Lightning can really be a pickle

[ComputerNut]

Just had a rather large T-Storm pass over us in S. Ontario. I remember coming back from work, saying to myself, "I'd better shut off my puny PII server, and any other PC's in the house". One of the many precautions I take in a T-Storm, the last thing I want is all the caps on my mobo, flying everywhere.

So I go to work the next day, which is an I.T Department. I realize that soon after I left the previous day, the place nearly went to hell! One of the primary server's mobo got shot by the lightning (the UPS didnt quite save it there), at least several networking devices got shot in the building units behind us, and we had lost internet that entire day. First thing I thought is "Good thing I shut off my PII server" :D. So the entire day was mainly a day of trying to get the office back together. Still waiting for the internet techs to fix the outside internet feed.

Just thought I'd share my story of how lightning mixed with low-voltage electronics always bears rotten fruit. :cool:

[David M]

Makes ya wonder if lightning can arc thousands of feet how a UPS or surge supressor is going to stop it.

[edfair]

Surge suppressors can't stop it. They absorb a certain amount, to their limits, and they self immolate. The average surge suppressor on the average power line uses up the protection ability in a year and needs replacing.
The only sure method of protecting a computer is to unplug it from the power and disconnect everything else from it. I've seen systems killed from the induced voltages from printer cables, modem cables, and network cables.
My favorite story was from the lazer show at a local park. They used serial devices to control part of the display. The induced voltages from thunderstorms in the local area caused them to buy line drivers and receivers by the thousands.

[mbossman2]

theoretically, a faraday cage will protect equipment, but even those can be overwhelmed or have a minute break in them and ZZZZZZZZZZZZAAAAAAAAAAAAAAAAPPPPPPPPPP!

[rjfvillarosa]

As you can imagine, down here in the Caribbean electrical storms are quite frequent especially through hurricane season (begining of June till end of October).
I have now lost three or four motherboards, several PSU's and countless modems of my own despite using good quality surge protectors, so that I now make a point of unplugging my machines from both the electrical and phone supply at the first sign of a storm.
For the few years that i worked as a telecom engineer in the UK I know from experience that when a surge protector is killed it doesn't allways cut the service and therefore you don't know that you are running without protection, as stated above most surge protectors are probably useless after their first year of operation and you have no idea that you are not protected because there is no indicator to tell you, that little green "OK" LED is a liar. :eek:



I am not complaining mind you, because changing motherboards and modems is what keeps me employed down here. ;)

[HAL9000]

You guys need to learn to bury a few more power lines... it becomes much less of an issue. Small towns around here still have overhead lines and yes, they kill a lot of computer equipment, within the city, most lines are below ground and I found we didn't get a lot of repairs due to lightning strikes.

[RevCLB]

2-3 years ago a friend told me that tying a knot in the power cord from wall to device would be protection. Supposedly a Government installation lost all equipment except for a computer with the knot during a thunderstorm. Is this just another myth? I have knots in all my cords. It doesn't cost anything to tie a knot!!! LOL. RevClb

[Kareeser]

That doesn't make much sense. Even if you tie a knot in the cord, the wire will still be uninterrupted, since only the rubber/plastic casing is coming in contact with itself...

[HAL9000]

A lot of things in life don't make senese... so I guess until that theory can be proven wrong, what does it hurt to try?

[ComputerNut]

A lot of things in life don't make senese... so I guess until that theory can be proven wrong, what does it hurt to try?


Well to try and prove that theory, you might have a bunch of running systems during a severe T-Storm, have a few with a knot on the cord, some dont. Hope that a bolt will surge all the power outlets in the building, and see what happens. But of course, such an experiment will be costly with all the PC's that will be damaged expectedly.

[David M]

I would not bet my electronics on what seems to be an urban legend.

[blue60007]

Doubt it'd work either...might work for a garden hose, but not for an electrical cord...

[HAL9000]

I'm not saying bet your electronics... I'm saying for what it's worth, why not put the knot in there... a spike could come through literally at anytime, not just during a thunderstorm..... I doubt it would work either... I'm just saying what is the harm in trying.. don't take it so literally.

[David M]

Blue...Hah!...perfect analogy. :)

[RevCLB]

I didn't mean to start a war. I like scientific stuff but don't understand most of it. I just wondered if anyone eles had heard this. SORRY!!

[blue60007]

I thought it was an interesting idea. I agree with HAL. It'd be interesting to hear that the one equipment hooked up to a knotted cord survived a lighting strike.

[edfair]

That was luck of the draw. It wasn't time for that that power supply to die.
Burying power and phone cables cuts down on the direct hits and does wonders for the tree limb problems but doesn't help the induced current problem. And the spikes from the major transmission lines come across about the same as before.

[David M]

Only as much current is drawn as is what is needed. The voltage spike is the problem.

[edfair]

Granted that the spike does the damage, but the spike generally is from a high current somewhere close with the field cutting the transmission line and acting like a transformer. It does not require a direct hit.

[Stuey]

I'd be very very VERY surprised if a knot in the power cable is what saved the system (if in fact all of this is true).

*Theoretically* (as in I'm trying to rationalize this) it *could* be possible but almost certainly not with a single simple knot.

[HAL9000]

I doubt it too... I was just stating that for what it's worth, it couldn't hurt now could it LOL.

[Stuey]

I doubt it too... I was just stating that for what it's worth, it couldn't hurt now could it LOL.
Actually, doing say *may* have adverse effects on the other signal/power/speaker cables nearby.

[HAL9000]

OK, going to doubt that one too because I have a lot of cable coiled up in my shelf with my cable box and home theater, no adverse affects.