Done a lot of mil communications stuff all SMPS, noise is not much different as previously stated. A well designed and layed out SMPS dose not add much if any noise to the envoironment. Power Line Communications, adds noise to the mains, again a adequatly designed SMPS should not add noise to the mains, that is basic EMC design.
SMPS efficient, little waste heat...
As to the most rugged design, I dont think there is anything to choose, though SMPS's tend to win on MTBF figures as they use smaller value capacitors (electrolytics being one of the limiting factors for MTBF calculations.
And as the output is DC (please no one go on about the tiny bit of AC content...) the following logic does not care
SMPS efficient, little waste heat...
As to the most rugged design, I dont think there is anything to choose, though SMPS's tend to win on MTBF figures as they use smaller value capacitors (electrolytics being one of the limiting factors for MTBF calculations.
And as the output is DC (please no one go on about the tiny bit of AC content...) the following logic does not care
And there is the problem, a well designed and built SMPS is expensive. To the point of being rare and out of the range of DIY'ers.
PC power supplies are not well designed, they don't have high quality components, they also do not adhere to EMI interference regulations.
Us diy'ers do not have the time for this, we do not own a test chamber, we do not have the money to debug a highly complex smps psu.
We also do not have the expertiese to even replace switching transistors with a substitute in a smps WHEN it fails.
There is also the issue of the switching transistors, which when blown up usually take out other components along with them, when this happens SMPS is like a rats nest trying to diagnose and repair it, Linear psu's are not.
Wether or not you can safely build a mains connected device such as a SMPS should also not be considered to be a ******* contest, smps has no place at all in a ham radio shack simply because it has been proven time and time again that the Part 5 FCC regulations are inadequate to offer enough protection.
PC power supplies are not well designed, they don't have high quality components, they also do not adhere to EMI interference regulations.
Us diy'ers do not have the time for this, we do not own a test chamber, we do not have the money to debug a highly complex smps psu.
We also do not have the expertiese to even replace switching transistors with a substitute in a smps WHEN it fails.
There is also the issue of the switching transistors, which when blown up usually take out other components along with them, when this happens SMPS is like a rats nest trying to diagnose and repair it, Linear psu's are not.
Wether or not you can safely build a mains connected device such as a SMPS should also not be considered to be a ******* contest, smps has no place at all in a ham radio shack simply because it has been proven time and time again that the Part 5 FCC regulations are inadequate to offer enough protection.
Last edited:
hardly any of your post reflects reality. you can fairly easily get a good SMPS supply these days, for not that much money and they DO have to design for EMI regulations. maybe not for piece of crap PC clones, but for instrumentation or just decent computers they have to. the SMPS will easily outlast the average computer lifetime and your computer is powered by SMPS whether you like it or not.
Last edited:
Again I have done numerous SMPS's design for military/aerospace and medical use (it is one area of PCB design I specalise in), a whole spectrum of uses and thier requirements for noise etc are a lot more stringent than any Ham radio, so a competantly designed SMPS will not give you any problems, and if you are worried filter the input and output to the SMPS.
There is a forum for power supply designs with plenty of SMPS design on thereand a thread started by yours truly on EMC design in relation to SMPS and Class D design.
Not all PC power supply designs are bad, its like all electronics, you do quite often get what you pay for and sometimes less, but there are supplies out there that are competantly designed, and its not only the SMPS design it is how it isintegratedinto the rest of the system.
There is a forum for power supply designs with plenty of SMPS design on thereand a thread started by yours truly on EMC design in relation to SMPS and Class D design.
Not all PC power supply designs are bad, its like all electronics, you do quite often get what you pay for and sometimes less, but there are supplies out there that are competantly designed, and its not only the SMPS design it is how it isintegratedinto the rest of the system.
where from?
I don't see any decent smps power supplies for sale on ebay, and the military standard smps units i've seen are $10-50 grand worth.
And who says that a computer has to last for 2-5 years? I've had 386's that are 30 years old which still work perfectly.
"wether YOU like it or not" your rude you know that.
I'm not even going to fight back with you on the first comment, because thats exactly what you would want, ******* trolls.
Oh btw your back on ignore.
