The Great Cable and Interconnect Swindle: An Etiology

[jabbr]

3 hours ago, gmgraves said:

Sure, if it’s bad enough and at a high enough level because the shielding on the interconnect wires is inadequate to keep it out of the interconnects, it can and will become a factor in compromising the sound of a system. But it doesn’t exist everywhere.

 

You need to think of this as a ground loop -- and this is exactly what causes ground loops. Common mode noise travels forward across an input and the ground forms the return loop. What I've talked about isn't EMI coming from outside. Surely you've encountered ground loops? (aka leakage currents)

 

There aren't always ground loops but just because you don't have a ground loop in a particular setup doesn't mean that you or someone else doesn't ...

 

https://web.mit.edu/jhawk/tmp/p/EST016_Ground_Loops_handout.pdf

[sandyk]

3 hours ago, gmgraves said:

I have a 100 MHz Tektronix ‘scope, yet at maximum sensitivity, I see no RFI/EMI On any interconnects on my system.

 George

Does the Tektronix have a higher sensitivity than most older CRT CRO's ?

 With my Preamp I had to use a battery powered low noise 10 x preamp as well to highlight stuff such as SMPS from the nearby TV to much more than a thick centre trace. Many have reported that even this level is audible on a resolving system.

 

Regards

Alex

[gmgraves]

18 minutes ago, jabbr said:

 

You need to think of this as a ground loop -- and this is exactly what causes ground loops. Common mode noise travels forward across an input and the ground forms the return loop. What I've talked about isn't EMI coming from outside. Surely you've encountered ground loops? (aka leakage currents)

 

There aren't always ground loops but just because you don't have a ground loop in a particular setup doesn't mean that you or someone else doesn't ...

 

https://web.mit.edu/jhawk/tmp/p/EST016_Ground_Loops_handout.pdf

Sure, of course I’ve encountered ground loops. I’m not saying that EMI doesn’t exist, or that it can’t influence the sound of a cable. But I am saying that it’s not present in every system. Which seems to be exactly what you are saying as well. 
Also, I don’t have a ‘scope camera and so I didn’t take any screen shots. I have tried to take ‘scope shots with my iPhone before, but that doesn’t really work very well. The iPhone won’t focus close enough to get a decent photo. Perhaps if I made a cardboard “tunnel” for a light shield, but it would have to be a long one.

Anyway, there’s really not much to see. It’s just a fuzzy, slightly wavy line. The fuzz is not resolvable because the time-base on the scope is limited to 100 MHz, and this fuzz is obviously much higher in frequency. It’s also (near as I can figure) about 0.0003 vpp. IOW, inconsequential background noise. I would like to see it’s composition, though.😉

[sandyk]

4 hours ago, gmgraves said:

Noise at -120 dB or less, is simply not going to “color” the sound

 I disagree.

 I have seen measurements taken of a front end modified good quality amplifier, where DIY Audio members have verified hearing differences in the -130dB area.

[fas42]

4 hours ago, gmgraves said:

All of what you say is true. But still, many contributors of this forum act as if every system suffers from RFI/EMI noise, and it just ain’t so. I also strongly doubt that RFI/EMI has anything to do with the “sound” of interconnects In the vast majority of situations. Sure, if it’s bad enough and at a high enough level because the shielding on the interconnect wires is inadequate to keep it out of the interconnects, it can and will become a factor in compromising the sound of a system. But it doesn’t exist everywhere. I have a 100 MHz Tektronix ‘scope, yet at maximum sensitivity, I see no RFI/EMI On any interconnects on my system. Blaming RFI/EMI for everything from the CV-19 virus to the sound of cables to the Isis insurgency in Syria is only muddying the issue. RFI/EMI Is not a panacea.

 

Au contraire ... 🙂.

 

I would quite happy to say that "every single audio system suffers from RFI/EMI noise" - I have always found it trivially easy to introduce some source of interference into the environment of a system, and hear the SQ degrade as a result ... if someone says a vehicle has fabulous suspension, just bring it to the right sort of road surface, or pothole - and feel the car fail to handle it. There are always limits; the important thing is not to pretend they don't exist - but to explore them, and have a full understanding of where they may have a bearing - crossing one's fingers is not the best engineering approach ... 😉.

[gmgraves]

1 hour ago, sandyk said:

 George

Does the Tektronix have a higher sensitivity than most older CRT CRO's ?

