Isolation transformer effect on sound experiences

[Arno Fennix]

For some years I had 2 1500VA isolation transformers laying around for trying these in my audio system.
For the mains supply to the audio there is a dedicated power line and for all small consumers these are fed via an Isotek filter unit.
Before I tried to also feed the 2 power amps (SET 300B) via the isotek high power outputs but this did kill the dynamics a lot.

So, yesterday, connected the SETs to the 1500VA isolation transformer and an extra cable for attaching the "earth" wire.
Listening to music and I am amazed by the difference. Harshness disappeared (high mids / highs). When connecting the "earth" some of that seemed to come back again.

First steps, and I am working towards some measurements and options to switch faster between with/without isolation transformer for further listening tests.

Anyone tried these kind of comparisons (W/WO) by listening as well as measuring. What are your experiences?

 

[syzygetic]

I use an isolation transformer too on my main system. I noticed no audible improvements, but it did prevent my highly automated, LED-lighting-filled home from interfering with and distorting the displays on my Meridian speakers... so I guess it did something! Again, nothing audible, though, I was only trying to stop the flickering and interference in the electronics.

 

[Arno Fennix]

I use an isolation transformer too on my main system. I noticed no audible improvements, but it did prevent my highly automated, LED-lighting-filled home from interfering with and distorting the displays on my Meridian speakers... so I guess it did something! Again, nothing audible, though, I was only trying to stop the flickering and interference in the electronics.

I can imagine the effect or purpose can be different per system as well.

The other thought on isolation transformers is, why would these make the difference that could not be made with an on-board mains (iso) transformer per audio component...after all, all got a mains transformer

 

[syzygetic]

Indeed, I’m not sure what wasn’t properly isolated to begin with, or perhaps simply forcing a common ground helped. I don’t know, and I agree it probably depends on the system and the environment, but I would think that only in the most severe cases of interference would you find an audible improvement.

 

[SIY]

For some years I had 2 1500VA isolation transformers laying around for trying these in my audio system.
For the mains supply to the audio there is a dedicated power line and for all small consumers these are fed via an Isotek filter unit.
Before I tried to also feed the 2 power amps (SET 300B) via the isotek high power outputs but this did kill the dynamics a lot.

So, yesterday, connected the SETs to the 1500VA isolation transformer and an extra cable for attaching the "earth" wire.
Listening to music and I am amazed by the difference. Harshness disappeared (high mids / highs). When connecting the "earth" some of that seemed to come back again.

First steps, and I am working towards some measurements and options to switch faster between with/without isolation transformer for further listening tests.

Anyone tried these kind of comparisons (W/WO) by listening as well as measuring. What are your experiences?

How did you implement basic controls?

 

[Arno Fennix]

How did you implement basic controls?

Can you explain a bit what you mean?
For now, the isolation transformers are in experiment phase for me

 

[SIY]

Can you explain a bit what you mean?
For now, the isolation transformers are in experiment phase for me

Randomization and blinding.

 

[Arno Fennix]

Randomization and blinding.

Oh, not there yet. Just checken if other people had experiences and investigated this any further

 

[Speedskater]

While an AC power line isolation transformer(S) might reduce hum, buzz and other low frequency power line noise, I don't see how it could cause: Harshness disappeared (high mids / highs)

 

[Chrispy]

While an AC power line isolation transformer(S) might reduce hum, buzz and other low frequency power line noise, I don't see how it could cause: Harshness disappeared (high mids / highs)

That depends on your expectations perhaps?

 

[SIY]

While an AC power line isolation transformer(S) might reduce hum, buzz and other low frequency power line noise, I don't see how it could cause: Harshness disappeared (high mids / highs)

Or a change of dynamics.

Bottom line: if you don't have line noise without the transformer, don't bother with the transformer.

 

[Arno Fennix]

That depends on your expectations perhaps?

I am aware of that always present potential risk indeed

 

[wwenze]

I can imagine the effect or purpose can be different per system as well.

The other thought on isolation transformers is, why would these make the difference that could not be made with an on-board mains (iso) transformer per audio component...after all, all got a mains transformer

Common-mode is common.

 

[Arno Fennix]

Playing around a bit. Listening to with/without there is a significant difference. Also when someone else switched between the normal/iso connection it was easily identifiable when the iso was in or not.

 

[Arno Fennix]

Interesting site:

https://ggianluca.wixsite.com/opamplifier/the-51st-anniversary-line-isolator

 

[SIY]

Interesting site:

https://ggianluca.wixsite.com/opamplifier/the-51st-anniversary-line-isolator

One of many with beautiful construction and poor engineering. Very common in faith-based fashion audio.

 

[Arno Fennix]

One of many with beautiful construction and poor engineering. Very common in faith-based fashion audio.

If there are additional fact-based diy project links, please add, I would be interested

 

[SIY]

If there are additional fact-based diy project links, please add, I would be interested

For what kind of project? That site had a mish mash of many things, some useful (if poorly engineered), some not.

Step 1: Define the problem you're trying to solve.
Step 2: Validate the existence of the problem (this step here is the stumbling block for faith-based fashion audiophiles, rendering the rest moot).
Step 3: Define the metric of success.
Step 4: Implement the attempted solution.
Step 5: Validate that the solution has solved the problem.

 

[Rock Rabbit]

Playing around a bit. Listening to with/without there is a significant difference. Also when someone else switched between the normal/iso connection it was easily identifiable when the iso was in or not.


View attachment 102192View attachment 102193

The readings on Alphalab EMI means some 2.4 dB "better" in the range over 10 kHz, and 8% precision over 20 kHz to 10 Mhz. EMI below 20 kHz (?) and measurable effect in 10-20 kHz from 10 MHz bandwidth is non conclusive...unless you could measure some of this EMI in loudspeaker cables.
Anyway a line filter is many dB better so I hope your pre/amp don't have a line filter for this effect to be measurable (a simple noise floor line level digital recording can show that difference)
https://www.mpe.co.uk/products/low-leakage-extended-performance-range-filters/
An example of a line filter with 100 dB @ 10 kHz.
But even without filter the noise rejection (audible range) of any internal power supply put that 2 dB difference (or less due to instrument error) below the amp noise

 

[Mnyb]

Also SET amps ? How are these designed ? It’s a factor .

“Normal” amps have property called PSRR “power supply rejection ratio” it can be in 100dB range so the circuit design itself of modern amps helps with defeating gunk from the power supply .

Botique brand audiophile amps with some kind of faith based design rather then electrical engineering may not behave a normally . The prior sentence may seen as absurd and slightly provoking. That’s not my intent .
But it’s actually what’s happening a very simple design with little or no global feedback may have low PSRR because the designers sacrificed this because “negative feedback is evil” or similar priorities a real engineer won’t have , the faith based designer will accept real problems in exchange of solving imaginary ones .

Isolation transformers or special filters should not be needed normally. That said I was once suckered inte buying an Isotek main strip it still in use .