2. pc-audio
  3. 1408304552

Asynchronous USB DAC and USB cable quality

I have converted to PC audio about 2 years ago and enjoying the hobby. I recently upgraded my DAC from a Benchmark DAC1 HDR to a DAC2 HGC mainly to download DSD files. I am now using a 'regular' Belden USB printer cable purchased at Office Depot which sounds great. I have performed A/B comparison between Toslink fiber input and USB input using well recorded Red Book files and could not hear any differences, which I kind of expected.

The reasoning behind this was that whatever jitter is generated by the PC/USB/Toslink cable is re-clocked by the Benchmark DAC which uses asynchronous clocking. However, would I gain anything sonically by purchasing an audio grade USB cable like a Furutech GT2-USB cable? I would need about 10 feet and do not want to pay a load of money (if any benefits). The Furutech would cost about $235 for that length.
128x128dasign
dtc's avatar
dtc
1,047 posts
 
Sorry I was a bit harsh. I am constantly surprised how many people never try to bring their treads to closure. You at least did, although in a different thread. I found the discussion helpful, so thank you for that. Good luck with the never ending hunt for better cables.
dasign's avatar
dasign OP
201 posts
 
Dtc,

I have learned a lesson here and thanks for your last thread, which make me feel better now.
kijanki's avatar
kijanki
4,428 posts
 
As I understand it, asynchronous USB uses internal DAC's clock
for the D/A converter. This clock is independent
(asynchronous) to USB data rate. For that reason data has to
be buffered. DAC has to request different number of samples
each frame to keep data buffer between underflow and overflow.
Since D/A converter uses data from the buffer (memory) with
internal clock, then quality of the cable lies only in good
shielding to prevent electrical noise contamination of the
DAC. In synchronous USB D/A converter clock comes from the
computer. Usually D/A converter runs from internal clock that
is synchronized with incoming USB clock by PLL (Phase Lock
Loop) but it is less than perfect. This scheme is used in
pretty much every CDP, except here incoming USB clock can be
very jittery.
audioengr
3,282 posts
 
Kij - Async USB is quite immune to jitter from the USB signals, however it is not immune to jitter caused from noise on the ground or the power wires in the cable.

This is why USB cable common-mode filters like the Short-Block are so effective:
http://www.empiricalaudio.com/products/short-block

I also now have a USB filter and power supply combo called the Power-Block. If your USB interface needs the 5V power, such as the Berkeley converter and others that have galvanic isolation, the Power Block can make a big difference in SQ:
http://www.audiocircle.com/index.php?topic=128620.0

The cable is still important in either the case of the cable-powered or device-powered USB interface.

Steve N.
Empirical Audio
More to discover