 With my Preamp I had to use a battery powered low noise 10 x preamp as well to highlight stuff such as SMPS from the nearby TV to much more than a thick centre trace. Many have reported that even this level is audible on a resolving system.

 

Regards

Alex

That’s a good question Alex. I have to say that I really don’t know. Frankly, until you asked the question, I never even thought about it!
The ‘scope is a Tektronix 465. It’s a 2 channel, 100 MHz, solid-state CRT-based scope I bought on E-bay. back in January. I had an older Tek scope before that, and I gave that away. I recently bought a battery powered hand-held scope that has a color LCD screen, and I find myself using that more and more. It’s just much easier to set-up and use!

[gmgraves]

51 minutes ago, sandyk said:

 I disagree.

 I have seen measurements taken of a front end modified good quality amplifier, where DIY Audio members have verified hearing differences in the -130dB area.

We’re gonna have to agree to disagree here. -130 dB is so far below the threshold of human hearing, that anyone who says that they can hear something so low that it’s below the thermal noise on any active device operating at audio frequencies has a great imagination.

[gmgraves]

17 minutes ago, fas42 said:

 

Au contraire ... 🙂.

 

I would quite happy to say that "every single audio system suffers from RFI/EMI noise" - I have always found it trivially easy to introduce some source of interference into the environment of a system, and hear the SQ degrade as a result ... if someone says a vehicle has fabulous suspension, just bring it to the right sort of road surface, or pothole - and feel the car fail to handle it. There are always limits; the important thing is not to pretend they don't exist - but to explore them, and have a full understanding of where they may have a bearing - crossing one's fingers is not the best engineering approach ... 😉.

Frank, a car with a fabulous suspension is SUPPOSED to feel every change in road surface and should feel even tiny potholes. That’s what a “fabulous suspension” is designed to do. “Smooth ride” is absolutely diametrically opposed to a “fabulous suspension”.

[sandyk]

51 minutes ago, gmgraves said:

We’re gonna have to agree to disagree here. -130 dB is so far below the threshold of human hearing, that anyone who says that they can hear something so low that it’s below the thermal noise on any active device operating at audio frequencies has a great imagination.

Several members are now routinely hearing differences of a small fraction of 1dB, and I think that you will find that Chris is now hearing extremely low level differences on very well measuring gear when using his RAAL-requisite SR1a headphones.

https://audiophilestyle.com/forums/topic/58443-raal-requisite-sr1a-has-landed/        #348

[jabbr]

1 hour ago, gmgraves said:

Sure, of course I’ve encountered ground loops. I’m not saying that EMI doesn’t exist, or that it can’t influence the sound of a cable. But I am saying that it’s not present in every system. Which seems to be exactly what you are saying as well. 

Yes. Really I am suggesting a (testable) possibility. I don’t spend endless amounts of time thinking about USB cables but lots of people seem to hear differences. It’s easy to completely eliminate the possibility of ground loops/leakage currents etc with fiberoptic Ethernet and it’s cheap so I use it. 

 

[sandyk]

23 minutes ago, jabbr said:

It’s easy to completely eliminate the possibility of ground loops/leakage currents etc with fiberoptic Ethernet and it’s cheap so I use it. 

 

 And the wider the Bandwidth of the fibreoptic Ethernet, the better it is for passing  along any UHF (and higher) RF/EMI that may be riding along with the Signal from the source device ?

 e.g. SSDs can inject noise into the PSU with harmonics that are capable of upsetting close proximity 2.4G devices and the EMI noise falling on 1.57 GHz band is excessively high for GPS applications

14A2-B4.pdf

[jabbr]

Just now, sandyk said:

 And the wider the Bandwidth of the fibreoptic Ethernet, the better it is for passing  along any UHF (and higher) RF/EMI that may be riding along with the Signal from the source device ?

 e.g. SSDs can inject noise into the PSU that is capable of upsetting close proximity 2.4G devices and the EMI noise falling on 1.57 GHz band

is excessively high for GPS applications

um no. the specification for 10Gbe Ethernet is very very tight and higher bandwidth even tighter. In any case you don’t call that EMI because fiber doesn’t carry that. The specs don’t allow noise. Ask Tektronix because they make big $$$ with test equipment. 

