Puritan Audio PSM156 Review (AC Filter)

[Yevhen]

Why the resistance will change with load. You have not answered that.

A real part of the load impedance (resistance) will change with the current consumption because of Ohm's law. You have constant voltage 110 or 220V, so with different currents filter will "see" different equivalent resistance of the load

 

[Yevhen]

I see that you have come here like a storm! Just joking.

First of all, I don't challenge Amirm's measurements. I find them quite valuable and have motivated me to star doing my own measurements. I consider myself a "deaf" person, in the sense that I unable to hear some things that people could consider obvious (compressed/uncompressed, tube warmth, digital filters...).

I also don't consider that Amirm is measuring the wrong parameters or that his measurements point us to a wrong direction. I believe that these kind of measurements help people understand that not all claims by the manufacturer should be true by default. By no means I am someone who would consider buying something like this, my most precious and most expensive device I have is a SMSL DA-9 I bought a month ago (250$). I can't think about buying something like this because of money reasons, and even if I could I wouldn't.

And about the noise I have between my AC line and interconnects, I have also measured it with an oscilloscope and the noise is there. It comes from the elevator of my building, when it is moving the noise appears. I can assure you the I have minimal 50 Hz noise in my setup as I always try to use balanced cables and avoid ground loops. I started noticing the 8 kHz tone while using some "audiophile grade" RCA cables that were unshielded, even with Mogami RCA cables the noise appears although much lower.

What I was selfishly thinking was for a setup to introduce these kind of noises into a test setup. Everyone can dream, I just brought possible scenarios to the table. By no means I want Amirm to change his measuring system, I have learned a lot from his findings and the comments of many people around this forum.

Interesting. Do you think you pick up that elevator noise via AC mains or through the air on RCA interconnects? Or maybe via GND?

 

[sarumbear]

A real part of the load impedance (resistance) will change with the current consumption because of Ohm's law. You have constant voltage 110 or 220V, so with different currents filter will "see" different equivalent resistance of the load

You are mistaken. The resistance of the filter changes due to the filter setup. Current has no effect. Write the formulae and see it yourself.

 

[Yevhen]

10 ohm is unrealistic, it would make 23A at 230V.

I'm just varying it between 10R and 1k, but indeed 10R is extreme.

 

[ShinMolina]

Interesting. Do you think you pick up that elevator noise via AC mains or through the air on RCA interconnects? Or maybe via GND?

I have already tested it multiple times. It is radiated noise from every mains cable near and appears stronger the closer the circuit is to the elevator. If any AC line goes close to an interconnect the noise appears. I forgot to say that the only link affected that causes the noise to be listenable is the interconnects from my turntable to my phono preamp since the latter amplifies the signal coming from the RCA by a large factor.

I managed to get around the problem by keeping my cables separated and using properly shielded RCA cables. I can even hear the 8 kHz noise if I put my ear close to the distribution panel of my house, but not as close as to touch the 220 V with my ear. I believe this could be due to the cables vibrating. I have contacted the elevator company but they have rejected every of my problems. They claim that the noise is within the allowed range.

 

[Deleted member 30699]

From the video I can see they have 3rd order filter. If I look at the guts of the 16A similar filter from Schaffner, I get following:

View attachment 150282
So supression is load-dependent, although a bit less than I expected.

In practice, you always dampen the resonances. Sometimes the inductor ESR is used.

 

[Deleted member 30699]

Not sure, the amp should behave more or less as an active load (resistor) if the cos Phi of the power supply is close to 1

Switched mode power supplies are everything but a resistive load. The power factor is tuned to be sufficiently high in steady-state operation, where you could use a resistor as equivalent circuit.
However, this equivalent circuit does not work well at higher frequencies and does not work at all for dynamics.
Same is true for speakers...

 

[restorer-john]

2. If you are not experiencing such interferences, then my measurements are completely applicable and show that these filters do not improve your audio performance.

DC offset can be present one day and not another. Transformer physical buzz in a toroid is not subtle. Transient line related issues are by their vey nature transient and unpredictable. Various intererferences manifest at all different times.

A sensible, effective line filter, surge protector and DC blocker is common sense for an expensive audio system.

