|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
95.223.228.204
In Reply to: RE: RPi-4B and Power ?? posted by tvr2500m on January 13, 2021 at 11:42:54
Any of the advise you've been given so far I consider highly questionable.
If you experience "hum" in the system it's usually bad grounding connection
or a broken device.
Check your cables and connectors and also introduce a proper grounding scheme,
such as a star grounding scheme. Keep your cables short.
Now there keeps the myth "SMPS is bad", "Linear is good" and "battery is better"
nonsense floating around. Well underlined by the other responses to this thread.
There are characteristics which clearly define the quality of a power supply, such as
* current capability
* ripple (periodic) and
* noise (random),
* transient response,
* impedance,
* accuracy,
* load regulation,
* temperature stability,
* and more
These characteristics will define the quality of a power supply.
You can achieve top quality results with any type of power supplies.
What matters is the implementation first of all. Usually designing
a high quality SMPS is more complex then designing a Linear supply though.
And that's the key factor. Complex = high cost and effort. That's why you'll find many low Q SMPS out there and SMPSes got a bad rep.
However. There's no reason why a well done (having a good spec according to above) SPMS with a
nice buffer on it's output shouldn't do as good or better as a high Q (according to above spec) linear supply or a high Q battery supply.
Again, what matters are the specs.
Unfortunately manufacturers do not tell you much about it.
Even if the "Noise" spec is shown, it won't tell you much.
Often you'll find noise specs shown for a 0-20kHz range.
That's simply not enough. Higher frequency noise can also very well impact attached devices.
And it's a noise spec produced in a lab environment. You wouldn't know how such a device would work in a real world setup. How much noise it generates on the mains etc in your home environment.
Manufacturers hiding these and most quality parameters IMO for a reason.
Not many people demand them! I once did demand them for the Allo devices.
The answer was like. "We provide what others are providing". All I can tell I didn't make friends by asking for it.
E.g. attaching a 5 foot cable and some el-cheapo connectors plus adapters to a PS will definitely kill every "transient response" characteristic.
That's IMO why manufacturers simply avoid showing it. iFi and Allo belong of course to the few companies who talk about the subject at all.
You might want to have look at quality Lab Power supply specs. Usually these offer more related info. I'd expect high quality (audio) power supply manufacturers to at least extend their datasheets on that lab supply level.
I also had to taken my choice, even without having plenty of specs around.
I've been running my RPis with modified iPower SMPS for years. My RPi 4 server with 2 USB SSDs attached still runs stable with one of these. This doesn't mean that a stable and rather clean $15 SMPS wouldn't be able do the same job.
Still. Nowadays I'd buy an Allo Nirvana SMPS. For that purpose. I can keep the cable short. That keeps the impedance low and betters the transient response all at a reasonable price tag.
Now the streamer.
I'd try to keep the RPi and audio interface PS separated. I'd use two Allo Nirvanas or one Allo Shanti - the Shanti only in my modded version (shorter cables, DC relays) though.
I did have iFi and other USB filters around. I still use one on my Khadas Toneboard.
On modern DACs like the D90 - according to Audio Science Review - these filters do not show any effect on their measurement equipment. This of course could differ from DAC 2 DAC.
The Allo devices also offer a nice starground point to cope with groundloop issues.
Bottom line.
Even this response can not be satisfactory. I know. It's because there's
simply not the one and only solution.
Good luck on your journey.
Enjoy.
Follow Ups:
Add power supply bandwidth. Most supplies have an upper frequency limit beyond which 'regulation' is ineffective. Typically, IC regulators are poorer than solid state high quality units.
Transient response somewhat relates to bandwidth.
https://www.electronicdesign.com/technologies/test-measurement/article/21805656/simple-transientresponse-measurement-determines-powersupply-bandwidth
I'd be happy to see any of these in a PS datasheet.
You mentioned a 5 ft. Cable Length. In your opinion, what is the longest cable you'd be comfortable with?Also would the distance between the Power Supply and the Raspberry Pi matter at all. In other words, could you place the Raspberry Pi on top of a Power Supply?
Edits: 01/14/21
Nowadays I'd buy an Allo Nirvana SMPS.
Allo disagrees with your love of switchers unless you have a restricted budget.
I'd go for a Shanti . ;)
"As explained, Shanti is not only shining because of the ultra-low differential noise and common mode noise, but because of incredible low impedance / deep reservoir of electrons at the output of the PSU."
I own and run a Shanti - I think I mentioned more than once.
Allo doesn't disagree. They "hear" a minor difference. That's how I recall it. Let's not talk under what conditions.
They want to sell the product that very well fits the picture of the crazy audio crowd out there. And that's gotta be a linear supply.
Top high end audio gear manufacturers use SMPSs since years. That linear thing is a myth.
Once more. What matters is the implementation.
If Allo would have attached the same supercaps to the Nirvana output, which wasn't done, and I guess wasn't done on purpose, we wouldn't have this discussion. That's simply been a marketing decision. They needed to stay in the iPower ballpark.
And the Shantis "incredible low impedance" - a nice try. Have you seen
any transient response measurements?
Yep. Supercaps can a have quite a low ESR. At the soldering pins. But what's left of it after a 5 foot cable and some el-cheapo connectors and adapters. I never received an answer on that question.
The first thing I've done, was replacing and shortening the Shanti cables and of course getting rid of the plugs.
Enjoy.
Top high end audio gear manufacturers use SMPSs since years
Perhaps to you. I'm always amused when folks play the role of apologist to a manufacturer's product decisions. I hear the same silly explanation for speakers having bi-wiring posts. :)
Have you seen any transient response measurements?
I couldn't care less about simplistic metrics. What I hear is far more important. As for DC cables, both of mine are one foot and use low inductance star quad construction. What I hear is better dynamics, especially at the PP end of the scale.
very important
there are many components which determine that.
Speakers having diaphragms with lower mass than the air around them is one important factor to me.
Yep. "You couldn't care less."
Sounds pretty ignorant to me.
You're fighting my arguments about lower impedance and better transient response as objective, thus measurable, quality parameters for power supplies and then you're telling us, that you actually did lower the impedance and improve the transient response in your setup by replacing the cable.
...obviously without realizing what you're doing.
Watch your attitude!
Enjoy.
to reading charts. There are few metrics with direct correlation to what we hear.
To each his own. Even grumps like you. ;)
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: