LinuxQuestions.org

LinuxQuestions.org (/questions/)
-   General (http://www.linuxquestions.org/questions/general-10/)
-   -   Fluorescent light (http://www.linuxquestions.org/questions/general-10/fluorescent-light-4175449756/)

sulekha 02-12-2013 12:53 AM

Fluorescent light
 
Hi all,

I was going through the book "Head first Networking" , in it it is written that Fluorescent light can cause noise in network cable(s).
How correct is this statement ?

rstewart 02-12-2013 09:45 AM

Yes, if they are unshielded and/or ungrounded and run in extremely close proximity to the lighting fixture. If the light fixtures use modern electronic ballasts not the older style magnetic ballasts and/or your networking cabling is modern cat5/cat6 then no not very likely.

SaintDanBert 02-12-2013 10:13 AM

Don't forget about the fans or damper motors that may be lurking in your overhead.
These, too, throw a serious field when they start and stop.

Since field strength is affected by the square of the distance, simply run your
cables as far away as is practical.

Cheers,
~~~ 0;-Dan

michaelk 02-12-2013 10:22 AM

Due to ethernet being a differential signal and unshielded twisted wiring is fairly resistant to Electromagnetic Interference (EMI). UTP can be susceptible to noise if it is damaged. In general you should not place signal wiring near electrical equipment that can generate large magnetic fields.

jlinkels 02-12-2013 12:07 PM

It seems to me that this again someone who writes a book and writes everything he has read in other places. It is the first time I hear about, and why would fluorescent lights cause noise. There are so many other sources of noise, and perhaps much stronger than FL.

jlinkels

SaintDanBert 02-12-2013 12:46 PM

Quote:

Originally Posted by jlinkels (Post 4890045)
...
why would fluorescent lights cause noise.
...
jlinkels

A florescent light is actually an electrical arc from end to end.
Very early radio transmissions used a "sparc gap". Have you ever
tried to listen for AM radio during a thunderstorm?

Modern florescents use electronic circuits to initiate the bulb.
Older devices were inductor+capacitor to create a large voltage
differential and "start" the arc.

DISCLAIMER -- This is how I understand things. The dusty details
may be different but this is close enough for mortal folks.

Cheers,
~~~ 0;-Dan

jlinkels 02-12-2013 12:54 PM

I have been a radio engineer for the past 30 years. I had EMC, noise and interference problems for breakfast. That is why I am asking.

Field strengths in our buildings were so high that fluorescent lights started to emit light without being switched on. Only in rare conditions this would interfere with UTP signals. So I am surprised with this statement.

jlinkels

SaintDanBert 02-12-2013 12:56 PM

Again -- inverse square law. I believe the trouble lies with UTP proximity to lighting
equipment or motors in the overhead.

73, de KI4MQ

jlinkels 02-12-2013 01:03 PM

My emphasis in the first answer was on fluorescent lights. I wonder why those were mentioned as only source. I know many causes of interference, I know the inverse square law, I know that UTP in general is not extremely susceptable.

Had the statement been that "noise caused by harmonics in electrical lines, noise cause by dimmers, switching power supplies, RF transmitters, PLC, switchgear, lightning, ADSL [...]" I would not have been triggered. Although I think it is a bit far fetched.

The susceptability of UTP, or the lack thereof is demonstrated by the low noise emittance of UTP (reprocity principle), as when compared to PLC or ADSL.

jlinkels

sundialsvcs 02-12-2013 01:48 PM

I suggest that you should start by ensuring that all of the equipment (routers, hubs, and so forth) that is connected to the network, as well as the various computers themselves, use high-quality UPS boxes.

Pragmatically speaking, I think it's far more likely that problems will come in through the power-supply cord than through induction. Computer equipment is remarkably intolerant of even the slightest electrical problem. A good UPS will provide stable power all the time.

jefro 02-12-2013 02:40 PM

The specs for networking include many tips on how to get the best performance such as how tight of bends, how much to open twists at connections and such. They also include how to limit interaction between noise. That limit may be to increase distance away or add in extra shielding. A lamp such as used in businesses and homes can have emf/rfi. Other sources could be motors, vfd's, radio and other electronics like microwave ovens. It is part of best practices even if the spec doesn't directly include it.

WiseDraco 02-14-2013 01:26 PM

not say about network cables, but i have approx 200 litres aquarium \ fish tank with two fluoriscient tubes building in his top cover. onnce i put my working netbook on that top, when lights is on, and see some glitches -lcd display start looks strange, mice cursor in kde walks without touchpad touch and so... after i put netbook on another place, all be ok.

trevoratxtal 02-17-2013 11:26 AM

In answer to How correct is this statement ?
The answer is faulty Florescent lights can cause interference to a lot of equipment.
First I will explain how Florescent lamps are gas discharge tubes they conduct electricity when the potential reaches the breakdown voltage of the gas or gasses in the tube, the breakdown voltage is influenced by temperature not only of the gas but the cathode (the filament).
A faulty tube will flicker, the breakdown voltage is very high to not possible this causes a heavy current to pass through the tube for a short time causing a spike.
It is this spike that cause a wide band of Radio Frequency harmonic interference to be transmitted.
Normal this is not a serious problem If and it is a big If , the shielding of the cabling is good, the earthing is not defective as in high resistance, the neutral of the line is at Earth potential.
If any or all are defective then problems can arise in any electronic equipment in the near vicinity.
The quickest and most effective test is to use a AM portable radio tuned to be off station.
Walk around the site that may have the problem and hunt down the source the noise will tell you the faulty fitting.
In my past life part of my remit as area trouble shooter was to track down interference to Radio and Tv signals.
One example sticks in my mind when a complete village lost its Tv reception, and I was called in to find why.
The local village council was convinced that a new installation of a Traffic controller was the culprit so starting there I turned it Off, the problem got worst.
A quick test of the neutral line voltage to local earth via a copper rod + bottle of water I kept in my tool box showed approx 120v floating.
So poor earth was one of the problems (this was a PME earth system as the local area was sandy soil.)
On careful inspection I found a business with fuel storage tanks that had just been removed in the process the earth straps had been cut.
How that came about was not for me to fix but handed to the local supply company.
However I did follow up the source of interference.
I tracked it down to a local butcher that said he had no florescent fittings on the premises.
I persisted and one of the staff over hearing said what about that not working light in the walk in fridge.
Yes it was a flickering florescent tube. Tube removed interference stopped.
And how was the chairman of the local council yes the butcher.
This is a long winded post for a good reason.
If all reasonable avenues have been tried, look for the plain stupid.
Net work cable run on low voltage one side of which is earthy, as is telephone cables if the earth floats you corrupt or lose signals.
Regards
Trev

SaintDanBert 02-27-2013 06:17 PM

Thanks, trevoratxtal. That was a fun read on several levels. All you needed was a poor dog chained to the ground (earth) stake who yipped when the phone rang...

I always enjoy stories of follow-the-rabbit fault analysis. Add to that the fact that I'm a yank and had to translate between our common languages to understand terms like "earthy." I've always preferred your term "earth" to our use of "ground" for both "mother earth" (arrow symbol drawn with three lines) as well as "chassis reference" (three slanted sticks with a flat top symbol) http://www.google.com/imgres?hl=en&s...iw=929&bih=462.

Until next time,
~~~ 0;-Dan
de KI4MQ
Austin, TX USA

jefro 02-27-2013 07:29 PM

All linear and non-linear devices could affect network quality. The issue of emi/rfi is not limited to "faulty" lamps.

The good old days of tube guitar amps would yield audible effects from many sources of noise including lamps and even thunderstorms. Any emf interaction with the cables can cause issues. Networking quality is usually a test of how well the wire can transmit an ac rf signal. The higher the speed the higher the frequency band.


All times are GMT -5. The time now is 05:28 AM.