When putting my DAC together, starting out on the cheap I found out that I had a couple of spare 12V/2A wall warts that came from external hard drives lying on a shelf. These were made by Asian Power Devices, and one performed better than the other - one was pretty good with little discernible noise; the other had a high-pitched oscillating whine that was detectable with nothing playing.
Having read on here that Leicke power supplies are good quality, I decided to try one out, and I'm very impressed with the overall package - it's smaller than the wall wart I was using before and barely heats up at all when in use. Best of all, with nothing playing, even with an ear right next to the speaker there is no noise at all at full volume compared to when the DAC is powered off. For the price (around €15 on Amazon) they're clearly a step up from something generic, but at the same time very good value.
Many of these Chinese DAC boards can be powered either by a single +12V or a dual +/- supply. I don't know the relative merits of these or if they make a difference to the sound quality, and I'd love to hear if anyone has any measurements that show that there should be a difference.
For now, though, I'm glad to have found the one or two posts on here that mentioned using Leicke power supplies, and trusted them enough to try them out. Since the posts I read didn't mention their relative merits, I thought it was worth saying something about them. I'm impressed, at least.
Having read on here that Leicke power supplies are good quality, I decided to try one out, and I'm very impressed with the overall package - it's smaller than the wall wart I was using before and barely heats up at all when in use. Best of all, with nothing playing, even with an ear right next to the speaker there is no noise at all at full volume compared to when the DAC is powered off. For the price (around €15 on Amazon) they're clearly a step up from something generic, but at the same time very good value.
Many of these Chinese DAC boards can be powered either by a single +12V or a dual +/- supply. I don't know the relative merits of these or if they make a difference to the sound quality, and I'd love to hear if anyone has any measurements that show that there should be a difference.
For now, though, I'm glad to have found the one or two posts on here that mentioned using Leicke power supplies, and trusted them enough to try them out. Since the posts I read didn't mention their relative merits, I thought it was worth saying something about them. I'm impressed, at least.
Wall warts and switching regulators often put out some noise in the MHz frequency range. Sometimes its more common mode noise rather than differential. Although it might not sound like an audio tone when used to power a dac or other audio device, such power supplies sometimes still have some adverse affect on sound. One pretty well understood mechanism is when RF is rectified in opamp input stage semiconductor junctions and or by opamp ESD input protection diodes. It can then cause varying DC offsets at the opamp output. It can also get multiplied with the audio signal causing a type of distortion/noise product. Bottom line, many of us only use linear voltage regulators and R-core transformers for dacs we want to have sound their best.
A +/- supply is always going to be better when there’s an opamp or push/pull circuit.
It is known for many decades that HF/RF stray-in in audio devices results in worse sound. The stupid and easy way of looking at matters is saying the human ear does not hear RF ..... whilst knowing that the electronics do "hear" the RF and that may keep them busy doing things they were not supposed to do 
.... Any evidence is not needed, you can just accept some knowledge from a member and then try out for yourself. Now I know this since CD players were introduced, today it is 2021 and RF is everywhere around us by choice. Shielding is even more important and what do we see? People building sensitive equipment in plastic or wooden cases
Do a simple test with a cell phone, put it on silent and lay it on an unshielded DAC and listen to music while a friend calls you. Simple evidence!
The majority of switchers stray out signals in the MHz range/die prematurely/are unreliable/are powered on 24/7. Linear PSU's dance rings around them and almost every test is in favor of them. Only the very well designed SMPS don't misbehave but these are rarely delivered with audio equipment. Even the brands that heralded the merits of SMPS have over average breakdown of products but of course after the warranty period.
Th fact that many tech people still defend SMPS while the actual facts are the opposite is mind boggling...I do work with pro SMPS in pro devices where one would not want to go back to linear PSU's but for low power audio applications the average SMPS as delivered simply is crap.
.... Any evidence is not needed, you can just accept some knowledge from a member and then try out for yourself. Now I know this since CD players were introduced, today it is 2021 and RF is everywhere around us by choice. Shielding is even more important and what do we see? People building sensitive equipment in plastic or wooden cases Do a simple test with a cell phone, put it on silent and lay it on an unshielded DAC and listen to music while a friend calls you. Simple evidence!
The majority of switchers stray out signals in the MHz range/die prematurely/are unreliable/are powered on 24/7. Linear PSU's dance rings around them and almost every test is in favor of them. Only the very well designed SMPS don't misbehave but these are rarely delivered with audio equipment. Even the brands that heralded the merits of SMPS have over average breakdown of products but of course after the warranty period.
Th fact that many tech people still defend SMPS while the actual facts are the opposite is mind boggling...I do work with pro SMPS in pro devices where one would not want to go back to linear PSU's but for low power audio applications the average SMPS as delivered simply is crap.
Last edited:
Its possible to use SMPS for audio, but it takes some careful engineering to get it right. OTOH wall warts are usually made to be cheap, not ultra-low noise. Also, for the highest performance I would still go linear. Benchmark DAC-3 uses SMPS, but Topping D90 with its linear supplies is the better sounding dac (or was back when there was still AK4499). D90 is much better sounding for 'space junkie' types. Seriously. DAC-3 soundstage is compressed towards the center and forward. D90 is very wide. I kept D90 as my backup dac and gave the DAC-3 to my daughter.
Last edited:
Wait long enough till they all have SMPS and then market your new DAC with a linear PSU
Magazines (if they still exist) will write about "blacker than black" etc. The Full Linear PSU DAC or FLiPD.They all will have SMPS as there are many reasons linear PSU's are not preferred but do the math and realize that quality is not among them. It is cost, weight, cheaper switches, EU standby power consumption demands, no need to certify the PSU yourself etc.
Measurable or audible performance? If only measurable but not audible, who cares other than those who want to brag about what they own.
Better sounding in replaying equipment, it's the level of fidelity that determines. Personal impression of the listener has no bearing on determination criteria unless you are doing a musical recital, as in live performance in front of a crowd.Benchmark DAC-3 uses SMPS, but Topping D90 with its linear supplies is the better sounding dac (or was back when there was still AK4499). D90 is much better sounding for 'space junkie' types. Seriously. DAC-3 soundstage is compressed towards the center and forward. D90 is very wide. I kept D90 as my backup dac and gave the DAC-3 to my daughter.
Completely different DAC chip but it must be the power supply that makes the difference?
If it is measurable and a significant difference is measured it is most often audible as well. Now the other way around is somewhat more problematic. Some would then say that the wrong measurement was done.
Anyway with todays below 5 µV noise linear regulators and SMPS with multiple tens of mV of noise AND HF/RF straying things are clear very quickly. The vast majority of standard SMPS can only be improved by adding extra filtering and/or low noise linear regulators.
All useless debate as BOTW, Farad, Keces etc linear PSU's are an immediate upgrade for SMPS. Even with older linear regs the situation improves over the standard wall wart SMPS devices.
Last edited:
Please explain what you mean with the pejorative seeming remark. I have done and do meanwhile countless tests and measurements on SMPS wall warts and also examine their failures. If you mean you picked device A by brand B that happens to have a good SMPS that may be so. It is about the average level and it is simply poor.
I work on audio on almost daily basis and build linear PSU's in all kinds of variants for various devices. Even an LT1084 based PSU outperforms the average SMPS for low power applications so wall warts. Now if it would only be me you could call me the city's fool but apparently there are more people experiencing the same. In audio reviews wall warts are replaced for linear PSU's and the differences are noted. Maybe you are closing your eyes for facts?
Connect one of those miracle devices and measure it and then take an average linear PSU and measure. Compare and tell us the merits of the SMPS.
I work on audio on almost daily basis and build linear PSU's in all kinds of variants for various devices. Even an LT1084 based PSU outperforms the average SMPS for low power applications so wall warts. Now if it would only be me you could call me the city's fool but apparently there are more people experiencing the same. In audio reviews wall warts are replaced for linear PSU's and the differences are noted. Maybe you are closing your eyes for facts?
Connect one of those miracle devices and measure it and then take an average linear PSU and measure. Compare and tell us the merits of the SMPS.
Last edited:
The amount of Studio gear I get through here with horribly noisy (read
cheap) SMPS's is increasing significantly. Sadly I am starting to see them
being marketed as a superior performance alternative to linear PS's.
You can install RF attenuation measures / boards which helps but it also depends on grounding arrangement.
TCD
Yes it seems the "newer is better" marketing strategy works OK with laymen. Or is it severe cognitive dissonance (as it can be tested very easily). Or is it trolling? Plain stupidity?
* In high(er) power applications SMPS are the de facto standard thankfully but then designs are way better and various measures are taken to make it quality devices.
In audio when things are below 25VA and SMPS wall warts are Chinese OEM devices ... even decades old industrial low power linear PSU's outperform the average OEM SMPS and not slightly. Surprising: even decades old Chinese or then often still Japanese OEM linear PSU's outperform their modern replacements So, with linear PSU's the average quality of the DC voltage/noise is reasonably to very good and the exceptions are bad ones. With SMPS this is the other way around.
* In high(er) power applications SMPS are the de facto standard thankfully but then designs are way better and various measures are taken to make it quality devices.
In audio when things are below 25VA and SMPS wall warts are Chinese OEM devices ... even decades old industrial low power linear PSU's outperform the average OEM SMPS and not slightly. Surprising: even decades old Chinese or then often still Japanese OEM linear PSU's outperform their modern replacements
So, with linear PSU's the average quality of the DC voltage/noise is reasonably to very good and the exceptions are bad ones. With SMPS this is the other way around.
Last edited:
Fair question. The particular chip is probably not most of the difference in this case (stereo illusion of space). However some other differences in overall topology and component selection are likely more significant factors. Likely the PS is a significant part of it too. Could be otherwise though.
Last edited: