• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Power grid interference: What does it look like measured with different tools?

RDacoustic

Member
Audio Company
Joined
May 19, 2021
Messages
13
Likes
46
Location
Roznov pod Radhostem, Czech Republic
Having experience from the field of electromagnetic interference in industrial applications, we incorporated HF noise filtration into our high end portfolio as a simple and essential means of increasing the distance between useful signal and noise. I would like to start a discussion about this kind of solutions: have you ever dealt with power grid interference, what is your experience/opinion?

We have several articles covering why we consider this important on our blog, I'll kick off with the latest one that also discusses the nanocrystalline materials used in filters, basic parts and some more background. If you're interested to see what power grid interference looks like, just scroll down and take a look at the videos where we measure with and without a filter using different tools.


A Tad Bit of Physics

The 19th century has brought a myriad of findings in the field of electricity. Before Nikola Tesla invented the electric engine and Graham Bell the telephone, in 1831, Michael Faraday described electromagnetic induction and so laid a foundation for the entire field of electrical engineering. Electromagnetic induction denotes a phenomenon when magnetic field occurs around a conductor due to current that flows through it.


Magnetic-field.svg_-300x123.png


Permeability

Magnetic field can be transferred using magnetically conductive materials. Permeability then is a kind of material constant the value of which depends also on other parameters, such as the size of the magnetic field, frequency, hysteresis, and others. Depending on permeability, we divide materials to paramagnetic materials and feromagnetic materials where μr is several times larger than 1. The race for the best magnetically conductive material could start. It was found that various alloys achieve higher permeability than iron or steel.


Molybdenum-300x200.png

Molybden

limonit_4-300x225.png

Iron ore

cobalt1-300x206.jpg

Cobalt

nickel-metal1-300x201.jpg

Nickel


Permalloy

Invented in 1914 by physicist Gustav Elmen at Bell Telephone Laboratories, permalloy is an alloy consisting of approximately 80 % nickel a 20 % iron. Permalloy generally achieves about 100x higher permeability compared to iron or steel. Supermalloy followed in 1948 as an alloy of 79 % Ni, 16 % Fe, and 5 % Mo.

The Present

Today, using complex simulation and thanks to developments in material engineering and superior manufacturing processes, nanocrystalline materials are used. Thousands of alloys with various specific properties are available. The range of magnetically conductive cores available is vast but as is usually the case in life, an absolute ideal for every use case is non-existent. One needs to find and test the most fitting material that will achieve the needed magnetic field intensity, it will be effective on the required frequencies, and will be accessible price-wise.

Our F4SX Power Filter

Our filter combines four filtration sections, which are designed to filter symmetrical and assymetrical interference in a wide frequency spectrum. The filter also includes surge protection against high energy pulses SURGE (lightning) and fast high frequency pulses BURST. We will not reveal the exact composition of the filter, but we will get you acquinted with the way it works. :)

Passive or Active Filtration?

What needs to be said first is that the phrase “active filter,” which one often hears in the audio field, is quite misleading. What it actually is is a switched-mode power supply that first “streamlines” the voltage to DC and then “compiles” a new sine wave from that DC voltage using pulse-width modulation. The modulation frequency ranges from 20 to 40 kHz and on each edge of the pulse, a large amount of further harmonic frequencies is created. Following this “active” part, there inevitably has to be a passive part present to at least partly suppress and filter out the spikes that occur on the sharp edges of the switching transistors.


img_vfd_41-300x215.jpg


Passive Filtration

Passive filtration stands firm on verified electrotechnical principles. In short, a range of different connection schemas can be implemented, each with its own advantages and disadvantages and with a set of basic parts. Each of the parts has its quality, efficiency, ideal working range, but also its parasitic properties, which can be only partly minimised in the manufacturing process. Let us have a brief look at the cornerstones of passive filters.


Choke coil (inductance) – accumulates energy in its magnetic circuit thanks to electromagnetic induction. Its parasitic property is internal resistance and redundant heat due to the flowing current. The stronger the wire used for the winding, the lower its internal resistance. The part however also becomes larger in size and more expensive.
Inductor_Symbols1-300x128.jpg
inductor1-300x198.jpg





Compensated choke coil – two inductors on a common magnetic circuit. If the magnetic circuit is well designed, it can cause the high frequency spikes of symmetrical interference to cancel each other out.


flux-compensated-inductor-292x300.png
compensated-inductor1-300x244.jpg





Capacitor (capacitance) – thanks to its capacity (dielectric) accumulates electrical energy. Every real capacitor has also parasitic properties, such as inductance and resistance.


capacitor-types1-300x59.jpg
capacitor1-300x300.jpg





Combining the basic parts, a connection schema working on a certain frequency can be created—such as an integrator, differentiator, LC resonant circuit and so on. Combining the parts in a specific way, we can even exploit some of their parasitic properties.


some-filter1-300x102.jpg


Interference Transmission in the Power Grid

A power grid is common for everybody in a given power circuit. If anybody’s devices cause interference at their home grid, that interference everybody else at the same transformer station circuit will have at their homes. Interference from all devices adds up. The biggest culprit here are switched-mode power supplies, converters, inverters, fluorescent tubes, which, apart from causing high frequency interference, also deform the sine wave shape (harmonic distortion of power, THD). Without going into details, the trend of increasing interference in the power grid is absolutely clear. Every appliance that causes interference should include an effective filter that eliminates its interference so as not to inhibit another device. The limits of tolerable interference are defined by EMC standards, but in practice, there is simply much more interference in the power grid than the standards allow for.

Why does power grid interference influence audio quality?

The easiest answer is that our hearing is very sensitive, having a range of 1:10-12, one to a trillion. The other significant cause is that high frequency components of interference have the ability to come through the power supply part of a device to the analog part. The more interference there is in the grid, the smaller the distance between signal and noise in the audio output.


Parasitic high frequency interference occurs on very high frequencies and because of this it acts in a completely different way than power grid voltage with the usual 50 Hz frequency. Parts and schemas that are designed for the usual frequency behave differently on higher frequencies. Mutual capacitance and inductance between conductors or circuit boards both have a much bigger effect; shielding is important. In filters then, it is essential to use the right material and parts designed specifically for one or another part of the high frequency spectrum.


High frequency interference occurs in the grid in the order of millivolts and is modulated on alternating current (230 V/50 Hz). It is not dangerous from the point of its size (except for short-term SURGE/BURST overvoltage), but it is mainly dangerous because of its character, due to which it can very easily influence useful signal, which is what we audiophiles try to avoid as much as possible.

Measuring Power Grid Interference

What needs to be said is that the amount of interference of course changes in time depending on the number and intensity of its sources, location, earthing quality, earthing impedance and power grid impedance. We measure in a given moment, in your surroundings, on the same power circuit, and the result depends on how many sources of interference are active and how much interference they produce.

EMI Line Meter

This is a simple non-calibrated gauge that is fully acceptable for home use. It measures in the frequency range of 10 kHz to 10 MHz. The high frequency interference is displayed in millivolts, the effective value of grid voltage in volts, and the device has an earphone using which the interference can be listened to.


Oscilloscope

Measures voltage values in time. Thanks to an oscilloscope, we can spot a deformation (harmonic distortion) of the sine wave as on the video below, caused predominantly by switched-mode power supplies that draw current from the top of the sine wave only. This deformation is not essential for audio quality. What is much more important here is what appears as tiny fringes, visible mostly on the peaks of the wave. These are the parasitic high frequencies we are most interested in.


More advanced oscilloscopes allow for the calculation of the frequency spectrum using a mathematical function called Fourier’s transformation (FFT) where the deformation due to uneven draw of current (only on the tops of the sine wave) is visible. The first peak is at 50 Hz and it should be the only one. Further peaks in the frequency spectrum are those higher harmonic components we do not wish to have. Comparing them we can calculate THD.


In the video below, there is captured a detail of the bottom peak of a power grid sine wave. The jagged, fringy part is what’s important. If we zoomed the graph in even more, on these fringes, further fringes would become visible, and so on. We will see this when measuring with a spectrum analyzer.


On the output of the F4SX filter, we can see the high frequencies completely suppressed. The energy of the high frequency interference is spread out to a much larger time base. The power supply parts of audio components will easily deal with such a rounded voltage waveform. The wave below is without load. With load (when current is flowing), the shape becomes much smoother.


Spectrum Analyzer

This is a device specifically constructed to analyse the frequency spectrum. For home use, it would be too expensive and because of its sensitivity, it can easily be damaged when not connected professionally. On axis X we can see frequency, on axis Y energy for a given frequency. In the bottom part, we can see the frequency spectrum and its immediate value in time, while the upper part shows a spectrogram that records the average of all measured values.


In the video, we can see power grid interference as captured by the spectrum analyzer. After the first 3 seconds, the output voltage of the F4SX audio filter is measured. We can notice a significant decrease in interference. 50 seconds later, the spectrum analyzer is connected back straight to the power grid and we see a significant increase in interference.
 

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,376
Likes
234,553
Location
Seattle Area
EMI Line Meter

This is a simple non-calibrated gauge that is fully acceptable for home use. It measures in the frequency range of 10 kHz to 10 MHz. The high frequency interference is displayed in millivolts, the effective value of grid voltage in volts, and the device has an earphone using which the interference can be listened to.
And that's the problem with it. The sounds it makes are not at all related to audible effect in electronics. It is down converting high frequency spectrum to audible band which does not occur in our stereos.
 

Propheticus

Senior Member
Joined
Nov 18, 2020
Messages
431
Likes
644
Location
Vleuten, Netherlands
Let's be frank, this is just (badly) concealed advertising. Right?
My snake oil alarm bells are going off. Some grounding in science, indeed doing something, but nothing that would be relevant for audible improvements.

We don't need before and afters of reduced Mhz range interference in the AC line. We need before and afters of the audio equipments output in the 20Hz-20KHz range.
 

Superdad

Active Member
Audio Company
Joined
Aug 6, 2017
Messages
149
Likes
91
Virtually the same mains noise "sniffer," showing here how a typical SMPS actually suppresses what comes out of the wall. ;)

And by the way, a typical power-factor-corrected Class VI desktop SMPS will distort the peaks of the AC mains far less than a standard transformer>diodes>capacitor>regulator linear power supply. So not sure what his graphs are showing in that regard.​
 

SIY

Grand Contributor
Technical Expert
Joined
Apr 6, 2018
Messages
10,386
Likes
24,752
Location
Alfred, NY
We don't need before and afters of reduced Mhz range interference in the AC line. We need before and afters of the audio equipments output in the 20Hz-20KHz range.

That is something they will never provide, at least any honest measurement. Mostly because it won't be useful for fleecing the gullible.
 

Blumlein 88

Grand Contributor
Forum Donor
Joined
Feb 23, 2016
Messages
20,524
Likes
37,057
Is this really much better than scanning the AM band with a handheld radio? Then supplementing that with some short wave bands.
 
OP
RDacoustic

RDacoustic

Member
Audio Company
Joined
May 19, 2021
Messages
13
Likes
46
Location
Roznov pod Radhostem, Czech Republic
Let's be frank, this is just (badly) concealed advertising. Right?
My snake oil alarm bells are going off. Some grounding in science, indeed doing something, but nothing that would be relevant for audible improvements.

We don't need before and afters of reduced Mhz range interference in the AC line. We need before and afters of the audio equipments output in the 20Hz-20KHz range.

Just give us some before and after measurements of decently working amp and DAC outputs.. Then we’ll talk.

Right. Well, the articles we write that I wish to post here are usually tied to a product of ours and I can't get rid of that link entirely, and yes, we want people to know what we make, but as for your hypothesis, in all fairness, I can hardly imagine this forum's members to be our target customer group. Likewise, if we liked magic saucers, this would not be the right place to present them to readers, vis-a-vis the responses or the very name of this forum. :)

Now, to the point: The reason to care about high frequencies is amplitude modulation and interference. High frequencies form low frequencies. The low frequency noise is what we want to suppress, because with that out of the way, details will stand out more and the recording will brighten up.

Most audio components already include a filter, but it is usually just a simple transformer in the source section. Due to its parasitic internal capacitance, high frequency interference still gets from the source part to the signal. Most electronics then address this with a strong feedback loop, which diminishes the detail we were looking for in music in the first place. Interference that occurs at the switching diodes already in the DC part for example, is usually not addressed at all if only because the solution would have to be large and expensive...

How much does this matter? It depends. Interference differs by your location and region. Sensitivity of the audio system and the speaker drivers differs also. In our experience with EMC norms in place in Europe and how they are respected, it absolutely matters, and a standalone filter is the best way to deal with this. And then there is our audio chain, where we go for high sensitivity, so (given the sensitivity of the human ear) the change is audible. If you use equalization for example, doing this kind of a thing will loose sense of course.

Truly, I'm not sure of an easy method to measure this and show it clearly at amp outputs. The best way to go about this would probably be blind tests. We will be glad to welcome anybody who wishes to try. :) What we did do in our listening room (good grounding and our own power circuit installations) is measure the background noise coming through a lower cost PrimaLuna amplifier with and without the filter. https://rdacoustic.cz/en/blog/2020/08/01/background-noise-in-audio/ The difference was some 3 dB, which is significant taking into account that dB is a logarithmic unit.
 
OP
RDacoustic

RDacoustic

Member
Audio Company
Joined
May 19, 2021
Messages
13
Likes
46
Location
Roznov pod Radhostem, Czech Republic
And that's the problem with it. The sounds it makes are not at all related to audible effect in electronics. It is down converting high frequency spectrum to audible band which does not occur in our stereos.

There is no frequency conversion involved (if I understand you right). How would that work? The noise is amplified, amplitude modulation is why we are concerned with high frequencies that we cannot hear.
 

SIY

Grand Contributor
Technical Expert
Joined
Apr 6, 2018
Messages
10,386
Likes
24,752
Location
Alfred, NY
There is no frequency conversion involved (if I understand you right). How would that work? The noise is amplified, amplitude modulation is why we are concerned with high frequencies that we cannot hear.

Do you have data showing this effect in the audible range with actual commercial engineered amplifiers?
 

AndrewC

Member
Joined
Jun 24, 2016
Messages
25
Likes
59
Location
Singapore

voodooless

Grand Contributor
Forum Donor
Joined
Jun 16, 2020
Messages
10,229
Likes
17,812
Location
Netherlands
https://rdacoustic.cz/en/blog/2020/08/01/background-noise-in-audio/ The difference was some 3 dB, which is significant taking into account that dB is a logarithmic unit.

I asked for a decent amp, not some ancient tube technology.. these things probably have little to no feedback and consequently bad PSRR. Speced snr is like 92 dB, which is already subpar. Measured snr by Stereophile was way worse: between 68 and 77 dB. That is just bad.

And we didn’t event start about the measurement method.
 
OP
RDacoustic

RDacoustic

Member
Audio Company
Joined
May 19, 2021
Messages
13
Likes
46
Location
Roznov pod Radhostem, Czech Republic
Virtually the same mains noise "sniffer," showing here how a typical SMPS actually suppresses what comes out of the wall. ;)

And by the way, a typical power-factor-corrected Class VI desktop SMPS will distort the peaks of the AC mains far less than a standard transformer>diodes>capacitor>regulator linear power supply. So not sure what his graphs are showing in that regard.​

No, it certainly does not. I would need to know the grid/grounding point impedance to be able to say what's happening in the video. My guess is that if the impedance is high (high resistance), what the line meter shows will be influenced by the SMPS's filter. Since the SMPS is not in load, only its filter that is connected right next to the line meter applies – due to high impedance in the grid, HF interference goes down in the part of the spectrum that the line meter is measuring. If we tried the same with a spectrum analyzer, the effect of the SMPS filter would be very small all over the spectrum. It's a hard to measure thing because the interference has different character in every part of the spectrum...

Anyway, line meters visualize only a fraction of the interference that influences audio and can be intercepted by our ears. They are just an aid that measures a limited frequency spectrum, i.e. only "something". There are more influences that we solve by filtering: the DC current component (loads transformers, causes overheating and therefore hum if the transformers are not vacuum impregnated), deformations in the voltage sine (filtration of high frequencies and their harmonic components), HF interference and last but not least, absolute surge protection.
 

Propheticus

Senior Member
Joined
Nov 18, 2020
Messages
431
Likes
644
Location
Vleuten, Netherlands
Anyway, line meters visualize only a fraction of the interference that influences audio and can be intercepted by our ears.
Blah blah. So when you use one to demonstrate something it's okay, but if someone else does it's suddenly invalid?

Truly, I'm not sure of an easy method to measure this and show it clearly at amp outputs.
I know one: do an AP test, heck even a simpler RMAA loop test on a normal* amplifier or AVR. Show a suite of tests, primarily noise and THD+noise (again, 20Hz to 20kHz range to proof low frequencies were formed out of high frequencies). Once with your AC-line EMI filter added, once without**.
You've got a fancy oscilloscope, you can also measure an amps output (pre-out or speaker terminal out, don't care) when you feed it a steady 1kHz test tone, and compare the output sine with and without AC-filter in the chain.

