Most of us understand the importance of cable selection and what characteristics are ideal. When selecting power cables for a “conditioner” unit we typically are concerned with quality of manufacturer, gauge, capacity, etc.
Has anyone ever considered what role the shielding plays in the sound of their system? Does anyone feel that a shield is necessary (or not) for power that’s being filtered or regenerated? Pros or cons are welcomed but don’t berate me if missed the class that day.
I use PSAudio AC5 on my two regenerators. Mostly vintage Cardas, DIY 12ga. Furutech (wire and AC connectors) & Cullen Cables on everything else. Two systems.
On the subject of shielding in AC cables and other related topics, from the guru himself:
I have spent a great deal of time and effort on this issue of power cable shielding and can share some of my thoughts with you. A conventionally shielded power cable where both ends of the outer shield are tied to ground is more an antenna than a shield. A
few cable manufacturers get it right - like our power cables - where the outer shield is tied only at the IEC receptacle ground and not at the male plug end. That’s for starters.
One other characteristic I have noted. If the shield is too close to the conductors we get a pinched character to the sound. You want the shield as far away as possible. Or, no should at all, like AudioQuest Dragons I like. Here they do a twist
in the conductors to help lower EMI.
The biggest problem you’re working to solve is not shielding against radiated noise coming from the wall socket, but rather the noise coming from the equipment itself. DACs, for example, a really noisy and if you don’t manage to take care with
cables, you can spew much around everything else in the system.
Power cable shielding acting as an antenna…
Leave it to the wisdom Of Paul to bring to my attention something that I would have never thought about before today. I had read somewhere about the negative effects of the shielding being too close to the conductors and now I feel that has been confirmed as well. Very useful and helpful information Paul!
Thanks,
Definitely interesting. I am saving for a TWL Obsession cord and doing my research now on this…good learning moment!
Higher capacitance is added with a shield. Lowering capacitance is a DISTANCE factor.
Grounding at the SEND end of a true shield, not an unbalanced cable ground path (coaxial cable) is called an antenna ground. This keeps the origination path noise grounded at that end and not out on the cable where it can inductively couple. RF is routed to ground with a low impedance, to RF, shield to the same ground point.
Low attenuation is important in a shield to lower inductive coupling to the core components. You want R to be LOW in a shield at RF. You don’t want voltage to be dropped across the shield, but into the ground point.
The FCC used to require you to TEST and certify that your devices don’t have emissions above class A or B requirements.
The home emissions level is stricter than the industrial limits, too. This helps offset poor cable or device placement the average person will use in a home.
If the device didn’t pass FCC emissions with UL’s worst case power cord you are required to have power cord that passes permanently (captive cord) attached. The power cord is part of the device.
It is no secret that PC, and PC monitors are emitters. This is why many PC’s and monitors had captive power cords with ferrite chokes at the ends. Stuff is better designed such that those awful retention force IEC, most all with no locking, outlets are now common.
If you can pass the test with a worst case situation, a UTP power cord, then an IEC outlet can be used.
Ground impedances at RF aren’t what people think. Proper termination at the ground point of the shield is more important than the shield! Most all AC plugs don’t allow a true and proper RF shield termination. It is all about what lowers the apparent contact resistance at each frequency.
Consider the device. If it has an IEC outlet, it should pass FCC emissions standards with ANY cord attached! That’s the facts. But, if you want to be careful consider what the component is. If it is a digital device, like Paul says, they are prone to being noisey. They are what cause the problems, not the ingress from the outside world. FCC looks at INGRESS and EGRESS, both. YOU may be the problem!
You may consider a shielded power cord on THOSE noisey devices. A pure analog device not so much.
Shields alter the ground plane reference values over distance. This makes a cables electricals values irregular based on the presision of the grounds physical proximity to the core wires. UTP cable are “shielded” but they are shielded to the reference world outside the cable core, that’s the earth as it may be. The true E field is stopped by the inner surface of the shield until the E field turns into a B field at very low frequencies. The B and E fields magnitude switch places as frequency drops. At RF, the E field inner shield distance determines the cables “impedance” across frequency by altering the capacitance.
A shield creates another version of earth but one that is far, far closer to the cable. We can determine what this earth is, and manipulate it to alter cable electrials (shield, wire and core size etc on a coaxial cable).
Since the core is so CLOSE, the effects of physically moving it is very pronounced. It is imparative at RF to manage the shields physical distance over length or it is actually WORST than no shield at all. This is why shields are best when it is verified that one is needed. An Unshielded Ethernet cable has a far better internal electrical NEXT, Near End Cross Talk, and ACR, Attenuation to Crosstalk Ratio, than the shielded version. Internal cross talk is 6 dB better on the UTP cable. A shield creates inductive coupling of the internally captured EM fields. UTP is also has a more stable electrical impedance cross frequency. This improves RL, Return Loss reflection, off a load at RF.
Why use a shield? Easy, when the noise coming in alters the signal to noise ratio MORE than the negative effects of using a shield. If I have noise below that threshold, it is better to use a UTP cable. If the cross talk is 6 dB worse with shield (it is), but ingress noise makes it 12 dB worse, a shield adds 6 dB of cross talk margin than a UTP, for instance.
The lower you go in frequency the less the shields physical perfection matters as the EM wave ignores the shield! Yes folks, a shield won’t stop low frequency noise. Only a low permeability path (a magnet will stick to it) shields the noise we see around motors and AC high current lines. Here we use passive cancellation (XLR type designs) as the most practical method to remove what we don’t and keep what we do. The next cheapest thing to do is DISTANCE as EM fields drop with distance squared.
Sure, you can put your cables into low permeability conduit and get the best of E and B field shielding. Just make sure you are good at tube bending!
Best,
Galen Gareis
Yes, and thanks to this question you inspired Friday’s post.
Thank you Galen Gareis, for sharing your wealth of knowledge, and experience with me (forum readers). Electrical engineering is something I’ve always marveled at and your explanations are always enlightening for a greenhorn such as myself. I confess that at times I struggle to digest all the scientific principles involved but I invariably leave with a greater understanding. The series that you wrote about speaker cable design for Copper Magazine was a considerable step forward for me as an evolving audiophile.
Thanks again!
Paul,
Well, maybe shielding is the answer, but as soon as I installed the Stellar PP3 in my system (not a rant on the PP3; it’s awesome!), I had a problem with a low-level popping noise in my monitor speakers whenever I turned off the room light switch. Was that switch a spark gap xmtr?
What solved the issue permanently was replacing the AC cords on the Stellar GDAC, m700 amps and Stellar PP3 (going into a standard AC wall outlet) with AudioQuest NRG-Z3 cords. So did shielding solve the popping? Whatcha think?
I use an unshielded power cord with Furutech copper connectors on my Bryston DAC. Plugged into a power distributor. No audible RF/EMI. If anything, I got an expanded soundstage and better transparency than with a shileded pc. ymmv as always. I might try unshielded on my power amp, but it’s really transparent already, and musical.
I have the same thing happen to a p20 when switching particular lights/ceiling fan. But I also have shielded power cables
I’d think replacing the switch with a higher quality one would be a simple solution.