Does balanced connection prevent USB/ground loop noise?

[Graph Feppar]

I plan to buy Fiio K3, a tiny 99€ dacAmp that gets power through USB, it doesnt have separate power connection, no brick.
I have read alot and I noticed these cheap small dacs and dacAmps dont have USB galvanic isolation and often users report or even measure some noise or morr specific, inharmonic idle tones that I believe are harmonics of AC power. I am no expert but I believe its some ground loop USB coupled crap.

Does balanced connection to headphones help with this problem? Mind you, I mean the K3 which is amp and dac in one box, I think it might be different if the amp and dac are separate and are connected by unbalanced connection. I believe the K3 is internaly balanced.

 

[solderdude]

Does balanced connection to headphones help with this problem?

No. it only increases the maximum available output voltage (doubles it for high impedance headphones)
A groundloop cannot not be caused via headphones as they are fully isolated by centimeters of foam and plastic.

A possible groundloop can exist when the amps audio source is connected to mains and the fiio is also connected to mains via a separate power supply or when the audio source is not properly grounded for instance.

I drew this a while ago to demonstrate the loop:

When the audio input and USB power are coming from the same PC/laptop a loop can occur between the USB ground and audioground if the ground plane/wiring to the board of the USB or audio path is not 'perfect'.
In the above drawing you can view the 'mains path' as the ground plane/USB cable.
The solution for this is feed the FiiO with an extra wallwart instead of USB power but that too can create a groundloop.
Depends on the situation and used equipment.

 

[L5730]

So what is the downside to just lifting the ground in the balanced cable?
Sure the digital side and the conversion to analogue will be at one ground potential and the amplifier will have a different ground potential but how much of a big deal is that, if they are plugged into the same power strip, and not plugged into another outlet elsewhere in the building?

The only thing I see is possible offset of the AC signal going into the amp, and therefor potential clipping and some possible phase issue.
I guess the speaker driver might be held slightly off from it's usual relaxed position, which would ordinarily be the zero volts point.
I confess, I dunno how an amp and speaker works too well.

I am using pseudo-balanced with lifted shield which is an unbalanced termination to balanced termination with the ground connection left floating, using balanced screened cable, with the screen connected at the unbalanced ground connection.

 

[solderdude]

That depends on the situation.

For instance whether or not an input- or output-transformer is used in the signal cable or that the in- and out-puts are just opamp circuits.
In the latter case when a screen is not connected and there is a common mode voltage difference between the 2 connected devices the input circuits may be overloaded by that voltage difference when it becomes higher than the opamps can handle.

Balanced usually is for long (low) signal cables that run over longer distances, sometimes crossing or running in parallel to mains or lighting cables.
Grounds over longer distances can have voltages across them.
For short distances there are little advantages over well screened SE cables.
But ... a groundloop could be one of them.

Especially when using metal microphones it can be problematic to have a ground lift in the pre-amp (console) when that part is not properly connected to safety ground.

When there is a groundloop somewhere (implying there are at least 2 grounds) then it is probably not a problem to use the ground lift.
It's a safety thing as well as a screening thing.

 

[DonH56]

Might be helpful: https://www.audiosciencereview.com/...d-balanced-unbalanced-and-all-that-jazz.1352/