# 3 phase 4 wire conductors radiated electromagnetic interference?

1. Aug 15, 2011

### djr3203

If you have a 3 phase 4 wire distribution where the 3 phase conductors and 1 neutral is inside a pvc conduit, how would you model the force of the electromagnetic field at a certain distance away from the conduit?

Assuming the loads are balanced, the currents would sum to zero on the neutral. Does this result in the magnetic fields would sum to zero as well?

Would twisting the wires reduce the amount of electromagnetic interference radiated by the conductors?

2. Aug 15, 2011

### Staff: Mentor

Welcome to the PF.

The sum of the power currents is zero, so you are not likely to get EM radiation away from this cable at power frequencies.

Instead, you can get EMI radiating away from the cable if there are RF common-mode (CM) currents flowing in the cable. The cable can act as a radiating antenna for CM RF currents flowing in the cable.

Do you have an EMI issue? What is the setup, and what frequencies are the problem?

3. Aug 15, 2011

### djr3203

Well someone has brought up the concern of EMI issues with the installation. The setup involves 8 parallel conductors for each phase both carrying approx. 345A all at a frequency of 60Hz. Someone has suggested twisting the wires to reduce EMI radiation, but I tend to think this wont have a significant impact.

I have read twisting can reduce EMI pickup on wires because it reduces the cross sectional area the EMI would be cutting through. Does this have any validity? However everything I read about twisted pair has to do with differential signals and reducing common mode noise pickup, not EMI radiation.

What would twisting the wires do, if anything?

4. Aug 15, 2011

### Staff: Mentor

Datacomm differential cables use twisted pairs to minimize B-field pickup. If the wires weren't twisted, the loop area can allow a net B-field pickup, inducing a differential noise signal.

But as you say, this is mainly an issue for data cables, not power cables.

I suppose it could be said that twisting your power cables could help to cut down on B-field coupling to data cables that may be run nearby, but those data cables should already be twisted anyway. Are there any other cables or electronics nearby that might have problems with the B-field from the power cables?

5. Aug 15, 2011

### djr3203

The power cables will be concrete encased and approx 10'-15' below some rooms that could have some sensitive digital equipment. Someone on another forum recommended running them in steel conduit and grounding the conduit that would provide some shielding.

6. Aug 15, 2011

### Staff: Mentor

Yes, steel conduit is a good idea for shielding B-fields. You want to use conduit that is continuous in the circumferential direction, which is the same direction that the B-field circulates around a current-carrying wire. Can you post what the steel conduit pieces may look like, and how they are joined? Grounding the metal tubing also will help shield external devices from any high-frequency CM noise that may be on the cables, like from noisy loads that you are powering (like variable frequency motors, etc.).

7. Aug 16, 2011

### EducationalC

Funny that I found this now, as I had been having the same problem. Just found PF today.

I moved to Massachusetts and we built a new structure for a building that housed servers. I tried forums everywhere and until I searched out a local Boston electrician as a consultant to find out that the metal tubing needed to be grounded; our noise was eliminated (or at least properly shielded). Was feeling pretty ignorant to say the least.

A related question... does the power cable twisting noted below really have anything to do with the B-field coupling? I always thought that was irrelevant because (as you said) the data cables should already be twisted.

8. Aug 16, 2011

### Staff: Mentor

Welcome to the PF.

If you're only worried about interference with datacables, then twisting the power cables won't change much. I was thinking more about sensitive circuits on circuit boards, where the designers may not have anticipated a lot of B-field noise being present. Magnetic components like inductors and transformers can pick up B-field noise, and if there are loops in traces or flying wires in the device, those loops will also pick up the noise.

9. Aug 19, 2011

### sophiecentaur

The situation wouldn't be a dynamic one so would you consider using 'hum-bucking' / cancelling to reduce hum fields? But how much would such low frequency hum be likely to affect GHz frequency circuits, actually?