Induced voltage (Phantom Voltage) vs. cable length

Status
Not open for further replies.
Good day everyone out there

I have a burning question that I wish I could really find an answer for. This is related to induced voltage from adjacent conductors (within the same multipair cable).

We have an instrument that requires 120 V ac to work and we mandate that the conductors length from control room (safe area) to the instrument do not exceed 1000 ft (305 meters). The reason is that we think if the conductors (within a multipair cable) exceed 305 meters, then induced voltage may appear on adjacent pairs.

I am not totally convinced as each pair is (of course) twisted and there is also an individual shield and an overall shield. The question here is:
Is there a relationship between the conductors/cable length and the induced voltage (given the fact that each pair is individually shielded and also twisted)?

Thank you
 

GoldDigger

Moderator
Staff member
Location
Placerville, CA, USA
Occupation
Retired PV System Designer
It is really pretty simple. If the load resistance in the tested pair is high the induced voltage will be directly proportional to the length over which the pairs are in contact.
If the pair is terminated by a low impedance (such as an op amp summing point) the induced current will be proportional to the length instead.

The effect will be smaller because of the twists and shielding, but will still be proportional to length.

By the way, the twisted pairs will only help capacitive effects if balanced differential signals are used and if the power wires are also balanced, as in 120-0-120.
Twisting is more important when you are worried about magnetic rather than capacitive effects.
For best effect the pitch of each twisted pair should be slightly different.
 

gar

Senior Member
Location
Ann Arbor, Michigan
Occupation
EE
151028-0806 EDT

GoldDigger gave you some useful information.

There is nothing magic about 1000 ft. So length alone does not define whether there is a problem or not. There is no threshold effect at 1000 ft such that at 999 ft you don't have a problem, but at 1001 ft a problem occurs.

You did not specify whether you were concerned with magnetic and/or capacitive coupling.

Electrostatic shielding, an aluminum foil or braided, greatly reduces capacitive coupling between what is inside the shield and what is outside the shield. This type of shield does virtually nothing for low frequency magnetic coupling.

Twisting wire pairs greatly reduces magnetic coupling from an identical current in the twisted pair to an adjacent conductive loop. The wire pair must be the only path for its source and return current. Each little twist in a twisted pair is a one turn coil. From one twist to the next the magnetic field direction reverses. If you have an adjacent single loop coil that is long compared to the twist pitch adjacent to the twisted pair, then fairly good cancellation of the reversing magnetic fields occurs. But if you have adjacent twisted pairs with exactly the same pitch, then you can have additive coupling. This is the reason GoldDigger mentioned a different twist pitch on adjacent pairs.

What problem do you want to solve? What are the different signals in your cable?

.
 
Induced voltage (Phantom Voltage) vs. cable length

Gentlemen

Thank you for your excellent replies. I truly admit I may have not explained further, but here it is:

What I am trying to solve is: I want to know the exact length of the home-run cable which I could not exceed before potential capacitive or inductive coupling could exist.

The situation I have is a multi-pair cable (12 pairs) each one of them is of the same signal type, 120 Vac digital output serving horns and beacons.

I would like to know how to figure out the maximum distance I could lay the cable for before something wrong could happen. I was told Phantom voltage may exist and cause problems. Please let me know in details and if you could share any other information, I would appreciate that.

Thank you
 

GoldDigger

Moderator
Staff member
Location
Placerville, CA, USA
Occupation
Retired PV System Designer
Gentlemen

Thank you for your excellent replies. I truly admit I may have not explained further, but here it is:

What I am trying to solve is: I want to know the exact length of the home-run cable which I could not exceed before potential capacitive or inductive coupling could exist.

The situation I have is a multi-pair cable (12 pairs) each one of them is of the same signal type, 120 Vac digital output serving horns and beacons.

I would like to know how to figure out the maximum distance I could lay the cable for before something wrong could happen. I was told Phantom voltage may exist and cause problems. Please let me know in details and if you could share any other information, I would appreciate that.

Thank you

With a signal level of 120VAC I think that you are really talking about powering the horns and beacons rather than just supplying a control signal to turn them on and off with the power provided by another circuit.
But in any case, there is no specific distance at which a problem will suddenly appear.
If the devices will work just as well on 100VAC as 120VAC and will not operate at an input of 20VAC, then it is very unlikely that there would be any crosstalk problems even at 1000 feet.
You are much more likely to get a problematic voltage drop in that length of wire from the wire resistance than a problematic stray voltage from coupling.

Two of us so far have told you that the problem is gradual in nature and even if you knew the exact details of the wiring and the exact noise sensitivity of the devices there would not be a simple cutoff point beyond which the devices would not work reliably.
And yet you come back and ask for just that.

