Then medical supplies are the way to go - they are designed not to fail because somebody could die. Cost more than run of the mill switchers, or even the equivalent big iron.
No one ever *died* if their stereo just quit due to a bad SMPS, so they don’t have to worry so much about reliability past the warranty period. As long as it just blows a fuse and not burn the house down.
They’re not designed not to fail (if you’ll pardon the double negative)
How can you design electronics not to fail ….. have a strong word with it as you’re building it?
Medical grade (IEC 60601 ) is about safety, leakage current etc
Well, SMPS's that are at the same cost point of Huge transformers/linear supply are mostly top quality and likely to last a very long time.
IMHO comparing at the same price point does make a difference. Going cheap with any type of supply is not likely to give long life.
Greg
IMHO comparing at the same price point does make a difference. Going cheap with any type of supply is not likely to give long life.
Greg
Yes, it is more about keeping dangerous voltage off the patient, even if the supply fails. Short of a direct lightning strike which would set the whole building on fire, it should keep from zapping a patient or doctor. But they can’t afford to go through power supplies like water, either. So they DO tend to be better than consumer electronics supplies. Not that they NEVER fail, but at rated load you’re not driving it to an early grave like you would be with the one in some cheap powered subwoofer. Those *are* designed so that if you play it loud long enough, you will be buying another subwoofer. Good for sales.
So why use smps ? One good reason Ive found with low power valve stages like phono's, pre amps and headphone amps is the lack of Low Frequency noise. Ive been driven
nearly mad with linear supply versions in the past trying to get completely rid of hum etc . This usually meant using two chassis to keep the transformers away from sensitive circuitry.
With smps this isnt a problem and they are totally hum free and much more compact. Of course now you can get HF noise instead but Ive found that that is easier to deal with
than LF noise. For me they measure ( noise) and sound better than my linear supply versions
For good reliability ( I hope!) Ive been using good quality ac/dc (240V/48V) supplies from Meanwell then a self built filtered dc/dc boost converter for the HT (240V) and Traco/XP
dc/dc converters with filtration for the heaters. Ive had a 2P2 9L headphone amp powered this way in use for over two years now. This amp is switched off and on regularly each day
when experimenting with other devices and so far its been very reliable.
At higher powers it might be a different matter
nearly mad with linear supply versions in the past trying to get completely rid of hum etc . This usually meant using two chassis to keep the transformers away from sensitive circuitry.
With smps this isnt a problem and they are totally hum free and much more compact. Of course now you can get HF noise instead but Ive found that that is easier to deal with
than LF noise. For me they measure ( noise) and sound better than my linear supply versions
For good reliability ( I hope!) Ive been using good quality ac/dc (240V/48V) supplies from Meanwell then a self built filtered dc/dc boost converter for the HT (240V) and Traco/XP
dc/dc converters with filtration for the heaters. Ive had a 2P2 9L headphone amp powered this way in use for over two years now. This amp is switched off and on regularly each day
when experimenting with other devices and so far its been very reliable.
At higher powers it might be a different matter
Putting the "reliability" concern aside for the moment....
Why are SMPS not preferred in audio amplifiers? They have way less weight and mechanical noise. I was reading randy sloanes amplifier construction guide and IIRC only downside (other than reliability) I can see he mentions is that they do not have the 'transient' ability to deliver current for the dynamic swings of music, but then provides no scientific backing for this. (I may be misquoting him, will have to go back to the book)
Is this true, that SMPS cannot deliver the same transient current swings that a linear supply can, provided the signal stays within the peak power capabilities of both supplies i.e. the music average is around 5W and never peaks above 200W, and both supplies are capable of supplying 1000W? I know the SMPS will "hard" limit at 1000W, whereas the linear supply will likely go much further than that and you get "power supply headroom", but does it matter if our peaks are only 200W max?
Is this why people say SMPS designs sound "flat"? Can anyone answer this, with some proof?
