Isolation Transformer and Balanced Power versus PS Audio Regenerator

[Kingrex]

Rex, this is a mirror image of my conditions. Yes, you’re correct about the phono stage(remember I contacted you about this awhile back and since have solved the problem.) We have a lot of thunderstorms and lightning here in the mid south so there’s whole house surge protection. I also have Jim Weil’s Sound Application in the system.

Thanks Dave, I remember. I have had quite a few people call and want to get rid of radio or other noise in their phono. I took a Torus to one guys house along with 2 phono stages I know to be quiet. I first put the Torus in but said its not going to help. And it did not help with the radio out the speaker. But it made the soundstage expand and lent a quiet to the room.
I then inserted My Lino and my Allnic phono stages and the noise went away. I tried to get this person to change some cables but he refused.

I mention the cable as another person called with the same issue. I pointed him to some mylar to buy as well as dead soft 22 awg silver wire from Rio Grande jewelry. Then told him to sleeve the cable from his TT to phono stage in it. The guy was shocked it eliminated all the radio noise from his system.

 

[Amir]

By Art Kelm

After working over 40 years in the professional recording industry and the past 10 years of that involved in the high end A/V market, I have lost track of the number of times I have heard the term "clean power" used to describe someone's power system.
From my personal experience, I know that power and grounding is the heart of any electronic system.

+1

I believe all debates like "tubes vs Solidstates" or "push pull vs SET" all are far less important than AC Power quality.
My TAD M700 Power amplifier is 350w/ch push pull low current feedback solidstate amplifier but when the AC quality is perfect the harmonics/tone is so much emotional like good tubes.

 

[Amir]

Kingrex
I am not against Torus but my general experience shows most audio companies do their best for degrading sound quality specially in killing dynamics for having better black background.

for example most vibration isolation products or most AC filters or most AC cables are not good for sound. David in Utah have lots of experience in this regard and he believes many products are not good for sound.
I have lost so much money for buying useless audio equipments since 2000.

I did not have Torus but I any audiophile product make me scare about the result.

 

[Atmasphere]

Would you please "unpack" this for us?

What was the measured electrical effect of the Everest?

What was the sonic effect of the Everest?

@Kingrex 's measurements point to a 5th harmonic, which is well-known for causing problems with power transformers (mechanical noise, greater heat, lower and distorted output voltages) and rectifiers (noisier), as well as syncronous motors (less torque, higher operating temperature, possibly also starting in the wrong direction in extreme cases), which audiophiles like to use. It is arguably the most pesky harmonic.

To be clear, to anyone using an isolation transformer, whether it be balanced or not: if you run the transformer past about 50% of its rated capacity, the 5th harmonic will start to show up in significant amounts. So if you really want that isolation transformer to bring benefit rather than tradeoffs or outright degradation, make sure you run it at less than 50% of capacity.

 

[Kingrex]

@Kingrex 's measurements point to a 5th harmonic, which is well-known for causing problems with power transformers (mechanical noise, greater heat, lower and distorted output voltages) and rectifiers (noisier), as well as syncronous motors (less torque, higher operating temperature, possibly also starting in the wrong direction in extreme cases), which audiophiles like to use. It is arguably the most pesky harmonic.

To be clear, to anyone using an isolation transformer, whether it be balanced or not: if you run the transformer past about 50% of its rated capacity, the 5th harmonic will start to show up in significant amounts. So if you really want that isolation transformer to bring benefit rather than tradeoffs or outright degradation, make sure you run it at less than 50% of capacity.

Interesting. But it was the Everest dumping noise onto the line. There is no Torus at that project yet. Work starts in about 2 months. I have seen the same thing with other filters. A furman was so bad, the circuit breaker was buzzing in the panel. We removed the Furman and all was quiet.

Why is it these type of filters poor so much noise back into the power.

I will keep it in mind loading a Torus. I pretty much send every project to Torus to size the system. I am starting to size my own now.

When I measure current with an isolation transformer, its usually maybe 4 to 6 amps max. There may be momentary peeks my fluke is not quick enough to read. I'm usually at 10% to 20%.

Commercially we loaded them to 70% and they ran happy as could be for 30 to 40 years.

 

[Onepoint5]

Had some time to play with the various scenarios in the front of the Equitech 1RQ today.
The effects of inserting a Topaz (isolation transformer) in front of the Equitech as recommended by Equitech. The Topaz output is balanced. Drawing attached hopefully makes the text easier to understand.

SQ judgement

Type A test
Bass was very wooly and lacking in control
Treble, veiled and tends to smear
Loss of front to back soundstage
When the 1RQ was initially energised, there's a loud buzz of current due rich harmonic content. The buzz settled down after the audio components switched on ready for operation.
I was hopeful to use the Topaz to provide a level of surge protection, but it hampers SQ significantly. So will have to use sacrificial surge devices at the main panel instead.

Type B test
Bass and Treble returns a lot clearer, bass returns to controlled.
The 20m cable is a 1.5mm extension cable. I have had dedicated lines before, but this is a new location for the audio gear, just wanted to observe a few things with this method.

Type C test
Here the AC input cable for the equitech was changed to a shielded power cable, generic industrial type. The soundstage returned to acceptable, not outstanding level, can live with this temporarily. Connected to the wall socket, there's no energisation buzz from the 1RQ, since the connection to the main panel is very low impedance, that is of the network itself.

Conclusion

1. The added impedance of the Topaz has adverse effects such as higher voltage drops when diodes are conducting and the continuous opposition to rise of currents cause dynamics or even regular playing to be clamped. To dimension a transformer so it avoids these impedances, would confirm experience as posted by @Amir a very large transformer is required.

2. Using two transformers such as different models of Equitech or Torus, both centre points of the secondaries must be connected together at one point. This will avoid different ground potentials when connecting audio equipment say pre-amp to amplifier.

3. The cable from the dedicated breaker in the panel to the audio equipment receptacle needs to be shielded and rated to cater for the power amp's current demand divided by 0.25.

I was trying to re-use the Topaz, since the price on a Torus RM unit is substantial. So there's not much choice now. Would still prefer the one unit over a pair of different isolation transformers.

 

[Onepoint5]

Interesting. But it was the Everest dumping noise onto the line. There is no Torus at that project yet. Work starts in about 2 months. I have seen the same thing with other filters. A furman was so bad, the circuit breaker was buzzing in the panel. We removed the Furman and all was quiet.

Why is it these type of filters poor so much noise back into the power.

I will keep it in mind loading a Torus. I pretty much send every project to Torus to size the system. I am starting to size my own now.

When I measure current with an isolation transformer, its usually maybe 4 to 6 amps max. There may be momentary peeks my fluke is not quick enough to read. I'm usually at 10% to 20%.

Commercially we loaded them to 70% and they ran happy as could be for 30 to 40 years.

Some filters are absorbing, others are reflective. The buzzing on the breaker could have been caused by the Furman absorbing too many harmonics and it is trying to filter the house and next door's install as well. If you put your tong tester on, it would be reading fine, but put on a scope like the Fluke 43 and you can see which harmonic is 'overloading' the breaker. The breaker may trip sooner than expected.
I find Furman power conditioners good for printers and computers and not for audio equipment.

 

[LL21]

Interesting. But it was the Everest dumping noise onto the line. There is no Torus at that project yet. Work starts in about 2 months. I have seen the same thing with other filters. A furman was so bad, the circuit breaker was buzzing in the panel. We removed the Furman and all was quiet.

Why is it these type of filters poor so much noise back into the power.

I will keep it in mind loading a Torus. I pretty much send every project to Torus to size the system. I am starting to size my own now.

When I measure current with an isolation transformer, its usually maybe 4 to 6 amps max. There may be momentary peeks my fluke is not quick enough to read. I'm usually at 10% to 20%.

Commercially we loaded them to 70% and they ran happy as could be for 30 to 40 years.

Good to know...on our Torus it reads a stable use of 5.0 amps on the reading, and the unit is designed for stable/steady to 16amps (hence model name Torus AVR16UK). That 5.0amp reading is also consistent with the calculation from the manufacturer about what the system draw should be. I have seen 5.5a even 6a but very very rarely.

 

[Kingrex]

I'm always glad when Atmasphere chimes in. It made me.go back and look at a project calculation Torus did reciently and now I see why they want me using a 10 kVA instead lf 6kVA. The amps are so massive, they each need their own core. Both on one would be bumping up to that 50%.

 

[dcathro]

Had some time to play with the various scenarios in the front of the Equitech 1RQ today.
The effects of inserting a Topaz (isolation transformer) in front of the Equitech as recommended by Equitech. The Topaz output is balanced. Drawing attached hopefully makes the text easier to understand.

SQ judgement

Type A test
Bass was very wooly and lacking in control
Treble, veiled and tends to smear
Loss of front to back soundstage
When the 1RQ was initially energised, there's a loud buzz of current due rich harmonic content. The buzz settled down after the audio components switched on ready for operation.
I was hopeful to use the Topaz to provide a level of surge protection, but it hampers SQ significantly. So will have to use sacrificial surge devices at the main panel instead.

Type B test
Bass and Treble returns a lot clearer, bass returns to controlled.
The 20m cable is a 1.5mm extension cable. I have had dedicated lines before, but this is a new location for the audio gear, just wanted to observe a few things with this method.

Type C test
Here the AC input cable for the equitech was changed to a shielded power cable, generic industrial type. The soundstage returned to acceptable, not outstanding level, can live with this temporarily. Connected to the wall socket, there's no energisation buzz from the 1RQ, since the connection to the main panel is very low impedance, that is of the network itself.

Conclusion

1. The added impedance of the Topaz has adverse effects such as higher voltage drops when diodes are conducting and the continuous opposition to rise of currents cause dynamics or even regular playing to be clamped. To dimension a transformer so it avoids these impedances, would confirm experience as posted by @Amir a very large transformer is required.

2. Using two transformers such as different models of Equitech or Torus, both centre points of the secondaries must be connected together at one point. This will avoid different ground potentials when connecting audio equipment say pre-amp to amplifier.

3. The cable from the dedicated breaker in the panel to the audio equipment receptacle needs to be shielded and rated to cater for the power amp's current demand divided by 0.25.

I was trying to re-use the Topaz, since the price on a Torus RM unit is substantial. So there's not much choice now. Would still prefer the one unit over a pair of different isolation transformers.

Hi ,

What was the size in KVA of the Topaz?

Also, what is the power rating of the amps and the efficiency of the speakers?

Just curious as to the power requirements of a system having a bearing on the KVA requirement for an isolation transformer.

My speakers are 94db efficient with an 8ohm load, driven by a 150W stereo amp, so I am probably not drawing that much current through my isolation transformer (rated at 5KVA).

 

[Onepoint5]

Hi ,

What was the size in KVA of the Topaz?

Also, what is the power rating of the amps and the efficiency of the speakers?

Just curious as to the power requirements of a system having a bearing on the KVA requirement for an isolation transformer.

My speakers are 94db efficient with an 8ohm load, driven by a 150W stereo amp, so I am probably not drawing that much current through my isolation transformer (rated at 5KVA).

Topaz is 2000VA, as per drawing.
Power rating of the amp is 250W/ch, speakers are 90db efficient. Maximum current input on the AC line to the Equitech didn't exceed 1.5A rms @ 233V line to line, 116.6 any line to ground. This level was very loud, don't normally play at this level.

The optimum for an Equitech or a Torus with balanced output is the symmetry of the secondary windings. To maintain that symmetry, very good regulation is required, better than 2%. The trade off is the transformer ends up being twice the size and weight compared to a standard transformer with 5% or higher regulation of the same electrical rating.

 

[Onepoint5]

Kingrex
I am not against Torus but my general experience shows most audio companies do their best for degrading sound quality specially in killing dynamics for having better black background.

for example most vibration isolation products or most AC filters or most AC cables are not good for sound. David in Utah have lots of experience in this regard and he believes many products are not good for sound.
I have lost so much money for buying useless audio equipments since 2000.

I did not have Torus but I any audiophile product make me scare about the result.

The filtering action of a Torus especially, is that it can filter noise above 50/60Hz, this is their NBT branding.. This can be due to the construction of the core, the type of windings, how they are wound, size of wire. Canadians are very good at creating filters with magnetics in transformers. Mirus for example, create a broadband harmonic filter (for industrial applications) up to the 50th harmonic, by using a single transformer and many capacitors. Classic methods to remove harmonics was to build a 5th, 7th, 11th etc. filter as a standalone unit, very messy and expensive to wire. The Mirus does it all in one go.

I have considered the UPS, but it's distortion is too high (5% to 2%) compared to the raw value from the street.

A standard industrial isolation transformer passes noise from primary to secondary very well, unless it has the shield between windings and very, very low capacitance between the windings. Added engineering requires the use of EMC filters to remove differential and common mode noise. By the time all that work in selecting and purchasing these components, it's easier to implement a Torus instead.

 

[Atmasphere]

The optimum for an Equitech or a Torus with balanced output is the symmetry of the secondary windings. To maintain that symmetry, very good regulation is required, better than 2%. The trade off is the transformer ends up being twice the size and weight compared to a standard transformer with 5% or higher regulation of the same electrical rating.

In audio, a true balanced connection does not have a center tap. That is because the CT cannot be exactly placed and so common mode rejection ratio will always be degraded by its use. So balanced connections don't use them.

It strikes me that placing a CT on an isolation transformer will introduce a similar problem. Ideally, it would seem that there should be no current in the CT at all (for example if tied to ground) but in practice (if you were to place an ammeter in series with the CT) I think you'll find there always is.

The real problem here is that grounding in audio equipment is poorly understood by designers. As a result, and in particular if you use single-ended equipment, grounding is pretty much the Wild West. IME its best to fix those problems at their source. In that way the AC ground connection can do its job by not being noisy- this in turn will allow the chassis (if there is one) to be a silent shield.

 

[Amir]

The filtering action of a Torus especially, is that it can filter noise above 50/60Hz, this is their NBT branding.. This can be due to the construction of the core, the type of windings, how they are wound, size of wire. Canadians are very good at creating filters with magnetics in transformers. Mirus for example, create a broadband harmonic filter (for industrial applications) up to the 50th harmonic, by using a single transformer and many capacitors. Classic methods to remove harmonics was to build a 5th, 7th, 11th etc. filter as a standalone unit, very messy and expensive to wire. The Mirus does it all in one go.

I have considered the UPS, but it's distortion is too high (5% to 2%) compared to the raw value from the street.

A standard industrial isolation transformer passes noise from primary to secondary very well, unless it has the shield between windings and very, very low capacitance between the windings. Added engineering requires the use of EMC filters to remove differential and common mode noise. By the time all that work in selecting and purchasing these components, it's easier to implement a Torus instead.

did you listened to those filters in a high performance system?

the positive subjective experience is the key

 

[Amir]

The real problem here is that grounding in audio equipment is poorly understood by designers.

this is an important subject

 

[Alrainbow]

In audio, a true balanced connection does not have a center tap. That is because the CT cannot be exactly placed and so common mode rejection ratio will always be degraded by its use. So balanced connections don't use them.

It strikes me that placing a CT on an isolation transformer will introduce a similar problem. Ideally, it would seem that there should be no current in the CT at all (for example if tied to ground) but in practice (if you were to place an ammeter in series with the CT) I think you'll find there always is.

The real problem here is that grounding in audio equipment is poorly understood by designers. As a result, and in particular if you use single-ended equipment, grounding is pretty much the Wild West. IME its best to fix those problems at their source. In that way the AC ground connection can do its job by not being noisy- this in turn will allow the chassis (if there is one) to be a silent shield.

CT???

 

[dcathro]

CT???

Center Tap.

 

[Another Johnson]

This has been a really interesting thread and I appreciate many of the insights shared.

Back to the regenerator question though … in my experience the regenerator’s contribution is dependent on the amp(s) it drives.

None of my CJ amps were even listenable, let alone good, when run on a regenerator. The ARC amps sprung to life with the regenerator set to 117 VAC. The Denon AVR couldn’t have cared less … it plodded along about the same with or without the regenerator.

My point is that blanket dismissals of regenerators based on experience in one setting is myopic. If you ever consider buying one, be sure to buy it from a vendor with a solid return policy. But if you’re lucky, you may not need it. Mine aren’t going back.

 

[Kris]

Explain to me what proprietary technology or known shielding technology has been employed in that commercial transformer. What of it is designed to eliminate noise that impacts audio equipment. Its probably 0. Its is most likely a simple Buck and Boost transformer. Stepdown from 100,000 volts to 12,480 volts. Or 12,480 volt to 480 volt.

I do not work for Torus. I am only 1 of many reps. I do communicate with them and asked their chief technology officer for more info on NBT technology. They of course are not going to publish their proprietary technology and design.. Their exclusive narrow bandwidth technology (NBT) enables their transformers to operate as low-pass filters, allowing the 50Hz or 60Hz wave from the utility to pass while filtering out the higher frequencies where noise, interference and distortion reside. And they have 2 types of surge suppression. Series and MOV. They are in hundreds of recording studios around the world. Those studios do not use a commercial transformer to feed their recording equipment. Maybe the lights and general use outlets.

So, to be blunt, absolutely Yes. I believe a Torus isolation transformer is leaps ahead of a commercial transformer when it comes to filtering power that goes to sensitive amplifying equipment. I believe a part of the misconception of the performance value a good isolation transformer bring is from people using a commercial unit and not finding it satisfactory, then claiming all isolation transformers operate the same. I still have a Topaz 2.4kVA with .0005 pf. 120 volt to 120 volt. Suppose to be the one to get for audio. It came from a film editing studio. You want it. I will give it to you for a good price. Give me $200 and pay for the shipping. I spent $500 to buy it. Try it, then do what you can to sell it because its not going to make you happy. Its a boat anchor in my garage.

Great reading your experience. Hope to learn something I dont know.

I agree about the grounding . This is more Important then the power itself In my system.
In my area We have all 3 phases + neutral delivered to homes , ground MUST be connected to Neutral ( Code requirement) so it is Very difficult to isolate noise one from the other . Also neutral carries some current So it affects the ground.

I do have exactly same Topaz like yours but not 2,4 but 5 kVA and yes it works quite well.
In my case 240 to 240 v isolation.
Fortunate my Topaz being manufactured in San Diego area can be wired this way.
Unfortunate for some reason the SQ lacks dynamics and something is missing From overall presentation.
Have tried 7 balanced trafos ( each for every equipement) . There are some good things related to it but overall Not happy with that.
Have tried 2 very advanced laboratory isolating trafos. Now they are taking dust in the garage .
Had a StromTank S5000 for 10 months and as good as it is unfortunate it kills dynamics badly and sucks the air complealtly . I love air around instruments so had to sell it.

I know this iso devices all work in some areas , but in the end the wall is better after midnight.

The big utility trafo IS NOT DESIGNED TO CLEAN ANYTHING .
no need as it is realtively the cleanest power you can get here.
It is just a way to stay away from other sources of noise that is generated by other houses, small business and farmers .

Will try everything as that is the most important thing in audio.

 

[Amir]

It seems in a high performance system going for Buying Isolation Transformers or any other AC solution is wasteful.
A big/rich company like TAD should put money for designing a good AC Regenerator.

I remember the only audio comany who used AC regenerator was Mark Levinson (NO.32 pre).

last night the AC quality was good between 8:30PM to 9:15PM so the tone was lovely, after 9:30 the AC quality was not good.