken Javor [SMTP:ken.ja...@emccompliance.com] Inviato: giovedì 7 settembre 2000 18.43 A: Paolo Roncone; 'eric.lif...@ni.com' Cc: 'emc-p...@ieee.org' Oggetto: Re: R: Conducted Emissions on Telecom Ports Although I don't work commercial EMC on a regular basis and I do not know over what frequency range the telecom port CE are controlled (I assume here 150 kHz - 30 MHz), I believe that there is a mistaken premise inherent in the comments to which I am responding. The purpose of controlling common mode CE on any port is not to protect equipment at the other end of the cable, or other co-sited cables, but rather to control radiated emissions in a frequency range in which CE are easier to measure than RE. In turn, the purpose of controlling RE is to protect broadcast radio reception. -- >From: Paolo Roncone >To: "'eric.lif...@ni.com'" >Cc: "'emc-p...@ieee.org'" >Subject: R: Conducted Emissions on Telecom Ports >Date: Thu, Sep 7, 2000, 9:45 AM > > > Hi Eric, > > I 100% agree with you. The scope of emissions standard should be to protect > the "outside" (i.e. public) environment from interference. So only ports > that connect to public telecom networks should be covered by the standard. > The problem is (as pointed out in one of the previous notes) that the new > CISPR22 / EN55022 standard clearly includes LAN ports in the definition of > telecommunications ports (section 3.6) no matter if they connect to the > "outside world" or not. > > Regards, > >
