Author Topic: How important is Power Supply Design?  (Read 5257 times)

Offline steve

  • Audiologist
  • *
  • Posts: 765
    • Reference components
How important is Power Supply Design?
« on: August 30, 2018, 08:19:10 PM »
Below is a link to a white paper I initially wrote in October of 1997, and recently updated. Although basic with little math, my paper deals with multiple purposes and considerations when designing an accurate power supply.

A power supply is much more difficult to design than most think, including designers/engineers I have seen on other boards. The power supply can very easily make or break a design.

http://www.sasaudiolabs.com/theory7a.htm

cheers

steve
« Last Edit: August 30, 2018, 08:36:30 PM by steve »
Steve Sammet (retired, owner, SAS Audio Labs)
"V" Very Low Capacitance ICs, 40pf 1 meter
SAS Audio Labs Test Phono Stage
SAS Audio 11A Tube Reference Preamplifier
SAS Audio Labs 25 W Triode Reference Monoblocks
2 way test Spkrs, 28 - 20khz  -3db (28hz)

Offline P.I.

  • Industry Participant
  • *
  • Posts: 795
  • Silence is that blackness beneath the music
    • P.I. audio group, LLC
Re: How important is Power Supply Design?
« Reply #1 on: August 30, 2018, 10:39:38 PM »
Thank you for posting this Steve.

Power is the heart of any system.  Without good, clean, consistent power we have bupkis.  A brain without a heart is dead.  Muscles without a good heart are worthless.

Clean power is essential, just like clean blood is essential to body health.

This is why I do what I do.  Sending abundant, quiet power to a well designed power supply is what drives me.

Awesome  :thumb:
"A man with an experience is never at the mercy of a man with an argument." - Hilmar von Campe

Offline steve

  • Audiologist
  • *
  • Posts: 765
    • Reference components
Something you may not know about Power Supply Design.
« Reply #2 on: August 31, 2018, 07:20:30 AM »
Thank you for posting this Steve.

Power is the heart of any system.  Without good, clean, consistent power we have bupkis.  A brain without a heart is dead.  Muscles without a good heart are worthless.

Clean power is essential, just like clean blood is essential to body health.

This is why I do what I do.  Sending abundant, quiet power to a well designed power supply is what drives me.

Awesome  :thumb:

Absolutely, minimizing problems from the AC power end is absolutely essential.

What is much more difficult in a DC power supply is optimizing the DC filtering system on the other end, where the musical signal "resides".

On other forums, very very few, if any, understand this aspect of the power supply's function. It is extremely difficult, virtually impossible to correctly design a power supply that will optimize the sonic qualities of a component without understanding this. In otherwards, virtually no component on the market has a properly designed DC power supply.
This includes not only the inexpensive models, but all the higher priced ones.

cheers

steve
« Last Edit: August 31, 2018, 08:46:54 AM by steve »
Steve Sammet (retired, owner, SAS Audio Labs)
"V" Very Low Capacitance ICs, 40pf 1 meter
SAS Audio Labs Test Phono Stage
SAS Audio 11A Tube Reference Preamplifier
SAS Audio Labs 25 W Triode Reference Monoblocks
2 way test Spkrs, 28 - 20khz  -3db (28hz)

Offline Nick B

  • Global Moderator
  • *****
  • Posts: 2277
Re: How important is Power Supply Design?
« Reply #3 on: August 31, 2018, 04:34:09 PM »
Below is a link to a white paper I initially wrote in October of 1997, and recently updated. Although basic with little math, my paper deals with multiple purposes and considerations when designing an accurate power supply.

A power supply is much more difficult to design than most think, including designers/engineers I have seen on other boards. The power supply can very easily make or break a design.

http://www.sasaudiolabs.com/theory7a.htm

cheers

steve

Thanks for posting your white paper,  Steve, and in a manner thatís understandable. Iím a big believer in noise reduction and thatís why I love Daveís Uber so much. It would have been such a treat to have heard your products when they were in production.
Nick
ICEpower 1200as2 amp
Audio Hungary APR 204 preamp
Fritz Carrera 7 BE speakers
Border Patrol SEi dac
Auralic Aries Mini & Mojo Audio lps
Roon, Tidal, Qobuz
PI Audio UberBUSS

Offline Folsom

  • Obsessively Audiophilic
  • ****
  • Posts: 574
Re: How important is Power Supply Design?
« Reply #4 on: September 01, 2018, 02:04:25 AM »
I think this shows a clear problem with a fundamental understanding of what switches (transistors/tubes) do. Geddes tried to go down this road, too, and it just isn't so.

Offline steve

  • Audiologist
  • *
  • Posts: 765
    • Reference components
Re: How important is Power Supply Design?
« Reply #5 on: September 01, 2018, 08:20:34 PM »
I think this shows a clear problem with a fundamental understanding of what switches (transistors/tubes) do. Geddes tried to go down this road, too, and it just isn't so.

What does your comment have to do with my white paper concerning better understanding of power supplies?

What is the clear problem with understanding of switches, transistors/tubes?
What do switches have to do with understanding a power supply/my white paper? 
What road did Geddes try to go down, as he is deep into speakers, not power supplies?
What "just isn't so"?
Please elaborate in depth on the above questions.

cheers

steve
« Last Edit: September 01, 2018, 10:29:45 PM by steve »
Steve Sammet (retired, owner, SAS Audio Labs)
"V" Very Low Capacitance ICs, 40pf 1 meter
SAS Audio Labs Test Phono Stage
SAS Audio 11A Tube Reference Preamplifier
SAS Audio Labs 25 W Triode Reference Monoblocks
2 way test Spkrs, 28 - 20khz  -3db (28hz)

Offline Folsom

  • Obsessively Audiophilic
  • ****
  • Posts: 574
Re: How important is Power Supply Design?
« Reply #6 on: September 02, 2018, 09:07:15 AM »
Signal frequency to PSU frequency impedance. It shouldn't be hard to find it in a DIYaudio search.

Offline steve

  • Audiologist
  • *
  • Posts: 765
    • Reference components
Re: How important is Power Supply Design?
« Reply #7 on: September 02, 2018, 04:52:19 PM »
Signal frequency to PSU frequency impedance. It shouldn't be hard to find it in a DIYaudio search.

That is only a small part of what I covered in my white paper. So where is your detailed explanation, reference material as I requested? By the way, how do you know the diyers/"experts" are accurate if you only partially understand the subject? (If you had understood the subject, you would have replied to all parts of my white paper.)

To give the readers a sense of the differences between our positions, we are not splitting hairs, or nit picking small differences.

My scientific white paper is based upon knowledge from the first few weeks of the first semester of electronics engineering. We are not discussing fourth year, nor third year, second year, nor end of first year engineering, but the first few weeks of the first semester in electronics engineering.

A couple of years ago, I encountered one with a masters in electrical engineering from MIT who designed speaker crossovers, yet who did not understand how an inductor/choke was non-linear. (The classic definition is amplitude linear. He simply learned a book definition.) Again, this is first few weeks in the first semester stuff.
Since you are dicing me, Folsom, please explain how an inductor/choke is non-linear.

Another with a masters in electrical engineering from Carnigie-Mellon university once attacked Blair Lamphear (from Niteshade Audio) by claiming that changing the speaker impedance did not alter the damping factor. Of course the attacker was wrong and altered his post after we nailed him.

Once again, I would appreciate your detailed response and explanation to these questions I posed to you in my last post.

What do switches/tubes, transistors have to do with understanding how a power supply affects analog audio and sonics?
How did Earl Geddes go down this road (as he is deep into speakers, not power supplies) and thus failed, connected to my white paper?

cheers

steve
« Last Edit: September 02, 2018, 08:50:03 PM by steve »
Steve Sammet (retired, owner, SAS Audio Labs)
"V" Very Low Capacitance ICs, 40pf 1 meter
SAS Audio Labs Test Phono Stage
SAS Audio 11A Tube Reference Preamplifier
SAS Audio Labs 25 W Triode Reference Monoblocks
2 way test Spkrs, 28 - 20khz  -3db (28hz)

Offline Folsom

  • Obsessively Audiophilic
  • ****
  • Posts: 574
Re: How important is Power Supply Design?
« Reply #8 on: September 03, 2018, 10:14:05 AM »
Signal has no dependency on power supply impedance for the given signal frequency. PSU impedance is only about attenuating noise that isn't signal. This has been beat to death a great many people on DIYa, and especially by the leaders on the forum (who know more than either of us).

I may have been wrong about an easy search, Geddes may have requested to have it all deleted since it was not true. He had graphs and everything.


Offline steve

  • Audiologist
  • *
  • Posts: 765
    • Reference components
Re: How important is Power Supply Design?
« Reply #9 on: September 03, 2018, 12:11:44 PM »
Signal has no dependency on power supply impedance for the given signal frequency. PSU impedance is only about attenuating noise that isn't signal. This has been beat to death a great many people on DIYa, and especially by the leaders on the forum (who know more than either of us).

I may have been wrong about an easy search, Geddes may have requested to have it all deleted since it was not true. He had graphs and everything.

Your reply shows your ignorance of basic electronics. So how could you possibly know the diyers fully understand the subject when you do not understand the subject? I see once again you could not provide any science from the diyers to support your, nor your diy associate/friends position. We are still waiting for your/diyers scientific information.

Not surprising since 1st semester basic electronics engineering teaches the opposite of your/diy position as my white paper clearly explains. Taking a semester of electronics engineering would help your diyer friends.

Claiming the leaders of the diy forum know more than both of us is obviously trolling; as you/diyers cannot even explain the basic science.

We do not appreciate your attempts to mislead the public. We are still waiting for your scientific explanation that you claim to have, folsom.

cheers

steve
« Last Edit: September 03, 2018, 12:54:52 PM by steve »
Steve Sammet (retired, owner, SAS Audio Labs)
"V" Very Low Capacitance ICs, 40pf 1 meter
SAS Audio Labs Test Phono Stage
SAS Audio 11A Tube Reference Preamplifier
SAS Audio Labs 25 W Triode Reference Monoblocks
2 way test Spkrs, 28 - 20khz  -3db (28hz)

Offline Folsom

  • Obsessively Audiophilic
  • ****
  • Posts: 574
Re: How important is Power Supply Design?
« Reply #10 on: September 03, 2018, 01:19:14 PM »
The title of the website does not describe the members found within as a whole. I will gladly take opinions from people that have had very successful ventures in the audio world, and the people who design electronics for much more complicated things that you (or I) have. The website has contributors from everywhere. Many members don't even declare whom they work for. There are loads of ex Harmon employees, physicists, employees at loads of other commercial audio companies, etc etc etc.

This is like the preamp thread at audiocircle where you claim your buffer driving an amplifier will provide zero sonic benefits over the source poorly driving the amplifier.

Let me see if I can explain it really plainly... if the impedance of the power supply had to be equal across the audio band to produce the audio band, it would be no different than saying the power supply must provide a matching frequency to the signal that is being amplified. It should not be hard to see how that cannot be true, has never been, and never will be. You would never see vanishing distortion from a classD based amplifier that's loaded full of inductors.

Offline steve

  • Audiologist
  • *
  • Posts: 765
    • Reference components
Re: How important is Power Supply Design?
« Reply #11 on: September 03, 2018, 02:55:40 PM »
The title of the website does not describe the members found within as a whole. I will gladly take opinions from people that have had very successful ventures in the audio world, and the people who design electronics for much more complicated things that you (or I) have.
Which means nothing, as

1. a thevenin equivalent circuit, Kirchoff, Norton etc, 1st semester basic electronics has proven they do not understand basic electronics in an analog application. If they don't understand the basics, how are they going to understand and properly design the complex? We are still waiting for your links to these diy "experts".

2. Successful means what? A large company with a huge marketing strategy. Cosmetics sells more than solid designs.
I think it is interesting to note that Martin DeWulf, a criminal defense attorney and music lover, investigated and wrote an article titled "Truth be Told" which discussed how shills and certain groups work in order to mislead the public to increase their own business.

3. You accept their opinions, but Thevenin equavilent circuits, Kirchoff, Norton, etc. are not opinions, but fact.
 
4. Since you don't understand even basic electronics, how do you know if they are marketing or providing real science.
You don't. Afraid you have been suckered my friend.

Quote
The website has contributors from everywhere. Many members don't even declare whom they work for. There are loads of ex Harmon employees, physicists, employees at loads of other commercial audio companies, etc etc etc.

At least you admit there are shills, conflicts of interest, employees on the diy forums. We have found such as well on Stereophile forums, AK forums etc. One company, a room acoustics company, demands their employees post and push agendas, to sell more room acoustic treatments since their jobs depend upon sales. 

Quote
This is like the preamp thread at audiocircle where you claim your buffer driving an amplifier will provide zero sonic benefits over the source poorly driving the amplifier.

Not a coherent statement. I also never manufactured, nor possessed a buffer of any kind over the decades.
No wonder no link. That is the second time I have caught you falsifying information. 

Quote
Let me see if I can explain it really plainly... if the impedance of the power supply had to be equal across the audio band to produce the audio band, it would be no different than saying the power supply must provide a matching frequency to the signal that is being amplified. It should not be hard to see how that cannot be true, has never been, and never will be.

What kind of a weird, non coherent, non-scientific statement is that? Are you kidding? It is obvious you don't understand a Thevenin Equivalent Circuit, Kirchoff, nor Norton, etc, all basic 1st weeks of electronics engineering. So how could you possibly understand my white paper.

It can easily be proven, (via Thevenin Equivalent Circuit, Kirchoff, Norton etc) that musical signal current "flows" through the power supply "filter" capacitor of an analog stage. Your comment ignores even that simple bit of science.
I am afraid the diy leaders have played you for a suckered folsom.

In the bottom diagram, musical signal current flows through both the series 0,47uf coupling capacitor and RG to ground, and through series RL and power supply capacitor C1 to ground. In the diagram, the plate of the tube is the musical AC signal source (C of transistor). Usually, RL,C1 has many times more musical signal current than through 0,47uf coupling cap and Rg. As such, the power supply filtering system will affect the sonics.

Sorry folsom, but that is how basic electronics works. I think you are killing the reputation of the diy community by pushing their snake oil. If they don't understand the basics, how can they competently design the complex.

I would be very wary of anything the diyers preach folsom. Diy forum shills, marketers, employees are there to push parts sales, and will have you spending for the rest of your life.

cheers
steve
« Last Edit: September 03, 2018, 08:16:17 PM by steve »
Steve Sammet (retired, owner, SAS Audio Labs)
"V" Very Low Capacitance ICs, 40pf 1 meter
SAS Audio Labs Test Phono Stage
SAS Audio 11A Tube Reference Preamplifier
SAS Audio Labs 25 W Triode Reference Monoblocks
2 way test Spkrs, 28 - 20khz  -3db (28hz)

Offline P.I.

  • Industry Participant
  • *
  • Posts: 795
  • Silence is that blackness beneath the music
    • P.I. audio group, LLC
Re: How important is Power Supply Design?
« Reply #12 on: September 03, 2018, 08:35:34 PM »
I donít have a dog in this fight.  Saying that Thevevin, Norton, Kirchhoff are not to be ignored.  Having built several of Folsomís P/S boards I can attest to their excellence performance.

But!

What Steve says is true IME.

I need to go back and look at the design methodology with Folsomís amps.  They do sound great!

Might I suggest a day of introspection to assess positions?

Whatever. Makes for a great read... :shock:
"A man with an experience is never at the mercy of a man with an argument." - Hilmar von Campe

Offline steve

  • Audiologist
  • *
  • Posts: 765
    • Reference components
Re: How important is Power Supply Design?
« Reply #13 on: September 03, 2018, 09:43:17 PM »
I donít have a dog in this fight.  Saying that Thevevin, Norton, Kirchhoff are not to be ignored.  Having built several of Folsomís P/S boards I can attest to their excellence performance.

You are not part of the dog fight, P.I. and thanks for the information. So folsom does have a financial stake. Interesting, and now I understand why the attack on my white paper.

By the way, that basic white paper could be used by any college class for source information, and contains information which is found in any electronics engineering textbook.

If he is building class D amps, the design is already public domain, so just copy.

He reminds me of a local company here, who initially copied others designs, then he found Dave, who does his designing.

Quote
But!

What Steve says is true IME.

You are quite correct P.I. They are absolutely critical if one wishes an accurate analog design. However, a novice could just copy any of a "million" analog schematics. It is done all the time by diyers.

The question is how accurate are the designs in terms of sonic quality? They do vary and some are pretty good. To be truly accurate is incredibly difficult, and time consuming. And it takes a lot of expertise. To me the question is, how close is pretty good to perfect, or near perfect.

Quote
I need to go back and look at the design methodology with Folsomís amps.

