[mkaza]

Hi mkaza,

I wish I had an equation for calculating the maximum frequency at which NR is "effective" vs. the pF capacitance rating, but all I can say is that 0.001 is pretty good, compared to most new isolation transformers available today. Still, the smaller the pF, the higher the audio frequencies at which effective noise reduction is audible. I just don't have any data points to share.

Many thanks @zilch0md for your reply. I agree, quantification is not easy - but your view that an IT with 0.001pf is pretty good - will make it worth a shot! Thanks.

@mkaza

I’d think the best sound quality would come from the wall/IT/component combination but since you already have the UPS it won’t hurt to give it a shot.

I had a Surgex SA1810 prior to my Topaz/MGE units. They are good but don’t offer the same level of performance as my ITs. In the purely surge protector category, I’d highly recommend trying one though.

A dedicated circuit will provide less noise for an IT to attenuate but there still will be noise on a dedicated circuit due to upstream power components outside of your home. If you are in the US it’s possible to use other 120v wire that comes into your house as a quieter and cheaper source for a “dedicated circuit.” It’s not truly dedicated since it has 240v components on it but the number of devices on it is smaller.

Many thanks @oneguy for your reply. Yes, even my Audio Note dealer insists to remove the UPS but I am comfortable keeping it to address any power outages. My requirement is more towards noise filtration than surge protection currently. SurgeX components incidentally provide inherent EMI/RFI filtration (apart from surge protection) but your experience of IT proving more beneficial than SurgeX is an important data point for me.

One other q - my UPS is rated 3KVA - so, should the IT be also rated 3KVA or can I go with something smaller, say 2KVA considering my components' peak demand would be around 700W only?

 

[oneguy]

Many thanks @zilch0md for your reply. I agree, quantification is not easy - but your view that an IT with 0.001pf is pretty good - will make it worth a shot! Thanks.



Many thanks @oneguy for your reply. Yes, even my Audio Note dealer insists to remove the UPS but I am comfortable keeping it to address any power outages. My requirement is more towards noise filtration than surge protection currently. SurgeX components incidentally provide inherent EMI/RFI filtration (apart from surge protection) but your experience of IT proving more beneficial than SurgeX is an important data point for me.

One other q - my UPS is rated 3KVA - so, should the IT be also rated 3KVA or can I go with something smaller, say 2KVA considering my components' peak demand would be around 700W only?

If the power draw tops out at 700W then an 2KVA IT should be fine if the UPS is t replenish itself. If it needs to replenish itself then you would have take that into consideration as well.

 

[mkaza]

If the power draw tops out at 700W then an 2KVA IT should be fine if the UPS is t replenish itself. If it needs to replenish itself then you would have take that into consideration as well.

Thanks much, appreciate it!

 

[oneguy]

I bought an EMI noise tester to see how the Topaz -69 does. The pics below are with the house air conditioner off. With it on it was roughly 350mV no IT and 8mV with the IT.

EMI tester specs:
Input voltage: AC 85Vac - 250Vac 50/60Hz (accuracy +/- 1 VAC)
Frequency range: 10KHz - 10MHz
Noise range: 1mV - 1999mV
Measurement accuracy: +/- 5% @ 1MHZ; +5% to -50% @ 10 KHz TO 10MHz
Noise indication: 4 digits yellow and blue two-color OLED digital dot matrix display
Dimensions: 132 X 91 X 41mm
Machine weight: 150g

 

[mkaza]

I bought an EMI noise tester to see how the Topaz -69 does. The pics below are with the house air conditioner off. With it on it was roughly 350mV no IT and 8mV with the IT.

Excellent - many thanks; a good confirmation that Topaz is doing its job pretty effectively. If you don't mind, what is the pF & CMNRR ratings for the '69'?

 

[oneguy]

Excellent - many thanks; a good confirmation that Topaz is doing its job pretty effectively. If you don't mind, what is the pF & CMNRR ratings for the '69'?