There is no simple formula for crosstalk without knowing exact details of wire sizes, spacing, twist, shielding efficiency, etc. What there will be is a rule of thumb from experience. The person you are talking to may be working from experience or knowledge of systems that had problems.
 

FionaZuppa

Senior Member
Location
AZ
Occupation
Part Time Electrician (semi retired, old) - EE retired.
whats the nature of the 120vac load of instrument (or is it just 120vac signaling)? i believe you are venturing into realm of mag coupling. how does the device get its power. if the signaling is all high impedance (very low current) the induced voltage from signaling mag fields will be very small.
http://www.eeweb.com/toolbox/twisted-pair
http://ecmweb.com/content/avoid-instrumentation-problems-properly-installing-low-voltage-wiring


i have not seen wire where the pair is twisted and then folded to reduce mag coupling (well, more accurately, reduces induced voltage).

another Q i have is, does the 120vac signaling have a pull-down feature when the signal is zero volts?
 
Last edited:
All data available to scrutnize this situation further

All data available to scrutnize this situation further

Good day Gentlemen

I collected some more information to see if one could look further and provide some help into figuring out if there will be a problem of inductive coupling at a distance of 3000 meters.

Details of the cable being used:

1- Multi-pair (12 pairs)
2- All conductors are in the same cable and will serve the same purpose (power and control Beacons and Horns through 120 V ac).
3-We are complying with IEEE 518 when it comes to spacing with other adjacent signals. In this case, there is enough spacing on the same cable tray between this multi-pair cable and adjacent ones (about 6 inches from signal level 2, and 8 inches from signal level 4).
4- Twist: 60 mm
5- Conductor: Circular copper,7 strands minimum; tinned per ASTM B33
6- Conductor size: 18 AWG

7- Primary Insulation material: thermoset cross-linked polyolefin, temperature rating 90 C.
8-
Primary insulation voltage: 300 Vac.
9- Primary Insulation thickness: 0.38 mm

I hope with this information, one could figure out if 3000 meter run is going to be a problem.

Thank you all.
 

gar

Senior Member
Location
Ann Arbor, Michigan
Occupation
EE
1521101-0742 EST

HSB_007:

You list yourself as an Instrument Engineer, but I don't believe you have an electrical engineering degree. Thus, for background I would judge you have some sort of electrical installation experience.

You have provided some information on your cable, but nothing on power voltage and current levels, nor signal levels, impedances, and sensitivity to coupled noise.

Magnetic fields. Any current flow produces a magnetic field. An electron beam in a cathode ray tube (CRT) (no metalic conductor, just a beam of electrons) produces a magnetic field. Thus, that beam can be deflected by another magnetic field. Bring a strong magnet near the face of a CRT and you will produce at least some movement of the image. On a color CRT there will be color distortion.

How strong is a magnetic field from a current? It is proportional to current, inversely related to the distance from the current, and influenced by the magnetic material between the current and the point of measurement.

Magnetic fields can add or subtract from each other. Cut two lengths of #14 Romex 12 ft long. Strip both ends so you can make various connections. Tape these together so the flat sides are tight together. Also keep the white wires adjacent which means the black wires are adjacent.

Form the the wire assembly to make a circle of about 4 ft in diameter. Mount on a large sheet of cardboard.

In the first experiment use a compass as a DC magnetic field sensor. Position the plane of the coil vertical and in a north-south direction. This makes the magnetic field from the coil perpendicular to the earth's field. Place the compass at about the center of the circle.

With no current in any wire of the coil the compass needle will point north and in the plane of the coil.

Apply about 10 A DC to one wire. The compass will deflect away from north. Reverse the current polarity. The compass needle will deflect about the same amount, but in the opposite direction.

Next add another wire in series with current flow in the same direction. Now the compass needle deflection will be greater, but not quite 2 times.

Reverse the current flow in one of the two wires. Now there is no needle deflection at any current level. Here the two coil generated fields cancel each other.

Rotate the coil assembly on a vertical axis by 90 degrees. Now any amount of DC in one coil will produce no needle deflection. This is because the coil field is in the same direction as the earth's field. The field intensity at the compass is changing, but the compass can not show us that.

More later. Some CAT-5 or other twisted cable will be required.

.
 

FionaZuppa

Senior Member
Location
AZ
Occupation
Part Time Electrician (semi retired, old) - EE retired.
is that "300vac" = "300vac rms" ?

as last post says, the answer to the Q cannot be derived from just the wire and length, need more info.

is the bundle (cable) with multi-twisted pair carrying 120vac(rms) for powering the device(s) on the far end, along with some pairs being used for signaling the remote devices?

to me this sounds like its leaning towards pairs that are (should be) wrapped in foil, then on each end all the foils are tied to a good gnd.
 

