The State of High End Audio

[amirm]

Makes no difference to Vladimir. In fact I have a 9 meter run of single ended cable from my preamp to my amp without any concern for loss of signal gain

Signal gain is not the problem. Raising of the noise floor is.

Anytime you have two pieces of equipment, you have a voltage differential between their chassis. This is unavoidable. It will be worse between two different circuits but even on the same circuit it will exist. That difference then causes current to flow between the pieces of equipment over the ground of your interconnect. Again, this is unavoidable as any voltage differential will cause current to flow.

UL safety standards in US allow up to 750 microamps of such current leakage. Let's say for ease of math, that you have half as much leakage or 316 microamps. Microamp is a millionth of an amp. Sounds like a pretty small number, right?

Let's compute the voltage drop across the shield using ohms law: AC induced noise voltage = current * resistance. If your interconnect uses 26 gauge shield then its resistance is about 1 ohm. So our AC noise voltage becomes = 316 uamp * 1 = 316 micro volt. Still a small number?

Let's continue the math camp. Consumer line output is at -10 dBV or 316 millivolt. We can now compute our signal to noise ratio = 20 * log (316 microvolt / 316 millivolt) = 60 db!

Yes, such a system has as signal to noise ratio worse than a cassette tape! It is nowhere near the S/N of CD that is 96 dB.

If you upgrade the interconnect to have a much lower resistance shield, you can improve this number up to 86 dB (Belden 8241). Still shy of CD.

[All of this math comes from Bill Whitlock]

Unbalanced interconnects came about in the days of analog where such signal to noise ratios were well above what the analog gear could do and it was assumed that the lengths would be much shorter, just going from one piece in a rack to another. Using them with high signal to noise ratio sources we have today and at long lengths completely invalidates these assumptions and makes unbalanced interconnects the bottleneck in the system.

Note that everything I said is based on simple and undeniable way electronics/physics operate. It is not subject to debate. If you insist on using unbalanced interconnects at such long lengths, better measure the shield resistance and leakage current between your gear. Otherwise you are operating blind and could easily have a very noisy system.

True balanced interconnection eliminates all of this because chassis ground between your equipment is NOT the audio signal reference.

 

[Geardaddy]

Technically balanced is the way to go,
https://en.m.wikipedia.org/wiki/Balanced_audio
but it is technically more complicated to design balanced components from input to output.
Keith.

More of these misguided and uninformed cable designers should consult Wikipedia before they venture forth with their foolish designs.....

Balanced makes some sense if you are the average joe "Pro audio" guy with a rat's nest of patch cords, etc and a need for long runs of cabling. Here is a useful guide: http://www.crutchfield.com/S-FOWSwDVu6JY/learn/pro-audio-live-sound-cable-shopping-guide.html

 

[Steve williams]

Take it up with Vladimir.. All I know is my system is free of any noise. He will never use a balanced cable

 

[amirm]

Take it up with Vladimir.. All I know is my system is free of any noise. He will never use a balanced cable

I don't need to. He doesn't know the voltage differential between your outlets, nor the resistance of your shield cable. As I said, none of this up to debate. You can appeal to any authority you want. You can't change any of these parameters.

 

[Purite Audio]

Cartridges are inherently balanced so balanced connection to the phono would also seem to be a good idea.
Keith.

 

[DaveC]

Not necessarily. There are a host of cable manufacturers who don't agree with you (for a variety of reasons). DaveC???

Lol, this little detour might be my fault but I try not to get into old tired debates online. I kinda forgot there are still people who think wire is wire, and some post on this forum.

