Yes it could but what is the purpose of using a switcher that is way more expensive than a good linear PSU (and still needs an LC filter)? The latter won't polluting mains voltage as well. Main selling points of SMPS are low price, low weight, high efficiency etc.
I know it is fashionable to do simple things the complicated and, most important, a "convenient" way but in the case of SMPS one has an HF/RF generating device both at AC and DC side. It really is simpler to do it linear for best results with regards to clean DC and low noise. If price is the decisive factor and good sound quality is not relevant then SMPS are the solution. Any low power device can be tested with both and that is convincing to most.
Do the test with a DAC, streamer or any good audio device under 25VA and the SMPS will be with the potato peels in the garbage can. Low price SMPS as delivered with such devices anyway, some mid priced SMPS might be acceptable but these will be outperformed by well designed low noise LPS as well.
I know it is fashionable to do simple things the complicated and, most important, a "convenient" way but in the case of SMPS one has an HF/RF generating device both at AC and DC side. It really is simpler to do it linear for best results with regards to clean DC and low noise. If price is the decisive factor and good sound quality is not relevant then SMPS are the solution. Any low power device can be tested with both and that is convincing to most.
Do the test with a DAC, streamer or any good audio device under 25VA and the SMPS will be with the potato peels in the garbage can. Low price SMPS as delivered with such devices anyway, some mid priced SMPS might be acceptable but these will be outperformed by well designed low noise LPS as well.
Last edited:
I did and these are well engineered workhorses.
I´m with Jean-Paul though, if I didn´t have these I would use a linear PSU for everything I do. Cleaner, less hassle & defects; only misses the regulated output voltage but even that is comparatively cheap and easy to do for linear PSUs as well.
A well designed linear PSU IS one that has regulated voltage preferably with a ultra low noise LDO design to limit energy loss and limit generated heat. Old proven low noise technology while trying to meet 2021's demands.
It really seems we are in a period that many haven't grown up with LPS in their youth and all of a sudden linear power supplies are strange beasts contrary to wall wart stuff that everyone seems to like (tip: use a scope and check how good it is). No, LPS can be state of the art and they're very worthwhile to DIY. These don't need extra output filters to filter their self generated pollution (why would one tolerate THAT in audio? It's like sh***ing in your pants deliberately and then be forced to clean it up later) and they need an input filter to filter the self generated garbage other SMPS in your home produce. Simple. The quality of any audio device is determined by the quality of the power supplies. Or is good audio just doing something the most easy and convenient way?
Again, for devices under around 25VA. I have very good experiences with industrial SMPS and JLH class A amplifiers but that is another area.
It really seems we are in a period that many haven't grown up with LPS in their youth and all of a sudden linear power supplies are strange beasts contrary to wall wart stuff that everyone seems to like (tip: use a scope and check how good it is). No, LPS can be state of the art and they're very worthwhile to DIY. These don't need extra output filters to filter their self generated pollution (why would one tolerate THAT in audio? It's like sh***ing in your pants deliberately and then be forced to clean it up later) and they need an input filter to filter the self generated garbage other SMPS in your home produce. Simple. The quality of any audio device is determined by the quality of the power supplies. Or is good audio just doing something the most easy and convenient way?
Again, for devices under around 25VA. I have very good experiences with industrial SMPS and JLH class A amplifiers but that is another area.
Last edited:
"if I didn´t have these I would use a linear PSU for everything I do"
That's the thing, isn't it.
They are so good that make linear supplies unnecessary in most applications.
Don't get me wrong, I like linear PSUs as much as the next guy, but if I can use a Vicor I will always prefer it.
And if one wants extremely low noise then a simple CapMx after it will do just that.
That's the thing, isn't it.
They are so good that make linear supplies unnecessary in most applications.
Don't get me wrong, I like linear PSUs as much as the next guy, but if I can use a Vicor I will always prefer it.
