USB & data

[Reid Malenfant]

Hi good people, I have a question.

Ignoring the positive & negative power feed in a USB cable & just looking at the data part of it, does anyone know if these are balanced? Frankly I don't think they are, & one would be the equivalent of ground while the "hot" connection squirts the data through...

I'm asking simply because the Musical Fidelity V-Link I feed from my laptop is now externally powered, so in theory I could just fit a data cable with maybe a data conductor & screen, the same goes for the HDD caddy I'm using, as it has an external PSU....


Or am I a country mile out here :D


Thanks in advance!

[StanleyB]

They are balanced. The positive and negative data lines feed directly into the USB chip, whichever make or model chip it is. Don't make the mistake to connect the negative data line to ground. I have fried a few chips that way.

[Reid Malenfant]

Ah, cheers Stan, that's even better news :D So I guess that they are balanced about the ground connection.

If so it looks like a balanced cable will do the business :)

Many thanks indeed!

[Covenant]

If you don't mind me going slightly off topic Mark, I was pondering why a USB cable between my Beresford Caiman and SB Touch sounds so much better than RCA. The RCA is a very good quality one but the USB one is the cheapest of the cheap. I guess it must be to do with the USB circuit rather than the cable itself.

[Reid Malenfant]

Possibly yes, as Stan mentioned, they are balanced in the USB circuit which means that the USB should have a bit of common mode rejection, where as the S/PDIF wont :)

I try to use balanced interconnects on as many circuits as possible, whether that happens to be analogue audio or AES/EBU digital.

I can say that a cheap MIDI lead I bought to try out AES/EBU digital (I fitted cheap XLR connectors to) peed all over an expensive S/PDIF, which amounts to the same thing :D

I'm going to be making my own USB cables with some high quality OFC balanced audio cable, I no longer use the 5V power link to any of my kit, so it makes sense :cool:

[Covenant]

:eyebrows:If you fancy making me one at the same time Mark I will cross your palm with silver.(No worries if you haven't got the time).

[Reid Malenfant]

You'd best get a PM off to Stan to find out if the DAC your using takes any power at all from the 5V line in the USB cable. I doubt it does, but best to check ;)

If it doesn't then you are laughing & I'll be more than willing to make you a cable :)

[Covenant]

Great stuff, I shall ask his Lordship.

[sq225917]

The data lines on USB are + and -, Gnd isn't referenced in assigning a bit value based on the voltage of the two signal lines, it is merely the difference in voltage between the two lines that decides whether a bit is a 0 or a 1.

So strictly speaking they aren't balanced, they are differential- there is no scope for common mode rejection within the USb data lines, one data line is not an inverted version of the other as they are with balanced signals.

[Reid Malenfant]

Still no harm in giving them a decent cable though, seeing as the 5V line is out of the window ;)

I don't understand it or I wouldn't have asked the question in the first place, but when I want to I generally do a bit of digging, & here is as good a place as any :)

Cheers Simon.

[sq225917]

Mark, no harm in giving them a good cable at all. Part of the USB spec is a fixed impedance, so might be worth looking that up as geometry could have an adverse effect. I'd be tempted to strip a CAT7 cable and solder that in place, I think the impedance is the same as USB as the shielding will be first rate.