On balanced vs single ended, 99.9% of the time in home systems unbalanced and unshielded cables work just great with no significant additions to the noise floor. You just have to be a little careful how the cables are routed. Phono level signals are another matter and here shielding will certainly provide a cleaner signal, this is easily measurable... But despite all that many people prefer unshielded cables because the capacitance is lower and shielding tends to dull the sound a bit. IMO, the only time to use balanced cables is in a balanced system and it's best not to mix single ended and balanced equipment. Balanced in the context of a home system makes very little sense to me, you need two of everything, what a waste. In fact, many like the sound of single ended better as it it doesn't cancel the even order harmonic distortion. For larger multichannel HT installations and pro audio balanced makes a lot more sense and that's what it was designed for... systems that need longer runs of cable in an environment that is prone to noise. But in a home system where there is not that much equipment, it's probably all on a rack or organized cleanly, and cables runs are under 3 meters... balanced offers questionable benefit here imo.

 

[Steve williams]

I don't need to. He doesn't know the voltage differential between your outlets, nor the resistance of your shield cable. As I said, none of this up to debate. You can appeal to any authority you want. You can't change any of these parameters.

I'm not debating nor am I changing parameters. I am merely stating what Lamm does. He does not use balanced cables. There is no noise in my system and life is good. I really don't care about the measurements as my room is silent

 

[DaveC]

Signal gain is not the problem. Raising of the noise floor is.

Anytime you have two pieces of equipment, you have a voltage differential between their chassis. This is unavoidable. It will be worse between two different circuits but even on the same circuit it will exist. That difference then causes current to flow between the pieces of equipment over the ground of your interconnect. Again, this is unavoidable as any voltage differential will cause current to flow.

UL safety standards in US allow up to 750 microamps of such current leakage. Let's say for ease of math, that you have half as much leakage or 316 microamps. Microamp is a millionth of an amp. Sounds like a pretty small number, right?

Let's compute the voltage drop across the shield using ohms law: AC induced noise voltage = current * resistance. If your interconnect uses 26 gauge shield then its resistance is about 1 ohm. So our AC noise voltage becomes = 316 uamp * 1 = 316 micro volt. Still a small number?

Let's continue the math camp. Consumer line output is at -10 dBV or 316 millivolt. We can now compute our signal to noise ratio = 20 * log (316 microvolt / 316 millivolt) = 60 db!

Yes, such a system has as signal to noise ratio worse than a cassette tape! It is nowhere near the S/N of CD that is 96 dB.

If you upgrade the interconnect to have a much lower resistance shield, you can improve this number up to 86 dB (Belden 8241). Still shy of CD.

[All of this math comes from Bill Whitlock]

Unbalanced interconnects came about in the days of analog where such signal to noise ratios were well above what the analog gear could do and it was assumed that the lengths would be much shorter, just going from one piece in a rack to another. Using them with high signal to noise ratio sources we have today and at long lengths completely invalidates these assumptions and makes unbalanced interconnects the bottleneck in the system.

Note that everything I said is based on simple and undeniable way electronics/physics operate. It is not subject to debate. If you insist on using unbalanced interconnects at such long lengths, better measure the shield resistance and leakage current between your gear. Otherwise you are operating blind and could easily have a very noisy system.

True balanced interconnection eliminates all of this because chassis ground between your equipment is NOT the audio signal reference.

Your assumptions can certainly be questioned. For example sources today produce 2V, not 316 mV. And I could go on but I don't have time. I will only say that in the context of a home system with short runs of cable, your assumptions are wrong and so is your conclusion.

Also, IF you use cables with low resistance ground connections and connect your equipment to the electrical system properly, your assumptions are very far from reality.

 

[DaveC]

I don't need to. He doesn't know the voltage differential between your outlets, nor the resistance of your shield cable. As I said, none of this up to debate. You can appeal to any authority you want. You can't change any of these parameters.

Right, except your assumptions about these factors are nowhere close to the reality that many people see inside their own homes with their own systems, as I said before.

 

[amirm]

Your assumptions can certainly be questioned. For example sources today produce 2V, not 316 mV. And I could go on but I don't have time. I will only say that in the context of a home system with short runs of cable, your assumptions are wrong and so is your conclusion.

The length of the cable was not an assumption, but part of the equation. Steve's cable length is 9 meters which is 30 feet. That is not short in my book. As to higher output level, you better have the volume at those levels, not just have the output be capable of 2 volts.

