Linear Power Supply

[novisnick]

Aug 30, 2018 23:10:04 GMT -5 klinemj said:

Aug 30, 2018 22:31:53 GMT -5 novisnick said:

klinemj , speaking of your LH Labs Geek Pulse xinfinity, what are you using as a power supply for it? I’ve missed it or don’t recall.

I have the base on that came with it, but I am about to try another option. Going to try an ifi power linear power supply that's actually a wall wart. Comes reco'd by others who use the Geek as being as good as many high priced options.

Mark

Mark, thats a lot of vinegar to swallow IMHO. I find it impossible for such a device to supply the type of difference in your DAC that will be a noticeable improvement in sound. It may/will supply the power required to operate your DAC but for a serious improvement In audiophile  sound I don’t believe that PS will do anything for you. 

I’ve found that a superior PS used on any component in the audio signal chain makes a huge difference, like using a set of $500.00 speakers as the endpoint of $10,000.00 in gear. Your not going to hear any improvement unless you improve those speakers. A system is only as good as its weakest link.

With all due respect, once you’ve tried a serious PS you won’t cheat yourself by not hearing everything your DAC has to offer. Of course a great PS can not improve the SQ of a poor DAC. I haven’t heard your DAC preform but I’m sure it must be of very good quality.

The SQ improvement of my DAC after installing the Sbooster PS was not minimal, there was a very big difference! Believe me, I didn’t think there was any room for much improvement at all in my digital system, there was! 

With love and respect my friend.
Nick

 

[klinemj]

Aug 31, 2018 3:40:05 GMT -5 novisnick said:

Aug 30, 2018 23:10:04 GMT -5 klinemj said:

I have the base on that came with it, but I am about to try another option. Going to try an ifi power linear power supply that's actually a wall wart. Comes reco'd by others who use the Geek as being as good as many high priced options.

Mark

Mark, thats a lot of vinegar to swallow IMHO. I find it impossible for such a device to supply the type of difference in your DAC that will be a noticeable improvement in sound. It may/will supply the power required to operate your DAC but for a serious improvement In audiophile  sound I don’t believe that PS will do anything for you. 

I’ve found that a superior PS used on any component in the audio signal chain makes a huge difference, like using a set of $500.00 speakers as the endpoint of $10,000.00 in gear. Your not going to hear any improvement unless you improve those speakers. A system is only as good as its weakest link.

With all due respect, once you’ve tried a serious PS you won’t cheat yourself by not hearing everything your DAC has to offer. Of course a great PS can not improve the SQ of a poor DAC. I haven’t heard your DAC preform but I’m sure it must be of very good quality.

The SQ improvement of my DAC after installing the Sbooster PS was not minimal, there was a very big difference! Believe me, I didn’t think there was any room for much improvement at all in my digital system, there was! 

With love and respect my friend.
Nick

 

I read up on them and they claim extremely low noise and quote numbers that look very good. In fact, at 0.000001 volt it's lower than the only number I can find for the uptone audio LPS-1 (0.004 v). So, I am going to try it.

Not much to lose!

Mark

[klinemj]

Just realized I called the ifi power a linear power supply - it's not. It's a SMPS, however the noise spec holds. It uses an "Active Noise Cancellation" system. They claim 20x lower noise than other "audiophile power supplies". When looking for data on others, I can rarely find it. The Ultracap LPS-1 says "It is true that low noise is important to power supply performance, and we use the best-in-class Texas Instruments TPS7A4700 regulators (0.004mV RMS noise) for the 1-amp output of the UltraCap LPS-1"

And the Wyred 4 Sound's PS-1 specs actually show "Noise: less than 5µV (no filter). Less than 2.5µV (20Hz - 1000Hz)" for 12v/0.75 amp.

So, it's lower in noise than 2 expensive linear power supplies.

And, in Stereophile, Kalman Rubinson said “Recordings with open, ambient soundstages sounded cleaner, and both instrumental and vocal music was more distinct. I intend to keep a suite of them on hand.”

In a ComputerAudiophile test of 8 different power supplies (some SMPS and some LPS), it came in 2nd overall in lab tests for "ripple" and performed so well it surprised them. They said: "Comparison of the two loaded bands, two bands coincide highly, very stable, even the peak of the highest point is not more than 1uV, the overall average of 150nV, measured data lower than iFi official data of 1uV, really low-key marketing. It can achieve this kind of performance with maximum loading, no wonder in the listening test it was much better than the LPS, my friends and I have heard it. Now the mystery has been solved. And later looked at the iFi's official website, may be this is the ANC technology at work, seems a very good “dark-art” technology, indeed should receive some praise."

I'll let you know how it works.

Mark

[KeithL]

I absolutely agree with novisnick: a system is only as good as its weakest link.

