In the last week, I made some changes to make things easier for better testing environment. I moved all the network gear and their power supplies at least 10 feet away from the audio rack to keep EMI out of the picture. From there I only ran one single ethernet cable that would plug into the BDP-1. This is the only ethernet cable that was in the vicinity of the audio rack. All power cables are shielded 14 AWG going into the Torus. The other cable in that area is the Grimm TPR for both analog and digital (AES) use. 3 feet of TPR in each cable.
The final switch is powered by Teradak LPS. The switch is DGS-1005 which is unshielded, so there cannot be ground loops regardless. Remember, BDP-1's ethernet port is shielded.
I tried 3 ethernet cables of the same length:
1) Cat 6 UTP
2) Cat 6a SSTP (grounds lifted on both side)
3) Cat 6a SSTP (ground connected to BDP-1)
The UTP had some grain to the sound with a slightly warm sound and felt the loudest. On the same track, I was comfortable at 9o clock.
With the #2 option of floating ground on both side, I think it was slightly quieter than the UTP. So, each time I felt I had to turn the volume up to 9:30 ish on the pot. It was a tiny bit more neutral sounding, a bit less grain, but still nothing that would be appreciable to most.
With the #3 option of ground connected on one side, the sound was distinctly different. Grain/hash in vocals was completely gone. In fact, it almost felt too smooth at first. I also had to turn the volume up on both the headphone amp (Audeze LCD-2C) and the speakers to 11:00. The soundstage became more fluid on the Audeze in particular.
I absolutely had to turn the volume up quite a bit to get the music going. However, as I continued to turn the volume up I never noticed any stress. I had to do a double take on the volume level. I'm never at these volume ranges with that music.
__________How big is the difference?
Between #1 and #2, I think some listeners might be able to pick up some differences, but I doubt anything would come out in a blind test. However, if you put up #1, or even #2 against #3, I think a lot of listeners here would be able to pass a blind test. It's THAT noticeable. That smoothness in the vocals will be a dead giveaway each time!
I remember reading that you should not have the switch near the audio rack as the cable leading up to it will still be producing noise. It's best to keep all the network gear and power supplies far away from the gear and only have the one cable going into the audio rack.
The fact is that both screens and shields and the copper balanced twisted-pairs in a UTP cable will behave as an antenna to some degree. The difference is that, as demonstrated by the simplified loop antenna model, the noise that couples onto the screen or shield is actually 100 to 1,000 times smaller in magnitude than the noise that is coupled onto an unshielded twisted-pair in the same environment.
As shown in figure 7, the coupling between two UTP cables (shown in black) is a minimum of 40 dB worse than the interaction between two properly grounded F/UTP cables (shown in blue). It should be noted that 40 dB of margin corresponds to 100 times less voltage coupling, thus confirming the modeled predictions. Clearly, the UTP cable is radiating and receiving (i.e. behaving like an antenna) substantially more than the F/UTP cable!
Very happy with the BDP-1 this way. Right now, it truly sounds and justifies its price. It has to be heard this way.
This is how I'd recommend reviewers to hear the BDP's at their best. I've done the simple and isolated flash drive connection with MPD and it doesn't compare.
I think I might be done with copper ethernet tweaking. Simple rules, some inexpensive upgrades to cables and power supplies and a lot of space between them.