Also, IF you use cables with low resistance ground connections and connect your equipment to the electrical system properly, your assumptions are very far from reality.

Again, none of those were assumptions but actual parameters in the equation. In this case, they were all measured in a real system (hence the reference I provided). If you don't know the parameters I spoke of and you repeat here, then you don't know how bad your system is. Advocating it as a good solution then is non-sequitur.

Message here is simple: chassis ground has no business being the reference for your audio signal. We do all these crazy things to improve the fidelity of the system from use of fancy racks, equipment built to such high standard only to allow the simple fact of voltage difference between two pieces of equipment in unbalanced interconnect ruin any hope of improvements from those tactics. Sticking to 50 year old interconnect scheme is a total mistake.

 

[amirm]

I'm not debating nor am I changing parameters. I am merely stating what Lamm does. He does not use balanced cables. There is no noise in my system and life is good. I really don't care about the measurements as my room is silent

If his systems use chassis as audio reference, then that is not the balanced systems we are talking about.

 

[amirm]

BTW, one of the "objective" reasons why interconnects may make an audible difference is what I just explained. Due to broken architecture of unbalanced systems, and the fact that no two homes are the same, and no two cables have the same impedance, and each equipment has different leakage current to chassis, it means there may very well be audible difference that a) vary from person to person's experience and b) cannot be verified in AB testing. Power cables may also enter the equation here. And outlets. And how you wired your house....

As I think Keith and others have said, once you go to true balanced system, then a lot of these variables disappear.

 

[esldude]

Amirm just read Tim's signature.

I guess these guys know you can have true noise cancelling balanced connection between two fully single ended pieces of gear.

With all the far out approaches some of the gear uses to think balanced connection is a bridge too far.... pretty funny that.

Just read Tim's signature.

The ridicule is nothing personal just toward the ideas expressed.

 

[DaveC]

The length of the cable was not an assumption, but part of the equation. Steve's cable length is 9 meters which is 30 feet. That is not short in my book. As to higher output level, you better have the volume at those levels, not just have the output be capable of 2 volts.


Again, none of those were assumptions but actual parameters in the equation. In this case, they were all measured in a real system (hence the reference I provided). If you don't know the parameters I spoke of and you repeat here, then you don't know how bad your system is. Advocating it as a good solution then is non-sequitur.

Message here is simple: chassis ground has no business being the reference for your audio signal. We do all these crazy things to improve the fidelity of the system from use of fancy racks, equipment built to such high standard only to allow the simple fact of voltage difference between two pieces of equipment in unbalanced interconnect ruin any hope of improvements from those tactics. Sticking to 50 year old interconnect scheme is a total mistake.

No, you took the max allowed leakage current and decided half that was about right. Well, ime it is not right. Not if you have your equipment connected to the same power distribution system that ties all it's grounds together, and not if you're using cables with reasonably sized ground connections (edit: for the noise calculations). Then maybe you go one step further and connect all your chassis together with heavy gauge cable and a copper bar, or one of those fancy grounding systems? And a 2V source. Now you're looking at conditions where the difference between balanced and single ended just doesn't matter anymore. Sorry, but you have to make some unrealistic assumptions to prove your point, and while your math may be correct your assumptions about the level of current likely to be seen and as a problem wrt SCIN are not anywhere close to correct, and further, does not match up with the reality of what people actually tend to experience in real life.

 

[JackD201]

So Jack, are you trying to say that this is the state of high end audio? *elbow into your ribs*

Tom

LOL! Ironically yes. In my baby food fed brain (wink wink "dude"), high prices are symptomatic of both extremes. There are not enough people that care and those that care perhaps too much. All this arguing when fact of the matter is, there's something for everybody.

 

[DaveC]

Amirm just read Tim's signature.

I guess these guys know you can have true noise cancelling balanced connection between two fully single ended pieces of gear.

With all the far out approaches some of the gear uses to think balanced connection is a bridge too far.... pretty funny that.

Just read Tim's signature.

The ridicule is nothing personal just toward the ideas expressed.

