Real USB loopback effects

[somebodyelse]

I don't understand your reasoning. To me the ethernet cable is about practicality not being 'hi-end' - the server happens to be in a different room to the hifi. The potential issues of connecting a computer (which is what a streamer is) to a DAC are unchanged.

I have an HTPC that is a client for music but a server for the satellite TV. That connects to the TV via HDMI, and to a USB audio interface with balanced ins and outs. The TV RCA outs go to the balanced ins on the interface, so there is a loop. Balanced out from the interface goes to balanced in on the amp. I have no ground related noise issues, so no need to use a USB isolator.

 

[phofman]

@widemediaphotography : IMO no need to complicate it. There are two conditions for a ground loop effect on audio device to occur:
1.

The cable shield of an unbalanced cable is the very point where the noise is actually created, that's the thing you have to understand.

Pefectly said. I would just add - the shield forms a return channel for the unbalanced signal, where the noise current adds up to the signal return current, creating noisy voltage on the receiver-side cable terminals.

2. A noise current must be actually running through the unbalanced return channel (the shield), i.e. a voltage difference between both sides of the unbalanced analog-signal return channel must exist. That's not obvious at first sight.

E.g. :

A) Your case (IIUC): earthed PC -> earthed (class I) USB DAC -> unbalanced analog -> amplifier: Very likely there is a noise ground-loop current throught the GND wire/shield of the digital-signal USB cable. But it does not affect return channel of the unbalanced analog line. If the amplifier is earthed (class I), a noise ground loop current could run through the analog-cable shield, but why should there be a voltage difference of GND between the DAC and the amp? There are no large currents running through the DAC GND traces which could create this voltage differential.

B) Another case is rather common: earthed PC -> unearthed (class II) USB DAC -> unbalanced analog -> earthed (class I) amplifier: Here the ground-loop currents run through the USB GND AND through the unbalanced analog shield (return channel), creating noisy voltage between the receiver-side terminals of the unbalanced analog connection.

C) The case above can be fixed with a balanced line instead: earthed PC -> unearthed (class II) USB DAC -> balanced analog -> earthed (class I) amplifier: The noise ground-loop currents still run through the shield of the balanced analog cable, but the shield does not serve as return channel for the analog signal and the created noise voltage is not added to the analog signal being transferred.

If only the amp has balanced inputs, often a simple passive unbalanced-balanced cable can turn the case B to C, eliminating thus the ground-loop noise effect.

Of course breaking the continuous ground connection with galvanic isolation solves the issue too.

 

[solderdude]

It really does not matter HOW the system is made and what is connected how.
When you don't hear unwanted noises everything is fine.
When you do ... find out where it comes from.

 

[widemediaphotography]

Why have you added HQ Player into the mix? No one has mentioned that in this thread.

You can find in many other Hi-Fi forums, as one reference audio chain to solve loopback issue:

1. PC/Server HQPlayer - - ethernet cable - - Raspberry/NUC NAA - - USB cable - - DAC
2. PC/Server HQPlayer - - ethernet cable - - UPnP network player - - DAC

Why do not connect pc- DAC with USB optic cable?

 

[Jimbob54]

You can find in many other Hi-Fi forums, as one reference audio chain to solve loopback issue:

1. PC/Server HQPlayer - - ethernet cable - - Raspberry/NUC NAA - - USB cable - - DAC
2. PC/Server HQPlayer - - ethernet cable - - UPnP network player - - DAC

Why do not connect pc- DAC with USB optic cable?

And what impact has anyone said or implied HQplayer (or any pc playback software for that matter) has on ground noise?

 

[pma]

It really does not matter HOW the system is made and what is connected how.
When you don't hear unwanted noises everything is fine.
When you do ... find out where it comes from.

This is highly questionable statement. Even if you do not directly “hear” the buzz or other kind of unwanted noises, there may still be an effect to sound quality. Only a direct comparison of perfect/compromised setup would tell you if it matters or not. Otherwise you know nothing.

 

[Killingbeans]

Why do not connect pc- DAC with USB optic cable?

Because jumping through silly hoops gives you an an excuse to buy more gear.

PC --> toslink --> DAC

That would be a cheap and effective solution, and it would most likely not degrade the signal in any degree worth mentioning.

 

[phofman]

It really does not matter HOW the system is made and what is connected how.
When you don't hear unwanted noises everything is fine.
When you do ... find out where it comes from.

