Audio-gd Digital Interface
May 21, 2012 at 1:54 PM Post #2,866 of 4,156

JulienM

100+ Head-Fier
Joined
Aug 9, 2010
Posts
114
Likes
12
Quote:
Also important not to get too caught up by cable voodoo!
(Speaks the guy who impulsively bought Wireworld 2M USB cable)
JulienM, where did you get the usb plugs and the canare cable? I could only find the usb plug on DoubleHelix but shipping is $40!
 

 
I found the USB connectors here:
http://www.pimfg.com/
 
And the Canare cable from here:
http://www.markertek.com/
 
USB cable comparison coming very soon! No cable voodoo or audiophile gimmicks, just very affordable pro stuff!
 
May 21, 2012 at 5:13 PM Post #2,868 of 4,156

JulienM

100+ Head-Fier
Joined
Aug 9, 2010
Posts
114
Likes
12
This morning I was able to give a serious audition to all the USB cables I've being building for the past few weeks. And also evaluate the effects of the Olimex isolator in my main system.
 
For your reference, the system consists of an Apple MBP 15" (latest model - also used for work) -> DI (with PS) -> NFB-8 DAC (with TCXO) -> C-3 (Final Version) preamp -> C-10 amp -> Thor speakers (fat cabinet, 100 liters).
Interconnects are all cast using Canare DA202 cable.
The system is plugged into a Agd power filter through DIY power cords (unshielded braided design).
 
The computer USB output is set for 96KHz and 24 bit integer. The DI is set to pass through, and not over sample. Even with a Tent Labs clock, I found this arrangement to give a more detailed and spacious sonic presentation.
 
I used 3 test tracks for evaluation. 
Julieta Venegas and Marisa Monte - Ilusión (Volume set to 27)
http://www.amazon.com/Julieta-Venegas-Mtv-Unplugged/dp/B0018OAPBQ/ref=sr_1_4?s=music&ie=UTF8&qid=1337630851&sr=1-4  
Beethoven Symphony No. 9, 4th and 5th movements - Karajan + Berlin Philharmonic Orchestra (Volume set to 42)
http://www.amazon.com/Beethoven-Symphony-No-%7E-Karajan/dp/B000001GJR/ref=sr_1_13?s=music&ie=UTF8&qid=1337630952&sr=1-13

And, once in a while, I would listen to various arias from Handel's Giulio Cesare just to check my perception, but I didn't use it for all cables.
http://www.amazon.com/Handel-Giulio-Cesare-George-Frideric/dp/B000094HLD/ref=sr_1_3?s=music&ie=UTF8&qid=1337631152&sr=1-3
All observations without the Olimex isolator.
 
Monoprice USB
Natural presentation with a good midrange. A bit dry and lacking finesse. Not as fluid as the Canare cables. A little congested on top. Capable of resolving a great deal of information but not as involving or dynamic. A bit boring. Competent cable, an excellent value!
Presence: 7.5
Sound stage: 8
Choral goose bump factor: 7
 
Canare DA202 (Blue)
Very detailed and lush. Excellent definition. Brighter, more extended, and slightly faster than the other Canare cables. A little analytical on the top end, but with more air.
Presence: 8 Sound stage: 9
Choral goose bump factor: 9 (quite intense)
 
Canare L-2T2S (Orange)
More body than the DA202 but not as lush. Great micro detail and dynamic. Slightly dryer but more natural. Great texture and depth on strings. Cellos are very seductive and capable of goose bumps. This is a great cable to use as a USB connector. Very impressive!
Building this cable is not so easy though, things get very tight at the USB terminations...
Presence: 9
Sound stage: 9
Choral goose bump factor: 9.5 (amazing!)
 
