Jeff,
Thanks for supplying the links - big help.
First, let me start by saying that meeting FCC Part 15, Class B requirements
falls way short of what is needed for RFI suppression when coupling into a
radio system is involved. There are people on this reflector who live and work
in the world of ITE and related equipment, and in which those Class B
requirements are routinely met by computers, routers, and switches. Then there
are others who, like myself, work (or did work before retirement) in the worlds
of military and aerospace systems and were concerned with MIL-STD-461 and/or
DO-160. Within the frequency range of 30 MHz to 1 GHz, there are differences
of as much as 30 to 40 dB between the FCC requirements and the aerospace
requirements. It is those differences in radiated emission levels that can
cause real problems when the FCC levels are not adequate for a given
installation.
>From the information you gave in your initial post on this subject, I can not
>tell what the threshold level is between your digital system and the 2m system
>that is the victim of the interference. (The threshold refers to that level
>at which the interference would cease to exist for the 2m system.) As noted
>in the exchanges between Jim Brown, K9YC, and me, there are some major issues
>to consider in planning for 10 to 20 dB of attenuation versus 40 dB (or more).
> As briefly as possible, my comments on your proposed products are as follows:
1. The shielded RF-45 connectors should be better than unshielded ones, but it
appears that the cable shield path to chassis will be via the locking clip.
The rest of the path to chassis is then determined by the RJ-45 jack on the
equipment. Please allow me to clarify a point: You say that the grounding on
the jacks is low impedance at DC (I presume as measured with an ohmmeter).
Please keep in mind that impedance ("Z") is an AC value, resistance is a DC
value. Z of any given path is a function of the actual ohmic values of the
various conductors (summed) + Xl of all lead lengths (summed), which is a value
that is directly proportional to frequency. Thus, the actual Z of any chassis
grounding path is a complex value that will vary with speed (rep rate and
risetime) of the data.
2. As K9YC mentioned, the issue here is not dependant upon an actual "earth
ground" connection, although it is nice to see that the switch's enclosure has
such a terminal. That terminal might be of use if multiple pieces of equipment
are connected to it as a common system reference point, but that can cause
problems as well as help them.
3. Any part of the system that runs at 10 Base-T will be a LOT quieter than
the parts running at 100 Base-T (or faster). Reference: School of hard data
(or was it hard knocks?).
4. "Small" gaps are not inconsequential in many cases. Such gaps form slot
radiators, which can be very effective at propagating RFI. It depends upon how
much conversion takes place between differential mode (signal between the "hot"
and "return" leads of a pair) and common mode (coupled and/or induced energy
that rides on each conductor equally) and at what frequency (usually determined
by the risetime of the data). Both the gap width and length can be critical.
5. If you procure the given connectors, I would strongly suggest that you
visit K9YC's web site and carefully review his information and data on ferrites
and means to control emissions to and from cables. Apply the ferrites before
you install the connectors (unless you plan to leave plenty of length so that
connectors can be chopped off later to allow choke beads and absorbers to be
added, if needed).
6. Be advised that any of those ports on the switch that are not terminated
with a shielded cable and plug should by covered with copper or aluminum foil
or tape to reduce their radiation. (The copper or aluminum foil MUST make
direct contact with the bare metal of the switch housing. If there is paint in
the path, it is a waste of time.)
7. Plug everything together and see what you get. If you are lucky, the
problem will have fallen below the threshold and you can carry on. If not,
apply the ferrite materials and try again. Be sure to address BOTH ends of
each cable. Beyond that, you are then getting into the area of equipment mods
and some major issues.
Good luck. Keep us posted on how it goes.
73, Dale
WA9ENA
-----Original Message-----
>From: Jeff Stevens <jeff@mossycup.com>
>Sent: Mar 2, 2013 5:20 PM
>To: rfi@contesting.com
>Subject: Re: [RFI] RFI suppression cores for 147 MHz
>
>Thanks to everyone who has contributed to the discussion on the use of
>shielded twisted pair. I'm interested in what folks think about connection
>of the following devices in regard to providing a circumferential shield.
>They are a fairly good representation of what is easily available 'off the
>shelf' (and at reasonable cost) to the home user. I've not used either of
>these two exact products -- simply similarly constructed products. My
>belief has been that they provide what is effectively a circumferential
>shield. Sure, there are actually some small gaps but the connection is
>certainly low impedance (I've proven so at DC anyway) and the gaps are
>small relative to VHF wavelengths so I've understood them to be
>inconsequential.
>
>First, the shielded RJ-45 connector:
>http://sewelldirect.com/Cat6-Shielded-RJ45-Connectors-Stranded-50pc.asp?gclid=CMKYjtSO37UCFUZa4AodLUoAHA
>
>This is assembled and crimped exactly like an unshielded connector with the
>addition of insuring the STP foil shield and drain wire are in contact with
>the metal shield (which extends around the base of the connector and covers
>much (but not all) of the interior.
>
>Second is the networking device:
>http://www.amazon.com/Netgear-ProSafe-16-Port-Ethernet-Desktop/dp/B000063UZW
>
>The above device has ports with metal tabs on the sides which make contact
>with the shielded jack. Again, this provides a partial (but not complete)
>shield where the plug meets the jack. Note that the entire port is metal
>shielded as well. The metal chassis comes with an terminal on the rear
>intended for earthing.
>
>In both the case of the RJ-45 connector and ports on the device above, I've
>understood the very small gaps to be inconsequential. Does anybody have
>any thoughts?
>
>-Jeff
>W7WWA
>_______________________________________________
>RFI mailing list
>RFI@contesting.com
>http://lists.contesting.com/mailman/listinfo/rfi
_______________________________________________
RFI mailing list
RFI@contesting.com
http://lists.contesting.com/mailman/listinfo/rfi
|