I would, however, add the kudo that you must always consider that in the context of the actual requirements of the system.

Perhaps the best way to explain that is simply to suggest that you always take the metaphor literally.

If you have a chain, and one link is weak, then that link will be what limits the strength of the entire chain.

So, if you have a ten foot length of one-inch chain, and you insert a single 1/4 inch link, that will limit the strength of the entire chain.

HOWEVER, you must also realize that adding a few inches of two inch thick chain, at one particular spot in your one inch thick chain, is NOT going to make the entire chain any stronger either.

(And your car won't come out any cleaner, or go any faster, if you wash it with $500 a gallon "lab grade ultra pure water" instead of the stuff that comes out of the tap.)

Certain types of circuitry are very sensitive to the amount of noise that is present on the power they run on; while, for others, it doesn't matter at all.

This is especially true for DACs.

DACs run on several power supply and reference voltages.

The accuracy and amount of noise present on some of the power supplies is critical, while others barely matter at all.

The tricky part is that the details depend very heavily on the particular design.

You CANNOT reasonably make general statements about this unless you know a lot about the details of the particular design involved.

In other words, if you want to make generalizations, then everyone is going to be both right and wrong.

On SOME DACs, the quality of the power supply is critical, and adding a more precise or quieter power supply may make a huge difference in performance.

Yet, on others, it may make no difference at all.

(In a well-designed DAC, the critical reference voltages are regulated and filtered separately, specifically so their quality is not dependent on the quality of the overall power supply.)

And, to make matters much more complicated, there is no single measure of "quality"; there are many parameters that define the performance of a power supply.

The reality is that your favorite company may have chosen their power supply based on how much they want to spend - and substituting a better one may improve performance significantly.

Or they may have carefully determined exactly the sort of performance they need in their power supply to get the best possible performance from the rest of the unit - and included that one already.

And, yes, sometimes there is a sort of "grey area" - where a huge increase in cost may produce a slight improvement in performance - and whether that's worthwhile comes down to being a value judgment.

Of course, sometimes, substituting a massively overdesigned and expensive power supply, that would work great in a different product, may actually make the performance of your unit worse instead of better.

Unfortunately, depending on the quality of the engineering involved, it can be difficult to figure out which situation you have at a given moment.

Audiophiles have a tendency to ASSUME that every product is cost-limited; and that they'll be able to get a significant improvement simply by upgrading their favorite component or components.

In reality, in order for this to work, you must first discover a weakest link, and then discover how to improve it in a meaningful way...

In a well-designed product, the links are all equal and adequate; and, if so, strengthening one of them alone is unlikely to get you much improvement.

In short, the specific improvements in performance you can hope to achieve with "design improvements" must be considered - well - specifically... in terms of the overall design.

If you don't believe this, try towing your favorite trailer with a few feet of anchor chain from an aircraft carrier, and see how far you get .

In the case of power supplies, you simply may or may not be able to improve the overall performance; it all depends on whether the power supply was the limiting factor to begin with.

(And, yes, there is a huge range of performance on both linear and SMPS; so it is foolish to assume that either is automatically better.)

If you really want to learn about this sort of thing, a good place to start is to download some of the Application Notes from the various DAC chip manufacturers.

They will usually tell you exactly which voltages are critical, and which ones don't matter much at all, and frequently offer suggestions about how to make sure their chips perform at their best.

And, if the guys who spent millions developing that chip have suggestions about the best way to use it, there's a very good chance they know what they're talking about.

(Some folks refer to this as "cookbook engineering", and talk about it in disparaging terms; the reality is that the chef who wrote the cook book probably really is a better cook than you, and taking his advice probably is a good idea.)

Aug 31, 2018 3:40:05 GMT -5 novisnick said:

Aug 30, 2018 23:10:04 GMT -5 klinemj said:

I have the base on that came with it, but I am about to try another option. Going to try an ifi power linear power supply that's actually a wall wart. Comes reco'd by others who use the Geek as being as good as many high priced options.

Mark

Mark, thats a lot of vinegar to swallow IMHO. I find it impossible for such a device to supply the type of difference in your DAC that will be a noticeable improvement in sound. It may/will supply the power required to operate your DAC but for a serious improvement In audiophile  sound I don’t believe that PS will do anything for you. 

I’ve found that a superior PS used on any component in the audio signal chain makes a huge difference, like using a set of $500.00 speakers as the endpoint of $10,000.00 in gear. Your not going to hear any improvement unless you improve those speakers. A system is only as good as its weakest link.

With all due respect, once you’ve tried a serious PS you won’t cheat yourself by not hearing everything your DAC has to offer. Of course a great PS can not improve the SQ of a poor DAC. I haven’t heard your DAC preform but I’m sure it must be of very good quality.

The SQ improvement of my DAC after installing the Sbooster PS was not minimal, there was a very big difference! Believe me, I didn’t think there was any room for much improvement at all in my digital system, there was! 

With love and respect my friend.
Nick

 

[klinemj]

KeithL

Do you have a good reference source for noise testing on various power supplies? I can't find much out there that's published, other than what I noted about the ifi power supply, the Uptone LPS-1, and the Wyred 4 Sound PS-1. All others I looked at "say" they are low noise but show zero data/zero specs. Therefore, it's hard to see which one might be a "stronger" link than what I am pretty sure is already a weak link.

Based on that, I'm willing to spend $49 to have 1 delivered to try...it may or may not make a difference, but I can return it is it does not.

Mark

[novisnick]

Aug 31, 2018 9:56:53 GMT -5 klinemj said:

KeithL

Do you have a good reference source for noise testing on various power supplies? I can't find much out there that's published, other than what I noted about the ifi power supply, the Uptone LPS-1, and the Wyred 4 Sound PS-1. All others I looked at "say" they are low noise but show zero data/zero specs. Therefore, it's hard to see which one might be a "stronger" link than what I am pretty sure is already a weak link.

Based on that, I'm willing to spend $49 to have 1 delivered to try...it may or may not make a difference, but I can return it is it does not.

Mark

Sometimes the specs just Don’t say it all. The “Power” supply and what it gives you can be heard the moment you power up. With a serious DAC your jaw will hit the floor. The Sbooster is such a PS. It is returnable but that thought NEVER crossed my mind. Sometimes you Do Get what you pay for.
My purse isn’t bottomless and I don’t care for waisting money as its hard enough to come by.

In the words of Forest Gump, “and thats all I’ve got to say about that”.

[KeithL]

Interesting.... my conclusion would be the exact opposite.

"With a good DAC the manufacturer will have already included whatever power supply you need to get the best performance".

So, if a different power supply improves it, then I would be wondering why the original one they provided was inadequate.

Aug 31, 2018 10:08:55 GMT -5 novisnick said:

Aug 31, 2018 9:56:53 GMT -5 klinemj said:

KeithL

Do you have a good reference source for noise testing on various power supplies? I can't find much out there that's published, other than what I noted about the ifi power supply, the Uptone LPS-1, and the Wyred 4 Sound PS-1. All others I looked at "say" they are low noise but show zero data/zero specs. Therefore, it's hard to see which one might be a "stronger" link than what I am pretty sure is already a weak link.

Based on that, I'm willing to spend $49 to have 1 delivered to try...it may or may not make a difference, but I can return it is it does not.

Mark

Sometimes the specs just Don’t say it all. The “Power” supply and what it gives you can be heard the moment you power up. With a serious DAC your jaw will hit the floor. The Sbooster is such a PS. It is returnable but that thought NEVER crossed my mind. Sometimes you Do Get what you pay for.
My purse isn’t bottomless and I don’t care for waisting money as its hard enough to come by.

In the words of Forest Gump, “and thats all I’ve got to say about that”.

[novisnick]

Aug 31, 2018 10:14:53 GMT -5 KeithL said:

Interesting.... my conclusion would be the exact opposite.

"With a good DAC the manufacturer will have already included whatever power supply you need to get the best performance".

So, if a different power supply improves it, then I would be wondering why the original one they provided was inadequate.

Aug 31, 2018 10:08:55 GMT -5 novisnick said:

Sometimes the specs just Don’t say it all. The “Power” supply and what it gives you can be heard the moment you power up. With a serious DAC your jaw will hit the floor. The Sbooster is such a PS. It is returnable but that thought NEVER crossed my mind. Sometimes you Do Get what you pay for.
My purse isn’t bottomless and I don’t care for waisting money as its hard enough to come by.

In the words of Forest Gump, “and thats all I’ve got to say about that”.

Because of price point. Why doesn’t Emotiva put the best of everything in the world and sell it, because Emotiva has to hit a price point obviously, just like every other manufacturer out there. Except for the price no object guys.

With all due respect, 

[KeithL]

The short answer is no.

There are many different types of noise; different circuits are more or less sensitive to different ones; and each is tested differently.

Therefore, before you can test for how well a certain power supply will perform in a certain application, you have to know the requirements of the application itself.

Both the requirements, and what's available, have also both evolved over time.

For example, in the old days, phono preamps tended to be very sensitive to hum and other audio frequency noise.

And, in the old days, most power supplies were linear, so the noise you would find would be 60 Hz hum - which is quite audible.

Therefore, in those days, you would want to design your power supply to produce very low levels of hum, and you would want to measure that.

Then, more recently, many switch mode power supplies operated at audible frequencies, so that was a concern.

