Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Is current cancellation within coaxial cable always true?

  1. Nov 23, 2015 #1

    goodphy

    User Avatar
    Gold Member

    Hello.

    I'm using coaxial cable to transfer power to the load with original aim that radiation from the coax would be low or almost nothing as current on both conductor (central conductor and outer shield layer - ground) are equal in amplitude and opposite in phase so their sum is zero; zero current means zero radiation.

    But my colleague said he was probably measuring significant radiation emitted from the cable, although measurement was done in very rough way so it was not for sure. But this makes me think about my simple assumption seriously.

    Is there a guaranteed currents are always cancelled out within coaxial cable in principle? If so, please give me proof or logic for this conclusion. If not, how can I make up such a special and ideal case? what kind of factors are to be considered?
     
  2. jcsd
  3. Nov 23, 2015 #2

    Svein

    User Avatar
    Science Advisor

    No. A coaxial cable is designed for RF signals with the outer screen at ground potential. If the outer screen carries appreciable current or is not at ground potential, it will radiate as any other conductor will.
     
  4. Nov 23, 2015 #3

    f95toli

    User Avatar
    Science Advisor
    Gold Member

    It will also depend on the quality and type of cable. In general, the more flexible a cable is the more it will radiate. Semi-rigid should radiate very little as long as everything is properly connected, but is of course less convenient to work with.
     
  5. Nov 23, 2015 #4

    BvU

    User Avatar
    Science Advisor
    Homework Helper
    Gold Member

    Time to ask what this is about specifically. This to be treated as a VHF transmission line that also transports power, or do you want to effectively shield a power cord in an audio application ?
     
  6. Nov 23, 2015 #5

    meBigGuy

    User Avatar
    Gold Member

    The differential currents between ground and center conductor will cancel. But, that is not the only thing to consider.

    If the ground and center conductor have common mode currents flowing from one end to the other, that will radiate.

    For example, if ground is changing voltage at one end compared to ground at the other, there will be a resulting current flow as if ground and center conductor were shorted as one wire.
     
  7. Nov 24, 2015 #6

    goodphy

    User Avatar
    Gold Member

    Hello.

    I'm so a little bit of confusing ground of shield of the coaxial cable. The cable is now connected to some instrument with commercial connector (7/16 DIN). Since chassis of the instrument is grounded thus I guess cable is also grounded. (However, few ampere are expected to flow through the cable and RF power frequency is of 13.65 MHz thus possible ground potential rise between the instrument and opposite end of the cable is possible. The cable length is 1.3 M. Does this disrupt "ground" of the cable seriously?)

    And if your conditions are satisfied, then current is cancelled within the cable? then how it is proved?
     
  8. Nov 24, 2015 #7

    goodphy

    User Avatar
    Gold Member

    Could you tell me concept of differential current? I know differential voltage (voltage difference between two wires) but hardly imagine current case. And for your comment, "if ground is changing voltage at one end compared to ground at the other", do you mean ground potential rise due to current level or frequency (or both) are high enough so even there is potential difference occurred over single line?

    I guess if some portion of return current (current of shield) is flowing on outer layer of the shield, it definitely becomes radiation source. Is there a way to confirm while return current within inner layer of the shield?
     
  9. Nov 24, 2015 #8

    anorlunda

    User Avatar
    Science Advisor
    Gold Member

    I know that coaxial cables are usually used for high frequency signals, but not always.

    In DC circuits we measure current using a shunt that generates several millivolts. The cable carrying this millivolt signal back to the display device must be shielded from stray magnetic fields and induced voltages. A twisted pair of wires is the usual solution, but I think a coaxial cable would work just as well.
     
  10. Nov 24, 2015 #9

    Svein

    User Avatar
    Science Advisor

    If the screen is used as a power ground, you are in trouble. If not, there is still a chance that some power return will flow through the screen. To get rid of that current, cut the connection to ground at the source end and insert a 100nF capacitor between the ground and the screen.
     
  11. Nov 24, 2015 #10

    sophiecentaur

    User Avatar
    Science Advisor
    Gold Member

    The currents in the outer are carried on the inside surface and, assuming the outer is unbroken and of low enough resistance, those currents will be equal and opposite to the currents on the inner.
    If you are feeding an antenna then the design of the termination (e.g. the balun) needs to ensure that no current flows on the outer or, as has been mentioned, that will involve radiation from the feeder as well as from the antenna and affect the pattern / match.
     
  12. Nov 24, 2015 #11

    meBigGuy

    User Avatar
    Gold Member

    When I said differential current I was referring to the current that flows out the center and back on the shield. That is, currents that are different and balanced between the shield and the center conductor. Essentially, the signal currents.

    What you are doing with the coax (among other things) is stopping the signal carried on the center conductor from radiating.

    If you connect a braided wire between the shield points of the two systems, and any current flows, that will cause radiation.
    That is the common mode signal I was referring to.
    It can be more complicated depending on your ground/shield architecture, but that is the gist of it.
     
  13. Nov 26, 2015 #12

    goodphy

    User Avatar
    Gold Member

    So, it looks like you suggested that the load also has to be grounded with direct additional connection to the lab ground in order to flatten potential of the shield (screen in your word) of the coaxial cable, is it right?

    I guess you mentioned returned power due to impedance mismatching at the load and in this case, inserting capacitance between cable shield and power source box ground will be solution but could you tell me why? capacitance plays short for high frequency thus returned current will be diverted to the ground?
     
  14. Nov 26, 2015 #13

    goodphy

    User Avatar
    Gold Member

    Anyway I derived that current summation should be zero for any type of transmission line as shown in the attached image. Please confirm this is right or not.
     

    Attached Files:

  15. Nov 26, 2015 #14

    f95toli

    User Avatar
    Science Advisor
    Gold Member

    You've shown that the current should sum to zero in any ideal TL. That is not in any way in dispute.
    However, in any real setup other effects will come into play meaning this is not always true.
     
  16. Nov 26, 2015 #15

    sophiecentaur

    User Avatar
    Science Advisor
    Gold Member

    The problem can be that an antenna is at an elevated position on a mast or building. Merely shorting the unbalanced current to the local ground may not help at all because the current can still radiate power. It's only when the antenna is mounted low down or on a large enough elevated ground plane that the unbalanced currents will not have any effect. You can eliminate the unbalanced current with a balun or a choke of some sort in the coax. A balun is the best solution because it eliminates the unbalanced current by design and is a more relaible way to produce the intended radiation pattern.
     
  17. Nov 26, 2015 #16

    meBigGuy

    User Avatar
    Gold Member

    Within the transmission line you are correct. But, that is not the issue. (BTW, Your diagram is a lumped model, which is a simplification.)

    I think you missed my point about common mode currents. You need to think about this or ask questions until it makes sense.

    This is a sloppy description, as I am short of time.

    Let's say you attached a noise source on the left side that induced identical noise voltages on both left terminals with respect to a virtual ground. Then put identical load resistors on the right two terminals with respect to the same virtual ground. Current from the noise sources would flow on both wires from left to right and on virtual ground from right to left. This means there would be a field between the two wires and virtual ground. I refer to the currents on the cable as being common mode since they are the same and in the same direction.

    These common mode currents flowing on the cable between the two units will radiate.
     
  18. Nov 27, 2015 #17

    meBigGuy

    User Avatar
    Gold Member

    Let me try again:

    Say you have box A with ground Ag and Box B with ground Bg. (both with misc electronics inside, of course) Connect a GROUND wire between them. If any AC currents flow between the boxes along that GROUND wire it will radiate. It flows from (say) A to B and then back from Bg to Ag. There are no "signals" involved, just the boxes and their grounds. If the boxes are perfectly sealed and the connections to ground are perfect and the two grounds are perfect, then there would be no current flow from A to B and no EMI. But, in the real world there are currents flowing on the outside of the box, currents flowing between the boxs and their "ground" and therefore, when you connect the boxes grounds together, there will be currents between the boxes.

    So, just by connecting two boxes grounds together with a braided cable, you can get EMI. You don't even need a "signal". Using a coax won't change that since we are not even hooking up the signal yet. Adding a ferrite to the cable can help attenuate those signals (you see these on PC cables a lot, especially old video cables and wall warts).

    You can test this by disconnecting the center conductor on your coax.

    On the other hand, not all coax cables are created equal. Some have 100% shields, some 70%. Some are foil, some are braid, and some are foil and braid. If the problem only happens when you connect the signal, then suspect your coax.
     
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook




Similar Discussions: Is current cancellation within coaxial cable always true?
  1. Coaxial Cable loss (Replies: 81)

Loading...