Isolation Transformer and Balanced Power versus PS Audio Regenerator

[Tranquility Bass]

Hi guys

What is the use of purifying the AC mains voltage when nearly every audio appliance that is connected to it uses some sort of peak rectification circuit either in the primary side or secondary side of the power supply ? The mains current is certainly not sinusoidal no matter how clean the AC mains voltage is. You will find it is this current which is responsible for a lot of ground loop problems in unbalanced interconnects and usually manifests itself as a buzz in the speaker. Like Amir said the use of these mains purifying device might actually make the problem worse ! The solution to all of this is to forget about using these mains "purifiers" and focus on using balanced interconnects on equipment that has a properly implemented balanced output and input stages with high CCMR and then the fidelity of the mains waveform will be totally irrelevant. You'll pay a lot less for balanced interconnects than some uber expensive unbalanced RCA cable which solves none of these problems and usually creates them !!

cheers
AP

 

[Chorus]

I tried a conditioner and it wrecked the sound stage. I have two 20amp dedicated ac lines. My lines come directly from the pole transformer.
I was advised to add a Balanced Power Transformer and did. The result was a lowered noise floor and overall improved sound. If I was to consider
any other conditioning it would be from Shunyata.

 

[Kingrex]

Good question.

I'm expecting a non-sense answer. It seems like the status quo is preferring bigger ground loops.

Is it your belief you should stack all your equipment on a single 10 awg wire and life is good.

Or did you install a #6 wire and believe that somehow made a single 20A circuit better.

It is always better to add circuits. If done right, you should read 0 milivolts between grounds or neutrals between receptacle around the room.

Take a scope and look at the waveform damage your audio equipment creates. It's horrible. Now take all your amps, preamps and digital gear and pile them onto a power strip and ask a single wire to handle that incredibly distorted wave form.

Its an internet lie that 2 or more circuits creates ground loops. It's piss poor wiring by the guy in the field that creates issues with multiple circuits.

 

[Atmasphere]

Hi guys

What is the use of purifying the AC mains voltage when nearly every audio appliance that is connected to it uses some sort of peak rectification circuit either in the primary side or secondary side of the power supply ? The mains current is certainly not sinusoidal no matter how clean the AC mains voltage is. You will find it is this current which is responsible for a lot of ground loop problems in unbalanced interconnects and usually manifests itself as a buzz in the speaker. Like Amir said the use of these mains purifying device might actually make the problem worse ! The solution to all of this is to forget about using these mains "purifiers" and focus on using balanced interconnects on equipment that has a properly implemented balanced output and input stages with high CCMR and then the fidelity of the mains waveform will be totally irrelevant. You'll pay a lot less for balanced interconnects than some uber expensive unbalanced RCA cable which solves none of these problems and usually creates them !!

cheers
AP

We were the first company anywhere to offer balanced line equipment for home audio use (1987) and we made the world's first balanced line preamps. I've been arguing some of your points for decades. So I get what you are saying about CMRR but I can tell you that in order to get the most out of a differential amplifier its beneficial to use a good constant current source (which is a regulator) and good bipolar regulated power supplies. You give the differential circuit the best environment you can and the improvement is easily heard and measured.

We don't regulate the power supplies in our power amps though. Its arguably the most beneficial where the amps are concerned to regulate the B+ supplies in the output section. But to do so introduces reliability issues! So it is beneficial to have clean AC power- the difference again is audible and measurable.

But the simple fact is audiophiles have yet to be convinced that doing everything balanced differential is better! I think this is because there is so much 'balanced' equipment in the high end audio market that does not support the balanced line standard (AES48) and sometimes has terrible performance so people get the mistaken impression that because they are listening to a balanced setup that they are hearing everything balanced offers. They often don't find out that such isn't the case!

Another aspect you didn't mention is the use of high feedback circuits, which can give you rejection of problems caused by poor AC power. Those circuits are not commonly used in high end audio. Usually the feedback (if its there at all) is far too low to be able to reject AC line artifact.

So a good line conditioner is still a good thing.