Canare L-2E5 (Black)
Not as involving as the L-2T2S, but very close. I'd say 95% as good as the L-2T2S. Very natural with a big mid range full of detail. Well balanced and natural presentation. 
Much easier to terminate than the L-2T2S due to the smaller gauge of the conductors.
Presence: 8.5
Sound stage: 8.5
Choral goose bump factor: 8.5
 
 
Canare L-2T2S with Olimex isolator
In my main system, somehow not as transparent or exciting! Goose bump factor went down. Robbed micro detail and slowed the system down. The sound is more "round", helping lesser recordings. 
It does work very well on my (more modest) family A/V system which consist of a mac mini feeding an Agd Fun using a chip amp to drive Seas coaxial speakers. A very musical system. With the Olimex in place, the system acquired more presence with a bigger midrange. I'm keeping the Olimex there for now!
 
 
My first choice being the Canare L-2T2S! I'll have to build another for my family system... One could experiment mixing different cables for data and power sections. I might try different arrangements in the future, but I feel that the data section is the more significant in this scenario since the DI is being powered by its own power supply.
 
As a quick disclaimer, these are my subjective opinions, experimentation is key. Be very carefully if you're building your USB cables. Make sure you test the cable for continuity and short circuits. It's easy to make a mistake and damage a component.


 
 
 
The USB cables:

 
The system:

 
May 22, 2012 at 2:12 AM Post #2,869 of 4,156

DarknightDK

500+ Head-Fier
Joined
Sep 25, 2011
Posts
665
Likes
27
Nice! thanks for posting your impressions.
 
May 22, 2012 at 3:07 AM Post #2,870 of 4,156

drez

Headphoneus Supremus
Joined
Sep 7, 2009
Posts
2,921
Likes
182
Quote:
 
I found the USB connectors here:
http://www.pimfg.com/
 
And the Canare cable from here:
http://www.markertek.com/
 
USB cable comparison coming very soon! No cable voodoo or audiophile gimmicks, just very affordable pro stuff!

 
frown.gif
  Why did I not find this seller earlier - I ordered my connectors for $3 each and paid $40 for postage.  I really should have shopped around more.  Ahh well I will know better if I ever have to order more connectors.  if I ever build another full length USB cable I will have to test the L - 2T2S - indeed I may have to try it out to compare against my Furutech GT2 which sells for around $125 just out of curiousity
biggrin.gif

 
For the latest USB cable I made I just took the conductors out of an audiophile cable thought sounded rubbish and used about 7cm to make a short cable, and what would you know the same conductors sound brilliant when they are 7cm long.  
 
Do I remember correctly that the L - 2T2S is the one with 88 Ohms impedance?  It looks like a very well made cable!  How did you connect the braided shields (one end or both)?  I think if I do build a shielded cable I will use twisted pair as then there is less inductance between the conductors and the shield due to the geometry.  I think shielding around parallel pair as Neo D+ USB cable uses must just have to high inductance and capacitance or something.  Either way the geometry of that Canare L - 2T2S is certainly doing something right.
 
May 22, 2012 at 5:05 AM Post #2,871 of 4,156

LosNir

100+ Head-Fier
Joined
Aug 20, 2008
Posts
158
Likes
10
Thanks JulienM, nice and short comparison! Just one thing I did not understand, you opted out of the Olimex in your main rig? In your family rig you are using Olimex + L-2T2S?

drez, you're pretty much lucky! DoubleHelix is my only option for me, as none of the other sellers ship outside of the U.S.
This looks fun and interesting. Maybe someone wants to help and sell me some connectors and wires? :)
 
May 22, 2012 at 7:52 AM Post #2,872 of 4,156

JulienM

100+ Head-Fier
Joined
Aug 9, 2010
Posts
114
Likes
12
Thanks JulienM, nice and short comparison! Just one thing I did not understand, you opted out of the Olimex in your main rig? In your family rig you are using Olimex + L-2T2S?
drez, you're pretty much lucky! DoubleHelix is my only option for me, as none of the other sellers ship outside of the U.S.
This looks fun and interesting. Maybe someone wants to help and sell me some connectors and wires? :)


If you're interested, I should be able to build you a USB cable using the L-2T2S. Send me a private message and we can discuss.
 
