Do high voltage power lines affect WiFi transmission?

[pradeep.ghimirey]

I am planning to set up a 2 Km WiFi link between two points. But there is a high voltage railway line running between the two points. Is there any effect due to this line running between the points?
Is there any effect of the Electromagnetic field created due to the electric cables on wifi transmission.
Might sound weird query. But have no clue about this. Can anyone help :roll:

[WHT]

In spite of "well informed" sources screaming that you should never shoot through power lines, there is no credible process or mechanism that would explain any interference with the signal.

[InoX]

works for me

[UBNT-Mike.Ford]

In spite of "well informed" sources screaming that you should never shoot through power lines, there is no credible process or mechanism that would explain any interference with the signal.

Except the metal from the transmission lines?

[WHT]

The effective aperture at more than a few feet would make them invisible. Along the same lines, if you mount an omni several wavelengths away from a tower leg, it won't cause any change in the omni pattern.

[UBNT-Mike.Ford]

True

[Shockware1]

I have seen issues when you run parallel to power lines. Perpendicular however seems to be of little consequence.

Shooting about 35 degrees across very high voltage mainline transmission lines we have seen a client have a "black hole" in terms of signal strength. He should be having -50 or so signal, however he only gets -76 and fades out every night. Not long distance, good LOS, little elevation change (AP was down-tilted actually, so nothing to do with signal shooting over his head).

I also had another guy I was checking signal for show issues with power lines. His neighbor had good signal strength, something like -66 or something like that. This guy was located approx. 50 meters south of the first guy shooting west to the same tower with the same LOS. He could only get in the low -80s. I could not install him. Tried other channels and the works. He had power lines going to the road, and the other guy did not.

Some lines seem to have little effect however. I suppose there was not much current draw/lower voltage on them at the time? On the whole we do NOT have huge problems with power lines, but we do look to avoid them if possible.

I guess you could say that I'm in the camp that believes power lines can have an effect. This is not because of a university degree or any sort of training/researching, but from personal experience.

[WHT]

He could only get in the low -80s. I could not install him. Tried other channels and the works. I can duplicate the same thing and without power lines in the path. Move his antenna up or down a few feet solves that.

Not to back pedal, but to qualify my answer - In the very rare situation where a CPE might be looking right down the bore path of some metallic lines, power or telephone or whatever, you may have some problems. Nevertheless, my answer to "shooting through (diagonally)" power lines not having an affect would still be valid.

[Shockware1]

At the time of installation we did a full survey of the site. This includes holding the antenna high in the air, a test from ground level, from outside any treeline if applicable... If I were to notice that I held my antenna a few feet higher, I'm sure I would not have mounted the antenna where it is.

I have seen that issue several times when installing a Trango dish on a metal roof. Move it 3 feet up and you get +25 more dB. Crazy stuff. But not here. Asphalt roof and same signal all over the yard. Other clients off of the same AP got normal signals to be expected for their location/distance. This guy got several dB lower than expected. Used different dishes, radios, cable, and locations all with very similar results.

At the time his signal level was close to acceptable and his packet test was fine, so we installed him. Not enough fade margin though.

Here is a top view representation of what we were doing.
Distance is no more than 3 miles. Probably closer to 2 1/2. Pathloss calculator says I should be getting about -58dB. I was getting -75 at BEST. Installed at final location at -77 I think. That's where he wanted it...
http://img29.imageshack.us/img29/1585/powerlinefade.jpg


The only conclusion we could draw was the power lines. The lines were probably more parallel to LOS than the picture but as you can tell I am no artist. :)http://img21.imageshack.us/img21/2465/powerlinefade2.jpg

[WHT]

Yeah....in that case, the aggregate cross section of the lines begin to approximate a solid or grid antenna. I wasn't thinking of that situation. When we did the testing two years ago, we were shooting diagonally.

In future comments about power line interference, I'll include that caveat about small angles to the lines.

[WHT]

On a different tilt, you can try this with Radio Mobile.

Move your CPE height up and down several meters and you'll see several sweet spots about umm...lemme think, about two meters wide, and then a dead spot about 10 dB less about two meters wide. Of course, that depends on the distance; but for usually under a few miles, that's what I've seen.

