Pimping your router

Main problem I have with my router is the bits keep getting stuck.

@mikem 

Ideally a DAC would take care of ALL external negative "influences" presented by the peripherals (i.e. switches, routers, etc.). The reality is quite different, and while the DAC manufacturers have taken tremendous steps towards isolation, there is still work to be done. 

Keep in mind,  much of perceived digital audio performance is dependent upon the quality of the clock signal entering the DAC chip and the presence of jitter.  Jitter causes distortions in the audio signal coming out of the DAC.

John Swenson has done quite a bit of work on this, and identified two kinds of external influences that can cause noise on a DAC's internal ground-plane: 1) jitter carried through digital data, and 2) leakage current (i.e. from power supplies of the various networking devices, router, switch, in your network).

Leakage current can flow through power cables, audio interconnects, digital cables, including Ethernet and USB cables. The leakage will flow from a power supply, through a cable to another device or component, then through that device or component’s respective power supply back to the AC mains.

Alternatively it can flow through a cable to another box and then through a safety ground (the “third pin” of an AC plug) and return back to the AC mains.

In fact, there are two types of leakage current, traditional low-source-impedance leakage, which is found in all power supplies, and high-source-impedance leakage, which only occurs in SMPSs. This high-source-impedance leakage is hard to block and can take surprising paths to get where it wants to go.

These two different forms of leakage currents have a deleterious effect on Ethernet in audio systems. All copper Ethernet connections have transformers in each device port, and these transformers will block DC and low-source-impedance leakage. However, they DO NOT block high-source-impedance AC leakage. This is particularly important for digital networks used for audio. Most all network systems used with audio use switching supplies (SMPS) to power the digital devices, routers, switches, computers, etc. The high-source-impedance leakage current from the SMPS will travel through the network equipment, the Ethernet cables, the end-points, and into the DAC where they will create noise on the ground-plane, and thus jitter in the DAC circuitry

I recently added an ENO Ag (fed by Monoprice CAT8) in front of my Roon Nucleus, that was already being powered by an HDPlex 300 LPS, which was also feeding 5v power to my Schiit Yggdrasil DAC with Unison USB, via a USB Disruptor.  It made a clear improvement.

So did replacing my 2013 MacBook Pro running Roon with the Nucleus, adding the LPS to the Nucleus, and adding the 5v LPS to the USB Disruptor (in decreasing order of effect size).  In all cases, the effect was a lowering of a noise floor that I thought was already non-existent, or at least below what I could hear before the change was made.  Low level information that was previously not there, became easily apparent.  The decay of a note, the decay of a room, the shuffle of a foot on the floor, the squeak of a chair.  Details and more details.  And guess what?  More of these details all add up to realism, and that's what I am aiming for.  I want a time machine that takes me back to the instant of the musicians playing the take on the recording.

So, to me, with no need to do double-blind testing, the effects are real.  Heck, even if they are pure placebo, I don't care, as long as I hear them.  And I do.  So that begs the next question, why did those changes make that improvement?  Some might want to know because they want to understand the world.  I just want to know "why", so I can make more such improvements, to make my time machine better.

So "improvements" might have to do with lowering some property of the electrical noise floor.  I don't know.  Show me some data. But I do know they make the music better.  The why is a mystery at this point.  One thing that I do wonder about: in every waveform analysis that I have seen (have not seen that many though), the digital 1's and 0's are far from perfect.  Overshooting, ringing, not perfectly vertical or horizontal waveforms.  So they are really an analog representation of a digital signal.  I know they work fine to deliver this overly long post full of conjecture, but I am more uncertain of the effect of non-perfect digital 1's and 0's on the reproduction of the very fine details of music.

Enjoy the music!!

-docknow

@lalitk @docknow @thynameThank you all for the responses ! I’m not quite convinced! But I appreciate the civil discussion and the chance to learn..