Nerd Alert: A Discussion Of Meaningful RF Interference on CAT5 cabling

This is off the blog topic, but interesting nonetheless. The principles discussed in this article are SWAGs (Simply a Wild-Assed Guess), but I'm pretty smart, I'm sure a lot of them are right. I could use some attention from an expert on the subject (Ed's Note: The picture Wikipedia uses is pretty dorky), so by all means, leave me comments.

Today I was at one of my customers' site diagnosing a network connectivity issue. This customer specializes in RF welding of exotic plastics. They produce delicate, non-bioreactive plastics, like the plastic devices used in surgery or artificial body parts; and ultra-durable plastics, like Kevlar and other pliable armors. RF welding is a highly-specialized form of welding used when the media being welded is either very tough, or very delicate. It utilizes ultra-high-frequency (UHF) radio waves focused tightly on plastics placed under the high-pressure of a press or a mold, heating the plastic at the molecular level causing it to make a very strong bond. They specifically create very-high-quality artificial limbs and Kevlar liners for U.S. Military helmets and flak jackets.

The customer's site is an RF nightmare: high-intensity UHF radio waves leak out of the RF welding devices causing meaningful interference on all types of communication devices, like Wi-Fi networks, Ethernet networks, power grids and telephone (mobile and land line) systems. The way we have countered this issue in the past is to install Faraday Cages on all the cables running from the back of the PCs. The Faraday Cage intercepts the RF signal before it has a chance to interfere with the equipment.

Today we were having an issue with a new PC that was just installed last week. The PC has the standard Faraday Cages installed on it to reduce meaningful interference, but the Ethernet connection consistently errors out, or disconnects completely at times. I was confused why this was occurring until I noticed a few weird things. The cables (power, mouse, keyboard, Ethernet, USB) running from the back of the PC were all warm to the touch. Excepting power cables, and Ethernet cables under very heavy loads, none of these cables should be above the ambient air temperature in the room they are in. This machine wasn't experiencing any heavy workloads, so none of the cables should have been at a higher temperature than the comfortable 74F the room's air temperature was. I thought this was odd, and since Ethernet cables' nominal operating temperatures are between 60-100F*, I thought maybe this could be our issue. I unplugged the Ethernet cable, and grabbed the one I had brought with me. I had unpacked it and left it sitting on the workbench, unplugged and still wrapped up in plastic by the manufacturer. As I started opening the package, I noted that this cable was also warm to the touch. This cable, in the previous hour, had been in my air-conditioned car, and previous to that, my air-conditioned shop; so it should be, at the warmest, room temperature.

Apparently, this new PC was setup in a room that had previously been a storage room, with no electronics in it. The room also shares a wall with the area of the warehouse where the RF welding machines are located. The RF energy leaking from the poorly shielded welders was powerful enough to pass-through the Faraday Cages and convert itself from RF to thermal energy, heating the cable past its nominal operating temperature. I only noticed this when the cable that had never been used was almost hot-to-the-touch, and had no reason to be that warm.

This RF-thermal conversion I experienced at this customer's site is very similar, albeit at a much lower power, to the EMP produced by a nuclear detonation. When nuclear weapons are detonated they produce a huge amount of RF radiation, which is normally converted to thermal energy when detonated close to the ground, in thick atmosphere. When nuclear weapons are detonated at high-altitudes, the lack of atmosphere allows that high-intensity bust of RF waves to remain RF energy and not convert to thermal, and then propagate along ANY long antennas - any long runs of radio-conducting material, like: telephone wires, electrical wires, the copper wrapped around alternators, cell-phone antennas and even some long sewer pipes. These long "antennas" will capture the RF burst as it reaches them, and with no protection will convert the high energy RF into thermal energy, over-heating and shorting out the antenna. The only way to fix this damage is to replace the antenna.

There is not really anything I can do to counter this since the cable has to be in close proximity to the RF producing interference. The only option this customer has is to either move the equipment, or spend hundreds of dollars installing heavy-duty copper mesh on the wall of the room, and an air-conditioning system to remove the converted thermal energy from the room. If you don't add AC to a Faraday Cage exposed to constant RF energy, it can make the air-temperature get uncomfortably warm, or make the Faraday Cage itself short-out.

*CAT5 will operate anywhere between 0-167F, but will have degraded performance outside of human-comfortable room temperatures - 60-100F.