I would also be interested as well.

Quote
Might I suggest a day of introspection to assess positions?

No need on my end. The white paper is simple 1st semester electronics engineering information, who principles etc are in college textbooks. Been known for a hundred years.

I deal with laws, and special proprietary setup listening tests that I developed, to actually check for component accuracy in absolute terms, not simply specs or auditions. In fact, I developed two listening testing methods, so I can compare and see if there is a discrepancy.

This situation reminds me of another individual/company out west, who used a professional reviewer to dice my preamplifier for several years on AC, even though the individual/company out sourced designing to Alan Kimmel.

The company owner could not design a lick, but Martin DeWulf, Bound for Sound, told me how this guy diced me, hoping Martin would purchase his inferior preamplifier. Martin still purchased my 11A as his reference.

Then there was A.M. online magazine (?), who when I first got started in 1997, 1st magazine, asked to review a test 10A line preamp. I sent it, they listened, and a month later wanted to purchase it for $300.00. I said I needed it for testing purposes, easy to change parts etc. I would pay return shipping and send a factory 10A for review.

I got the test 10A back before I could send shipping money, and someone had hammered the entire front, top, back of chassis, bent the rack front plate into a V, sheered volume control nut off, inside pc board demolished. When I complained, I did receive a check for $300.00. (List price for a 10A was $1299.00 at the time.)
Never did get the review.

I was then informed by one of my customers of a forum called A.A., someone was attacking my amps. When I defended the tubes I used, I received as many as 36 virus attacks via emails in one day. This went on for weeks. I soon left AA, and learned to back up my computer hard drive.

I got scammed by R**ze, a professional reviewer for P.F. Wrote his review before I even got home. Never got another phone call up from Milwaukee way after he was done with me. Hid the fact that he was a reviewer until the evening before I traveled up for a group audition. That is the 2nd reviewer.

A speaker company, out of Canada, kept seeing negative reviews from a guy (from USA) on forums, and he never  auditioned one of their speakers. Finally, he did audition the speakers, and then attempted to sell them without paying for them. Company caught him and contacted United States law inforcement, who did act. To keep out of jail, he had to delete all his negative posts. The speaker company then dropped the charges, from memory.

I am glad the viewers here got to see behind the scenes of what really happens in the audio world. Reviewers who shill and attempt to extort money/goods for reviews, attacking opponents with viruses etc. Been interesting I will say.

Now I get attacked over a basic white paper that any college class would use as source information, and contains information which is found in any electronics engineering textbook. Surprise, surprise.  :lol:

cheers

steve
« Last Edit: September 03, 2018, 10:51:07 PM by steve »
Steve Sammet (retired, owner, SAS Audio Labs)
"V" Very Low Capacitance ICs, 40pf 1 meter
SAS Audio Labs Test Phono Stage
SAS Audio 11A Tube Reference Preamplifier
SAS Audio Labs 25 W Triode Reference Monoblocks
2 way test Spkrs, 28 - 20khz  -3db (28hz)

Offline TomS

  • Certifiable
  • ***
  • Posts: 193
  • Yeah, it's just a hobby...
Re: How important is Power Supply Design?
« Reply #14 on: September 04, 2018, 06:58:39 AM »
Steve,

The white paper is good stuff. Thanks also for clear thinking and speaking the truth, it's most refreshing.

I remember those first EE courses well. Those fundamental laws remain the same and need to be thoroughly understood to do good work that passes muster. Prof's William Hayt and Gerry Neudeck or their TA's personally drilled it into me back then. The creative art comes in when one ventures off the script of application notes and textbook reference designs (other people's work), and the designer tries something original or unique. Their exams always had a few of these "ringers" on them you'd never seen before, just to make sure we used our brains to think for ourselves. Fortunately, the analysis using those laws, remained the same, and never let you down. Thinking you could somehow "trick" or conveniently ignore them never worked out well  :) Then, of course, there is the effect of real world factors - components, circuit layout, variability, etc. to deal with. As you know, that's part of the fun in it.

For the benefit of all, I hope everyone tries to keep it a respectful and informative dialog. Please carry on.
PS Audio Directstream DAC, PS Audio BHK Signature Preamp, Nagra Classic Amp, Focal Sopra No 2's, SVS SB16 Ultra