• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required as is 20 years of participation in forums (not all true). There are daily reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Home noisy cabling setup: can I improve it without changing the cables?

tonycollinet

Major Contributor
Joined
Sep 4, 2021
Messages
3,644
Likes
5,659
Location
UK/Cheshire
If headphones on the focusrite sound OK, then a jensen ISOMAX or similar device onthe last page could be what you need.
You could have two of the Chinese jobs clown earlier, and 10M of XLR between them, with 1M on each end.
Bear in mind that balanced connection doesn't eliminate the ground loop. That still exists in the screen of the balanced cable.

It simply makes *that* (balanced) connection immune to the noise. Any remaining unbalanced connections - including the 1m on each side - could still pick up the noise. So the 1m on each side should be more like 5cm on each side :)
 

Holmz

Major Contributor
Joined
Oct 3, 2021
Messages
1,796
Likes
1,047
Location
Australia
Bear in mind that balanced connection doesn't eliminate the ground loop. That still exists in the screen of the balanced cable.

It simply makes *that* (balanced) connection immune to the noise. Any remaining unbalanced connections - including the 1m on each side - could still pick up the noise. So the 1m on each side should be more like 5cm on each side :)

Yeah, good points Tony, but I thought that the ISOMAX specifically have a transformer that decouples the ground.

The data sheet for one of the devices says, “hum and ground loops” in it.
 

tonycollinet

Major Contributor
Joined
Sep 4, 2021
Messages
3,644
Likes
5,659
Location
UK/Cheshire
Yeah, good points Tony, but I thought that the ISOMAX specifically have a transformer that decouples the ground.

The data sheet for one of the devices says, “hum and ground loops” in it.
Yep - the isomax should break the ground loop. (in that circuit - one would be needed for each problematic loop if more then one exists) They are pricey though.

I was referring to the unbalanced to balanced to unbalanced conversion devices - which as far as I know, don't isolate the ground.
 

Martin Takamine

Active Member
Forum Donor
Joined
Sep 8, 2019
Messages
223
Likes
474
Location
East Coast
If you intend the Desk PC, PC to HP directly it's because with the HPs I can listen to the PC output with no need to switch on the Rotel; and Rotel to HP it's because that let's me listen to the all the other sources that goes to the Rotel, NUC included. One output of the Rotel goes to a Rotel RMB 1075 power amplifier that drives the hi-fi floor loudspeakers.

If you intend the NUC, yes I think it is not, as it is powered with a brick PS with a two poes plug.

Sorry but I don't understand this question.

To try to give you a better understang of my setup here's a "map":

View attachment 248415
My first goal is to be able to listen to the NUC via my headphones when I'm mixing at the desk PC, controlling the NUC with Remote desktop,
and the second goal is to be able to listen to the NUC through the room loudpeakers. Both goals with no noise, RFI/EMI or HUMs.

As a workaround, for the first goal I will try & use the "send sounds to the host" Desktop restore function, to see if there will be no quality degradation and no latency.
For the second I will try to use balanced cables from NUC to Rotel + a conveter.

But I can't think there's no way to track the source down and to tame it. Nobody have to use balanced cable in a house, haven't they?
f.
If you haven't done so yet, move the the Focusrite interface to desk PC and interface with USB. Use the same RCA cables and use an extension cord to use the same AC outlet currently being used. This should determine whether the NUC or Focusrite is the issue.
 

tonycollinet

Major Contributor
Joined
Sep 4, 2021
Messages
3,644
Likes
5,659
Location
UK/Cheshire
If you haven't done so yet, move the the Focusrite interface to desk PC and interface with USB. Use the same RCA cables and use an extension cord to use the same AC outlet currently being used. This should determine whether the NUC or Focusrite is the issue.
With earth loops it is not a particular component that is the issue - it is the system and how it is wired.
 

Holmz

Major Contributor
Joined
Oct 3, 2021
Messages
1,796
Likes
1,047
Location
Australia
Yep - the isomax should break the ground loop. (in that circuit - one would be needed for each problematic loop if more then one exists) They are pricey though.

