Shielding components from EMI/RFI... Help please

03-04-12: Almarg
...for shielding to be effective at high frequencies, as Dr. Johnson indicated the shield should be grounded at both ends. But I believe that the key element of what he is referring to by "grounding" is a connection at each end between the shield and the metallic structure of the component, rather than a connection to some external ground point. I believe that circulation of noise currents from the cable shield into the metallic structure of the components will dissipate their energy significantly, although perhaps less so in the case of the network switch due to its small size.

I've been giving more thought to this, and I decided to experiment further with an alternative grounding scheme. I disconnected the ground wire between the ethernet switch and the power conditioner (thereby removing the switch's connection to earth ground), and I added a braided grounded strap between the ethernet switch and a 6" x 12" aluminum plate. My hope was that, by "enlarging" the surface area of the ethernet switch, it would provide more adequate dissipation of RFI or other noise picked up by the long ethernet cable between the computer and the switch.

The results are at least as good, and possibly better, than the result of grounding the ethernet switch to the power conditioner (which, in turn, was grounded to the preamp, the amp, and earth).

Bryon

Sounds good! With respect to placement of the plate, if you haven't already done so it might be worth trying it positioned directly under the switch.

Best,
-- Al

Yes, Al, the plate is directly under the switch. I placed it there to keep the braided ground strap as short as possible, to minimize any antenna effect. Out of curiosity, is there another reason why it should be directly under the switch?

Also, I neglected to mention that I also changed another variable: I replaced the 1' unshielded ethernet cable between the switch and the Sonos with a SHIELDED version. So prior to this change, the configuration was...

computer -> 50' shielded Cat6 cable -> ethernet switch (grounded to power conditioner/earth) -> 1' unshielded Cat6 cable -> Sonos -> reclocker...

And my continuity results were...

Ethernet switch to Sonos input… NO continuity
Ethernet switch to Sonos output… NO continuity
Ethernet switch to reclocker input… NO continuity
Ethernet switch to reclocker output… continuous
Ethernet switch to Meridian G68… continuous
Ethernet switch to Pass amp… continuous

So the break in continuity was due to (a) the galvanic isolation in the reclocker and (b) the UNSHIELDED ethernet cable between the switch and the Sonos.

Now the configuration is...

computer -> 50' shielded Cat6 cable -> ethernet switch (not grounded to power conditioner/earth, but grounded to the "grounding plate") -> 1' SHIELDED Cat6 cable -> Sonos -> reclocker...

And the continuity results are the "OPPOSITE"...

Ethernet switch to Sonos input… continuous
Ethernet switch to Sonos output… continuous
Ethernet switch to reclocker input… continuous
Ethernet switch to reclocker output… NO continuity
Ethernet switch to Meridian G68… NO continuity
Ethernet switch to Pass amp… NO continuity

So now the upstream half of the system (computer, ethernet switch, Sonos, reclocker input) is discontinuous from the downstream half (reclocker output, preamp, amp).

In this new arrangement, the high frequencies are improved, which is consistent with my listening results from 3/3, where the presence of a SHIELDED ethernet cable between the switch and the Sonos produced the best high frequency results. At the time, I didn't go with a shielded ethernet cable between the switch and the Sonos because the bass was lacking pitch definition with that arrangement. But with the "grounding plate" for the switch in place, the bass pitch definition is excellent.

My conclusions...

1. The variation in the bass was a function of grounding the ethernet switch to a sufficiently large chassis, which was originally provided by the power conditioner and is now provided by the grounding plate. The variation in the bass was NOT due to the presence or absence of EARTH grounding, as I originally believed.

2. The variation in the highs was a function of the presence or absence of shielding in the ethernet cable between the switch and the Sonos. I suspect that shielding that cable reduces noise, either...

(a) noise transmitted from upstream components (computer, switch) through the ethernet cables, or...

(b) noise transmitted from nearby components (amp, preamp, Sonos, reclocker) through the air, or...

(c) noise transmitted from downstream components (preamp, amp, power conditioner) through the ground wire between the switch and the power conditioner (that ground wire is now gone, but it was present on 3/3, when I experienced the same variation in the high frequencies).

Not sure which of these was the culprit.

Bryon

The plate is directly under the switch. I placed it there to keep the braided ground strap as short as possible, to minimize any antenna effect. Out of curiosity, is there another reason why it should be directly under the switch?

Just my vague intuitive feeling that placing it there might enhance the effectiveness of the switch's internal ground plane, perhaps reducing the amount of digital noise generated within the switch that might end up coupling or radiating to points downstream.

With regard to the well thought out summary you presented in your last post, my only comment is that it would seem to make sense that the best results were obtained via a configuration that does not provide any paths that bypass the reclocker and its galvanic isolation. As you noted, such a path previously existed, via the ground wire from the switch to the conditioner, and from there via power cords to the downstream components.

Well done! Best,
-- Al

I agree that the current configuration seems to maximize the benefit of the reclocker's galvanic isolation.

I have a follow up question for you, Al, about the dissipation of noise in an equipment's chassis... Is the dissipation potential of a chassis determined by its surface area, its mass, its material, its conductivity, its magnetism...? Several of the above?

bc