(Modern op amp circuits are quite immune to power supply noise; but noise can still leak into the circuitry in other ways.)

This is why so many phono preamps, especially older designs, still use a well filtered linear power supply.

Nowadays, most modern switch mode power supplies operate at ultrasonic frequencies (so, if a tiny bit of noise leaks in, it will be inaudible).

Therefore, by all the measurements, modern SMPS often perform far better than linear power supplies.

Our XPS-1 separate phono preamp obtains it's power from a little wall wart - which contains a switch mode power supply.

Being a switch mode power supply, it produces virtually no hum at all.

And, because it operates at ultrasonic frequencies, any tiny amount of noise it might generate is inaudible anyway.

However, interestingly, I still hear from people wondering whether using "a high quality linear power supply" to power the XPS-1 would improve it's performance.

The funny part is that the SMPS wall wart is simply providing raw power to power the internal power supplies inside the XPS-1 - which have their own filtering.

And, what most people don't realize, is that the internal power supplies in the XPS-1 also also switch mode power supplies.

And, since the internal power supplies perform very well on their own, the quality of the external supply doesn't matter all that much at all.

(That's not to say that a badly designed, or improperly grounded, external supply couldn't manage to introduce noise; but just that "improving" it beyond what's really needed isn't going to improve the overall performance.)

Therefore, adding a superb quality well filtered super expensive linear power supply to the XPS-1 is unlikely to affect its performance at all; it won't hurt; but it won't make a positive difference either.)

Whether it's likely to improve the performance of a DAC will depend on the design of the DAC itself... and on it's current power supply... and on how well the current power supply fulfills that DAC's particular requirements.

I would expect it to make a huge difference on some - and no difference at all on others.

Of course, we're assuming that a particular "upgrade power supply" actually improves the particular parameters that are likely to improve that particular DAC.
This is why I would always warn against assuming that an upgrade that improves one particular DAC will do likewise for a different one.

I should also point out that it isn't especially expensive to design a high quality power supply; and many circuits really will perform perfectly with a simple $5 power supply.

(As the engineering joke goes: "Not only isn't it rocket science; but rocket science isn't all that hard these days anyway."

Finally, switch mode power supplies are relatively new technology.

As with all new technology, some of the early designs didn't perform very well.

And, beyond that, many engineers simply aren't familiar with how to make good designs using the new technology (the "ground rules" and "tricky bits" are quiet different).

Someone who went to engineering school twenty years ago is more likely to be familiar with how to design a decent linear power supply than with how to design a good SMPS.

(Therefore, given his or her choice, that's the option they're going to choose... and, quite possibly, the one that they'll be able to achieve good results with.)

Aug 31, 2018 9:56:53 GMT -5 klinemj said:

KeithL

Do you have a good reference source for noise testing on various power supplies? I can't find much out there that's published, other than what I noted about the ifi power supply, the Uptone LPS-1, and the Wyred 4 Sound PS-1. All others I looked at "say" they are low noise but show zero data/zero specs. Therefore, it's hard to see which one might be a "stronger" link than what I am pretty sure is already a weak link.

Based on that, I'm willing to spend $49 to have 1 delivered to try...it may or may not make a difference, but I can return it is it does not.

Mark

[novisnick]

And , not being able to keep my mouth shut please let me add one more thing.
Many big dollar pieces of gear prefer to have the power supply separate from the device. This is why some produce a separate PS and others give you the option of adding on a better PS by including a port for it alongwith the specs needed to match the gears minimum requirements.

Logic, at least to me would want to put The noisy PS awake from my signal carrying my music, IE minimal proximity.
I may have slept at a Holiday Inn Express

[KeithL]

Personally, I find it naive to assume that you can ALWAYS purchase better performance by spending more money.

Is there a "price no object" power supply that will do a better job of powering an AD1955 DAC chip than Texas Instrument's latest $5 chip regulator?

Perhaps... but I would not bet on it.

The power supply you need for a DAC has some more or less specific requirements.

This can make it somewhat tricky to design, and leaves the door open for both design errors and design improvements.

(So even some really expensive ones may still be badly designed.... but the problem is not the cost limitation.)

I'll take our DC-1 DAC as an example.

If you were to ask me to redesign it, but limit the cost of the power supply to $5, there would be some serious compromises.

However, if you were to offer the opportunity to design a $5000 power supply for it, I very much doubt you could improve the measured performance or sound quality by doing so.

(You could design a power supply that measured better by itself, and certainly one that cost more, and might last for a million years, but it would NOT improve the overall performance of the DAC.)

I would assert that the "design ceiling" for a typical DAC power supply is a few hundred dollars at most.... and you simply cannot get a MEANINGFUL improvement by spending ANY amount above that.