The 91905-69 has the same 0.0005pf capacitance as the -31 and -32 transformers and the exterior construction is the same as well. The difference is it can accept 100, 120, 200, 220, or 240v input and outputs 120 or 240v. The only one I’ve seen is the one I have and unfortunately I don’t have any other specs sheets for it.

I tested this one because it’s outlets are not being used and thus nothing on the output side to pollute the readings.


At @Energy ‘s request, I plugged the 91905-69 in series with the 91005-31 (another 5KVA IT) to see what would happen. Here are the results:
From the wall

After the 91005-31

After the 91905-69


Things to consider: the 91005-31 had all of my audio stack plugged into it likely contributing to the high reading of 25mV. The things plugged in that are powered on are the Holo Spring KTE in standby mode, LPS-1.2 (SPMS powered), iBox-V100 (SPME powered), SU-1 (powered by LPS-1.2), EtherRegen (power by LH Labs LPS-4), LH Labs Pulse Blue (power by LH Labs LPS-4), Emotiva XPS-1 (power by LH Labs LPS-4). The SMPS powering my turntable was removed so I had a free outlet.


My takeaways:

-Components on a common strip on the output side can pollute the output side which I think we all knew that, the only question is to what degree and is it audible. I’ll eventually unplug everything from the 91005-31 and get baseline readings for it.

-The measurement of 2 IT’s of 0.0005pf each is series resulted in the same 8mV reading which leads me to believe the actual noise is at the floor of what this EMI reader can detect. So, the 8mV reading earlier for just the single IT is likely below that but the meter isn’t precise enough to measure it. (Makes sense given the high amount of noise attenuation of these units).

*NOTE: All of my transformers are wired for balanced power output.

 

[rsbrsvp]

I have a question for all you experts...

Should my entire system be hooked into the topaz or should my PC or my LPS for my regen iso NOT be plugged into the topaz?

 

[Zenvota]

I have a question for all you experts...

Should my entire system be hooked into the topaz or should my PC or my LPS for my regen iso NOT be plugged into the topaz?

if youre using the iso regen with galvanic isolation and the uptone ultracap lps has the ground shunted smps i would not put the pc and the smps on the topaz.

 

[rsbrsvp]

I'm using the regen ISO with galvanic isolation but NT the LPS from uptone...

 

[Zenvota]

hmm, so the uptones lps + the shunted smps provide complete isolation breaking high and low impedance leakage loops. You could try with the pc off the topaz but the regens lps on the topaz, but I think leaving the pc and lps on the topaz would be better, having leakage loops when one component is on the topaz and one off increases impedance and noise. Your power cables and regenative supplies should offset any line noise or harmonics from the pcs smps

 

[oneguy]

A couple updates:
I finally got around to unplugging everything from my mainstay the 5KVA-31 to see how it would compare to the 5KVA-69. The results are below. All measurements were done at the power strip. The 5KVA-69 and 2.5KVA-31 have powerstrips constructed nearly identically. The 5KVA-31’s power strip is constructed similarly to the other two just with one extra plug and the rhodium plated plugs from a different manufacturer. The T-100R is using stock wire and plugs just rewired by me for balanced power.


Baseline from the wall: 315mV
5KVA-31: 36mV
2.5KVA-31: 20mV
1.8K-T100R: 20mV
5KVA-69: 8mV

Takeaways:
The 5KVA keeps measuring 8mV which is the lowest. Interestingly, the 5KVA-31 last measured 25mV with a baseline of 290mV from the wall so it was actually worse this time. Although not likely, I thought the equipment being plugged into the 5KVA-31’s original test may have skewed the measurements for the better so I plugged everything that was into the 5KVA-31 into the 5KVA-69 and measured. The -69 remained at 8mV. So given this, I have been promoted from collecting dust status to main system use.

The the 1.8KVA-T100R and 2.5KVA came in at the same value but one without the silverplated copper wire wire and rhodium plated receptacles or plugs. Probably not a factor for this test but something worth noting. I have 2 more rhodium plated receptacles that I can wire into the 1.8KVA-T100R if I have time before I move.