[fas42]

1 hour ago, gmgraves said:

Frank, a car with a fabulous suspension is SUPPOSED to feel every change in road surface and should feel even tiny potholes. That’s what a “fabulous suspension” is designed to do. “Smooth ride” is absolutely diametrically opposed to a “fabulous suspension”.

 

Ah, so a fabulous suspension is intrinsically incapable of providing a smooth ride ... ?

 

BTW, pleased to hear you're still happy with recent hand held scope unit - one should be arriving here, very shortly.

[sandyk]

4 minutes ago, jabbr said:

In any case you don’t call that EMI because fiber doesn’t carry that.

I said RF/EMI out of force of habit. It should then also pass RF from SSDs etc. that may be riding along with the signal.

[jabbr]

2 minutes ago, sandyk said:

I said RF/EMI out of force of habit. It should then also pass RF from SSDs etc. that may be riding along with the signal.

It doesn’t. 
 

If you read up on “stressed eye pattern” you can understand 

[sandyk]

7 minutes ago, jabbr said:

It doesn’t. 

 How does it reject it if it's part of the signal?

[Audiophile Neuroscience]

so what you may gain in galvanic noise EMI isolation you may lose in UHF RF noise?

[fas42]

The only measuring that's worthwhile is the level of rejection by the audio system, as a complete unit, over the full range of possible frequencies, of outside activity - internally generated noise is a bugger that gets resolved by improving the engineering; measuring it is pointless, when the real goal is to get rid of the muck being there.

[jabbr]

3 minutes ago, Audiophile Neuroscience said:

so what you may gain in galvanic noise EMI isolation you may lose in UHF RF noise?

No. This is a possibility which could be true save for the actual details of the specification : read about “stressed eye pattern”

 

Consider that Ethernet is now up to 800 Gbe ... hmm back to the car analogy, as the race car becomes faster, little bumps on the road become more significant. There is literally no physical room to carry out of band information in a 100Gbe signal let alone a 800 Gbe signal. If noise were carried along, the system literally wouldn’t work, the Internet as we know it couldn’t  work. Yet here we are.

[Audiophile Neuroscience]

1 minute ago, jabbr said:

No. This is a possibility which could be true save for the actual details of the specification : read about “stressed eye pattern”

 

Does the 10Gbe Ethernet specification exclude High RF transmission

 

i will look at "stressed eye pattern" but am more concerned with "stressed ear pattern" 😉...kidding

[Audiophile Neuroscience]

Okay, waaay to premature for me to be making any conclusions but stressed eye test is a receiver jitter tolerance test about not making a mistake.

 

One could argue that leakage currents in USB audio have been addressed by various implementations/ error correction mechanisms.

 

The appeal for fiberoptic (for me) is it is said to be not dependent on error tolerance at all (for EMI). It doesn't have to tolerate or fix anything if its not present. This does not appear to be the case for RF induced jitter?

 

Error tolerance to me suggests possibility of failure. Not necessarily, just possibly, maybe.

[fas42]

Interesting experimenting I'm doing with the new active speakers,driven by boring Toslink from a Samsung DVD player - both units need mains filtering; and latest exercise has shown the that the speakers need lots of it - chaining filtering improves things. But the Samsung can't do without - if left with negligible conditioning the SQ suffers ... which indicates that jitter related behaviour over the optical link does matter.

 

Using a brilliant optical link most likely won't make that go away - so, shield the source component more, to attenuate that noise being generated.

[jabbr]

1 hour ago, sandyk said:

 How does it reject it if it's part of the signal?

It is absolutely NOT part of the signal. The bits are the signal. The system is designed to pass the bits and nothing else. Lots of super smart people have designed it and it works. Not only do they say it works but there is test equipment that verifies it. If it didn’t work we wouldn’t be communicating. Probably the single greatest invention of the 20th century. Hmm ... do you use quantum mechanics every day? You use the Internet every day  

[sandyk]

33 minutes ago, jabbr said:

The system is designed to pass the bits and nothing else

So is USB and wired Ethernet !

Now we are back to the usual A.S.R. style putdowns of Planes falling out of the sky etc. if the system didn't work as designed.

[sandyk]

50 minutes ago, fas42 said:

... which indicates that jitter related behaviour over the optical link does matter.

Using a brilliant optical link most likely won't make that go away - so, shield the source component more, to attenuate that noise being generated.

 +1