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

Shielding & ground loop for audio

  1. Feb 14, 2015 #1
    I have a 3 conductor wire that are all wrapped in a foil shield with wire conductor, for a total of 4 wires.

    I am connecting the 3 inside wires to audio left, right, and ground. This is an unbalanced setup, right?

    Where/how do I connect the shielding?
    Do I connect both sides (i think no)? Which side? To where?
    My choices for where to connect the shield: audio ground, chassis, nothing

    I would like to avoid all ground loops and noise.

    If I felt like putting on some EMI protection and have some braided EMI shielding roll, should I bind the EMI shielding to the foil shield to make one shielding conductive layer or should I separate the shield layers with insulation? any pros and cons to each method?
    Do I connect both sides (i think no)? Which side? To where?
    My choices for where to connect the shield: audio ground, chassis, shield (wherever that may be connected to), nothing

    I feel like I've seen some manufacturers with multi-layer shielded wires that have separated layers, others that are together, and some that have a combination. What are the benefits?
    Last edited: Feb 14, 2015
  2. jcsd
  3. Feb 14, 2015 #2


    User Avatar
    Science Advisor

    I would normally ground the foil at the reference ground for the signal but it's one of those things that depends on the circuit and application. Some general information about shielding.
  4. Feb 14, 2015 #3
    ground the foil with ground at which end?
  5. Feb 14, 2015 #4


    User Avatar
    Science Advisor

    The signal source.
  6. Feb 14, 2015 #5
    Could you explain? What do you expect the noise to do? Do you not introduce noise through the ground??
  7. Feb 14, 2015 #6


    User Avatar
    Science Advisor

    Did you read the link I provided on my first response? Look at figures 2 & 3, the shield provides an alternative current loop path for the external noise signal instead of the signal ground loop on the wire inside your cable that I assume is fairly short due to the unbalanced nature of your connections so hopefully the noise levels will be low.
  8. Feb 15, 2015 #7
    Yes I have seen both of your links before and read them both. Sorry but I just wanted to have a more complete answer to my post in hopes of some clarification rather than regurgitation. I dont quite fully understand the text(s) as it is. I also am having a hard time applying it to my situation with the 3 audio: left, right, & ground; and shield on ground vs chassis vs nothing on either side.

    So then the shield should be separate from the signal ground and should attach to the chassis on the input side only?
  9. Feb 15, 2015 #8

    jim hardy

    User Avatar
    Science Advisor
    Gold Member

    Well, That was rude.

    You haven't even said what two devices your wire connects.

    The basic premise of shielding is:
    Shield intercepts capacitively coupled currents. You want those currents to flow in the shield wire not in the signal wires. That's because current in unbalanced signal wires produces voltage drop that directly adds to signal.
    You also want the shield wire as close as possible to same potential as signal wires so that there'll be little capacitive coupling between shield and signal wires.
    That's why you tie shield at signal end of the wire.
    Ideally the shields would form a Faraday cage around the whole measurement system , grounded at the point of measurement.
    Usually one finds shields are connected less than optimally but " we get away with it"..

    I think if you dig through Mr Spook's most excellent links that's what it'll boil down to.
  10. Feb 18, 2015 #9
    Sorry, I tend to be a blunt speaker. No offense meant, but the content of what I said is the truth.
    To reiterate, I have seen most of the documents you can find online, but I am still having trouble fully understanding. I am posting here in hopes of a complete answer or pin-pointing within documents I have not yet seen.
    I appreciate all contributions, whether right or wrong, as long as the real facts are being sought with open minds; I just want to attempt to find the truth with the community, and I naturally assume fellow science fielders would not react emotionally when in the pursuit of the truth in a fair manner.

    Well, you haven't even asked?
    And since you point that out, can I ask if it even matters at this point? Audio signal with conventional 3.5mm audio jack connections going from input to output, relatively standard with no special equipment. My questions were fundamentally: "where is the correct place to connect the wire's shield" & "if I wanted to add EMI (as opposed to RFI) protection (specifically in the form of EMI protection tape that I currently have) on the wire, what is the best way? and then what if we were to take into consideration the existing RFI protection (foil + drain, which is related to my first question)?"
    Will having a real world application to this "problem" really help us resolve them? Once again, not trying to offend, just trying to understand your reasoning.

    Intercepts currents? Does that include magnetic fields? Or do you mean the shielding reacts by generating currents? Not quite sure what you mean, could you explain what "capacitively coupled currents" are? Sorry I am not EE.

    Is this as opposed to balanced signal wires? or, in balanced signal wire, the added signal eventually gets eliminated, so it doesn't matter as much as opposed to unbalanced signal wires? In essence, all signal wires will exhibit this behavior, just that the balanced has its signal cleaned out at a later stage? Saying that this happens "...in unbalanced signal wires..." can be a misleading statement to those that are less knowledgeable.