FWIW, I've never in my personal life had a SMPS fail touch wood. Be it laptop charger, phone charger, computer PS, shaver charger you name it. Where I have seen loads of failures is in my work life - and that is the little power supplies for LED lights (strips or pot/down lights). Here is where they are made with the absolute cost as number one priority, these things are a joke when you open them up. They are ridiculously cheap.
Why are SMPS not preferred in audio amplifiers? They have way less weight and mechanical noise. I was reading randy sloanes amplifier construction guide and IIRC only downside (other than reliability) I can see he mentions is that they do not have the 'transient' ability to deliver current for the dynamic swings of music, but then provides no scientific backing for this. (I may be misquoting him, will have to go back to the book)
Is this true, that SMPS cannot deliver the same transient current swings that a linear supply can, provided the signal stays within the peak power capabilities of both supplies i.e. the music average is around 5W and never peaks above 200W, and both supplies are capable of supplying 1000W? I know the SMPS will "hard" limit at 1000W, whereas the linear supply will likely go much further than that and you get "power supply headroom", but does it matter if our peaks are only 200W max?
Is this why people say SMPS designs sound "flat"? Can anyone answer this, with some proof?
FWIW, I've never in my personal life had a SMPS fail touch wood. Be it laptop charger, phone charger, computer PS, shaver charger you name it. Where I have seen loads of failures is in my work life - and that is the little power supplies for LED lights (strips or pot/down lights). Here is where they are made with the absolute cost as number one priority, these things are a joke when you open them up. They are ridiculously cheap.
Last edited:
They are preferred in many audio amplifiers today (unfortunately). Even at the A brands they do not have a long service life compared to linear PSU's. As they need to be cheap in consumer electronics they are also often too low rated in power. So they are at the crossing point of economics and decline in demand of audio equipment. Ultra cheap ones are nice transmitters of garbage as well. I have seen waste containers full with hundreds of defective ones at Internet service providers. I experienced a defective one of an expensive LED desk lamp only last week. Had to pay 55 Euro for a new one, said I thought that was odd to have to pay after only a year using it and got 1 sent to me for free. Pfff.
You'll always hear that very good ones exist. They do exist. Just not in your 500 Euro costing amplifier.
If you build stuff and wish it to last .. Build linear PSUs. if you build class A power amplifiers SMPS can be beneficial as the hum will not be there. For low power applications like DACs, preamps, media players... linear PSUs are easily outperforming most SMPS. If you take care to design low loss then the PSU will be relatively green as well, greener than it cousin SMPS that ends on landfill after 2 years. Noise of good linear PSUs is expressed in µV and that of SMPS in tens of mV. Stuff that switches at high frequencies and such 24/7 for years as no power switch is used and that in an airtight wall wart. What do we expect?
For some unknown reason the trend is to see SMPS as superior and linear PSUs suddenly are the lesser kind. Can't figure out why but it sure does not correlate with reality. Here on diyaudio some projects are done with SMPS. This is not because of their quality but for safety, lower cost and easier building. Proof of their quality is that they need an aftermarket filter to filter the noise
PS hum issues as described in post #26 depend on building skills. If linear PSU's would be the cause then many brands would not have sold their equipment the past 50 years.
You'll always hear that very good ones exist. They do exist. Just not in your 500 Euro costing amplifier.
If you build stuff and wish it to last .. Build linear PSUs. if you build class A power amplifiers SMPS can be beneficial as the hum will not be there. For low power applications like DACs, preamps, media players... linear PSUs are easily outperforming most SMPS. If you take care to design low loss then the PSU will be relatively green as well, greener than it cousin SMPS that ends on landfill after 2 years. Noise of good linear PSUs is expressed in µV and that of SMPS in tens of mV. Stuff that switches at high frequencies and such 24/7 for years as no power switch is used and that in an airtight wall wart. What do we expect?
