Re: [SI-LIST] : Cavity Resonant frequecy and radiated EMI

About this list Date view Thread view Subject view Author view

From: Charles Grasso (chasgra@msn.com)
Date: Mon Sep 25 2000 - 11:28:12 PDT


Hi - I must agree with Steves analysis that the source of the problem
is not the cavity resonance but associated with the circuit.

Although there is no actual data signal on the wire itself, it is extremely
easy to conduct signal noise from the circuit board out of a shielded
enclosure either directly along the wire from the circuit board or have
noise coupled
 to the wire that is internal to the box and then conducted outsideof the
box.

A common error in shield rooms, for example, is to penetrate the room
with a shielded cable and not tie the shield to the room walls. The same
effect occurs.

Common-mode chokes designed for power supply filtering do not perform
well at these higher frequencies. Their performance is limited by the
material
of the ferrite and the interwinding capacitance. A more common approach is
is use a "balun". To wind one your self, twist two thin wires together and
wind them over a core of Type 28 material. That will make it work a hell
of a lot better. There are little tricks to be aware of in the installation
and construction so beware.

For other things to try:
    a) shield the portion of the cable that is internal to the box [This is
not
    recommended as it is difficult to implement sometimes. ]
    b) Add a capacitor to both the wires to the chassis at the point of exit
fr omthe
        box. (Try a 0.01uF)
    c) Isolate the device that is radiating onto the cable, if any, and add
a shield
    over the device.
    NOTE: Adding a capacitor may cause ESD/Immunity issues if the case
    is not properly designed.

----- Original Message -----
From: "KHOO,KENG-KOK (HP-Singapore,ex4)" <keng-kok_khoo@hp.com>
To: <si-list@silab.eng.sun.com>
Sent: Monday, September 25, 2000 12:21 AM
Subject: RE: [SI-LIST] : Cavity Resonant frequecy and radiated EMI

> Hi all,
>
> Thanks to Doug Smith and Steve for their input. I have came acrossed the
> Doug Smith discussion on impedance inversion where ferrite actually
increase
> EMI in a certain way.
>
> Maybe I should put in more detail for this problem. There are two cables
> running out of the box. Both are unshielded and there is no signal running
> along the cables because it's open all the time(it's a cable used for
ON/OFF
> purpose).
>
> I have tried soldering a common choke between the PCA and the cables but
> there was no effect on the EMI.
>
> Thanks a lot.
>
>
> Best Wishes,
> KHOO KK
>
>
> -----Original Message-----
> From: S. Weir [mailto:weirsp@atdial.net]
> Sent: Monday, September 25, 2000 2:42 PM
> To: si-list@silab.eng.sun.com
> Subject: Re: [SI-LIST] : Cavity Resonant frequecy and radiated EMI
>
>
> Keng-Kok,
>
> The resonant frequency of the box gives some measure of how difficult
> common mode suppression will be at certain frequencies. The closer you
get
> to the box resonance, the less attenuation you will get. However, your
> problem does not sound like box resonance.
>
> 100-150MHz emission is pretty broad-band. Usually, emissions will line up
> either on clocks and their harmonics, or on carriers.
>
> There are a couple of reasons that ferrite inside the box didn't
> work. Clamp-on ferrites work to defeat common mode noise between the sum
> of the conductors, and shield in the cable and earth. If your cable is
> shielded to the box beyond the ferrite, then you get attenuation between
> the box, and the driving end of the cable. But, since the ferrite is
> inside the box, you do not get attenuation between the box common mode and
> earth.
>
> If the cable is not well shielded to the box, then the answer may be in a
> brain-teaser presented by Doug Smith about 8 months back. Your 430mm
cable
> extension represents close to 1/4 wavelength in the 100-150MHz range for
> Cu. By placing a ferrite on the cable at the box exit, the cable has a
> high impedance at that end, and the 1/4 wavelength causes an impedance
> inversion at the other. So, the free end of the cable ends-up with a low
> impedance which does not couple well to free-space, and your EMI gets
> attenuated. If you double your cable length on the outside, you may find
> that you EMI goes up by more than 10db, with the same ferrite
configuration.
>
> Regards,
>
>
> Steve.
> At 10:12 AM 9/25/00 +0800, you wrote:
> >Hi all,
> >
> >I have somne questions on the cavity resonant frequencies. I found a
> formula
> >in Noise Reductions Techniques in Electronics System by Henry Ott Pg.199
> >that says
> >
> >f = 212/L,
> >
> >where:
> >L is the largest dimension of the cavity in meters
> >f the lowest resonant frequency in MHz
> >
> >I have a rectangular box with L=140mm so f = 1.514GHz.
> >
> >There is a cable running out of the enclosure and the cable lenght inside
> >the enclosure is 210mm and 430mm on the outdside of the enclosure.
Radiated
> >EMI measurement shows cable emission around 100-150MHz.
> >
> > Enclosure
> > 210mm | 430mm
> > --------------|----------------------------------
> > Inside | Outside
> >
> >Ferrite core works well in reducing the 100-150MHz emission when it is
put
> >on the outside of the enclosure but did NOT have any effect when put on
the
> >inside portion of the cable.
> >
> >Questions:
> >How do I relate the 1.5GHz resonant frequency to the 100-150MHz emission
> >from the cable?
> >Have I interpreted the formula wrongly?
> >Where did the 212/L formula came from?
> >
> >Thanks a lot.
> >
> >KHOO KK
> >
> >**** To unsubscribe from si-list or si-list-digest: send e-mail to
> >majordomo@silab.eng.sun.com. In the BODY of message put: UNSUBSCRIBE
> >si-list or UNSUBSCRIBE si-list-digest, for more help, put HELP.
> >si-list archives are accessible at http://www.qsl.net/wb6tpu
> >****
>
>
> **** To unsubscribe from si-list or si-list-digest: send e-mail to
> majordomo@silab.eng.sun.com. In the BODY of message put: UNSUBSCRIBE
> si-list or UNSUBSCRIBE si-list-digest, for more help, put HELP.
> si-list archives are accessible at http://www.qsl.net/wb6tpu
> ****
>
> **** To unsubscribe from si-list or si-list-digest: send e-mail to
> majordomo@silab.eng.sun.com. In the BODY of message put: UNSUBSCRIBE
> si-list or UNSUBSCRIBE si-list-digest, for more help, put HELP.
> si-list archives are accessible at http://www.qsl.net/wb6tpu
> ****
>
>

**** To unsubscribe from si-list or si-list-digest: send e-mail to
majordomo@silab.eng.sun.com. In the BODY of message put: UNSUBSCRIBE
si-list or UNSUBSCRIBE si-list-digest, for more help, put HELP.
si-list archives are accessible at http://www.qsl.net/wb6tpu
****


About this list Date view Thread view Subject view Author view

This archive was generated by hypermail 2b29 : Tue May 08 2001 - 14:29:34 PDT