• 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!

Puritan Audio PSM156 Review (AC Filter)

sarumbear

Master Contributor
Forum Donor
Joined
Aug 15, 2020
Messages
7,604
Likes
7,314
Location
UK
Well, I measured the AC mains with an oscilloscope and the only noticeable peak was that 8 kHz one. I measured up to 20 MHz if I remember correctly. I also think that it is quite strange, 8 kHz noise couldn't radiate as easily from a power cord.

Likewise, the ferrite choke didn't help at all, the noise is there with or without it.
It is indeed strange. I never heard a case where there’s an 8kHz RF signal nor it is transmitting through air. As I said the wavelength of that signal is 35km. Where’s the antennae? Even at 1/1000th wavelength it should be 35 metres long!
 
OP
amirm

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,368
Likes
234,389
Location
Seattle Area
This is the schematics of Schaffner FN2030-4 line filter. Current rated up to 4A.
Did you bother reading their site with respect to impedance used to measure transfer function? I assume not despite the fact that I pointed that out to you. Here is their white paper on "50" ohm test per CISPR 17: https://www.schaffner.com/fileadmin...note/Schaffner_AN_CISPR17_Measurements_E8.pdf

1630349934412.png


They go no to say:

1630350161349.png


So no wonder the "industry" likes to use 50 ohm, damn the reality of it.

It was disingenuous of @KSTR for not mentioning this and instead saying, "It's not immediately clear if Amir's insertion loss measurement used the industry standard input and output termination impedance of 50Ohms? If not, it is not really useful. "

Not really useful? Following an industry practice that uses fictitious impedances is useful instead?

Even if we cared to follow the industry to sell such products, these measurements are not aimed at audio frequencies. The lowest range specified in the standard is 9 kHz. Again from the same doc:

1630351300500.png


See how the lowest frequency is 10 kHz (often written as 0.01 Megahertz) and goes up to 1 MHz? They do that because the intent is to get rid of emissions that cause you fail regulatory certification and interfere with radio operation. Not because they are trying to examine efficacy for filtering noise in audio band. If I had followed the "standard" you would not see anything below 9 kHz which would make such measurements useless.

And it is not like you are given any such measurements from the companies whose products I am testing. Where do you find a single graph showing the transfer function? If they had such, sure, we would try to match their test conditions to verify. But in absence of that, using any test means is fine as long as we are consistent and are using it as a first order approximation of whether they device is doing any filtering and if so, by what order magnetite.

So stop following someone else's futile attempt to justify these boxes without understand what it means. You are filling all these review threads with nonsense of no value....
 

pma

Major Contributor
Joined
Feb 23, 2019
Messages
4,591
Likes
10,727
Location
Prague
Effect of generator impedance (0 ohm vs. 50 ohm) to HF mains filter frequency response (measurement).

mainsfilter_generatoreffect.png


Similar saddle as with 50 ohm (brown) one can see in the post #1.
 
D

Deleted member 30699

Guest
Well, I measured the AC mains with an oscilloscope and the only noticeable peak was that 8 kHz one. I measured up to 20 MHz if I remember correctly. I also think that it is quite strange, 8 kHz noise couldn't radiate as easily from a power cord.

Likewise, the ferrite choke didn't help at all, the noise is there with or without it.

At these frequencies magnetic coupling is more likely than RF propagation. If correct, putting the power cord orthogonal to the RCA cable should solve it. Grounding the turntable should also help.
We can probably help more if you happen to know where the 8kHz signal originates.
 

sarumbear

Master Contributor
Forum Donor
Joined
Aug 15, 2020
Messages
7,604
Likes
7,314
Location
UK
Effect of generator impedance (0 ohm vs. 50 ohm) to HF mains filter frequency response (measurement).

View attachment 150491

Similar saddle as with 50 ohm (brown) one can see in the post #1.
It’s obvious you are trying to prove something but you are not doing a good job. Why would the Q of a filter will matter in the context we are discusing?
 
D

Deleted member 30699

Guest
There are a few things at play here.
The most important thing here is that aside from voltage drop and flattening of the AC voltage we must realize the power is not coming from the voltage but from the drawn current.

This argument is deceptive but unfortunately wrong. So deceptive that even IEEE (the guys defining US grid standards) thought this and made this the norm for a long time.
Latest research shows that current harmonics are largely inconsequential _as_long_as_ you don't distort the grid voltage. This is relevant _because_ this means that you don't affect other devices on the same network.
IEC (the guys defining EU grid standards) have already adopted it and IEEE is largely following suit.
 

sarumbear

Master Contributor
Forum Donor
Joined
Aug 15, 2020
Messages
7,604
Likes
7,314
Location
UK
At these frequencies magnetic coupling is more likely than RF propagation. If correct, putting the power cord orthogonal to the RCA cable should solve it. Grounding the turntable should also help.
How can magnetic coupling work on non-magnetic elements? Audio cables are made of copper or silver, neither has magnetic properties. Not to mention that you need huge energy to oscillate a magnetic element. Does MRI rings a bell?
 

sarumbear

Master Contributor
Forum Donor
Joined
Aug 15, 2020
Messages
7,604
Likes
7,314
Location
UK
Latest research shows that current harmonics are largely inconsequential _as_long_as_ you don't distort the grid voltage.
What you say is against Ohm’s law.
1- Harmonic are part of the distortion.
2- If there’s distortion on the voltage, there will be distortion on the current as well.
 
