USB cables

[RBFC]

Aren't the "split" cables made by running the power leg down a separate insulated conductor path, and only connected at both ends?

Lee

 

[Bruce B]

Aren't the "split" cables made by running the power leg down a separate insulated conductor path, and only connected at both ends?

Lee

I think that's the way it was explained to me.


Welcome Gavin!

 

[CKKeung]

Aren't the "split" cables made by running the power leg down a separate insulated conductor path, and only connected at both ends?
Lee

It is Acoustic Revive in Japan who discovered the benefit of separating the DC power wires & signal wires of an USB cable.
It was in 2009 and they was given innovation awards by several Japanese hifi magazines in 2010 & 2011.
However they did not go for copyright/patent application so this practice has been adopted by numerous audiophile cable makers all over the world.

Acoustic Revive's USB cables are still one the the best in the market now.
http://www.acoustic-revive.com/english/pcaudio/usb_cable_01.html

 

[gavn8r]

Aren't the "split" cables made by running the power leg down a separate insulated conductor path, and only connected at both ends?

Hello again and thanks for your warm welcome. I think I've properly set up my signature.

Both configurations have separated power and signal legs. Standard LightSpeed is used similarly to most USB cables: one connector on each end. Split Lightspeed has one USB-B connector and two USB-A connectors: one for signal, one for power. This configuration can be used with self-powered DACs, such as Da Vinci DAC.

 

[Julf]

Hi, Gavin,

each LightSpeed cable goes through a very thorough final test and measurement. We include each cable's unique test report in the box.

I would love to hear what parameters you measure.

 

[gavn8r]

LightSpeed USB Cable Test Report

Hi Julf,

Here's a PDF test report for one of the last prototypes before we went into production.

View attachment Test Result EP3-030.pdf

-GF

 

[Julf]

Here's a PDF test report for one of the last prototypes before we went into production.

Thanks! Very informative!

Any comments on how the parameters affect sound quality?

 

[gavn8r]

I'm not sure I quite understand the question, Julf. Can you be a little more specific?

Oh, and Jack - I'm glad you'll be getting yours in a couple weeks. The box is indeed on the dock ready to be picked up.

 

[Julf]

I'm not sure I quite understand the question, Julf. Can you be a little more specific?

Well, the measurements show, at at minimum, that the cable performs as specified. But in your measurements, have you found any correlation between the measured parameters and actual changes in sound, and if so, what effect do the diffeernt parameters have? As an example, you measure the absolute impedance - but would having a 70 ohm impedance instead of 90 change the sound, and how?

 

[Larry Ho]

Well, the measurements show, at at minimum, that the cable performs as specified. But in your measurements, have you found any correlation between the measured parameters and actual changes in sound, and if so, what effect do the diffeernt parameters have? As an example, you measure the absolute impedance - but would having a 70 ohm impedance instead of 90 change the sound, and how?

Hi, Jeff

This is a very good question. And actually maybe need a lot of people, more than us to do more exploration. Some parameter's impact should be easier to find, some should not.

In the test report, we make the regulation especially tight. For example, 90Ohm impedance is the USB cable standard. So for example, if one cable only has 70 characteristic impedance, and it will cause impedance mismatch. So in the beginning and end of this USB cable will generate reflection of digital square waves. It will interfere/cancel/enhance the original signal which will make USB sender IC and
receiver IC work hard to correct it. The correction process at least will create jitter, the worse case, will create error... (e.g. 1 become 0, or 0 become 1)

So in the test report and my point of view, digital related parameters, more accurate to its original specs it should be, the better. For analog related one, it needs a lot of experiments.
Impedance needs to be as close as possible to 90Ohm. When there is a very close to perfect match. I think what Bruce found in sound quality improvement will show up. Timing accuracy, more natural and spectral information correctness.

Hope this helps!

Larry

 

[Larry Ho]

Another tips here... ;-)

If want to get the max separation result, you can connect TWO USB-A of Lightspeed to computer first. After USB done the protocol handshake, even while music is playing, you can "un-plug" the Power LEG.
Music will keep playing there... Just no power loop between computer and your USB DAC...

I would love to know if someone get some listening tests for above. Only little trouble is: Need to plug and unplug each time USB DAC needs to be discovered by computer...