Last edited:
Thank you! I now know where to go for competently designed smps designs and for properly filtering the input and output of smps power supplies.
This information isn't easy to find if you don't know what to look for.
Heres the link:
http://www.diyaudio.com/forums/power-supplies/193705-switcher-emc-design.html
The trouble with SMPS's is that they are unforgiving due to the higher frequency and high dI/dt, whereas linear are more benign and will tolerate bad layout etc better (or with less noise).
http://www.diyaudio.com/forums/power-supplies/193705-switcher-emc-design.html
The trouble with SMPS's is that they are unforgiving due to the higher frequency and high dI/dt, whereas linear are more benign and will tolerate bad layout etc better (or with less noise).
Today you can buy switching regulators that run at 4MHz. At my job we have opamps with 60db of gain running from DC-DC converter bricks. With a bit of filtering they are pretty quiet. Most SMPS today have very little ripple on the output, maybe 50mv p-p at full load. This is not the 1970s anymore.
like ebay is the yardstick for quality SMPS...
haha interesting phrase to be offended by (as well as mispell and add your own capitals like I was shouting) quite the complex youve got going on there. 'whether you like it or not' simply referred to the fact that the board is littered with local SMPS, powering it with a linear power supply is a bit redundant. not one iota of personal commentary.
I like it how you state i'm on ignore like its something I should care about...
i'm using a fully regulated 16A +/-55v (2 x 600W) SMPS for the amps i'm building, its from Audiopower, the noise is ridiculously low, as are the emissions.
it operates at >90% efficiency, with 150nS power response and in the span of up to 500kHz the highest peak is -84dB at 150.4kHz
but I gather they also do work for industry as well as MIL, so even though I havent looked for a low voltage version, i'm sure its available from someone for less than 5-10K... . these are ~$200ea. used to power a class AB poweramp, THD+N is 0.0002% and full power 16A +/-55v pulse performance is not like anything ive ever seen before from a linear supply.
as mentioned, its not the 20th century anymore.
haha interesting phrase to be offended by (as well as mispell and add your own capitals like I was shouting) quite the complex youve got going on there. 'whether you like it or not' simply referred to the fact that the board is littered with local SMPS, powering it with a linear power supply is a bit redundant. not one iota of personal commentary.
I like it how you state i'm on ignore like its something I should care about...
i'm using a fully regulated 16A +/-55v (2 x 600W) SMPS for the amps i'm building, its from Audiopower, the noise is ridiculously low, as are the emissions.
it operates at >90% efficiency, with 150nS power response and in the span of up to 500kHz the highest peak is -84dB at 150.4kHz
but I gather they also do work for industry as well as MIL, so even though I havent looked for a low voltage version, i'm sure its available from someone for less than 5-10K... . these are ~$200ea. used to power a class AB poweramp, THD+N is 0.0002% and full power 16A +/-55v pulse performance is not like anything ive ever seen before from a linear supply.
as mentioned, its not the 20th century anymore.
Hi
I wish to state that i start this project because i have most of the parts in my bin and is aware of the debate about smps and linear dc for pc audio.
My smps is is cheap product and not what you called audiophile standard.
This is not an exercise to convince anybody else but myself of the sound benefits or not.
My opinion is only my opinion which is very subjective. After all, this is a diy forum and anyone can share the opinions.
If you think that i am on a futile exercise, i can accept that.
One of the aims of my posts is to gain knowledge about regulated dc power supply as i am hoping to use one of this supply for an amp i am planning to build in the near future.
I apologies if i have irritated anybody by my posts
cheers
regards
kp93300
I wish to state that i start this project because i have most of the parts in my bin and is aware of the debate about smps and linear dc for pc audio.
My smps is is cheap product and not what you called audiophile standard.
This is not an exercise to convince anybody else but myself of the sound benefits or not.
My opinion is only my opinion which is very subjective. After all, this is a diy forum and anyone can share the opinions.
If you think that i am on a futile exercise, i can accept that.
One of the aims of my posts is to gain knowledge about regulated dc power supply as i am hoping to use one of this supply for an amp i am planning to build in the near future.