May 22, 2012 at 8:04 AM Post #2,873 of 4,156

JulienM

100+ Head-Fier
Joined
Aug 9, 2010
Posts
114
Likes
12
:frowning2:   Why did I not find this seller earlier - I ordered my connectors for $3 each and paid $40 for postage.  I really should have shopped around more.  Ahh well I will know better if I ever have to order more connectors.  if I ever build another full length USB cable I will have to test the L - 2T2S - indeed I may have to try it out to compare against my Furutech GT2 which sells for around $125 just out of curiousity:D

For the latest USB cable I made I just took the conductors out of an audiophile cable thought sounded rubbish and used about 7cm to make a short cable, and what would you know the same conductors sound brilliant when they are 7cm long.  

Do I remember correctly that the L - 2T2S is the one with 88 Ohms impedance?  It looks like a very well made cable!  How did you connect the braided shields (one end or both)?  I think if I do build a shielded cable I will use twisted pair as then there is less inductance between the conductors and the shield due to the geometry.  I think shielding around parallel pair as Neo D+ USB cable uses must just have to high inductance and capacitance or something.  Either way the geometry of that Canare L - 2T2S is certainly doing something right.


Yes, it's the 88 ohms cable, which is within USB 2.0 specification. It is a twisted pair with a braided shield (all 3 Canare are). The specification asks for twisted pairs for the data pair.

I connect the shield at both ends. The connectors have to be adapted a bit for all that shield...

I'll take pictures of the building process when I build my next set.
 
May 22, 2012 at 11:00 AM Post #2,876 of 4,156

kr0gg

100+ Head-Fier
Joined
Mar 18, 2009
Posts
439
Likes
17
nobody's convincing anyone in anything here.
you don't believe our experience? no problem.
try for yourself and see if your experiment will give you any difference.
yes, it might not give any difference at all. but even then you would have no reason to convince people that they cannot possibly hear it.
even if it is only in our minds - i don't see any problem with that.
we enjoy it and as far as i know we don't hurt anyone with our enjoyment.
 
on second thought, actually the only one here that is trying to convince anyone is you.
 
May 22, 2012 at 11:13 AM Post #2,877 of 4,156

drez

Headphoneus Supremus
Joined
Sep 7, 2009
Posts
2,921
Likes
182
Quote:
I hate to be 'that guy', but its hard enough convincing people that analog signal cables make a difference, now we're saying that digital cables make a difference? So the 10101101 signal in one cable sounds better then it does in the other?

 
Yes that's why USB has a minimum specification for data cables to try to ensure that the digital signal is of a good enough standard.  In fact PCB design also has to take into account similar consideration of noise and jitter when products are designed.  Its the same reason we have asynchronous USB devices, SPDIF receivers with PLL's etc because digital signals are an analog waveform which is treated conceptually as 1 and 0.  The reason cables matter is that USB and digital input receivers are not perfect, and ferrite chokes and galvanic isolation are not perfect either.  A good USB cable will make it easier for the USB receiver to do it's job by presenting as square a digital waveform as possible with as little latency or delay as possible.  This is just my opinion of why USB cables might make a difference.
 
There are quite a few threads on this in the sound science form but I would suggest not bothering to read into it too much especially when you can build an USB cable for around 5 dollars and test it out for yourself.  If you are wondering the consensus from the objective side is that USB cables should not make a difference, but I would take that with a grain of salt.  It is much more sensible to test for your self, controlled or not depending if you trust yourself.
 
May 22, 2012 at 9:11 PM Post #2,880 of 4,156

LosNir

100+ Head-Fier
Joined
Aug 20, 2008
Posts
158
Likes
10
Found some Canare L-2T2S @ Redco Audio
http://www.redco.com/shopexd.asp?id=539
 

Users who are viewing this thread

Top