Yes, and that’s why double blind is best, as even the person making the “switch” doesn’t know which is which, so they can’t send you subtle (and maybe unintentional) signals.
1 Like
LAN connections are galvanically isolated at either end by default - every RJ45 socket in a piece of networked equipment has isolating and common mode transformers built in. You don’t need additional galvanic separation.
3 Likes
Hmmm, did you read the article linked above? Benchmark Audio’s AHB2 power amp has a switched mode power supply. It bests every LPSU on the market in scientific tests. SNR is state of the art by a large margin. You can’t achieve those results with a LPSU.
“LPSUs are clearly better than SMPSUs” is an audiophile myth and it’s been busted.
4 Likes
Thanks for that correction.
I own and use an AHB2. I’m very familiar with it and its specs. That doesn’t mean you can"t get better performing SQ out of linear supplies. Especially when replacing the myriad of garbage switching supplies out there.
Linear Power Supply makes no sense if your DAC has enough power through the signal USB cable. The same is true about the “audiophile” USB cable: if your current USB cable is good enough (thick and good screened) - you don’t need “special” USB cable for streaming.
It’s only “zeroes and ones” and nothing magic involved… But real magic is inside your DAC that should be expensive enough with good algorithms.
You made a broad statement that LPSUs are better than SMPS. I responded to that statement, because it’s not true and I cited the AHB2 as an example.
Sure, there are some cheap SMPS which are supplied with audio gear and they may benefit from replacement.
Note, I said may, because it depends on the design of the component as to whether it is sensitive to the quality of power supply.
We’re getting back into the realms of subjectivity vs objectivity again. You can measure a power supply and tell whether it’s good or bad. If it’s good, it won’t affect sound quality. If it’s bad, it might. It depends on the noise levels and their frequencies as to whether it can make an audible difference.
1 Like
I did some power supply measurements specifically on the RPi Gen 4. They are posted elsewhere on the forum. Take a look if you are interested.
Bottom line: with most endpoints and computing devices like NUC’s, your linear power supply is directly feeding an SMPS! Audiophile dammit!
Off line 120-240 Vac to 5 Vdc or 12 Vdc, etc. SMPS having two-wire mains plugs do dump line current into their grounds and outputs. This ends up on your system ground.
All you have to do to solve this particular issue is run a ground wire from your SMPS (-) output to earth ground on your mains outlet.
Done. No need to worry about ground loops if your DAC is properly designed with transformer or optical coupling.
Far be it for me to say that using a linear supply won’t make a difference in what you hear. It’s your money, your system, your unbridled listening pleasure.
It’s totally ok that power supply differences in these systems may exist solely in your auditory cortex. That does not make them any less real.
1 Like
I’m curious - why the caution when there is no safety ground to begin with on two wire mains supplies?
As-is, any fault currents will end up in your signal ground. Adding a safety ground to the output of an offline SMPS provides a fault path other than your signal ground. Hum is not an issue since the inputs of a well designed DAC are galvanically isolated.
You make a good point. All component SQ isn’t worse off because it has an SMPS. The AHB2 is a good example of that. Could Benchmark improve upon it with an LPS ? Do you know if it was a consideration for them?
I’ve found most companies who produce in mass use SMPSs. I can only guess as to why. I do know that in the numerous components where I’ve substituted an LPS the SQ has improved. It is very subjective and limited in scope but the improvements were there. The LPSs I have are not cheap and they are a bit larger.
They cite noise as the reason they stopped using linear supplies five years ago and further state that they would never have achieved 135dB SNR in the AHB2 with a linear supply.
Generally power amplifiers never achieve anywhere near as low a noise floor as this because of magnetic coupling into the signal path. Also the rectified ripple frequency is either 100 or 120Hz depending on region which as we all know is very much in the audio band. Noise at 50/60Hz and 100/120Hz is very much a thing.
With a well-designed SMPS the noise is a long way outside the audio band. SMPS are also lighter and more efficient. With a much stronger focus on environmental credentials across all industries, I think we’ll see many more components utilising SMPS.
3 Likes
Thanks everyone for the advice. I just got my NUC yesterday and got it all set up. Running a Raspberry Pi4 with no special things into my Chord Qutest as a source, and it sounds great. I bought a fancy USB cable from Curious but it didn’t work (as per my other threads) so I’m keeping it simple.
I’m eventually getting a Sugden amp (ordered with a two month wait), so that may be a bit more revealing of my sources than the Brio but for now I’m satisfied.
2 Likes
Good info. Actually “linear” is a misnomer because unless the power supply has a choke input (most solid state supplies do not use chokes) then the current through the transformer secondary, rectifiers, and into the filter capacitor is highly non-linear. The resulting noise is very hard to get rid of.
It’s only a misnomer if you link the word ‘linear’ to the current quality. I was under the impression ‘linear’ in the name refers to the lack of switching involved.
Correct Ben. The term linear wrt power supplies usually refers to the method of regulation of the DC output, linear being a supply that drops voltage across a series pass element. Whereas a switching power supply contains an electronic switch used to control the current through an inductor as a means to control output voltage.
But linear also has other meanings in electronics, like mathematically linear relationship between current and voltage.
How linear in audio came to be interpreted as ‘good’ and switching power supplies ‘bad’ is uncertain.
If something is linear, it is better, right? That appears to be the promise under which a lot of linear power supplies are marketed and sold.
Some would say that’s tendentious bumf.
Then it’s not really a misnomer, but a misunderstanding of the use of the term, and that’s what causes the camps to form. It’s like with amplifier classes. But as the song goes, ‘it’s not what you do it’s the way that you do it, and that’s what gets results.’