GoldDigger

Moderator
Staff member
Location
Placerville, CA, USA
Occupation
Retired PV System Designer
A multi-pair cable intended for signal and data use should have specifications about pair to pair crosstalk at various frequencies. That combined with information on the noise sensitivity of the most sensitive circuit will give the go-no go answer.
 

gar

Senior Member
Location
Ann Arbor, Michigan
Occupation
EE
151101-2150 EST

Continuation of experiment, but with AC 60 Hz.

I added a 20 ft length of Belden 8723 cable tightly taped to the pair of Romex cables. This 8723 is longer than the Romex to provide shielding to get away from the Romex terminations.

8723 has two individually Beldfoil shielded twisted pairs, and a drain wire. Beldfoil is an aluminum shield wrapped around a twisted pair, but insulated so as to not produce a shorted turn around the pair. This provides very effective electrostatic shielding.

This I refer to as a living room experiment, others are referred to as kitchen counter types. Attached are two photos. One shows the loop, and the other the magnetic field sensor.

PICT3858.jpg
PICT3859.jpg
Thru one conductor in the loop 10 A at 60 Hz was applied. At the loop center 0.095 gauss was measured. Adding one more turn and this increased to 0.195 gauss as expected. The magnetic sensor calibration at 60 Hz is 20 mV per gauss. Calculating the theoretical flux density we get 2*Pi*I*10^-7*10^4/R gauss = 2*3.1416*10^-2/(24/39.34) = 0.062832/0.6101 = 0.103 gauss for the 4 ft diameter 1 turn circular coil with 10 A flowing.

Next the two turns were connected with the same but opposing currents. The reading dropped to 0.01 gauss, and really much lower because this was the meter noise level.

The induced current between two one turn coils was 0.07 A out for 10 A in. The open circuit induced voltage was 0.015 V for 0.34 V input.

Next were tests with the 8723 cable. Red and black are one twisted pair. Red and black were shorted at one end. The voltage at the other end of the pair between red and black was less than 10 microvolts (meter noise level) with no current in a one turn loop. No change in the ouptut reading with a 10 A loop current. Note the shield was connected to nothing. The electrostatic shield does nothing relative to magnetic fields.

The next tests are for capacitive coupling. Care has to be taken with meter leads to prevent unwanted coupling. Twist the leads full length and run so there is minimum coupling from 120 hot.

120 V 60 Hz was applied between the Romex black lead and the combined white and ground wires of the Romex closest to the Belden 8723. Neutral/EGC were connected to house neutral. Measurement voltage reference point is the Neutral/EGC of the first Romex.

Measuring meter is a 10 megohm input impedance DVM.

21 V to EGC of other second Romex. Szme to black and white.

17 V to drain wire of 8723.
15 V to 8723 black and red inside shield, but shield floating.

10 microvolts from 8723 red and black relative to drain and Neutral/EGC when drain was connected to Neutral/EGC with or without 120 V applied to the black of the first Romex.

A long cable will have higher values than these. You must understand basic electrical theory and apply it to your problem. You need to know power voltage and current levels, signal levels, types of noise signals, and what is the susceptibility of your instruments to various noise inputs.

You have to determine the threshold points of problems and how much protective margin you need. So far you have provided no information on your power and signal levels, or the nature of the signals, and what might be acceptable margins, nor any load impedances.

.
 
Induced voltage (Phantom Voltage) vs. cable length

1521101-0742 EST

HSB_007:

You list yourself as an Instrument Engineer, but I don't believe you have an electrical engineering degree. Thus, for background I would judge you have some sort of electrical installation experience.

You have provided some information on your cable, but nothing on power voltage and current levels, nor signal levels, impedances, and sensitivity to coupled noise.

Magnetic fields. Any current flow produces a magnetic field. An electron beam in a cathode ray tube (CRT) (no metalic conductor, just a beam of electrons) produces a magnetic field. Thus, that beam can be deflected by another magnetic field. Bring a strong magnet near the face of a CRT and you will produce at least some movement of the image. On a color CRT there will be color distortion.

How strong is a magnetic field from a current? It is proportional to current, inversely related to the distance from the current, and influenced by the magnetic material between the current and the point of measurement.

Magnetic fields can add or subtract from each other. Cut two lengths of #14 Romex 12 ft long. Strip both ends so you can make various connections. Tape these together so the flat sides are tight together. Also keep the white wires adjacent which means the black wires are adjacent.

Form the the wire assembly to make a circle of about 4 ft in diameter. Mount on a large sheet of cardboard.

In the first experiment use a compass as a DC magnetic field sensor. Position the plane of the coil vertical and in a north-south direction. This makes the magnetic field from the coil perpendicular to the earth's field. Place the compass at about the center of the circle.

With no current in any wire of the coil the compass needle will point north and in the plane of the coil.

Apply about 10 A DC to one wire. The compass will deflect away from north. Reverse the current polarity. The compass needle will deflect about the same amount, but in the opposite direction.

Next add another wire in series with current flow in the same direction. Now the compass needle deflection will be greater, but not quite 2 times.

Reverse the current flow in one of the two wires. Now there is no needle deflection at any current level. Here the two coil generated fields cancel each other.

Rotate the coil assembly on a vertical axis by 90 degrees. Now any amount of DC in one coil will produce no needle deflection. This is because the coil field is in the same direction as the earth's field. The field intensity at the compass is changing, but the compass can not show us that.

More later. Some CAT-5 or other twisted cable will be required.

.

Mr. gar

I do have a master degree from University of Southern California with a grade point average of 3.8/4.0

I am an a qualified instrumentation engineer that has been working in the oil and gas sector for over 16 years. I am also a P.E. in the state of Texas in control systems. I do see no value in degrading someone without addressing the real solution from a practical real world application. What counts is experience, not pure theoretical conversation.

We all know as electrical engineers when a current passes through a piece of conductor that a magnetic field is generation (magnetic induction, from uncle Faraday's law). We have learned this way back in school. That does not address the problem.

I agree I forgot to mention some important data, but that was not intentional as I am bombarded with so many activities, and I may have overlooked this data. Anyway, respect is important more than knowledge.

Here is more information so you can come up with a solution:

Current: 0.33 A

Voltage : 120 Vac 60 HZ

Impedance of the conductor: 15.9 ohm/km

Loop impedance: 31.8 ohm/km

Cable size: 16 AWG

If I missed anything, please let me know. We all can learn from this situation rather than showing off muscles.

Thank you!
 
Induced voltage Problem

Induced voltage Problem

is that "300vac" = "300vac rms" ?

as last post says, the answer to the Q cannot be derived from just the wire and length, need more info.

is the bundle (cable) with multi-twisted pair carrying 120vac(rms) for powering the device(s) on the far end, along with some pairs being used for signaling the remote devices?

to me this sounds like its leaning towards pairs that are (should be) wrapped in foil, then on each end all the foils are tied to a good gnd.

I believe the 300 Vac is not rms. This is the insulation voltage rating of the cable.

I did give more data in my response to gar. I know that as the cable (which houses 12 pairs of 16 AWG conductors) will have more induced voltage as it gets longer, but my personal thought is that even that voltage induced will not be able to drive the adjacent horns because they would require more current. If we assume the induced voltage is say 20 vac, will the current generated be enough to drive the adjacent horns? My feeling is that it would not unless the overall cable length is way longer than say 3km.

I do understand the concept, but do not know what equations to use to figure out this problem. We are using a very good multi-pair cable which have individual shields on each pair with an overall shield. The very same conductors are all twisted which would minimize or cancel out the capacitive coupling. I would tend to feel that the real problem as indicated by Mr. GoldDigger may be that voltage drop more than inductive coupling.

Thank you all anyway for the help.
 

mivey

Senior Member
I do see no value in degrading someone without addressing the real solution from a practical real world application. What counts is experience, not pure theoretical conversation.
...
If I missed anything, please let me know. We all can learn from this situation rather than showing off muscles.
I believe what you missed was that there was no insult intended on gar's part. I believe he is trying to determine how to respond. Without knowing someone's background it is hard to know how much background information one should include.

As far as muscle, gar has plenty of theoretical, practical, and experimental muscle. Not much showing off that I recall but he does use them to selflessly help people here.

Anyway, respect is important more than knowledge.
Swallow your own medicine and don't be so easily offended. Fortunately, gar also has thick skin to go along with his muscles so I suspect he will continue to help you.
 

mivey

Senior Member
We all know as electrical engineers when a current passes through a piece of conductor that a magnetic field is generation (magnetic induction, from uncle Faraday's law). We have learned this way back in school.
and for the record, current and the resulting magnetic field are described in Maxwell's equations and are related through Ampere's Law. Faraday's Law has to do with the relationship between magnetic fields and electric fields.
 
Do you have an answer?

Do you have an answer?

I believe what you missed was that there was no insult intended on gar's part. I believe he is trying to determine how to respond. Without knowing someone's background it is hard to know how much background information one should include.

As far as muscle, gar has plenty of theoretical, practical, and experimental muscle. Not much showing off that I recall but he does use them to selflessly help people here.

Swallow your own medicine and don't be so easily offended. Fortunately, gar also has thick skin to go along with his muscles so I suspect he will continue to help you.

Mivey

Do you have an answer to my question?
 
Status
Not open for further replies.
Top