USB SQ, what can one expect

Willie wrote:

"Asynchronous USB interfaces are not sensitive to the quality of the USB cable.

I have personally experienced better USB cables do make a discernible difference i.e. in my system!"

It is true that Async eliminates the problem of jitter from the computer getting into the master clock of the Async interface, however there are other mechanisms that can still cause jitter due to the USB computer interface and the USB cable.

The main mechanism is common-mode noise. The USB interface is differential, so in theory all that matters in the signal is the difference between the two signal wires. However, in practice any noise on both signals (common-mode noise) is not rejected completely by the Async receiver. Because this noise is not completely rejected, it adds to jitter in the Async interface.

There are a couple of ways to eliminate or reduce this noise:

1) filter the USB interface
2) use a galvanically isolated USB interface

I have solutions for both of these:
1 Short Block
http://www.empiricalaudio.com/products/short-block

2 Power Block
http://www.audiocircle.com/index.php?topic=128620.0

Another mechanism that can cause USB cables to differ is the error rate. It is best to get a high-quality USB cable. Same for S/PDIF coax. Use a high-quality cable there too.

Steve N.
Empirical Audio

One other way to reduce/eliminate noise on the USB cable is to use the dual USB cables such as the one I identified on my first post of this thread

And we come full circle :-)

Back to the tunes

Steves the man! He has all the solutions. You can put all
the lipstick you want on the usb pig that it is. Willie, I
will read through this. I have been posting my position
about usb on agon for years. I would just like for people
to take a peak at USB and how it works. The very basic
premise of it transmiting packets of data instead of a
stream is a big problem. The second is that is is just what
is is - a 'universal' seriel bus and you just dont want
your music going through this. All these implementations
to better usb are getting better but you are starting with
the inferior medium of usb. Just try a good soundcard
and you can easily hear what usb doesnt do. I will read
the links. Thanks,

Hi Willie, that took me two seconds. I have read both of
those in the past. It fails to illustrate the inherent
problem with USB. You can compare diffent systems. I am
talking about the poor music interface of USB due to its
design. You can see more and more with new designs that
USB is flawed (after ALL these years, now there is a dual
usb cable with separate power? How may understand that
USB, with all the problems, even has power going through
that cable - and some designs add more power and some
dont?) There is NO standard cause its, excuse the pun,
"wilie nillie". See, Willie, back when
computers started getting into everybodys house, customer
sevice was out of control and too big an expense because
users struggled with attaching printers, and whatnot.
So, that made it idiot proof and created USB. And now
all us idiots 'assume' thats the best place to hook our
$100k hifi to. me says no. Me says take a look at it
and make an intelligent decision based on science and
knowledge, not based on what Steve Nugent wants to sell
us.

Cerrot - you are correct...

Simply put...
- USB does send data in packets
- each packet contains a CRC - Cyclic Redundancy Check
- the CRC tells the DAC whether the packet received was identical to the packet sent
- Asynch USB allows the DAC to request packet re-transmission
- if incomplete the DAC requests the same packet until the CRC indicates the packet is complete
- even if the CRC is corrupted the DAC simply requests the packet again
- the DAC then assembles the data stream once complete and plays it.

If this did not work the DAC would simply stop playing or skip - which in my case it does not.

This is the same technology used to transfer the spreadsheets and accounting info of Fortune 500 companies - and it works - otherwise computer systems in general would fail miserably - they don't

S/PDIF does not have CRC checks and balances and therefore you NEVER know if bits are getting dropped.

Yes - there may be other issues with USB, but transferring the correct data stream to an Asynch DAC is not one of them

Here's another link
http://www.thewelltemperedcomputer.com/KB/USB.html

Is it the perfect solution - maybe not

But it is NOT as bad as you are portraying it

I think it really depends on which USB approach a specific DAC manufacturer has implemented.

I have observed power issues with DAC's that power the USB circuitry via the USB cable, which most quality USB DAC's tend to avoid

Regards