long cable length, analogue or USB?

[captain pugwash]

Just wondered what the best way to connect a usb dac to my amp, the 2 being about 5m apart. Is it best to go with a long USB cable and short Phono cable or vica versa?

 

[SprockeT1597]

A shorter analog would be better. usb can be up to 5m anyway.

 

[captain pugwash]

Thanks, I'm guessing there's less signal loss over the long usb cable than there would be with a 5m analogue signal carrier?

 

[magnoman]

Captain

I'd vote for running the 5m USB cable.
It's a nice terminated, balanced, low level digital interface.
Less chance of pick-up and introduction of ground currents etc.

Thanks
-Antonio

 

[SprockeT1597]

Signal loss over usb dosnt matter. The digital information going in is the same as whats coming out. But with analog, theres voltage drop due to the resistance of the wire, possible interference, and im guessing capacitance etc too

 

[sofaspud]

A DAC inputs a digital signal and outputs an analog signal. The amp inputs an analog signal and outputs an analog signal.
It seems the choice is between a long phono cable and moving the DAC or amp.

 

[captain pugwash]

thanks for the replies folks - I forgot to mention that the USB dac is powered only by the pC via the usb input (it's an ODAC) does that make any difference to the length of the USB cable?

 

[jcx]

the ODAC uses the USB/PC gnd as the analog output reference ground - so in either configuration gnd loops are possible with PC, AMP plugged into different outlets


and while the PC, USB port presumably had to pass some conducted emissions standards there is the issue of PC noise on any of its outputs

but this specific noise aspect really doesn't change much with either RCA gnd conducting the noise, or the USB cable

the PC noise may or may not be a audible issue - and it may be impossible to tell without the right measurement equipment, also depends on the amps sensitivity to RF, input filtering...

quality audio line transformer isolation could be "best" - breaks gnd loops, has limited bandwidth, cutting conducted RF noise - but cheaper ones introduce distortion, limit audio bandwidth


the gnd loop issue can be dealt with by other means - though the room wiring may end up looking strange

...
Or, select plan B or C from my pic, see below.


IEEE-1050 was great, although retired.

Here's the pic. Option A is typical of small home systems with no powered sub. If you use a separate outlet for the source, you extend the ground loop to the service panel, worse.

Option B is very simple, wrap the source line cord around the IC's, then around the amp power cord to the outlet.

Option C is a small variation of B. I cannot guarantee it will work for long runs however, as I have only done if for 125 foot runs of unbalanced line level between the source and the amp, with 15 kW of scr based lights, a quarter million volt vandygraf, and a toro 700 leafblower. Houses typically are worse...

jn

 

[captain pugwash]

Interesting stuff - I had guessed that there must be many set ups similar to mine with a PC several meters from an amp but my searches found little comment on the best combination of cables to join the 2, I then assumed that it made little difference or was blindingly obvious! - I think I will give the 5m USB cable a try