Why Do Different USB Cables Sound Different

Hi Steve,

I have a friend who has made his own USB cable. The design is novel in only that he has addressed the "N" part of the equation with the design. He states that the change has been dramatic in overall SQ. Just like any digital cable there are proven principles in place. I 'll attach a review of the digital coaxial cable I use from DAC to transport....I think you will find the article interesting.

http://www.6moons.com/audioreviews/nirvana/digital.html

Steve Williams said:

I have seen threads here on different ethernet cables sound different and now we read here that USB cables sound different. How is this possible when all that is being transmitted are 0's and 1's.

I am far from a skeptic because my listening experiences do suggest there is an audible difference


I ask this question because now along comes the Master Built Ultra Ultra USB cable which I will be auditioning in the coming weeks. From everything that I have been reading this MB USB cable sonically takes things to a totally different level that is leaving listeners speechlesss. I am looking forward to the audition and sharing my experiences here. My overall experience with my MB Ultra loom has taken my sound system to a totally different level so I am more than curious about their version of a USB cable even though it has an MSRP in the ozone

I understand that the MB engineers have done something which isolates the data but this apparently is nothing new

So my question is whether others have been hearing improved sonics with different makes of USB cables and if so what differences are you noting

Click to expand...

The Digital Link has always played a very important part since the early days. It's just as important as your phono cables, 0s & 1s seem to be just as picky!

david

I too have heard some differences, not always for the better. I think the MB is a bit too rich for my wallet but if they want to send me a loaner, I will surely report my findings.

Steve Williams said:

I have seen threads here on different ethernet cables sound different and now we read here that USB cables sound different. How is this possible when all that is being transmitted are 0's and 1's.

I am far from a skeptic because my listening experiences do suggest there is an audible difference


I ask this question because now along comes the Master Built Ultra Ultra USB cable which I will be auditioning in the coming weeks. From everything that I have been reading this MB USB cable sonically takes things to a totally different level that is leaving listeners speechlesss. I am looking forward to the audition and sharing my experiences here. My overall experience with my MB Ultra loom has taken my sound system to a totally different level so I am more than curious about their version of a USB cable even though it has an MSRP in the ozone

I understand that the MB engineers have done something which isolates the data but this apparently is nothing new

So my question is whether others have been hearing improved sonics with different makes of USB cables and if so what differences are you noting

Click to expand...

The issue isn't the ones & zeroes - they are delivered as expected. What seems to be different between cables is the signal waveform of the digital signal arriving at the receiver.
There are a number of ways this signal waveform can be different but the USB standard specifies a range of acceptable distortion in this waveform - basically the USB standard defines a 'safe' level of distortion in the signal waveform that will avoid bit errors i.e. the ones & zeroes will be correctly received by the receiver.

Now a USB cable doesn't just carry the USB signal on two wires within the cable - it also has 2 additional wires in the cable, ground & 5V power & finally has a shield. These two wires & shield further complicate matters when it comes to signal integrity. All can carry noise between the PC & the USB receiving device - noise is just another signal waveform distortion as is jitter - neither of these issues affect the delivery of the digital bits (once within the acceptable range of the USB standard).

In my experience when there are changes in these factors it can often have an audible effect. Cables can change the noise spectra on the cable, on the cable shield, on the signal wires, on the ground & power wires - all/some of which can have audible effects.

I haven't listened to expensive USB cables but have used a cable which is made to suppress EMI noise - http://kgs-ind.com/wp-content/uploads/products/pdf/LNC-Noise-Suppression-USB20-Cable.pdf
This cable is not an audiophile offering - it is made for use in airlines, I believe & consists of a flexible ferrite compound fully covering surrounding the 4 USB wires. For those addicted to measurements, look at the noise measurements on the linked page - it shows the reduction in noise in the frequency range 50MHz to 1,000MHz. Now how could reduction in noise in these frequencies have audible effects in the audible range? Objectivists need to be inquisitive enough to investigate but they don't have the necessary motivated or skills, it would appear

This was noticeably quieter than my other USB cables but had the side effect of somewhat audibly softening high frequencies. This is not that surprising, as those who have tried ferrous rigs on USB will no doubt attest to.

RogerD said:

(...) I 'll attach a review of the digital coaxial cable I use from DAC to transport....I think you will find the article interesting.

http://www.6moons.com/audioreviews/nirvana/digital.html

Click to expand...

What is the technical interesting part of the article?

BTW, what is the N part of the equation?

microstrip said:

What is the technical interesting part of the article?

BTW, what is the N part of the equation?

Click to expand...

There is no technical explanation in the article,just comments on the construction that I thought Steve might find interesting.
The N stands for noise.

jkeny said:

Now a USB cable doesn't just carry the USB signal on two wires within the cable - it also has 2 additional wires in the cable, ground & 5V power & finally has a shield. These two wires & shield further complicate matters when it comes to signal integrity. All can carry noise between the PC & the USB receiving device - noise is just another signal waveform distortion as is jitter - neither of these issues affect the delivery of the digital bits (once within the acceptable range of the USB standard).
.

Click to expand...

The main issue is that.
5 V. usually supplied by computer's switching Power Supply which are well known to create Spikes.

caliaripaolo said:

The main issue is that.
5 V. usually supplied by computer's switching Power Supply which are well known to create Spikes.

Click to expand...

That's what a lot of people think but it's not the case, based on my testing - unless you have very bad ground loop noise, the main route for noise intrusion is actually via the USB signal wires D+ & D-

I don't use the 5V USB power as I consider this a basic flaw to use this for powering any audio device - not only is this power of undetermined quality (depending on the PC it originates from) but it is also being carried in a wire alongside the two USB signal wires as are the return currents being carried on the ground wire in the USB cable. A disastrous design but easy to avoid - don't use USB audio devices that use USB power to drive them.

The next best approach is to use a USB cable that has the power wire separated from the signal wires but this still leaves the ground wire with it's PS return currents next to the signal wires. Another variable As I said I have tested lifting ground & in my setup (which is ground loop free) it made no audible difference but isolating the D+ & D- USB signals.

Another approach is to inject a 'clean' 5V supply instead of the USB 5V - still not optimal as there are varying current draws on this PS wire which can affect the USB signal wires if it is running next to them in the cable.

Devices which use the 5V not for power but only for sensing when a USB device is connected are somewhat better

To my mind the only real solution is avoid any use of USB 5V power - this means the ground wire is no longer returning cur.rents from the 5V supply - it is the return path for the USB signal currents.

As I said, when all these PS issues are dealt with, the noise on the USB signal wires is still evident - isolate these wires & you will hear an improvement (you are likely removing what is called common mode noise) -regenerate the signal after this isolation (as it adds jitter to the waveform) & you will hear an improvement.

USB cables can only change the sound (not always improve it) to a certain extent by changing the noise profile on the ground & shield which ultimately intrudes the signal wires.

jkeny said:

That's what a lot of people think but it's not the case, based on my testing - unless you have very bad ground loop noise, the main route for noise intrusion is actually via the USB signal wires D+ & D-

Click to expand...

It's not just a people's thought, is the limit of computer based system.
I may believe on your test, but after I've replaced switching PS of my MacMini (with external PS-2 from Uptone), then selected an USB cable with separate 5V. wires and with an high isolation of D+ and D- wires (Curious USB cable) and finally by connecting an Entreq groundbox via USB cable (Ground wire pin only) connected to a spare USB port of MacMini, I can tell you which the improvement has been great.

caliaripaolo said:

It's not just a people's thought, is the limit of computer based system.

Click to expand...

Huh?

I may believe on your test, but after I've replaced switching PS of my MacMini (with external PS-2 from Uptone), then selected an USB cable with separate 5V. wires and with an high isolation of D+ and D- wires (Curious USB cable) and finally by connecting an Entreq groundbox via USB cable (Ground wire pin only) connected to a spare USB port of MacMini, I can tell you which the improvement has been great.

Click to expand...

Sure, I'm not doubting that all these changes made an improvement in sound -they are pretty much what I outlined in my post. If your USB audio device is using the 5V USB power for connection sensing or powering then the 5V power is probably the most crucial area but I also said that any USB audio device which uses USB 5V power is, IMO, flawed to start with. What USB audio device is connected to your MacMini?

BTW, replacing the SMPS of the MacMini (a notoriously bad designed PS for leakage currents) is not just providing a 'cleaner' 5V it's also reducing the leakage currents on the ground wire & signal wires of the USB cables. Ask John Swenson about this aspect of the effects of replacing the MacMini's PS.

What I said above is that if we avoid the flawed design of USB audio devices that use the USB 5V then the main source of noise travels along the USB signal wires.

Now that you have minimised the effect of 5V & ground noise with your approach, have you tried a USB isolator?

jkeny said:

The issue isn't the ones & zeroes - they are delivered as expected. What seems to be different between cables is the signal waveform of the digital signal arriving at the receiver.

Click to expand...

Is this true of all digital links John, spdif, res/ebu, toslink, glass, etc. because I hear very clear differences between the wires and connectors?

jkeny said:

USB cables can only change the sound (not always improve it) to a certain extent by changing the noise profile on the ground & shield which ultimately intrudes the signal wires.

Click to expand...

IME digital links are extremely important and can really destroy the sound of even the best digital combinations, they also define the character of one's digital front end more than analog wires and sadly DACs are just as susceptible to audiophile power cord distortions as any other component.

david

jkeny said:

What USB audio device is connected to your MacMini?

Click to expand...

Totaldac reclocker.

jkeny said:

Now that you have minimised the effect of 5V & ground noise with your approach, have you tried a USB isolator?

Click to expand...

Yes, I used Berkeley Alpha USB and now I use Totaldac Reclocker.
However, the biggest improvement has been achieved by solving the Power issues.
Cheers,
Paolo

ddk said:

Is this true of all digital links John, spdif, res/ebu, toslink, glass, etc. because I hear very clear differences between the wires and connectors?

Click to expand...

Well, yes & no - USB is particularly prone as it has a PC at the transmission end of the cable & this comes with all sorts of noise issues. The other factor is that USB cables are a particular bad design with their signal wires alongside power & ground wires all wrapped in a shield. There are many areas of noise interaction in such a configuration - noise from power wire, ground wire, shield, - interactions of signal wire currents & other wires in the cable.

The other digital links suffer from their own issues & need specific analysis as to their weaknesses. SPDIF, for instance has different weaknesses - the main one being that is relies on extracting the clock from the signal timing, Toslink, the same but with the added distortion of optical to digital conversion, etc.

Lest anyone think that I'm down on USB - it can be the best sounding of all the digital links once it's weaknesses are recognised & address - it has the major advantage of having a local clock timing which means lowest jitter

IME digital links are extremely important and can really destroy the sound of even the best digital combinations, they also define the character of one's digital front end more than analog wires and sadly DACs are just as susceptible to audiophile power cord distortions as any other component.

david

Click to expand...

Sure, people cite DACs for not being immune to all the upstream issues & also for not being immune to PS quality & we often hear the over-used phrase from those who have never designed an audio device (not to mention digital audio device), "properly designed DACs". The exposure of this posturing is revealed by asking such people to nominate such DACs & they go silent.

The reality is that this immunity is a very difficult thing to achieve as it is multi-dimensional & when noise/distortion is intertwined with signal waveforms, it's very difficult to remove it & leave the signal waveform untouched. However, each individual aspect that can audibly affect digital audio is being addressed & gradually worked through.

caliaripaolo said:

Totaldac reclocker.



Yes, I used Berkeley Alpha USB and now I use Totaldac Reclocker.
However, the biggest improvement has been achieved by solving the Power issues.
Cheers,
Paolo

Click to expand...

Neither of these devices truly isolate the USB signals. The Berkeley uses the same flawed approach as many claiming "galvanic isolation" - they isolate after the USB signal has been received by the USB receiver chip - too late - the isolation needs to happen on the D+ & D- signal not on the I2S signal out of the USB receiver which is what the Berkeley does. i have tried this approach & it is less effective than true isolation on the D+ & D- signals

The Totaldac recklocker doesn't seem to isolate the USB signals, either - it converts everything to DSD

So no, I don't think you have tried a true USB isolator from what I can see.

Thanks John!

What about Firewire, same issues as USB?

david

jkeny said:

Well, yes & no - USB is particularly prone as it has a PC at the transmission end of the cable & this comes with all sorts of noise issues. The other factor is that USB cables are a particular bad design with their signal wires alongside power & ground wires all wrapped in a shield. There are many areas of noise interaction in such a configuration - noise from power wire, ground wire, shield, - interactions of signal wire currents & other wires in the cable.

The other digital links suffer from their own issues & need specific analysis as to their weaknesses. SPDIF, for instance has different weaknesses - the main one being that is relies on extracting the clock from the signal timing, Toslink, the same but with the added distortion of optical to digital conversion, etc.

Lest anyone think that I'm down on USB - it can be the best sounding of all the digital links once it's weaknesses are recognised & address - it has the major advantage of having a local clock timing which means lowest jitter

Sure, people cite DACs for not being immune to all the upstream issues & also for not being immune to PS quality & we often hear the over-used phrase from those who have never designed an audio device (not to mention digital audio device), "properly designed DACs". The exposure of this posturing is revealed by asking such people to nominate such DACs & they go silent.

The reality is that this immunity is a very difficult thing to achieve as it is multi-dimensional & when noise/distortion is intertwined with signal waveforms, it's very difficult to remove it & leave the signal waveform untouched. However, each individual aspect that can audibly affect digital audio is being addressed & gradually worked through.

Click to expand...

ddk said:

Thanks John!

What about Firewire, same issues as USB?

david

Click to expand...

I've no experience with FW & haven't looked into it so can't comment.

jkeny said:

The Totaldac recklocker doesn't seem to isolate the USB signals, either - it converts everything to DSD

Click to expand...

Yes correct the Reclocker doesn't act as an USB isolator. It is a USB to SPDI converter which use to store data into a buffer.
Beside, the Reclocker doesn't convert everything to DSD!!!
I tested an Intona usb isolator (Which had been lent to me), but I didn't liked the result achieved. Too veiled.

caliaripaolo said:

It's not just a people's thought, is the limit of computer based system.
I may believe on your test, but after I've replaced switching PS of my MacMini (with external PS-2 from Uptone), then selected an USB cable with separate 5V. wires and with an high isolation of D+ and D- wires (Curious USB cable) and finally by connecting an Entreq groundbox via USB cable (Ground wire pin only) connected to a spare USB port of MacMini, I can tell you which the improvement has been great.

Click to expand...

I use a Berkeley Alpha USB as well and have tried about 6 USB cables. I agree the ones with the separate 5v wires were CLEARLY superior to the others.