And if one wants extremely low noise then a simple CapMx after it will do just that.
Yes, the stupidity of that occurred to me last night late as I was researching LC for 4A 230V mains! (cost GBP50-200 per inductor depending on current vs the cost for digital 😱)
I can see why now people use a choke on the lower current tubes and resistor on the higher current.
Last edited:
Well the application is "cost-effective and high-performance DIY" here mostly, so..
Two flatpacks (200W each) which together can create a symmetrical supply cost roughly 2x600€.
Of course they are compact and light but at a cost.
Transformer+rectifier+LDOs+heatsinks can probably be had for a good 1/10th of that price.
With 400W of needed power I would nowadays not even think of linear PSU's. There is SMPS king both with specs and with price/quality ratio. The other solution in the case of amplifiers (to avoid SMPS) is to build small amplifiers and again use LPS with regulated voltage.
This confusion probably makes people say that SMPS are "so good that they make linear supplies unnecessary in most applications". SMPS are not good at all in low power applications when one checks noise and ripple. Add low output impedance to that and those are the 3 most important parameters for best results in such low power applications. SMPS is higher in ripple and noise by design....
In a class A power amplifier however LPS and regulation are very inefficient and it will even make heat output way higher. With such high power draw ripple will be hard to get rid off and it will easily be heard. Regulation could solve the ripple but that would be a hefty regulator indeed. There an SMPS will be beneficial as hearing ripple is annoying which the SMPS will not produce that as much as an LPS. Therefor the distinction between low/high power applications.
This confusion probably makes people say that SMPS are "so good that they make linear supplies unnecessary in most applications". SMPS are not good at all in low power applications when one checks noise and ripple. Add low output impedance to that and those are the 3 most important parameters for best results in such low power applications. SMPS is higher in ripple and noise by design....
In a class A power amplifier however LPS and regulation are very inefficient and it will even make heat output way higher. With such high power draw ripple will be hard to get rid off and it will easily be heard. Regulation could solve the ripple but that would be a hefty regulator indeed. There an SMPS will be beneficial as hearing ripple is annoying which the SMPS will not produce that as much as an LPS. Therefor the distinction between low/high power applications.
Last edited:
Here's a good solution for el-cheapo SMPS:
PO89ZB , an inline DC filter for SMPS wall warts . Preamps, HPA, Korg NuTube, etc
PO89ZB , an inline DC filter for SMPS wall warts . Preamps, HPA, Korg NuTube, etc
Yeah, like I said earlier: it's like sh***ing in your pants deliberately and then be forced by yourself to clean it up later. And what about the mains side?
I am surprised not many build such low power devices with a LPS. One would think that when a SMPS needs an extra filter to be acceptable choosing a LPS would make it a shorter path to best results straight away. Well, make that a fact.
I am surprised not many build such low power devices with a LPS. One would think that when a SMPS needs an extra filter to be acceptable choosing a LPS would make it a shorter path to best results straight away. Well, make that a fact.
Last edited:
jean-paul,
There is no confusion here.
Not every SMPS, or SMPS in general, but Vicor. Devices that are employed in satellite comms equipment where ultra noise and stability is extremely important.
So it will be good if one does not make general statements.
Fully agree with the rest of your post though.
It is a bit of a pointless discussion here because as per usual it depends on application and quality of LPS/SMPS
I for example would definitely consider a 400W LPS. Why not? It´s cheaper, much more robust than your typical Meanwell and if my amp´s PSRR is high enough I can even skip the regulator, so why not? Look at all the posts from people describing noise/hiss or whatever due to application of SMPS in their amps.
Do you find many people here use SMPS for their class-AB amps?
Do you find many people here use regulated LPS for their amps?
The answer is clearly 2x NO.
A SMPS for sensitive low-power-applications? Take Jan Diddens silent switcher. It´s probably close to a battery with regards to noise if you do it right. (like cabling, shielding etc.)