(And, yes, if I spend $1000 for a DAC, and the designer was competent, I expect them to have already included a power supply that good.)

Likewise, if you were a millionaire, and wanted to build a "cost no object" home, you could have your house built with pure 24k gold roofing nails.

And you might even convince someone that gold was best.

However, the reality is that plain old steel roofing nails will work significantly BETTER than solid gold ones.

(The ideal choice would probably be stainless steel or titanium... although plain old galvanized steel has a longer track record of successful use.)

Aug 31, 2018 10:50:00 GMT -5 novisnick said:

Aug 31, 2018 10:14:53 GMT -5 KeithL said:

Interesting.... my conclusion would be the exact opposite.

"With a good DAC the manufacturer will have already included whatever power supply you need to get the best performance".

So, if a different power supply improves it, then I would be wondering why the original one they provided was inadequate.

Because of price point. Why doesn’t Emotiva put the best of everything in the world and sell it, because Emotiva has to hit a price point obviously, just like every other manufacturer out there. Except for the price no object guys.

With all due respect, I find your question rather naive @keithl

[KeithL]

There are two sides to that logic.... one quite positive and one quite NEGATIVE (ignoring the added cost).

By isolating the power supply, both by distance, and by using a separate box, you separate the AC line itself, and the power supply components, from the audio components.

This helps minimize noise leakage from the AC line, and from the power supply circuitry, into the audio components.

HOWEVER, especially with modern circuitry, there is also compelling reason to minimize wire lengths.

By moving the power supply two feet from the audio circuitry you have ADDED two feet of wire between them.

If you read up on modern circuit design, you'll notice that, in many situations, you are supposed to MINIMIZE the distances between components.

If you're designing a DAC circuit board, an extra few MILLIMETERS of circuit board between the clock chip and the DAC chip may double the jitter.

Likewise, those decoupling capacitors that are there to reduce interference from high frequency noise must be located within one or two millimeters of the chip they work with.

That extra two feet of wire you've added has increased the impedance and inductance of your power supply connection significantly.

It may also act as an antenna, and actually pick up MORE noise (you're trading noise from being near the line cord to noise from your neighbor's cell phone).

This is comically obvious with CD players.

Assuming you're designing JUST a CD player, with analog outputs, and a built-in DAC....... 

Best practice circuit design demands that the DAC circuitry be as close as possible to the digital output of the transport.

The best, cleanest, lowest jitter form of signal transport possible is a few millimeters of circuit trace.

ANYTHING ELSE is adding complexity and compromising performance.

The two foot cable, with the fancy high-speed buffers, and the high end shielding is NOT better.

If you're really lucky it isn't too much worse.

(And, if you're even luckier, the performance you've lost doing that makes up for the noise improvement from using separate boxes.)

And, yes, some simply offer those options for the same reason that your favorite car company offers the fancy trim options.

(Hint, it's NOT because it makes the car go faster; it's simply because some people like options are are willing to pay for them.)

I could even be very sinister and wonder whether the company who offers that expensive "power supply upgrade"....DELIBERATELY included a poor quality power supply with the unit to FORCE you to pay for the "upgrade".

(You will find many instances where features that cost only a few cents to include, like an extra input connector, are omitted from "bottom of the line" equipment - simply to get you to buy the next one up.)

Aug 31, 2018 11:23:23 GMT -5 novisnick said:

And , not being able to keep my mouth shut please let me add one more thing.
Many big dollar pieces of gear prefer to have the power supply separate from the device. This is why some produce a separate PS and others give you the option of adding on a better PS by including a port for it alongwith the specs needed to match the gears minimum requirements.

Logic, at least to me would want to put The noisy PS awake from my signal carrying my music, IE minimal proximity.
I may have slept at a Holiday Inn Express

[novisnick]

Aug 31, 2018 11:51:48 GMT -5 KeithL said:

Personally, I find it naive to assume that you can ALWAYS purchase better performance by spending more money.

Is there a "price no object" power supply that will do a better job of powering an AD1955 DAC chip than Texas Instrument's latest $5 chip regulator?

Perhaps... but I would not bet on it.

The power supply you need for a DAC has some more or less specific requirements.

This can make it somewhat tricky to design, and leaves the door open for both design errors and design improvements.

(So even some really expensive ones may still be badly designed.... but the problem is not the cost limitation.)

I'll take our DC-1 DAC as an example.

If you were to ask me to redesign it, but limit the cost of the power supply to $5, there would be some serious compromises.

However, if you were to offer the opportunity to design a $5000 power supply for it, I very much doubt you could improve the measured performance or sound quality by doing so.

(You could design a power supply that measured better by itself, and certainly one that cost more, and might last for a million years, but it would NOT improve the overall performance of the DAC.)

