I had been using the D-Link DGS-1005 plastic switch. Yesterday, I picked up the metal D-Link DGS-108 with shielded jacks. Stock D-Link power supplies on both.
Speed has never been the problem since the ethernet system is closed and used only for Roon and the distances are minimal. I use Roon with an iMac desktop system. The audio gear is a meter away from the iMac on a rack. So unfortunately, I cannot have the ideal setup of having the computer in another room as the iMac is needed. Even if the Roon was on another computer in a room far away, the iMac would still be in use. So might as well run the Roon on the iMac. However, the iMac and other power supplies for the router and switch are plugged outside the Torus. All the audio gear on Torus isolation transformer. Stock 14 AWG shielded power cables for all the gear.
The three cables are 6a S/STP 1 feet, 5 feet and 7 feet long. Hardly challenging given the load and distance.
The 7 footer goes from the iMac to a Linksys E3000 router that has its wireless disabled along with other stuff. The Linksys doesn’t have shielded jacks. Then a 1 footer from the Linksys to the DGS-108. Then a 5 footer from the DGS-108 to the BDP-1. Both the DGS and BDP-1 have shielded plugs. So that’s the only shielded connection front to back in the chain. The DGS-108 has a ground screw at the back, however, I haven’t utilized that at the moment.
Anyways, it’s hard to conclusively say whether it’s the power supplies interaction or the airborne crap from the iMac in near proximity or both that might be affecting other gear. Not the most controlled setting.
I will say that having gone from the 5e and 6 UTP to 6a S/STP made a positive difference. Switching out the DGS-1005 to the DGS-108 also seems to have made a positive difference. Although, it’s been a day so I won’t reserve my impressions as final that came with the change in network switch.
I’ve tried many combinations with the routing and cabling and there has never been a difference on the BDP-1’s CPU’s readout. It holds steady at the same number for most part as expected.
The change in sonics that I experienced with ethernet tweaking follows roughly the same changes in sonics that I experienced when experimenting with the BDP-1 with hard drive storage in MPD. For example, powering a portable hard drive off the BDP-1 vs. a flash drive, or using an external hub, with or without a Jitterbug.
The improvement in clarity almost always seems to accompany changes in how deep or dynamic I think the bass feels. With noisy combinations, the bass seems flat and doesn’t go as deep. The benchmark for the worst was feeding the BDP-1 from a Powerline adapter. The high-end extension goes out the window. Bass goes flat and soundstage closes in. The tonality stays the same.
On the Amphions, I do find with vocals there is better phase accuracy. The phantom center holds more steady. As the system gets progressively ‘noisier,’ the image starts to deviate and has a bit of an artificial haze.
I do have an incoming Teradak U9 and a spare Jitterbug that I might try with the switch. Even if it doesn’t make any improvements directly or indirectly, I still think it’s a good idea to remove the stock switching away from the proximity of audio gear.