I was referring to the unbalanced to balanced to unbalanced conversion devices - which as far as I know, don't isolate the ground.

I think we are on the same page.
But I only assumed that the Chinese version shown earlier was like an ISOMAX.


If you haven't done so yet, move the the Focusrite interface to desk PC and interface with USB. Use the same RCA cables and use an extension cord to use the same AC outlet currently being used. This should determine whether the NUC or Focusrite is the issue.

Alternatively use headphones at the Focusrite, and don’t run RACs anywhere…
If there is noise and hum there, or not, we have some big clues…

Hum usually does not get lower going through more devices, so it is not bad to start at one end and step through it.
headphones and RCA to headphone or earbuds are handy tool for that.
 

Destinyberry

New Member
Joined
Dec 8, 2022
Messages
2
Likes
0
If the problem is with the ground plane, you should, in my opinion, improve the ground connection by using a wire with a larger diameter on one side of the twisted pair.
 

tonycollinet

Major Contributor
Joined
Sep 4, 2021
Messages
3,644
Likes
5,659
Location
UK/Cheshire
If the problem is with the ground plane, you should, in my opinion, improve the ground connection by using a wire with a larger diameter on one side of the twisted pair.
Are you replying to the correct thread here?

There has been no ground plane discussed, and as far as I know, there are no twisted pairs.
 

Holmz

Major Contributor
Joined
Oct 3, 2021
Messages
1,796
Likes
1,047
Location
Australia
Are you replying to the correct thread here?

There has been no ground plane discussed, and as far as I know, there are no twisted pairs.

it looks a similar concept as in post #23 with a larger wire on the ground/sheild side of a RCA.
 
OP
F

fabiospark

Member
Joined
Dec 4, 2022
Messages
28
Likes
1
Start with max simplification - one single cable path from nuc to HP amp.
I don't use any audio output on the NUC. It is only connected with the Focusrite via USB.
Nuc to Interface to Rotel To HP amp. EVERYTHING ELSE COMPLETELY DISCONNECTED FROM THE SYSTEM. If the Nuc has a monitor, disconnect it and switch it off. Control Nuc from remote desktop - ideally via wifi (No ethernet cables to Nuc)
As stated above, here can only be Interface to Rotel and Rotel to HP amp

Anyway, I did even better, I think.
I resumed an analog Mackie mixer to use it as a balanced to unbalanced converter.
I put it beside the Rotel and I converted the old two stereo RCA "flat" cables into two mono unbalanced TRS terminated cables.
Then I connected the RGB stereo RCA cable I made yesterday from the interface to one stereo input (unbalanced) of the Mackie
and the new two mono TRS balanced cables I've just made, to another stereo input (balanced) of the Mackie, then I moved my headphones
to the HP out of the Mackie, leaving the HP ampli Head Box out of the equation.
On the Mackie, when I open only the channel with the unbalanced connection, with the NUC on, the RF noise is clearly present in the HPs;
when I open only the channel with the balanced connection there is no noise at all.

Scan 387.jpg

To hear the sound on the floor loudspeakers I connected a short RCA stereo cable (1m) from the Mackie tape out to the Rotel Stage input
and when I send the Stage input to the loudspeakers through the Rotel RMB 1075 (another RCA stero 1m) I can here that the RF noise is there, even if weaker than before.

To me it seems that each RCA unbalanced cable in this room is going to pick up the RF emissions of the NUC, regardless the position and the length of the cable...
so in the end I will have to find a way to stop the emissions from the NUC in some way (Faraday cage, I can only think...).
 
Last edited:
OP
F

fabiospark

Member
Joined
Dec 4, 2022
Messages
28
Likes
1
...
Any remaining unbalanced connections - including the 1m on each side - could still pick up the noise. So the 1m on each side should be more like 5cm on each side :)
I read this after I posted my last post (#50) and I can confirm that even a 1m RCA meters away from the NUC picks up the RF noise, even if maybe at a lower level.
 