I would assert that the "design ceiling" for a typical DAC power supply is a few hundred dollars at most.... and you simply cannot get a MEANINGFUL improvement by spending ANY amount above that.

(And, yes, if I spend $1000 for a DAC, and the designer was competent, I expect them to have already included a power supply that good.)

Likewise, if you were a millionaire, and wanted to build a "cost no object" home, you could have your house built with pure 24k gold roofing nails.

And you might even convince someone that gold was best.

However, the reality is that plain old steel roofing nails will work significantly BETTER than solid gold ones.

(The ideal choice would probably be stainless steel or titanium... although plain old galvanized steel has a longer track record of successful use.)

Aug 31, 2018 10:50:00 GMT -5 novisnick said:

Because of price point. Why doesn’t Emotiva put the best of everything in the world and sell it, because Emotiva has to hit a price point obviously, just like every other manufacturer out there. Except for the price no object guys.

With all due respect, I find your question rather naive @keithl

Absolutely Keith! And for the record, I’ve already omitted the uncalled for naivity remark. I was out of line. Now I think we’re swinging in extreme directions, gold nails and all. Yes, I was discussing adding a $300. PS to a $30. DAC. Come on my friend lets be a little more reasonable. 
There are more then a couple of manufacturers that include an External PS port for just such a reason. They have included an adequate PS but also thought that their device would benefit from a better/different PS. 

Question, is it beneficial, in some/many cases to have the PS away from the DAC and its signal carrying our music? I would think so but I Am NOT an engineer.
cross post error, asking while your answering, 😋 Oops. 

[DYohn]

Here's my take. Any reasonable manufacturer will always supply/create/build-in whatever power supply is best suited to achieve the intended performance specs of the gear. Indeed the specs they list, the specs they guarantee, the specs they use to sell products will always be met or exceeded by whatever is in the box they ship.

Are there instances where "better" performance can be attained using "better than spec" power supplies? Sure. Is this always the case? No. Do some manufacturers offer various P/S options? Yes they do. Why do they do it? Because it might make a difference to the user, although it will not affect performance to the listed specs.

So in my humble opinion, if you like to tweak and experiment and seek that sometimes mythical extra 2% then by all means do whatever you like and can afford to try and squeeze it out. But you can also rest assured that no legitimate manufacturer will sell you crap that they know is crap. The last thing they want to do is harm their own reputation.

Oh and as to the question is a separate P/S better, the absolute answer is maybe.

[KeithL]

I agree absolutely.

At the lower end of the price spectrum the cost of the DAC circuitry and the power supply are probably quite similar.

And, if you're trying to minimize cost, then it makes sense to offer a minimal power supply with the unit - and then offer a better one as an upgrade.

I've also seen DACs which can be operated on USB power, or with an optional AC power supply, where the AC power option was a major improvement.

(Using USB power is the "no cost option", but there are ground noise issues, as well as limitations on the overall power available.)

However, in other cases, because of the design of the product itself, it is simply already functioning "optimally".....

And no amount of external upgrades are going to make a significant difference.

Unfortunately, there are also many companies who seem to offer "premium upgrades" simply because there is a market for them.

And there are a lot of audiophiles who are a bit too eager to assume that every low cost product can become a performance champion - if only you can find the right upgrade.

But, yes, when it comes to more expensive products, I do expect the original designer and manufacturer to have used all the right parts to begin with.

And, in engineering terms, a power supply really is relatively simple....

Therefore, for a device like a DAC, which doesn't use that much power, you really should be able to make a near-perfect power supply for a few hundred dollars.

(There really isn't much you can do beyond that, at any price, that is going to make it perform usefully better.)

As I mentioned in another post there are some benefits of moving the power supply into a separate cabinet.

However, there are also drawbacks, and the benefits have been dwindling with modern designs.

The main difference these days is more often one of design convenience.

You may be able to design a perfect internal power supply - but it's easier, and the design requirements are less picky, if you separate it.

There are also other details which aren't at all obvious.

For example, every unit that is powered from the AC line must pass a variety of safety requirements and certifications.

Therefore, when you design a new AC-powered device, you must send it off, pay for the tests, and wait for it to pass.

And, if you update it next year, you'll have to repeat the entire process.

BUT, if you use a power brick, the device itself is no longer AC powered...

All that needs to be certified for AC operation is the power supply.

You can purchase one that's already been certified from your favorite vendor.

Or you can design your own - and go through the certification process.

But, if that power supply works with multiple models, you only have to get it certified once.

And, if you update the main device next year, but continue to use the same power supply... no need to get a new certification.

This makes it very easy to design your own DAC, buy a power supply from another manufacturer, and avoid most of the pesky certifications.

It's also the main reason why the vast majority of small devices these days use a separate power brick.

At another level, by putting the separate functions in separate boxes, you can treat them separately.

You can have one team designing the power supply, and another team designing the device itself, and each can be tested and updated separately.

(This only makes sense if the project is big enough to involve multiple teams and major upgrades.)

The short answer to your question is that it is sometimes beneficial to put the power supply in a separate box - but the technical reasons aren't especially compelling most of the time these days.

Aug 31, 2018 12:17:01 GMT -5 novisnick said:

Aug 31, 2018 11:51:48 GMT -5 KeithL said:

Personally, I find it naive to assume that you can ALWAYS purchase better performance by spending more money.

Is there a "price no object" power supply that will do a better job of powering an AD1955 DAC chip than Texas Instrument's latest $5 chip regulator?

Perhaps... but I would not bet on it.

The power supply you need for a DAC has some more or less specific requirements.

This can make it somewhat tricky to design, and leaves the door open for both design errors and design improvements.

(So even some really expensive ones may still be badly designed.... but the problem is not the cost limitation.)

I'll take our DC-1 DAC as an example.

If you were to ask me to redesign it, but limit the cost of the power supply to $5, there would be some serious compromises.

However, if you were to offer the opportunity to design a $5000 power supply for it, I very much doubt you could improve the measured performance or sound quality by doing so.

(You could design a power supply that measured better by itself, and certainly one that cost more, and might last for a million years, but it would NOT improve the overall performance of the DAC.)

I would assert that the "design ceiling" for a typical DAC power supply is a few hundred dollars at most.... and you simply cannot get a MEANINGFUL improvement by spending ANY amount above that.

(And, yes, if I spend $1000 for a DAC, and the designer was competent, I expect them to have already included a power supply that good.)

Likewise, if you were a millionaire, and wanted to build a "cost no object" home, you could have your house built with pure 24k gold roofing nails.

And you might even convince someone that gold was best.

However, the reality is that plain old steel roofing nails will work significantly BETTER than solid gold ones.

(The ideal choice would probably be stainless steel or titanium... although plain old galvanized steel has a longer track record of successful use.)

Absolutely Keith! And for the record, I’ve already omitted the uncalled for naivity remark. I was out of line. Now I think we’re swinging in extreme directions, gold nails and all. Yes, I was discussing adding a $300. PS to a $30. DAC. Come on my friend lets be a little more reasonable. 
There are more then a couple of manufacturers that include an External PS port for just such a reason. They have included an adequate PS but also thought that their device would benefit from a better/different PS. 

Question, is it beneficial, in some/many cases to have the PS away from the DAC and its signal carrying our music? I would think so but I Am NOT an engineer.
cross post error, asking while your answering, 😋 Oops. 

[klinemj]

Aug 31, 2018 10:08:55 GMT -5 novisnick said:

Aug 31, 2018 9:56:53 GMT -5 klinemj said:

KeithL

Do you have a good reference source for noise testing on various power supplies? I can't find much out there that's published, other than what I noted about the ifi power supply, the Uptone LPS-1, and the Wyred 4 Sound PS-1. All others I looked at "say" they are low noise but show zero data/zero specs. Therefore, it's hard to see which one might be a "stronger" link than what I am pretty sure is already a weak link.

Based on that, I'm willing to spend $49 to have 1 delivered to try...it may or may not make a difference, but I can return it is it does not.

Mark

Sometimes the specs just Don’t say it all. The “Power” supply and what it gives you can be heard the moment you power up. With a serious DAC your jaw will hit the floor. The Sbooster is such a PS. It is returnable but that thought NEVER crossed my mind. Sometimes you Do Get what you pay for.
My purse isn’t bottomless and I don’t care for waisting money as its hard enough to come by.

In the words of Forest Gump, “and thats all I’ve got to say about that”.

While it is true that specs don't say it all, it is nice to actually see some specs/data. And, noise is only 1 factor. Power ripple is another. I can barely find data on any power supplies for either, but I can for the ifi and the data looks better on noise than 2 higher priced linear power supplies. On power ripple, I found a Computer Audiophile test (see link --->HERE!<---) that shows is lower than everything they tested except a 9v battery in the audible spectrum. It was compared to some lower end LPS's and some SMPS's.

With the data I have seen, it looks pretty good to me. I'd love to see some data/specs on other options. While specs are not the "end all" factor, if something is meaningfully better or worse on a key spec - it's worth noting. And, in the cases of "noise", that seems to be what all the power supply makers talk about. But...few show specs or data.

It looks like my Geek needs a 12v power supply. I have a message in to LH Labs to confirm specs, and if I hear what I think I will from them - I will order the ifi iPower 12v power supply and try it.