I apologies if i have irritated anybody by my posts
cheers
regards
kp93300
kp93300, you have not irritated anyone (I may have ).
I think this is a very valid discussion, and I have asked my question above in all seriousness as I want to determine what some see the benefits of linear over SMPS for PC's used for playing audio.
SMPS's have had a bad press over the years and still do, this is unfounded to an extent as the technology and techniques have improved and changed greatly over the last 20+ yeards. Also our understanding and methods of EMC prevention have grown and improved over the same period.
What has also changed over the same period in the world of digital is higher density of switching nodes on a single device (Moores law) and more improtantly Tr (the rise time, and fall time) of digital signals, this has increased the noise produced by the switching itself, simultaneous switching noise.
More later goto go to work.
).I think this is a very valid discussion, and I have asked my question above in all seriousness as I want to determine what some see the benefits of linear over SMPS for PC's used for playing audio.
SMPS's have had a bad press over the years and still do, this is unfounded to an extent as the technology and techniques have improved and changed greatly over the last 20+ yeards. Also our understanding and methods of EMC prevention have grown and improved over the same period.
What has also changed over the same period in the world of digital is higher density of switching nodes on a single device (Moores law) and more improtantly Tr (the rise time, and fall time) of digital signals, this has increased the noise produced by the switching itself, simultaneous switching noise.
More later goto go to work.
Sound comparison-linear power vs smps power for pc audio
I did a sound comaprison between linear power and smps power for my atom based audio pc running on linux.
I organize 2 sessions with 2 guests per session. This was done just now and yesterday .
My guest for 4 middle age man between 40 to 60 yr and have been playing hi fi for more than 20yrs and own expensive equipments .
I am the only one to know the type of power supply in use.
The 4 listerners are unanimous in concluding that there is a difference and all 4 are in favour of linear power !
The description of the sound difference is also remarkedly similiar.
I will try to describe.
Linear ps gives a darker background , better silence between noted with more natural decay of piano and guitar . Harmonics are more realistic.
Bass seems more with smps but it is more homogenous and separation between the different types of drum is blurred.
Vocal is more natural in linear with better depiction of microdynamic .
The high notes are cleaner with linear ps.
Sound stage is better and hold up better when the music rise in tension with linear ps.
BTW,
I have a very modest set up with Buffalo 2,
RJM buffer circuit which is from one of our member here,
amp is here
http://www.diyaudio.com/forums/class-d/217903-ti-tpa3100d2-amazing-class-d-amp.html
speaker is V2 from GR research V-2
cheers
kp93300
PS I do not expect expect everyone to agree with my observations and I accept dissenting views .
I did a sound comaprison between linear power and smps power for my atom based audio pc running on linux.
I organize 2 sessions with 2 guests per session. This was done just now and yesterday .
My guest for 4 middle age man between 40 to 60 yr and have been playing hi fi for more than 20yrs and own expensive equipments .
I am the only one to know the type of power supply in use.
The 4 listerners are unanimous in concluding that there is a difference and all 4 are in favour of linear power !
The description of the sound difference is also remarkedly similiar.
I will try to describe.
Linear ps gives a darker background , better silence between noted with more natural decay of piano and guitar . Harmonics are more realistic.
Bass seems more with smps but it is more homogenous and separation between the different types of drum is blurred.
Vocal is more natural in linear with better depiction of microdynamic .
The high notes are cleaner with linear ps.
Sound stage is better and hold up better when the music rise in tension with linear ps.
BTW,
I have a very modest set up with Buffalo 2,
RJM buffer circuit which is from one of our member here,
amp is here
http://www.diyaudio.com/forums/class-d/217903-ti-tpa3100d2-amazing-class-d-amp.html
speaker is V2 from GR research V-2
cheers
kp93300
PS I do not expect expect everyone to agree with my observations and I accept dissenting views .
Hi kp93300
Nothing particularly surprising in your findings. I know nothing about Atom boards but quite possibly you can take this a step further by disabling on-board switching regulators one by one and supplying your own regulated voltages. Are you able to identify the regulators and the corresponding circuits they supply?