Last edited:
OP
F

fabiospark

Member
Joined
Dec 4, 2022
Messages
28
Likes
1
If you haven't done so yet, move the the Focusrite interface to desk PC and interface with USB. Use the same RCA cables and use an extension cord to use the same AC outlet currently being used. This should determine whether the NUC or Focusrite is the issue.
As soon as I finish testing cables and configurations I will surely follow your advice. And I will try to change USB port on the NUC too.
 

tonycollinet

Major Contributor
Joined
Sep 4, 2021
Messages
3,644
Likes
5,659
Location
UK/Cheshire
I read this after I posted my last post (#50) and I can confirm that even a 1m RCA meters away from the NUC picks up the RF noise, even if maybe at a lower level.
Again - this is a ground loop issue - not a cables picking up RF issue.

Your ground loop acts like a single turn secondary of a transformer. The primary of the transformer is any source of magnetic fields. (In this case probably your NUC). When those mag fields pass through the loop (and your loop is large so a large part of that field will pass through the loop) it induces the noise current in the loop. That goes right around the loop (it doesn't matter how far away the cable is from the NUC - you just have currents flowing right around the closed loop now) - and those currents cause a voltage difference from one end of the cable to another.

Shielding the NUC is not going to help here.

You need to use balanced cables throughout - or break the loop. You break the loop by using a form of connection that doesn't connect the ground - eg an optical connection (hence the suggestion of optical from NUC to interface - or anywhere else in the loop - eg interface to rotel), OR by putting the isolation transoformers suggested upthread in there. You could also use a USB isolator device between NUC and Interface (assuming that loop is the only one you have)

Note that the ground connection in your usb lead can also form part of the loop - though I am not sure it does if the NUC is not grounded. If it is, you could also use a USB isolator device between NUC and Interface.

You may well need to break all loops in your system if you have more than one.
 
OP
F

fabiospark

Member
Joined
Dec 4, 2022
Messages
28
Likes
1
Again - this is a ground loop issue - not a cables picking up RF issue.

Your ground loop acts like a single turn secondary of a transformer. The primary of the transformer is any source of magnetic fields. (In this case probably your NUC). When those mag fields pass through the loop (and your loop is large so a large part of that field will pass through the loop) it induces the noise current in the loop. That goes right around the loop (it doesn't matter how far away the cable is from the NUC - you just have currents flowing right around the closed loop now) - and those currents cause a voltage difference from one end of the cable to another.

Shielding the NUC is not going to help here.

You need to use balanced cables throughout - or break the loop. You break the loop by using a form of connection that doesn't connect the ground - eg an optical connection (hence the suggestion of optical from NUC to interface - or anywhere else in the loop - eg interface to rotel), OR by putting the isolation transoformers suggested upthread in there. You could also use a USB isolator device between NUC and Interface (assuming that loop is the only one you have)

Note that the ground connection in your usb lead can also form part of the loop - though I am not sure it does if the NUC is not grounded. If it is, you could also use a USB isolator device between NUC and Interface.

You may well need to break all loops in your system if you have more than one.
I disconnected the NUC and connect a notebook to the interface, with the same USB cable that was connecting the NUC, and with this configuration there is no RF noise around the room. The NUC it is still on and connected to the LAN through an ethernet cable but no USB connection to the interface.
Does this help in any way to narrow down the possible solutions? USB isolator could be enough?
Thanks.
 

tonycollinet

Major Contributor
Joined
Sep 4, 2021
Messages
3,644
Likes
5,659
Location
UK/Cheshire
I disconnected the NUC and connect a notebook to the interface, with the same USB cable that was connecting the NUC, and with this configuration there is no RF noise around the room. The NUC it is still on and connected to the LAN through an ethernet cable but no USB connection to the interface.
Does this help in any way to narrow down the possible solutions? USB isolator could be enough?
Thanks.
Is the laptop also connected to Ethernet? Is the Ethernet cable to the NUC screened? (If it is screened it will have connectors with a metal shell). Was the laptop also connected to the HP amp with analogue - or did you remove that as well?

But yes - using the laptop confirms that somehow the nuc is forming part of the ground loop, and almost certainly isolating the USB will remove the noise. But bear in mind if the laptop was not connected to the HP amp, and the NUC was - this will also have an influence on the results

It might be cheaper to do this using a USB to Toslink converter, if you don't need to also get music from the interface over USB. EG:

The cheaper form of USB isolators are limited to the lower speed of USB, and I have no personal experience so I've no idea how they work - eg:
 
OP
F

fabiospark

Member
Joined
Dec 4, 2022
Messages
28
Likes
1
Is the laptop also connected to Ethernet?
No.
Is the Ethernet cable to the NUC screened?
No.
Was the laptop also connected to the HP amp with analogue - or did you remove that as well?
Only the interface audio outs are connected to the HP amp. No audio goes out of the NUC or the notebook.
But yes - using the laptop confirms that somehow the nuc is forming part of the ground loop, and almost certainly isolating the USB will remove the noise. But bear in mind if the laptop was not connected to the HP amp, and the NUC was - this will also have an influence on the results
Again, no audio connections from NUC or notebook.
It might be cheaper to do this using a USB to Toslink converter, if you don't need to also get music from the interface over USB.
Not sure what you mean here with "get music from the interface over USB".
I need analog audio from the outputs of the Focusrite to send it to the headphones and/or to the power ampli to listen to it through the floor loudspeakers.
And I'm usually doing this unbalanced from interface to the Rotel and now, temporarily, balanced to the the Mackie and then unbalanced from Mackie to the Rotel.

Later I will try to lift the groung of the NUC power supply and report back.

Thx.
 
OP
F

fabiospark

Member
Joined
Dec 4, 2022
Messages
28
Likes
1
I moved the ethernet cable from the NUC to the laptop and connected the USB from laptop to interface and there's no noise around.
 

tonycollinet

Major Contributor
Joined
Sep 4, 2021
Messages
3,644
Likes
5,659
Location
UK/Cheshire
Again, no audio connections from NUC or notebook.

Not sure what you mean here with "get music from the interface over USB".
I need analog audio from the outputs of the Focusrite to send it to the headphones and/or to the power ampli to listen to it through the floor loudspeakers.
And I'm usually doing this unbalanced from interface to the Rotel and now, temporarily, balanced to the the Mackie and then unbalanced from Mackie to the Rotel.

Later I will try to lift the groung of the NUC power supply and report back.

Thx.
You have a usb connection from the Nuc/notebook to the interface. That connection has a ground. That ground can form part of the ground loop - just as the ground connections in the mains cables (if they exist) can. Digital cables don't suffer the noise though, due to the inherrent noise immunity of digital signals. The cables however can still conduct the noise downstream to devices which do have susceptible analogue connections.

EDIT - and I asked the wrong question earlier. I should have been asking about the desktop PC connection to the HPamp, not the NUC/Laptop. Does that connection still exist?
 
OP
F

fabiospark

Member
Joined
Dec 4, 2022
Messages
28
Likes
1
EDIT - and I asked the wrong question earlier. I should have been asking about the desktop PC connection to the HPamp, not the NUC/Laptop. Does that connection still exist?
Not at the moment and during this testing, just interface to Mackie - Mackie to Rotel and Rotel to HP ampli.

A moment ago I tried to lift the ground in the USB cable from NUC to interface placing a bit of isolating tape on the first stripe to the left in the USB A plug, but nothing changed.
 

arvidb

Member
Joined
Dec 1, 2022
Messages
37
Likes
35
Location
Sweden
Try to just touch the shell of the USB plug to the NUC and see what happens with the noise. Also, if you're an electrician, use a DVM to check continuity between the NUC case and mains ground to see if the NUC is grounded. Edit: obviously with the USB cable disconnected.
 
Last edited:
Top Bottom