 

[restorer-john]

@pma, you have been a major pain to me for a long while. You have dismissed my comments previously because in your opinion it is "philosophical" and not facts and test results. Instead of taking such cheap shots, why not explain why you disagree with what I wrote? The point I'm making now is that audiophiles give manufacturers a free pass by not insisting that manufacturers back up their claims with test results, that it is largely because of this that there is need for Amir to do most of what he does, and that when Amir does what he does, very many people pile on him in looking for ways to criticize his tests when what they really should be doing is insisting that the manufacturers back up their claims with published test results. If you do not agree with this, please explain why you do not agree with it. In any case, please cease and desist from making jackhole comments such as the one you have made here and now, where you criticize me for "long posts with philosophical content". I think that between this post of yours and the others you have directed at me in the past, you have given me more than ample reason to say are an extremely rude person, with no tact or manners. I wouldn't say something like this except in extreme cases where it is well deserved, and this is an extreme case where it is well deserved. You are routinely rude and dismissive of other people, telling other people that they are "ignorant" and so one, and the minute that anyone is even a little bit critical of something you've written, you take great insult. I recall that I looked back through some of posts a few months ago and one thing that was conspicuous, forming an obvious pattern, is that even in a given thread you would argue with one person by taking a particular position, then less than a day later you would take the exact opposite position in order to argue with someone else. I've observed you do this on multiple occasions. Have a nice day.

Instead of such a childish, public showing of discontent, press the ignore button if he bothers you so much. It's not difficult. Then we won't have to endure another 358 word post on why you are butthurt next time.

@pma doesn't tolerate fools. He can be brash, but his content is truly excellent. He is one of the reasons I and others stay here.

 

[Bob from Florida]

That would be a truly bad design, and I'd rather replace such a bad amp by a better one than "fix" it with such an expensive power line filter.

I hope you understand that I was theorizing a possible candidate for the line filter to become useful. Not promoting anything at all.

 

[MC_RME]

I hope @amirm is not too much pissed off by all the discussion here. IMHO it has and still brings up a lot of interesting points and insight. Personally I am torn between getting similar stuff that cleans my AC and knowing it won't make a difference at all. I have experimented with several kinds of filters, both for DC and AC, and so far have not found any situation where an existing problem was solved by them. Nor did the measurements at the DAC output ever show any change. It is more the typical 'peace of mind' that makes me look at these AC filters.

That said I also know that filters not only block one way. If your device connected to such a filter relies on a super-low impedance source to have all the 'shit' that it generates short-circuited at its input, then a filter will prevent that to happen, and chances are that 'shit' travels a different way and suddenly does become measurable or even audible. It's worse when there is only one filter and then distributes that to several outlets, but valid even with separared filters per output. As this effect was not mentioned here (unless I overlooked it) I thought I add it.

 

[amirm]

A sensible, effective line filter, surge protector and DC blocker is common sense for an expensive audio system.

Disagree. They make you feel good but do nothing. I have plenty of these things but don't use any of them. The only power product I use is a UPS for my rack and home infrastructure until the generator kicks in.

 

[Yevhen]

In practice, you always dampen the resonances. Sometimes the inductor ESR is used.

I see no resistors on PCB next to the filter. Just C-L-C filter for each channel.

Also, if you check the datasheets for the commonly used filters like https://asset.conrad.com/media10/ad...-250-vac-3-a-l-x-b-48-mm-x-225-mm-1-stuks.pdf or https://www.emea.lambda.tdk.com/uk/KB/B84771-Full-Datasheet.pdf you'll see 100k to 1000k resistor, which is not able to dump the resonances properly. It is also shown in the measurements from datasheet.

 

[pma]

Disagree. They make you feel good but do nothing.

DC blocker is good to get rid of annoying acoustical hum from large toroidal transformers in case that your mains power net has DC (usually like 1V or so). The DC component varies during the day depending on instant power consumption conditions on the power net.

 

[solderdude]

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.

ALL audio equipment use rectifiers which only do 'something' near the voltage peaks (positive and negative) and conduct only shortly.
In that small time period the voltage across reservoir caps need to be replenished by these currents.
This means there has to be a wide enough bandwidth (well over 50/60 Hz) to be able to handle the fast current transient.
Below current in the secondary windings (and thus similar shaped but different in value on the primary side) of a transformer.
The upper trace shows the voltage (ripple) across the reservoir caps being replenished and then discharged by the load (the absolute voltage levels are given next to the trace.
Lower trace shows the current in the transformer. Note that the circuit used draws a constant 3.3A (82W) yet the current from the transformer is 24A (peak) because of the short time it has to replenish. The current (24A peak) in this case is limited by the transformer itself.

Below the spectrum of these currents:

What we see here is that up to around 500Hz we don't want any filtering otherwise the current can't rise fast enough.
Filtering of mains below. The big inductors do this, the small common mode ones are starting to do something well above the audible range.

500Hz... check. Optimal filtering for audio equipment. So well done Puritan Audio. This is far lower than most other conditioners.

DC-blocking ... check.. can be handy when certain equipment in the room sometimes hums audible and at other times doesn't.
Nice to have in there all the time.

