But we havn't even scratched the surface of the audio improvements you can get by using a USB hub especially when it is disguised as an audiofool device and surrounded by some wonderful mystique... Though if I was cynical I would suggest that someone has seen and created a niche market so they can suck more money out of gullible audiophiles and every man and his dog is jumping on the bandwagon to bring devices to market, but I am not so will bask in the wonderful anecdotes of the vast improvements these devices bring even when measurements and sanity would suggest otherwise.
Ground plane and power plane noise data?
USB PHY current surge/power-related noise data during bit-determination with the usual signal integrity issues?
And this during async USB real-time streaming, which isn't the same thing as file transfer contrary to what you erroneously implied?
Someone was asking for someone else who knows to enlighten, here is one. barrows knows.
Interesting stuff about the clock, barrows. I've not yet experimented with those, but I've done some really easy variations around signal integrity as well as the ground plane where the SQ effects are obvious and higher than what one would expect.
None of which matters a jot if the output from the DAC is clean down to -120dB. If there is a problem in the input it will be seen on the output. Not seen then no problem, other than in your mind!
Async USB audio just sends data, protected by a CRC. Something has to be seriously wrong to get data corruption, and it would be very noticeable when it happens. If it is not happening, then the data is perfect. Whether the DAC then does a good job of turning perfect data into an analogue signal is a separate issue, and would be unaffected by USB cables etc. (apart from the usual audio issues like ground loops).
This is a boring story, and will not lead to sales or exciting magazine articles.
This is a boring story, and will not lead to sales or exciting magazine articles.
I didn't miss it, it just seems to be irrelevant. Do you have any data beyond stories and handwaving?
I can not believe the blind ignorance this thread is promoting...
USB audio signal is a digitally encoded analog signal stream in nature, sent down the twisted pair transmission line cable.
The only digital signal that I know of is either ON or OFF signal, like in PLC digital input cards, where +24V means ON, and 0 volts means OFF.
Nick
Care to explain to us idiots for whom USB appears to be very digital caring little about absolute levels as long as they are within the spec. For USB2 0.38V and 0.42V are both 1. Nothing analog about it.
Then send that digital signal of yours down the figure 8, un-shielded cable, like one you'd use for speakers. According to you knowledgeable fellas, it is only USB digital, right?
How is a digital stream of ones and zeros coming out of the USB a different 'digital' that the stream of ones and zeros in your PLC then?
Jan
One could discuss the intricacies of bit stuffing working nrzi encoding, but I don't think that's central to your point. You can still look at USB protocols as edge clocked, so that says digital to me.
Google is your friend…
USB 1/2 are half duplex using one twisted differential pair and 3 is full duplex using two twisted differential pairs. The signaling is digital using differential drivers and receivers.
There are a number of parameters that are measured to confirm a cable meets requirements but differential impedance, skew and insertion loss are the most important for proper operation especially for USB 3.0.
The differential impedance of the cable assembly (twisted pairs + connector) is supposed to be tightly controlled. The certification mask for the connector to cable interface and the connector to connector interface loosens a bit. A differential TDR or Vector Network Analyzer with special SW is used) The time skew between the + and – wires of each pair must be tightly controlled. This is measured with special equipment. If the cable loss is too high due to crappy quality cable or too long a cable, you will have problems. If the skew between the + or – wires of a pair is too high, you will have problems.
Building cables to meet all the requirements in the certification document and at the same time being cheap is nearly impossible. You can start out with good cable and connectors, but the fabrication time required to meet certification is too high to meet cost. In general companies make “gold standard cables” to be submitted for certification.
USB 1/2 are half duplex using one twisted differential pair and 3 is full duplex using two twisted differential pairs. The signaling is digital using differential drivers and receivers.
There are a number of parameters that are measured to confirm a cable meets requirements but differential impedance, skew and insertion loss are the most important for proper operation especially for USB 3.0.
The differential impedance of the cable assembly (twisted pairs + connector) is supposed to be tightly controlled. The certification mask for the connector to cable interface and the connector to connector interface loosens a bit. A differential TDR or Vector Network Analyzer with special SW is used) The time skew between the + and – wires of each pair must be tightly controlled. This is measured with special equipment. If the cable loss is too high due to crappy quality cable or too long a cable, you will have problems. If the skew between the + or – wires of a pair is too high, you will have problems.
Building cables to meet all the requirements in the certification document and at the same time being cheap is nearly impossible. You can start out with good cable and connectors, but the fabrication time required to meet certification is too high to meet cost. In general companies make “gold standard cables” to be submitted for certification.
You do know that USB 1.x and 2.0 is un-shielded? However as it is a system rather than a cult the cable parameters are clearly defined in terms of twisted pair and characteristic impedance. Meet those and it will work.
USB 3.0 to hit gigabit rates has Shielded cable.
Crosstalk is the main reason the two USB 3.0 twisted pairs in the cable are shielded. The USB 1/2 twisted pair in the same cable is not shielded. All require a DC common to keep CM low.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.