Measuring power line noise with the TriField EM100, and the impressive effect of Shyunyata products

[ack]

Just so you know, while I can see that device being very handy, it is limited to 2MHz and a lot of offending noise happens well above that. For example if you have a cheap class D plate amp in a subwoofer, it can make noise well into the FM band (even though its marked with the CE mark; the manufacturer might be using an old technique called 'lying'). If that noise gets rectified by something in your setup it can still become audible, even though the Trifield says it isn't there.

Specs say 10kHz to 10MHz ("+/- 8% @ 20 kHZ to 2 MHz; within a factor of 2 from 10 kHz to 10 MHz"), identifies noise from my desktop computer's SMPS and of course the lights. And I can confirm that it doesn't detect anything below 10kHz because I measured one of those MIT Magnum Z-Traps which are add-ons to power cords, and it told me it didn't reduce noise, and I know it does in terms of 50/60 Hz harmonics (6 filter poles altogether for 6 harmonics)

 

[Atmasphere]

Specs say 10kHz to 10MHz ("+/- 8% @ 20 kHZ to 2 MHz; within a factor of 2 from 10 kHz to 10 MHz"), identifies noise from my desktop computer's SMPS and of course the lights. And I can confirm that it doesn't detect anything below 10kHz because I measured one of those MIT Magnum Z-Traps which are add-ons to power cords, and it told me it didn't reduce noise, and I know it does in terms of 50/60 Hz harmonics (6 filter poles altogether for 6 harmonics)

Pardon me- yes they do. And 10MHz isn't nearly high enough.

 

[ack]

And 10MHz isn't nearly high enough.

I would agree with that. I am counting on the Denali in this case to cover anything higher:

Noise Suppression
Input to Output (100 kHz – 30 MHz): > 50 dB reduction
Zone to Zone (100 kHz – 30 MHz): > 60 dB reduction

I guess it does not filter FM frequencies...

 

[ack]

@Kingrex So it looks like your recommended MA-1000 Lutron is really for incandescent lights



The ones I have are for the electronic 12V transformers... and as it turns out, one of the dimmers I replaced was the old MAW-600 which was adding some noise (see dimmer-vs-rocker-vs-new-dimmer test). Their dimmers are improving, but the fundamental problem is those pesky noisy electronic transformers. I am hunting for compact low-noise replacements...

 

[Atmasphere]

I would agree with that. I am counting on the Denali in this case to cover anything higher:

Noise Suppression
Input to Output (100 kHz – 30 MHz): > 50 dB reduction
Zone to Zone (100 kHz – 30 MHz): > 60 dB reduction

I guess it does not filter FM frequencies...

I had one of those cheap subs I mentioned in my bedroom system and I could pick it up on my FM tuner easily enough. It made FM reception terrible both in my bedroom system and my main system. I thought I was having bad multipath problem. When I replaced it with a better sub, the FM reception was vastly improved!

I found that is was responsible for noise in my main system too (not on FM), which is on a different floor and circuit.

 

[ack]

I had one of those cheap subs I mentioned in my bedroom system and I could pick it up on my FM tuner easily enough. It made FM reception terrible both in my bedroom system and my main system. I thought I was having bad multipath problem. When I replaced it with a better sub, the FM reception was vastly improved!

I found that is was responsible for noise in my main system too (not on FM), which is on a different floor and circuit.

Ditto here, I can pick up interference on my FM when I turn on the computer; it's also why I never bothered to upgrade my LPS-based sub all these years

 

[Atmasphere]

Ditto here, I can pick up interference on my FM when I turn on the computer; it's also why I never bothered to upgrade my LPS-based sub all these years

Just FWIW, that stuff isn't supposed to do that

I replaced the cheap Dayton Audio sub with an older M&K sub I found on Craigslist for $150.00. It was vastly better in every way. And I got my FM back.

 

[adyc]

Get the Trifield meter today. Test out the noise in my circuit. It is not great.



However, I have power conditioner Eurobuss

https://piaudiogroup.com

The noise reduction is very effective.



Furthermore, this power conditioner not just reduce the noise of the outlets of the power conditioner. All the outlets parallel to this device are also reduced to similar readings. Very impressive. I do not claim it sounds better than Shunyata products as I do not have Shunyata products to compare.

 

[K3RMIT]

I suspect there is a line noise filter (small cap) near the low noise outlet. However resonances in the power network could be amplifying the noise. The connection to earth/ground will have little effect on the noise (except when its helping the antenna properties of the wiring). Romex is pretty lossy above 30 MHz but plenty good below. Every mismatch or connection in the cabling (junction box or breaker box) is a place where the transients can radiate from the wiring. Steel conduit is the best for controlling radiation and noise but not used much in residential construction.