Unless you have modulation or similar effects (like in post 2) the typical SMPS-noise/-frequencies also are far away from audio frequencies,
so again it always depends.
I for example would definitely consider a 400W LPS. Why not? It´s cheaper, much more robust than your typical Meanwell and if my amp´s PSRR is high enough I can even skip the regulator, so why not? Look at all the posts from people describing noise/hiss or whatever due to application of SMPS in their amps.
Do you find many people here use SMPS for their class-AB amps?
Do you find many people here use regulated LPS for their amps?
The answer is clearly 2x NO.
A SMPS for sensitive low-power-applications? Take Jan Diddens silent switcher. It´s probably close to a battery with regards to noise if you do it right. (like cabling, shielding etc.)
Unless you have modulation or similar effects (like in post 2) the typical SMPS-noise/-frequencies also are far away from audio frequencies,
so again it always depends.
The Silent Switcher is called like this as it uses ultra low noise linear regulators to be silent 🙂 That is a hybrid with best of both so to speak.
True but Vicor is one of the exceptions. The average SMPS as used by many is far from OK compared to even an average LPS. The point I was trying to make is that it depends on the application but for low power you almost can not go wrong with an LPS and there are many chances things don't work out that OK with an SMPS.
True but Vicor is one of the exceptions. The average SMPS as used by many is far from OK compared to even an average LPS. The point I was trying to make is that it depends on the application but for low power you almost can not go wrong with an LPS and there are many chances things don't work out that OK with an SMPS.
Last edited:
Make me a linear 12V 100A power supply with 20mV ripple on the output and 90% efficiency. Make it fit in an ATX PSU Chassis. I'll wait. 🙄
Right now, I'm using SMPS and DC-DC boost for a tube phono stage. If I turn the volume up to max I can hear some noise. At that volume, the record would drive the amp connected to the circuit well into clipping, so who cares?
I plan on adding some sort of power filter (probably either an LC filter or a linear regulator) but for most people, it's clean enough.
Right now, I'm using SMPS and DC-DC boost for a tube phono stage. If I turn the volume up to max I can hear some noise. At that volume, the record would drive the amp connected to the circuit well into clipping, so who cares?
I plan on adding some sort of power filter (probably either an LC filter or a linear regulator) but for most people, it's clean enough.
While you wait we buy an SMPS for that 12V 100A task. Using an SMPS and DC-DC boost (followed by a future linear regulator to make it silent....) for a phonopre is just eh.....a complicated way of doing things but probably a wise thing to do in some circles 🙂 When I was involved in phono stuff somewhere in the eighties we used to make power supplies as silent as possible as then phono was very low output voltage and very susceptible to noise. We thought it to be a real joy when we heard undistorted music and as least noise as possible.
Last edited:
I mean sure, some SMPS are noisy crap.
I have one that dumps so much noise out of it that if I have the USB scope connected, the mouse and touchscreen stop working properly! It was so bad the DAC would stop outputting if it was connected to the same ground. It's the only one I've seen do that and so I feel it's defective.
Changed it out for a different model and voila, no more noise or oscillations. Yes, I know that making a 12V 8A linear supply isn't terribly hard, but it's way more costly, heavier, physically larger, and the linear 8A regulator would need a heatsink and take up even more space.
There's a reason we don't use Tungar rectifiers anymore, too 🙂
http://www.r-type.org/pdfs/tungar.pdf
I have one that dumps so much noise out of it that if I have the USB scope connected, the mouse and touchscreen stop working properly! It was so bad the DAC would stop outputting if it was connected to the same ground. It's the only one I've seen do that and so I feel it's defective.
Changed it out for a different model and voila, no more noise or oscillations. Yes, I know that making a 12V 8A linear supply isn't terribly hard, but it's way more costly, heavier, physically larger, and the linear 8A regulator would need a heatsink and take up even more space.