Followed by individual common mode chokes followed by glass fuses per output. Switching noises coming from connected gear (usually high in frequency) thus are filtered twice between each connected device.
However, only for HF noise.

Audio frequency common mode ground loop currents are not attenuated so when there are several devices with different ground loop currents and SE (RCA) connectors are used these pesky common mode currents that induce small voltages across the shielding of RCA cables thus still can induce small voltages.
Depending on the shield resistance (cable quality) and ground loop currents these can reach audible levels.

So no... it won't fix all gremlins.

The problems some folks talk about and blame on not so clean mains that become audible to them means they are either using audio equipment with poor PSRR (sensitive to power supply fluctuations) or have ground loop issues they are not aware off.
Those pesky 'spikes' they hear caused by heavy machinery are most often common mode. If they were differential mode (between L and N) then many devices that have MOV's or other peak voltage limiters may have equipment fail pretty often.

To those people I would advise to check each device separately, look for gear that is connected using balanced cables or use a regenerative UPS.

When troubled with ticks and noises... first get the equipment and interlinks sorted out (not easy to do) and only then start with power supply conditioning.
One can only gain so much starting out with mains filtering.

See mains filtering as a last resort instead of a starting point for improving a flawed audio system.

The claims of 'always improving' audio quality is what Amir measures here. In that sense it is limited in what is measured.
One as to realize that there is not one single EMC test that can show what happens in anyone's home and audio system where the usual values that are used for testing are exceeded. This likely is the case here and there.

So folks... stop bitching about how Amir does not show your or anyone else's particular conditions.
Just realize what is measured here is the (always false) claim conditioners ensure 'audible improvements' and NOT immunity to common mode crap which differs in EACH condition in each home.

 

[amirm]

DC blocker is good to get rid of annoying acoustical hum from large toroidal transformers in case that your mains power net has DC (usually like 1V or so). The DC component varies during the day depending on instant power consumption conditions on the power net.

I know what it is for. Last time I had transformer buzz it was on a gulf cart charger. Tightened the screws on it and it improved it some. Once in a blue moon I run into audio gear with such him but it is extremely rare. Such is also the case with experiences people share. To stick one in every system makes no sense to me. As I keep saying, wait until you have a problem before you apply remedy. You don't take an aspirin without a headache to avoid one. Same here.

 

[Mart68]

Especially when the aspirin costs fifteen hundred dollar.

 

[Yevhen]

I know what it is for. Last time I had transformer buzz it was on a gulf cart charger. Tightened the screws on it and it improved it some. Once in a blue moon I run into audio gear with such him but it is extremely rare. Such is also the case with experiences people share. To stick one in every system makes no sense to me. As I keep saying, wait until you have a problem before you apply remedy. You don't take an aspirin without a headache to avoid one. Same here.

I also see that Purifi doesn't recommend using IEC AC connectors with integrated EMI filters in their 2x400W amp demo. Ghent Audio used these filters in the first version of their DIY set, but later, after the review, replaced with a simple connector. So these filters are just useless or even harmful for SMPS. I hope those "audiophile" toroidal transformers become obsolete and I will be able to buy good class AB amp with SMPS.

 

[solderdude]

I also see that Purifi doesn't recommend using IEC AC connectors with integrated EMI filters in their 2x400W amp demo

The proper filtering is already done in their power supply. Indeed, adding extra filtering may make things worse in certain specific situations. In other situations it may not have adverse effects. The biggest issue is not knowing what gear reacts well and which one doesn't and what possible consequences can be in specific cases.

 

[KSTR]

It is almost that. The differential probe I use has 50 ohm input impedance (output of it is high impedance so don't get confused by that). The input to the AC filter is 40 ohm because that is the closest impedance AP has for balanced mode. For the purposes of what we are doing, this is more than good enough.

Thanks for clarification. I see you've mentioned one (or several?) differential probe several times. Care to give us exact make and model and how exactly you used it here?
I'm asking because
i) differential probes typically have attenuators which are not actually required with your test level of 8Volts or any test level the AP can produce, rather it would restrict dynamic range.
ii) differential probes very rarely have a 50Ohm differential input termination (except some exotic microwave stuff), normally this is an external add-on.
iii) it doesn't seem straightforward to use a differential (active?) probe here at all. The AP has differential inputs, a 100k input impedance and paralleled XLR/Banana inputs, and a banana to BNC adapter fitted with a 50Ohm terminator (with sufficient wattage) would be the natural choice.

Indeed you can see its correctness when I use it in situ.

Where exactly can we see this, wrt to unloaded operation, and in the light of the above questions?