Ethernet cables

Ethernet cables are directional whether they are marked or not- at least the silver plated wire ones are.  So if it sounds bad, try turning it around.  I bought some silver plated wire Pangea ethernet cables about a year ago.  They made my previous NUC based music server sound better.  They were not marked but they were directional.  It wasn't hard to tell the difference.  The highs were edgy and grating one way and smooth the other way.  I had the same thought at first- these silver plated wire cables sound worse than my basic cheap ethernet cables.  Then I thought to reverse them and they sounded much better.  I don't think it took much time to tell the  difference.  So I had to try out each cable both directions, and I marked them; but by the third one I could tell right away.  Also, the cable closest to the server is most noticeable with the cable between modem and router the least noticeable.  The Pangea are very reasonable offering silver plated Cardas wire.  I later bought some AQ ethernet cables with thicker silver plating.  It improved the sound but I'd say the improvement didn't match the increase in cost.  I kept the AQ because cost is a secondary issue for me at this point.

Hearing a difference in ethernet cables is system dependent.  The better cables make the TV picture look better but I don't hear a difference with my HT system setup.  With my stereo system the difference is night and day.  I have found a handful of songs that have bad mic clipping with my stereo system.  I can't play those songs on my stereo, it's too grating.  Play the same songs on my HT system or in my car and I can't hear the clipping at all.  The detail just isn't there.  I can hear guitar strings rattle in some songs on my stereo that cannot be heard in the car or on the HT system.  You just don't know it's there until you hear it in a more critical system.  So, careful what you wish for.

I think lugsy is mocking us audiophiles.  

Everyone knows that filling ice cube trays with hot water freezes faster.

I found the wire in the ethernet cables matter. Bits are bits but the high frequency internet data transmission rates behave like an analog wave. The Signal to Noise ratio of digital is superior to analog but it is still not impervious to noise. My theory is that the 5-80 MHz ethernet signal is affected by the skin effect of wire. Therefore, silver plated ethernet wires might be sufficient for improved sound vs. full silver wire. The required thickness of silver plating for the skin effect can be calculated based on the frequency of the ethernet signal. My theory could be all wet but it seems to work in practice. Pangea ethernet cables are relatively inexpensive- about 8 times the cost of a basic budget cable. For the money however, you get silver plated cardas wire and good, robust connectors. I found them effective at improving the sound with my NUC based music server, but as I said, they are not marked but are directional.

Another good experiment that was discussed in these forums a few years back is to have someone disconnect the ethernet cable from your music server while listening to the music. Since the music is delivered in packets and buffered, it will continue to play for several seconds. Many of us noted an improvement in the sound for those few seconds of disconnected playing. Interesting, especially since that last packet of music was already downloaded. That implies the ethernet cables are carrying in noise that affects the server and DAC. It would seem that my NUC based music server that I had at the time had no isolation. So I added a fiber optic link to my ethernet to disconnect my music server galvanically- or electrically. The sound improved. Next I tried the Pangea silver plated ethernet cables and found more improvement. And finally I did the big step and bought an audio grade network switch. I didn’t expect much but found it a good step up in sound. I also found it performed better without the fiber optic link.

To summarize: The ethernet behaves like an analog wave. It’s not analog but the high frequencies have characteristics that mimic analog.

Streaming music files behave differently than downloaded music files.

Line noise/EMI is a factor in streaming music.

Silver conductors seem to work better with high frequencies- i.e. above the audible range. Maybe.

The better servers, streamers and DACs are moving towards isolation and reclocking. My new DAC and Music server/streamer have double isolation and reclocking. They sound fantastic but the downside is that I cannot turn them off. Otherwise it takes a couple of days for the clocks to stabilize for optimum sound. Hard for me to reconcile since I grew up in the midwest with a Dad who didn’t allow lights to be left on in a room unattended. And another thing, it took me about 20 years to stop eating all of the food on my plate.  So maybe in a few years I'll get over having to leave half of my stereo on all the time.

The tendency is to compare digital signal transmission to a telegraph.  After all, the dot-dash DC circuit of the telegraph is similar to the 0-1 of digital.  And like the DC circuit of a telegraph the voltage of a digital signal is either 0 or 5 Volts.  We tend to imagine an analog wave form as something like a +/- sine wave or combination of sine waves overlapping.  Digital signals are waves too.  They do not cross zero but they have a bias of 2.5 Volts.  ie. over time the average voltage of a square wave of a digital signal will be something around 2.5 Volts.  That means the digital square wave "zero crossing" is at 2.5 Volts.  Given that these square waves are propagating at 5 MHz and higher, they share many properties of analog waveforms.  They can experience phase shift due to capacitance or inductance and they can pick up distortion although their tolerance for distortion is much higher than analog.  On the other hand, digital signals for streaming are less tolerant to phase shift or timing.

 Ethernet and USB cables have an impact on the sound of streaming music that many claim is impossible.  Meanwhile, audio hobbyists may not fully understand why or how but find ways to improve and move forward the state of the art.

Maybe I'm talking out of my hat here but much of what we experience with streaming hifi manifests itself more as a wave than not.

They are transmitting digital data not analog signals- False

If you tap into your ethernet cable do you see 0's and 1's?  You do not.  You see a square wave ranging from 5-80 MHz.  A square wave is.....wait for it.... ANALOG.  It is a wave.  It is shaped different than a sine wave but it is a wave.  It is analog.  Now since the square wave is interpreted at the receiving end as 1's and 0's it is not as susceptible to analog noise- ie. static and hiss, but the square wave's weakness is distortion causing timing and phase shifts.  And EMI or RFI can get in the line and ride all the way into your DAC, and that will affect the performance of a D/A convertor.  

No one will be able to tell the difference- False.  Better cables do a better job at interference noise rejection.  That is quite noticeable to the performance of D/A convertors, especially since lower cost DACs and music servers have little to no noise rejection on their inputs.  The higher end, more expensive DACs and Music servers have better noise rejection, isolation of the incoming signal and more often now, re-clocking.  At least they should.

I didn’t say that about IEEE and you know it.  I said a square wave is analog.  Agree or disagree and explain why.

You are free to buy cheap cables.  They will make sound.  Why are you on this forum is all you want is sound?  This forum is for the pursuit of hifi.

This is a summary of about 3 years of study on the subject and experimentation on my part to confirm or dispel others’ claims. I can’t cite a single source that would explain it all. Much of the information is scattered about. I’ve not had technical courses or been tested on this subject matter so I am open to corrections and clarifications. But you will find references to digital signal transmission having an analog component. I think a lot of the information is held closely as trade secrets and to hold a competitive advantage. The IEEE specs for data transmission has protocols and standards that the industry follows, but beyond that it is up to engineers and designers to finds ways to beat out everyone else. In time more information will find its way into classrooms.

When it comes to cables- Caveat Emptor. Yes, a lot of hocum in the world of cables. I stick to the more established brands with a pedigree. They often cost more but they also resell better. If you research wire, only a handful of wire makers exist in the USA with the experience and know- how to draw wire for the best sound in audio. The drawing process is crucial. Finding out who the companies are would be a real feat. The cable makers hold their cards close to their chest. Beyond that, materials, weave types and assembly processes play a crucial role as well to the sound of cables.  Look for companies with the technical resources to develop cables.  Not just put something together that looks pretty.  They need to have a lab with the proper test equipment.  Another thing is to take note of what cables vendors use at audio shows like Axpona.  Not the rooms specifically selling a brand of cable but the rooms that are showing off their speakers, amps or sources.  I actually saw a lot of rooms using the same few brands.

As for Ethernet and USB cables, 4 years ago I did not believe they could impact the sound at all. Digital has come a long way. CD players up to the early 90s couldn’t hold a candle to vinyl’s low level detail and the ability to create a 3D soundstage. Now my digital rig is as good as my best vinyl. That is is all done digitally- creating lifelike images in space like analog totally amazes me. But just like with vinyl, everything matters.

I can be specific on what exactly the silver plated wire ethernet cables did for my system with my previous NUC based music server.  The clarity of the music improved.  The highs smoothed out- less of that glare.  The first time I plugged in one of the better cables the highs sounded worse.  I let it break in for a day but it still sounded worse than a cheap cable.  Then I thought to turn the cable around and the sound changed for the better- much better.  It made streaming sound closer to my CD Transport going into my DAC.

I have since moved up to a better DAC which then caused me to have to upgrade my music server as well.  My theory is that this DAC and music server, because they both have isolation and re-clocking, they should be less dependent on the digital cables (Ethernet and USB) for good sound.   Only right now my system sounds too good to tamper with it.  I’m just not motivated to do any additional experimenting.  You know, if it is not broken, don’t fix it.  Sometime I will, out of curiosity, put some cheap cables in my system to see how this DAC and music server perform.

I have seen videos of Cardas assembling their cables.  They do not draw the wire themselves.  They say the wire is drawn in the Northeast from copper mined in the Southwest.  Perhaps the diagonal (ie from Southwest to Northeast) has something to do with the sound. :)  Good luck finding a video of audio wire being made (drawn).  Everyone holds their cards close to their vest when it comes to turning copper/silver/gold into wire for audio.

Cold welding or crimping is the preferred method for attaching terminals to audio cables.  I can believe that.  Heat based welding could affect the grain structure and temper of the copper wire.  Soldering not so much but that is an expensive process compared to crimping.  I have experience with crimping from my automotive days.  Due to some issues I learned more about crimping wires than I ever cared to learn.  The most important fact is that the crimping process must be very well controlled and closely monitored.  Insufficient crimping leads to oxidation and failure while too much crimping can crack the housing of the terminal leading to even faster oxidation and failure.  That means periodic audits of crimp joints- sectioning and viewing under magnification is required.  Something that I have not seen mentioned by any cable makers in their videos.  In fact I have only seen cable makers using hand crimp tools which have more variation than automation.

The less you flex your cables, the longer they will last.  Stress on the crimp joints can loosen the wires over time.  Proper crimp joints can last hundreds of thousands of cycles.  Less than proper- well, we saw degradation fairly quickly.  The only way to tell if a crimp joint is good is to section it.  

I try not to think about it.  Typing this isn't helping.