Zen and the Art of Lightning Protection

Dont let Lightning and power surges eat your PC or electronic equipment. Folks love to sell you some Very Fine $$$urge arre$tor$ to protect your equipment. But the amount of protection you get usually has nothing to do with the amount of money you spend on the surge protector. The surge protection specs are fine, but they are taken in laboratory conditions on a standard test bench configuration.

The real secret to surge protection is installing your wiring and equipment so it looks just like that laboratory bench ! ! !

Actually its not even a secret, Your phone company operates expensive switching equipment connected to wires coming in from miles around. How come they dont have to buy new equipment every time it thunders ? Back when the Earth was cooling, Ma Bell spent a fortune studying how to avoid costly lightning and surge damage, and BELCORE published the findings in the Network and Building Standards (NEBS). Its deadly dull reading, even for engineers, about how to get the best value for your money when protecting the CO (Central Office) from corrosion, fire, flood, vandalism, earthquakes and yes, lightning and surges.

How does NEBS say to prevent lightning and surge damage ? The little birdies told me.

The little birdie on the left has no problem standing on a power line carrying V1=4215432632 volts. The little birdie on the right is having a bad day because he is standing across two different lines, V1=4215432632 volts and V2=2364543 volts, or anything thats not the same as V1. Your equipment is a birdie, and its wiring are its feet. Ever notice how things with only one wire coming in almost never get lightning damage ?

CyberBirdies have tiny PCs and tiny generators so they can play Doom. These two are sitting on different wires, and want to play a game of Multi Doom using their tiny modems. What do you supose will happen when the modem cable is connected ???

These CyberBirdies have done well in the bond market and bought very expensive surge protectors. But ironically, the surge arrestors do not see any surges when the modem line is connected. The AC surge Protector on the left sees only the tiny generator voltage between any of its 3 wires. The modem line protector sees only the modem signal. The Bad Thing which is about to happen will go entirely unnoticed by any of the surge protectors, because it is a COMMON MODE surge, or as the NEBS says, a LONGITUDINAL SURGE. The problem is not a shortcoming of the surge protectors but of the equipment installation.

Good earth grounding is not the solution. How does an airplane work ? Not with an Earth Ground ! A Very Bad Thing would happen if someone tossed the birdies a Good Earth Ground.

These CyberBirdies have only cheap surge protectors, but their modem and power lines are routed thru PROTECTIVE GROUND WINDOWs. The Protective Ground Window (PGW) does not even need to be connected to Earth at all in order to give protection. Its function is to make sure ALL conductors are near the SAME voltage.

Its easy to design a PGW. Put the entire system you want to protect in a big sack, WITHOUT disconnecting any wires. The mouth of the sack will become your PGW. EVERY SINGLE CONDUCTOR entering or leaving the sack MUST be connected or protected directly to the PGW. Take a metal plate and connect every ground wire or shield to the plate. Connect a gas arc tube, sidactor, zener diode or whatever DIRECTLY from EVERY non grounded conductor to the plate. The plate is now your PGW, and everything in your system will be at the same voltage as the PGW. Whether or not the PGW is connected to Earth, all your birdie's little feet are now on the same wire. Happy Birdie.

Zen and the Art of EMC

ORIGINALLY WRITTEN TO HELP DSL USERS ISOLATE ELECTRICAL NOISE PROBLEMS AFFECTING RESIDENTIAL DSL SERVICE. OTHER ELECTROMAGNETIC COMPATIBILITY (EMC) FAQs TO FOLLOW< BUT RIGHT NOW ITS TIME FOR A BEER

ELECTRICAL POWER NOISE

Very common, caused by brush type motors, lighting ballasts, dimmers, controls, fence chargers, plus too many others to list here. The biggest enemy of my homebrew VLF receiver turned out to be a small battery charger for a DeWalt cordless drill. Dilligent detective work is needed to pinpoint the source.

Home in with an AM or shortwave radio. These waves are many meters to hundreds of meters long and they couple efficiently into long phone and power lines. ("shortwave" is a misleading term. Back when the earth was cooling, 1 MHz was considered a high frequency) At low freqs "foil windows" etc, are not directional, but the common "loop stick" antenna (ferrite rod wound with wire) has a far field NULL along the axis of the rod. Gimbal the radio to MINIMIZE the signal and the rod is pointing to the source. Strictly speaking this null applies only to a distant point source, but still its somewhat useful in near field conditions (at close range).