Edit: even simpler....send one in for testing by Amir.

*no crazy boutique stuff or ancient tube tech. e.g. take a regular Denon AVR or NAD amp or equivalent.
**In actuality most decent equipment already comes with an EMI filter in the C14 power socket. So it's more like testing with 2 EMI filters, or with 1.
1627585178623.png
 

Cbdb2

Major Contributor
Joined
Sep 8, 2019
Messages
1,530
Likes
1,485
Location
Vancouver
No reply from rd? No surprise. Must have taken his BS to the more gullible.

There wrong conclusion says it all

"Conclusion

Some might say “Oh, but 3 dB don’t mean anything.” We are however concerned with the smallest details that brighten the recording up, make it cleaner. The noise distance difference with RD EMI Neutralizer is hearable here and the overall dynamics of the instruments becomes more pronounced."

If my amp noise floor drops 3db from inaudable to inaudable ( then I play some music with an even higher noise floor) those filters will do nothing but drain my bank account, and I don't even know the ridiculous price. And wtf is "noise distance"!
 
Last edited:

Cbdb2

Major Contributor
Joined
Sep 8, 2019
Messages
1,530
Likes
1,485
Location
Vancouver
Uh-oh.... looks like another mains isolation transformer with all the safety concerns that come along with those... https://rdacoustic.cz/en/high-end-audio/rd-emi-neutralizer-power-filter/

RD_EMI_Neutralizer4-scaled-1-1240x800.jpg
And the wood enclosure on a RF filter screams of audiophool. And don't toroids leak more RF from primary to secondary because of there higher interwinding capacitance? But they don't look as cool. And why am I paying for those fancy transformer covers nobody sees? But then we woudnt have this fancy picture, and everybody knows you buy audio gear with your eyes, and marketing is much more important than function.
 

AnalogSteph

Major Contributor
Joined
Nov 6, 2018
Messages
3,338
Likes
3,278
Location
.de
Nothing says "audiophile" like a cheesy cursive font... although to be fair, this one seems more tasteful than the '70s Marantz style that is still occasionally found (ugh). I wonder what a design pro's take on the matter would be, but cursive generally just doesn't have a place on technical equipment IMHO.
 

Audiofire

Addicted to Fun and Learning
Forum Donor
Joined
Feb 8, 2022
Messages
625
Likes
347
Location
Denmark
High frequency interference occurs in the grid in the order of millivolts and is modulated on alternating current (230 V/50 Hz). It is not dangerous from the point of its size (except for short-term SURGE/BURST overvoltage), but it is mainly dangerous because of its character, due to which it can very easily influence useful signal, which is what we audiophiles try to avoid as much as possible.
Audiophiles should worry much more about impedance, since that is what can cause voltage sags in the power grid. AC from wall sockets is normally really good/low impedance, that is why electric motors that are used in a lot of equipment can be run from it (people in the construction industry generally don't need expensive power conditioners for their power tools for example). It is however possible that you have bad impedance where you live, but that should not be considered as a normal/acceptable situation.

So AC regenerators, especially an uninterruptible power supply often do not give an audible improvement. If the music sounds lighter and is missing bass with a power conditioner, it is most likely because it has worse/higher impedance.

All you had to do was record some music from the line level output of hi-fi equipment with and without a power conditioner, but the reason we don't see that is because decent hi-fi equipment smooths out the DC and removes the AC interference (called ripple current) by using capacitors or some other electronic component.

You're probably better off upgrading your capacitors than buying some questionable AC regenerator. Just know that they can store enough energy to kill you, know what you are doing before capacitors are replaced.

But please use a surge protector as voltage spikes can actually be an important problem for a lot of capacitors. AC interference is not necessarily important. You don't have to believe in me, because Paul McGowan from PS Audio that makes the expensive power plants claims the same in the video below.

Virtually the same mains noise "sniffer," showing here how a typical SMPS actually suppresses what comes out of the wall. ;)
I don't think that looks normal at all. There could be something seriously wrong with the grounding of his house wiring. SMPS should be very noisy, above 75 watts is where some have power factor correction that is not noisy. Does anybody have a better explanation?

How important is clean power?

 
Last edited:
Top Bottom