how about USB cable?

Which analog IC's are you using between U24 and amp?

margo

Yes, USB cable quality (and perhaps length) can affect sound quality. They are impedance controlled differential transmission lines and will experience the same issues you get on S/PDIF cables. Termination quality, connections, impedance match are all critical.

The USB receivers we use take the 1ms packet rate to regenerate the 44.1k word clock. The cleaner the USB data signals, the cleaner the output clock. Bad edges means jittered reception.

That's why I do a reclocking AFTER all of the USB stuff. In fact, I put my cleanest clock right on the DAC chip, where it makes the most difference.

jh

I have a Waveterminal U24 USB device. When I first started experimenting with it, I had to run a 15 ft USB cable to reach my gear. I initially thought that the sound was rather harsh (bright). I then tried using my laptop and a shorter (3 ft) USB cable and that changed everything - no more glare. This is while using a standard USB (2.0) cable that came with a card reader I had. I did pick up a Dynex brand cable of the same length with gold connectors for like $7 online and tried it for giggles. No difference after about a month of use.

From what I've read in here I do not believe that people consider a proper burn in for these USB devices before giving up on them. I bought my U24 from a guy who had less than 10 hrs on it. For me this device did not come into its own until after more than 200 hrs of use. It wasn't until then that the sound bested my Adcom GCD-700 (which probably isn't saying much).

So in my experience there is no difference in USB cabling other then the length that is utilizied and would say that less than 6' is probably better. More important consideration should be given to IC's between your USB device and gear. Also consider a reasonable amount of burn in time for your USB device.

I am using Monster Cable for my Computer to DAC connection.
There was a test afew years ago in Maximum PC comparing the Monster Cable, a QVC, and a generic cable. The Monster was marginally faster than the QVC (which is my standard for most applications), and both blew away the generic.

Of course, this probably has no real-world effect:)