D

Deleted member 30699

Guest
How can magnetic coupling work on non-magnetic elements? Audio cables are made of copper or silver, neither has magnetic properties. Not to mention that you need huge energy to oscillate a magnetic element. Does MRI rings a bell?
Electromagnetism?
Every cable has a magnetic field (albeit small) and can couple with every other cable.
 
D

Deleted member 30699

Guest
What you say is against Ohm’s law.
1- Harmonic are part of the distortion.
2- If there’s distortion on the voltage, there will be distortion on the current as well.
Since when is the grid resistive?
.. and no it is not, you can draw a pulse current from a DC powe supply...
 

solderdude

Grand Contributor
Joined
Jul 21, 2018
Messages
15,891
Likes
35,912
Location
The Neitherlands
This argument is deceptive but unfortunately wrong

It's neither deceptive nor wrong.

Latest research shows that current harmonics are largely inconsequential

The 'flattening' of the top of the sine wave is caused by devices having a rectifier. What will be defined is how much flattening is allowed.
And no this is not a problem.
 
Last edited:

sarumbear

Master Contributor
Forum Donor
Joined
Aug 15, 2020
Messages
7,604
Likes
7,314
Location
UK
Electromagnetism?
Every cable has a magnetic field (albeit small) and can couple with every other cable.
Electromagnetism or electromagnetic (EM) radiation is radio frequency (RF) hence what I explained above stands.
 

ShinMolina

Active Member
Joined
Jun 4, 2021
Messages
113
Likes
98
Location
Madrid
At these frequencies magnetic coupling is more likely than RF propagation. If correct, putting the power cord orthogonal to the RCA cable should solve it. Grounding the turntable should also help.
We can probably help more if you happen to know where the 8kHz signal originates.
Yeah, the crossing angle of the cables is important of course. I'm limited because the cables come parallel at the back, I've done what I can. The turntable is grounded and all possible ground loops have been avoided.
Could your 8khz noise be cooling from a usb dac conection somehow ?
The noise was there even before I incorporated a DAC into my setup. The problematic chain is turntable -> preamp -> amplifier -> speakers. And I'm sure that the noise enters the system through the RCA cables between the turntable and the preamp.

Thanks to everyone for digging into this issue, but I considered it solved long ago. The current level of noise is low enough to be labeled as background. I'm only curious into finding if a filter like the measured on this thread or a cheaper one like many have shown could solve issues like this that present real listenable artifacts.
 
Last edited:

sarumbear

Master Contributor
Forum Donor
Joined
Aug 15, 2020
Messages
7,604
Likes
7,314
Location
UK
Last edited:
OP
amirm

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,368
Likes
234,389
Location
Seattle Area
It’s obvious you are trying to prove something but you are not doing a good job. Why would the Q of a filter will matter in the context we are discusing?
He doesn't know because if he knew, he would see that such "filters" can actually amplifier input noise & harmonics depending on your power supply impedance! Yes, make it worse than better. He seems to think running a trivial spice simulation makes him look good while digging a hole for himself in not realizing what he is posting.
 

LTig

Master Contributor
Forum Donor
Joined
Feb 27, 2019
Messages
5,760
Likes
9,442
Location
Europe
How can magnetic coupling work on non-magnetic elements? Audio cables are made of copper or silver, neither has magnetic properties. Not to mention that you need huge energy to oscillate a magnetic element. Does MRI rings a bell?
I think it could couple into the pickup. Might be interesting to check whether an RCA cable terminated with a 4.7 kOhm (for MM; for MC rather 20 Ohm) at the open end also picks up the 8 kHz noise.
 
D

Deleted member 30699

Guest
Electromagnetism or electromagnetic (EM) radiation is radio frequency (RF) hence what I explained above stands.
By your definition ever electric motor or overhead line is a RF device.
Science tends to distinguish between high frequency (rf, emc, etc) and low frequency. While Maxwell's laws apply to both, different mechanisms are dominant.
8kHz is low frequency.
 

MC_RME

Addicted to Fun and Learning
Technical Expert
Audio Company
Joined
May 15, 2019
Messages
854
Likes
3,564
I'm only curious into finding if a filter like the measured on this thread or a cheaper one like many have shown could solve issues like this that present real listenable artifacts.

Easy to check for near no money:

https://www.aliexpress.com/item/32880545504.html?spm=a2g0s.9042311.0.0.27424c4dYIvajh

https://www.amazon.com/Filtering-Frequency-Two-stage-Low-pass-Optional/dp/B07T55RCLW?th=1

But only recommended for people who know what they do. High chance of getting electrocuted when fiddling with this stuff!
 
Top Bottom