Your best bet for AM and shortwave homing is by signal strength. Since receiver AGC circuit prevents audio level from representing signal strength you need an RF attenuator. Wrap a long cardboard tube with foil, leaving one end open. Dangle a small receiver inside by a string. Whenever you find a large area of interference, lower the radio deeper into the tube, the evanscent mode of this open shield attenuates the signal rapidly, making the area of reception smaller and location more precise. You can locate the source to house(s) or building(s) served by a single transformer, then use circuit breakers to isolate further.

POWER LINE DISTRIBUTION NOISE

Sometimes power noise does not isolate to a single electrical load. The distribution network itself generates noise due to cracked or dirty insulators, loose or corroded splices, switches, fuseholders, missing bond wires, transformer winding-to-core leakage, etc.

Major aircraft manufacturer has a Flight Test Ground Station in Cobb, which was plagued by HF (shortwave) interference on its long range radios. Our 'trolls called it "The Crawlies" On a daily basis in summer there was a harsh AC buzz about 1 MHz wide starting around 3 MHz in the morining, crawling up to around 8 or 12 MHz by afternoon, and crawling back down in the evening. Heres the chase and fix: Using a portable SW receiver I located the source to within a half mile or so, field strength has several near-field lobes. Took a sledge hammer and knocked the &%#!& out of each pole suspected. Hitting a certain pole made the RFI go wild. Called the Atlanta Rd engineering office and gave them the street and the pole number (its on the pole) they were out the same day and fixed a bad splice. No More trouble.

Also a communications radio in our corp jet trailer was plagued by noise apparently from a fenced power yard. A week later Marietta had a major outage for several hours while a very large xfmr was replaced. No more trouble. 'nuff said.

These are very common problems, tricky to trace down and too many to fix them all. Dont point the finger until you can prove EXACTLY whats wrong. Be nice, but be Very Persistent.

AM BROADCAST INTERFERENCE

AM radio causes this type problem in dialup, DSL, audio, PA systems, etc. At AM broadcast wires (phone lines) make good antennas, and the AM signal is efficiently demodulated (received) by nonlinear junctions like corroded metallic joints, junction of dissimilar metals, semiconductor junction in electronic component. I spent weeks trying to pinpoint an intermittent problem only to realize it was caused by the aluminum ladder I was using to hunt for it. Telephone line balance is not much help, its the common mode (longitudinal mode for you telco types) which causes nearly all these problems.

Typically the broadcast transmitter is clean and within FCC rules. US law has no requirement for consumer devices to be immune from radio interference.

Ive cured some "impossible" cases of AM broadcast interference by wrapping all wires to the victim device thru a core of LOW FREQUENCY ferrite. This fixed modem problems near a Class 1 broadcaster where even sewing machines, blenders, light fixtures and fences were "receiving" the station in question.

EMI ferrite acts as an inductor at low freqs, and above a transition freq acts like a resistor. You want the transition freq to be below the freq of the interfernce. Dont waste your time with common hi-freq ferrite (type 43 from Fair-Rite) made to stop harmonics of digital devices. To STOP interference due to AM broadcast you need low-freq (Fair-Rite type 73), but its special order. So where can you get some ? There is ONE VERY COMMON APPLICATION for large "beads" of low frequency ferrite alloys: video deflection yokes. Yank out the yoke of an old monitor or TV, rape off the wire and you find a magnificent bead of low freq ferrite, it will cure AM broadcast RFI that no store-bought EMI filter can help in the slightest.

No, this wont affect your intended connection. The intended signals are normal mode (transverse mode for you telco types) so winding the whole cable thru the core makes an impedance that only appears to the common (longitudinal) mode, leaving the normal (transverse) signal unaffected.

MICROWAVE

DSL and modem problems are probably not microwave, those freqs dont propagate well in ordinary wires. A notable exeption is equipment mounted high up in line with large powerful radar where the PRF (Pulse Repetition Frequency) is demodulated by a nonlinear junction (corroded metallic junction, semiconductor junction in electronic component) then the PRF video (audio) can propagate in ordinary phone and power wiring. Listen for a short, harsh buzz or beep every time the dish rotates past you.

CATV LEAKAGE

Not a problem to DSL service but I have one tip anyway. Leaks in a closed system work both ways, in and out. When youre certain youve proved a CATV leak exists, call the carrier pretending to be a customer and complain that you "hear airplanes talking" on your TV. You will be amazed how fast they react when properly motivated.

Altair Home emc.html, Copyright © October 2001, Altair. EMail altair@altair.org