Next update:

On the audiophile style a while ago the question was brought up about safe distances from transformers. I bought a GQ EMF-390 to take some measurements and here are the results:


(From the 5KVA-69 with my system not being used to make music)

Distance from core (inches) EMF(mG)

27——————————————2.5
10.25—————————————30
5.75—————————————100

For EF and RF no significant increase at any distance.

Other:

-EF was 3V/m at about 3” from my unshielded power cords with my system at idle and increased with the amount of power being used.

-cords with their components switched off still affected the EF tester. Unplugging the cord caused it to not affect the tester.


-putting my hand in front of the tester was enough to significantly shield the tester from EF. For EMF my hand in front of the tester did not have a noticeable effect.

At my listening position with the system not making music I measured 0.4mG, 1.0V/m and 125-151pW/cm2. My listening position is approximately 8.5 ft from the transformer and behind by stack of vinyl.

CAVEAT: These are all approximate measurements and only data on small sample size of units. This is not meant to be definitive in anyway. Just food for thought.

 

[Zenvota]

A couple updates:
I finally got around to unplugging everything from my mainstay the 5KVA-31 to see how it would compare to the 5KVA-69. The results are below. All measurements were done at the power strip. The 5KVA-69 and 2.5KVA-31 have powerstrips constructed nearly identically. The 5KVA-31’s power strip is constructed similarly to the other two just with one extra plug and the rhodium plated plugs from a different manufacturer. The T-100R is using stock wire and plugs just rewired by me for balanced power.


Baseline from the wall: 315mV
5KVA-31: 36mV
2.5KVA-31: 20mV
1.8K-T100R: 20mV
5KVA-69: 8mV

Takeaways:
The 5KVA keeps measuring 8mV which is the lowest. Interestingly, the 5KVA-31 last measured 25mV with a baseline of 290mV from the wall so it was actually worse this time. Although not likely, I thought the equipment being plugged into the 5KVA-31’s original test may have skewed the measurements for the better so I plugged everything that was into the 5KVA-31 into the 5KVA-69 and measured. The -69 remained at 8mV. So given this, I have been promoted from collecting dust status to main system use.

The the 1.8KVA-T100R and 2.5KVA came in at the same value but one without the silverplated copper wire wire and rhodium plated receptacles or plugs. Probably not a factor for this test but something worth noting. I have 2 more rhodium plated receptacles that I can wire into the 1.8KVA-T100R if I have time before I move.

Next update:

On the audiophile style a while ago the question was brought up about safe distances from transformers. I bought a GQ EMF-390 to take some measurements and here are the results:


(From the 5KVA-69 with my system not being used to make music)

Distance from core (inches) EMF(mG)

27——————————————2.5
10.25—————————————30
5.75—————————————100

For EF and RF no significant increase at any distance.

Other:

-EF was 3V/m at about 3” from my unshielded power cords with my system at idle and increased with the amount of power being used.

-cords with their components switched off still affected the EF tester. Unplugging the cord caused it to not affect the tester.


-putting my hand in front of the tester was enough to significantly shield the tester from EF. For EMF my hand in front of the tester did not have a noticeable effect.

At my listening position with the system not making music I measured 0.4mG, 1.0V/m and 125-151pW/cm2. My listening position is approximately 8.5 ft from the transformer and behind by stack of vinyl.

CAVEAT: These are all approximate measurements and only data on small sample size of units. This is not meant to be definitive in anyway. Just food for thought.

Is there anyway to reduce emf? faraday cages? iron cases?

i measure 0.0mg at my seating positions, but its a cheap $10 tester and i sit about 4' from a unit.

I realised an area with a bunch of switching supplies radiates a 10mg field on the other side of the wall. interestingly uptones ultracaps have almost no radiation, whereas the linear supplies in my headamp have a fairly strong field but not that large. cheap switching supplies have very high emf but very small fields

today i tried removing the topaz from the headphone system and it was immediately clear the difference in sound quality, couldnt listen as loud, less defined treble quality, impacted imaging and staging, tried to ignore it and acclimate but after a while i plugged the topaz back in and again the difference was immediately noticeable, able to crank the volume back up, purely defined cymbals etc.