Whatever. It's ridicule and has no place here imo.

Your "facts" are assumptions and aren't correct. Sorry. Just because you think they are facts does not make it so.

Your post and Tim's sig are exactly why I didn't respond to this. If that's how you're going to behave then I'm done. Maybe take a clue from Amir about what's appropriate to say, even if he is wrong at least he doesn't denigrate the people he's talking to like that. Your BS and Tim's sig is unacceptable imo.

And your last sentence is exactly like when someone says "No Offense", then goes on to offend thinking it's ok. Well, sorry. You're offensive. If you want to contribute how about making a reasonable argument, or is that beyond your technical and social abilities?

 

[JackD201]

Amir

AES48-2005

Chassis for pin 1 (case to chassis optional) is the officially recommended method.

 

[amirm]

No, you took the max allowed leakage current and decided half that was about right.

That was actually a measured value by Bill.

Well, ime it is not right. Not if you have your equipment connected to the same power distribution system that ties all it's grounds together, and not if you're using cables with reasonably sized ground connections (edit: for the noise calculations).

Well, share some data with us please.

Now you're looking at conditions where the difference between balanced and single ended just doesn't matter anymore.

I am responding to Steve's case of a 30 foot cable. Do you have measurements to show that is not a problem?

Sorry, but you have to make some unrealistic assumptions to prove your point, and while your math may be correct your assumptions about the level of current likely to be seen and as a problem wrt SCIN are not anywhere close to correct, and further, does not match up with the reality of what people actually tend to experience in real life.

I have not seen any real-life measurements other than the one I shared.

Can you optimize the situation? Of course. Now folks know what goes into the equation. But who has done this before? Nobody. If you want to close your eyes and get good performance, then better use balanced interconnects. The moment you use this antiquated scheme known as unbalanced, then you better learn a whole host of parameters before even dreaming about matching CD's performance let alone high-res music.

 

[DaveC]

Also, it's not like I'm not putting my money where my mouth is and sticking my neck out by making these claims. Here's Andre Marc's review that says my products gave him the best resolution he's experienced and lowered the noise floor of his systems:

http://www.avrev.com/home-theater-a...cables-and-surgex-zenwave-edition-review.html

Yes, it's anecdotal and there are no measurements but tons of people have had his exact experience and they all describe it the same way. Never had an explanation for that by someone who thinks cables can't make a difference.

Another point, the truth is the ground (pin 1) of an XLR cable can make a large audible difference. Ground is still a voltage reference in the component and XLR doesn't make the system immune to grounding issues. It's simply a way to reject common mode noise, which isn't an issue in 99.9+% of home audio systems.

 

[DaveC]

That was actually a measured value by Bill.


Well, share some data with us please.


I am responding to Steve's case of a 30 foot cable. Do you have measurements to show that is not a problem?


I have not seen any real-life measurements other than the one I shared.

Can you optimize the situation? Of course. Now folks know what goes into the equation. But who has done this before? Nobody. If you want to close your eyes and get good performance, then better use balanced interconnects. The moment you use this antiquated scheme known as unbalanced, then you better learn a whole host of parameters before even dreaming about matching CD's performance let alone high-res music.

Steve's case is not usual and I agree a long run of single ended cable is much more likely to cause issues, but that doesn't mean it will. If the ground potential differences on each side of the 9 meter cable are minimal and there are no EM fields the cable will couple with on the way then it'll be fine...

As I said though, you can't just close your eyes and get good results with balanced... it's subject to the many of the same issues as single ended equipment wrt grounding issues. The potential differences between grounds should still be minimized for best results. Sure, you don't need to worry much about interference and CMMR is great... but is this really worth it? Is the complexity of having to have two of everything... so 4 amplification stages instead of two for the entire system... and cables with two runs of signal wire instead of one... this is almost like doubling the parts count of your system. For performance advantages that are seldom a true advantage in home systems.

As far as measurements I don't have any to share, as for this kind of stuff it's very easy to just listen to the system... I've seen the equations and done the math so it's not like the results are going to be a surprise.