You can have issues with separate power runs from the main panel for amps vs the other electronics. Biggest problem if they are on different phases ,that would mean 240V between hot leads and lots of potential leakage. Second largest is if the chassis grounds are different returns and develop a voltage drop, another source of potential leakage and hum+noise. The ground return even with no current can develop a voltage from its close proximity to the power conductors. If its not perfectly equidistant to them so the mag fields cancel you get a 1 turn transformer. This is a real problem in a surge event since all three conductors will have a lot of voltage on them. And also for broadband noise coupling.

Does typical bx cable or steel armor MC cable count for metal tubing ?

 

[K3RMIT]

So, for the last few days I've been measuring noise over and over again, and the only culprit is the dimmable lights. The 120V dimmable LED flood lights can be replaced with the older incadescent style I had for years, if I can find them; but not worried about that, they are in the basement. The real problem is the 12V electronics transformers, and although I can just turn them off when listening, I'd rather find good old analog transformers...

So, does any know where to find analog 12V transformers but compact in size to fit inside light fixtures, 50W-60W? I have s
Your friend has shared a link to a Home Depot product they think you would be interested in seeing.

https://www.homedepot.com/p/Progress-Lighting-150-Watt-Landscape-Lighting-Transformer-P8

So, for the last few days I've been measuring noise over and over again, and the only culprit is the dimmable lights. The 120V dimmable LED flood lights can be replaced with the older incadescent style I had for years, if I can find them; but not worried about that, they are in the basement. The real problem is the 12V electronics transformers, and although I can just turn them off when listening, I'd rather find good old analog transformers...

So, does any know where to find analog 12V transformers but compact in size to fit inside light fixtures, 50W-60W? I have searched high and low...

Transformers make noise too I think , have you tried using a switch and check for noise. Also dim them and check

 

[ack]

Get the Trifield meter today. Test out the noise in my circuit. It is not great.

View attachment 78793

However, I have power conditioner Eurobuss

https://piaudiogroup.com

The noise reduction is very effective.

View attachment 78794

Furthermore, this power conditioner not just reduce the noise of the outlets of the power conditioner. All the outlets parallel to this device are also reduced to similar readings. Very impressive. I do not claim it sounds better than Shunyata products as I do not have Shunyata products to compare.

Very nice. The manual says below 50 is low noise. But you are now enabled to investigate and address, which is the point of this thread

 

[wil]

I have the same meter branded as Alpha Labs. I also have the Entech device which according to Caelin G. of Shunyata, measures a narrower band which is more applicable to noise which effects audio quality.

 

[adyc]

Very nice. The manual says below 50 is low noise. But you are now enabled to investigate and address, which is the point of this thread

Found the culprit. It is my QNAP NAS producing huge amount of nosies. Although NAS is far away from listening room, it is affecting the outlets in my listening room. Surprisingly, all the modems or routers do not produce much noise. Also my 2014 iMac is also pretty quiet.

 

[ack]

A measurable case for star-grounding?

Let me see if I can articulate this experiment well enough. It goes like this:

1) First, plug most components into the Denali which is fed by Wall Outlet1, but one Denali outlet feeds an MIT Z-Strip1 Distributor; MIT Z-Strip1 hosts the FM Tuner and TriField EM100 device; all audio components are effectively star-grounded, but not exactly 100% because there is this MIT Z-Strip1 and the FM Tuner's ground path is slightly longer

2) All noise measurements are done on this Z-Strip1

3) Measure noise on MIT Z-Strip1; it reads ~25mVp-p

4) Move the FM Tuner to another MIT Z-Strip2, fed by an adjacent Wall Outlet2 on the same circuit - no Denali there

5) We have clearly broken star-grounding by moving the FM Tuner off of Outlet1, and now the relative ground path length of the FM Tuner is much longer than any other component's

6) Measure again noise on MIT Z-Strip1; it reads ~30mVp-p; unplug the FM Tuner from Z-Strip2, and Z-Strip1 reads again ~25mVp-p

7) Repeat by moving the FM Tuner back and forth between Z-Strips, and the readings on Z-Strip1 are consistent all the time: ~25 vs ~30

Q: Why the experiment with the FM Tuner?
A: It's the noisiest of all of my components because of its poorer power supply design

Q: Why the need for Z-Strip1? Why not plug the FM Tuner and EM100 into two different Denali outlets?
A1: I wanted the FM Tuner to always be plugged into the same kind of distributor - MIT Z-Strips
A2: I wanted to put as many of my components directly on the Denali as I can, to have as many ground paths as possible, and also mimic every-day configuration

To summarize, it goes like this:

1) Plug all components as follows:
Outlet1 --> Denali --> [Preamp, Phono, DACs, [MIT Z-Strip1 distributor --> FM Tuner, EM100]]
Outlet2 --> [MIT Z-Strip2 distributor --> (nothing)]
Noise reading on MIT Z Strip1: ~25mVp-p

