Balanced interconnection and ground

[Geert]

That's in line with the diagram I posted above. For low frequencies your connection is like floating at both ends. Insufficient shielding against mains frequency. Input pin 1 to ground will work in this case since the source is floating (as long as you don't connected other unbalanced sources to the Yamaha).

 

[GXAlan]

That's in line with the diagram I posted above. For low frequencies your connection is like floating at both ends. Insufficient shielding against mains frequency. Input pin 1 to ground will work in this case since the source is floating (as long as you don't connected other unbalanced sources to the Yamaha).

Got it. What is the benefit of grounding the amp versus just connecting the ground lug of the Yamaha to the Earth ground so that both the amp and source share the same earth ground?

 

[Geert]

Connecting the source end to ground creates a ground loop, theoretically and in this case only for high frequencies so it won't matter. If you don't experience any audible issues, then it's fine as it is.

 

[Sokel]

Looking at the Meyer Sound Amie manual:

View attachment 343530

Are you sure that the internal connection of your cable is connected to XRL's casing (the forth upper pin) ?
A lot of them don't.

 

[nagster]

My Yamaha CX-A5100 (two prong power) via XLR to my Meyer Sound Amie (three prong power) generates a ground loop that’s audible and annoying even though it’s a presumably a proper balanced connection as both companies make PA gear.My Yamaha CX-A5100 (two prong power) via XLR to my Meyer Sound Amie (three prong power) generates a ground loop that’s audible and annoying even though it’s a presumably a proper balanced connection as both companies make PA gear.

I can remove the hum by taking the phono lug of the Yamaha to the ground lug of my line conditioner. The ground lug of my line conditioner is attached to earth safety ground.

Ideas? I have tried a few XLR cables. It’s not a big deal to use the grounding wire, but I am curious. Panamax, Belkin, Monster Power all have ground lugs on their power conditioners (with Panamax being the only one still making power conditioners)

When connecting XLR cables with floating shells to Meyer Sound powered speakers, noise will often occur.
One easy solution is to short the speaker's XLR terminal shell and pin 1. I have several Image plugs, and if I don't have time, I plug them into the loop out terminals of speakers.

 

[ldarieut]

I use these:

https://www.neutrik.com/en/neutrik/products/xlr-connectors/xlr-chassis-connectors/a-series?c=audio

and more precisely, this one:
Female connector
-1 = mating connector shell to pin1 and front panel

This way, I never had ground loop issue: pin1 connects to the shell and the chassis with the metal screw.

Using these:

https://www.neutrik.com/en/neutrik/products/xlr-connectors/xlr-chassis-connectors/d-series?c=audio

You have to solder a tiny wire from pin 1 to the ground latch, which is connected to the shell that connects to the chassis. If you don't do this, you will probably hear a hum.
Worst case is when you do only half of the above: better not ground your XLR socket at all, than doing half the job.

 

[RHO]

I'm planning on building new balanced cables for my new set-up. Currently everything is connected single ended.
I've measured GND connections on the DAC (SMSL RAW-MDA1) and powered sub (KRK S10.4). On both I measure chassis ground to be the connected to signal ground on the SE outputs/inputs. Chassis ground is also connected to pin1 on the XLR outputs/inputs.
I measured this with a simple multimeter. It looks like the connection from pin1 to chassis is just a little more low-ohmic then the connection to signal gnd from the SE connections. I will not open the devices since they are still under warranty.
So it could be that the pin1 from the balanced terminals is connected to chassis very close to the terminal and signal gnd a bit further away, on the PCB. Would this be the correct way for them to do it?
I think I can connect pin1 on both sides since there is a "gnd lift" switch on the receiving side (subwoofer).
Initially I will go SE out from the sub to an amp, that's powering my speakers. This will eventually be replaced by active monitors (with balanced inputs).

I'm currently struggling with digital noise from a pc on the analog audio signal.
There was no noise when the sub was not between the DAC and speaker amp. It got introduced by adding the sub in the signal path.
The noise only appears when that specific PC comes out of sleep mode. From the moment it goes back in sleep mode the noise is gone. (there are several laptops running in this room, creating no such issues)
Hoping to solve this with the balanced connections.

 

[antcollinet]

So it could be that the pin1 from the balanced terminals is connected to chassis very close to the terminal and signal gnd a bit further away, on the PCB. Would this be the correct way for them to do it?

With balanced IO there is no signal ground as such. Pin 1 is still the reference - or should be connected to the 0V reference on the PCB via the chassis.

Pins 2 and 3 carry the hot and cold balanced signals, which are the inverse of each other.

Yes you should connect pin 1 both ends. With balanced there should be no need for the ground lift.

I'm currently struggling with digital noise from a pc on the analog audio signal.
There was no noise when the sub was not between the DAC and speaker amp. It got introduced by adding the sub in the signal path.

That might be due to the sub being earthed (if the amp is not). If that is the case then the balanced cable to the sub should help.

However it could also be just due to the long cables you'd need to go DAC to Sub and back to amp. In which case you might still get noise on the amp when it is connected unbalanced to the sub.

The best solution to this (if you get it) would be to use an isolated connection from PC to DAC. The best of these is Toslink, and if you don't have a toslink out from your PC, inexpensive USB to Toslink DDCs are available.

Some people don't like Toslink because they think it can't do hires audio. Rest assured. There is no benefit from going higher than Toslink can easily achieve.

 

[RHO]

With balanced IO there is no signal ground as such. Pin 1 is still the reference - or should be connected to the 0V reference on the PCB via the chassis.

Pins 2 and 3 carry the hot and cold balanced signals, which are the inverse of each other.

Yes you should connect pin 1 both ends. With balanced there should be no need for the ground lift.

I understand. It's that the chassis/pin1/shield is also connected to signal ground from the SE connections. And since I have those (SE connections) in use now with the digital noise issue I'm a bit afraid the problem will persists when I switch to balanced connections. When disconnecting the shield/pin1/chassis om 1 side it could open the ground loop. At least that was my thinking. But that could depend on how the signal ground is connected to chassis/pin1 inside the device. Right?

That might be due to the sub being earthed (if the amp is not). If that is the case then the balanced cable to the sub should help.

However it could also be just due to the long cables you'd need to go DAC to Sub and back to amp. In which case you might still get noise on the amp when it is connected unbalanced to the sub.

That's my fear. But ultimately the SE amp will also be swapped for active monitors with balanced inputs.
The external supply for the amp is grounded.

The quality and length of the current SE cables could indeed be an issue too.

The best solution to this (if you get it) would be to use an isolated connection from PC to DAC. The best of these is Toslink, and if you don't have a toslink out from your PC, inexpensive USB to Toslink DDCs are available.

Some people don't like Toslink because they think it can't do hires audio. Rest assured. There is no benefit from going higher than Toslink can easily achieve.

The noisy PC is connected to the DAC by optical cable.
The laptops are connected via a USB switch. The DVD player via digital coaxial. No issues with high-res audio.

 

[antcollinet]

And since I have those (SE connections) in use now with the digital noise issue I'm a bit afraid the problem will persists when I switch to balanced connections.

It won't. The whole point of a balanced connection is not that it does't have ground noise, it is that it doesn't care. The noise is added equally to the hot and cold signals then when they are subtracted from one another at the input to your amp, it cancels out.

The noisy PC is connected to the DAC by optical cable.

Then that is unlikely to be the noise source. Unless noise is going from it to the mains socket, and then from there backup through the DAC and down the analogue interconnect to your sub/amp earth connections.

(I'm assuming the noise is ground noise at the moment)

 

[RHO]

It won't. The whole point of a balanced connection is not that it does't have ground noise, it is that it doesn't care. The noise is added equally to the hot and cold signals then when they are subtracted from one another at the input to your amp, it cancels out.

But the noise is carried from the one device to the next IF signal GND is also connected to the shield. Noise put on the signal wires wil be canceled out. But the way the ground seems to be connected noise on the shield will be transferred to signal ground.
If, eventually all connections become balanced, this should indeed solve the problem. But as long as the connection to the speaker-amp is SE I'm not certain.

Then that is unlikely to be the noise source. Unless noise is going from it to the mains socket, and then from there backup through the DAC and down the analogue interconnect to your sub/amp earth connections.

When the sub is removed from the chain, there is no digital noise. So I think it's the signal ground that is picking it up trough the RCA cable. Either the one running from the sub to the speaker amp or the one from the DAC to the sub.
When I had the DAC connected to the speaker amp directly there was no audible noise from the PC. The cables were much shorter (further away from the PC) and higher quality.
I'm thinking about running the "long" RCA cable from the DAC to the speaker amp, next to the PC, removing the sub from the equation. Just to hear if it is the cable.
I suspect that it is the sub or the cable connecting the sub to the speaker amp, since I don't think there is any noise coming out of the sub itself. (need to check that again)

(I'm assuming the noise is ground noise at the moment)

I think so too. Somewhere picked up by the signal ground going from sub(-amp) to the speaker amp. But like I said, signal- and chassis-ground and shield seem to be all connected, which I think creates the problem when using SE connections.

 

[antcollinet]

If, eventually all connections become balanced, this should indeed solve the problem. But as long as the connection to the speaker-amp is SE I'm not certain.

Correct - in this case, the signal ground noise is transferred via the balanced connection and onto the sub. From there, it will be on the signal ground of the unbalanced connection to the amp. This is why I made my point in post number 96 that you may still get noise going to the amp.

It is possible that you won't, IF the noise current is going through a ground loop via the mains ground AND the sub has a ground connection to mains. In this case it is possible the ground current from DAC to Sub may be drained to ground (or some of it may) and not make its way along the unbalanced interconnect to cause problems there.

If you still get noise after going balanced to the sub, then you need to find the source of that noise and - if it is possible - isolate it.

 

[RHO]

It is possible that you won't, IF the noise current is going through a ground loop via the mains ground AND the sub has a ground connection to mains. In this case it is possible the ground current from DAC to Sub may be drained to ground (or some of it may) and not make its way along the unbalanced interconnect to cause problems there.

The sub indeed has mains ground. It is connected to chassis, as it should. Let's see what the results are when I go balanced from DAC to sub.

If you still get noise after going balanced to the sub, then you need to find the source of that noise and - if it is possible - isolate it.

Going balanced to active speakers should improve things even further.
But first I'll build me some nice balanced cables.

Thanks for the help!

 

[Speedskater]

We are co-mingling:

a] Safety Ground/Protective Earth
b] audio circuit common
c] metal chassis
d] shield

It's very confusing.

 

[RHO]

We are co-mingling:

a] Safety Ground/Protective Earth
b] audio circuit common
c] metal chassis
d] shield

It's very confusing.

What's confusing?
They are all the same.

 

[somebodyelse]

Pin 1 is still the reference - or should be connected to the 0V reference on the PCB via the chassis.

Pin 1 should not be the reference for a balanced input. If it is then you're doing it wrong. The signal is the difference between hot and cold, not referenced to pin 1. The interconnect should still work if the shield is disconnected at one end (the receiving end for best performance) - this is one way to break a ground loop if the receiving end has a 'Pin 1 Problem'. Pin 1 should be connected to the chassis as close as practical to the connector per AES48.

What's confusing?
They are all the same.

They aren't, and that confusion is one of the reasons why people keep making equipment vulnerable to ground related noise issues.

 

[antcollinet]

We are co-mingling:

a] Safety Ground/Protective Earth
b] audio circuit common
c] metal chassis
d] shield

It's very confusing.

Sort of - but mostly they are connected together.

Protective earth (on class I devices) is connected to chassis.
Audio circuit common (audio ground) should be connected to chassis
As should cable shield.
(AES 48)

While (audio) ground loops mostly involve 2 or more connections back to PE, they can occur without any, if there are looping interconnect. (EG Audio from audio interface out to amp plus recording back to interface.

(obviously there are some exceptions such as when the chassis is not metal, but then there should be an internal star point for everything to come together)

 

[RHO]

They aren't, and that confusion is one of the reasons why people keep making equipment vulnerable to ground related noise issues.

I'm very well aware that they shouldn't be. But in my situation they are. (They are all "shorted". They have a very low ohmic connection to each other.)

 

[sam_adams]

From.

 

[RHO]

Pin 1 should not be the reference for a balanced input. If it is then you're doing it wrong. The signal is the difference between hot and cold, not referenced to pin 1. The interconnect should still work if the shield is disconnected at one end (the receiving end for best performance) - this is one way to break a ground loop if the receiving end has a 'Pin 1 Problem'. Pin 1 should be connected to the chassis as close as practical to the connector per AES48.

That's why I assume the ground lift switch could help.
But also the way they are connected inside the device could matter. This I cannot check.
My multimeter isn't accurate enough to see the difference between a very short connection between pin1 and chassis, and a longer one connecting signal ground on the pcb to chassis.