Don, Decoupling the "high-speed" bus, while injecting the AM signal into the mains or any other "low-speed" interface, can be achieved in different ways", but it is more or less something you got to learn by experience. Although I never had any problems with decoupling devices if they were used on interfaces on which the AM signal was NOT coupled.
- Use common mode chokes (or ferrite clamps), but they only decouple and do not constitute the required 150 ohm CM impedance to ground, unless you configure your cabling in a way that the 150 ohm impedance is achieved (at higher freq.!) - Use commonly available networks like ESH3-Z4 (2-wire)or EZ-10 (4-wire) of R&S (other suppliers will also have these networks available). If your frequency range fits into the Passband of these networks, than you decouple and constitute a 150 ohm CM-mode impedance as required bij EN61000-4-6. For these networks, the 6dB Passband frequency is >10MHz@100 ohm symmetrical load, thus depending mainly on the DM termination impedance of your circuit. The 2nd option is quite suitable for products with a limited number of wires per interface. The first option might be usefull if the product has more wires per interface, but doesn't work quite well at he lower frequencies. Also consider the fact that it is not feasable to follow the requirements for decoupling - as stated in EN61000-4-6 - rigourously, if your product is equipped with a few hundred (or even more) wires. In this case, I wonder which laboratory has the capability to do so. A practical solution is then to leave the cabling as it is, use eventually ferrite absorbing clamps, do not implement the 150 ohm impedance to ground, but suppose that the cabling will introduce some CM impedance to ground and put all these data in the testreport. The problem you raise here is not only related to IEC61000-4-6, but also exists when applying p.e. surge-pulses on "high-speed" interfaces. Unfortunately, existing EMC standards are not taking into account these issues. Hope this helps! Chris On Wed, 02 August 2000, [email protected] wrote: > Chris, > > Thank you for your reply. > > In the situation where you are injecting into the mains but must still be > connected to the peripheral through the bus, how do you decouple the bus? > > Taking this one step further, it is my understanding that the ferrite clamp > does not decouple well at low frequencies, such as 150 kHz. It would seem > that the objective of the decoupling is twofold -- to protect the peripheral > from upset, and to insure that the peripheral does not corrupt the test > signal being coupled to the EUT. It does not seem that a ferrite based > decoupling clamp would be very good for these 2 points at frequencies below > about 10 MHz. What has been your experience in this regard? > > Even more curious, > > Don Umbdenstock > > > > ---------- > > From: Chris Collin[SMTP:[email protected]] > > Sent: Tuesday, August 01, 2000 7:12 PM > > To: [email protected] > > Cc: [email protected] > > Subject: Re: EN 61000-4-6 CDNs for a high speed bus. > > > > Don, > > > > If no CDN networks are suitable - and indeed they are not for high > > communication lines -, you should consider the possibility to use the > > EM-clamp as described in EN61000-4-6. That's the way we perform tests > > without degradation of the EUT signals. > > > > Regards, > > > > Chris > > > > On Tue, 01 August 2000, [email protected] wrote: > > > > > > > > > > > Hello Friends, > > > > > > We are getting into faster interconnects that can be affected by the > > CDNs > > > called out by EN 61000-4-6. > > > > > > Has anyone tested a system consisting of an EUT connected to various > > > peripherals by USB, Firewire, 10 Base T or other bus configurations > > > considered "high speed" today? It seems decoupling by an inductor > 280 > > uH > > > @ 150 kHz as indicated in the various CDNs in Annex D should cause > > > considerable signal integrity issues even before the test signal is > > applied. > > > > > > If you have tested high speed interconnects to EN 61000-4-6, how did you > > > manage the signal integrity issues? Special CDNs? Special test setup? > > > This seems like a topic everyone would be interested in. > > > > > > Just call me curious, > > > > > > Don > > > > > > (not George) > > > > > > ------------------------------------------- > > > This message is from the IEEE EMC Society Product Safety > > > Technical Committee emc-pstc discussion list. > > > > > > To cancel your subscription, send mail to: > > > [email protected] > > > with the single line: > > > unsubscribe emc-pstc > > > > > > For help, send mail to the list administrators: > > > Jim Bacher: [email protected] > > > Michael Garretson: [email protected] > > > > > > For policy questions, send mail to: > > > Richard Nute: [email protected] > > > > _______________________________________________________________________ > > > > Free Unlimited Internet Access! Try it now! > > http://www.zdnet.com/downloads/altavista/index.html > > > > _______________________________________________________________________ > > _______________________________________________________________________ Free Unlimited Internet Access! Try it now! http://www.zdnet.com/downloads/altavista/index.html _______________________________________________________________________ ------------------------------------------- This message is from the IEEE EMC Society Product Safety Technical Committee emc-pstc discussion list. To cancel your subscription, send mail to: [email protected] with the single line: unsubscribe emc-pstc For help, send mail to the list administrators: Jim Bacher: [email protected] Michael Garretson: [email protected] For policy questions, send mail to: Richard Nute: [email protected]