2) Move the FM Tuner to another adjacent outlet and second MIT Distributor; both outlets are on the same circuit, but I suspect one has a longer ground path than the other:
Outlet1 --> Denali --> [Preamp, Phono, DACs, [MIT Z-Strip1 distributor --> EM100]]
Outlet2 --> [MIT Z-Strip2 distributor --> FM Tuner]
Noise reading on Z-Strip1: ~30mVp-p (FWIW, noise on Z-Strip2 is higher)

I suspect this experiment measurably demonstrates the value of star-grounding, though the measured difference in noise on the same Z-Strip1 is rather small. But I'd like to hear counter-arguments or other comments that would break down my theory. Equally important, can someone else confirm or deny the results of such an experiment, with the understanding that all of our systems are different - we are not exactly trying to write a scientific paper here...

-ack

 

[Marcus]

Get the Trifield meter today. Test out the noise in my circuit. It is not great.

However, I have power conditioner Eurobuss

The noise reduction is very effective.

Furthermore, this power conditioner not just reduce the noise of the outlets of the power conditioner. All the outlets parallel to this device are also reduced to similar readings. Very impressive.

For even better results, try the Buss Depot which works in parallel(!!) and suppress noise on the whole line...

Off:


On:

 

[Lee]

I have not done these measurements but the addition of an Everest greatly lowered the noise floor in my system.

 

[Kingrex]

A measurable case for star-grounding?

Let me see if I can articulate this experiment well enough. It goes like this:

1) First, plug most components into the Denali which is fed by Wall Outlet1, but one Denali outlet feeds an MIT Z-Strip1 Distributor; MIT Z-Strip1 hosts the FM Tuner and TriField EM100 device; all audio components are effectively star-grounded, but not exactly 100% because there is this MIT Z-Strip1 and the FM Tuner's ground path is slightly longer

2) All noise measurements are done on this Z-Strip1



3) Measure noise on MIT Z-Strip1; it reads ~25mVp-p

4) Move the FM Tuner to another MIT Z-Strip2, fed by an adjacent Wall Outlet2 on the same circuit - no Denali there

5) We have clearly broken star-grounding by moving the FM Tuner off of Outlet1, and now the relative ground path length of the FM Tuner is much longer than any other component's

6) Measure again noise on MIT Z-Strip1; it reads ~30mVp-p; unplug the FM Tuner from Z-Strip2, and Z-Strip1 reads again ~25mVp-p

7) Repeat by moving the FM Tuner back and forth between Z-Strips, and the readings on Z-Strip1 are consistent all the time: ~25 vs ~30

Q: Why the experiment with the FM Tuner?
A: It's the noisiest of all of my components because of its poorer power supply design

Q: Why the need for Z-Strip1? Why not plug the FM Tuner and EM100 into two different Denali outlets?
A1: I wanted the FM Tuner to always be plugged into the same kind of distributor - MIT Z-Strips
A2: I wanted to put as many of my components directly on the Denali as I can, to have as many ground paths as possible, and also mimic every-day configuration

To summarize, it goes like this:

1) Plug all components as follows:
Outlet1 --> Denali --> [Preamp, Phono, DACs, [MIT Z-Strip1 distributor --> FM Tuner, EM100]]
Outlet2 --> [MIT Z-Strip2 distributor --> (nothing)]
Noise reading on MIT Z Strip1: ~25mVp-p

2) Move the FM Tuner to another adjacent outlet and second MIT Distributor; both outlets are on the same circuit, but I suspect one has a longer ground path than the other:
Outlet1 --> Denali --> [Preamp, Phono, DACs, [MIT Z-Strip1 distributor --> EM100]]
Outlet2 --> [MIT Z-Strip2 distributor --> FM Tuner]
Noise reading on Z-Strip1: ~30mVp-p (FWIW, noise on Z-Strip2 is higher)

I suspect this experiment measurably demonstrates the value of star-grounding, though the measured difference in noise on the same Z-Strip1 is rather small. But I'd like to hear counter-arguments or other comments that would break down my theory. Equally important, can someone else confirm or deny the results of such an experiment, with the understanding that all of our systems are different - we are not exactly trying to write a scientific paper here...

-ack

Are you landing the interconnect on the tuner to the preamp?. Or just plugging it into a power outlet.

 

[ack]

Are you landing the interconnect on the tuner to the preamp?. Or just plugging it into a power outlet.

All components are plugged into the preamp as usual, therefore the system is unperturbed. This is all about moving the power cables around.

 

[Kingrex]

All comports are plugged into the preamp as usual, therefore the system is unperturbed. This is all about moving the power cables around.

How about trying the same thing with a different powet cable to the tunner . Like one of your NR cables.

 

[ack]

How about trying the same thing with a different powet cable to the tunner . Like one of your NR cables.

So the tuner already has an NR cable