In actual practice, I haven't seen 10 dB loss. It might have been the fade refraction working in my favor at the moment. That's why I use a 20 dB fade margin for my plotting.

[oneofthefew]

ideal link scenario means no infringement within 60% of the links fresnel

consider using a MIMO radio (802.11n) if your link is not performing well, but it primarily comes down to how much throughput you want on the link

also you may want to think about frequency choice

5GHz is better at getting round obstacles than 2.4GHz due to the carrier size being smaller:

5.1GHz ( 0.059m )
2.4GHz ( 0.125m )

also please note that you may be receiving a multipath signal which could cause you issues when you try and push large amounts of data over the link

would suggest checking both uplink and downlink RSSI to ascertain the possibilities of a multipath signal

hope this helps

:D

[Shockware1]

the AP radio does not show signal strengths unfortunatly. Should change the out but have not yet.

What I did was I put a nano5L on the tower and a bullet5 on his house with a 26dB grid using 5hz channel size. The client seems fine now.

An upgrade like this for us in the past would cost over $1000 in radios alone. (RAD link most likely) Now we can do it for pretty cheap in order to keep clients on board.

This guy in question sits around -65 now with negotiated rates hanging around 12 to 13.5.

[pradeep.ghimirey]

Thanks everybody for the elaborate response. :D have got a good clarity regarding the issue.
Since my deployment is perpendicular to the transmission link, I hope not to face many issues.

[cape-connect]

Have a look at this (http://www.ubnt.com/forum/attachment.php?attachmentid=91&d=1253481104) image.

Those are 400,000V powerlines in the background.

The AP is about 7.5km away.

The client's probably connected to the internet right now. Our management system shows she has virtually zero downtime.

[PCaddict]

I can give u a situation i see by myself. Not far away of one of my network they have a powerline of 750kv (Hydro Quebec) that cross the road. If i go under when i talk with my cell phone i lost the connection all the times , and many of my clients at this spot report me the same issue. (With any wireless provider) so if the cell phone stop working under the powerline , i expect that the wifi signal have no more chance :P But i will try during this winter to make a connection between my ap and the Hydro Quebec powerline to see waht kind of issue i can face off

[Dave-D]

High-voltage power lines themselves are
probably not a real problem. But the condition
of the lines can be. It's not unusual to have a
crack in one of those big hanging ceramic
insulators, and when that crack gets a bit of
moisture, it makes a nice corona.

Even a very foggy day with dripping moisture
can cause corona on the insulators. You can
hear the 'frying' noise, and sometimes see
the long tendrails of sparks.

That corona creates a very wide-spectrum
rf noise. The power company searches for
bad insulators with a portable 'microwave'
hand-held receiver with a parabolic dish.

If your power company is current on its
maintenance, great. Otherwise, not great. Dave

[WHT]

That corona creates a very wide-spectrum
rf noise. The power company searches for
bad insulators with a portable 'microwave'
hand-held receiver with a parabolic dish.
Eight of the nine electric utilities I've worked with use a an eight inch or so dish antenna that has a microphone to pick up the ultrasonic corona discharge noise. Some have RF detectors, but never used to detect corona discharge.

[Dave-D]

WHT, how about comments to my point?
Compare to your post #2. Dave

[WHT]

My #2 post (and subsequent posts) were related to the physical presence of an electric power line, much the same way as a picket fence, tall shrub, and Marie Antoinette wigs.

Other noise sources tend to peter out past 100 MHz. If indeed a 100 MHz noise source was getting into your radios, that would be the least of your worries.

[Dave-D]

If a high-tension power line were a passive element
like the ones you mention, there'd be no discussion.

I think you're implying that corona discharge (which
is a high-voltage spark) has no important rf components
in the WiFI bands (or above 100Mhz). Do you know that
to be the case? Dave

[WHT]

Do you know that
to be the case?Yes, I can sit under a noisy power line and the FM components will will wipe out my 46 MHz two-way, hardly heard on my 152 MHz mobile phone, not heard at all on my 460 MHz radios.

[Dave-D]

You may well be correct, WHT.

But you can't prove anything by listening on
your 2-way. The frequency aperture of 10Khz
compared to 40Mhz provides about 4000 times
less susceptability to interference.

Yep: spread-spectrum and all that, but there
is an enormous difference in bandwidth. Dave