Nothing particularly surprising in your findings. I know nothing about Atom boards but quite possibly you can take this a step further by disabling on-board switching regulators one by one and supplying your own regulated voltages. Are you able to identify the regulators and the corresponding circuits they supply?
Qusp, things are not so obvious or straightforward. I am sure you are aware of sata filters, bespoke CAT5 and USB cables and other illogical entities.
At the time of the first asynchronous USB device most listeners expected the end of "audiophile" digital cabling, dependence on PC OSs, software and hardware. It wasn't to be. Next step was the ram buffer, optical isolation and subsequent recklocking. Did it eliminate the PC from the list of variables in the sound chain? Hardly.
For me it's far from clear what the linear supply achieves. The improvement in sound may be entirely coincidental and have nothing to do with the dac but rather with the level of RF pollution in the room. I recently used a class D amp to amplify a sine wave for my turntable motor. It achieved that admirably but made the sound of my phono setup extremely unpleasant just by spraying the room with RF.
No one in their right mind expects a 317 to have any bandwidth beyond a few kHz. So, what?
At the time of the first asynchronous USB device most listeners expected the end of "audiophile" digital cabling, dependence on PC OSs, software and hardware. It wasn't to be. Next step was the ram buffer, optical isolation and subsequent recklocking. Did it eliminate the PC from the list of variables in the sound chain? Hardly.
For me it's far from clear what the linear supply achieves. The improvement in sound may be entirely coincidental and have nothing to do with the dac but rather with the level of RF pollution in the room. I recently used a class D amp to amplify a sine wave for my turntable motor. It achieved that admirably but made the sound of my phono setup extremely unpleasant just by spraying the room with RF.
No one in their right mind expects a 317 to have any bandwidth beyond a few kHz. So, what?
I give up, a linear PSU will have only a little difference to the noise spectrum you will find in your PC...layout, decoupling etc are the main contributers as a lot of the noise is generated by the circuitry switching. Get a de3cent scope and do some measurements you may get a shock.
What this does confirm to me is that the high frequency noise superimposed on the analogue signal is not audible.
As to sptraying the room with RF, get somthing thgat has been designed properly and dosn't spray the room with rf, thats common sense.
It is only after the DAC that we get analogue, up to that point the digital streem dose not care whaqt the PSU is and linerar or SMPS will not make any difference to the digital data flow.
Anyway its been an interesting discussion, I was hoping others would have some empirical data (measurements) that I could compare to measurements I have (I can not post them as they are covered by NDA's and other restricions), from these findings and subsequent tests developemnt, linear supplies lost out, due to bulk heat and noise already in the audio spectrum.
I must ask what is the difference between audiophile digital cabling and normal digital cabling, digital cabling is digitalk cabling, it very nicely follows the transmission line theorys first developed by Heaviside, so can and is engineered adequatly without any audiophile esoteric features.
A lot of the stuff being sold regarding digital audio is pure snake oil, it is a whole new market to pray and influence people fear of this new (!!!) digital age. Myths like solid state disks sounding better!!! Digital requires proper engineering practices and measurements to detect and solve problems...
What this does confirm to me is that the high frequency noise superimposed on the analogue signal is not audible.
As to sptraying the room with RF, get somthing thgat has been designed properly and dosn't spray the room with rf, thats common sense.
It is only after the DAC that we get analogue, up to that point the digital streem dose not care whaqt the PSU is and linerar or SMPS will not make any difference to the digital data flow.
Anyway its been an interesting discussion, I was hoping others would have some empirical data (measurements) that I could compare to measurements I have (I can not post them as they are covered by NDA's and other restricions), from these findings and subsequent tests developemnt, linear supplies lost out, due to bulk heat and noise already in the audio spectrum.
I must ask what is the difference between audiophile digital cabling and normal digital cabling, digital cabling is digitalk cabling, it very nicely follows the transmission line theorys first developed by Heaviside, so can and is engineered adequatly without any audiophile esoteric features.
A lot of the stuff being sold regarding digital audio is pure snake oil, it is a whole new market to pray and influence people fear of this new (!!!) digital age. Myths like solid state disks sounding better!!! Digital requires proper engineering practices and measurements to detect and solve problems...
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.