Cheers,

Larry

 

[Julf]

Larry,

So in the test report and my point of view, digital related parameters, more accurate to its original specs it should be, the better.

In principle, yes.

Impedance needs to be as close as possible to 90Ohm. When there is a very close to perfect match. I think what Bruce found in sound quality improvement will show up.

Unfortunately 90 ohm is not exactly right - it is just right on the average. If the USB standard specifies an allowed tolerance of +- 10 %, the computer end of the interface might be
81 ohm or 99 ohm. If it is 81 ohm, then a cable with an impedance of 81 ohm would supposedly sound better than one with 90 ohm. Same goes with the DAC end, of course.

 

[Larry Ho]

Larry,



In principle, yes.



Unfortunately 90 ohm is not exactly right - it is just right on the average. If the USB standard specifies an allowed tolerance of +- 10 %, the computer end of the interface might be
81 ohm or 99 ohm. If it is 81 ohm, then a cable with an impedance of 81 ohm would supposedly sound better than one with 90 ohm. Same goes with the DAC end, of course.

Well... from the beginning to the end of whole chain should be: USB transceiver IC in computer ---> PCB trace route ---> USB-A connector of computer ---> USB-A connector of USB cable ---> USB Cable wires ---> USB-B Connector of cable --> USB-B connector of DAC ---> PCB trace route ---> USB transceiver IC or other processor at DAC side.

Each element of above will impact the whole transfer process, but in my experience, IC usually very tightly in specs. for the best possible compatibility. USB connectors depends on what connectors that computer and USB DAC use, that will be one of the key point need to check. For PCB trace route, every capable CAD software and PCB manufacturer could control within 3% ~ 5%, averagely precisely at 90 Ohm.

The assumption here is people don't use USB Hub in between!

The usual suspect, well, is the long running cable... Especially the long long long one... (Luckily USB standard has 5m limitation there)

So target precisely at 90Ohm I think still the best way to make the USB cable, because we can not control what connectors they use.

And for SPDIF or AES/EBU interface, this impedance match stuff seems cause more troubles.. Luckily, we don't need to worry about that now....

 

[Julf]

So target precisely at 90Ohm I think still the best way to make the USB cable, because we can not control what connectors they use.

Of course. Just pointing out that there is no "absolute right", so now audiophiles can start obsessing about matching their USB cable impedance...

 

[dcc]

I have chosen a sligthly different approach. I have a MacBook Pro with Audirvana Plus. Though my DAC (Audio Aéro La Source) is fitted with a high quality USB input, I am not using a USB cable all the way as I have a digital interface in between.

Basically, there is a USB Wireworld Platinum from the Mac to the digital interface and then a AES/EBU Wireworld Platinum from the interface to the DAC. The interface handles signals up to 24/192 and has a dedicated power supply which may be cut from the mains via a built in battery buffer.

There a clear sonic improvements compared to the sole use of a USB cable. There is of course a cost associated to this option.

 

[Larry Ho]

Sorry to tell you, with AES/EBU or SPDIF, you can go up to 24/192K only like you wrote.

And I'm the believer of asynchronous interface so DAC could be the clock master, not your computer. If you use AES/SPDIF in between the digital chain, you will literally let computer source as your clock master and that should not be good to your sound quality.
With a proper USB interface and cable, you can go highest to 32/768K or 32/384K and DSD256... So why add AES/SPDIF in between?

Well, this is my general talk, according to different system situation, result would be different. But I would suggest you use USB only for a test, maybe this is better!

 

[Shaffer]

Soooo... no one has a recommendation for me, as described in the original post?

 

[JackD201]

Sorry dude. The blue Belkin I got from Best Buy works. Far better than one far east audiophile brand which glitched like crazy.

 

[Bruce B]

Soooo... no one has a recommendation for me, as described in the original post?

Your best bet would be to contact "The Cable Company". They have dozens of USB cables they will loan you at all price points. Use what's best for your system and ears. It's the only way!

 

[Shaffer]

Your best bet would be to contact "The Cable Company". They have dozens of USB cables they will loan you at all price points. Use what's best for your system and ears. It's the only way!

Thank you for the great advice.