BTW the statement I highlighted in your post above is false. One benefit of balanced operation is supposed to be elimination of ground loops (which are not caused by distortion in the AC mains current waveform) because in a proper balanced line system ground is ignored. You can prevent ground loops in single-ended equipment by insuring that the chassis, which should be grounded to the AC mains ground, isn't the same as the audio circuit ground. If this is done correctly the AC current waveform can do anything and the audio circuit will not have a ground loop. But if there is noise on the AC line such as the 5th harmonic, that can cause diodes and transformers to misbehave, causing a buzz that sounds like a ground loop.

 

[Alrainbow]

And I wonder who understands your great post lol. Doesn't ps audio make a fully bal lol

 

[wil]

PSA has always touted balanced, no?

 

[Alrainbow]

PSA has always touted balanced, no?

Paul ha s not ever made a true balanced amp or preamp Brito tells us idiots it is. even though it’s only at input then uses ic chips to covert to se but hey you get to use bal interconnects lol. google it Watch Paul lie in videos lol.

 

[Tranquility Bass]

BTW the statement I highlighted in your post above is false. One benefit of balanced operation is supposed to be elimination of ground loops (which are not caused by distortion in the AC mains current waveform) because in a proper balanced line system ground is ignored. You can prevent ground loops in single-ended equipment by insuring that the chassis, which should be grounded to the AC mains ground, isn't the same as the audio circuit ground. If this is done correctly the AC current waveform can do anything and the audio circuit will not have a ground loop. But if there is noise on the AC line such as the 5th harmonic, that can cause diodes and transformers to misbehave, causing a buzz that sounds like a ground loop.

The balanced system doesn't eliminate ground loop currents. The currents still flow around the chassis connected shielding however it is the high CMRR of the balanced input and output stages that suppresses any common mode interference on the two signal lines.

The inherent problem with the unbalanced system is the way it shares the safety ground with the signal ground. By separating the signal ground from the safety ground within the chassis the shielding on the unbalanced cable becomes ineffective at suppressing RF interference and allows RF noise to be injected back into the equipment unless there is a low impedance path to ground for RF. Some designs use some tricks by lifting the signal ground from the safety ground by connecting a low value resistor and capacitor to chassis ground etc. And when you have a stereo or multi-channel pair sharing the same signal ground inside the chassis and using separate Left and Right unbalanced cables that are not co-located together this can also setup a ground loop that will pick up the AC current (usually from a mains transformer) and create the buzz again. Best to use a properly designed balanced system right from the start.

cheers
AP

 

[kodomo]

I own a Torus avr30. I have moved my house and the new place although very scenic has problems with the electricity. It is not pollution, it is complete black outs! I am now planning to get an online UPS for the whole house. I found one with less than thd 1.5% output. If I add my Torus isolation transformer at the output of the ups (input of my listening room), would that be good enough?

*As a note the UPS will have more capacity than the whole house needs.

 

[Atmasphere]

The balanced system doesn't eliminate ground loop currents. The currents still flow around the chassis connected shielding however it is the high CMRR of the balanced input and output stages that suppresses any common mode interference on the two signal lines.

The inherent problem with the unbalanced system is the way it shares the safety ground with the signal ground. By separating the signal ground from the safety ground within the chassis the shielding on the unbalanced cable becomes ineffective at suppressing RF interference and allows RF noise to be injected back into the equipment unless there is a low impedance path to ground for RF. Some designs use some tricks by lifting the signal ground from the safety ground by connecting a low value resistor and capacitor to chassis ground etc. And when you have a stereo or multi-channel pair sharing the same signal ground inside the chassis and using separate Left and Right unbalanced cables that are not co-located together this can also setup a ground loop that will pick up the AC current (usually from a mains transformer) and create the buzz again. Best to use a properly designed balanced system right from the start.

cheers
AP

The sentence is true. But the second sentence is not. The reason balanced lines prevent ground loops is because ground is ignored, meaning that there is no signal current imposed on the ground (like you see in a single-ended circuit). A good CMRR value certainly helps!

I agree that doing balanced right from the start is the way to go- we made the first balanced amps and preamps for home use- we've been at it a long time.

 

[Folsom]

Is it your belief you should stack all your equipment on a single 10 awg wire and life is good.

Or did you install a #6 wire and believe that somehow made a single 20A circuit better.

It is always better to add circuits. If done right, you should read 0 milivolts between grounds or neutrals between receptacle around the room.

Take a scope and look at the waveform damage your audio equipment creates. It's horrible. Now take all your amps, preamps and digital gear and pile them onto a power strip and ask a single wire to handle that incredibly distorted wave form.

Its an internet lie that 2 or more circuits creates ground loops. It's piss poor wiring by the guy in the field that creates issues with multiple circuits.

Electricity is multiplication and division. There’s no add and subtract. If the path exists then something on it will be present. You will have loops so long as the equipment is safety grounded. They may not be super obvious and super bad, but they exist.

What really comes into play isn’t just their hash they put on the line but how much of that stuff rings. Having multiple lines is a double edged sword because the higher resistance helps but distance does not. You’re basically trading, but which one is worse?

If you are truly worried about what you put back onto the line then use filtration. Cut down the common mode and differential mode noise with attenuation, not hokus pokus ideas about being able to “drain” the noise (that actually increase the collection loop of noise).

 

[wil]

The sentence is true. But the second sentence is not. The reason balanced lines prevent ground loops is because ground is ignored, meaning that there is no signal current imposed on the ground (like you see in a single-ended circuit). A good CMRR value certainly helps!

I agree that doing balanced right from the start is the way to go- we made the first balanced amps and preamps for home use- we've been at it a long time.

@Atmasphere Is "truly balanced" synonymous with "differentially balanced"? And must a balanced audio system comply with the AES48 to truly balanced?

Can one test with a Multimeter? And if you get resistance between any two pins on the XLR (that is, any reading at all) does that mean the XLR lines are connected to ground and hence, Un-Balanced?

If that is the case, would there be no Common Mode Rejection and the XLR connections would serve no purpose other than another way to run an interconnect?

 

[Alrainbow]

There are so many non audio terms used in audio I get lost lol.
in simple terms bal means no signal is grounded and had a POS and negative path not grounded . But it's more complex then just my words in pure truth is what is meant . Paul just says a chip to convert a bal signal to se in a non truly bal amp or pre amp is bal. Wow just cut out the complete doubling of all the components needed saves him money. But this still means he right ?
You cant use a meter but let me say this. A mono block se amp compared to a truly non Paul amp that's balanced you need one se stereo amp to make one. Complete one all parts. Man the one chip really must be great lol.
Now I'm not saying bal is better but it has a double size voltage swing while a also a double noise figure too.

 

[Alrainbow]

Is "truly balanced" synonymous with "differentially balanced"? And must a balanced audio system comply with the AES48 to truly balanced?

Can one test with a Multimeter? And if you get resistance between any two pins on the XLR (that is, any reading at all) does that mean the XLR lines are connected to ground and hence, Un-Balanced?

If that is the case, would there be no Common Mode Rejection and the XLR connections would serve no purpose other than another way to run an interconnect?

A ic chip can be used to lower common mode noise on interconnects . This is not the point of my posts

 

[Atmasphere]

@Atmasphere Is "truly balanced" synonymous with "differentially balanced"? And must a balanced audio system comply with the AES48 to truly balanced?

Can one test with a Multimeter? And if you get resistance between any two pins on the XLR (that is, any reading at all) does that mean the XLR lines are connected to ground and hence, Un-Balanced?

If that is the case, would there be no Common Mode Rejection and the XLR connections would serve no purpose other than another way to run an interconnect?

You can have balanced operation and not support AES48.

The traditional means of driving a balanced line is with an output transformer that has one side of the output winding tied to pin 2 and the other side to pin 3. There's no center tap; the winding floats with respect to ground. In this manner you can see that the pin 2 signal is generated with respect to pin 3 and vice versa. Ground is merely the chassis.

It gets tricky if a transformer isn't involved. In this circumstance the circuit driving the balanced line has to have equal impedance to ground (pin 1) from the two outputs (pins 2 & 3). The tricky bit is you can't just use dual single ended outputs that are out of phase with each other. What happens then is each output is using the shield of the interconnect as a return path and this can open up the circuit, even though its balanced, to ground loops (it does not take much of a mismatch of components for this to happen!). However, if the receiving end has a very high Common Mode Rejection Ratio you stand some chance of minimizing the ground loop noise. In practice, you can't count on that: some inputs do not employ an input transformer and might have a low CMRR simply out of design fault. As a result there really aren't that many ways to drive a balanced line correctly because one output must be able to sink current from the other output and vice versa; ground only being a shield. Some might see this as a bit of a nuance but the proof is in the pudding. It won't wash if you run balanced lines everywhere and still wind up with a ground loop! That's not supposed to happen with balanced line equipment.

 

[Alrainbow]

See you always teach us all. Ty for this.
i have both se and balanced setups I can’t say one is better but if you need longer interconnects all bets are off. You go a good distance using a good bal cable keeping both neg n pos inside the twisted set grounding only on one side.
i truly me my comments as honest praise to you. just in case some may think I might being disrespectful I’m am not

 

[wil]

You can have balanced operation and not support AES48.

The traditional means of driving a balanced line is with an output transformer that has one side of the output winding tied to pin 2 and the other side to pin 3. There's no center tap; the winding floats with respect to ground. In this manner you can see that the pin 2 signal is generated with respect to pin 3 and vice versa. Ground is merely the chassis.

It gets tricky if a transformer isn't involved. In this circumstance the circuit driving the balanced line has to have equal impedance to ground (pin 1) from the two outputs (pins 2 & 3). The tricky bit is you can't just use dual single ended outputs that are out of phase with each other. What happens then is each output is using the shield of the interconnect as a return path and this can open up the circuit, even though its balanced, to ground loops (it does not take much of a mismatch of components for this to happen!). However, if the receiving end has a very high Common Mode Rejection Ratio you stand some chance of minimizing the ground loop noise. In practice, you can't count on that: some inputs do not employ an input transformer and might have a low CMRR simply out of design fault. As a result there really aren't that many ways to drive a balanced line correctly because one output must be able to sink current from the other output and vice versa; ground only being a shield. Some might see this as a bit of a nuance but the proof is in the pudding. It won't wash if you run balanced lines everywhere and still wind up with a ground loop! That's not supposed to happen with balanced line equipment.

So, if there is resistance between pins when measured with multimeter does that mean the circuit is Un-balanced? Or, as in the explanation above, maybe a suspect and problematic Balanced operation?

 

[Tranquility Bass]

The sentence is true. But the second sentence is not. The reason balanced lines prevent ground loops is because ground is ignored, meaning that there is no signal current imposed on the ground (like you see in a single-ended circuit). A good CMRR value certainly helps!

I agree that doing balanced right from the start is the way to go- we made the first balanced amps and preamps for home use- we've been at it a long time.

I wasn't talking about signal current. I am talking about ground loop induced currents which have nothing to do with the signal. And it doesn't matter if the output is referenced to ground. A balanced system will still work perfectly even if one of the signal wires is connected to ground. It just will by definition since the proceeding input stage only reacts to the difference in potential. Mind you the output impedances of both hot and cold should still be indentical for proper balanced operation

cheers
AP

 

[Kingrex]

I wasn't talking about signal current. I am talking about ground loop induced currents which have nothing to do with the signal. And it doesn't matter if the output is referenced to ground. A balanced system will still work perfectly even if one of the signal wires is connected to ground. It just will by definition since the proceeding input stage only reacts to the difference in potential. Mind you the output impedances of both hot and cold should still be indentical for proper balanced operation

cheers
AP

How do you have a ground loop without a signal wire?

 

[Kingrex]

Electricity is multiplication and division. There’s no add and subtract. If the path exists then something on it will be present. You will have loops so long as the equipment is safety grounded. They may not be super obvious and super bad, but they exist.

What really comes into play isn’t just their hash they put on the line but how much of that stuff rings. Having multiple lines is a double edged sword because the higher resistance helps but distance does not. You’re basically trading, but which one is worse?

If you are truly worried about what you put back onto the line then use filtration. Cut down the common mode and differential mode noise with attenuation, not hokus pokus ideas about being able to “drain” the noise (that actually increase the collection loop of noise).

Sounds flowery and nice but does not play out as such in the real world. Use multiple circuits.

Filters can help. They can also hinder.
I have never seen a filter help a ground loop. A device like an Isomax can. But a isolation transformer in the signal path will change the sound.

You can also use a device like a HumX that is back to back diodes in the power ground. That will break a ground loop.

Al told me years ago, try Mogami cable for interconnects. I am using a Mogami microphone cable as a single ended interconnect. The shield is only terminated on the source end. The cable is 37 feet long. I could not hear a difference between it and other 2 to 5 foot interconnect in my system. Or its so close I am not concerned. It even has a lamp cord on a dimmer laying parallel to it for about 4 feet.