[RFI] router RFI

[Dale]

-----Original Message----- 
From: Dale 
Sent: Mar 4, 2013 10:55 AM 
To: jim at audiosystemsgroup.com, rfi at contesting.com 
Cc: ka5s at earthlink.net, paul at n1bug.com 
Subject: Re: [RFI] router RFI 


Jim,

Like Cortland, I have no info available to me on the work I did on Ethernet radiated emissions. That was early in my stint at Collins, and the actual data was considered to be proprietary. That said, I do recall the following:
 
1. Our only concern was the spectrum from 30 MHz to 1 GHz.  Thus, I can not comment on birdies, or spurs, in the HF spectrum.  I do not hear any digital "noise" on the HF bands at all, but bear in mind that all of my HF antennas are 50 to 250 feet from the LAN cables and hardware.  (Note: I have not done a full spectrum sweep from 1.8 MHz to 30 MHz to determine if, in fact, there are any spurs audible at my QTH.)    
 
2.  I do vividly recall that overall emissions were lower in amplitude and easier to control when the Ethernet was set to run at 10 Base-T speeds.  For example, one could obtain fairly decent performance of the shield on STP by use of the drain wire soldered to the metal shell of the connector, as opposed to the full circumferential shields needed for 100 Base-T systems.  I also recall that we did NOT use any RJ-45 jacks at all - interconnects were done with either Sub-D 9 pin connectors or mil-style circular connectors with pin inserts.  All connectors had back shells and everything was metal.
 
3.  We did find plenty of spurs in the VHF and UHF spectrum when running 100 Base-T.  I do not recall any specific frequencies.  As I reported in one of my posts, my LAN has a definite spur on 145.39 MHz, and I truly suspect that there are plenty more.  I just have not done a detailed run to see and log where all of them occur.  
 
Paul, that is most unfortunate about the situation at your County EC person.  Separation distance is an effective tool to reduce the effects of unwanted emissions on victim receivers.  If that person can not move the antennas and LAN further apart, then the few remaining choices, other than turning off the LAN, would include forcing the LAN to run only in 10 Base-T (if that is possible) or shielding the equipment associated with the LAN.  Question: Do you know if the problem exists because of radiation from the cables, or could it be both cables and hardware?  
 
Let me conclude by saying that reduced emissions from data networks and systems MUST start with the design of the hardware itself.  Unfortunately, general purpose ITE equipment that is designed for FCC Part 15 Class B or Class A service misses the mark for good control by a country mile.  It's simple economics: anything done that exceeds the required spec is money wasted.  It requires a lot of effort (spelled "$$$$") to get hardware into MIL-spec or aerospace spec compliance, although the actual cost can be reduced considerably by intelligent design of circuit cards, enclosures, interconnects, and choice of clock speeds.  The name of that concept is "Design to Pass the Requirements".  Part 15 devices do not meet those requirements - period.  Radio systems that must co-exist with high speed data systems have to take these considerations into mind and allow plenty of separation distance between the data hardware and the radio system antennas.  When there are limitations on the separation distance, sometimes the use of directional antennas can help, or mounting the antennas even higher than at present may help.  
 
I wish I had more hard data, but you have what I can recall.  Lots of experience taught me the bad news about data systems and ham radio.  As I said, I did not even bother with STP here at my QTH.  It's all UTP, Cat 5, and working fine.  In my case, the HT on my belt only squawks when I get within about 1 meter of the LAN cabling or hardware.  I can live with that.  
 
73, Dale
WA9ENA 
 



-----Original Message-----
>From: Jim Brown 
>Sent: Mar 3, 2013 11:34 PM
>To: rfi at contesting.com
>Subject: Re: [RFI] router RFI
>
>On 3/3/2013 8:01 PM, Roger (K8RI) wrote:
>> computers on cat6 megabit, as well as one net book, one lap top, and 
>> one printer on wireless. NADA, but I do hear two signals that appear 
>> to be routers spread across about 14 MHz clear up into the VHF range.
>
>Hi Roger,
>
>There are several common spurs that show up in CW bands, which is where 
>I mostly operate. They are good "beacons" for Ethernet trash. They are 
>around 14029.5, 21052.5, and the very low end of 10M (like 28014, etc.) 
>Each switch is free-running on it's own clock, so if you have neighbors 
>you'll hear these spurs for each of them on a slightly different 
>frequency. I heard at least 3-4 in Chicago, but out here in the woods, 
>with antennas pretty far from the house, I barely hear my own router 
>(which, of course, is choked).
>
>Dale and Cortland -- something you could do for us is provide a good 
>description of the spectra that typically results from Ethernet traffic 
>so that we can identify it. Those slightly modulated carriers (plus a 
>lot more than those listed here), and broadband hash on 2M, is all that 
>I've identified myself.
>
>73, Jim K9YC
>_______________________________________________
>RFI mailing list
>RFI at contesting.com
>http://lists.contesting.com/mailman/listinfo/rfi