CAT5 and induced voltage from lightening- is it possible?

Texun

Platinum Member
Oct 21, 2001
2,058
1
81
I work with cable modems, routers hubs and CAT5 every day. Each time there is an electrical storm I see an increase in dead NIC's and occasionally, a bad hub or modem port. A theory that I am researching is whether or not a nearby lightening strike can induce enough voltage in the unshielded CAT5 to blow a port. In almost every situation, the connectivity stops at the port. The NIC shows fine in the PC but will not ping, or the modem will show that it is not connected to the hub\router\PC when it actually is.

Looking at the cable I can see that the coax is shielded and the system grounded. The AC on the utility side is also as per spec. The only commonality is that the failure occurs one either side of the unshielded CAT5. Any thoughts on my theory of an induced charge ?


 

RemyCanad

Golden Member
Sep 28, 2001
1,849
0
0
At one of the campuses at work we have metal awanings over all of the walking paths. Because of lighting strikes to them and near by durring a large storm we fan loose up to 6 nics in the same fashion your discribing. The nic card is good only the port is bad. Sadly they are iMacs so it now motherboard time for them!
 

Menelaos

Senior member
Oct 10, 1999
251
0
0
Check out this link: EN61000-4-5 surge testing

Read the part in section 3.1 underneath figure 3A: "For example, surge testing ... their overhead line".

The voltage level for signal wires (as in CAT5) should be less, but you get the point. IIRC their are filters available for RJ45 ports, sorta like the UPS for mains.

Menel.
 

capybara

Senior member
Jan 18, 2001
630
0
0
Originally posted by: Menelaos
Check out this link: EN61000-4-5 surge testing

Read the part in section 3.1 underneath figure 3A: "For example, surge testing ... their overhead line".

The voltage level for signal wires (as in CAT5) should be less, but you get the point. IIRC their are filters available for RJ45 ports, sorta like the UPS for mains.

Menel.
i doubt any filter (at any reasonable cost anyway) can stop a bolt of lightning from running down 8 stands of awg 24 wire to ground, or if the wire is shielded, the shield is a fine
lightning conductor also. the damage done is usually cheaper to fix than to prevent.
 

RemyCanad

Golden Member
Sep 28, 2001
1,849
0
0
Buying a proctector for all 100+ imacs on this campus alone would be extremly more expincive than just replace the 500 dollars motherboard on them.

 

unclebabar

Senior member
Jun 16, 2002
360
0
0
I don't know if this is an urban legend or not, but the story is: people have been electrocuted during electrical storms while talking on the telephone. I imagine that RJ-45/Cat-5 cable will carry more current than RJ-11/telephone cable.

Now I'm not sure what you mean by inducing voltage, having failed out of EE school a long time ago, but I'd be confortable wagering money that the storm was a contributing factor in the disablement of at least some of the mentioned electrical equipment. Is there a deeply grounded lightning rod on the building housing the equiment? Does the awning you mentioned later have a safe path to ground? You could probably (and cheaply) solve the problem right there. Get the physical plant people to ground the awnings. otherwise, anti-static mats might work.

My (ridiculous, uninformed) theory is you're getting wicked static electricity due to the awnings and people passing through pick up a charge, which they will unwittingly transfer to your iMacs, and violia, fried apples.
 

RemyCanad

Golden Member
Sep 28, 2001
1,849
0
0
Hmm they could bring static electricty in with them but they never touch the iMac it self really just the mouse and keyboard....
 

unclebabar

Senior member
Jun 16, 2002
360
0
0
once again, purely mad speculation, but perhaps, just the introduction of a voltage potential (a person carrying a static charge) nearby is enough to cause a damaging current fluctation. Correct me if I'm wrong, but if you set two Van de graf generators next to each other, won't you get a spark? In the case of a person, the voltages must be orders of magnitude lower, but if there is a non-zero chance and with multiple occurances (multiple people using the lab) maybe it's enough to cause problems. You've got the iMac screen, that's got to have a static field. If it's higher potential than the person, then you'll get electrons jumping towards it to the closest conductor. Which probably is bad news if it's an unshielded cable.
 

Texun

Platinum Member
Oct 21, 2001
2,058
1
81
Thanks everyone for the discussion. I can't say it's static from the person. This has happened many times, and each time there are random differences as to the status of the equipment at the time the NIC or modem lost a port. By random I mean the PC may be on or off, and in the case of a LAN the network may be disconnected or connected to a modem. The only constant has been that a port blows somewhere in the LAN or at the modem when the CAT5 is connected and there is a lightening in the immediate area.

This is why I believed the weak link was the lack of shielding in the CAT5 and the fact that either the PC's or the modem would be a source for ground. I have no other explaination but I am open to other suggestions. Anyone else ever hear or see this after a storm???

 
Jun 26, 2002
185
0
0
Well considering the fact the lighting is a bridge from the positive electric field in the clouds to the negative field on the earth it might be the problem, if the lighting is really close. A huge static charge gets pulled from the earth up into the air before lighting stikes. This is why they say if your hair starts standing up in a storm run like hell because lighting is going to strike soon. So if a charge is pulled from the earth around the unshielded part of the cable then it might cause a problem. But I think the storm would have to start pulling the charge really close, like 10 feet to cause a problem. But I guess in the same train of thought the charge could then be drained though the Cat 5 cable to a ground, then preventing the lighting stike.

The shielding on the cable should prevent any charge being induced in the wire from the outside, but then again to shielding would have to be a perfect electrical conductor to prevent an induced charge. Since a PEC does not exist at room temp then if the electric field was strong enough then a charge could be induced in the line, but big enough to fry a NIC I don't think so because even with todays shielding over 95% of the voltage should be induced in the shielding.
 

ScottMac

Moderator<br>Networking<br>Elite member
Mar 19, 2001
5,471
2
0
Cat5 (UTP- unshielded twisted pair), with rare exception, is not rated for external use. In any case where copper is used outside, there should be entrance protection at both ends.

Entrance protection usually take the form of a "gas popper" or other breakdown device. Entrance protection provides a grounding path / drain in case of unusual power or static surges.

Avaya makes an external-rated Cat5, and offers matching entrance protection devices.

This is being / has been discussed a couple times in the Networking Forum.

FWIW

Scott
 

Ben50

Senior member
Apr 29, 2001
421
0
0
Everything I've heard says you should run fiber instead of copper when using it for outdoor situations.
 

Peter

Elite Member
Oct 15, 1999
9,640
1
0
Cat5 cabling is useless between buildings anyway, you got an allowed cable length of 100 meters, which isn't much. Cat5 cabling is shielded btw, so was the good old Cheapernet cable ... but there's still enough induction coming in from a lightning to easily zap a NIC's delicate physical interface circuitry. Same for modems - except that the phoneline cabling traditionally is unshielded.

regards, Peter
 

Peter

Elite Member
Oct 15, 1999
9,640
1
0
Cat5 cabling is useless between buildings anyway, you got an allowed cable length of 100 meters, which isn't much. Cat5 cabling is shielded btw, so was the good old Cheapernet cable ... but there's still enough induction coming in from a lightning to easily zap a NIC's delicate physical interface circuitry. Same for modems - except that the phoneline cabling traditionally is unshielded.

regards, Peter
 

ScottMac

Moderator<br>Networking<br>Elite member
Mar 19, 2001
5,471
2
0
The vast majority of Cat 3,4, 5, 5e, and 6 cabling is not shielded.

Certainly there are some shielded cabling sysems that are category rated (I think Ortronics and IBM teamed up with a "new generation" shielded that made Level 7/Cat 6), but mostly not.

Shielding a cable adds a whole new dimension to the possible problems of improper termination. Newbies should absolutely avoid using shielded cabling.

FWIW

Scott
 

JustinLerner

Senior member
Mar 15, 2002
425
0
0
Any cable can be 'shielded', but most Ethernet cable is UTP (unshielded twisted pair) because it has less crosstalk, higher frequency response, greater bandwith and is easier to use because of the lack of shielding and the propensity to do STP incorrectly. Like ScottMac said, STP should be avoided except by professionals.

As ScottMac, Peter and Ben# mentioned, fiber is the correct use between buildings or whenever going outside. So Remy, your situation is a big problem. Replacing $500 boards, 6 at a time per storm is not a reasonable solution. This is attrition and waste through inadequate management and planning.

Fiber should also be used WITHIN the same buildings between floors and between sections of the building having different electrical services, transformers, major ground services, etc.

Electrical systems in buildings can have voltages where they should not, including voltages on ground (yes, I've measured this before). This is often one of the causes of problems. When grounds throughout the building are not equivalent, then problems occur to equipment within the building. Supposedly, using isolated grounding is better, but if the isolated grounds differ between other common grounding then problems occur.

Electrical storms can cause individuals to transmit electrostatic potentials to PC's, but as you said, the most likely cause is the lightning strikes and MODEMS or other intenet connection via the CO or WAN connections. Voltages and potentials can travel anywhere where there is a path of little resistance. So if the CO or a line somewhere gets hit, it could possibly affect your connections.

Since I don't know enough about electical service, I believe having power conditioning equipment like induction tranformers (large ones) to condition electrical service from the electric company might work well in conjuction with largee UPS systems with generator backup and active switching for each building to provide conditioned and protected electricity to outlets in areas requiring protection (ie, not the bathrooms, halls, etc).

It is possible to test electrical power service in buildings.

What about professional lightning rod installations where needed to reduce lightning strikes near the buildings?