Home Computer Audio Asylum

Music servers and other computer based digital audio technologies.

RE: USB cable length

The assumption on which your Positive Feedback article was based what that the time of signal transition mattered. This is correct with SPDIF, because this is how the audio clock is encoded. Unless some kind of reclocking is used jitter in this clock signal will make it directly to the clock used to do the digital to analog conversion. With regard to USB there is no relationship between the bit clock used to transmit individual bits in a USB packet. Jitter on this clock is only relevant if it becomes large enough to close the eye pattern at the receiver and cause data errors.

Going into a little more detail on how SPDIF works, it is useful to contrast it to the simpler I2S. Clock transitions on SPDIF depend on reflections from previous signals on the wire. These signals are constantly changing, because of changing audio data bits. This means that the time of transition at the receiver will be constantly varying due to cable reflections. Now contrast this with I2S. With I2S there is a separate clock wire. Reflections on this clock wire will not cause jitter, because the clock signal is periodic. The most any reflections will do is to create a constant phase shift in the transition time of the clock signal at the receiver. This won't add any jitter to the received clock.

USB is vastly more complex than SPDIF and I2S, being a packet based system. USB transmits audio data ischronously, with one frame sent every 1 msec, with the number of samples in each frame depending on the audio sample rate. The frame rate must be kept synchronized with the DAC master clock.

In the original USB audio, the frame rate was controlled by the source and the DAC forced to derive its master clock from the frame rate. Note that the frame rate has nothing to do with the data rate used to transmit bits on the USB, which is a separate clock altogether. The USB data clock does not have to be synchronized to the USB frame rate. For this reason, jitter on the USB data is irrelevant to the audio clock, except as to determination of the start of frame. This is handled by the USB protocol with each frame starting with a fixed header. At the time a frame is transmitted, the USB cable is completely idle so there will be no reflections. As the header is transmitted there will be reflections, but as the same header is transmitted with each frame the impact of these reflections will be constant. Hence cable reflections will not be a source of jitter. (There is a much worse source of jitter when running at 44.1 kHz, namely the fact that the audio clock is not an exact multiple of the frame rate, causing serious audio quality issues with early USB receiver chips.)

With the newer asynchronous USB audio, the connection between the frame clock and the DAC master clock is broken. The DAC master clock runs at its own rate. The computer's frame clock runs at its own rate, and the size of data in each packet is adjusted to keep buffers from underflowing or overflowing. With this mode there is no connection between the timing of signals on the USB cable and the DAC master clock. If transport or cable changes affect the sound it is not due to jitter, its must be due to other factors, such as noise coupling between portions of the DAC circuitry.

As to rise times. The data rate on USB is much higher than SPDIF (except the older USB 1.1 which is not suitable for hi-res audio). This requires fast rise times, otherwise there won't be an open eye pattern. With the slower SPDIF it is debatable whether fast risetimes are desirable. They reduce the jitter impact of noise at the receiver, but at the expense of creating high frequency noise that may be hard to keep away from critical DAC circuitry. Better to have a system architecture that avoids this dilemma. Both I2S and USB meet this criteria.



Tony Lauck

"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar



This post is made possible by the generous support of people like you and our sponsors:
  Schiit Audio  


Follow Ups Full Thread
Follow Ups
  • RE: USB cable length - Tony Lauck 09/13/1411:28:27 09/13/14 (1)

FAQ

Post a Message!

Forgot Password?
Moniker (Username):
Password (Optional):
  Remember my Moniker & Password  (What's this?)    Eat Me
E-Mail (Optional):
Subject:
Message:   (Posts are subject to Content Rules)
Optional Link URL:
Optional Link Title:
Optional Image URL:
Upload Image:
E-mail Replies:  Automagically notify you when someone responds.