IMO understanding the causality of the issue is crucial. If a mouse is audible, the initial source is obvious. But the solution is not to avoid using the mouse or somehow try to reduce the currents flowing through the GND traces in the PC.

 

[widemediaphotography]

Because jumping through silly hoops gives you an an excuse to buy more gear.

PC --> toslink --> DAC

That would be a cheap and effective solution, and it would most likely not degrade the signal in any degree worth mentioning.

Toslink has theoretical limit of the bandwidth to 24/96 KHz... out of the game!

 

[widemediaphotography]

And what impact has anyone said or implied HQplayer (or any pc playback software for that matter) has on ground noise?

The problem lies in the use of upsampling that all DACs perform. PCs can do this much better than any DAC, as long as you use DACs that are truly native DSD. That is, a signal upscaled beyond a certain value is bypassed by the DAC.
All DACs apply filters when upsampling. HQplayer is very specialized in this.
However there are few native DSD DACs where internal upsampling can be bypassed.
I hope I have clarified your doubts

 

[Jimbob54]

The problem lies in the use of upsampling that all DACs perform. PCs can do this much better than any DAC, as long as you use DACs that are truly native DSD. That is, a signal upscaled beyond a certain value is bypassed by the DAC.
All DACs apply filters when upsampling. HQplayer is very specialized in this.
However there are few native DSD DACs where internal upsampling can be bypassed.
I hope I have clarified your doubts

No. What's HQplayer got to do with usb related ground him. Which I understood to be the point of the thread.

 

[Killingbeans]

Toslink has theoretical limit of the bandwidth to 24/96 KHz... out of the game!

144dB dynamic range, and 48kHz treble limit.

What kind of strange game are you playing?

 

[widemediaphotography]

No. What's HQplayer got to do with usb related ground him. Which I understood to be the point of the thread.

1. Fanless NUC with a 'bit-perfect' player (I like foobar2000) connected to a discrete DAC like the SMSL D0300 via a USB cable.

That is my config.

​

Why don't you say the same about Foobar2000? These are solutions that can be found in other Hi-Fi forums as recommended configurations. If you have any adversity towards HQplayer I will try not to mention it... Have you tried upsampling bypassing the oversampling of the internal DAC? Do you recommend it?

 

[widemediaphotography]

144dB dynamic range, and 48kHz treble limit.

What kind of strange game are you playing?

A lot kind of DSD files!

 

[Jimbob54]

Have you tried upsampling bypassing the oversampling of the internal DAC? Do you recommend it?

Not when I'm trying to avoid usb related ground loops.

And you seemed to just mention it. Again.

 

[Verig]

You can find in many other Hi-Fi forums, as one reference audio chain to solve loopback issue:

1. PC/Server HQPlayer - - ethernet cable - - Raspberry/NUC NAA - - USB cable - - DAC
2. PC/Server HQPlayer - - ethernet cable - - UPnP network player - - DAC

Why do not connect pc- DAC with USB optic cable?

Even easier:

Streamer --> amp

Just stream wirelessly, no need for cables, servers, players, formats or upsampling etc. boxes to consider fooling around with.

A streamer is just a pc. But built so that it doesn't come with typical pc problems. At least good ones are. Another thing is that there's no way I'll have a pc in my living room.

 

[Killingbeans]

A lot kind of DSD files!

Alright... this ties into the stuff you wrote above about upsampling?

Sounds sort of religious to me. I have a feeling that any appeal to reason wil have zero effect.

 

[widemediaphotography]

Even easier:

Streamer --> amp

Just stream wirelessly, no need for cables, servers, players, formats or upsampling etc. boxes to consider fooling around with.

A streamer is just a pc. But built so that it doesn't come with typical pc problems. At least good ones are. Another thing is that there's no way I'll have a pc in my living room.

If a streamer is a PC, why would I have 2 PCs in my audio chain? Therefore, an audio PC that can do everything is better, but not a streamer! Maybe PC - optical USB cable - DAC

 

[widemediaphotography]

Alright... this ties into the stuff you wrote above about upsampling?

Sounds sort of religious to me. I have a feeling that any appeal to reason wil have zero effect.

If music or sounds are sort of a religious, why should the DAC be allowed to manipulate it? If there is something that can handle music better than the DAC why not use it. The oversampling and filters applied by a PC seem to be better than those possible with a DAC, why deprive yourself of them, why not try before establishing a sentence?