Is ripple the only spec we care for in an "audio source" PSU?

Hi there, you know how ppl are...they always want more, more, more!

When I upgraded my computer PSU to a Corsair w/ the lowest ripple measured on a famous hardware site(7mV ripple on the 12V rail in idle), I would swear I heard a SQ improvement on several internal soundcards I had at the time(using LT1028/AD797 opamps and Sanyo OS-Con PSU filtering for the most part).

I currently run a Firestone Spitfire DAC on its 24VDC "Supplier" PSU fed w/ toslink, SQ is amazing to my ears...and most ppl agree that the SQ improvement from that PSU is quite drastic: http://www.head-fi.org/forum/thread/410199/firestone-audio-supplier#post_5438537

so that's my 0.25A PSU inside shots: http://www.head-fi.org/forum/thread/286131/diy-supply-for-firestone-spitfire#post3664105

the only technical specs I could find for that 0.5A 220V gray transformer: http://www.bingzi.com/UploadFiles/201011416362630.pdf

Some linear PSU's can have a VERY high ripple, this one has 100mV:

its datasheet: http://www.velleman.eu/downloads/2/ps13xxgbnlfresd.pdf

All this to come to my question: Would there be any point for me to upgrade from the "Supplier" PSU to this thing?

They boast about 1mV ripple: http://www.niemeijer.nu/misc/manualps3003.pdf

I could get another cheaper laboratory PSU rated for 2A:

but they talk about 5mV ripple in its datasheet: http://www.produktinfo.conrad.com/datenblaetter/500000-524999/511806-an-01-ml-LIN_NETZGERAET_PS_1302D_de_en_fr_nl.pdf

A friend of mine tells me that ripple is not the only part of the story  and that opamps carry a high PSRR anyway...but the other chips in my DAC(S/PDIF receiver CS8416, PCM1793 DAC chip) might still exhibit the ripple improvement?

Some ppl on the link I posted above say that the Supplier PSU is just "a simple linear regulator IC"...and my friend tells me: "the Supplier does not just have a transformer & 2 caps: it has a regulator too =) and that's what counts for a DAC PSU (not a power amp, which would most probably have an unregulated power supply)."

I should also say that my favorite LT1363 opamp has a rather low PSRR at 88dB..a far cry from the 120/130dB found in other chips, improving the PSU section is mandatory on that chip...it can sounds quite horrid if poorly fed in many ppl's experience.

Firestone's techsupport told me this, but I can't make anything out of it:

done! a friend of mine ordered that PSU and I'm wondering whether it'd do any good to my Spitfire

well the Spitfire has a 0.2A fuse so that wouldn't be a problem I guess.
 

Originally Posted by cobaltmute 

I'd suggest you look at this and then look back at your 7mV noise figure:

 

http://tangentsoft.net/elec/psu-tests/

 

You can get into the uV noise range.  amb's sigma supplies get into that range as well.

 

For the digital side, such as clocks, I have read that flicker (1/f) noise is important as well.  Your limiting factor in the chase for low noise will be on board regulators generating noise.

noise and ripple are different things, aren't they?

I got the 7mV ripple figure from that site: http://translate.google.com/translate?u=http%3A%2F%2Fwww.canardpc.com%2Fdossier-36-450-Corsair_CX_400_Watts.html&sl=fr&tl=en&hl=&ie=UTF-8

They measured the rails ripple in idle/load and also the harmonics levels(3rd/5th level, etc)

or is the noise measured on your link is a mix of ripple/harmonics levels etc etc?

So I could find some 0.00x mV ripple PSU for my DAC? but this would be very pricey I'd guess..and is ripple the only spec we care for? not likely? harmonics levels come to mind.

ok, but then it means that this 1mV ripple spec is meaningless? and w/o knowing the different levels of harmonics, it's also impossible to know which one would "sound" better?

I can use a soldering iron and I'd like to get more into DIY...I'll check your link more thoroughly, but apparently the two best ones in this test are unavailable..I need a 230V 24VDC 0.5A, will check if the TREAD can do that.

yes, LT1363 has a 88dB PSRR, but the CS8416/PCM1793 would also most likely benefit from a less noisy PSU?

so the top choice in that review costs $100 in parts, the second not sure...but no kit is available for either of these...and the 3rd choice(the TREAD) is apparently quite cheap.

if I were to build a YJPS, it'd be a keeper and hopefully I could change its output voltage if I ever want to get a new DAC...and it'd never be a bottleneck to SQ. I like the sound of that(pun intended)

Well, the Supplier PSU is rated for 0.5A but the Spitfire DAC has a 0.2A fuse anyway...I guess it's about running a bigger PSU so it doesn't get hot and offers a better "throughput".

Yes, I checked the FAQ for the YJPS, and it's said that if you don't use the right wire gauge and so...there will be no point over using a TREAD anyway. There's a few electronics repair shops around my place, I might just ask them to check the ripple on that Supplier PSU and take it from there.

Thanks for pointing this out! Instead of wasting my time on a -more than likely- very subtle *audible* improvement off the Spitfire DAC(besides, its matched "Supplier" PSU sounds fantastic to my ears ), I think I'd better get myself a TREAD PSU for the Monitor 01USD...the SQ improvement would prolly be much higher, apparently these units really lack in the power filtering department: http://hifiduino.wordpress.com/2010/03/29/summary-of-musiland-01-mini-mods/

I'd need to hack the power lines off a USB cable, and supply something like 0.5A 5V.