Ground loop breaker and unbalanced connections

[saint gillis]

Yes, the one on the right, but the symbols are not what I'm accustomed to. The triangle would be the audio reference (a.k.a. "ground" or V), the thing that looks like a rake or a broom would be chassis or earth.

Yes, sorry...  And thanks for the lesson!

What about a Hi-Z unbalanced input?  Should we apply the same principle ?

 

[saint gillis]

.

I think there was a mistake on my right hand drawing, here is the correct version :

PRR you show an inverting stage?

 

[Speedskater]

Over two decades ago, Neil Muncy (RIP) wrote about this problem in:
"Noise Susceptibility in Analog and Digital Signal Processing Systems"
AES  June 1995

and yet the discussion continues.

 

[ruffrecords]

Over two decades ago, Neil Muncy (RIP) wrote about this problem in:
"Noise Susceptibility in Analog and Digital Signal Processing Systems"
AES  June 1995

and yet the discussion continues.

But that primarily discusses balanced connections not unbalanced.

Cheers

Ian

 

[abbey road d enfer]

But that primarily discusses balanced connections not unbalanced.

+1. According to Muncy, unbalanced connections should not exist.
Many papers are quite pedantic about the so-called "Pin 1 problem", but fail to give any practical advice about unbalanced connections.
Many mixers have unbalanced connectors hard-wired to the chassis; that leaves the burden of breaking the ground loop to the sources. Turntables are easy because they are usually floating (Class 2), but for other sources, very often there is a conflict between the safety regulations of Class 1 equipment, that imposes that the audio ground is connected to the earth (chassis).
Often, the solution is to make sure the sources are very close to the mixer, or to use dangerous practises, such as lifting the earth.

 

[ruffrecords]

+1. According to Muncy, unbalanced connections should not exist.
Many papers are quite pedantic about the so-called "Pin 1 problem", but fail to give any practical advice about unbalanced connections.

Very true. I have only seen one reasonably good treatment of it. I will see if I can find it. I rememeber it because it advocated adding more 0V copper between items of equipment connected by unbalanced signals on the basis that the lower the impedance the lower the noise voltage (and backed it up with actual measurements IIRC) which is the antithesis of the ground lift approach.

Cheers

Ian

 

[JohnRoberts]

+1. According to Muncy, unbalanced connections should not exist.

For professional audio interfaces between separate chassis, 3 wire (2 circuit differential audio) and a shield, make sense. Inside chassis that many wires are inefficient. 

Many papers are quite pedantic about the so-called "Pin 1 problem", but fail to give any practical advice about unbalanced connections.

Its good that pin 1 problem is now old news, it wasn't always...  :-[

For unbalanced connections (like most hifi poop),, we want to manage circulating ground currents (say between two chassis) to prevent those currents from corrupting audio signals with a voltage drop in the wire resistance.

Many mixers have unbalanced connectors hard-wired to the chassis; that leaves the burden of breaking the ground loop to the sources.

more effective RF rejection that way... Grounds (audio low and shield) on unbalanced sends can be lifted from hard ground by a modest resistance, and then the audio signal differentially referenced to that external audio 0V.

In some cases for short send/returns, like unbalanced console insert points we don't have much choice but to share a common hard ground.  Gear routinely used in those applications know this, or should. 

Turntables are easy because they are usually floating (Class 2), but for other sources, very often there is a conflict between the safety regulations of Class 1 equipment, that imposes that the audio ground is connected to the earth (chassis).
Often, the solution is to make sure the sources are very close to the mixer, or to use dangerous practises, such as lifting the earth.

Audio transformer isolation is less dangerous than lifting safety grounds (but more expensive.)

JR

 

[saint gillis]

Can I apply this same principle for a HI-Z input?

Just using a very hi impedance input op amp and for instance R1=R3=100Ohm and R2=R4=1MOhm .

Or does this work?

 

[JohnRoberts]

With high impedance inputs (like instrument amps) the noise from the high impedance resistors can become a factor.

Perhaps google " instrumentation amps" 

JR

 

[saint gillis]

So you would do something of this kind?

 

[abbey road d enfer]

You may try that; a little less complicated.

 

[saint gillis]

You may try that; a little less complicated.

That's interesting, thank you! Why R17 and C9 to chassis and not to the board's ground?

 

[abbey road d enfer]

That's interesting, thank you! Why R17 and C9 to chassis and not to the board's ground?

That's because the board's "ground" is connected to chassis somewhere. That's needed to make the simulation show what I want to see. Indeed R17/C9 should be connected to 0V on the board

 

[saint gillis]

Ok!

 

[saint gillis]

To add a Hi-Z input to a BJT input transformless microphone preamp I'd like to try something like this, do you think it makes sense?

 

[abbey road d enfer]

To add a Hi-Z input to a BJT input transformless microphone preamp I'd like to try something like this, do you think it makes sense?

It should work!

 

[saint gillis]

Thanks! I'll give a try

 

[Bo Deadly]

To add a Hi-Z input to a BJT input transformless microphone preamp I'd like to try something like this, do you think it makes sense?

This might not be ideal for a guitar. A guitar should have a really good ground connection. If you're connecting another unbalanced piece of gear that has it's own power supply and earth connection, then this circuit might help. But otherwise, the 10R might just be a liability for a floating instrument like guitar.

 

[abbey road d enfer]

This might not be ideal for a guitar. A guitar should have a really good ground connection. If you're connecting another unbalanced piece of gear that has it's own power supply and earth connection, then this circuit might help. But otherwise, the 10R might just be a liability for a floating instrument like guitar.

Electric guitar is sensitive to electrostatic fields because of its high impedance, typically about 100kohms. In order to minimize induced hum and buzz, it must be grounded via a significantly lower impedance. 10 ohms is negligible against 100k; I would thing even 100 ohms would do the job.
Grounding is irrelevant agains magnetic fields.

 

[saint gillis]

So I tried it with a TL071, with a 680n cap between input and R11 to AC couple the input. only I changed R12 from to 100K to 1Meg, I not sure if it is subjective but I feel there was a bit more high-end (tested with a guitar). It works very well and sounds good, it's even strange not to get hum when switching to single coils on a strat.