Its especially noticeable with the music i listen to which is heavily distorted and just becomes a wall of unintelligible sound, its been really surprising how beneficial all this equipment has been to audio quality with lofi recordings, i never though hifi was something that would benefit me ^^

i like these things but i do worry about the emf. Though i had read studies years ago that documented negative effects of emf were only from very large power generators and not from magnetic transformers, power cables, etc.

Im going to try removing them from my multi channel system and projectors next. I dont think ill see a difference with the projectors due to the nature of the units(dmd + current driven leds) but i do like the effects of noise reduction on the units in terms of a power supply heat and processing as well as the surge supression. But i have a feeling its going to be very noticeable with the multi channel system(it was adding it, imaging and treble quality drastically improved) so i may just try and find a better location for that unit where ill have less contact with it, since i need to rearrange all those switching supplies for the hdmi extenders anyway.

 

[oneguy]

Is there anyway to reduce emf? faraday cages? iron cases?

i measure 0.0mg at my seating positions, but its a cheap $10 tester and i sit about 4' from a unit.

I realised an area with a bunch of switching supplies radiates a 10mg field on the other side of the wall. interestingly uptones ultracaps have almost no radiation, whereas the linear supplies in my headamp have a fairly strong field but not that large. cheap switching supplies have very high emf but very small fields

today i tried removing the topaz from the headphone system and it was immediately clear the difference in sound quality, couldnt listen as loud, less defined treble quality, impacted imaging and staging, tried to ignore it and acclimate but after a while i plugged the topaz back in and again the difference was immediately noticeable, able to crank the volume back up, purely defined cymbals etc.

Its especially noticeable with the music i listen to which is heavily distorted and just becomes a wall of unintelligible sound, its been really surprising how beneficial all this equipment has been to audio quality with lofi recordings, i never though hifi was something that would benefit me ^^

i like these things but i do worry about the emf. Though i had read studies years ago that documented negative effects of emf were only from very large power generators and not from magnetic transformers, power cables, etc.

Im going to try removing them from my multi channel system and projectors next. I dont think ill see a difference with the projectors due to the nature of the units(dmd + current driven leds) but i do like the effects of noise reduction on the units in terms of a power supply heat and processing as well as the surge supression. But i have a feeling its going to be very noticeable with the multi channel system(it was adding it, imaging and treble quality drastically improved) so i may just try and find a better location for that unit where ill have less contact with it, since i need to rearrange all those switching supplies for the hdmi extenders anyway.

Based in the follow owing graph you could make a faraday cage or box out of 0.067” thickness or greater 410 steel to shield the primary 60hz frequency. Also, adding distance between you and the source will help mitigate.

https://en.m.wikipedia.org/wiki/Skin_effect
(The red line represents 50hz)

 

[Thenewguy007]

Anyone know how this would compete with Topaz isolation transformers?
https://www.aliexpress.com/item/4001285233703.html
or
https://www.aliexpress.com/item/4000176601682.html

Price is much cheaper & is much more recent than used Topaz.
What should the input & output selection be for North America? Both 110v?

 

[darkfireblade25]

Is the chart in the OP still accurate in that you need to get 2 sets of transformers? I'm currently confused as to how you're supposed to connect your system with computer/dac/amp. There's a lot of stuff said in the thread but none are summarized into one recommendation after being discussed so far.

I just have a computer/Schiit Bifrost 2/and THX 887 amp. The Bifrost 2's Unison USB claims to have electrostatic+electromagnetic isolation so I'm not sure if I need to move the PC off the Topaz or not. I see a lot of Topazs coming with two receptacles. Do you need to put every component on a strip off of 1 receptacle or can you mix and match?

(Edited for question clarifying)