Real USB loopback effects

[Roland68]

Pin-4 which is the - for the power supply and the reference for the data lines (for 4-pin USB connections)

I wanted to know from @KSTR which of the two ground connections he meant.

 

[Roland68]

Mine is a technical question, to understand what happens beyond real practice. I simply need to raise the volume of the amplifiers to the maximum to verify that both the headphones and the speakers emit real zero noise even when I move the mouse on my audio-pc connected to the SMSL DO300 DAC
However, these are just my experiences, not based on quantitative but only qualitative phenomena.
If the USB cable does not carry the 5V voltage, because it has the pin disconnected, or it is a split cable, could the loopback phenomenon exist? Is it the same if I have pin +1 connected?

You can cover the + pin with a small piece of sticky tape or something similar and try it out.

 

[solderdude]

I wanted to know from @KSTR which of the two ground connections he meant.

That's what he meant, not the screen.
Somewhere, at some point, the screen and pin-4 are usually connected on the either or both sides of the connection as for a shield to be effective it has to be connected to the enclosure. Preferably the shield should be connected directly to that enclosure while pin-4 might be connected to the USB circuit.
In most cases the common (- of the 5V connection) is connected to the enclosure somewhere so via a different 'routing'.
So cutting the minus wire (when it is separate) might still be bridged by the screen (if that isn't cut as well).
Not recommended to cut the 4-pin but could (when shielding is done properly) still 'work' in some rare cases where USB leakage currents can enter the PCB on an unwanted spot.
So... highly situation dependent and not recommended for a few reasons.

You can cover the + pin with a small piece of sticky tape or something similar and try it out.

That won't do anything in this case as the ground connection is (most likely) the culprit and that remains so when you cut the 5V.

Whenever these kinds of problems exist the culprit is a not desirable current, due to 2 devices not having the same ground potential and as a result a current will flow, some of that current, due to the way a PCB layout is done or internal wiring is done at the point where the digital and analog section are 'combined' a small current at the wrong spot can create a small voltage in the output signal.
As that part of the circuit/layout can not be altered the solution is to ensure that 'equalization current' can not flow.
An USB isolator (with DC/DC converter or its own (separated) in and output circuit for 5V or a pure optical connection (so not optical data and still a 5V power supply that is not DC/DC converted will help.

And even the usage of a DC/DC converter might not fully help with RF leakage currents. Toslink will with certainty.

 

[widemediaphotography]

In the meantime, I would like to thank everyone who attend however something still doesn't add up. If the culprit is an undesirable current, due to the fact that 2 devices do not have the same ground potential and consequently a current will flow, it would be sufficient to be sure that the DAC and the PC have an active connection on the same system ground electric.
Observing the PIN-out of a USB port, if there are no incompatibility problems by eliminating the 5V voltage (PIN 1), while the earths (PIN4) remain connected, we should find ourselves in the condition that the PC only sends data to the DAC, exactly with the connection toslink which is immune to loopback noise. The DAC would only have its potential and the PC no potential, not being electrically connected as is the case with TOSLINK. Then when I read about the harmfulness of excluding the voltage, I wonder what could happen if I connect a DAC via USB, turn on the DAC and never turn on the PC. Every time this happens do we risk breaking the DAC?

 

[phofman]

If the culprit is an undesirable current, due to the fact that 2 devices do not have the same ground potential and consequently a current will flow, it would be sufficient to be sure that the DAC and the PC have an active connection on the same system ground electric.

In case of PC the problem is not a different ground potential for the PC and the amp, but different ground potentials of the USB GND (or integrated soundcard output GND or PCI(-e) port GND for added internal soundcard) point on the motherboard and the PC inlet GND, caused by huge currents flowing through the motherboard GND traces/planes. That's why often the ground loop manifests itself as noise related to CPU work, mouse/keyboard work, HDD work, GPU work (e.g. I could hear letters being printed in linux GUI terminal with one particular ground loop)

 

[KSTR]

@widemediaphotography, personally I can only recommend USB isolators like Topping HS01 etc. It just works, and is good enough isolation to prevent typical ground-loop problems. You could still interrupt the +5V line coming out of the isolator (but keep the GND/Shield!) but as noted, not all devices will work and bus-powered devices can't work anyway. I' using an Intona USB2.0 and a few HS01's without troubles (the HS01 is a bit picky, sometimes).

 

[widemediaphotography]

@widemediaphotography, personally I can only recommend USB isolators like Topping HS01 etc. It just works, and is good enough isolation to prevent typical ground-loop problems. You could still interrupt the +5V line coming out of the isolator (but keep the GND/Shield!) but as noted, not all devices will work and bus-powered devices can't work anyway. I' using an Intona USB2.0 and a few HS01's without troubles (the HS01 is a bit picky, sometimes).

As I already mentioned, an HS01 is installed inside my case, but if I remove it I don't notice any disturbances and/or differences. That doesn't mean the insulator doesn't do something, maybe I can't hear everything just with my ears, maybe I could measure it and realize the differences are out of reach of my old ears! I'm interested in understanding how much science there is on this phenomenology...
Let's limit the analysis to the case of the USB port of the PC and the DAC, in my case if I raise the volume to the maximum (200watt) of my system, or of the headphones (Hifiman He1000), the DAC does not emit any buzz but always remains absolutely silent . Even when I move the PC mouse.
If you changed your PC or DAC and noticed a hum, it would be easy to figure out that the problem is with your PC or DAC. So how do you accept that there may be DACs or PCs that have this problem? I have connected at least 4 PCs/laptops and 1 Raspberry to my DAC (smsl DO300), without ever noticing a hum. It's a lucky DAC, I have 5 lucky PCs? I say that in my opinion there are no ghosts and magic and therefore it should always be possible to establish in advance if that DAC or that PC has loopback problems. If I don't connect the GND of the turntable it's mathematical that I will only hear noise!

So, what would be so difficult about connecting all the “critical” GNDs together and being sure that at each point the potential is always the same?
I therefore deduce that this cannot be the cause of the phenomenon and even more so that the USB PC - USB DAC potential will always be the same, if the pair of pin 4 of the cable are always connected, including any shielding of the cable.
The fact that isolators sometimes work and sometimes don't already shows that we still don't have a great understanding of this phenomenology!

 

[AnalogSteph]

As I already mentioned, an HS01 is installed inside my case, but if I remove it I don't notice any disturbances and/or differences. That doesn't mean the insulator doesn't do something, maybe I can't hear everything just with my ears, maybe I could measure it and realize the differences are out of reach of my old ears!

A cursory look at your posts shows you seem to have an SMSL DO300 and a Violectric HPA-V222, presumably connected via XLR (it would be silly not to).

Both of these devices are IEC Class I (earthed), so the DO300 would be dumping most of the ground current coming from the USB into its mains PE connection right away, with little ground potential difference remaining between both DAC and amp. Since I assume neither unit has a Pin 1 Problem to contend with, the rest is then taken care of by the balanced input's CMRR. Guess what, balanced connections will work exactly as advertised if done right. (You generally wouldn't even have any major issues with unbalanced connections in such a scenario though.)

 

[widemediaphotography]

A cursory look at your posts shows you seem to have an SMSL DO300 and a Violectric HPA-V222, presumably connected via XLR (it would be silly not to).

Both of these devices are IEC Class I (earthed), so the DO300 would be dumping most of the ground current coming from the USB into its mains PE connection right away, with little ground potential difference remaining between both DAC and amp. Since I assume neither unit has a Pin 1 Problem to contend with, the rest is then taken care of by the balanced input's CMRR. Guess what, balanced connections will work exactly as advertised if done right. (You generally wouldn't even have any major issues with unbalanced connections in such a scenario though.)

I confirm that the DAC and the headphone amplifier are connected with a balanced XLR cable, but the DAC is also connected to the YAMAHA RX-V1200 amplifier with unbalanced Focal RCA cables and in any case I've zero noise with maximum volume at 4 Infinity tower speakers. In some points of the 230V electrical system in my house I can measure NO ZERO volts between L and N and between Line and GND

 

[solderdude]

Do I understand correctly that you do not have any hum or other issues with your DAC and your PC at all ?
Neither with nor without the isolator ?
Were you just trying to learn why weird sounds can happen and what would be the cause ?

 

[AnalogSteph]

I confirm that the DAC and the headphone amplifier are connected with a balanced XLR cable, but the DAC is also connected to the YAMAHA RX-V1200 amplifier with unbalanced Focal RCA cables and in any case I've zero noise with maximum volume at 4 Infinity tower speakers.

Which is quite plausible if there is nothing else connected to the AVR that would introduce another earth connection. Being classic hi-fi gear, it's an IEC Class II (double insulated) device, and most traditional sources tend to follow the same design paradigm - you want your system to work with all unbalanced connections after all. The most likely place to introduce another earth connection would be either an external FM antenna (though you can get isolation transformers for that) or a satellite receiver, as this model AVR does not sport any Ethernet yet.

 

[solderdude]

O.K. so there don't seem to be any ground loops that have become audible here.

One must realize that in the vast majority of cases audio systems (even the ones with RCA only) are not affected by any ground loops or too high leakage currents.
Whether or not they happen depends on many factors.
Chances are there is a small leakage current anyway but when the DAC/amp is well designed (PCB layout, schematic) that current will not make it into the audio path.

 

[widemediaphotography]

Do I understand correctly that you do not have any hum or other issues with your DAC and your PC at all ?
Neither with nor without the isolator ?
Were you just trying to learn why weird sounds can happen and what would be the cause ?

My intent is to understand the phenomenon well, because I am convinced that the best digital source is made up of a PC or something similar to it, and to hear that a streamer is preferable because it acts as an isolator and that the best connection is the one ethernet in 2024 cannot be listened to!
However, in this long examination, no incontrovertible cause of the phenomenon has emerged. For each of those proposals I could give counterexamples, such as the equipotentiality of GNDs (very simple to achieve) or the infallibility of insulators (I have one that I find rather useless). Why shouldn't we be able to establish in advance whether or not a device has loopback problems?

 

[solderdude]

My intent is to understand the phenomenon well, because I am convinced that the best digital source is made up of a PC or something similar to it, and to hear that a streamer is preferable because it acts as an isolator and that the best connection is the one ethernet in 2024 cannot be listened to!

Its nonsense. There is more nonsense floating around on the web than accurate info.

However, in this long examination, no incontrovertible cause of the phenomenon has emerged. For each of those proposals I could give counterexamples, such as the equipotentiality of GNDs (very simple to achieve) or the infallibility of insulators (I have one that I find rather useless). Why shouldn't we be able to establish in advance whether or not a device has loopback problems?

You can't do that up front. It is highly circumstance dependent. RCA cables, DAC, PC/laptop power supply, other devices connected to the same system, usage of 'power conditioners' or filtered power strips, even different sockets gear is connected to or even other gear connected to mains can cause audio gremlins.
In most cases, however, there are no audible issues. Not even with gear that is reported to have such issues.
Too many variables.

To understand it do the following:
Connect a few devices to mains.
Connect them to nothing else but mains.
Connect a multimeter between the ground of an RCA of the device and the other wire to safety ground.
Measure the voltage and note that voltage.
Now reverse the plug (if possible) and measure again.
Also measure AC current (mA setting) in these conditions.

Do this for multiple devices.

Now you will see all devices connected to mains 'leak' current. Connect those devices and they will 'connect' via interlink cables and thus all have the same potential .... but ... currents will flow.
When they are high enough AND RCA cables are used some of that current will cause a very small voltage across the ground wire. That may become audible if high enough.
Not only the cable resistance (of the shield) but also the layout and schematic inside the components may or may not be optimal for this.

Depending on what currents flow, how insensitive the connected devices are (how they are built), how crappy the cable is, how high the signal is there may or may not be any noise audible. It may even well be there but below audible levels.

So no.... you can't say up front whether or not you might be plagued by this or not. Using balanced connections helps prevent it. Using USB isolators or TOSLINK or even SPDIF can prevent this from happening but all is pointless IF there isn't an audible problem to begin with.

 

[somebodyelse]

My intent is to understand the phenomenon well, because I am convinced that the best digital source is made up of a PC or something similar to it, and to hear that a streamer is preferable because it acts as an isolator and that the best connection is the one ethernet in 2024 cannot be listened to!

You should probably read Bruno's The G Word and/or Jensen Transformers' AN007. They explain the potential for ground-relate noise and how to avoid it. The most common ground-relate noise issues are hum (classic ground loop) and 'USB noise' (leakage current modulated by PC activity like moving the mouse, changing GPU load etc.) Single ended interconnects make it a risk not a certainty. Properly engineered balanced interconnects avoid it, but not all are properly engineered (see 'Pin 1 problems' in the previously mentioned docs.) A networked streamer is one possible method to break a ground loop, but not the only one. Note that a networked streamer is really just another computer, and the same risk of ground relate noise may apply to it as to a PC.

or the infallibility of insulators (I have one that I find rather useless).

If you don't have a ground-related noise issue (effects are below audible levels) an isolator can't help.

Why shouldn't we be able to establish in advance whether or not a device has loopback problems?

That's sort of in the nature of ground related noise issues. There are well established ways to avoid them, used on a daily basis in the pro audio world because they want to connect things and have them 'Just Work' rather than waste time and money fixing avoidable noise issues. The HiFi world has chosen not to use these methods because most of the time it isn't a problem in a domestic setting...except when it is, at which point we get people asking how to fix it. AN007 above presents a step by step route to fixing such problems. If you don't have a problem then there's nothing to fix.

 

[wwenze]

The amount of ground noise can be approximated to the typical performance of onboard sound

If you don't hear noise from onboard sound (roughly -90dB typical, maybe -70 worst case), you situation is fine. And this is why this is often being the accepted performance level by equipment manufacturers - It's beyond audibility most of the time. Some of us just want -120dB for measurement or academic purposes.

As long as your equipment draws power or has some form of voltage reference from the PC, it will have the same noise, be it an internal soundcard or an external DAC.

Streamers are nice and all since they do not generate as much noise as a PC, but what if I told you streamers are sort of a mini PC themselves. And sometimes their power adapter is even noisier.

 

[KSTR]

I confirm that the DAC and the headphone amplifier are connected with a balanced XLR cable, but the DAC is also connected to the YAMAHA RX-V1200 amplifier with unbalanced Focal RCA cables and in any case I've zero noise with maximum volume at 4 Infinity tower speakers.

My intent is to understand the phenomenon well

Applied to above connection scenario, a ground-loop would manifest itself only in the unbalanced RCA connection from the DAC to the amp but because your Yamaha amp is 2-prong (class-II) the classic ground loop cannot form.
There will be some small capacitive balancing current but that is not large enough to cause significant noise voltage drop along the RCA cable shield.

Ground that amp and you'll probably get some noise but since the DAC is grounded too -- and this ground dominates the other noisy PC ground via USB -- it all about that ground connection vs the amp's ground connection, not the ground at the PC.

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.

 

[Killingbeans]

I don't think a double electrical/optical/electrical conversion is necessary for a single meter of audio connection with the double result of introducing jitter and limiting the files to 24/192 KHz. and make the clock travel together with the data. Toslink is an old standard from 1980, when chips were expensive!

Sounds like pure FOMO to me.

The risk of any of those three things degrading the signal audibly is close to absolute nil.

Nail guns has been around since the 50's. Sometimes an old tool is all that's needed for an undemanding job

 

[widemediaphotography]

The discussion is heading in an interesting direction, revealing that the phenomenon related to loopback/EMI/RFI disturbances is rather murky, where solutions that we might sometimes define as bizarre thrive.

Let's take an audio chain as a reference, consisting of:

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

Due to what we are discussing, the so-called hi-end variants emerge:

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

When it would suffice to have:

PC/Fanless NUC Foobar2000/HQPlayer - - USB optical cable - - DAC

I'm sure I've written something wrong, leave your opinions...
Thanks

 

[Jimbob54]

The discussion is heading in an interesting direction, revealing that the phenomenon related to loopback/EMI/RFI disturbances is rather murky, where solutions that we might sometimes define as bizarre thrive.

Let's take an audio chain as a reference, consisting of:

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

Due to what we are discussing, the so-called hi-end variants emerge:

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

When it would suffice to have:

PC/Fanless NUC Foobar2000/HQPlayer - - USB optical cable - - DAC

I'm sure I've written something wrong, leave your opinions...
Thanks

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