For some unknown reason the trend is to see SMPS as superior and linear PSUs suddenly are the lesser kind. Can't figure out why but it sure does not correlate with reality. Here on diyaudio some projects are done with SMPS. This is not because of their quality but for safety, lower cost and easier building. Proof of their quality is that they need an aftermarket filter to filter the noise
PS hum issues as described in post #26 depend on building skills. If linear PSU's would be the cause then many brands would not have sold their equipment the past 50 years.
Last edited:
Some may not be aware that the diyAudio Store sells an SMPS post-filter for USD 10. It's a very small PCB with nine electronic components. Max current is 3 amperes and max voltage is 48V; it's aimed at preamps, line stages, headphone amps, that kind of thing. Total resistance from unfiltered DC input to filtered DC output is 0.1 ohms. There are 14 enthusiastic reviews on the Store's page, and hundreds of positive comments in the Forum thread. The Store provides a link to that thread.
I'm confident that accomplished designers can scale the basic design up or down, to meet their own specific needs.
I'm confident that accomplished designers can scale the basic design up or down, to meet their own specific needs.
I hope you realize that there are over a thousand DIYers who have built the "ACA" class-A amplifier designed by Nelson Pass, which uses a SMPS for its power supply, right here on diyAudio. And >99% of them are thrilled with the sound quality it produces.
There are also 200 DIYers who have built Nelson Pass's "VFET" class-A amplifier, which also uses a SMPS for its power supply, right here on diyAudio. And those folks are ecstatic with its sound quality.
And there are even more DIYers who have built the "ACA+" headphone amp + preamp which uses a SMPS for its power supply. They love its sound too.
Add to that the Korg NuTube preamp, the Noir headphone amp, the Butte headphone amp, the list of audio projects powered by SMPS goes on and on. DIY fanatics love the sound these devices produce.
There are also 200 DIYers who have built Nelson Pass's "VFET" class-A amplifier, which also uses a SMPS for its power supply, right here on diyAudio. And those folks are ecstatic with its sound quality.
And there are even more DIYers who have built the "ACA+" headphone amp + preamp which uses a SMPS for its power supply. They love its sound too.
Add to that the Korg NuTube preamp, the Noir headphone amp, the Butte headphone amp, the list of audio projects powered by SMPS goes on and on. DIY fanatics love the sound these devices produce.
It's a myth to think a "big iron" power supply can outperform a regulated smps. It is a myth propagated by those with little knowledge of electronics engineering. Unfortunately this lack of understanding has pressured many manufacturers to keep providing old school products despite the cost and weight disadvantages. The old American adage for a car purchase was more lbs/$. Perhaps that exists in the amplifier world too.
High reliability electronics engineering simply comes down to over speccing everything that gets hot. I used to work in the aerospace industry where there were 30 y/o smps's that had been running 24/7 their entire lives without failure. Failure rates were usually lower than their linear counterparts that usually ran hotter because of their inefficiency.
High reliability electronics engineering simply comes down to over speccing everything that gets hot. I used to work in the aerospace industry where there were 30 y/o smps's that had been running 24/7 their entire lives without failure. Failure rates were usually lower than their linear counterparts that usually ran hotter because of their inefficiency.
(sorry if I bodged the quote -- had to do it manually . . )mainframe99 said:
Yes, but there's more to it than just raw Watts: If the output devices are saturated delivering 200W to the load, the power supply's hard limiting at that power (or some higher power) won't matter in the slightest.
Just my 2¢
Cheers
It wasn't untill not so long ago, when the pwm of the smps was to low so it affected sound and the filtering was not aquedate yet.But since more than while those issues were solved and most good smps's are aboslute silent in the audio passband and perfectly fit. But the hifi world is very conservative, so...
But i would love to have my tube amps with smp's, they are lighter, cheaper and often better regulated than even a regulated linear psu these days. For me there is no technical reason anymore to keep using linear psu's, just make sure your smps is overspec'ed for a tube amp so it can deal with the voltage swings. I know some are busy with it, even for SET amps. It's just a matter of time and those designs will appear on the market.