    Could you explain this further?
    I've read in other documents that the shield would not interfere with whatever is inside it, due to something about electricity radiating outward or another. (I forget the physics of it now, if I find it again I'll post)
    Some say things like: "Electrical rules require the center conductor and inside of the shield to always carry equal and opposite RF currents." (when the shield has electrically equivalent isolated outer and inner layers).

    What exactly is the reason why? Unclear to me from what I have read from you so far...

    What is considered the "point of measurement" for a Faraday cage? Just any point of measurement that moves around anytime you measure it?

    As I have stated in my previous post, I have seen those links.

  11. Feb 18, 2015 #10

    jim hardy

    User Avatar
    Science Advisor
    Gold Member

    magnetic field can't be shielded , it goes right through copper or aluminum shield.
    Electric field can be stopped though by a shield.
    Any two conductors separated by a dielectric form a capacitor.
    A voltage source in proximity to your signal wires can pass current into them through that capacitance.
    The shield forms one plate of a capacitor surrounding your signal wire.
    By earthing the shield you siphon off that current before it reaches your signal wires.


    Again yes. In balanced the noise voltage induced in each wire is the same so it cancels at the differential input.

    Sorry if you felt misled.
    Yes, a balanced scheme with differential input should cancel any noise that arrives in BOTH wires in phase and equal in amplitude.
    Such noise is called "Common Mode" because it's common to both wires , and is referred to signal common.
    The caveat here is no amplifier has perfect common mode rejection .

    If there is voltage between the shield and signal wires there will be noise current coupled into the signal wires.
    Since it is unlikely that both signal wires have equal impedance to ground, there'll be unequal noise current flowing in the signal wires.
    Those unequal currents produce unequal voltage drops along the length of the signal wires, meaning even if you have zero volts of noise between the wires at measuring end you no longer have zero volts of noise between them at receiving end.
    Noise volts between the wires is called "Normal mode noise" as opposed to noise volts between signal wires and signal common(ground).
    The mechanism i just described is called "Conversion of common mode noise to normal mode noise"

    So that shield and signal wires are at same potential, ie there's minimum voltage between them.

    The physical location of your signal's origin.

    That's the ideas behind shielding.
    But you run into practical complications, usually from capacitive coupling.
    What if the amplifier itself is a strong source of common mode voltage?
    You might be better off grounding shield at amplifier so common mode current is shunted to ground there and doesn't flow in your shield all the way out to signal source.

    High level audio typically sent via 3.5mm jacks is low enough impedance that one gets away with murder because the capacitances involved are small.
    Connecting a Walkman to a 1 volt stereo input will be way more forgiving of inattention to shielding details than connecting a 2 millivolt magnetic phonograph to same stereo's phono preamp input.

    Lastly there are cases where the shield is not grounded , instead it's actively driven to same voltage as signal wire. That's called "driven shield" . As above it is intended to minimize capacitive current in signal wires.
    Bottom line is one wants to understand the basics so he can decide how much rigor is warranted.

    Thirty years ago I spent a lot of time in this book
    the author does a good job with the underlying physics
    and i see he's published a couple editions since so others must have found it worthwhile.

    since you have an interest i'd recommend you dig into the subject. It's logical enough once you grasp the basics above.
  12. Feb 19, 2015 #11


    User Avatar
    Science Advisor
    Gold Member

    There is no simple answer to the question. There are some good rules of thumb (ground only at one end at the source , no ground loops etc), but in real situations you will almost always end up having to try a few different options before you find the optimum combination; especially in large systems with many components and cables.
    If you want to dig deeper into the theory you will find that the answer will depend on the signal frequency (which is known in this case), output and input impedance, type of cable, type of signal (balanced or unbalanced),. level of external noise, which frequencies you are most worried about (for shielding high frequency noise it might be better connect the shield to ground at both ends) etc.
    In most situations you will also find that there are other constraints..

    That said, the rules of thumb mentioned above are usually a good starting point.
  13. Feb 19, 2015 #12
    Magnetic shields made of combinations of these metals and mu metals are very effective against low frequency and static magnetic fields. Superconductors can shield 100% , but their use is unpractical today due to requirement of low temperatures.
  14. Feb 19, 2015 #13


    User Avatar
    Science Advisor
    Gold Member

    In most cases you don't have to shield magnetic fields (=inductive pickup) as long as a well designed twisted pair is used. The twisting minimizes the area between the conductors which is what matters in this case.
    Also, static fields do not cause any noise unless the cable is actually moving, so the first thing to do is to make sure that can't happen.
  15. Feb 19, 2015 #14

    jim hardy

    User Avatar
    Science Advisor
    Gold Member

    Indeed, wrapping with something ferromagnetic will help with magnetic interference,

    I'd advise OP to rely on his equipment being designed well

    and simply connect his three signal wires to tip & rings, shield to sleeve.


    ....being sure to slide the cover up the cable before soldering the wires in place.
    Last edited: Feb 19, 2015
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook