AC Filtering, Grounding Boxes, Linear PSU and Balanced Power.

[Speedskater]

On 7/27/2017 at 2:46 PM, One and a half said:

That item is not suitable for after the Transformer,  but after the RCBO.

Something like this, will need a box and cable and wiring. DIY not recommended.

With USA compliance the leakage current is around 10ma, not sure maybe @Speedskater can advise.

https://au.rs-online.com/mobile/p/mcbs/7721303/

The common Ground Fault Circuit Interrupter circuit breaker or receptacle trips at about 4 to 6 mA. But there are other types of GFCI breakers that have different ratings.

 

For most consumer products, UL has a line-to-ground leakage current limit of 0.5 mA.

Worldwide the limit may vary from 0.5 mA to 5.0 mA.

The 0.5 mA requirement limits capacitors to 0.01 uF.

[One and a half]

8 hours ago, Speedskater said:

The common Ground Fault Circuit Interrupter circuit breaker or receptacle trips at about 4 to 6 mA. But there are other types of GFCI breakers that have different ratings.

 

For most consumer products, UL has a line-to-ground leakage current limit of 0.5 mA.

Worldwide the limit may vary from 0.5 mA to 5.0 mA.

The 0.5 mA requirement limits capacitors to 0.01 uF.

Ah ok,  for AU, UK domestic is 30mA. 

 

Industrial feeders are 500mA and work down from there to discriminate. 

 

Often rfi caps to earth are top large and trip the breaker first round, but hold in once the cap is charged. A kludge and nuisance especially for domestic or unattended locations.

 

[Speedskater]

Yes a typical US type GFCI is much more sensitive than an RCB. This sometimes leads to nuisance trips when power conditioner's dump the noise into the Protective Earth/Safety Ground.

Some US pro-audio guys write about other types of GFCI's used in large TV and movie studios. These GFCI's may have different current ratings or response times. In addition to multi-channel audio the have large dangerous theatrical lighting systems and motorized devices to watch out for.

[YashN]

On 7/27/2017 at 9:40 AM, Johnseye said:

 

I probably know as much as you whether his measurements were "well-made". I asked you to elaborate on why you called his measurements pseudo science 2 weeks ago and am still waiting for your reply.  Here it is again.

 

I'm not sure why you would call the measurements he took pseudo science, and what he did wrong.  Please elaborate.

 

I don't think you know as much as I do at all.

 

If you're convinced by blatant mis-measurements, that's your problem. If you think just by showing a graph of mis-measurements, that this is more important than actually knowing whether the measurements are well made and whether the grounding was properly done using the AP gear, that's your problem too, not mine.

 

The reasons have been discussed at length elsewhere and there's also an excellent post about how making proper measurement can be hard by iFi in this very forum.

 

mmerrill can chime in if he wants to, but I doubt you'll understand what he says either.

[zilch0md]

On 7/26/2017 at 9:49 PM, archtype said:

 

 

I realize this thread may be on its last leg, but I wanted to thank you all for your assistance in helping this newbie make some remarkable improvements in my SQ.  

 

I ended up ordering three Isolation Transformers (and returning two of them), and A/B testing them all.  (I did so using the Tripp Lite Outlet strip SS3612, the one without surge protection, with everything plugged into it).  

 

These are the transformers I compared:  Tripp Lite IS1800HG, Topaz 91001-22 1kVA 0.001pF, and Topaz 91018-32 1.8kVA 0.0005pF.  

 

I found that the Tripp Lite yielded only marginally better SQ than plugging directly into the wall.  The 0.001pF Topaz sound was a major improvement over the Tripp Lite.  The 0.0005pF Topaz sound was a noticeable step up in clarity from the 0.001pF Topaz.  I realize that the smaller Topaz has less potential for high current draws, but even with quiet listening, the larger Topaz with 0.0005pF offered a much broader soundstage and greater depth. To me the difference was startling.  If I were to give them each a SQ Score (scale 1-10, poor to incredible), I would rank them…
AC Mains—4
Tripp Lite—4.5
Topaz 0.001pF—7
Topaz 0.0005pF—8

 

In the end I returned both the Tripp Lite and the Topaz 91002-22, and am delighted with the sound I am getting out of the Topaz 91018-32.  

 

My take home?  The Topaz 0.0005pF IT gave me significantly better sound than either the Tripp Lite Hospital Grade IT or the Topaz 0.001pF unit.  The difference is enough for me to recommend upgrading to the 0.0005pF unit from either of the other units.  

BTW, this is a listing of the specs and details on the 0.0005pF Topaz transformers… http://www.pacificparts.com/vends/mge/images/t1.pdf

 

Lastly, I tried inserting a B&K Precision 1604A IT with a floating secondary into my system.  First I tried it upstream from my CD player (modified with a dedicated linear power supply) and could notice no difference.  I then tried it upstream from my Freya preamplifier.  Here I noticed a small but significant improvement in SQ.  (On my scale of 1-10, I would rate it 8.5).  In doing so however, I discovered that the 1604A significantly stepped up the secondary voltage (from 124V to 130V).  This was too high for my comfort, so I added an APC LE1200 voltage regulator upstream from the Topaz unit and set the regulator at 107V.  With this setup, my sound is magic.  

 

Thanks again to you all.  
Ben
 

 

 

Thank you Ben, for encouraging me to look into replacing my Tripp-Lite is500 isolation transformer with a Topaz 91095-32 (500VA, 0.0005 pF) "Ultra-Isolator." 

 

I had followed all of the positive talk of great results had with the low-capacitance Topaz transformers, but it was your diligence, in comparing them to a Tripp-Lite is1800HG, that got me off the fence.  I had assumed that my Tripp-Lite transformer, with grounded-neutral secondaries was lowering the noise floor just as well as any Topaz transformer.  Not so!  

 

That's the insidious thing about noise - you don't know your noise floor can go lower, until you lower it further still.  Ouch!  The Topaz 91095-32 has greatly clarified the high frequency content, with longer decays and better definition of brushes on cymbals, for example.  It's so remarkable, it leaves me wondering how much lower the noise floor could yet go.  

 

For the record, my 500VA Topaz 91095-32 increases the input voltage by about 2.7% and the downstream B&K Precision 1604A (a 1.25 Amp, floating neutral secondary, increases the voltage by another 6.7%) - attributable, in part, to the fact that neither transformer is loaded to its rated capacity, with my Oppo Sonica DAC pulling only 20.5 Watts (which is about 1/14th of the 1604A's rated capacity and only 1/24th of the Topaz' rated capacity.  I am therefore have the Topaz plugged into the same APC LE1200 voltage regulator I had been using with the Tripp-Lite is500, with the regulator set to an output voltage of 107.  This allows the Sonica DAC to see a voltage of 122.0, down from 133.7 in the absence of the voltage regulator (with a mains voltage of 123.0). 

 

Every component other than my Sonica DAC is DC-powered, running on independent LiPo battery packs (no AC noise and no ground loops). 

 

:-) 

 

Mike

[Johnseye]

On 8/1/2017 at 0:26 PM, YashN said:

 

I don't think you know as much as I do at all.

 

If you're convinced by blatant mis-measurements, that's your problem. If you think just by showing a graph of mis-measurements, that this is more important than actually knowing whether the measurements are well made and whether the grounding was properly done using the AP gear, that's your problem too, not mine.

 

The reasons have been discussed at length elsewhere and there's also an excellent post about how making proper measurement can be hard by iFi in this very forum.

 

mmerrill can chime in if he wants to, but I doubt you'll understand what he says either.

 

Prove it.

[marce]

On 27/07/2017 at 6:24 AM, YashN said:

 

I conclude therefore that you are incapable of showing his numerous and blatant mistakes in his so-called measurements.

 

Better called mis-measurements, over at the audiosciencemockery forum. Feel free to join there, and do us all a favour too, stay there.

Missed this one, been away for a while... why should I stay there...

Rude mate, rude. LOL

[mikicasellas]

Hello all,

 

Has any tried using 2 Topaz Ultra Isolator in series?:

 

One 2.4 KVA Topaz with a .0005 pF of capacitance and the other 1 KVA Topaz Ultra Isolator with .005 pF

Could that setup make an improvement or this should not be done?

 

Thanks!

[archtype]

 I have not tried it, so I could only give you my thoughts on what it might be like.

 

 I can think of no reason why it couldn't be done.  Theoretically, it  should help reduce the high frequency harmonics that degrade sound quality.  However the 0.0005 pF transformer is so effective, I would be surprised if the addition of the 0.005 pF transformer would add enough to make a difference in sound quality. However I can't prove that.

 

 If you decide to give it a try,   it is possible that you may end up with higher voltage coming out of the two of these in series than is desirable for your system.  The voltage is stepped up by small degree with each isolation transformer. Putting them in series would increase this obviously. If you were going to give this a try, before you hook your equipment up to it, I would probably check your voltage going in and out of each transformer using a Kill A Watt.  

 

 If you decide to give it a try, please report your results. I think it would be very interesting to  hear your observations. 

[zilch0md]

1 hour ago, mikicasellas said:

Hello all,

 

Has any tried using 2 Topaz Ultra Isolator in series?:

 

One 2.4 KVA Topaz with a .0005 pF of capacitance and the other 1 KVA Topaz Ultra Isolator with .005 pF

Could that setup make an improvement or this should not be done?

 

Thanks!

 

Hi mikcasellas,

 

I would just think it's unnecessary.

 

A single Topaz Ultra-Isolator provides -146 dB attenuation of common-mode noise at frequencies as high as 20kHz, which equates to a 20 Million-to-1 reduction. It will give you better than -150 dB of attenuation all the way to 8000 Hz. 

 

Putting two of them in series would theoretically provide (20 Million * 20 Million)-to-1 attenuation, which is 400,000,000,000,000-to-1.  Even in the highest treble frequencies, you're not likely to hear any improvement.

 

There's also the fact that with my  small, 500VA Topaz 19095-32 and archtype's much larger 2.4kVA Topaz, the output voltages we have measured are higher than the input voltages.  If you were to string two of them together in series, your final output voltage would suffer the gain imposed by both of them.  In my case, I'm seeing about a 2.7% increase in voltage, before loading, but it doesn't drop much, even when loaded, as I'm using a headphone system that doesn't pull a lot of watts.

 

Archtype and I are both using APC LE1200 voltage regulators on the upstream side of our respective Topaz Ultra-Isolator transformers, to keep the final voltage seen by our components at around 120V.  But for this voltage regulator, set to 107V at the wall outlet, we wouldn't be able to use B&K Precision 1604A floating-neutral isolation transformers, in series, downstream from our respective Topaz grounded-neutral transformers.

 

Whether or not one subscribes to the idea that a floating-neutral IT can attenuate common-mode noise (voltage transients that reside between Ground and Hot OR between Ground and Neutral), just as well as grounded-neutral ITs, there's still the benefit offered only by floating-neutral ITs, of preventing backwash noise (from a component that is using a SMPS, especially) from polluting other components that are plugged into the same powerstrip that's fed by the Topaz.  

 

So, in my opinion, it's far more beneficial to serially connect a floating-neutral IT to the output of a grounded-neutral IT, than it would be to serially connect two grounded-neutral ITs.  Keep in mind that I only use the floating-neutral ITs to power digital components (source and DAC).  AC-powered analog components can be plugged directly into the Topaz power strip (unless you know that one of them uses a SMPS.)  

 

Mike

[zilch0md]

:-)  I crossed with archtype - posting similar comments.  :-)

[mikicasellas]

1 hour ago, zilch0md said:

 

Hi mikcasellas,

 

I would just think it's unnecessary.

 

A single Topaz Ultra-Isolator provides -146 dB attenuation of common-mode noise at frequencies as high as 20kHz, which equates to a 20 Million-to-1 reduction. It will give you better than -150 dB of attenuation all the way to 8000 Hz. 

 

Putting two of them in series would theoretically provide (20 Million * 20 Million)-to-1 attenuation, which is 400,000,000,000,000-to-1.  Even in the highest treble frequencies, you're not likely to hear any improvement.

 

There's also the fact that with my  small, 500VA Topaz 19095-32 and archtype's much larger 2.4kVA Topaz, the output voltages we have measured are higher than the input voltages.  If you were to string two of them together in series, your final output voltage would suffer the gain imposed by both of them.  In my case, I'm seeing about a 2.7% increase in voltage, before loading, but it doesn't drop much, even when loaded, as I'm using a headphone system that doesn't pull a lot of watts.

 

Archtype and I are both using APC LE1200 voltage regulators on the upstream side of our respective Topaz Ultra-Isolator transformers, to keep the final voltage seen by our components at around 120V.  But for this voltage regulator, set to 107V at the wall outlet, we wouldn't be able to use B&K Precision 1604A floating-neutral isolation transformers, in series, downstream from our respective Topaz grounded-neutral transformers.

 

Whether or not one subscribes to the idea that a floating-neutral IT can attenuate common-mode noise (voltage transients that reside between Ground and Hot OR between Ground and Neutral), just as well as grounded-neutral ITs, there's still the benefit offered only by floating-neutral ITs, of preventing backwash noise (from a component that is using a SMPS, especially) from polluting other components that are plugged into the same powerstrip that's fed by the Topaz.  

 

So, in my opinion, it's far more beneficial to serially connect a floating-neutral IT to the output of a grounded-neutral IT, than it would be to serially connect two grounded-neutral ITs.  Keep in mind that I only use the floating-neutral ITs to power digital components (source and DAC).  AC-powered analog components can be plugged directly into the Topaz power strip (unless you know that one of them uses a SMPS.)  

 

Mike

 

Mike,

 

Thanks for the detailed feedback.

 

Do you think the 2.4 KVA .0005pF would performance better than the 1 KVA with .005pF?

 

Miki

 

 

[zilch0md]

14 minutes ago, mikicasellas said:

 

Mike,

 

Thanks for the detailed feedback.

 

Do you think the 2.4 KVA .0005pF would performance better than the 1 KVA with .005pF?

 

Miki

 

 

 

Hi Miki,

 

I honestly don't know if that difference would be audibly discernible.  All of the Topaz Ultra-Isolators are rated as providing -146 dB of normal-mode noise attenuation at frequencies as high as 20kHz, so I wouldn't worry about 0.005 pF vs. 0.0005 pF.   

 

It's more important to select a VA rating that's at least 2x the total VA of all of the loads you intend to plug into the Topaz transformer.  

 

A Kil-A-Watt can be used to measure the power consumed by each of your loads, in VA or in Watts.  

 

Before selecting a Topaz model, you could plug the Kil-A-Watt into a wall outlet, then plug a power strip into the Kil-A-Watt, then plug all of your components into that power strip, operating your amp at the maximum level you would normally use.   That would allow the Kil-A-Watt to report the total VA being consumed by all components.  

 

You can then go looking for a Topaz model with a VA rating that's at least twice that amount.

 

@archtype  Did you come up with a figure for the total VA consumed by all of your components?   And, what fraction is that of your giant Topaz' VA rating?

 

(I would agree that it doesn't hurt to go over 2x the amount consumed.)

 

Thanks!

[mikicasellas]

1 hour ago, zilch0md said:

 

Hi Miki,

 

I honestly don't know if that difference would be audibly discernible.  All of the Topaz Ultra-Isolators are rated as providing -146 dB of normal-mode noise attenuation at frequencies as high as 20kHz, so I wouldn't worry about 0.005 pF vs. 0.0005 pF.   

 

It's more important to select a VA rating that's at least 2x the total VA of all of the loads you intend to plug into the Topaz transformer.  

 

A Kil-A-Watt can be used to measure the power consumed by each of your loads, in VA or in Watts.  

 

Before selecting a Topaz model, you could plug the Kil-A-Watt into a wall outlet, then plug a power strip into the Kil-A-Watt, then plug all of your components into that power strip, operating your amp at the maximum level you would normally use.   That would allow the Kil-A-Watt to report the total VA being consumed by all components.  

 

You can then go looking for a Topaz model with a VA rating that's at least twice that amount.

 

@archtype  Did you come up with a figure for the total VA consumed by all of your components?   And, what fraction is that of your giant Topaz' VA rating?

 

(I would agree that it doesn't hurt to go over 2x the amount consumed.)

 

Thanks!

 

 

I will measure it again, i already have a Kill-aA-Watt and i believe when i measured it was something like 400VA aprox.

 

Someone somewhere here was telling that he got a lot of better results with the 2.4 KVA .0005pF vs his 1 KVA .005pF

 

Thanks

 

 

[Ralf11]

If I buy a TOpaz or other item, can I stick it outside on my deck and punch a hole thru the wall for conduit to make up run to an outlet for it?

 

Or would this violate every NEC ever made?

[archtype]

Miki

In a previous post, I described my experience using a 1KVA topaz 0.001 pF transformer and comparing it to a 1.8 KVA 0.0005 pF topaz transformer. My experience was that the transformer with lower capacitance made a discernible improvement in my sound quality. This was even when listening to sounds at a very low amplitude. 

 Regarding the question as to how much power I was drawing, using my Kill A Watt, at peak  listening levels with my two 500 W monobloc's, I was drawing about 1100 W at peak. Most of the time at softer or more usual levels of listening, I am drawing around 400 W. 

Based on my experience, I believe the lower capacitance does make a difference, and if you have a chance to choose between the two transformers, I would choose the one with the lowest capacitance. 

[zilch0md]

11 hours ago, Ralf11 said:

If I buy a TOpaz or other item, can I stick it outside on my deck and punch a hole thru the wall for conduit to make up run to an outlet for it?

 

Or would this violate every NEC ever made?

 

As long as the transformer isn't exposed to the weather and the hookup is done to code, I'd say 'go for it.'

 

By the way, these specs can be found for the Topaz Ultra-Isolators:

 

Operating temperature 0°C to 50°C

Storage temperature -40°C to 85°C

[zilch0md]

Here's another PDF that may be of interest to those considering a Topaz Ultra-Isolator:

 

http://www.surplussales.com/Transformers/Pdfs/tp-91092-31_N.pdf

 

It has some interesting discussion about adding capacitors on the downstream side of a Topaz, as well as substituting the mains earth ground with one of your own making.  Scary.  :-)

 

It also provides hookup instructions for the models that have terminal strips instead of ready-to-use, three-prong outlets.

[zilch0md]

14 hours ago, zilch0md said:

 

Hi Miki,

 

I honestly don't know if that difference would be audibly discernible.  All of the Topaz Ultra-Isolators are rated as providing -146 dB of normal-mode noise attenuation at frequencies as high as 20kHz, so I wouldn't worry about 0.005 pF vs. 0.0005 pF.   

 

[snip]

 

Correcting my previous post...

 

All of the Topaz Ultra-Isolators are rated as providing -146 dB of common-mode noise attenuation (not normal-mode, as I had written above).

 

I don't know if archtype's 1kVA Topaz 0.001 pF transformer was designated as an "Ultra-Isolator" (and I don't recall the model number he gave previously), but given his experience in comparing that 0.001 pF transformer to his 1.8 kVA 0.0005 pF transformer,  it only makes sense to go for the lowest capacitance.  

 

Further research has revealed... (I'm still learning...) that only the Topaz transformers with part numbers ending in -31 or -32 have a 0.0005 pF capacitance.  So, that's the easiest way to go shopping for a 0.0005 pF Topaz, given that sellers don't always provide a sharp photo of the rear label.

 

[zilch0md]

Another discovery:  

 

The trademark "Ultra-Isolator," which can be seen on Topaz, Daitron, MGE, and Eaton isolation transformers,  is not synonymous with a capacitance of 0.0005 pF.

 

 

[Solstice380]

20 hours ago, zilch0md said:

Another discovery:  

 

The trademark "Ultra-Isolator," which can be seen on Topaz, Daitron, MGE, and Eaton isolation transformers,  is not synonymous with a capacitance of 0.0005 pF.

 

Definitely get a -3x model.  Pricing doesn't seem to follow with the lower capacitance.  Not sure the sellers (mostly used available anyway, although I did get a great buy recently on a new 750VA 0.0005C!) are concerned or even aware of the importance of the capacitance. 

[zilch0md]

Here's an amazingly relevant video, for those seeking a better understanding of the relationship between an isolation transformer's inter-winding capacitance and leakage currents:

 

Stray-Currents, compensating-currents and ground-loops in Transformers

 

 

[Speed Racer]

2 hours ago, zilch0md said:

Here's an amazingly relevant video, for those seeking a better understanding of the relationship between an isolation transformer's inter-winding capacitance and leakage currents:

 

Stray-Currents, compensating-currents and ground-loops in Transformers

 

 

 

The Topaz has the "shield" to lower the effect of stray capacitance between the primary and secondary windings.

 

In the video example, there was 160pF stray capacitance between the primary and secondary windings. Is that the same as 160pF interwinding capacitance? Does that mean that my Topaz has .0005pF stray capacitance?

 

If that is the case, while the isolation transformer in the example video "leaked" 12µA at 230v and 50Hz, my Topaz leaks .0000226µA at 120v and 60Hz. That's a whole hell of a lot better! No wonder it has 146dB of common mode noise reduction. Wired up to deliver balanced power, there is supposed to be about 20dB more common mode noise reduction. 

[zilch0md]

2 hours ago, Speed Racer said:

 

[snip]

 

In the video example, there was 160pF stray capacitance between the primary and secondary windings. Is that the same as 160pF interwinding capacitance? Does that mean that my Topaz has .0005pF stray capacitance?

 

[snip]

 

Wired up to deliver balanced power, there is supposed to be about 20dB more common mode noise reduction. 

 

Yes!

 

Even the Tripp-Lite ITs have Faraday shields, but their design/construction is not anywhere near as effective, with only 1000:1 (60 dB) CMNR vs. 20,000,000:1 (146 dB) with the Topaz 0.0005 pF models.

 

You are the second person, I know, to comment that balanced-mode configuration of the screw-terminal models offer an additional 20 dB of CMNR.  Do you have a source for that information?  I'm not sure I actually need more than 20,000,000:1 attenuation, but I'm curious.  :-)

[Speed Racer]

The balanced configuration cancels X amount of common mode noise at the device. So, whatever makes it through the interwindings and on to the power cable is partially canceled.

 

Even if I weren't getting better common mode noise reduction, I like the fact that the isolation transformer is running both quiter and cooler.