Noise and potential ground loops problems

Status
Not open for further replies.

bzartsky

Member
I am trying to solve noise problem at one telecom site in Florida.
The system noise appears on all telephone pairs in the office and traced down to the ground and potential ground loops.

The noise is coming through the ground conductor carried by entrance cable onto CEGB (Cable Entrance Ground Bar).

Here is the configuration:

Copper pairs cables come to the outside pedestal where it is spliced and carried further into a stand along telecom hut (remote terminal or RT). The shelds of both cables, incoming to the pedestal and outgoing to the hut, are bonded to the ground bar in the pedestal which then is bonded to the ground rod. The rod is connected to the RT ground field that is within 30 feet.

The cable that comes from a pedestal enters the hut and spliced into the tip cable right at the entrance. The shield of the entrance cable is connected to the CEGB. The CEGB is insulated from the bay frame. CEGB is then connected to the MGB (Master Ground Bar) in the P-portion of the bar. The shield of the tip cable is connected to the MDFGB (Main Distribution Frame Ground Bar) which then too connected to the MGB.

MGB, of course, is bonded, on one hand to the MGN and, on the other, to the ground field.

There is also a power transformer about 200 feet away from the aforementioned pedestal and HV power line not far away either.

I believe that noise is injected into the cable from that transformer and power line.

I think that grounding of the copper pair cable in the pedestal through the ground rod and its subsequent connection to the ground field (GF) is fine. However I am concerned that grounding the entrance cable in two places, one at the ground rod in pedestal and CEGB-MGB-GF, is a potential problem of creating a ground loop. It is possible though that grounding of the entrance cable in the pedestal is required by the code (though I cannot find a reference to that) because we do not want to bring any "garbage" from outside into the office.

Could anyone please help me to clarify this issue?

I also would appreciate information on methods of how to isolate ground loops and solve the noise problem.

By the way. Should we connect the transformer and a HV power tower to the Remote Terminal ground field as well? What is the minimum distance that requires this scheme?

Yours truly,
Boris Zaretsky
 

dereckbc

Moderator
Staff member
Location
Plano, TX
Re: Noise and potential ground loops problems

Originally posted by bzartsky:
I also would appreciate information on methods of how to isolate ground loops and solve the noise problem.
This is the first step. Right now forget about loops. There are going to multiple loops in a telephone office ground system until you hit the MGB where the integrated ground plane (intentional and unintentional loops) ends, and the isolated single point ground plane begins, it is part of the design. Loops are not necessarily bad, depends on where they are located.

You need to isolate where the noise is being generated from by lifting ground cables one at a time until the source has been identified. Until you do that you are chasing ghost and wasting time. The problem sounds like either something being induced, bad connections improperly configured grounding arrangement, or something is floating like a transformer.
 

dgamble

Member
Re: Noise and potential ground loops problems

Hi,
I hope you have your problem solved. I would be intrested in your findings. If not, may I ask some questions? If I understand your post your Circuit Noise goes down if you undo the sheath bond at the Network Interface. I assume we are talking pots trouble. What is the ac current on the tele sheath measures with a clamp on ampmeter at the NI? What is your Power Influence readings at the RT and NI? Is your RT copper or fiber fed? What are your CN readings?
Ground rods and grids are great for lightening protection but do little for noise mitigation. Power mgn is required for this.
 

karl riley

Senior Member
Re: Noise and potential ground loops problems

I would want to know if your "noise" is 60Hz including the usual harmonics.

As Dereck says, you can lift grounds but you can also measure the noise on the grounds with a clamp-on ammeter without having to disconnect.

Grounding the Tformer to the ground grid would set up a parallel path for neutral current back to the Tfrmr and create a heck of a lot more noise. Not to do.

Karl
 

hurk27

Senior Member
Re: Noise and potential ground loops problems

A loss MGN at the transformer or somewhere before it can also cause current to flow on your GEC grid. Try running off the utility grid to see if it comming in from them. Of course this wont work if other loads are also on this transformer.
 

dgamble

Member
Re: Noise and potential ground loops problems

Karl,
I am required to bond a tele cable at the beginning and end and every 1/4 mile and whenever within 10ft of power, to a power co mgn. Yes, this sets up an alternate return path. I use this return ac current to oppose the influences of the ac going to the customer.IEEE talks about how much this current can be. Usually when I fix bonds and grds I watch the CN and PI(noise-to-grd) go down when I achieve a measurable ac current. This problem intrests me because it is the opposite.
I have had trouble where the ac current on the tele sheath was high and caused noise. This ended up being either a lousy or open netural.
Unless there is trouble with the tele copper pairs, CN is usually made up of harmonics of 60hz. The idea using the formula of PI-LB=CN (LB is longitidunal balance) is to get the CN at 20dbrnc or less. IEEE says LB should be 60db or higher, leaving the wild card of the power co influence. The PI can usually be reduced thru good bonding & grounding. As with anything exceptions abound.
 

karl riley

Senior Member
Re: Noise and potential ground loops problems

I would want to refer your post back to Dereck, since I am not really following this.
Karl
 

dgamble

Member
Re: Noise and potential ground loops problems

I sure went over like a lead balloon. I'll try again. Power lines produce large fields inducing a faraday circuit into surrounding metal structures. It would be better to keep this induced currrent? on the outer metal sheath and less on the twisted copper. Question-why are some electronic circuits placed inside grounded metal boxes?
 
Status
Not open for further replies.
Top