Data Network causing Radio Interference

[daronwilson]

Just getting ready to track this one down but thought I would ask the group first.

I have an issue at home as I expanded my network and computers now I have a nice overwhelming noise issue on a couple VHF radios in the house. I've not traced it down or tried to eliminate it yet, but I just got a call from a customer who has a 40 or so drops in their building, and a base station radio with the antenna on the roof. It has slowly increased in noise with the computer use (according to the customer) and now is a major concern.

The customer explained that first thing in the morning before everyone turns their computers on it is relatively quiet. If they shut down the switches, it pretty much goes away. As the day starts up and more people get on the network and move data, it gets so noisey the radio is unusable.

Anyone experienced this and managed to mitigate it? I'm thinking of starting with snap on torroids on the patch cables at the switch and see if that decreases it.

Maybe it will make more sense when I'm there and look at it, just thought someone might have already solved a problem like this.

thanks

It is only my opinion, based on my experience and education...I am always willing to learn, educate me!
Daron J. Wilson, RCDD
daron.wilson@lhmorris.com

 

[franklin97355]

Sounds like the antenna cable is not properly grounded. There is a fair amount of noise put out by a computer. You may need to make sure your grounds in the electrical system are also good.

 

[buitenhek]

Daron,
Start small. I had the same with a building in San Diego. The customer called us in because he had problems from 6 AM to 10 PM (7 days per week). Huh?

I was standing on the 4th floor thinking, when I noticed the small airport nearby. Yes the hours of op were 6-10.

To boil it down, my first thought was shielded cable. However, I decided to start investigating specific areas first. We informed all users on each "hub" leg that they would be down for 2 minutes at the top of each hour. Initially all user were cautious and did not do any processing at the top of the hour so our "noise" figures were bogus. After that we were getting pretty good figures that boiled down the location to kitchens and microwave ovens. We also found 3 faulty (dirty) PC power supplies that we put on Oneacs. The customer could live with the remaining low level noise so the total time was 2-3 days and no new cable was required. Bottom line, a naughty I-beam acting like an antenna, was suspect.

One final thing...this company moved out and another moved in with you guessed it, "music & buzz". The building manager recommended us for the fix. Boy did we look good. Sorry customers, I have charged this fix now 4 times.


Regards
Peter Buitenhek
Profit Developer.com

 

[netmanrick]

Remember with switching power supplies there may be switching transients induced on the ground AND the neutral legs.
Get the building electrician to verify the security of the terminations in the recptacles AND the branch feed panels.
Loose bonding on neutral and ground "legs" in the feed panels will drive you batty.

Rick Harris
SC Dept of Motor Vehicles
Network Operations

 

[wires]

The first thing I would do is beg, borrow, eBay or rent a spectrum analyzer. Trying to troubleshoot RF problems with out one is like trying to find wires without a toner and probe.

Once you have "found" the offending frequency you can use several techniques:

1. Turn off one piece of equipment at a time and see if the RF level at the offending frequency goes down.

2. Turn everything off then power up one piece of equipment at a time and see if the RF level at the offending frequency goes up.

3. Use a directional (at the offending frequency) antenna to "fox and hound" the problem.

4. Change wire lengths or insert torroids on connections that exhibit increased RF level at the offending frequency.

I am currently using a LP Technologies (

http://www.lptech.com)

LPT-1250 which has decent PC based control and recording software. The peak hold function makes RF site surveys very easy. Just set it up and let it run. Come back later and see if the spectrum you are interested in is "clean".

If you are doing RF work do yourself a favor and get a spectrum analyzer. Avcom (

http://www.avcomofva.com)

units sell on eBay for around $1000 all the time. While these are not "lab grade" they are much better than no spectrum analyzer at all.

 

[daronwilson]

We set up a spectrum analyzer and looked at the site. With a local antenna in the building, the broadband noise (not frequency specific) is up 9dB above the ambient noise floor when the network switches and PC's are turned off. During large file transfers, peaks in the noise (not quite so broadband) were near 15 dB above the noise floor.

Testing was started with all equipment off, measurement of the noise floor. Servers were turned on, switches, PC's all incrementally showed a slight increase in the noise floor, but nothing particularly frequency specific, just broadband noise.

We went through after this and bonded the rack, switches, servers, conductive water pipes, 2nd ground rod, utility ground, telco ground, radio chassis and feedline. None of this is my installation, and the switches weren't grounded and the actual building ground was not as good as I wanted, so we added a second ground rod and bonded everything. Next plan is to repeat the noise tests to see if we made any improvement.

Torroids on 46 network drops is certainly a possibility, just didn't want to sell that to them to find that it didn't help their installation.

Currently they have an omni antenna right above the roof. The radio guys told me they first put a small yagi up there, but it caused the customer problems on their network when they transmitted (good clue).

I suspect after bonding everything we may see some reduction, but in the long run the antenna needs to be a gain omni antenna with a pattern that focuses signal above the building, and it need to be up in the air above the building.

Hopefully we'll get a chance to sort through it this week and see, thanks for the ideas.


It is only my opinion, based on my experience and education...I am always willing to learn, educate me!
Daron J. Wilson, RCDD
daron.wilson@lhmorris.com

 

[buitenhek]

Daron,
Very Interesting. Please let us know the results.


Regards
Peter Buitenhek
Profit Developer.com