Mark

[novisnick]

Aug 31, 2018 14:36:28 GMT -5 klinemj said:

Aug 31, 2018 10:08:55 GMT -5 novisnick said:

Sometimes the specs just Don’t say it all. The “Power” supply and what it gives you can be heard the moment you power up. With a serious DAC your jaw will hit the floor. The Sbooster is such a PS. It is returnable but that thought NEVER crossed my mind. Sometimes you Do Get what you pay for.
My purse isn’t bottomless and I don’t care for waisting money as its hard enough to come by.

In the words of Forest Gump, “and thats all I’ve got to say about that”.

While it is true that specs don't say it all, it is nice to actually see some specs/data. And, noise is only 1 factor. Power ripple is another. I can barely find data on any power supplies for either, but I can for the ifi and the data looks better on noise than 2 higher priced linear power supplies. On power ripple, I found a Computer Audiophile test (see link --->HERE!<---) that shows is lower than everything they tested except a 9v battery in the audible spectrum. It was compared to some lower end LPS's and some SMPS's.

With the data I have seen, it looks pretty good to me. I'd love to see some data/specs on other options. While specs are not the "end all" factor, if something is meaningfully better or worse on a key spec - it's worth noting. And, in the cases of "noise", that seems to be what all the power supply makers talk about. But...few show specs or data.

It looks like my Geek needs a 12v power supply. I have a message in to LH Labs to confirm specs, and if I hear what I think I will from them - I will order the ifi iPower 12v power supply and try it.

Mark

Nothing to lose my friend. 😁 🎶🎶🎶🎶🎶

[geeqner]

Make it even MORE complex:

Something like a DAC may have a mix of both Digital and Analog Inputs and Outputs.

The Power Supply provided by the manufacturer is not always the "BEST POWER SUPPLY FOR THE APPLICATION" - they usually provide the Best Power Supply WITHIN THE COST and PERFORMANCE targets that they are attempting to achieve  (higher-end stuff will usually be better while "joe consumer" stuff will usually cheap-out big-time....)

If an Audio Device has a mix of Digital and Analog circuits - some isolate / separate the "sections", so that the Digital circuitry does not introduce interference into the Analog sections. Sometimes, this is done with a single Power Supply and Isolation / further regulation for the various Sections, and other times there may be SEPARATE Internal Power Supplies feeding the various Sections.

In some of these cases, replacing the "Main" DC Power Supply with a high-quality Linear MAY make a difference / in other cases, if the sub-isolation and/or secondary supplies are properly engineered - you will get NO difference.

If you want to look into Linear Power Supplies with actual specifications and do some experimenting - I would recommend buying them from Electronic Wholesalers.
I have used supplies from Condor (Now owned by SL Power) - [Just search "Condor Linear Power Supplies"]
They manufacture some really well-regulated supplies for a variety of uses / purposes.
you can check them out via suppliers like Allied Electronics, Mouser, and Digikey
They manufacture a wide variety of units with common Voltage Outputs at a variety of Current levels
You would need to provide your own enclosure, AC Line Cord, and DC Power Cable / Connector to go to whatever device you are trying to power.

One LAST Point - you may need to investigate if the circuitry in the device you are attempting to power has an Isolated Ground (Earth / Chassis SEPARATE from "-") or if they are common. Some Power Supplies do not like to be run one way or the other.

[Lonnie]

With all due respect to everyone here I’ve been working with and or designing both linear and switching supplies for more then 25 years. While linear supplies are simple and easy to do they have one huge disadvantage. 60 and 120Hz noise that you can’t get around. It modulates both on the ground planes and the power lines even with all the secondary filtering and storage. Modern day switchers are running up in the megahertz range so second, third and even forth harmonics are way outside the audio band and therefore offer a huge increase in S/N due to the reduced noise floor. It is for this exact reason why we are using them in all new designs. They simply offer better audio performance.

Lonnie

[novisnick]

Sept 10, 2018 20:01:25 GMT -5 Lonnie said:

With all due respect to everyone here I’ve been working with and or designing both linear and switching supplies for more then 25 years. While linear supplies are simple and easy to do they have one huge disadvantage. 60 and 120Hz noise that you can’t get around. It modulates both on the ground planes and the power lines even with all the secondary filtering and storage. Modern day switchers are running up in the megahertz range so second, third and even forth harmonics are way outside the audio band and therefore offer a huge increase in S/N due to the reduced noise floor. It is for this exact reason why we are using them in all new designs. They simply offer better audio performance.

Lonnie

I use both types of power supplies, both high quality products and I cannot hear any noise from either one of them. I trust that the proper engineering will result in a superior product.

At this time I do not have any switching supplies for my speaker amps as I pray my R and X series amps survive me. 😲