Eric Brasseur Home    |    Links    |    Contact    |   

Insufficient lightning protection for electronic equipment

A friend owns sensitive high quality Hi-Fi equipment. He wanted to protect it against lightning strike surge peaks. I advised him towards my favorite local electronics store to buy a lightning protection device. A few time later his home was struck by lightning and one of his apparatus got damaged. Probably without the protection system things would have been way worse. Nevertheless I was quite annoyed the protection had not been perfect. So I started studying the problem. My friend's protection device could not be opened so I focused on the build-in lightning protection of most computer devices. Every computer I opened contains the same protection components: two blue 470 V varistors. One between neutral and ground and one between phase and ground:

The role of such a varistor is to become a shortcut should the tension between its two wires exceed 470 Volts. So the tension between the phase wire and the ground wire never can exceed +470 V or -470 V. As well as between the neutral wire and the ground wire. Any tension on the neutral or phase wire that would be lower than -470 V or higher than +470 V, whatever the source of that tension pike, whether it be a lightning strike or a condensator discharge, will be drained to the ground by the varistors. The current they can let through can be very high provided it has a short duration. The maximum tension present between 240 V AC normal mains wires is 339 V. So a 470 V varistor protection means the tension is not allowed to go 40% higher than the maximum normal value.

At first glance the protection seems perfect. Yet is is not. Because what matters for an electronic device power source is the voltage between the neutral and the phase. If the tension difference between neutral and ground will be held to a maximum of 470 V and the tension difference between ground and phase is held to a maximum of 470 V too, this means the tension difference between neutral and phase can rise up to 2 x 470 = 940 V. For example suppose at a given moment the phase tension rises to its normal maximum of +310 V. A lightning strikes and would tend to make the neutral wire go down to a tension of -10,000 V. The varistor between neutral and ground trades in and cuts the neutral wire tension back to a mere -470 V. This makes the tension between the phase wire and the neutral wire be 310 - -470 = 780 V. That is 150% higher than the maximum normal value! A though device will withstand such a tension pike. Yet a not a sensitive or fragile device.

Something needs to be done to force the tension between phase and neutral not to exceed say 470 V. The solution is simply to solder a third 470 V varistor between the neutral and phase wires:

Such a theory has to be tried out. I didn't want to wait for a lightning strike so I used an old VGA computer display that makes my computer crash. Sometimes, when I switch that old display on, the ATX power source of my computer goes into protection and suddenly halts the computer. This is due to tension peaks on the mains produced by the old VGA monitor being switched on. The two varistors my computer power source contains are not sufficient to protect the power source. Actually they do protect but not enough to prevent the power sources chooses to switch off. So I put a varistor between the neutral and phase wires of my computer and tried to switch the old VGA display on and off many times. The computer power source never switched off. So clearly a third varistor is a good thing.

Maybe there are technical reasons not to do so. But I did not think or hear of any serious one till now. A kind reader informs me a reason can be in some countries the neutral is wired to the ground, so the third varistor would be useless.


My computer system is now well protected against lightning strikes. Both the monitor, printer and computer are protected. But when I hear thunder I tend to unplug the modem. Because the path from the mains towards the phone line can be a way for a high current lightning strike. I'm not sure about this but I prefer not to take any risk. So I unplug the modem. Following the same reasoning I unplug my computer from the local network. Because there could be a lightning path between the networked computers. Good network and modem devices contain protections against lightning. But who knows. I could not check them to be 100% effective and I'm not ready to risk my computer to test this out. Finally, I make sure the computer, monitor and printer are on the same mains wall plug. That way there is no lightning path between two different wall plugs (for example should the computer be on one wall plug and the monitor on another wall plug situated another side of the room).

Eric Brasseur  -  Novembre 19 2002  till  May 5 2004