This page is an archive from our previous website. Please check out our new website where you can read new COMMENTARY eNewsletters, TELL IT LIKE IT IS blog posts or Press Releases.
The old radio guys need to know more about IP, but more importantly, I think, the next generation with its knowledge of all things IP needs to understand more about RF

Fighting Interference

Monday, February 02, 2009

I have been volunteering some of my time to help both the Santa Barbara Amateur Radio Club and The Nature Conservancy (TNC) to track down some radio interference they have been experiencing. It has been a rewarding experience, though dealing with problems such as these can be very frustrating as well. Since we have been discussing interference as it relates to new spectrum uses, I thought it appropriate to share my recent experiences with you, my readers.


The main radio site on Santa Cruz Island is on Mt. Diablo (elevation 2,474 feet) approximately 25 miles from Santa Barbara, and it was first developed by the original landowner to provide telephone service to the island. In his will, he left the bulk of the island to The Nature Conservancy and the site is now home to a Verizon Wireless micro cell, a lot of ham radio equipment, several weather stations, a U.S. Park Service radio, several aircraft remote radios, and a tower-mounted camera that transmits over Wi-Fi back to the mainland. This camera can be controlled and rotated via the Internet. A new VHF repeater was recently installed on the island for use by the people who are tracking the foxes, and those replacing non-native vegetation with native trees and shrubs.


All of the equipment is run by batteries with solar panels to charge the systems, and Verizon has an onsite generator. There are only two ways to reach this site. The first is to take a boat or small fixed-wing aircraft to the island and then hike up a very steep trail (not for me). The second is to take a helicopter over to the mountain where it lands about 100 yards away and this is the only practical way to get equipment to the site.
 I became involved in the project after TNC's repeater had been installed and the receiver was hearing a lot of noise on the channel. The radio was brought back to the mainland and two of us went through it from top to bottom, ran it for awhile here, and declared it ready to return to the site. This time I was able to go along. It was my first trip to the island and the helo held four of us including the pilot, the radio, and our test equipment.

We lugged the radio up the hill and re-installed it in the rack, connected the duplexer, and turned it on. To our dismay, we had the same problem-heavy interference on the receive channel that seemed to be there all the time, indicating it was not being generated by a transmitter that was only on intermittently. We started troubleshooting the problem-we swept the antenna, changed it out with an antenna from another radio, and checked the duplexer, but still could not find the problem. Since the helo was being paid for by the hour by a company with an aviation radio on the site, we tended to that radio and made our way back to the helo, disappointed that we had not cured the problem. The radio was usable, but not really to handhelds on the far side of the island.


We can't rent a helo and go to the site anytime we want, we have to find someone to pay for the trip, so we knew we had to find the problem somehow before we returned to the site. A gentleman who runs a not-for-profit company spends a lot of time on the island with volunteers replacing the vegetation is also a ham. The next time he went out to the island, he hiked to the radio site and we began trying to isolate the interference.


First he unplugged the weather station, no change in the noise level; then he shut off all the ham radio gear, no change; then he turned off each radio at the site with the same results, no change. We were in radio contact with him during this process and one ham was generating a radio signal from his radio room using a signal generator with a power amp to transmit a tone at a specific, low level so we would hear it when we found the problem. We tried a few other things and then he pulled the A/C power strip that powers the camera with its Wi-Fi transceivers and the router/switch.


Instantly, the noise stopped! Finally, we had a clue as to what was causing the problem. He then put the plugs back into the strip one at a time until the noise began again, and identified the source as the Ethernet router/switch. Now we knew the cause of the problem, but not how to fix it. On the next trip back to the site, the router was wrapped in aluminum foil and that helped reduce the noise, but not enough. When the Ethernet cables were unplugged one at a time, we found that they all contributed to the noise but the one to the camera was causing the most noise. Since all of the Wi-Fi nodes can be remotely controlled via a radio link from the mainland, we were later able to verify our findings.


We have pinpointed the problem, but the fix will be interesting. We will take a new router/switch out on our next trip and, hopefully, some filters for the Ethernet cables, and we will also take some shielded Ethernet cables in case we have to run all new ones including the one to the camera and the weather station mounted on one of the towers. But we are convinced we will be able to fix the problem.


Interference is one of the major reasons radio systems stop working or degrade. Most of the time, the interference comes and goes, making it even more difficult to find, since it can be caused by a transmitter that has gone out of tune, a combination of transmitters that are mixing on a rusty bolt, rusty barbed wire, etc., or any number of situations. When the interference is constant, as it was in this case, it was difficult to track down, but we knew when we found the problem without having to wait for the offending signals to appear again as with intermittent interference.


All of the radio transmitters, receivers, Wi-Fi nodes, and other devices at the site are within the parameters specified by the FCC, yet we still had a major case of interference. The Ethernet cables are acting as antennas and spewing the low power noise generated by the router/switch around the inside of the radio room. Why it is affecting only one receiver is part of the black magic of wireless. Had we been able to take a spectrum analyzer with us, we might have been able to see the noise, but we couldn't because of weight limitations and we had to do things by the seat of our pants, so to speak.


I have described what we had to go through to solve a simple interference problem to make a point. As we add more wireless systems, which are now full of computers and chips and most likely will be connecting via Ethernet cables as well as antenna cables, the chances for interference at any given radio site will increase. Those from the world of IT and computers certainly know how to put together a network, but they haven't been taught to think about RF interference. It is assumed that since the devices meet FCC specifications for spurious emissions, they will work as advertised and potential interference will be non-existent. In reality, as we found out the hard way, cables can act as antennas receiving a signal and interfere with data flowing down the cable, or as in our case, acting as an antenna to radiate noise generated by a router/switch that is well within FCC specifications.


How are we going to solve such problems once the RF-savvy "old timers" retire and are replaced by people who know more about IT and IP than RF? Some will say, "Cognitive Radios!" Okay, cognitive radios are designed to find usable channels and might someday really work as advertised, but they will not solve problems of interference, they will simply mask them. Today's radios are basically computers that transmit and receive RF energy, and if there is a problem, you simply pull out a board and replace it. The new generation of RF field personnel is not being trained to trouble shoot, but rather to replace modules.


The old radio guys need to know more about IP, but more importantly, I think, the next generation with its knowledge of all things IP needs to understand more about RF and such things as interference, why radio coverage changes depending on the time of year, why some places that had coverage yesterday don't today, and other common knowledge the old timers grew up with. Radio, especially, is part science and part black magic, which must be very difficult for someone coming out of the IP world to comprehend.


Andrew Seybold



COMMENTS: This is an archived post. Commenting is no longer available.

dave maples - 02/04/2009 06:56:09

Andrew: Two thoughts:

1.Be sure to take some ferrite cores with you to clobber common-mode interference from each of the IP devices. Take enough to put two or three cores on each end of each cable.

2. The FCC's spurious-emissions standard was written in a day when there were no clocks and no switchmode power supplies. Those rules need to be seriously revised to take into account current conditions. Manufacturers will fight this every way they can; they must be forced to lose.

Andrew Seybold - 02/04/2009 19:02:46

Dave--thanks, and the response to this article has been very interesting, both here and on the two-way radio site I post on (Private Wireless Forum on Yahoo). The number of other instances of the same problems is interesting, and the general feeling is that it is only going to get worse before it gets better.
The Part 15 rules which limit the amount of radiation which can be emitted from a computer, or router or what ever, were written, as Dave says, a long time ago, and today if you build a product you do not have to certify it, you only have to say that your meet the rules. The result is a LOT of equipment that does not meet the rules, and of course, when you plug in an Ethernet cable the cable just acts like an antenna.
More and more radio sties have switching power supplies, computers, processors, routers, switches and lots of Ethernet cable in them. I am willing to bet that the noise floor in most of these installations is high but no one has really noticed yet.
We are going to replace the router, and use shielded Ethernet cables, and also the ferrite cores Dave mentioned aboue and we are going to use drain wires at each connector and make sure that they are tied to a good solid ground. I will let you know how this works out and thanks fo all of you for your simialr stories and some really good advice about fixing the proglem, I even had an offer from an engineer who offered to come to the site on his own dime to help, and the offer of the loan of a handheld Sprectrum Analizer to take with us on our next trip to the island.