There's a reason we don't use Tungar rectifiers anymore, too 🙂
http://www.r-type.org/pdfs/tungar.pdf
Most of the consumer grade SMPS and/or the standard wall warts are noisy crap yes.
Well it is a true art to build LPS the best you can as they are more costly, heavier, physically larger, and they would need a large heatsink if you would not design uLDO. The challenge goes on to make them as small, efficient AND ultra allow noise as possible. The results are rewarding. I built around 13 in lockdown and every single one of them is way better than the various SMPS I tested. My goal may differ though, I am in the hobby for maximum quality I can make with my own hands at an acceptable price/quality ratio and not for absolute lowest price and plug and play convenience.
About the tungar... I use a 1922 mercury rectifier of that kind to charge my iPhone. As stubborn some may choose the absolute worst kind of PSU for low power (even highly sensitive analog) audio, some persist in even more stupid choices 😀
Well it is a true art to build LPS the best you can as they are more costly, heavier, physically larger, and they would need a large heatsink if you would not design uLDO. The challenge goes on to make them as small, efficient AND ultra allow noise as possible. The results are rewarding. I built around 13 in lockdown and every single one of them is way better than the various SMPS I tested. My goal may differ though, I am in the hobby for maximum quality I can make with my own hands at an acceptable price/quality ratio and not for absolute lowest price and plug and play convenience.
About the tungar... I use a 1922 mercury rectifier of that kind to charge my iPhone. As stubborn some may choose the absolute worst kind of PSU for low power (even highly sensitive analog) audio, some persist in even more stupid choices 😀
Last edited:
At my old job we made equipment that had Vicor power bricks in the supply. Using their active filter modules the output was extremely clean. Absolutely no noise problems with analog circuits having 50-60db of gain in the audio range.
Sadly, there are crap linear and SMPS's available. Good ones of both types exist and are available.
However, as I am somewhat concerned with the environment, I stick with the better SMPS's. Your choice, but using far less power does matter to many of us and keeps our electrics bills less as well. This is DIY, use whatever makes you happy.
However, as I am somewhat concerned with the environment, I stick with the better SMPS's. Your choice, but using far less power does matter to many of us and keeps our electrics bills less as well. This is DIY, use whatever makes you happy.
Yes we want to make a point by mentioning leading class SMPS when the average DIYers uses the generic/cheap ones 😉 Again, even an average performing LPS performs better in the for audio relevant areas compared to low and mid priced SMPS.
If one is really concerned with environmental issues then one should take a look at the landfill where there are heaps of SMPS. A linear PSU is also normally built with a real mains switch (too expensive!) and does not use any energy when switched off contrary to supposedly green stealth energy consumers that consume 24/7 for years until they go out with a bang. I visited a customer a few months ago where a container was situated with 1000's of recalled wall warts. A linear PSU in general has a longer service life than 2 years.
In my branch we have linear 48V rectifiers/PSU's with passive heatsink we only replace when there is a need for more power than the system can deliver. These rarely fail and are all older than 23 years. Their modern modular actively cooled SMPS brothers fail more often. Anecdotical but I see this so often in various companies that I can not ignore it. Simple technology outlives complex switching technology easily and such often with better specs as well.
If one is really concerned with environmental issues then one should take a look at the landfill where there are heaps of SMPS. A linear PSU is also normally built with a real mains switch (too expensive!) and does not use any energy when switched off contrary to supposedly green stealth energy consumers that consume 24/7 for years until they go out with a bang. I visited a customer a few months ago where a container was situated with 1000's of recalled wall warts. A linear PSU in general has a longer service life than 2 years.
In my branch we have linear 48V rectifiers/PSU's with passive heatsink we only replace when there is a need for more power than the system can deliver. These rarely fail and are all older than 23 years. Their modern modular actively cooled SMPS brothers fail more often. Anecdotical but I see this so often in various companies that I can not ignore it. Simple technology outlives complex switching technology easily and such often with better specs as well.
Last edited: